Dr. Chen is a postgraduate year 4 radiology resident at Baystate
Medical Center in Springfield, MA. She graduated from Tufts Medical
School in 1994 and completed several years of surgery residency at
Rhode Island Hospital before switching to radiology.
The Eagle Pub on Bene't Street in Cambridge, England, has the
distinction of once being the haunt of a brilliant, brash young
Englishman named Francis Crick and a brilliant, even brasher and
younger American named James Watson. On the last day of February
1953, Crick burst through the pub's doorway and announced to all
within earshot that he and Watson had found the secret of life. The
secret that Watson and Crick (or Crick and Watson, as they are
referred to on the Eagle's side of the Atlantic) had uncovered was
the shape of a now familiar molecule called deoxyribonucleic acid
A Swiss scientist had isolated the chemical from the white blood
cells of pus in 1869, but neither he nor anyone else had any idea
of its importance to life. It was just before Watson and Crick
began their collaboration that an American bacteriologist named
Oswald Avery demonstrated that DNA was the stuff of
genes--discrete, inheritable packets of information aligned along
the chromosomes inside the cell's nucleus.
DNA is the basic genetic material that determines our genetic
makeup and can be reproduced and passed on to subsequent
generations. Within the nucleus, the DNA is packaged into 23 pairs
of chromosomes, half of which are derived from each parent.
Chromosomal DNA is wrapped tightly around histone proteins, which
allows it to remain condensed within the nucleus.
A gene is a stretch of DNA that contains the blueprint for the
sequence of amino acids making up a particular protein. Cellular
homeostasis is maintained by genes encoded for structural proteins
and cytoskeletal elements along with genes responsible for cellular
replication, protein, and nucleic acid. Mammalian genes consist of
multiple noncontiguous segments of DNA coding sequences (exons)
separated by varying lengths of noncoding intervening sequence DNA
(introns). Both upstream and downstream regulatory elements control
which genes are actively expressed and in which cell type (figure
1). These elements, called promoters or enhancers, are species-,
tissue-, and cell-type specific and can be a target for the
disruption of regulated cell growth. Such disruption can lead to
neoplastic transformation of the cell.
If a particular gene is mutated, it may not be able to produce
protein or it may function poorly or even too aggressively. This
disrupts cell and tissue functions that depend on the normal gene
product and causes abnormal cell behavior, leading to symptoms of
disease. Simply put, gene therapy is the deliberate transfer of DNA
for correction of mutated genes, which can be carried out by
introduction of a normal gene into the cell or by direct repair.
Science of the construct
Certain viruses have evolved to exploit cellular nucleic acid
and protein synthesis to ensure their own replication and survival.
Viruses are ideal vectors, being naturally suited for delivery of
genetic material into target cells. DNA viruses may exist
extrachromsomally as self-replicating episomes or may integrate
into the host genome. Ribonucleic acid (RNA) viruses, termed
retroviruses, reverse the normal sequence of events by converting
their RNA genome back into DNA by using a viral enzyme called
reverse transcriptase. The DNA thus produced is integrated into the
genome; host transcription and translation machinery is then
usurped for viral production. Discovery of reverse transcriptase
was a Nobel prize-winning landmark in molecular biology and allowed
DNA coding sequences to be derived from cellular messenger RNA
(mRNA), thus paving the way for gene cloning.
Gene cloning exploits the fact that coding sequences have been
separated or spliced from intervening sequences during the process
of transcription and mRNA production. After purification of mRNA
from the cell, single-stranded mRNA molecules are reverse
transcribed back into double-stranded DNA by using retroviral
reverse transcriptase. Reverse transcriptase then synthesizes a
second complementary DNA (cDNA) strand by using the first as a
fresh template. This process yields a double-stranded DNA molecule
that contains all of the relevant coding sequences present in the
original genomic DNA gene sequence, but without the often millions
of bases of noncoding intervening sequence DNA. Polymerase chain
reaction is currently the method of choice for amplifying the
cloned DNA into sufficient quantities (figure 2).
