Current Pharmaceutical Design, 2003, 9, 0000-0000
1
Gene Therapy-Mediated Modulation of Immune Processes in the Central
Nervous System
Roberto Furlan*, Stefano Pluchino and Gianvito Martino
Neuroimmunology Unit, Dept. of Neurology, San Raffaele Scientific Institute Via Olgettina,
58 –20132 Milano - Italy
Abstract: Selective interference with immune processes in the central nervous system (CNS) is a
very difficult task because of the limitations associated with the delivery of immuno modulatory
molecules across the blood brain barrier. Systemic administration of immune-mediators, either by
conventional routes or by intramuscularly or intravenous gene therapy, is hampered by severe
side effects and alters immune-system functions also in peripheral organs. To overcome these
problems, different gene therapy strategies have been developed to deliver immuno modulatory
molecules directly within the central nervous system. The use of engineered CNS antigenspecific circulating cells as selective delivery vehicles, the direct injection of gene vectors into the brain parenchyma, or
also the ependymal route, have been proposed as possible alternative gene therapy protocols to selectively interfere with
immuno-pathological processes in the CNS. We will review the use of these CNS-targeted gene therapy protocols for the
treatment of experimental autoimmune encephalomyelitis (EAE), the prototypical experimental immune-mediated disease
of the CNS, and therefore discuss the relevance of these results for the therapy of multiple sclerosis (MS) the most
common, immune-mediated, demyelinating disease of the CNS in humans.
Key Words: gene therapy, central nervous system, multiple sclerosis, experimental autoimmune encephalomyelitis, cytokines,
and growth factors.
INTRODUCTION
Immunomodulatory therapies targeting the central
nervous system (CNS) are a crucial challenge to future
medicine. In fact, besides some pathogeneticallycharacterized immune-mediated diseases of the CNS that
await more efficient therapeutic tools for a better
management, such as multiple sclerosis (MS), other neuro
degenerative and neuro pathological syndromes, have a
recently recognized immune-mediated component that might
represent a potential therapeutic target. Alzheimer’s disease
(AD) [1], amiotrophic lateral sclerosis (ALS) [2], other
neuro-degenerative diseases [3], and even stroke [4] have
recently been considered as diseases candidate for a novel
immunomodulatory therapeutic approach, along with neuroprotective and if possible etiological treatments. However,
the immune system and the CNS are in a very peculiar
relationship, when compared to the rest of the body. In fact,
the immune system surveys the CNS as any other district in
the living organism. However, immune reactions in the CNS
usually follow distinct rules because of anatomical and
functional peculiarities that, collectively, have allowed the
CNS to be defined as an immune-privileged organ. The
anatomical, functional, and regulatory peculiarities of the
immune system in the CNS are still under investigation and
only partially understood. These studies are difficult also
*Address correspondence to this author at the Neuroimmunology UnitDIBIT, San Raffaele Scientific Institute, Via Olgettina 58, 20132 Milan,
Italy; Tel: +39 02 2643 4867; Fax: +39 02 2643 4855;
E-mail: furlan.roberto@hsr.it
1381-6128/03 $41.00+.00
because targeting molecules to the brain is extremely limited
by the presence of the blood-brain barrier (BBB) [5]. To
date, most of conventional neurotherapeutic agents are
supposed to cross the BBB because of their smaller size.
However, more than 98% of small molecules don’t cross the
BBB either [6]. We can then conclude that a specific
delivery within the CNS of molecules able to modulate the
local immune response would be therefore useful to
investigate the features of the immune reaction in the brain,
thus representing a novel therapeutic tool for several
heretofore untreatable CNS-restricted inflammatory diseases.
Unfortunately, the selective modulation of the immune
system in the CNS by a conventional peripheral therapeutic
approach seems, at present, an unreal task.
This review will focus on gene therapy as a possible
alternative approach to target immunomodulatory molecules
selectively to the CNS permitting an in situ regulation of
immune reactions.
IMMUNOMODULATORY GENE THERAPY IN THE
CENTRAL NERVOUS SYSTEM
Gene therapy has been recognized first as a tool for the
correction of genetic disorders, and only in a second instance
as an alternative drug delivery system. This might be the
reason why immunomodulation in the CNS by gene therapy
is a recent, although rapidly growing, field of investigation.
The prototypical, chronic, immune-mediated demyelinating
disease of the CNS is multiple sclerosis (MS). MS is of
unknown etiology, and it is characterized by the presence of
© 2003 Bentham Science Publishers Ltd.
