- Email: [email protected]
Oxygen carrying microbubbles for enhanced sonodynamic therapy of hypoxic tumours
Oxygen Carrying Microbubbles for Enhanced Sonodynamic Therapy of Hypoxic Tumours Conor McEwan, Joshua Owen, Eleanor Stride, Colin Fowley, Heather Nesbitt, David Cochrane, Constantin.C. Coussios, M. Borden, Nikolitsa Nomikou, Anthony P. McHale, John F. Callan PII: DOI: Reference:
S0168-3659(15)00094-2 doi: 10.1016/j.jconrel.2015.02.004 COREL 7553
To appear in:
Journal of Controlled Release
Received date: Revised date: Accepted date:
9 October 2014 20 December 2014 3 February 2015
Please cite this article as: Conor McEwan, Joshua Owen, Eleanor Stride, Colin Fowley, Heather Nesbitt, David Cochrane, Constantin.C. Coussios, M. Borden, Nikolitsa Nomikou, Anthony P. McHale, John F. Callan, Oxygen Carrying Microbubbles for Enhanced Sonodynamic Therapy of Hypoxic Tumours, Journal of Controlled Release (2015), doi: 10.1016/j.jconrel.2015.02.004
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Oxygen Carrying Microbubbles for Enhanced Sonodynamic Therapy of Hypoxic Tumours
SC R
IP
T
Conor McEwan1, Joshua Owen2, Eleanor Stride2, Colin Fowley1, Heather Nesbitt1, David Cochrane1, Constantin. C. Coussios2, M. Borden3, Nikolitsa Nomikou4, Anthony P. McHale1 and John F. Callan1*.
MA
NU
1. Biomedical Sciences Research Institute, University of Ulster, Coleraine, Northern Ireland, U.K. BT52 1SA. 2. Oxford Institute of Biomedical Engineering, University of Oxford, UK, OX3 7DQ. 3. Department of Mechanical Engineering, University of Colorado, 1111 Engineering Drive, Boulder, CO 80309, USA. 4. Division of Surgery & Interventional Science, Medical School, University College London, London W1W 7EJ, UK
TE
D
* Corresponding author: Ph: 00442870123059; Email: [email protected]
Abstract: Tumour hypoxia represents a major challenge in the effective treatment of solid
CE P
cancerous tumours using conventional approaches. As oxygen is a key substrate for Photo/ Sono-dynamic Therapy (PDT / SDT), hypoxia is also problematic for the treatment of solid tumours using these techniques. The ability to deliver oxygen to the vicinity of the tumour
AC
increases its local partial pressure improving the possibility of ROS generation in PDT / SDT. In this manuscript, we investigate the use of oxygen-loaded, lipid-stabilised microbubbles (MBs), decorated with a Rose Bengal sensitiser, for SDT-based treatment of a pancreatic cancer model (BxPc-3) in vitro and in vivo. We directly compare the effectiveness of the oxygen-loaded MBs with sulphur hexafluoride (SF6)-loaded MBs and reveal a significant improvement in therapeutic efficacy. The combination of oxygen-carrying, ultrasoundresponsive MBs, with an ultrasound-responsive therapeutic sensitiser, offers the possibility of delivering and activating the MB-sensitiser conjugate at the tumour site in a non-invasive manner, providing enhanced sonodynamic activation at that site. Key words: hypoxia, microbubbles, SDT, Rose Bengal, cancer, pancreatic. 1
ACCEPTED MANUSCRIPT 1.0 Introduction: Hypoxia within solid tumours is a key determinant in the effectiveness of radiation- and chemotherapy-based treatments and arises as a consequence of the atypical vasculature that is characteristic of growing tumours [1]. Essentially, this atypical vasculature
IP
T
results in the development of areas within the tumour where oxygen demand outstrips oxygen supply. Hypoxic fractions ranging from 10 to 30 % are present in most solid tumours
SC R
regardless of size [2]. Because hypoxia negatively influences the effectiveness of conventional therapies, it is usually associated with poor prognosis. Pancreatic ductal
NU
adenocarcinoma (PDA) is an extremely aggressive malignancy with fewer than 20% of those diagnosed being eligible for curative treatment [3]. PDA possesses a pronounced hypoxic
MA
tumour microenvironment, with high levels of expression of the hypoxia biomarker HIF-1 (hypoxia inducible transcription factor) serving as a predictor of poor clinical outcome [4]. In
D
addition, while PDA shows initial sensitivity to the benchmark chemotherapeutic drugs 5-
TE
fluorouracil and gemcitabine, it is usually followed by rapid development of resistance [5]. Therefore, novel therapies that may improve the therapeutic outcome in patients with PDA or
CE P
conditions with similar morbidities are of great interest. Photodynamic therapy (PDT) has emerged as an alternative to conventional chemo-
AC
and radiation-based therapies in the treatment of certain cancers such as skin, head and neck, oesophageal and prostate cancer [6]. PDT involves the simultaneous use of a sensitising drug, light of an appropriate wavelength and molecular oxygen to generate singlet oxygen and other Reactive Oxygen Species (ROS) that result in cytotoxic effects. The main attraction of PDT as a clinical treatment is that production of the cytotoxic species can be localised to the desired area through control of the light source reducing collateral damage to surrounding healthy tissue [7]. However, a major factor prohibiting the more widespread use of PDT is that the light used to excite currently approved sensitisers is in the visible region of the electromagnetic spectrum limiting penetration through mammalian tissue [8]. This restricts use of the approach to the treatment of superficial tumours and limits its effective clinical use in treating more deeply seated lesions.
2
In addition to the above
ACCEPTED MANUSCRIPT challenges and since PDT is fuelled by the presence of oxygen, hypoxia can limit the effectiveness of PDT. Indeed it has been demonstrated that use of PDT can exacerbate hypoxia as a result of photodynamically-driven conversion of oxygen to singlet oxygen [9].
IP
T
Since hypoxia is associated with poor prognosis, this may not be desirable, particularly in the
SC R
treatment of larger lesions.
More recently, our group and others, have demonstrated that several sensitisers activated by light can also be activated by ultrasound, an approach known as Sonodynamic
NU
Therapy (SDT) [10,11]. SDT offers significant advantages over PDT because ultrasound is widely accepted as a cost effective and safe clinical imaging modality and, unlike light, can
MA
be tightly focused with penetration in soft tissue up to several tens of centimetres depending on the frequency used. We have previously shown by attaching the sensitiser Rose Bengal
D
(RB) to the surface of a lipid stabilised microbubble (MB), that the resulting conjugate
TE
produced more singlet oxygen and was more effective in vitro and in vivo than the sensitiser alone at the same concentration [12]. MBs are gas filled, lipid or polymer stabilised bubbles
CE P
of micrometre dimension and are approved as contrast agents in diagnostic ultrasound [13]. At low acoustic pressures, MBs undergo repeated volumetric oscillations, a process known
AC
as stable cavitation [14]. However, at higher acoustic pressures, MBs undergo violent expansion and compression typically resulting in destruction of the MB, a process known as inertial cavitation [15]. The process of inertial cavitation can be accompanied by the emission of light (sonoluminescence) and / or a significant localised increase in temperature [16,17]. We have reasoned that placing sensitisers in close proximity to MBs undergoing inertial cavitation could enhance their efficiency at generating ROS through sono-luminescence or pyrolysis- mediated processes. Moreover, attachment of the sensitiser to the MB surface enables the potential of ultrasound-mediated delivery and activation of the sensitiser at the tumour site in a minimally invasive manner.
3
ACCEPTED MANUSCRIPT As an extension of the above approach, in this manuscript we investigate the ability of an oxygen-carrying MB-sensitiser conjugate to improve the sonodynamic effect under hypoxic conditions. To this end we assess the ability of the conjugate to deliver enhanced
IP
T
ultrasound-mediated production of ROS in hypoxic cell free systems and examine its ability to provide enhanced ultrasound-mediated cytotoxic effects on pancreatic cancer cells
SC R
cultured in an hypoxic environment. Finally, we assess ultrasound-mediated therapeutic effects using the conjugate as a sensitiser and human pancreatic xenograft BxPc-3 tumours
NU
in mice as a target.
MA
2.0 Materials & Methods
2.1 Reagents and Equipment: Rose Bengal sodium salt, 2-bromoethylamine, NHS-biotin, MTT, avidin, FITC avidin and ethanol were purchased from Sigma Aldrich (UK) at the
D
highest grade possible. Singlet oxygen Sensor Green (SOSG) was purchased from
TE
Invitrogen (UK). 1,2-Distearoyl-sn-glycero-3-phosphocholine (DSPC), 1,2-distearoyl-sn-
CE P
glycero-3-phosphoethanolamine-N-(polyethylene glycol)-2000 DSPE-PEG(2000), and 1,2distearoyl-sn-glycero-3-phosphoethanolamine-N-[biotinyl(polyethylene glycol)-2000] DSPEPEG(2000)-biotin were purchased from Avanti Polar Lipids (Alabaster, Alabama, USA).
AC
Sulphur hexafluoride and Oxygen were purchased from BOC,Industrial Gases, UK. Phosphate Buffered Saline was purchased from Gibco, Life Technologies, UK.