Preparation of the reverse transcription-polymerase chain
reaction cDNA for cloning into an expression vector is performed by
adding specific linker sequences onto the cDNA. Vectors are DNA
constructs (usually derivatives of bacterial plasmids ["naked
DNA"]) that allow sequences contained within them to be shuttled
about between various cell types (figure 3). After cloning in the
expression vector, transfection into bacteria such as
allows large-scale manufacture of milligram-to-gram quantities of
the cDNA-vector complex. The cloned cDNA must first be purified
from the bacteria, after which it is suitable for transfer into the
gene delivery vehicle of choice (figure 4).
Vectors: A Trojan horse
An ideal vector should theoretically have the following
properties. It should provide ease of penetration and sustained
expression. It should deliver an adequate amount of therapeutic
genes with high enough concentration to allow many cells to be
infected. It should be able to integrate within the active regions
of the genome that replicate and should be responsive to
manipulation of its regulatory elements. Naturally, it should pose
no hazard to the patient. It should be efficient and selective; it
should be able to target the desired type of cell and have no
components to elicit an immune response. Finally, it should be
conveniently produced in large quantities.
Genes in the form of DNA are typically delivered to a target
cell by one of three methods: (1) packaging them into a virus
(adenovirus or DNA virus, retrovirus or RNA virus); (2) directly
applying them in the form of a plasmid or a plasmid in a synthetic
delivery system ("nonviral vector"); or (3) by delivering them
through physical means such as direct injection, electric pulse
discharge (electroporation), gene guns, or radiofrequency pulses.
Once inside the cell, transgenes can be expressed in different
locations: within the cytoplasm, in the nucleus, or episomally
integrated into the host genome. Cytoplasmic expression is a useful
technique of gene transfer since the construct does not need to be
transported into the cell nucleus. However, the co-delivery of RNA
polymerase is required to initiate transcription. All other
categories of transgene expression necessitate DNA penetration into
the nucleus. Once the gene is incorporated into the genome, the
duration and level of foreign gene expression is usually long.
Current gene technology has focused primarily on the use of viral
vectors, which can provide highly efficient transduction and high
levels of gene expression, as several viruses have efficient
mechanisms for transferring genetic material to target cells. The
precise design of a particular viral vector depends largely on the
type of virus used.
Table 1 lists specific disease applications of various viral
Adenoviruses contain double-stranded DNA and enter the cell by
receptor-mediated endocytosis. Their genomes do not integrate into
the host and therefore have no oncogenic potential. Adenoviral
vectors are capable of high-level transduction in cells of multiple
lineages, including hepatocytes.
Because of the relative safety of replication-deficient infectious
adenoviruses and the ability to produce high titers in vitro,
adenoviruses are widely used in gene transfer. They are easy to
construct and are generally not perceived as toxic to cells.
Nonetheless, adenovirus is quite immunogenic, and antiviral
immunity is the primary factor limiting successful
re-administration. Also, because adenoviruses do not integrate into
the host genome, an important limitation is unstable expression
resulting from the loss of the recombinant gene from the transduced
In addition, hepatic side effects have been reported after in vivo
administration. These side effects are apparently due to expression
of vector-encoded viral sequences and production of associated
Retroviruses contain a single-stranded RNA genome. They are RNA
viruses that replicate through a DNA intermediate synthesized by
viral reverse transcriptase. Retroviral-cDNA constructs, like
adenovirus, are replication-deficient and must be transfected into
packaging cell lines to allow encapsulation of viral sequences into
an infectious retroviral virion.
These viruses enter the cell by direct fusion to the plasma
membrane and then integrate in stable form into the host chromosome
during cell division. The high efficiencies of retroviruses in
transduction and stable gene delivery make them attractive vehicles
for gene delivery.
However, retroviruses infect only dividing cells, a property that
limits potential delivery of therapeutic cells to nondividing
cells, slowly growing tumors, or nonmitotic fractions of the tumor
populations. Retroviral vectors also have a restricted size
capacity of approximately 8 kilobases (8000 bases), which limits
its use with large genes. Further important limitations include
inactivation by serum complement, low serum titers of vector after
in vivo administration, and the appearance of replication-competent
retroviruses during large-scale production.
More recently, herpes simplex virus (HSV)-derived amplicons have
been used as an alternative method of gene delivery.