2 Current Pharmaceutical Design, 2003, Vol. 9, No. 20
perivascular inflammatory infiltrates in the CNS containing
T and B cells and activated macrophages, thus suggesting
that MS is a T cell mediated, CNS-confined, chronic,
inflammatory, demyelinating disease in which the ultimate
effector cells are activated macrophages [7]. The
inflammatory process, leading to multifocal patchy
demyelination and axonal loss, is mainly sustained by proinflammatory cytokines that along with chemokines,
adhesion molecules, and metalloproteinases modulate at
different levels the immuno pathogenic process underlying
MS [8]. Due to their central role in MS pathogenesis,
"inflammatory" molecules might represent suitable
therapeutic targets [9]. No currently available treatment for
MS enables a satisfying control of disease evolution, and MS
patients, usually young adults, accumulate various degrees of
disability over several years of disease [10]. MS is modeled
in animals by experimental autoimmune encephalomyelitis
(EAE), which can be induced in rodents and non-human
primates by active immunization with whole myelin, myelin
proteins, myelin peptides, or non-myelin antigens [11-12;
reviewed in 13]. Depending on the animal species, or strain,
and the antigen used for active immunization, different
disease courses (i.e. acute monophasic, relapsing-remitting
or chronic), resembling clinical subtypes of human MS, can
be obtained. Despite being performed on a more accessible
model than the human CNS, the immunological studies on
EAE in rodents have not produced conclusive data on the
role of various immune mediators in the pathological process
leading to lesion formation in the CNS [reviewed in 14].
However, numerous gene therapy protocols aimed at the
delivery of immunologically active molecules in the brain
have been published so far (Table 1). Three main approaches
have been attempted: (i) genetically engineered cells, usually
lymphocytes specific for CNS antigens, used as “Trojan
horses” able to find their target and deliver the engineered
therapeutic molecule; (ii) direct injection of biological or
non-biological gene therapy vectors into the CNS
parenchyma; (iii) injection of viral gene therapy vectors in
the cerebrospinal fluid circulation, namely the “ependymal
route”. These three different approaches have been tested in
different EAE models and gene therapy has been
administered either at the time of disease induction
(preventive treatment) or close to, or after, the disease onset
(therapeutic treatment).
CNS GENE THERAPY
ENGINEERED CELLS
USING
CIRCULATING
The perivascular inflammatory infiltrates in the CNS the pathological hallmark of MS/EAE - are thought to be the
cause of lesions formation and are renewed several times a
day [15]. This florid traffic of lymphocytes from the
circulation to the CNS can be exploited to take a therapeutic
molecule selectively across the BBB. Several groups have
used retroviruses as gene therapy vectors to permanently
insert therapeutic genes into T lymphocytes that then have
been re-infused to EAE mice. To ensure a specific delivery,
the T cells engineered to release therapeutic molecules were
specific for CNS antigens. The re-infused CNS-antigen
specific T cells are supposed to home preferentially into the
damaged area of the CNS, were inflammation is actively
ongoing, and, therefore, to release the protective agent
Furlan et al.
exactly at the site and time were it is needed. Myelin basic
protein (MBP)-specific T cells have been retrovirally
transduced with the interleukin (IL)-4, IL-10 or tumor
necorsis factor (TNF)α gene and have been transferred into
syngeneic (PL/J x SJL/J)F1 mice affected by chronicremitting, MBP-induced, EAE. IL-4-engineered cells, reinfused after disease onset, were able to ameliorate the
disease course, whereas IL-10-transduced cells were
ineffective, and TNF-α-engineered cells worsened the
clinical signs of EAE [16, 17]. In a different study, a MBPspecific BALB/c T helper 1 clone has been retrovirally
transduced with the latent form of transforming growth
factor (TGF)β, and transferred into (SJL x BALB/c)F1 mice
affected by acute monophasic EAE induced by immunization
with the proteolipid protein (PLP) peptide 139-151. This
treatment was effective when administered both at the time
of immunization and of the disease onset, although only
partially beneficial when engineered T cells were infused 3
days after EAE onset [18]. Activated PLP139-151-specific T
cells have also been retrovirally transfected with an
expression cassette in which the IL-2 promoter drives murine
IL-10. In this way, transcription of the transgene is induced