DisteNMR
spectra were recorded on a Varian 500 MHz spectrometer. ESI-MS characterisation of was achieved using a LCQTM quadrupole ion-trap mass spectrometer (Finnigan MAT, San Jose, California, USA) utilising electrospray ionisation (ESI). Fluorescence measurements were recorded on a Varian Cary Eclipse Spectrometer using 10 mm quartz cuvettes. Optical microscope images were taken with an optical microscope (Leica DM500 optical microscope) and fluorescence images were taken with a Nikon Eclipse Ti fluorescence microscope, FITC filter set. Dissolved oxygen was measured using a Thermo Scientific Orion Star A216 bench top. Oxygen exchange was CDI Blood Parameter Monitoring System 500 (Terumo UK Ltd. Tamesis, Surrey, UK). 4
ACCEPTED MANUSCRIPT 2.2 Synthesis of Biotin-Rose Bengal: Rose Bengal amine was prepared following a literature procedure [18]. To a solution of Rose Bengal amine (74 mg, 74µmol) in 2 mL DMSO, NHS Biotin (25mg, 74 µmol) dissolved in PBS buffer 2 mL (pH 9) was added.
IP
T
Triethylamine (10 µl, 0.71µmol) was added to the mixture which was then stirred at room temperature for 12 hrs. The resulting solution was placed in a solution of hexane:chloroform
SC R
(200ml 8:2) and stirred for 24 hrs at room temperature . The solution filtered and the filtrate retained. The filtrate was washed three times with diethyl ether and dried in a vacuum oven
NU
to yield the product as a red powder (14 mg, 10%).1H NMR (500 MHz, DMSO-d6): 1.28 (m, CH2, 2H), 1.45 (m, CH2, CH, 3H ), 1.57 (m, CH, 1H ), 1.60 (t,CH2, 2H), 2.52 (q, CH, 1H),
MA
2.58 (m,CH, 1H), 2.79 (q, CH,1H), 2.97 (t, CH2, 2H), 3.10 (m, SCH,1H), 3.85 ( t, OCH2, 2H), 4.09 (t,CH,1H), 4.28 (t, CH,1H), 6.34 (s, NH,1H), 6.41 (s, NH,1H), 7.46 (s, ArH, 2H), 7.56 (brs, NH, 1H).
13
C NMR (125 MHz, DMSO-d6): 8.98, 9.03, 25.52, 28.57, 29.45, 31.72,
TE
D
35.52, 37.34, 46.07, 53.13, 55.86, 59.66, 61.48, 64.65, 76.50, 97.69, 110.56, 124.271, 129.37, 130.48, 132.34, 134.79, 135.50, 136.41, 139.79, 157.42, 163.13, 163.57, 172.23,
CE P
172.68, 228.31, 230.86 –ve ESMS: calculated for C32H23Cl4I4N3O7S = 1243.0 found =1241. I.R. vmax (cm-1) 3425, 2924, 2369, 1683, 1576, 1541, 1342. M.P. 188-191°C 2.3 Preparation of O2 and SF6 loaded Microbubbles: DSPC, DSPE-PEG (2000), DSPE-
AC
PEG(2000)-biotin dissolved in chloroform were mixed in a glass vial giving a ratio of 82:9:9 mole % using 3.5 mg, 1.35 mg and 1.45 mg of each lipid respectively. This was covered with parafilm pierced with holes and placed on a hot plate for over 12 hours to evaporate the chloroform. PBS (5 ml) was then added to the dried lipid film followed by heating above the lipid phase transition temperature (>70 °C) under constant magnetic stirring for 30 minutes. The suspension was then sonicated with a Microson ultrasonic cell disruptor, 100 Watts, 22.5 kHz at power setting 4 for 90 seconds, the headspace was filled with SF6 and the gas liquid interface was sonicated (power 19) for 20 seconds producing MBs. The MB suspension was cooled in an ice bath for approximately 10 minutes. To ensure that the MB surface was biotin functionalised the MBs were mixed with FITC avidin, centrifuged and analysed by fluorescence microscopy. The MBs were observed to fluorescence against the 5
ACCEPTED MANUSCRIPT background indicating that avidin had successfully bound to the surface as shown in Figure 2.
T
50 μl of Avidin dissolved in water (10 mg/ml) was then added to the cooled MB
IP
suspension and allowed to mix for 10 minutes. The suspension was then centrifuged (300
was divided into two freeze drying vials.
SC R
RCF, 10 minutes) and resulting microbubble “cake” was concentrated into 1 ml of PBS. This For the SF6MBs the vials were then crimped
(sealed with a metal cap). To create OxyMBs the headspace of the vial and the MB
NU
suspension was exposed to a positive pressure of oxygen for 2 minutes. The exchange efficiency was determined by measuring the increase in ppO2 when 1 mL of Oxy MBs was
MA
added to 3 mL of deionised water using a CDI Blood Parameter Monitoring System 500. An increase of 37 kPa in ppO2 was observed for the OxyMBs while no change was detected
D
when the same volume of SF6MBs or vehicle alone was added. This suggests the volume of
TE
OxyMBs added (1 mL) contains ~2 mol of O2, which is the same approximate amount as the total gas content in 1 mL of the OxyMBs determined from the concentration and size
CE P
distribution measurements in Figure 1. Once the gas exchange was complete, the vial was then crimped and the MBs were used directly for experiments within 3 days. Following
AC
preparation as described above, MB samples were imaged under conventional optical microscopy to determine their size distribution and concentration. 10 μl was removed from each suspension and diluted in 90 μl of PBS followed by examination on a haemocytometer (Bright-Line, Hausser Scientific, Horsham, PA, USA). Images were obtained with at 40 x objective lens with a Leica DM500 optical microscope. The bubble size distribution and concentration were then obtained using purpose written image analysis software in Matlab (2010B, The MathWorks, Natick, MA, USA)[19].
6
ACCEPTED MANUSCRIPT 2.4 Singlet Oxygen Generation in Cell Free Systems: To a solution of SOSG (2.5 µM) in degassed PBS (4 mL) was added the OxyMB-RB conjugate (5µM RB, 2.5x107 MB/ml). The solution was exposed to ultrasound for 30 minutes (frequency = 1MHz, 3.0W/cm2, 50% duty
IP
T
cycle using a pulse repetition frequency of 100 Hz) while being kept in the dark. The fluorescence intensity of SOSG (EX =505nm) was recorded at 525 nm before and after
SC R
ultrasound exposure. This protocol was repeated for the SF6MB-RB conjugate and Biotin-RB under otherwise identical conditions. The protocol was also repeated for RB alone in non-
NU
degassed PBS solution.
2.5 In vitro cytotoxicity experiments: Human primary pancreatic adenocarcinoma (BxPC3)
MA
cells were maintained in RPMI 1640 medium supplemented with 10% (v/v) foetal bovine serum in a humidified 5% CO2 atmosphere at 37oC. These cells were plated into the wells of
D
a 96-well tissue culture plate at a concentration of 5x103 cells per well and incubated for 21
TE
hours at 37oC in a humidified 5% CO2 atmosphere before being transferred to a hypoxic chamber at 37oC (O2/CO2/N2, 0.1 : 5 : 94.9 v/v/v) for 3 hours. The medium was then
CE P
removed from each well and replaced with 50 µl each of the OxyMB-RB conjugate in PBS:DMSO solution (99:1 v:v). Individual wells were then placed in direct contact with the
AC
emitting surface of the ultrasound transducer of a Sonidel SP100 sonoporator and an ultrasound gel was employed to mediate contact. Each well was treated with ultrasound for 30 secs, using a frequency of 1MHz, an ultrasound power density of 3.0Wcm -2 and a duty cycle of 50% (pulse frequency = 100Hz). These US dosimetry conditions for the in vitro and in vivo were chosen as they had previously been demonstrated by us to yield sonodynamic effects with RB conjugates. Cells were retained in the hypoxic environment for a further 3 hours before each well was emptied, washed with PBS and re-filled with fresh medium (200uL per well). Plates were then transferred back to in a humidified 5% CO2 atmosphere at 37oC for 21 hours. Cell viability was then determined using an MTT assay [20]. A similar procedure was repeated for the SF6MB-RB conjugate, the OxyMBs alone and the SF6MBs alone at the same concentrations as used for the OxyMB-RB conjugate. A control population
7
ACCEPTED MANUSCRIPT of cells was also established for ultrasound treatment alone. In the in vitro and in vivo experiments error was expressed as ± SEM while statistical comparisons were made using a
T
student’s t-test.
IP
2.6 In vivo experiments: All animals employed in this study were treated humanely and in
SC R
accordance with the licenced procedures under the UK Animals (Scientific Procedures) Act 1986. BxPc-3 cells were maintained in RPMI-160 medium supplemented with 10% foetal calf serum. Cells were cultured at 37oC under 5% CO2 in air. BxPc-3 cells( 1 x10 7) were re-
NU
suspended in 100µl of matrigel and implanted into the rear dorsal of male Balb/c SCID (C.B17/IcrHan®Hsd-Prkdcscid) mice. Tumour formation occurred approximately 2 weeks after
MA
implantation and tumour measurements were taken every other day using calipers. Once the tumours had reached an average volume of 256 mm3calculated from the geometric mean
Following induction of anaesthesia (intraperitoneal injection of
TE
into 4 groups.