Amplicons are plasmids that can be packaged into recombinant viral
particles and used to transfer anticancer genes in vivo and in
vitro, although the vector itself does not possess oncolytic
functions. Amplicons can be packaged as multiple copies into each
viral particle, making it a more efficient method of plasmid
delivery. Another advantage of amplicons is that they are
maintained extrachromosomally (episomally) in replicating cells,
assuring passage to dividing daughter tumor cells. Finally,
amplicons, like adenoviruses, show little toxicity.
Other viral delivery systems:
A number of other viral delivery systems are currently under
development. These include poxviruses, such as the Sindbis and
Semliki Forest viruses. These allow viral protein expression
without viral replication and would therefore be useful for
vaccination. Lentiviruses, including human immunodeficiency virus
and simian immunodeficiency virus, have the advantages of high
efficiency of expression, trophism for immune cells, and stable
expression in nondividing cells.
Lastly, adeno-associated viruses (AAV) are simple, single-stranded
DNA viruses; AAV vectors can infect nondividing cells, do not
elicit an immune response, have long-term expression, and can
integrate into chromsomes at specific sites. Their disadvantages
are: they are smaller viruses with restricted usefulness because
they can carry only a small genetic payload; they insert themselves
randomly into chromosomes with risk of disrupting functional genes;
and, finally, they are difficult to manufacture in large
Nonviral, artificial vectors--
Despite the ability to transfect cells efficiently, viral methods
for human gene therapy have several limitations. Viral particles
are often unstable and can have low titers. Secondly, viral
particles are often immunogenic. Thirdly, a pathologic
replication-competent virus may be generated. Finally, oncogene
activation and cancer development may occur. To circumvent these
potential problems, transfection methods using nonviral artificial
vectors carrying plasmid DNA (naked DNA) have been developed. Other
nonviral synthetic vectors for gene therapy have included a variety
of compounds--for example, cationic liposomes, poly-L-lysine-DNA
complexes, DNA-coated microprojectiles (gold particles),
dendrimers, or even free-plasmid DNA. Liposome-mediated gene
transfer has been utilized extensively in in vitro studies
("lipofection"). Liposomes are formed from phospholipids similar to
those found on cell membranes of living organisms. These synthetic,
lipid bubbles act as carriers for plasmids whose original genes
have been replaced by therapeutic ones. Because liposomes consist
of a double layer of lipid molecules similar to those of animal
cells, a liposome can fuse with a cell and deliver its contents
into the cellular interior.
A drawback is that the internal diameter of a liposome is much less
than the longest DNA plasmid. This presents a size restriction for
incorporation of "long" DNA plasmids into the carrier liposome and
limits this method of transfer.
Finally, ionic poly-L-lysine DNA complexes have been attached to
target molecules such as asialoglycoprotein or transferrin.
Although these nonviral vector model systems seem to work some
degree ex vivo, gene transfer is generally lower in vivo.
Gene therapy applications
In the setting of genetic deficiences, the goal of therapy is
restoration of a particular function in cases in which a genetic
mutation has led to deficient or absent expression of a critical
gene product. This has been applied to several diseases (e.g.,
cystic fibrosis [transmembrane conductance regulator], adenosine
deaminase deficiency, or Gaucher's diseases [beta-glucocerebroside
Table 2 further elaborates candidate diseases for gene therapy and
corresponding approaches for treatment. Additionally, chronic
pathologic processes may be amenable to gene therapy with the goal
being delay or cessation of chronic progressive disease, i.e.,
peripheral vascular disease and vascular gene therapy.
Fischer has presented an extensive review of vascular gene therapy
application and delivery.
Gene therapy for cancer
Many gene therapy protocols to date have concentrated upon
treatments for cancer. Experts predict that the major emphasis in
human gene therapy experimentation over the next few years will be
directed toward the treatment of cancer.
Though many cancers have a genetic predisposition, they all involve
acquired mutations, and as they progress, their cells become less
differentiated and more heterogenous with the respect to the
mutations they carry. In general, cancers have one mutation to a
proto-oncogene (yielding an oncogene) and at least one to a tumor
suppressor gene, allowing the cancer to proliferate. The range of
different cancers and the mutations they carry have led to a
variety of strategies for gene therapy, namely oncogene
inactivation, tumor suppressor gene replacement,
immunopotentiation, molecular chemotherapy, and drug resistance
genes. Emphasis will be on genetically modifying cells of the
body's own immune system to fight tumors. In this way, the same
genes may be used to treat many different forms of cancer.
