- Email: [email protected]
Assay of hydroxyl radicals generated by focused ultrasound
Ultrasonics Sonochemistry 16 (2009) 339–344
Contents lists available at ScienceDirect
Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultsonch
Assay of hydroxyl radicals generated by focused ultrasound L. Villeneuve a,*, L. Alberti c, J.-P. Steghens d, J.-M. Lancelin e, J.-L. Mestas b a
Hospices Civils de Lyon, Pôle d’Information Médicale, Evaluation et Recherche, Lyon F-69003, France Inserm, U556, Lyon, F-69003, France; Université de Lyon, Lyon, F-69003, France INSERM, U590, Centre Léon Bérard, Université Lyon 1, Lyon F-69008, France d Hospices Civils de Lyon, Laboratoire de Biochimie et de Biologie, Lyon F-69003, France e CNRS, UMR 5180, Université de Lyon, Lyon F-69003, France b c
a r t i c l e
i n f o
Article history: Received 29 July 2008 Received in revised form 15 September 2008 Accepted 16 September 2008 Available online 10 October 2008 PACS: 43.35 Keywords: Cavitation Ultrasound Sonolysis Terephthalic acid Hydroxyl radicals Fluorescence
a b s t r a c t Water sonolysis leads to the formation of hydroxyl radicals (OH). Various techniques are used to detect the OH production and thus to assess the level of ultrasound-mediated cavitation generated in vitro. In this study, we used terephthalic acid (TA) as an OH trap. This method is based on the fluorescent properties of hydroxyterephthalic acid (HTA) formed by the reaction of TA with OH and used as an indicator of the degree of inertial cavitation caused. The experimental system is comprised mainly of a focused piezoelectric ultrasound transmitter and a measurement cell containing 1X PBS/TA diluted solution. In the first part, we aimed to characterize the most appropriate experimental conditions (TA dosimeter solution, irradiation time) in order to optimize the resulting HTA fluorescence values. Then, we could determine that the HTA production increased with the level of the cavitation phenomenon caused by the acoustic power from which OH production may be estimated. Ó 2008 Elsevier B.V. All rights reserved.
1. Introduction The presence of ultrasound-induced inertial cavitation manifests itself by very intense local physicochemical effects: microstreaming, sonoluminescence, brutal increase in local temperature, formation radical-containing chemical species [1–4]. The energy levels achieved are such that they can lead to the homolytic cleavage of covalent bonds between the oxygen and hydrogen atoms comprising the water molecules present. Water sonolysis leads to the formation of reactive oxygen species (ROS) such as hydroxyl radicals (OH) and hydrogen peroxide (H2O2) [2] according to the following equations:
H2 O ! H þ OH H þ O2 ! HO2 HO2 þ HO2 ! H2 O2 þ O2 OH þ OH ! H2 O2 H þ H2 O ! H2 OH * Corresponding author. Tel.: +33 472 681 944; fax: +33 472 681 931. E-mail address: [email protected] (L. Villeneuve). 1350-4177/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ultsonch.2008.09.007
Currently, various techniques are used to detect the production of ROS and thus to assess the level of ultrasound-mediated cavitation generated in vitro within a reaction volume [5]. In addition to electron spin resonance recognized as the reference technique for assaying OH produced by ultrasound-mediated cavitation [3,6], a chemiluminescent method is known for its enhanced sensitivity and lower detection limits [7]. A number of other techniques are less sensitive, but easier to implement. Thus, as iodometry, based on the Weissler reaction [8], can be used to detect hydrogen peroxide (H2O2), or whereas the Fricke dosimeter is based on the oxidation of Fe2+ ions by OH, HO2 and H2O2 compounds present in the medium [2,8]. These techniques are used to detect compounds generated by reactions involving OH radicals. They are called indirect methods and tend to lose their sensitivity when the irradiated medium contains other organic compounds with which the radicals can interact. Beyond their very high chemical reactivity, OH radicals are characterised by their very short life span (circa 109 s). Their detection therefore requires that they be captured as soon as possible after their formation. In this respect, the method presented in this study, using terephthalic acid (TA) as a hydroxyl radical trap, is an interesting alternative. Initially developed by
340
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
a electrical power applied to the focused ultrasound source of between 40 mW and 200 W, corresponding to a focal point pressure of between 0.07 and 6.32 Mpa (Fig. 2). A composite ceramic dish with a resonant frequency of 1 MHz, mounted on a very high acoustic impedance medium (backing) acts as a transducer dedicated to reception of the acoustic signal produced by the cavitation phenomenon in the propagation medium. This transducer, with a diameter of 100 mm and a radius of curvature of 100 mm, is placed perpendicular to the sound axis of the ultrasound transmitter and positioned such that its centre of curvature is aligned with the centre of the measurement cell. The received signal is fed through a wide-band isolation transformer, amplified (NF Electronic InstrumentsÒ BX31) and then sent to a spectrum sweep-analyzer (Hewlett PackardÒ ESA L 1500 A). The examined frequency range is 0.1–7.1 MHz, the frequency resolution bandwidth 10 kHz and the duration of analysis 50 ms. Ten successive power spectra are averaged to reduce noise. A polyethylene pipette (CMLÒ; bulb dimensions: diameter = 9 mm; length = 20 mm), mechanically integral to the excitation source, is used as a measurement cell and contains the dosimeter solution (1.4 ml) to be exposed to the ultrasound. The posterior part of the measurement cell is sealed to isolate its contents from the outside medium. It is then directed along the sound axis and centred on the ultrasound transmitter’s centre of curvature (focal point). Ultrasound wave attenuation by the measurement cell wall (thickness: 0.3 mm) is less than 0.1 dB.
