Microbubble-sonosensitiser conjugates as therapeutics in sonodynamic therapy

Article abstract of DOI:10.1039/c2cc33913g

A Rose Bengal sonosensitiser has been covalently attached to a lipid microbubble and the resulting conjugate shown to produce higher levels of singlet oxygen, enhanced cytotoxicity in a cancer cell line and a greater reduction in tumour growth than the sonosensitiser alone.

Full text of DOI:10.1039/c2cc33913g

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 8332–8334
www.rsc.org/chemcomm
COMMUNICATION
Microbubble–sonosensitiser conjugates as therapeutics in sonodynamic
therapyw
Nikolitsa Nomikou,a Colin Fowley,b Niall M. Byrne,b Bridgeen McCaughan,b
Anthony P. McHale*b and John F. Callan*b
Received 31st May 2012, Accepted 20th June 2012
DOI: 10.1039/c2cc33913g
A Rose Bengal sonosensitiser has been covalently attached to a lipid
microbubble and the resulting conjugate shown to produce higher
levels of singlet oxygen, enhanced cytotoxicity in a cancer cell line and
a greater reduction in tumour growth than the sonosensitiser alone.
sonosensitiser drugs in the vicinity of collapsing bubbles experience
such high local temperatures that ROS are generated through
pyrolysis reactions.8
The potential benefits of attaching sonosensitising drugs to
microbubbles are threefold: (1) it enables the site specific delivery
of the sonosensitisers by selectively destroying them with a non-
invasive ultrasonic stimulus. (2) Due to the tissue attenuation of
ultrasound, deep seated tumours can be accessed. This is a
particular advantage over PDT where the low tissue penetration
capability of visible light limits this technique to the treatment of
superficial tumours.9 (3) The close proximity of the sonosensitiser
to the MB should enhance the possibility of ROS generation by
either of the aforementioned mechanisms.7,8 Combined, these
benefits mean that a completely non-invasive approach for the
treatment of deep seated tumours is possible.
Microbubbles (MBs) are small (typically 1–8 mm in diameter) gas
filled microspheres that are currently used as contrast agents in
ultrasound based imaging. Clinical ultrasound uses MBs to help
opacify blood filled cavities and to visualise tissue perfusion. The
core of a MB is filled with an inert gas with low aqueous solubility
such as perfluorocarbons, coated with a stabilising shell consisting
of different materials such as phospholipids, proteins (especially
albumin) and biocompatible polymers.1 The behaviour of MBs
depends on the amplitude of ultrasound to which they are
exposed. At low acoustic pressures the MB oscillates in a
relatively symmetrical linear fashion.2 At high acoustic pressures
the MB undergoes forced expansion and compression which
results in destruction of the MB by either outward diffusion of
the gas during the compression phase, defects in the stabilising
shell, or by a complete fragmentation of the gas and shell.3 The
ability to selectively destroy MBs using an ultrasound stimulus of
appropriate energy has led to them being studied as potential
drug/gene delivery vehicles in recent years.3–5
In this manuscript we covalently attach a Rose Bengal sono-
sensitiser to the surface of a lipid based MB using a carbodiimide
based coupling protocol. We evaluate the ability of the resulting
conjugate to produce singlet oxygen and determine its cytotoxic
potential in a cancerous cell line upon ultrasound irradiation.
Finally, we examine the cytotoxic effect of this conjugate in vivo
using a human prostate tumour model in SCID mice by measuring
the % tumour growth before and after ultrasound irradiation.
Lipid MBs were prepared by sonication of an aqueous dispersion
of the lipid-based reagents in the presence of a perfluorobutane gas
stream (see ESIw). Amine functionality in the MB shell was
accomplished by the incorporation of distearoylphosphatidyl
ethanolamine-polyethylene glycol-amino (DSPE-PEG) at 5%
of the total lipid concentration (Scheme 1). These amino
functionalised MBs were characterised by optical microscopy and
observed to have an average diameter of 1.7 Æ 0.5 mm. To enable
covalent amide bond attachment of the sonosensitiser to the MB
shell it was first necessary to derivatise the commercially available
Rose Bengal sodium salt (RBNa) with carboxylic acid functionality.