only when the IL-2 promoter is active - at the time of cell
activation - supposedly when the cell is on the inflammatory
site. These cells, transferred into PLP139-151-immunized
(SW x SJLJ)F1 mice affected by relapsing-remitting EAE,
were able to ameliorate both clinical and neuropathological
signs of the disease [19]. Growth factors have also recently
been engineered into myelin antigen-specific T cells with the
aim to foster endogenous repair mechanisms and considering
in parallel their possible activity on immune cells. In a first
study, PLP139-151-specific T lymphocytes were engineered
with an antigen-inducible transgene coding for plateletderived growth factor (PDGF)-A. The transfer of these cells
in PLP139-151-immunized (SW x SJLJ)F1 mice was able to
ameliorate ongoing relapsing-remitting EAE [20]. In a
second study, MBP-specific CD4+ T cells were transduced
with a recombinant retrovirus coding for nerve growth factor
(NGF). These myelin antigen-specific T cells, when reinfused in Lewis rats, were unable to mediate EAE and,
furthermore, inhibited acute monophasic EAE when cotransferred with non-transduced MBP-specific T cells. This
latter result was associated not only with the induction of
tissue repair, but also with the reduction of inflammatory
infiltrates in the CNS, and especially of infiltrating monocytes, thus suggesting an unexpected immunomodulatory
role for NGF [21]. An original and recent variation of this
approach has been proposed lately. Using engineered B cells,
EAE susceptible mouse strains were made resistant by
specific down-regulation of the pathogenic auto-reactive T
cell clone. Starting from the concept that antigen
presentation from B cells is usually tolerogenic, normal B
cells were retrovirally transduced to express PLP139-151
fused to a lysosomal target sequence, to ensure loading on
MHC-II molecules. Transformed B cells transferred into
PLP139-151-immunized (SJL x BALB/c)F1 mice were able
to prevent relapsing-remitting EAE in the majority of mice,
and to delay onset and reduce severity in the affected ones
[22]. In a second recent study, on the same lines, B cells
were retrovirally transformed to express an IgG-MBP fusion
protein. Transformed cells were then co-transferred
with MBP-specific encephalitogenic T cells into syngenic
Gene Therapy-Mediated Modulation
Table 1.
Current Pharmaceutical Design, 2003, Vol. 9, No. 20 3
Experimental Gene-Therapy Trials in Rodent and Non-Human Primate Models of EAEa
Delivery route
Gene vector
Therapeutic gene
Animal strain
Immunization
Administration
scheduleb
Clinical
efficacy c
Refs
MBP-specific T cells
Retrovirus
IL-4
(PL/JxSJL)F1 mice
MBP
Therapeutic
+
16
MBP-specific T cells
Retrovirus
IL-4
(PL/JxSJL)F1 mice
MBP
Therapeutic
+
17
MBP-specific T cells
Retrovirus
IL-10
(PL/JxSJL)F1 mice
MBP
Therapeutic
-
16
PLP-specific T cells
Retrovirus
IL-10
SWXJ mice
PLP139-151
Therapeutic
+
19
MBP-specific T cells
Retrovirus
TNFα
(PL/JxSJL)F1 mice
MBP
Therapeutic
worsening
17
B cells
Retrovirus
PLP100-154
(SJLxBALB/c)F1 mice
PLP139-151
Preventive
+
22
B cells
Retrovirus
MBP-IgG1
(PL/JxSJL)F1 mice
bMBP + MBP1-17
Preventive
+
23
B cells
Retrovirus
MBP-IgG1
(PL/JxSJL)F1 mice
bMBP + MBP1-17
Therapeutic
+
23
B cells
Retrovirus
MBP-IgG1
(PL/JxSJL)F1 mice
PLP139-151
Therapeutic
-
23
B cells
Retrovirus
MBP-IgG1
(PL/JxSJL)F1 mice
bMBP + PLP139-151
Therapeutic
-
23
PLP-specific T cells
Retrovirus
PDGF-α
SWXJ mice
PLP139-151
Therapeutic
+
20
MBP-specific T cells
Retrovirus
TGF-β
(SJLxBALB/c)F1 mice
PLP139-151
Therapeutic
+
18
MBP-specific T cells
Retrovirus
NGF
Lewis rats
MBP
Preventive
+
21
Intracerebral
Liposomes
IL-4
Biozzi AB/H mice
SCH
Therapeutic
+
27
Intracerebral
HSV-1 134.5
IL-4
BALBc
SCH
Preventive
+
31
Intracerebral
HSV-1 134.5
IL-10
BALBc
SCH
Preventive
-
31
Intracerebral
Liposomes
IL-10
Biozzi AB/H mice
SCH
Therapeutic
+
27
Intracerebral
Retrovirus
IL-10
Biozzi AB/H mice
SCH
Therapeutic
+
29
Intracerebral
Adenovirus
IL-10
Biozzi AB/H mice
SCH
Therapeutic
-
29
Intracerebral
Liposomes
IFNβ
Biozzi AB/H mice
SCH
Therapeutic
+
28
Intracerebral
Liposomes
p55 TNFR-Ig
Biozzi AB/H mice
SCH
Therapeutic
+
27
Intracerebral
Liposomes
p75 dTNFR
Biozzi AB/H mice
SCH
Therapeutic
+
28
Intracerebral
Retrovirus
p75 dTNFR
Biozzi AB/H mice
SCH
Preventive
+
30
Intracerebral
Liposomes
TGF-β
Biozzi AB/H mice
SCH
Therapeutic
+
27
Intracisternal injection
HSV-1
IL-4
Biozzi AB/H mice
MOG40-55
Preventive
+
34
Intracisternal injection
HSV-1
IL-4
Biozzi AB/H mice
SCH
Therapeutic
+
35
Intracisternal injection
HSV-1
IL-4
Macaca mulatta
Whole myelin
Therapeutic
+
36
Intraventricular inject.