D
diameter using the equation tumour volume = 4πR3/3, animals were randomly distributed
Hypnorm/Hypnovel), a 60 µL aliquot of PBS containing 1.5 x 107 microbubbles and 91.6 µM In cases where animals were treated with
CE P
RB was injected directly into each tumour.
ultrasound, within 5 min of injection, ultrasound gel was administered to the tumour to
AC
ensure contact with the ultrasound transducer and animals were treated with ultrasound using a Sonidel SP100 sonoporator for 3.5 min. at an ultrasound frequency of 1 MHz, an ultrasound power density of 3.0 Wcm-2 (ISATP; spatial average temporal peak)
and using a
duty cycle of 30% at a pulse repetition frequency of 100 Hz. After treatment animals were allowed to recover from anaesthesia and tumour volume was monitored at the indicated times. The % increase in tumour volume was calculated employing the pre-treatment measurements for each group.
8
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Scheme 1 Schematic representation of the OxyMB-RB conjugates. S = sensitiser = Rose
AC
CE P
TE
D
MA
Bengal.
9
ACCEPTED MANUSCRIPT 3.0 Results and Discussion. Phospholipid coated sulphur hexafluoride MBs (SF6MBs) were prepared following the method developed by Zhao et al. [21]. DSPE-PEG(2000)-biotin was incorporated within the shell of these MBs to introduce biotin functionality. The average
IP
T
size of the MBs was 1 - 2 μm with a concentration of approximately 1 x109 MB/ml as determined by analysis of optical microscopy images (Figure 1). Avidin was then mixed with
SC R
the SF6MBs for 10 minutes and the resulting avidin coated SF6MBs isolated by centrifugation. Surface attachment of avidin was confirmed by adding fluorescently labelled
NU
avidin (FITC-avidin) to the SF6MBs and examining the isolated MBs by fluorescence microscopy (Figure 2). The results of this derivatisation are shown in Figure 2 and reveal the
MA
presence of fluorescent MBs against an otherwise dark background. Oxygen filled MBs (OxyMBs) were obtained by gas exchange through sparging the suspension of SF6 filled microbubbles with oxygen (see Section 2.3). In parallel, biotin functionalised Rose Bengal
TE
D
was prepared by reaction of Rose Bengal amine with NHS-biotin. The avidin coated O2 and SF6 filled MBs, surface functionalised with avidin, were mixed with RB-biotin to produce the
AC
CE P
final OxyMB-RB and SF6MB-RB conjugates. (Scheme 1)
10
ACCEPTED MANUSCRIPT
(a)
NU
SC R
IP
T
(b)
Figure 1 Optical microscope images taken with a 40 x objective lens of biotin microbubbles
MA
after dilution (1:10) in PBS. Scale bar is 10 μm; (b) size distribution of biotin microbubbles after centrifugation obtained from analysis of 30 optical microscope images (the unfilled boxes at the left hand side of the graph represent microbubbles that were detected by the
AC
CE P
TE
D
image analysis software but smaller than 450 nm, the optical resolution of the system).
11
SC R
IP
T
ACCEPTED MANUSCRIPT
Figure 2 Fluorescence microscope image (x 40 objective) at 100 ms exposure time using a FITC filter set of SF6MBs
after incubation with FITC-avidin and centrifugation. The
AC
CE P
TE
D
MA
microbubble surface. (Scale bar = 10 μm)
NU
fluorescence of the MBsagainst the background indicates the binding of avidin to the
12
ACCEPTED MANUSCRIPT To determine the ability of the MB conjugates to generate singlet oxygen upon exposure to ultrasound, solutions of each conjugate were prepared containing 1x107 MBs with 5 µM RB attached in degassed PBS buffer (pH 7.4 ± 0.1) and Singlet Oxygen Sensor Green
T
The container was then sealed and protected from light during the
IP
(SOSG) (2.5 µM).
exposure time. The % dissolved oxygen in the PBS solution was measured before (96.3 %)
SC R
and after (22.6%) degassing using a dissolved oxygen meter. These solutions were exposed to ultrasound for 30 min at a frequency of 1 MHz, a power density of 3.0W/cm2 and using a
NU
duty cycle of 50% at a pulse repetition frequency of 100 Hz. SOSG is a specific fluorescent probe for singlet oxygen with low fluorescence in the reduced form and high fluorescence
MA
upon interaction with singlet oxygen [22]. A plot of percentage increase in fluorescent intensity of SOSG before and after ultrasound treatment for RB alone (5 µM), SF6MB-RB and OxyMB-RB is shown in Figure 3. The data reveal a significant increase in SOSG
TE
D
fluorescence for the OxyMB-RB conjugate (276%) when compared to either RB alone (5%) or SF6MB-RB (11%). These results suggest oxygen- loaded MBs effectively deliver oxygen
CE P
to the PBS solution to fuel singlet oxygen generation in the presence of ultrasound. Indeed, when RB alone was treated with ultrasound in normal PBS (i.e. non-degassed) the OxyMBRB in the hypoxic system was still more than twice as effective at producing singlet oxygen.
AC
This supports the hypothesis that the MBs were delivering O2 during the sonication process, thereby enhancing the singlet oxygen quantum yield.
13
Figure 3
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Plot of % increase in SOSG fluorescence for RB, SF6MB-RB and OxyMB-RB
TE
D
conjugate in de-gassed PBS upon exposure to ultrasound. The final column shows the % increase of SOSG fluorescence for RB in non-degassed solution. Error bars represent + the
AC
CE P
standard error where n = 2.
14
ACCEPTED MANUSCRIPT Having established the ability of the OxyMB-RB conjugate to generate singlet oxygen in a simulated hypoxic cell free environment, the next step was to determine the cytotoxic effect this would produce using pancreatic cancer cells (BxPc-3) as a target. BxPc-3 cells
IP
T
were cultured in 96 well plates and incubated in a hypoxic chamber at 37oC (O2/CO2/N2, 0.1 : 5 : 94.9, v/v/v) for three hours. Then, RB alone, SF6MB-RB and OxyMB-RB were added to
SC R
selected wells at concentrations of 1 and 5 µM (with respect to RB concentration). Control wells containing just cells and cells treated with either SF6MBs or Oxy MBs alone were used
NU
for comparative purposes. Wells were then exposed to ultrasound treatment for 30 seconds using a Sonidel SP100 sonoporator (frequency of 1MHz, an ultrasound power density of
MA
3.0Wcm-2 and a duty cycle of 50%, pulse frequency = 100Hz). Following irradiation, cells were incubated for a further 3 hours in the hypoxic chamber. The treatment solution was then removed, cells were washed with PBS and fresh media subsequently added. The cells
TE
D
were then incubated in normoxic conditions in a humidified 5% CO2 atmosphere at 37oC for a further 21 hours before cell viability was determined using a MTT assay [19]. The results
CE P
are shown in Figure 4 and they indicate a significant dose-dependent increase in ultrasoundmediated cytotoxicity for the OxyMB-RB conjugate relative to either the SF6MB-RB or RB alone at both the 1 and 5 µM concentrations. Indeed, at a concentration of 5 µM, cell viability
AC
for cells treated with the OxyMB-RB conjugate reduced by 17% following ultrasound irradiation, while cells treated with SF6MB-RB conjugates at the same concentration reduced by only 4%. This increase in cytotoxicity can be directly attributed to the presence of oxygen delivered by the OxyMBs enhancing SDT efficiency. Indeed, the difference in cytotoxicity for cells treated with the OxyMB-RB conjugate and ultrasound was signinficantly greater than those treated with the OxyMB-RB in the absence of ultrasound (p < 0.05), and those treated with either the SF6MB-RB conjugate or RB with ultrasound (p < 0.01 and p < 0.05 respectively). The observed toxicity for the OxyMB-RB conjugate at 5 µM in the absence of ultrasound is most likely due to inadvertent activation of the sensitiser by ambient light, since RB can also serve as a photosensitiser. Indeed, the control for the OxyMBs alone showed no noticeable increased toxicity while incubation of target cells with RB alone did exhibit 15
ACCEPTED MANUSCRIPT limited toxicity with the magnitude of this increase appearing to be dose dependant with
AC
CE P
TE
D
MA
NU
SC R
IP
T
respect to RB concentration.
16
ACCEPTED MANUSCRIPT (a)
(b)
δ
Figure 4
SC R
IP
T
**
Plot of % cell viability for BxPc-3 cells grown in hypoxic conditions treated with
NU
(a) 1 µM or (b) 5 µM RB, OxyMB-RB or SF6MB-RB conjugate. Controls for untreated cells and those treated with just OxyMB or SF6MB are also included. Error bars represent + the
MA
standard error where n = 3. p<0.05, **p<0.01 and δp<0.05 for OxyMB-RB compared to
AC
CE P
TE
D
OxyMB-RB – US, SF6MB-RB + US and RB + US respectively.