Cell growth versus apoptosis: Shifting the balance
Cellular proliferation is an intricate balance between signals
driving cell cycle progression, signals maintaining quiescence, and
signals initiating the pathway for cellular destruction
(apoptosis). Signal transduction is the process of converting
extracellular signals into cellular function and is tightly
regulated. Apoptosis is a normal process by which senescent cells
undergo programmed cell death. Apoptosis is critical for the body
to ensure that aged or potentially damaged cells are targeted for
elimination from the system (figure 5). Immunologic protection
often involves induction of apoptosis (e.g., in infected or
neoplastic cells). Development of malignancy has been shown by
numerous studies to involve aberrant or absent apoptosis of
transformed neoplastic cells.
One signal transduction approach to therapy involves shifting the
balance back in favor of cell death by inserting genes, such as the
fas ligand gene, that are involved in triggering apoptosis.
Antibodies directed against fas ligand expressed on the tumor cell
may also mimic fas/fas ligand interactions and trigger the
apoptotic death of the tumor (figures 6 and 7).
Gene therapy approaches to cancer--
Oncogenes are cellular genes that play a role in regulating normal
cell growth and proliferation. Expression of certain oncogenes
related to normal growth factors or their receptors, such as the
erb-2B and ras oncogenes, may result in disregulated intracellular
signaling, which leads to uncontrolled neoplastic proliferation.
Oncogene activity can also be inhibited with antisense RNA,
intracellular antibodies, catalytic RNA molecules (ribo-zymes) that
inactivate or digest oncogene RNA, or antioncoprotein antibodies
designed to block aberrant signal transduction.
Tumor suppressor gene replacement:
Anticoncogenes, also known as tumor suppressor genes, are
implicated in the neoplastic transformation of many tumors. These
genes are termed antioncogenes because it is their absence rather
than their presence that leads to malignancy.
Figure 8 illustrates the role that one particularly ubiquitous
tumor suppressor gene, p53, plays in regulating tumor cell
Preliminary results from one phase I trial in which retroviral-p53
gene complexes were injected into refractory, p53-deficient,
primary nonsmall-cell lung carcinoma showed tumor regression in
one-third of patients and stabilization of disease in an additional
Moreover, expression of p53 is synergistic with chemotherapeutic
drugs, such as cisplatin,
and adjacent tumor cells that have not been transduced are killed.
This is termed the
Other tumor suppressor genes include BRCA1sv, the retinoblastoma
(RB) gene (also found in prostate and bladder tumors), and the
Wilms' tumor (WT) gene, also implicated in acute myelogenous
The aim of immunopotentation is to enhance the response of the
immune system to cancers, thereby leading to their destruction.
Passive immunotherapy aims to increase the pre-existing immune
response to the cancer, while active immunotherapy initiates an
immune response against an unrecognized or poorly antigenic tumor.
Passive immunotherapy usually involves harvesting
tumor-infiltrating lymphocytes and treating them to express
increased cytokines. Various immunostimulatory cytokines have been
inserted into tumors and shown to increase antitumor immunity.
These cytokines include IL-2, GM-CSF, Il-4, Il-12, tumor necrosis
factor (TNF), and interferon (INF). Paracrine secretion of cytokine
by the tumor elicits local immune activation, which enhances
antitumor immunity. Alternative strategies modify the host immune
cells directly. Lymphocytes are known to infiltrate tumors but
somehow remain suppressed and anergic. Gene transfer approaches are
being explored to enhance the activity of tumor-infiltrating
lymphocytes, as well as peripheral blood lymphocytes, by inserting
immunostimulatory cytokine genes such as IL-2, TNF, and INF into
the lymphocytes, thus allowing both autocrine and paracrine
activation of antitumor immunity. Data from early clinical trials
have indicated augmented antitumor immune reactivity in the
gene-modified lymphocytes, although few cases of sustained tumor
remission have been documented.
Active immunotherapy genetically modifies tumor cells to
increase expression of antigen presenting molecules/costimulatory
molecules, local concentrations of cytokines, or tumor antigens.