Adenovirus
IL-10
CSJLF1 mice
SCH
Therapeutic
+
43
Intracisternal injection
HSV-1
IFNγ
C57BL/6 mice
MOG35-55
Preventive
+
37
Intracisternal injection
HSV-1
IFNγ
C57BL/6 mice
MOG35-55
Therapeutic
+
37
Intracisternal injection
HSV-1
FGF-II
C57BL/6 mice
MOG35-55
Therapeutic
+
38
Engineered cells
Direct injection
Ependymal route
a
Abbreviations: p55 TNFR-Ig, p55 TNF receptor-Ig fusion protein; p75 dTNFR, p75 TNF dimeric receptor; SCH = spinal cord homogenate; bMBP = bovine myelin basic protein.
b
Preventive, at the time of immunization or shortly after; Therapeutic, close to or after disease onset. c +, disease amelioration; -, no effect.
4 Current Pharmaceutical Design, 2003, Vol. 9, No. 20
Furlan et al.
(PL/J x SJL)F1 mice, preventing relapsing-remitting EAE,
even when injected after the disease onset. The effect was
antigen-specific, since MBP-Ig transformed B cells were
unable to prevent EAE induced by co-immunization with
PLP and MBP [23]. Thus, several different approaches of
cellular gene therapy can be used to modulate immunemediated pathological processes in the CNS injecting
engineered cells as “trojan horses”. Unfortunately, several
drawbacks limit the applicability of this approach to MS.
While pathogenic auto-antigens are known in the
experimental model, the epitopes driving the immune
response in human MS are still unknown. What should,
therefore, be the proper antigenic specificity if the T cell
approach is taken in humans? Furthermore, T cells need to
be activated in vitro in order to properly home to
inflammatory sites in the CNS. This may lead someway to
disease exacerbation if the balance between beneficial effects
of the delivered therapeutic molecule and the detrimental
contribution of the myelin antigen-specific T cell to the
pathogenic process is unfavorable. Human trials with altered
peptide ligands have already demonstrated that undesired
activation of T cells with myelin antigens may lead to
disease worsening [24, 25]. Furthermore, studies using
engineered peripheral lymphocytes have failed to
demonstrate, in a definitive way, that transformed cells are
able to travel to the site of lesion within the CNS and there
release the active molecule. It could well be that the
beneficial effects may derive from the release of the putative
protective molecule in the periphery, therefore raising again
the issue of peripheral side effects. The above-described B
cell approach, on the other hand, is also novel and very
interesting, but, again, it is difficult to envisage antigenspecific therapies in MS where the putative auto-antigen is
still unknown.
GENE THERAPY THROUGH THE
INJECTION IN THE CNS PARENCHYMA
DIRECT
To overcome the above-mentioned concerns, many
groups have taken a different approach by injecting gene
vectors directly into the CNS parenchyma. Retroviral
vectors, which are able to transform only cycling cells, and
are preferred for ex vivo gene therapy protocols, cannot be
employed for direct injection in the brain because the very
low number of replicating CNS resident cells would lead to a
negligible efficiency of gene transfer. Among the different
vectors used in the following studies, lentiviral vectors, able
to permanently insert their genome into resting cells, are
missing. No doubt that in the near future lentiviral vectors –
already used for other experimental CNS gene therapy
protocols [26], will be also used to modulate CNS-confined
immune-mediated diseases. Naked plasmids, containing
expression cassettes coding for IL-4, TGFβ, interferon
(IFN)β, p55TNF receptor-Ig, p75TNF receptor-Ig, or IL-10,
have been injected intracerebrally into the right frontal lobe
of Biozzi AB/H mice affected by relapsing-remitting EAE
induced by immunization with spinal cord homogenate. This
approach proved completely ineffective on the disease
course. If, however, the same plasmids were complexed with
lipofectin (DNA-liposome complexes), then IL-4, IFNβ,
TGFβ, p55TNF receptor-Ig, and p75TNF receptor-Ig,
injected three days before disease onset, were able to
ameliorate EAE, while IL-10 remained ineffective [27, 28].