This was accomplished by the nucleophilic substitution reaction
between 8-bromooctanoic acid and RBNa (Scheme 2).10 This
product (RB1) was then covalently attached to the pendant
amino groups of the MBs using standard carbodiimide coupling
techniques (Scheme 2).11 After purification by centrifugation
the MB–RB conjugates were isolated as a pink coloured milky
suspension that floated on top of the PBS solution.
Sonodynamic therapy (SDT) involves the use of ultrasound,
a sonosensitising drug and molecular oxygen to generate reactive
oxygen species (ROS). While the mechanism for ROS production
in photodynamic therapy (PDT) is well understood, the mechanism
for the generation of ROS in SDT is less clear. One suggestion is
that the process of ultrasound inertial cavitation, which involves the
formation, oscillation and collapse of gas filled bubbles in
samples irradiated with ultrasound is responsible for initiating the
generation of ROS in SDT.6 When this cavitation phenomenon
becomes dominated by inertial forces, the bubbles collapse violently
leading to sonoluminescence emission.7 This luminescence emission
may subsequently excite the nearby sonosensitiser by an energy
transfer process resulting in the generation of ROS by the very same
mechanism as in PDT. Another possible explanation is that
a Sonidel Ltd., Dublin 5, Republic of Ireland
b Department of Pharmacy and Pharmaceutical Sciences, School of
Biomedical Sciences, The University of Ulster, Northern Ireland
BT52 1SA. E-mail: j.callan@ulster.ac.uk, ap.mchale@ulster.ac.uk
w Electronic supplementary information (ESI) available. See DOI:
10.1039/c2cc33913g
Optical microscopy analysis showed the presence of pink
coloured bubbles with a concentration of 1 Â 109 MBs per ml
c
8332 Chem. Commun., 2012, 48, 8332–8334
This journal is The Royal Society of Chemistry 2012

View Article Online
conjugate) also failed to produce significant quantities of singlet
oxygen upon ultrasound irradiation. Initially, we assumed this
experiment would function as a positive control for the singlet
oxygen potential of RB1 upon ultrasonic activation. Indeed, the
sodium salt of Rose Bengal (RB) did produce comparable
singlet oxygen levels as the MB–RB conjugate when present
at a much higher concentration13 (5 mM). This suggests that at a
concentration of 1.23 mM (which was the concentration of RB1
in the MB–RB conjugate), the yield of singlet oxygen produced
by RB1 alone is too low to be measured by this assay. However,
it does indicate a powerful synergistic relationship between the
MB and RB1 in MB–RB that significantly enhances the singlet
oxygen capability of the sonosensitiser. We postulate that this
enhanced singlet oxygen production results from the ultrasound
mediated inertial cavitation of the MBs that ultimately results in
their collapse. The collapsing bubble facilitates the production
of singlet oxygen by either a sonoluminescence or pyrolysis
(or both) mediated process mentioned earlier. Indeed, Gaitan
et al., have previously demonstrated that bubbles trapped in a
standing ultrasonic wave produced repeated flashes of sono-
luminescence attributed to the occurrence of inertially dominated
bubble collapse.14,15 Shi et al. experienced a similar phenomenon
when using a very dilute suspension of a commercially available
microbubble contrast agent.16 As the efficiency of an energy
transfer process is distance dependent,17 the covalent attachment
present in MB–RB ensures a close MB–sonosensitiser distance
enabling efficient sonosensitiser excitation upon ultrasound
activation of the MB–sonosensitiser conjugate. Similarly, pyrolysis
mediated production of singlet oxygen would also benefit from a
close MB–sonosensitiser distance, as one would expect the local
temperature to be greater closer to the surface of the collapsing
MB. We are currently developing methods to ascertain which
of these two mechanisms most likely contribute to singlet
oxygen production in these systems.
