Neuroblastoma targeting by c-myb-selective antisense
oligonucleotides entrapped in anti-GD2 immunoliposome: immune
cell-mediated anti-tumour activities.
Liposome encapsulation of anticancer agents results in reduced
side effects of the entrapped drug and improved therapeutic efficacy.
The external surface of the lipidic envelop can be coupled with
antibodies directed against tumour-associated antigens. The resulting
immunoliposomes allow to increase the therapeutic index of cytotoxic
drugs while minimising their systemic toxicity. In this regard,
the disialoganglioside GD2 is a very promising tumour-associated
antigen since it is expressed at high intensity on human neuroblastoma
cells, but is detected only in normal cerebellum and peripheral
nerves. Immunoliposomes can be used as vectors to deliver antisense
oligonucleotides to cancer cells with the aim to modulate oncogene
expression. Furthermore, antisense oligonucleotides have attracted
much interest because of their ability to stimulate immune responses.
Here, we will describe a novel experimental therapeutic approach
for Neuroblastoma based on anti-GD2 liposomal c-myb-selective
antisense oligonucleotides.
Introduction
Neuroblastoma (NB) is one of the most common solid tumours of
paediatric age1 and, due to its unfavourable clinical and biological
behaviour in more than half of the cases, represents a formidable
therapeutic challenge. Despite of aggressive treatment regimens,
the overall survival rate, for patients suffering from advanced
stage NB, has not been prolonged in a satisfactory manner2.
In this scenario, the need of developing new therapeutic approaches
is an emerging query.
The identification of different gene expression profiles in malignant
cells and their normal counterparts has highlighted the importance
of tumour-associated genes or mutated genes as potential target
in cancer treatment. The c-myb proto-oncogene is the best characterised
member of a family of transcription factor genes. Its expression
has been reported in several solid tumours of different origin,
including NB, in which it is linked to cell proliferation and
differentiation3. The use of antisense oligonucleotides (asODNs)
to selectively target tumour-associated genes without affecting
the normal ones appears to be a promising new approach. Furthermore,
it is already well known that antisense oligonucleotides can exert
anti-tumour effects via an indirect, immune-stimulatory mechanism.
Indeed, asODNs can be designed to contain unmethylated cytosine-guanine
(CpG) motifs able to trigger cytokine secretion and to stimulate
the innate immune system4-6. These features point to CpG-containing
asODNs as potentially useful immune adjuvants in cancer therapy.
The clinical application of asODNs has been limited by their high
sensitivity to cleavage by ubiquitous nucleases in vivo. This
problem can be circumvented by the synthesis of chemically modified
asODNs7. However, structural modifications of the asODNs molecules,
e.g., addition of phosphorothioate groups to increase their resistance
to nuclease degradation, may compromise their biological activities
or lead to non-specific binding.
Lipid-based systems have been shown to improve the in vivo delivery
of asODNs through increase of blood stability, cellular uptake
and specificity8-11. Moreover, sterically stabilised immunoliposomes
displaying tumour-directed antibodies on their external surface
appear very effective at selectively delivering drugs or diagnostic
agents to target cells12,13. The disialoganglioside GD2 is a tumour-associated
antigen very appealing for using in target approaches. Indeed,
it is widely expressed by cancer cells of neuroectoderma origin,
while its expression in normal tissues is confined to cerebellum
and peripheral nerves at very low levels14.
We recently demonstrated that c-myb asODNs entrapped within GD2-targeted
liposomes (targeted-liposome-myb-as) inhibited the growth of NB
cell lines by specifically triggering the down-modulation of c-myb
protein expression15 and increased the survival of mice grafted
with human NB cells16.
Here, we report an overview on the application of anti-GD2 liposomal
c-myb antisense oligonucleotides as a novel therapeutic approach
for Neuroblastoma.
In vitro effects of anti-GD2 liposomal c-myb selective antisense
oligonucleotides.
In the last decade, different approaches have been adopted to
allow the therapeutic delivery of antisense oligonucleotides.
Lipoplexes, obtained by mixing antisense oligonucleotides with
various cationic lipids, showed improved biological stability
and cellular uptake, but their large size and excess positive
charge accelerated their in vivo clearance by the mononuclear
phagocyte system17. Subsequently, the passive entrapment of asODNs
into liposomes produced using neutral lipids was found to enhance
the biological performance, however with very low binding capability
in vivo18,19. Recently, we produced charge-neutralized liposome-asODNs
particles following a novel technological method20 with some modifications15.