Also an adenoviral vector, engineered to contain an IL-10
expression cassette, and injected intracerebrally in mice
affected by the same EAE model, failed to inhibit the disease
[29]. On the contrary, intracranial injection of fibroblasts,
retrovirally transformed to release high amounts of IL-10,
inhibited spinal cord homogenate-induced EAE in Biozzi
AB/H mice. This latter treatment was surprisingly associated
with increased recruitment of B cells, and CD8+ T cells in
CNS inflammatory infiltrates, rather than with a decrease of
leukocyte trafficking [29]. CNS intraparenchymal injection
of retroviral-transformed fibroblasts was therapeutically
effective, ameliorating relapsing-remitting EAE in Biozzi
AB/H mice also when the delivered molecule was p75TNF
receptor-Ig [30]. Finally, replication attenuated herpes
simplex virus type-1 (HSV-1) vectors, depleted for the
gamma 34.5 gene, have been used for direct intracerebral
injection. These depleted herpetic vectors, despite being
partially replication competent, are supposed to be of
attenuated neuropathogenicity. HSV-1-derived vectors
coding for IL-4, but not for IL-10, induced protection from
acute monophasic EAE obtained in BALB/c mice by
immunization with spinal cord homogenate. Disease
inhibition was associated with decreased leukocyte
infiltration, demyelination, and axonal loss [31].
Injecting gene therapy vectors directly in the CNS is a
promising approach too. However, injection of naked DNA
or DNA-liposome complexes leads to a short-term (i.e.,for a
few days) gene expression, a relevant limitation in a chronic
disease like MS. The use of partially replicating viruses, such
as HSV-1 gamma 34.5 deletion mutants, is interesting in
experimental settings, but raises unaffordable safety issues if
transferred in human protocols. Finally, direct CNS intraparenchymal injection of the gene therapy vectors employed
in these studies allows only limited spreading from the
injection site, another major issue in MS, a disease
multifocally affecting the whole CNS.
THE EPENDYMAL
THERAPY
ROUTE
FOR
CNS
GENE
Recently, a novel strategy has been established for the
delivery of genes into the CNS. Based on the injection of
non-replicative viral vectors, this approach employs the
cerebrospinal fluid (CSF) as the driving force to deliver
therapeutic molecules to the entire CNS. Viral vectors
delivered into the CSF circulation, either through injections
into the cisterna magna (i.c.) or by stereotaxic
intracerebroventricular (i.c.v.) injections, will infect the
ependymal and leptomeningeal cells facing liquoral spaces.
Ependymal and leptomeningeal cells, infected with the viral
vector, will then produce the heterologous protein and, in
case the transgene codes for a secreted protein, release it in
the CSF circulation. From the CSF, the secreted protein can
travel within the CNS parenchyma and exert there its
beneficial effects. HSV-1-derived vectors engineered with
cytokine genes have been especially successful [32, 33].
HSV-1-derived vectors were used to deliver IL-4 in mice
with EAE, either before or after the disease onset. I.c.
injection of an IL-4-coding HSV-1 vector was able to inhibit
Gene Therapy-Mediated Modulation
chronic-remitting EAE development in Biozzi AB/H mice
immunized with the myelin oligodendrocyte glycoprotein
(MOG) 40-55 peptide [34]. Disease prevention was
associated with a decreased recruitment within the CNS of
monocyte/macrophages from the peripheral circulation. The
i.c., HSV-1-mediated, IL-4 delivery was able also to
ameliorate ongoing relapsing-remitting EAE in spinal cord
homogenate-immunized Biozzi AB/H mice, determining, in
this case, a significant modulation of the local cytokine
milieu, leading to down-regulation of pro-inflammatory
cytokines and chemokines [35]. This latter approach has
been tested also in non-human primates affected by a very
acute, invariably fatal, form of EAE induced by
immunization with whole myelin. Sixty percent of monkeys
i.c.-injected at the time of disease onset with an HSV-1
vector engineered with the human IL-4 gene, were
completely protected from EAE signs and symptoms [36].