17
ACCEPTED MANUSCRIPT To determine the therapeutic efficiency of the MB–RB conjugate in vivo, tumours were induced in BALB/c SCID mice using the BxPc-3 cell line. BxPc-3 cells (1 x10 7) were resuspended in 100µl of matrigel and implanted into the rear dorsal of SCID mice. Tumour
IP
T
formation occurred approximately two weeks after cell implantation and the oxygen partial pressure of the tumours was measured once the volume reached 200 mm3 and found to be
SC R
2 mmHg < 1% indicating a significant degree of hypoxia. Once the tumours had reached 256 mm3, the mice were separated into four groups. Group 1 received the OxyMB-RB conjugate
NU
(60µL containing 1.5 x 107 MB; 91.6 µM RB, administered via intratumoral injection) and ultrasound treatment (for 3.5 min. at a frequency of 1 MHz, a power density of 3.5 Wcm -2 a
MA
duty cycle of 30% using a pulse repetition frequency of 100 Hz; group 2 received the SF6MB-RB conjugate and ultrasound treatment as described above; group 3 received the OxyMB-RB conjugate only and group 4 received the SF6MB-RB conjugate only.
It should
TE
D
be noted that in these experiments ultrasound conditions employed have previously been shown to have no effect on tumour growth [23]. The results are shown in Figure 5 and reveal
CE P
a significant reduction in tumour growth for groups 1 (p < 0.05) and 2 (p < 0.01 and p < 0.001) relative to their respective controls (groups 3 & 4). Indeed, five days after treatment, tumours in group 1 reduced in volume by 45% from their original pre-treatment volume while
AC
those in group 2 had increased in volume by 35% over the same time period. This data suggest that the differential in tumour growth (80%) (p < 0.05) can be directly attributed to the presence the OxyMBs providing additional oxygen as a substrate for ROS production by the sonodynamic effect. By contrast, the volume of tumours in mice treated with the conjugates only (groups 3 & 4) increased to in excess of 180% over the same time period.
18
**
**
IP
***
**
SC R
*
**
*
***
NU
Figure 5
*
*
T
ACCEPTED MANUSCRIPT
Plot of % tumour growth against time for mice bearing BxPc-3 tumours and
MA
treated with (i) OxyMB-RB conjugate + ultrasound (filled diamonds) (ii) OxyMB-RB conjugate - ultrasound (open squares) (iii) SF6MB-RB conjugate + ultrasound (filled triangles) (iv)
D
SF6MB-RB conjugate - ultrasound (filled squares). Error bars represent + the standard error
TE
where n = 3. *p<0.05, **p<0.01 and ***p<0.001 for (i) compared to (ii) and (iii) compared to
AC
CE P
(iv). p<0.5 for (i) compared to (iii).
19
ACCEPTED MANUSCRIPT 4.0 Conclusions: In conclusion, we have investigated the potential of using oxygen loaded MBs in combination with a sensitiser drug to enhance the sonodynamic effect in hypoxic environments. This approach was successful in generating significantly more singlet oxygen
IP
T
in a cell free system than similar conjugates loaded with SF6. Moreover, a greater cytotoxic effect was observed when BxPc-3 cells, cultured in an hypoxic environment, were treated
SC R
with the OxyMB-RB conjugate and ultrasound compared to cells cultured under similar conditions and treated with the SF6MB-RB conjugate and ultrasound. Finally, mice bearing
NU
human xenograft pancreatic BxPc-3 tumours treated with the OxyMB-RB conjugate and ultrasound showed a 45% reduction in tumour volume five days after treatment while the
MA
volume of tumours in mice treated with the conjugate only increased by 180% over the same time period. Overall these results demonstrate that SDT may be significantly enhanced by incorporating an approach that involves ultrasound- and microbubble-mediated delivery of
TE
D
oxygen to hypoxic tumours. Our approach could provide the basis for less invasive treatment of more deeply-seated lesions, together with addressing hypoxia which limits sensitiser-
CE P
based approaches and is associated with less favourable prognoses. It also provides the basis for a novel treatment option for pancreatic cancer which is perhaps one of the most
AC
recalcitrant forms of cancer and where existing therapeutic options are extremely limited. Acknowledgements: JC thanks Norbrook Laboratories Ltd for an endowed chair. Thanks are also expressed to Dr J Worthington for a helpful discussion.
20
ACCEPTED MANUSCRIPT References [1] Vaupel P, Kallinowski F, Okunief F.
Blood Flow, oxygen and nutrient supply, and
T
metabolic microenvironment of human tumors: A Review. Cancer Res 1989;49: 6449-
IP
65.
SC R
[2] Moulder JE, Rockwell S. Hypoxic fractions of solid tumors: experimental techniques, methods of analysis, and a survey of existing data. Int J Radiat Oncol Biol Phys
[3] Cancer
Research
UK.
Pancreatic
NU
1984;10,695-712.
cancer:
UK
incidence
statistics.
2011.
MA
http://info.cancerresearchuk.org/cancerstats/types/pancreas/incidence. [4] Brown JM, Giaccia AJ. The unique physiology of solid tumors: opportunities (and
TE
D
problems) for cancer therapy. Cancer Res 1998;58,1408-16. [5] Kim MP, Gallick GE. Gemcitabine resistance in pancreatic cancer: picking the key
CE P
players. Clin. Cancer. Res 2008;14,1284-5. [6] See Cancer Research UK website: http://www.cancerresearchuk.org/cancer-help/about-
AC
cancer/treatment/other/photodynamic-therapy-pdt. [7] Triesscheijn M, Baas P, Schellens JH, Stewart FA. Photodynamic therapy in oncology. Oncologist 2006;9,1034-44. [8] Starkey JR, Rebane AK, Drobizhev MA, Meng, Gong A, Elliot A, et al. New two-photon activated photodynamic therapy sensitizers induce xenograft tumor regressions after near-IR laser treatment through the body of the host mouse. Clin Cancer Res 2008;14,6564–73. [9] Busch TM, Hahn SM, Evans SM, Koch CJ. Depletion of tumor oxygenation during photodynamic therapy: detection by the hypoxia marker EF3 [2-(2-nitroimidazol-1[H]-yl)N-(3,3,3-trifluoropropyl)acetamide ]. Cancer Res 2000;60,2636–42, 21
ACCEPTED MANUSCRIPT [10] McCaughan B, Rouanet C, Fowley C, Nomikou N, McHale AP, McCarron PA, et al. Enhanced ROS production and cell death through combined photo- and sono-activation
T
of conventional photosensitising drugs. Bioorg Med Chem Lett 2011;21,5750-2
IP
[11] Yumita N, Okudaira K, Momose Y, Umemura S. Sonodynamically induced apoptosis
SC R
and active oxygen generation by gallium-porphyrin complex, ATX-70. Cancer Chemother Pharmacol 2010;66,1071-8.
[12] Nomikou N, Fowley C, Byrne NM, McCaughan B, McHale AP, Callan JF. Microbubble-
NU
sonosensitiser conjugates as therapeutics in sonodynamic therapy. Chem. Commun
MA
2012;48,8332-4.
[13] Shashank S, Borden M. Microbubble compositions, properties and biomedical
D
applications. Bubble Sci Eng Technol 2009;1,3-17.
CE P
2006;79,188–94.
TE
[14] Stewart VR, Sidhu PS. New directions in ultrasound: microbubble contrast. Br J Radiol
[15] Hernot S, Klibanov AL. Microbubbles in ultrasound-triggered drug and gene delivery.
AC
Adv Drug Delivery Rev 2008;60,1153–66. [16] Umemura S, Yumita N, Nishigaki R, Umemura K. Mechanism of cell damage by ultrasound in combination with haematoporphyrin. Jpn J Cancer Res 1990;81,962–6. [17] Misik V, Riesz P. Free radical intermediates in sonodynamic therapy. Ann N Y Acad Sci 2000;899,335–48. [18] Fowley C, Nomikou N, McHale AP, McCarron PA, McCaughan B, Callan JF. Water soluble quantum dots as hydrophilic carriers and two-photon excited energy donors in photodynamic therapy. J Mater Chem 2012;22,6456–62.
22
ACCEPTED MANUSCRIPT [19] Sennoga CA, Mahue V, Loughran J, Casey J, Seddon JM, Tang MX, Eckersley RJ. On sizing and counting of microbubbles using optical microscopy, Ultrasound in Medicine
T
and Biology 2010; 36, 2093-96
IP
[20] McHale AP, McHale L. Use of a tetrazolium based colorimetric assay in assessing
SC R
photoradiation therapy in vitro. Cancer Lett 1988;41,315–21
[21] Zhao S, Borden M, Bloch SH, Kruse D, Ferrara KW, Dayton PA. Radiation-force
NU
assisted targeting facilitates ultrasonic molecular imaging. Mol Imaging 2004;3,135-48 [22] See:http://www.lifetechnologies.com/uk/en/home/references/molecular-probes-the
MA
handbook/probes-for-reactive-oxygen-species-including-nitric-oxide/generating-anddetecting-reactive-oxygen-species.html.
Nomikou N, Arthur C, Sterrett C, McCaughan B, Callan JF, McHale AP. The effects
D
[23]
TE
of ultrasound and light on indocyanine-green-treated tumour cells and tissues. Chem
AC
CE P
Med Chem 2012;7,1465-71
23
ACCEPTED MANUSCRIPT Oxygen Carrying Microbubbles for Enhanced Sonodynamic Therapy of Hypoxic Tumours
SC R
O2
IP
T
Conor McEwan1, Joshua Owen2, Eleanor Stride2, Colin Fowley1, Heather Nesbitt1, David Cochrane1, Constantin. C. Coussios2, M. Borden3, Nikolitsa Nomikou4, Anthony P. McHale1 and John F. Callan1.