The cells are then irradiated prior to being returned to the
patient, preventing the reintroduction of replication-competent
tumor cells. These approaches have been termed
. Tumor vaccines are genetically modified tumor cells or peptide
products of tumor that are used to "vaccinate" patients against
their own tumors.
A strategy currently in clinical trials for metastatic melanoma
is based on the observation that creating an immunostimulatory
environment in and around a tumor can activate lymphocytes that
have infiltrated the tumor but have remained quiescent. This
strategy involves partial tumor resection and insertion of a gene
that codes for an immunostimulatory cytokine (e.g., IL-2 or
granulocyte-macrophage colony-stimulating factor (GM-CSF) into the
resected tumor cells. After several cycles of cell division in the
laboratory to allow for expression of the cytokine gene, the
modified tumor cells are irradiated to prevent further cell
division and returned to the patient with a depot injection. The
modified tumor cells are still able to synthesize and secrete
cytokine, which produces paracrine effects on infiltrating immune
To circumvent difficulties in maintaining tumor cells in the
laboratory, autologous fibroblasts have also been transduced with
cytokine genes and injected together with irradiated tumor cells.
Another way tumor cells are genetically modified for active
immunotherapy is increasing the expression of costimulatory
molecules. One way in which tumors evade the immune system is by
means of loss or alteration of these important costimulatory
The tumor-lymphocyte interaction is facilitated by the
costimulatory ligand-receptor (B7-CD28) interaction. B7 is absent
in many tumors, a condition that leads to tumor-specific anergy and
thus deficient antitumor immunity. Re-establishing B7 expression by
insertion of the B7 gene into B7-deficient tumors has been shown to
facilitate immune recognition by tipping the balance away from
anergy and back in favor of protective immunity.
An alternative approach to increasing tumor immunogenicity is to
express foreign HLA antigens in the tumor by means of direct
intratumoral transfer, which causes the tumor to be "rejected" by
the host immune systems just as a foreign-tissue transplant might
be rejected. Phase I trials in which this approach has been used in
patients with metastatic melanoma have demonstrated significant
antitumor activity and, thus, the feasbility of this approach. In
some cases, complete regression of residual disease has been
Molecular chemotherapy and drug resistance genes:
Molecular chemotherapy and drug resistance genes direct
chemotherapy drug production by the tumor itself. A suicide gene
that encodes a prodrug converting enzyme, such as herpes simplex
virus-thymidine kinase (HSV-TK), is inserted into tumor cells.
A specific nontoxic prodrug (ganciclovir) is then administered
systemically. On uptake by suicide gene-expression tumor cells, the
prodrug is converted into a toxic antimetabolite, which leads to
tumor cell death.
Two systems have been extensively studied. The first is HSV-TK,
in which intratumoral expression of the HSV-TK allows ganciclovir
to be converted into a toxic triphosphate derivative, a process
that leads to early DNA chain termination. The second system
involves cytosine deaminase in which the cytosine deaminase gene
expressed by tumor cells allows 5-fluorocytosine to be converted
into the antimetabolite 5-fluorouracil. Local release and bystander
effects on adjacent cells offer a theoretical solution to often
inefficient gene transfer into cells. Indeed, Ogawa et al
reported significant antitumor activity when a cytosine
deaminase-based gene transfer system was used in colorectal cancer.
The bystander effects include tumor lysis, which changes the tumor
microenvironment from inhibitory to immunostimulatory and leads to
infiltration of the tumor by immune cells. Preclinical data
obtained with an experimental model of hepatic metastasis from
colorectal carcinoma have demonstrated that combinations of suicide
gene therapy and cytokine gene therapy may produce synergistic
Irrespective of the specific type of the gene delivery vehicle
used, it must be delivered efficiently to its intended target in
order for the gene to be expressed at therapeutic concentrations.
This may include ex vivo, in situ, and systemic in vivo routes of
therapy. Therefore, the interventional radiologist will play a
crucial role as part of the gene therapy team. We are uniquely
positioned to offer expertise in choosing approaches and sites for
gene delivery. Although systemic IV administration of vectors have
been performed, it has not proven to be as efficacious as local
delivery. Angiographic guidance will be of use in localizing local
tumor blood supply and directing the targeted intra-arterial
delivery of genes of interest so that vector-DNA complexes can be
delivered with accuracy and specificity. This approach offers
considerable advantages over systemic approaches in which much of
the vector-DNA complex may be lost in the first pass through the
pulmonary and hepatic reticuloendothelial systems. Embolization
techniques may also aid in prolonging vector contact with the
target cells by delaying washout and thereby further enhancing
target cell uptake. Other delivery strategies include image-guided
delivery to target tissues or tumor through stereotactically placed
needles and cavitary administration using US, CT, or MR guidance.