The ependymal route using HSV-1 derived vectors has been
employed also to deliver the IFNγ gene, which was able to
both inhibit or treat MOG35-55-induced chronic EAE in
C57BL/6 mice, through the induction of in situ apoptotic
death of encephalitogenic T cells [37]. Finally, HSV-1mediated intracisternal delivery of the fibroblast growth
factor (FGF)-II gene was able to induce oligodendrocyte
precursors proliferation and migration, thus ameliorating
ongoing chronic EAE in MOG35-55-immunized C57BL/6
mice [38]. HSV-1-derived vectors, however, have a low
chance to be employed in a human clinical setting due to: a)
the short term transgene production (i.e., up to 4 weeks); b)
their possible immunogenicity; and c) their derivation from a
virus potentially very dangerous for its selective
neurovirulence. Other vectors may be, therefore, more
suitable for the “ependymal route” of administration of the
transgene. Previous data have shown that intraventricular
injection in naive mice and rats of first generation
replication-defective recombinant adenoviral vectors
(Ad.Svβgal, Ad-α1AT, and AxCAHBG) leads to the
expression of reporter genes in ependymal cells [39, 40], and
to production into the CSF of proteins coded by the
transgenes, such as β-glucoronidase and α1-antitrypsin [40,
41]. Moreover, a recent report confirmed that the injection
into the CSF space, either sub-occipitally (within the cisterna
magna) or more caudally by lumbar puncture, of an
adenoviral vector containing the β-gal reporter gene
determined a widespread infection of ependymal cells
surrounding the CSF space in non-human primates [42].
Adenoviral vectors were validated also in EAE for the
ependymal delivery of IL-10, through i.c.v. injection, in this
case resulting in the amelioration of an acute monophasic
disease induced in CSJLF1 mice by immunization with
spinal cord homogenate [43].
The viral vector-mediated CNS delivery of cytokine and
growth factor genes through the ependymal route presents
the following major advantages: (a) availability of high
cytokine and growth factor levels virtually in all areas of the
CNS, thus representing an ideal approach for multifocal
diseases, such as MS; (b) persistent therapeutic effect for
over 4 weeks after a single vector administration,
irrespective of the injection site into the CSF circulation (i.e.
ventricular route or lumbar route); and (c) absence of CNS or
peripheral unpredictable and undesirable side effects. This
Current Pharmaceutical Design, 2003, Vol. 9, No. 20 5
latter point has been especially addressed from an
immunological point of view. Neither the delivery of IL-4
[34-36], nor of IFN-γ [37], affected the functions of antigenspecific encephalitogenic T cells in the periphery, and nonhuman primates, followed for one month after i.c. injection
of an HSV-1 vector engineered to produce the human IL-4
gene, did not display any sign of CNS and peripheral toxicity
[36].
CONCLUSIONS
MS is a chronic, multifocal, immune-mediated disease of
the CNS. Thus, a non-etiological therapeutic approach has to
be able to interfere with immune processes in the whole CNS
for a prolonged time. Any peripheral route of administration
has been to date limited by the difficulty to achieve active
drug concentrations across the blood brain barrier, without
causing peripheral side effects and without interfering with
normal immune system functions. As a consequence, most
clinical trials based on the systemic administration of
immunomodulatory molecules in MS have, so far, failed [4447]. Gene therapy offers the tools to overcome these
limitations by directly targeting the brain. The different
therapeutic approaches that we have discussed here, despite
needing further development, deserve consideration as
possible alternatives in the future MS therapeutic scenario.
Engineered cells re-infused in the blood stream of
patients might represent a non-invasive approach, in which
the ex-vivo genetic manipulation could allow a tight quality
control of the gene transfer. However, along with the current
lack of any indication of their fate in vivo, and the risks
involved in the re-infusion of activated, potentially
encephalitogenic, cells, they have a limited life span. At the
moment, this approach satisfies the requirement for a
multifocal, although not chronic, disease.
Direct injection into the CNS of gene therapy vectors,
also is a promising approach. However, it requires the
intervention of a neurosurgeon and carries the risks involved
in any CNS surgery, such as bleedings or unwanted CNS
lesions. Currently available gene therapy vectors, such as
lentiviral vectors, would allow long-term expression of the
therapeutic molecule from CNS resident cells, which have a
very long life span. However, spreading from the injection
site is limited to few millimetres and the diffusion range of
the released molecule will necessarily be insufficient to
target the whole CNS. Thus, this gene therapy approach
satisfies the requirements for a chronic, although not
multifocal, CNS disease.
The ependymal route is, at the moment, the only gene
delivery protocol satisfying the requirements for a CNS
disease, which is - like MS - at the same time chronic and
multifocal. Ependymal cells, in fact, infected with helperdependent adenoviral (HD-Ad) vectors, are able to produce
an heterologous protein for over 5 months [33], and the CSF
circulation will allow the released molecule to reach virtually
all areas of the CNS. The demonstrated absence of peripheral
side effects is an additional favourable feature of this
approach. However, not all released molecules may be able
to travel from the CSF circulation to the CNS parenchyma
6 Current Pharmaceutical Design, 2003, Vol. 9, No. 20
Furlan et al.
with the same efficiency, thus limiting the nature of the
transferred gene that can be employed with this approach.