Scheme 1 Schematic representation of the lipid microbubbles used in this
study. DSPC = distearoylphosphatidyl choline; PEG-40 stearate = (poly-
ethylene glycol)-40-stearate. Also shown is the chemical structure of DSPE-
PEG (distearoylphosphatidyl ethanolamine-polyethylene glycol)-amino.
Scheme 2 Synthesis of (a) RB1 and (b) the MB–RB conjugate.
(Fig. S1, ESIw). Using a standard calibration curve, we determined
the concentration of the attached Rose Bengal to be 92.7 mM. This
represents a grafting efficiency of RB1 to the MB of 37.5%. In
addition to the MB–RB conjugate a control reaction was also
performed where a solution of RB1 was mixed with the MBs
under the very same conditions and subjected to the same
purification protocol, without the presence of the coupling agents
EDC and sulfo-NHS. The purpose of this control reaction was to
determine if there would be any non-covalent interactions between
RB1 and the MBs through a self assembly process.
To determine the potential of the MB–RB conjugate to
produce singlet oxygen under ultrasound stimulation we utilized the
photo-oxidation of 1,3-diphenylisobenzofuran (DPBF) as a quan-
titative method.12 Specifically, 2 mL of the MB–RB conjugate was
added to an aerated solution of DPBF (10 mM) in an EtOH : H2O
(50 : 50) solvent system. The solution was then irradiated with
ultrasound delivered by a Sonidel SP100 sonoporator (emitting at
a frequency of 1 MHz, using an power density of 1.5 W cmÀ2
for 60 min and a 50% duty cycle at a pulse repetition
frequency of 100 Hz). This procedure was repeated for the
control MB sample and for un-conjugated RB1 alone at a
similar concentration to that present in the MB–RB conjugate.
The results are shown in Fig. 1a and reveal a significant loss in
the DPBF absorbance at 410 nm for the MB–RB conjugate
indicating significant singlet oxygen production. However, no
noticeable reduction in DPBF absorbance was observed
for the control MB sample suggesting the presence of covalent
bonds between the MB and sonosensitiser are necessary for
high yields of singlet oxygen to be produced. Surprisingly
though, the solution containing only RB1 (i.e. the sonosensitiser
alone at the same concentration as that present in the MB–RB
Fig. 1 (a) Plot of relative absorbance of DPBF at 410 nm against time for:
MB–RB conjugate (green n), control MBs (i.e. no EDC/S-NHS used) (red
&), MBs alone (purple X) and RB1 alone (blue }). (b) Plot of % cell
viability for RIF-1 cells: exposed to US only (U/S only); incubated with RB1
without US (RB1 only); incubated with RB1 and US (RB1+ U/S); with
MBs and US (MB + U/S); incubated with RB1 and MB non-covalently
linked and US (RB1+MB+ U/S); and incubated with the covalently linked
MB–RB conjugate exposed to US. (c) Plot of % tumour growth against
time for mice treated with the MB–RB conjugate with (n) and without (&)
US treatment. (d) Bar chart representing the % change in tumour growth
for treated and untreated tumour 4 days post-treatment.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 8332–8334 8333

View Article Online
To determine if the singlet oxygen generated would have the
desired cytotoxic effect in tumour cells, we carried out a similar
experiment using RIF-1 cells as a target in a tissue culture-based
bioassay. RIF-1 cells were cultured in 96-well plates and MB–RB
conjugate has no effect on tumour growth in the absence of the
ultrasound stimulus and it should also be noted that from
previous studies, the ultrasound conditions employed had no
effect on tumour growth.19,20 Since neither the conjugate nor
the stimulus exhibit toxicity, the system essentially comprises
the best attributes of a targeted therapeutic system.
was added to selected wells at a MB concentration of 2 Â 107 mlÀ1
.