Charge interactions between a cationic lipid and negatively charged
asODNs were optimized and a layer of neutral lipids was employed
to coat the outer surface of the cationic lipid-asODNs particles
figure 1. These coated cationic liposomes were
subsequently coupled to a monoclonal antibody (mAb) against the
tumour-associated antigen GD2, in order to target them selectively
to NB tumour cells. From now onwards, these liposomes will be
referred to as targeted-liposomes.
 |
Figure 1
Representative image of the novel lipid-based carrier for
antisense oligonucleotides: antibody-targeted coated cationic
liposomes. |
Targeted-liposome-myb-as15,20 were used in experiments of cell
growth inhibition after their specific binding to GD2 positive
NB cell lines. Two Neuroblastoma cell lines (GI-LI-N and HTLA-230),
positive for the expression of both GD2 (evaluated by FACS analysis)
and the c-myb protein (tested by Western Blot analysis), were
used to asses the efficacy of our liposomal formulation. Targeted
liposome-myb-as were efficacious at inhibiting the cell growth
of both NB cell lines used, while they did not affect cellular
proliferation of GD2 negative, c-myb protein positive, cell lines
(HeLa and Jurkat) that were tested as control figure 2.
Furthermore, inhibition of cell proliferation was directly related
to the down-modulation of c-myb protein expression. A time-dependent
modulation of c-myb protein was observed9,15. Targeted-liposomes
containing a scrambled sequence oligonucleotides had no effect
on both NB cell lines tested, demonstrating the sequence specificity
of the c-myb-as oligonucleotide9,15.
 |
Figure
2
Cell growth inhibition of two GD2-positive NB cell lines
(GI-LI-N and HTLA-230) and two GD2-negative cell lines (HeLa
and Jurkat), by treatment with targeted-liposome-myb-as.
Cells were treated with the liposomal formulation at a concentration
of 100 mg/ml of oligonucleotides from the start of the experiments
and every other day for 8 days. Two hours after each treatment
cells were washed to remove unbound asODNs and transferred
to fresh and complete medium. Finally, cells were detached,
stained with trypan blue and counted by microscope. Data
are expressed as mean ± STD of three different experiments. |
These results clearly show that the encapsulation of c-myb asODNs
within GD2-targeted liposomes can protect non-targeted cells from
the potential toxicity of the entrapped asODNs while enhancing,
the sequence specific inhibitory effects of c-myb asODNs against
the target cell population.
Non-antisense effects of targeted-liposomes containing oligonucleotides.
Antisense oligonucleotides containing so called CpG islands in
their backbone have attracted attention in the last decade due
to their ability to stimulate of the innate immune system. More
recently, it has been demonstrated that encapsulation of CpG-containing
asODNs within liposomes enhances the aforementioned immune-stimulating
effects21.
CpG-containing myb-as oligonucleotides entrapped within targeted
liposomes (targeted-liposome-CpG-myb-as) were used both in in
vitro and in ex vivo immune stimulation studies. We evaluated
the expression of the CD69 antigen, an early marker of cellular
activation, on various immunologically relevant cell types stimulated
by our targeted formulation16. Briefly, the ex vivo studies were
performed on splenocytes collected from tumour-bearing nude mice
that received CpG-myb-as encapsulated in either non-targeted or
targeted liposomes. Flow cytometric analysis indicated that targeted-liposome-CpG-myb-as
were able to induce a very rapid increase of CD69 expression on
monocytes, B lymphocytes and Natural Killer (NK) cells. The non-targeted
formulations, conversely, did not produce any activation of these
cell populations. Subsequently, the evaluation of the in vitro
stimulation of spleen cells, collected from tumour-free nude mice,
confirmed that targeted-liposome-CpG-myb-as lead to the up-regulation
of the CD69 antigen, as well as on granulocytes and dendritic
cells.
Further evidence of immune system activation came from the detection
of various pro-inflammatory cytokines in the serum of tumour-bearing
nude mice injected intravenously with targeted-liposome-CpG-myb-as.
Thus, serum levels of IL-12, IFN-g, IL-1b and TNF-a indicated
a strong activation of the innate immune system16. The kinetic
secretion profiles, together with the results of cellular activation,
suggested that quick production of IL-12 was likely attributable
to activated macrophages. IL-12 triggers the activation of NK
cells that, in turn, secrete IFN-g through a cascade mechanism.
The above effects are consistent with previous reports showing
that asODNs possess immune stimulating activities that are strictly
dependent on the presence of unmethylated CpG motifs5 but unrelated
to the specific antisense effects4,22. Indeed, bacterial and mammalian
DNA differ for the presence in the former of mostly methylated
CpG motifs23. Immune effector cells have evolved pattern recognition
receptors that, by binding CpG-motifs, are able to evoke protective
immune responses against bacteria. Specific recognition of CpG
motifs is mediated by the TLR9 receptor, which belongs to the
super-family of Toll-like receptors24. Furthermore, it has been
recently demonstrated that the encapsulation of CpG-containing
oligonucleotides in lipidic particles greatly increases their
immunostimulatory effects21.