[8]
Some of the above-mentioned limitations, may be
overcome by future available technology. The development
of our knowledge on hematopoietic stem cells, for example,
may in future allow us to permanently genetically modify
subpopulations of non-pathogenic lymphocytes - able to
selectively home within the CNS - and to use them as vehicle
for local drug delivery. On the other hand, some advantages
will come from the development of innovative gene therapy
vectors able to spread over larger areas of the CNS.
[9]
There are, however, some concerns - common to any
delivery route for selective CNS gene therapy - that must be
overcome before these protocols will be applied in human
trials. Firstly, there is the immunogenicity of the gene
therapy vectors and their products: repeated, or prolonged,
exposure of foreign viral molecules to the immune system
may lead to a specific immune reaction against the gene
therapy vector, thus making infected cells the target of an
undesired auto-immune reaction [48]. In this respect, the
ependymal approach seems very safe, since neither with
HSV-1 nor with first generation adenoviral vectors - both
highly immunogenic - an immune reaction against the viral
vector has been observed in non-human primates, up to one
month after injection in the CSF circulation [36, 42].
Furthermore, genetic material originated from bacterial
plasmids, contained in gene therapy constructs, is, per se,
immunogenic and able to modulate immune and autoimmune
responses [49, 50]. Secondly, there is the impossibility to
regulate the level and timing of gene expression. Availability
of regulatable promoters, suitable for the use in humans,
would enhance the safety profile of this approach, thus
allowing to produce the desired transgene only during
specific phases of the disease (i.e., release of an antiinflammatory agent during the clinical relapse or release of a
neurotrophic growth factor during recovery phases). Despite
it has been almost ten years from the first human gene
therapy protocol, this technology has still to solve enormous
problems. It represents, however, one of the most
conceivable approaches to therapeutic challenges, such as
those posed by MS, that appear difficult to face with
conventional pharmacological tools.
ACKNOWLEDGEMENTS
Our work is supported by Minister of Health (Progetti
Finalizzati), Minister of University and Research (MIUR)
and Italian National Multiple Sclerosis Society (AISM).
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
Hull M, Lieb K, Fiebich BL. Curr Med Chem 2002; 9: 83.
McGeer PL, McGeer EG. Muscle Nerve 2002; 26: 459.
Allan SM. Ann N Y Acad Sci 2000; 917: 84.
Kumar K. Curr Opin Investig Drugs 2001; 2: 1748.
Saunders NR, Habgood MD, Dziegielewska KM. Clin Exp
Pharmacol Physiol 1999; 26: 11.
Pardridge WM. Arch Neurol 2002; 59: 35.
Martino G, Hartung HP. Curr Opin Neurol 1999; 12: 309.
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
Kieseier BC, Storch MK, Archelos JJ, Martino G, Hartung
HP. Curr Opin Neurol 1999; 12: 323.
Martino G, Adorini L, Rieckmann P, Hillert J, Kallmann B,
Comi G, et al. Lancet Neurol 2002; 1: 499.
Keegan BM, Noseworthy JH. Annu Rev Med 2002; 53:
285.
Swanborg RH. Immunol Rev 2001; 184: 129.
Brok HP, Bauer J, Jonker M, Blezer E, Amor S, Bontrop
RE, et al. Immunol Rev 2001; 183: 173.
Uccelli A. In From Basic Immunology to immune-mediated
demyelination, Martino G. and Adorini L, Eds. SpringerVerlag: Milano 1999; pp. 233-245.
Lassmann H. Clin Neurol Neurosurg 2002; 104: 168.
Gold R, Hartung HP, Lassmann H. Trends Neurosci 1997;
20: 399.
Shaw MK, Lorens JB, Dhawan A, DalCanto R, Tse HY,
Tran AB, et al. J Exp Med 1997; 185: 1711.
Dal Canto RA, Shaw MK, Nolan GP, Steinman L, Fathman
CG. Clin Immunol 1999; 90: 10.
Chen LZ, Hochwald GM, Huang C, Dakin G, Tao H,
Cheng C, et al. Proc Natl Acad Sci USA 1998; 95: 12516.
Mathisen PM, Yu M, Johnson JM, Drazba JA, Tuohy VK. J
Exp Med 1997; 186: 159.