MB
NU
OxyMB-RB
MA
Oyxgen Loaded Microbubbles with Rose Bengal attached to the surface (OxyMB-RB) produced more singlet oxygen and were more cytotoxic to a hypoxic murine model upon ultrasound (US) irradiation than similar conjugates filled with SF6 gas.
AC
CE P
TE
D
Graphical abstract
24
S0168-3659(15)00094-2 doi: 10.1016/j.jconrel.2015.02.004 COREL 7553
To appear in:
Journal of Controlled Release
Received date: Revised date: Accepted date:
9 October 2014 20 December 2014 3 February 2015
Please cite this article as: Conor McEwan, Joshua Owen, Eleanor Stride, Colin Fowley, Heather Nesbitt, David Cochrane, Constantin.C. Coussios, M. Borden, Nikolitsa Nomikou, Anthony P. McHale, John F. Callan, Oxygen Carrying Microbubbles for Enhanced Sonodynamic Therapy of Hypoxic Tumours, Journal of Controlled Release (2015), doi: 10.1016/j.jconrel.2015.02.004
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Oxygen Carrying Microbubbles for Enhanced Sonodynamic Therapy of Hypoxic Tumours
SC R
IP
T
Conor McEwan1, Joshua Owen2, Eleanor Stride2, Colin Fowley1, Heather Nesbitt1, David Cochrane1, Constantin. C. Coussios2, M. Borden3, Nikolitsa Nomikou4, Anthony P. McHale1 and John F. Callan1*.
MA
NU
1. Biomedical Sciences Research Institute, University of Ulster, Coleraine, Northern Ireland, U.K. BT52 1SA. 2. Oxford Institute of Biomedical Engineering, University of Oxford, UK, OX3 7DQ. 3. Department of Mechanical Engineering, University of Colorado, 1111 Engineering Drive, Boulder, CO 80309, USA. 4. Division of Surgery & Interventional Science, Medical School, University College London, London W1W 7EJ, UK
TE
D
* Corresponding author: Ph: 00442870123059; Email: [email protected]
Abstract: Tumour hypoxia represents a major challenge in the effective treatment of solid
CE P
cancerous tumours using conventional approaches. As oxygen is a key substrate for Photo/ Sono-dynamic Therapy (PDT / SDT), hypoxia is also problematic for the treatment of solid tumours using these techniques. The ability to deliver oxygen to the vicinity of the tumour
AC
increases its local partial pressure improving the possibility of ROS generation in PDT / SDT. In this manuscript, we investigate the use of oxygen-loaded, lipid-stabilised microbubbles (MBs), decorated with a Rose Bengal sensitiser, for SDT-based treatment of a pancreatic cancer model (BxPc-3) in vitro and in vivo. We directly compare the effectiveness of the oxygen-loaded MBs with sulphur hexafluoride (SF6)-loaded MBs and reveal a significant improvement in therapeutic efficacy. The combination of oxygen-carrying, ultrasoundresponsive MBs, with an ultrasound-responsive therapeutic sensitiser, offers the possibility of delivering and activating the MB-sensitiser conjugate at the tumour site in a non-invasive manner, providing enhanced sonodynamic activation at that site. Key words: hypoxia, microbubbles, SDT, Rose Bengal, cancer, pancreatic. 1
ACCEPTED MANUSCRIPT 1.0 Introduction: Hypoxia within solid tumours is a key determinant in the effectiveness of radiation- and chemotherapy-based treatments and arises as a consequence of the atypical vasculature that is characteristic of growing tumours [1]. Essentially, this atypical vasculature
IP
T
results in the development of areas within the tumour where oxygen demand outstrips oxygen supply. Hypoxic fractions ranging from 10 to 30 % are present in most solid tumours
SC R
regardless of size [2]. Because hypoxia negatively influences the effectiveness of conventional therapies, it is usually associated with poor prognosis. Pancreatic ductal
NU
adenocarcinoma (PDA) is an extremely aggressive malignancy with fewer than 20% of those diagnosed being eligible for curative treatment [3]. PDA possesses a pronounced hypoxic
MA
tumour microenvironment, with high levels of expression of the hypoxia biomarker HIF-1 (hypoxia inducible transcription factor) serving as a predictor of poor clinical outcome [4]. In
D
addition, while PDA shows initial sensitivity to the benchmark chemotherapeutic drugs 5-
TE
fluorouracil and gemcitabine, it is usually followed by rapid development of resistance [5]. Therefore, novel therapies that may improve the therapeutic outcome in patients with PDA or
CE P
conditions with similar morbidities are of great interest. Photodynamic therapy (PDT) has emerged as an alternative to conventional chemo-
AC
and radiation-based therapies in the treatment of certain cancers such as skin, head and neck, oesophageal and prostate cancer [6]. PDT involves the simultaneous use of a sensitising drug, light of an appropriate wavelength and molecular oxygen to generate singlet oxygen and other Reactive Oxygen Species (ROS) that result in cytotoxic effects. The main attraction of PDT as a clinical treatment is that production of the cytotoxic species can be localised to the desired area through control of the light source reducing collateral damage to surrounding healthy tissue [7]. However, a major factor prohibiting the more widespread use of PDT is that the light used to excite currently approved sensitisers is in the visible region of the electromagnetic spectrum limiting penetration through mammalian tissue [8]. This restricts use of the approach to the treatment of superficial tumours and limits its effective clinical use in treating more deeply seated lesions.
2
In addition to the above
ACCEPTED MANUSCRIPT challenges and since PDT is fuelled by the presence of oxygen, hypoxia can limit the effectiveness of PDT. Indeed it has been demonstrated that use of PDT can exacerbate hypoxia as a result of photodynamically-driven conversion of oxygen to singlet oxygen [9].
IP
T
Since hypoxia is associated with poor prognosis, this may not be desirable, particularly in the
SC R
treatment of larger lesions.
More recently, our group and others, have demonstrated that several sensitisers activated by light can also be activated by ultrasound, an approach known as Sonodynamic
NU
Therapy (SDT) [10,11]. SDT offers significant advantages over PDT because ultrasound is widely accepted as a cost effective and safe clinical imaging modality and, unlike light, can
MA
be tightly focused with penetration in soft tissue up to several tens of centimetres depending on the frequency used. We have previously shown by attaching the sensitiser Rose Bengal
D
(RB) to the surface of a lipid stabilised microbubble (MB), that the resulting conjugate
TE
produced more singlet oxygen and was more effective in vitro and in vivo than the sensitiser alone at the same concentration [12]. MBs are gas filled, lipid or polymer stabilised bubbles
CE P
of micrometre dimension and are approved as contrast agents in diagnostic ultrasound [13]. At low acoustic pressures, MBs undergo repeated volumetric oscillations, a process known
AC
as stable cavitation [14]. However, at higher acoustic pressures, MBs undergo violent expansion and compression typically resulting in destruction of the MB, a process known as inertial cavitation [15]. The process of inertial cavitation can be accompanied by the emission of light (sonoluminescence) and / or a significant localised increase in temperature [16,17]. We have reasoned that placing sensitisers in close proximity to MBs undergoing inertial cavitation could enhance their efficiency at generating ROS through sono-luminescence or pyrolysis- mediated processes. Moreover, attachment of the sensitiser to the MB surface enables the potential of ultrasound-mediated delivery and activation of the sensitiser at the tumour site in a minimally invasive manner.
3
ACCEPTED MANUSCRIPT As an extension of the above approach, in this manuscript we investigate the ability of an oxygen-carrying MB-sensitiser conjugate to improve the sonodynamic effect under hypoxic conditions. To this end we assess the ability of the conjugate to deliver enhanced
IP
T
ultrasound-mediated production of ROS in hypoxic cell free systems and examine its ability to provide enhanced ultrasound-mediated cytotoxic effects on pancreatic cancer cells
SC R
cultured in an hypoxic environment. Finally, we assess ultrasound-mediated therapeutic effects using the conjugate as a sensitiser and human pancreatic xenograft BxPc-3 tumours
NU
in mice as a target.
MA
2.0 Materials & Methods
2.1 Reagents and Equipment: Rose Bengal sodium salt, 2-bromoethylamine, NHS-biotin, MTT, avidin, FITC avidin and ethanol were purchased from Sigma Aldrich (UK) at the
D
highest grade possible. Singlet oxygen Sensor Green (SOSG) was purchased from
TE
Invitrogen (UK). 1,2-Distearoyl-sn-glycero-3-phosphocholine (DSPC), 1,2-distearoyl-sn-
CE P
glycero-3-phosphoethanolamine-N-(polyethylene glycol)-2000 DSPE-PEG(2000), and 1,2distearoyl-sn-glycero-3-phosphoethanolamine-N-[biotinyl(polyethylene glycol)-2000] DSPEPEG(2000)-biotin were purchased from Avanti Polar Lipids (Alabaster, Alabama, USA).
AC
Sulphur hexafluoride and Oxygen were purchased from BOC,Industrial Gases, UK. Phosphate Buffered Saline was purchased from Gibco, Life Technologies, UK.