These were then treated with ultrasound for 30 s, using a frequency
of 1 MHz, an ultrasound power density of 1.5 W cmÀ2 and a duty
cycle of 50% (pulse frequency = 100 Hz). Control wells containing
RB1 alone and those containing MBs and RB1 at the same
concentration as in the MB–RB conjugate were used for
comparative purposes. We also included controls for the effect
of ultrasound alone and MBs alone for completeness. Following
irradiation the cells were incubated for 24 hours before the
viability was determined using an MTT assay.18 The results show
a 72% reduction in cell viability for the MB–RB conjugate upon
ultrasound irradiation while the un-conjugated MB/RB1 solution
was significantly less effective with a 51% reduction in cell
viability. Given RB1 alone produced a similar cytotoxic effect at
44% reduction, this further emphasises the importance of a
covalent interaction between the MB and sonosensitiser for
enhanced cytotoxicity. The fact that RB1 alone produced a
cytotoxic effect when present at the same concentration as in
MB–RB while it produced negligible singlet oxygen in the DPBF
assay may be due to two reasons (i) the DPBF assay is not
sufficiently sensitive to measure levels of singlet oxygen that are
cytotoxic in a cellular environment or (ii) other ROS in addition to
singlet oxygen may be generated that are not detected by the
DPBF assay. Nonetheless, these results validate those from the
DPBF assay in that a significant enhancement in toxicity is
obtained by covalent attachment of the sononsensitiser to the MB.
To determine the therapeutic efficiency of the MB–RB
conjugate in vivo, tumours were induced in BALB/c SCID
mice using the modified human prostate cell line LNCaP-Luc.
Once the tumours were 1.24 cm3, a 30 ml aliquot of the
MB–RB conjugate (2 Â 108 MB mlÀ1) was injected into the
tumour. In these experiments, intratumoral injection was chosen
as the administration route in order to preclude variables resulting
from systemic delivery. The tumours were then treated with
ultrasound for 3 min using a frequency of 1 MHz, an ultrasound
power density of 3.5 W cmÀ2 and 30% duty cycle (100 Hz pulse
frequency). Control mice that were administered the MB–RB
conjugate but not exposed to ultrasound irradiation were also
used for comparative purposes. The results are shown in Fig. 1c
and d and reveal a significant reduction in tumour size for those
animals treated with the MB–RB conjugate and ultrasound
compared to the MB–RB conjugate alone. In fact, 4 days after
treatment, tumours on animals treated with the MB–RB conjugate
and ultrasound actually regressed and were found to be 18%
smaller than the original pre-treatment size (Fig. 1d). On the other
hand tumours on animals treated with MB–RB in the absence of
the ultrasound stimulus had increased in size by 50% on day 4. It
was interesting to note that it was not until day 10 that tumours
treated with MB–RB and ultrasound reached their pre-treatment
size, whereas those treated with MB–RB in the absence of
ultrasound had increased to 100% that of the pre-treatment
tumour size (Fig. 1c). Essentially these results dramatically
demonstrate the therapeutic potential of our approach and
highlights the necessity for a combination of ultrasound and
the conjugate at the target site. We demonstrate here that the
In summary, a MB–sonosensitiser conjugate has been prepared
and observed to produce significant quantities of singlet oxygen
and be more cytotoxic to a cancerous cell line when irradiated with
ultrasound compared to the un-conjugated sonosensitiser at the
same concentration. In addition, ultrasound irradiation of animals
treated with the MB–RB conjugate significantly reduced tumour
growth when compared to those that received the drug but no
ultrasound. We attribute these effects to either a sonoluminescence
or pyrolysis mediated production of ROS. Either of these processes
would benefit from a close MB–sonosensitiser separation as offered
by the covalent linkage present in MB–RB. This approach offers
the potential to deliver sonosensitiser drugs to deep seated tumours
and activate them in a non-invasive manner. Furthermore, the
covalent attachment of the sonosensitiser not only improves the
singlet oxygen quantum yield but also reduces the possibility of
the sonosensitiser leaching from the bubble prior to being activated
at the target site. This not only improves the likelihood of a greater
proportion of the drug reaching its target site, but also reduces the
inadvertent activation of the sonosensitiser with ambient light,21 as
Rose Bengal is also known to be a potent photosensitiser.