Using ultrasound, the process of liposomal vector delivery can
be monitored directly because lipid vesicles are echogenic. Such
monitoring is of considerable importance in demonstrating accurate
delivery to sites of interest. After injection, acoustically
activatable carriers (ultrasound "contrast agents" such as
microbubbles carrying genes and/or drugs) can be monitored by
ultrasound or MR (liposomes encapsulating gadopentetate), and once
at the region of interest, high-frequency ultrasound pulses can be
applied to increase the vascular permeability and rupture the
microbubbles. It also can increase the permeability of cell
membrane and facilitate the diffusion of foreign DNA into the cell.
Lastly, guided biopsy of transduced tissues for histopathologic
analysis after gene delivery should improve confidence in the
evaluation of gene expression. The radiological applications of
this technique are self-evident and could be incorporated into the
clinical practice of ultrasound imaging.
Radiotherapeutic uses of ionizing radiation for gene therapy are
also in development. With radiation-sensitive inducible promoters
of gene transcription, the use of targeted gene delivery to a tumor
followed by activation of expression of the therapeutic gene with
stereotactically guided radiation therapy may be possible. One such
system has linked expression of the tumor necrosis factor gene to a
radiation-inducible promoter. Ionizing radiation is thus used in
both temporal and spatial regulation of gene therapy. Other
examples of radiation-sensitized genes include p53 and fas, both of
which play a role in promoting radiation-triggered apoptosis.
More recently, additional interventional strategies have been
employed to improve gene delivery, including electroporation,
microinjection, particle bombardment or gene gun therapy,
high-frequency ultrasound, and other mechanical or chemical
disruptions that improve local delivery. Electroporation creates
small temporary openings in the cell membrane, allowing
introduction of DNA through application of an electrical field for
a few microseconds to milliseconds. Microinjection is an old
technique performed manually under a microscope. A syringe is used
to pierce the cellular membrane and insert the genetic material
into the cytoplasm. Lastly, particle bombardment or "gene gun"
technology involves the direct bombardment of DNA-coated gold
particles through the cell membrane into the cytoplasm.
Since Francis Crick "winged" into the Eagle in Febrary 1953, the
spiraling double helix has become one of the most famous scientific
images of our times. With the discovery of DNA, the genetic code
was cracked, thereby paving the way for gene therapy. Gene therapy
places us at a vital moment in human as well as medical history.
Genetics is to our generation what nuclear physics was to the last:
only the atomic bomb is an appropriate analogy for the explosive
and implosive implications of gene therapy. Increasingly, gene
therapy is used experimentally and clinically to replace defective
genes and/or impart new functions to cells and tissues. With the
recent advances in vector design, and improvements in transgene and
prodrug activation strategies, gene therapy has been applied to a
wide variety of diseases, tissues, and organ systems. The
contemporary explosion in molecular probe design, imaging
technology, and genetic engineering is expected to generate new
imaging concepts and new affinity ligands, which will most likely
drive the development of novel imaging technology. The new frontier
of "molecular imaging" will expand our current capabilities and
will complement advances in other fields of molecular medicine. It
will revolutionize current approaches to screening, detecting,
characterizing, and treating different diseases.
Finally, it has always been clear that our specialty will play a
critical role in gene therapy and its clinical applications,
particularly in delivering genes and vector products by minimally
invasive interventional techniques. Radiology will be involved at
every level of gene therapy. We will characterize the disease and
its process, choose sites for targeting therapy, contribute to gene
delivery, monitor uptake in tissues of interest, and, finally,
evaluate levels of target gene expression and clinical response.
Consequently, it is especially important that we have an
understanding of the molecular biology and basic techniques used to
clone a gene, current applications for gene therapy, and
contemporary delivery systems in order to ensure our integration
into the gene therapy team as this field rapidly and furiously