DisteNMR
spectra were recorded on a Varian 500 MHz spectrometer. ESI-MS characterisation of was achieved using a LCQTM quadrupole ion-trap mass spectrometer (Finnigan MAT, San Jose, California, USA) utilising electrospray ionisation (ESI). Fluorescence measurements were recorded on a Varian Cary Eclipse Spectrometer using 10 mm quartz cuvettes. Optical microscope images were taken with an optical microscope (Leica DM500 optical microscope) and fluorescence images were taken with a Nikon Eclipse Ti fluorescence microscope, FITC filter set. Dissolved oxygen was measured using a Thermo Scientific Orion Star A216 bench top. Oxygen exchange was CDI Blood Parameter Monitoring System 500 (Terumo UK Ltd. Tamesis, Surrey, UK). 4
ACCEPTED MANUSCRIPT 2.2 Synthesis of Biotin-Rose Bengal: Rose Bengal amine was prepared following a literature procedure [18]. To a solution of Rose Bengal amine (74 mg, 74µmol) in 2 mL DMSO, NHS Biotin (25mg, 74 µmol) dissolved in PBS buffer 2 mL (pH 9) was added.
IP
T
Triethylamine (10 µl, 0.71µmol) was added to the mixture which was then stirred at room temperature for 12 hrs. The resulting solution was placed in a solution of hexane:chloroform
SC R
(200ml 8:2) and stirred for 24 hrs at room temperature . The solution filtered and the filtrate retained. The filtrate was washed three times with diethyl ether and dried in a vacuum oven
NU
to yield the product as a red powder (14 mg, 10%).1H NMR (500 MHz, DMSO-d6): 1.28 (m, CH2, 2H), 1.45 (m, CH2, CH, 3H ), 1.57 (m, CH, 1H ), 1.60 (t,CH2, 2H), 2.52 (q, CH, 1H),
MA
2.58 (m,CH, 1H), 2.79 (q, CH,1H), 2.97 (t, CH2, 2H), 3.10 (m, SCH,1H), 3.85 ( t, OCH2, 2H), 4.09 (t,CH,1H), 4.28 (t, CH,1H), 6.34 (s, NH,1H), 6.41 (s, NH,1H), 7.46 (s, ArH, 2H), 7.56 (brs, NH, 1H).
13
C NMR (125 MHz, DMSO-d6): 8.98, 9.03, 25.52, 28.57, 29.45, 31.72,
TE
D
35.52, 37.34, 46.07, 53.13, 55.86, 59.66, 61.48, 64.65, 76.50, 97.69, 110.56, 124.271, 129.37, 130.48, 132.34, 134.79, 135.50, 136.41, 139.79, 157.42, 163.13, 163.57, 172.23,
CE P
172.68, 228.31, 230.86 –ve ESMS: calculated for C32H23Cl4I4N3O7S = 1243.0 found =1241. I.R. vmax (cm-1) 3425, 2924, 2369, 1683, 1576, 1541, 1342. M.P. 188-191°C 2.3 Preparation of O2 and SF6 loaded Microbubbles: DSPC, DSPE-PEG (2000), DSPE-
AC
PEG(2000)-biotin dissolved in chloroform were mixed in a glass vial giving a ratio of 82:9:9 mole % using 3.5 mg, 1.35 mg and 1.45 mg of each lipid respectively. This was covered with parafilm pierced with holes and placed on a hot plate for over 12 hours to evaporate the chloroform. PBS (5 ml) was then added to the dried lipid film followed by heating above the lipid phase transition temperature (>70 °C) under constant magnetic stirring for 30 minutes. The suspension was then sonicated with a Microson ultrasonic cell disruptor, 100 Watts, 22.5 kHz at power setting 4 for 90 seconds, the headspace was filled with SF6 and the gas liquid interface was sonicated (power 19) for 20 seconds producing MBs. The MB suspension was cooled in an ice bath for approximately 10 minutes. To ensure that the MB surface was biotin functionalised the MBs were mixed with FITC avidin, centrifuged and analysed by fluorescence microscopy. The MBs were observed to fluorescence against the 5
ACCEPTED MANUSCRIPT background indicating that avidin had successfully bound to the surface as shown in Figure 2.
T
50 μl of Avidin dissolved in water (10 mg/ml) was then added to the cooled MB
IP
suspension and allowed to mix for 10 minutes. The suspension was then centrifuged (300
was divided into two freeze drying vials.
SC R
RCF, 10 minutes) and resulting microbubble “cake” was concentrated into 1 ml of PBS. This For the SF6MBs the vials were then crimped
(sealed with a metal cap). To create OxyMBs the headspace of the vial and the MB
NU
suspension was exposed to a positive pressure of oxygen for 2 minutes. The exchange efficiency was determined by measuring the increase in ppO2 when 1 mL of Oxy MBs was
MA
added to 3 mL of deionised water using a CDI Blood Parameter Monitoring System 500. An increase of 37 kPa in ppO2 was observed for the OxyMBs while no change was detected
D
when the same volume of SF6MBs or vehicle alone was added. This suggests the volume of
TE
OxyMBs added (1 mL) contains ~2 mol of O2, which is the same approximate amount as the total gas content in 1 mL of the OxyMBs determined from the concentration and size
CE P
distribution measurements in Figure 1. Once the gas exchange was complete, the vial was then crimped and the MBs were used directly for experiments within 3 days. Following
AC
preparation as described above, MB samples were imaged under conventional optical microscopy to determine their size distribution and concentration. 10 μl was removed from each suspension and diluted in 90 μl of PBS followed by examination on a haemocytometer (Bright-Line, Hausser Scientific, Horsham, PA, USA). Images were obtained with at 40 x objective lens with a Leica DM500 optical microscope. The bubble size distribution and concentration were then obtained using purpose written image analysis software in Matlab (2010B, The MathWorks, Natick, MA, USA)[19].
6
ACCEPTED MANUSCRIPT 2.4 Singlet Oxygen Generation in Cell Free Systems: To a solution of SOSG (2.5 µM) in degassed PBS (4 mL) was added the OxyMB-RB conjugate (5µM RB, 2.5x107 MB/ml). The solution was exposed to ultrasound for 30 minutes (frequency = 1MHz, 3.0W/cm2, 50% duty
IP
T
cycle using a pulse repetition frequency of 100 Hz) while being kept in the dark. The fluorescence intensity of SOSG (EX =505nm) was recorded at 525 nm before and after
SC R
ultrasound exposure. This protocol was repeated for the SF6MB-RB conjugate and Biotin-RB under otherwise identical conditions. The protocol was also repeated for RB alone in non-
NU
degassed PBS solution.
2.5 In vitro cytotoxicity experiments: Human primary pancreatic adenocarcinoma (BxPC3)
MA
cells were maintained in RPMI 1640 medium supplemented with 10% (v/v) foetal bovine serum in a humidified 5% CO2 atmosphere at 37oC. These cells were plated into the wells of
D
a 96-well tissue culture plate at a concentration of 5x103 cells per well and incubated for 21
TE
hours at 37oC in a humidified 5% CO2 atmosphere before being transferred to a hypoxic chamber at 37oC (O2/CO2/N2, 0.1 : 5 : 94.9 v/v/v) for 3 hours. The medium was then
CE P
removed from each well and replaced with 50 µl each of the OxyMB-RB conjugate in PBS:DMSO solution (99:1 v:v). Individual wells were then placed in direct contact with the
AC
emitting surface of the ultrasound transducer of a Sonidel SP100 sonoporator and an ultrasound gel was employed to mediate contact. Each well was treated with ultrasound for 30 secs, using a frequency of 1MHz, an ultrasound power density of 3.0Wcm -2 and a duty cycle of 50% (pulse frequency = 100Hz). These US dosimetry conditions for the in vitro and in vivo were chosen as they had previously been demonstrated by us to yield sonodynamic effects with RB conjugates. Cells were retained in the hypoxic environment for a further 3 hours before each well was emptied, washed with PBS and re-filled with fresh medium (200uL per well). Plates were then transferred back to in a humidified 5% CO2 atmosphere at 37oC for 21 hours. Cell viability was then determined using an MTT assay [20]. A similar procedure was repeated for the SF6MB-RB conjugate, the OxyMBs alone and the SF6MBs alone at the same concentrations as used for the OxyMB-RB conjugate. A control population
7
ACCEPTED MANUSCRIPT of cells was also established for ultrasound treatment alone. In the in vitro and in vivo experiments error was expressed as ± SEM while statistical comparisons were made using a
T
student’s t-test.
IP
2.6 In vivo experiments: All animals employed in this study were treated humanely and in
SC R
accordance with the licenced procedures under the UK Animals (Scientific Procedures) Act 1986. BxPc-3 cells were maintained in RPMI-160 medium supplemented with 10% foetal calf serum. Cells were cultured at 37oC under 5% CO2 in air. BxPc-3 cells( 1 x10 7) were re-
NU
suspended in 100µl of matrigel and implanted into the rear dorsal of male Balb/c SCID (C.B17/IcrHan®Hsd-Prkdcscid) mice. Tumour formation occurred approximately 2 weeks after
MA
implantation and tumour measurements were taken every other day using calipers. Once the tumours had reached an average volume of 256 mm3calculated from the geometric mean
Following induction of anaesthesia (intraperitoneal injection of
TE
into 4 groups.