Notes and references
1 S. Shashank and M. Borden, Bubble Sci., Eng., Technol., 2009, 1, 3.
2 V. R. Stewart and P. S. Sidhu, Br. J. Radiol., 2006, 79, 188–194.
3 S. Hernot and A. L. Klibanov, Adv. Drug Delivery Rev., 2008, 60,
1153–1166.
4 C. R. Mayer and R. Bekeredjian, Adv. Drug Delivery Rev., 2008,
60, 1177–1192.
5 C. R. Mayer, N. A. Geis, H. A. Katus and R. Bekeredjian, Expert
Opin. Drug Delivery, 2008, 5, 1121–1138.
6 S. Umemura, N. Yumita, K. Umemura and R. Nishigaki, Cancer
Chemother. Pharmacol., 1999, 43, 389–393.
7 S. Umemura, N. Yumita, R. Nishigaki and K. Umemura, Jpn. J.
Cancer Res., 1990, 81, 962–966.
8 V. Misik and P. Riesz, Ann. N. Y. Acad. Sci., 2000, 899, 335–348.
9 M. Wainwright, Therapeutic Applications of Near-Infrared Dyes,
Color. Technol., 2010, 126, 115–126.
10 N. Sugita, K. Kawabata, K. Sasaki, I. Sakata and S. Umemura,
Bioconjugate Chem., 2007, 18, 866–873.
11 I. Yildiz, E. Deniz, B. McCaughan, S. F. Cruickshank, J. F. Callan
and F. M. Raymo, Hydrophilic, Langmuir, 2010, 26, 11503–11511.
12 S. O. McDonnell, M. J. Hall, L. T. Allen, A. Byrne, W. M. Gallagher
and D. F. O’Shea, J. Am. Chem. Soc., 2005, 127, 16360–16361.
13 B. McCaughan, C. Rouanet, C. Fowley, N. Nomikou, A. P. McHale,
P. A. McCarron and J. F. Callan, Bioorg. Med. Chem. Lett., 2011,
21, 5750.
14 D. F. Gaitan, L. A. Crum, C. C. Church and R. A. Roy, J. Acoust.
Soc. Am., 1992, 91, 3166–3183.
15 D. Hammer and L. Frommhold, J. Mod. Opt., 2001, 48, 239–277.
16 W. T. Shi, F. Forsberg, A. Tornes, J. Ostensen and B. B. Goldberg,
Ultrasound Med. Biol., 2000, 26, 1009–1019.
17 J. R. Lakowicz, in Principles of Fluorescence Spectroscopy, Springer,
New York, 2006.
18 A. P. McHale and L. McHale, Cancer Lett. (Shannon, Irel.), 1988,
41, 315–321.
19 A. M. Rollan, A. Smyth, P. Hughes, C. N. Reid and A. P. McHale,
Cancer Lett. (Shannon, Irel.), 2005, 222, 1273–1280.
20 J. Larkin, D. Soden, C. Collins, M. Tangney, J. M. Preston,
L. J. Russell, A. P. McHale, C. Dunne and G. C. O’Sullivan,
Eur. J. Cancer, 2005, 41, 1339–1348.
21 M. Ochsner, J. Photochem. Photobiol., B, 1996, 32, 3–9.
c
8334 Chem. Commun., 2012, 48, 8332–8334
This journal is The Royal Society of Chemistry 2012