In vivo anti-tumour effects of targeted-liposome-CpG-myb-as
in an experimental murine model of human Neuroblastoma.
A clinically and biologically relevant experimental mouse model
of human Neuroblastoma, already established in our laboratory16,25,
has been used to assess the potential usefulness of GD2-targeted
formulations entrapping CpG-containing asODNs.
In the experimental therapeutic setting, NB-bearing nude mice
were subjected to intravenous injection of different targeted
or non-targeted formulations, entrapping either asODNs or scrambled
(CpG-myb-scr) sequences (all of them CpG-containing). Briefly,
treatment schedule started four hours after cell inoculation,
four doses per week, followed by three day rest (8 doses total).
Survival curves showed that only mice treated with targeted-liposome-CpG-myb-as
lived longer than control mice and those injected with targeted-liposome-CpG-myb-scr
figure 3. However, also non-targeted-liposomes
(containing either CpG-myb-as or CpG-myb-scr) produced anti-tumour
effects, inducing almost the same increase in life spans, whereas
free CpG-myb-as or CpG-myb-scr did not mediate any anti-tumour
activities (data not shown).
These therapeutic results clearly suggest that the efficacy of
targeted-liposome-CpG-myb-as is not attributable exclusively to
the down-modulation of the c-myb protein, carried out by the sequence
selectivity of asODNs. Indeed, also targeted-liposome containing
a scrambled sequence produced anti-tumour effects (even if at
lower extent). The most likely explanation for our findings is
that both direct modulation of c-myb protein expression and indirect
mechanisms of immune system stimulation were involved.
In our immunocompromised murine model devoid of T cells, the only
effector cells that could be recruited as putative effectors of
tumour cells killing are NK cells. The confirmation of our hypothesis
derives from standard 51Cr release assay, in which NK cells (effectors),
collected from NB-bearing mice stimulated with targeted-liposome-CpG-myb-as,
were able to specifically lyse HTLA-230 target cells16.
 |
| Figure
3
Survival of NB-bearing nude mice after injection of oligonucleotides,
either free or encapsulated within liposomal formulations.
Nude mice were injected intravenously with 3.5 X 106 HTLA-230
neuroblastoma cells. After 4 hours, each mouse received
50 mg of oligonucleotides, either free or encapsulated in
targeted or non-targeted liposomes. Control mice received
HEPES-buffered saline. Data are representative of three
different experiments.
|
Therapeutic experiments conducted in SCID-bg mice (lacking T
lymphocytes, B lymphocytes and NK cells) definitively confirmed
the central role of NK cells in tumour cell killing. Indeed, the
ablation of NK cells resulted in a complete loss of anti-tumour
activity in mice treated with targeted-liposome-CpG-myb-scr, while
targeted-liposome-CpG-myb-as maintained a partial efficacy, due
to the specific activity of asODNs sequence specific for c-myb16.
In conclusion, the encapsulation of c-myb antisense oligonucleotides
in GD2-targeted liposomes increased their therapeutic activity
against NB cells by a direct mechanism related to downregulation
of c-myb proto-oncogene expression and, on the other hand, by
an indirect mechanism that involves the stimulation of the innate
immune system, carried out by CpG-motifs.
We believe that targeted-liposomes-CpG-myb-as deserve clinical
evaluation as adjuvant treatment in the conventional therapy for
advanced-stage Neuroblastoma or for residual disease following
incomplete surgery or early micrometastatic lesions.
Perspectives
The potential relevance of the experimental therapeutic approach
described herein could be assayed in a tumour-bearing host with
an intact immune system. It may be envisaged that the availability
of large amount of IFN-g, consequently to CpG stimulation of the
innate immune system has a strong impact on the tumour cells themselves.
Indeed, in human neuroblastoma, IFN-g can sensitise tumour cells
to apoptosis26 or induce their differentiation, especially in
conjunction with TNF-a27. Alternatively, IFN-g can increase expression
of the class I major histocompatibility complex antigen on the
cell surface28, thus allowing neuroblastoma cells killing by tumour
specific cytotoxic T lymphocytes. In this scenario, it is apparent
that administration of CpG-containing oligonucleotides-liposomal
in a tumour-bearing immunocompetent host would give rise to multiple
opportunities for the host immune system to destroy cancer cells.
Aknowledgements
We thank Professor Allen T. M. for helpfull discussion. Work supported
by Italian Neuroblastoma Foundation and Italian Association for
Cancer Research. Brignole C. is a recipient of a FIRC fellowship;
Pastorino F. is a recipient of an Italian Neuroblastoma Foundation
fellowship.
Brignole C., Pastorino F.,
Marimpietri D., Di Paolo D.,
Zancolli M., Pagnan G., Ponzoni M.
Servizio per la terapia di differenziazione,
Laboratorio di oncologia,
Ospedale pediatrico G. Gaslini,
Genova, Italia.