D
diameter using the equation tumour volume = 4πR3/3, animals were randomly distributed
Hypnorm/Hypnovel), a 60 µL aliquot of PBS containing 1.5 x 107 microbubbles and 91.6 µM In cases where animals were treated with
CE P
RB was injected directly into each tumour.
ultrasound, within 5 min of injection, ultrasound gel was administered to the tumour to
AC
ensure contact with the ultrasound transducer and animals were treated with ultrasound using a Sonidel SP100 sonoporator for 3.5 min. at an ultrasound frequency of 1 MHz, an ultrasound power density of 3.0 Wcm-2 (ISATP; spatial average temporal peak)
and using a
duty cycle of 30% at a pulse repetition frequency of 100 Hz. After treatment animals were allowed to recover from anaesthesia and tumour volume was monitored at the indicated times. The % increase in tumour volume was calculated employing the pre-treatment measurements for each group.
8
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Scheme 1 Schematic representation of the OxyMB-RB conjugates. S = sensitiser = Rose
AC
CE P
TE
D
MA
Bengal.
9
ACCEPTED MANUSCRIPT 3.0 Results and Discussion. Phospholipid coated sulphur hexafluoride MBs (SF6MBs) were prepared following the method developed by Zhao et al. [21]. DSPE-PEG(2000)-biotin was incorporated within the shell of these MBs to introduce biotin functionality. The average
IP
T
size of the MBs was 1 - 2 μm with a concentration of approximately 1 x109 MB/ml as determined by analysis of optical microscopy images (Figure 1). Avidin was then mixed with
SC R
the SF6MBs for 10 minutes and the resulting avidin coated SF6MBs isolated by centrifugation. Surface attachment of avidin was confirmed by adding fluorescently labelled
NU
avidin (FITC-avidin) to the SF6MBs and examining the isolated MBs by fluorescence microscopy (Figure 2). The results of this derivatisation are shown in Figure 2 and reveal the
MA
presence of fluorescent MBs against an otherwise dark background. Oxygen filled MBs (OxyMBs) were obtained by gas exchange through sparging the suspension of SF6 filled microbubbles with oxygen (see Section 2.3). In parallel, biotin functionalised Rose Bengal
TE
D
was prepared by reaction of Rose Bengal amine with NHS-biotin. The avidin coated O2 and SF6 filled MBs, surface functionalised with avidin, were mixed with RB-biotin to produce the
AC
CE P
final OxyMB-RB and SF6MB-RB conjugates. (Scheme 1)
10
ACCEPTED MANUSCRIPT
(a)
NU
SC R
IP
T
(b)
Figure 1 Optical microscope images taken with a 40 x objective lens of biotin microbubbles
MA
after dilution (1:10) in PBS. Scale bar is 10 μm; (b) size distribution of biotin microbubbles after centrifugation obtained from analysis of 30 optical microscope images (the unfilled boxes at the left hand side of the graph represent microbubbles that were detected by the
AC
CE P
TE
D
image analysis software but smaller than 450 nm, the optical resolution of the system).
11
SC R
IP
T
ACCEPTED MANUSCRIPT
Figure 2 Fluorescence microscope image (x 40 objective) at 100 ms exposure time using a FITC filter set of SF6MBs
after incubation with FITC-avidin and centrifugation. The
AC
CE P
TE
D
MA
microbubble surface. (Scale bar = 10 μm)
NU
fluorescence of the MBsagainst the background indicates the binding of avidin to the
12
ACCEPTED MANUSCRIPT To determine the ability of the MB conjugates to generate singlet oxygen upon exposure to ultrasound, solutions of each conjugate were prepared containing 1x107 MBs with 5 µM RB attached in degassed PBS buffer (pH 7.4 ± 0.1) and Singlet Oxygen Sensor Green
T
The container was then sealed and protected from light during the
IP
(SOSG) (2.5 µM).
exposure time. The % dissolved oxygen in the PBS solution was measured before (96.3 %)
SC R
and after (22.6%) degassing using a dissolved oxygen meter. These solutions were exposed to ultrasound for 30 min at a frequency of 1 MHz, a power density of 3.0W/cm2 and using a
NU
duty cycle of 50% at a pulse repetition frequency of 100 Hz. SOSG is a specific fluorescent probe for singlet oxygen with low fluorescence in the reduced form and high fluorescence
MA
upon interaction with singlet oxygen [22]. A plot of percentage increase in fluorescent intensity of SOSG before and after ultrasound treatment for RB alone (5 µM), SF6MB-RB and OxyMB-RB is shown in Figure 3. The data reveal a significant increase in SOSG
TE
D
fluorescence for the OxyMB-RB conjugate (276%) when compared to either RB alone (5%) or SF6MB-RB (11%). These results suggest oxygen- loaded MBs effectively deliver oxygen
CE P
to the PBS solution to fuel singlet oxygen generation in the presence of ultrasound. Indeed, when RB alone was treated with ultrasound in normal PBS (i.e. non-degassed) the OxyMBRB in the hypoxic system was still more than twice as effective at producing singlet oxygen.
AC
This supports the hypothesis that the MBs were delivering O2 during the sonication process, thereby enhancing the singlet oxygen quantum yield.
13
Figure 3
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Plot of % increase in SOSG fluorescence for RB, SF6MB-RB and OxyMB-RB
TE
D
conjugate in de-gassed PBS upon exposure to ultrasound. The final column shows the % increase of SOSG fluorescence for RB in non-degassed solution. Error bars represent + the
AC
CE P
standard error where n = 2.
14
ACCEPTED MANUSCRIPT Having established the ability of the OxyMB-RB conjugate to generate singlet oxygen in a simulated hypoxic cell free environment, the next step was to determine the cytotoxic effect this would produce using pancreatic cancer cells (BxPc-3) as a target. BxPc-3 cells
IP
T
were cultured in 96 well plates and incubated in a hypoxic chamber at 37oC (O2/CO2/N2, 0.1 : 5 : 94.9, v/v/v) for three hours. Then, RB alone, SF6MB-RB and OxyMB-RB were added to
SC R
selected wells at concentrations of 1 and 5 µM (with respect to RB concentration). Control wells containing just cells and cells treated with either SF6MBs or Oxy MBs alone were used
NU
for comparative purposes. Wells were then exposed to ultrasound treatment for 30 seconds using a Sonidel SP100 sonoporator (frequency of 1MHz, an ultrasound power density of
MA
3.0Wcm-2 and a duty cycle of 50%, pulse frequency = 100Hz). Following irradiation, cells were incubated for a further 3 hours in the hypoxic chamber. The treatment solution was then removed, cells were washed with PBS and fresh media subsequently added. The cells
TE
D
were then incubated in normoxic conditions in a humidified 5% CO2 atmosphere at 37oC for a further 21 hours before cell viability was determined using a MTT assay [19]. The results
CE P
are shown in Figure 4 and they indicate a significant dose-dependent increase in ultrasoundmediated cytotoxicity for the OxyMB-RB conjugate relative to either the SF6MB-RB or RB alone at both the 1 and 5 µM concentrations. Indeed, at a concentration of 5 µM, cell viability
AC
for cells treated with the OxyMB-RB conjugate reduced by 17% following ultrasound irradiation, while cells treated with SF6MB-RB conjugates at the same concentration reduced by only 4%. This increase in cytotoxicity can be directly attributed to the presence of oxygen delivered by the OxyMBs enhancing SDT efficiency. Indeed, the difference in cytotoxicity for cells treated with the OxyMB-RB conjugate and ultrasound was signinficantly greater than those treated with the OxyMB-RB in the absence of ultrasound (p < 0.05), and those treated with either the SF6MB-RB conjugate or RB with ultrasound (p < 0.01 and p < 0.05 respectively). The observed toxicity for the OxyMB-RB conjugate at 5 µM in the absence of ultrasound is most likely due to inadvertent activation of the sensitiser by ambient light, since RB can also serve as a photosensitiser. Indeed, the control for the OxyMBs alone showed no noticeable increased toxicity while incubation of target cells with RB alone did exhibit 15
ACCEPTED MANUSCRIPT limited toxicity with the magnitude of this increase appearing to be dose dependant with
AC
CE P
TE
D
MA
NU
SC R
IP
T
respect to RB concentration.
16
ACCEPTED MANUSCRIPT (a)
(b)
δ
Figure 4
SC R
IP
T
**
Plot of % cell viability for BxPc-3 cells grown in hypoxic conditions treated with
NU
(a) 1 µM or (b) 5 µM RB, OxyMB-RB or SF6MB-RB conjugate. Controls for untreated cells and those treated with just OxyMB or SF6MB are also included. Error bars represent + the
MA
standard error where n = 3. p<0.05, **p<0.01 and δp<0.05 for OxyMB-RB compared to
AC
CE P
TE
D
OxyMB-RB – US, SF6MB-RB + US and RB + US respectively.
17
ACCEPTED MANUSCRIPT To determine the therapeutic efficiency of the MB–RB conjugate in vivo, tumours were induced in BALB/c SCID mice using the BxPc-3 cell line. BxPc-3 cells (1 x10 7) were resuspended in 100µl of matrigel and implanted into the rear dorsal of SCID mice. Tumour
IP
T
formation occurred approximately two weeks after cell implantation and the oxygen partial pressure of the tumours was measured once the volume reached 200 mm3 and found to be
SC R
2 mmHg < 1% indicating a significant degree of hypoxia. Once the tumours had reached 256 mm3, the mice were separated into four groups. Group 1 received the OxyMB-RB conjugate
NU
(60µL containing 1.5 x 107 MB; 91.6 µM RB, administered via intratumoral injection) and ultrasound treatment (for 3.5 min. at a frequency of 1 MHz, a power density of 3.5 Wcm -2 a
MA
duty cycle of 30% using a pulse repetition frequency of 100 Hz; group 2 received the SF6MB-RB conjugate and ultrasound treatment as described above; group 3 received the OxyMB-RB conjugate only and group 4 received the SF6MB-RB conjugate only.