Mathisen PM, Yu M, Yin L, Johnson JM, Kawczak JA,
Nishiyama A, et al. J Autoimm 1999; 3: 31.
Flugel A, Matsumuro K, Neumann H, Klinkert WE,
Birnbacher R, Lassmann H, et al. Eur J Immunol 2001; 31:
11.
Chen C, Rivera A, Ron N, Dougherty JP, Ron Y. Blood
2001; 97: 886.
Melo ME, Qian J, El-Amine M, Agarwal RK, Soukhareva
N, Kang Y, et al. J. Immunol 2002; 168: 4788.
Bielekova B, Goodwin B, Richert N, Cortese I, Kondo T,
Afshar G, et al. Nat Med 2000, 6: 1167.
Kappos L, Comi G, Panitch H, Oger J, Antel J, Conlon P,
Steinman L. Nat Med 2000; 6: 1176.
Consiglio A, Quattrini A, Martino S, Bensadoun JC,
Dolcetta D, Trojani A, et al. Nat Med 2001; 7: 310.
Croxford JL, Triantaphyllopoulos K, Podhajcer OL,
Feldmann M, Baker D, Chernajovsky Y. J Immunol 1998;
160: 5181.
Triantaphyllopoulos K, Croxford J, Baker, D, Chernajovsky
Y. Gene Ther 1998; 5, 253.
Croxford JL, Feldmann M, Chernajovsky Y, Baker D. J
Immunol 2001; 166, 4124.
Croxford JL, Triantaphyllopoulos KA, Neve RM, Feldmann
M, Chernajovsky Y, Baker D. J Immunol 2000; 164, 2776.
Broberg E, Setala N, Roytta M, Salmi A, Eralinna JP, He B,
et al. Gene Ther 2001; 8, 769.
Martino G, Poliani PL, Marconi PC, Comi G, Furlan R.
Gene Ther 2000; 7: 1087.
Martino G, Furlan R, Comi G, Adorini L. Trends Immunol
2001; 22: 483.
Furlan R, Poliani PL, Galbiati F, Bergami A, Grimaldi LM,
Comi G, et al. Hum Gene Ther 1998; 9: 2605.
Furlan R, Poliani PL, Marconi PC, Bergami A, Ruffini F,
Adorini L, et al. Gene Ther 2001; 8: 13.
Poliani PL, Brok H, Furlan R, Ruffini F, Bergami A, Desina
G, et al. Hum Gene Ther 2001; 12: 905.
Gene Therapy-Mediated Modulation
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
Furlan R, Brambilla E, Ruffini F, Poliani PL, Bergami A,
Marconi PC, et al. J Immunol 2001; 167: 1821.
Ruffini F, Furlan R, Poliani PL, Brambilla E, Marconi PC,
Bergami A, et al. Gene Ther 2001; 8: 1207.
Akli S, Caillaud C, Vigne E, Stratford-Perricaudet LD,
Poenaru L, Perricaudet M, et al. Nat Genet 1993; 3: 224.
Bajocchi G, Feldman SH, Crystal RG, Mastrangeli A. Nat
Genet 1993; 3: 229.
Ohashi T, Watabe K, Uehara K, Sly WS, Vogler C, Eto Y.
Proc Natl Acad Sci USA 1997; 94: 1287.
Driesse MJ, Kros JM, Avezaat CJ, Valerio D, Vecht CJ,
Bout A, et al. Hum Gene Ther 1999; 10: 2347.
Cua DJ, Hutchins B, LaFace DM, Stohlman SA, Coffman
RL. J Immunol 2001; 166: 602.
Calabresi PA, Fields NS, Maloni HW, Hanham A, Carlino
J, Moore J, et al. Neurology 1998; 51: 289.
Current Pharmaceutical Design, 2003, Vol. 9, No. 20 7
[45]
[46]
[47]
[48]
[49]
[50]
The Lenercept Multiple Sclerosis Study Group and
University of British Columbia MS/MRI Analysis Group.
Neurology 1999; 53: 457.
van Oosten BW, Barkhof F, Truyen L, Boringa JB,
Bertelsmann FW, von Blomberg BM, et al. Neurology
1996; 47: 1531.
The IFNB Multiple Sclerosis Study Group. Neurology
1993; 43: 655.
Thomas CE, Schiedner G, Kochanek S, Castro MG,
Lowenstein PR. Proc Natl Acad Sci USA 2000; 97: 7482.
Tsunoda I, Tolley ND, Theil DJ, Whitton JL, Kobayashi H,
Fujinami RS. Brain Pathol 1999; 9: 481.
Boccaccio GL, Mor F, Steinman L. Int Immunol 1999; 11:
289.