It should
TE
D
be noted that in these experiments ultrasound conditions employed have previously been shown to have no effect on tumour growth [23]. The results are shown in Figure 5 and reveal
CE P
a significant reduction in tumour growth for groups 1 (p < 0.05) and 2 (p < 0.01 and p < 0.001) relative to their respective controls (groups 3 & 4). Indeed, five days after treatment, tumours in group 1 reduced in volume by 45% from their original pre-treatment volume while
AC
those in group 2 had increased in volume by 35% over the same time period. This data suggest that the differential in tumour growth (80%) (p < 0.05) can be directly attributed to the presence the OxyMBs providing additional oxygen as a substrate for ROS production by the sonodynamic effect. By contrast, the volume of tumours in mice treated with the conjugates only (groups 3 & 4) increased to in excess of 180% over the same time period.
18
**
**
IP
***
**
SC R
*
**
*
***
NU
Figure 5
*
*
T
ACCEPTED MANUSCRIPT
Plot of % tumour growth against time for mice bearing BxPc-3 tumours and
MA
treated with (i) OxyMB-RB conjugate + ultrasound (filled diamonds) (ii) OxyMB-RB conjugate - ultrasound (open squares) (iii) SF6MB-RB conjugate + ultrasound (filled triangles) (iv)
D
SF6MB-RB conjugate - ultrasound (filled squares). Error bars represent + the standard error
TE
where n = 3. *p<0.05, **p<0.01 and ***p<0.001 for (i) compared to (ii) and (iii) compared to
AC
CE P
(iv). p<0.5 for (i) compared to (iii).
19
ACCEPTED MANUSCRIPT 4.0 Conclusions: In conclusion, we have investigated the potential of using oxygen loaded MBs in combination with a sensitiser drug to enhance the sonodynamic effect in hypoxic environments. This approach was successful in generating significantly more singlet oxygen
IP
T
in a cell free system than similar conjugates loaded with SF6. Moreover, a greater cytotoxic effect was observed when BxPc-3 cells, cultured in an hypoxic environment, were treated
SC R
with the OxyMB-RB conjugate and ultrasound compared to cells cultured under similar conditions and treated with the SF6MB-RB conjugate and ultrasound. Finally, mice bearing
NU
human xenograft pancreatic BxPc-3 tumours treated with the OxyMB-RB conjugate and ultrasound showed a 45% reduction in tumour volume five days after treatment while the
MA
volume of tumours in mice treated with the conjugate only increased by 180% over the same time period. Overall these results demonstrate that SDT may be significantly enhanced by incorporating an approach that involves ultrasound- and microbubble-mediated delivery of
TE
D
oxygen to hypoxic tumours. Our approach could provide the basis for less invasive treatment of more deeply-seated lesions, together with addressing hypoxia which limits sensitiser-
CE P
based approaches and is associated with less favourable prognoses. It also provides the basis for a novel treatment option for pancreatic cancer which is perhaps one of the most
AC
recalcitrant forms of cancer and where existing therapeutic options are extremely limited. Acknowledgements: JC thanks Norbrook Laboratories Ltd for an endowed chair. Thanks are also expressed to Dr J Worthington for a helpful discussion.
20
ACCEPTED MANUSCRIPT References [1] Vaupel P, Kallinowski F, Okunief F.
Blood Flow, oxygen and nutrient supply, and
T
metabolic microenvironment of human tumors: A Review. Cancer Res 1989;49: 6449-
IP
65.
SC R
[2] Moulder JE, Rockwell S. Hypoxic fractions of solid tumors: experimental techniques, methods of analysis, and a survey of existing data. Int J Radiat Oncol Biol Phys
[3] Cancer
Research
UK.
Pancreatic
NU
1984;10,695-712.
cancer:
UK
incidence
statistics.
2011.
MA
http://info.cancerresearchuk.org/cancerstats/types/pancreas/incidence. [4] Brown JM, Giaccia AJ. The unique physiology of solid tumors: opportunities (and
TE
D
problems) for cancer therapy. Cancer Res 1998;58,1408-16. [5] Kim MP, Gallick GE. Gemcitabine resistance in pancreatic cancer: picking the key
CE P
players. Clin. Cancer. Res 2008;14,1284-5. [6] See Cancer Research UK website: http://www.cancerresearchuk.org/cancer-help/about-
AC
cancer/treatment/other/photodynamic-therapy-pdt. [7] Triesscheijn M, Baas P, Schellens JH, Stewart FA. Photodynamic therapy in oncology. Oncologist 2006;9,1034-44. [8] Starkey JR, Rebane AK, Drobizhev MA, Meng, Gong A, Elliot A, et al. New two-photon activated photodynamic therapy sensitizers induce xenograft tumor regressions after near-IR laser treatment through the body of the host mouse. Clin Cancer Res 2008;14,6564–73. [9] Busch TM, Hahn SM, Evans SM, Koch CJ. Depletion of tumor oxygenation during photodynamic therapy: detection by the hypoxia marker EF3 [2-(2-nitroimidazol-1[H]-yl)N-(3,3,3-trifluoropropyl)acetamide ]. Cancer Res 2000;60,2636–42, 21
ACCEPTED MANUSCRIPT [10] McCaughan B, Rouanet C, Fowley C, Nomikou N, McHale AP, McCarron PA, et al. Enhanced ROS production and cell death through combined photo- and sono-activation
T
of conventional photosensitising drugs. Bioorg Med Chem Lett 2011;21,5750-2
IP
[11] Yumita N, Okudaira K, Momose Y, Umemura S. Sonodynamically induced apoptosis
SC R
and active oxygen generation by gallium-porphyrin complex, ATX-70. Cancer Chemother Pharmacol 2010;66,1071-8.
[12] Nomikou N, Fowley C, Byrne NM, McCaughan B, McHale AP, Callan JF. Microbubble-
NU
sonosensitiser conjugates as therapeutics in sonodynamic therapy. Chem. Commun
MA
2012;48,8332-4.
[13] Shashank S, Borden M. Microbubble compositions, properties and biomedical
D
applications. Bubble Sci Eng Technol 2009;1,3-17.
CE P
2006;79,188–94.
TE
[14] Stewart VR, Sidhu PS. New directions in ultrasound: microbubble contrast. Br J Radiol
[15] Hernot S, Klibanov AL. Microbubbles in ultrasound-triggered drug and gene delivery.
AC
Adv Drug Delivery Rev 2008;60,1153–66. [16] Umemura S, Yumita N, Nishigaki R, Umemura K. Mechanism of cell damage by ultrasound in combination with haematoporphyrin. Jpn J Cancer Res 1990;81,962–6. [17] Misik V, Riesz P. Free radical intermediates in sonodynamic therapy. Ann N Y Acad Sci 2000;899,335–48. [18] Fowley C, Nomikou N, McHale AP, McCarron PA, McCaughan B, Callan JF. Water soluble quantum dots as hydrophilic carriers and two-photon excited energy donors in photodynamic therapy. J Mater Chem 2012;22,6456–62.
22
ACCEPTED MANUSCRIPT [19] Sennoga CA, Mahue V, Loughran J, Casey J, Seddon JM, Tang MX, Eckersley RJ. On sizing and counting of microbubbles using optical microscopy, Ultrasound in Medicine
T
and Biology 2010; 36, 2093-96
IP
[20] McHale AP, McHale L. Use of a tetrazolium based colorimetric assay in assessing
SC R
photoradiation therapy in vitro. Cancer Lett 1988;41,315–21
[21] Zhao S, Borden M, Bloch SH, Kruse D, Ferrara KW, Dayton PA. Radiation-force
NU
assisted targeting facilitates ultrasonic molecular imaging. Mol Imaging 2004;3,135-48 [22] See:http://www.lifetechnologies.com/uk/en/home/references/molecular-probes-the
MA
handbook/probes-for-reactive-oxygen-species-including-nitric-oxide/generating-anddetecting-reactive-oxygen-species.html.
Nomikou N, Arthur C, Sterrett C, McCaughan B, Callan JF, McHale AP. The effects
D
[23]
TE
of ultrasound and light on indocyanine-green-treated tumour cells and tissues. Chem
AC
CE P
Med Chem 2012;7,1465-71
23
ACCEPTED MANUSCRIPT Oxygen Carrying Microbubbles for Enhanced Sonodynamic Therapy of Hypoxic Tumours
SC R
O2
IP
T
Conor McEwan1, Joshua Owen2, Eleanor Stride2, Colin Fowley1, Heather Nesbitt1, David Cochrane1, Constantin. C. Coussios2, M. Borden3, Nikolitsa Nomikou4, Anthony P. McHale1 and John F. Callan1.
MB
NU
OxyMB-RB
MA
Oyxgen Loaded Microbubbles with Rose Bengal attached to the surface (OxyMB-RB) produced more singlet oxygen and were more cytotoxic to a hypoxic murine model upon ultrasound (US) irradiation than similar conjugates filled with SF6 gas.
AC
CE P
TE
D
Graphical abstract
24