Armstrong et al. and used in radiation chemistry [9–10], this method is based on the fluorescent properties of hydroxyterephthalic acid (HTA), formed by the reaction of TA with OH. Indeed, contrary to TA, that has no fluorescent properties its hydroxylation by product (HTA) emits a fluorescent signal that can be observed at 424 nm when excited by a light source at a wavelength of 323 nm [8]. The present study consisted in assessing the production of OH radicals based on the assay of HTA, used as an indicator of the degree of inertial cavitation caused by a focused ultrasound system within a reaction volume. It was conducted in two steps. The first consisted in assessing the influence of TA concentration and irradiation time on the fluorescence of the treated medium; the second step focused on precisely characterising the degree of cavitation according to the acoustic power of the ultrasound source used. 2. Materials and methods 2.1. Experimental system The experimental system is comprised mainly of a focused ultrasound transmitter and a measurement cell containing the reaction medium (Fig. 1). Both elements are contained in a tank filled with degassed water (bath temperature: 20–25 °C; dissolved oxygen content: 3 ± 0.5 mg/L, measured by an oxymeter (Oxi 340 WTWÒ). The ultrasound transmitter is a focused piezoelectric device (Channel Industries Inc.Ò, C 5500) with a diameter of 100 mm, a radius of curvature of 100 mm and a resonant frequency of 473.7 kHz. The focal volume, defined by an acoustic pressure greater than 50% of maximum pressure, is 35 mm deep and 4 mm in diameter. The use of a sine wave generator (Hewlett-PackardÒ 33120 A; setting range: 22–22 dBm), followed by an attenuator (16 dB) and an amplifier (AdeceÒ, 200 W, 50 dB) generates
2.2. Dosimetry 2.2.1. Chemical solutions A reference solution containing 2 lM HTA in a 1X PBS (phosphate-buffered saline) solution was prepared fresh prior to each
Spectrum analyser
20 dB
IEEE Bus
Acoustic receiver Silicon Transducer Sine generator Amplifier
Focal volume
473,7 kHz
PC controller
Degassed water
Focal volume
Measurement cell
Fig. 1. Experimental system.
Focalized transducer
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
Power
10
Pressure
100
10
1
1
0.1 -15
-12
-9
-6
-3
0
3
6
9
12
15
Pressure (MPa)
Acoustic Power (W)
1000
0.1 18
Psp (dBm) Fig. 2. Influence of the power set point (Psp) on the acoustic power [11] and peak pressure measured on the focal point (hydrophone GEC Marconi Y-33-7611).
fluorescence measurement campaign, to check and calibrate the measurement instruments. The HTA was taken from a 20 mM stock solution prepared according to the method described by Mason et al. [3]. This stock solution was stored at +4 °C in the dark; its pH was systematically checked prior to use (pH = 7.3 ± 0.1; Radiometer CopenhagenÒ, PHM 210 standard pH meter). The reference solution excitation and emission spectra were measured by spectrofluorimetry. The maximum excitation and emission wavelengths were, respectively, kex = 323 nm and kem = 424 nm. A 20 mM TA stock solution (Acros OrganicsÒ) was prepared according to the method of Mason et al. [3] and stored at +4 °C in the dark for several weeks. Prior to use, its pH (7.3 ± 0.1) was systematically checked.
341
2.2.3. Experimental protocol The established protocol consisted in irradiating samples of 1X PBS/TA solution diluted to a fixed ultrasound power, for a determined time. Experiments were scheduled and test media were pooled before irradiating them successively. After treatment, samples were shielded from surrounding light to avoid any risk of untimely excitation, then frozen. The freezing step serves to block any chemical reactions that could taint subsequent fluorescence measurements. This step has no influence on fluorescence measurements and removes measurement variability caused by varying times between irradiation and fluorescence measurement. It also allows a large number of irradiated samples to be collected before conducting a measurement campaign. The first study focused on the influence of TA concentration and irradiation time on fluorescence. Three series of media with concentrations ranging from 0 to 18 mM TA were irradiated for 30, 60 and 120 s, respectively. Ultrasound power was set to 30 W, which corresponds to an electrical power level setpoint (Psp) of 8 dBm. The media were dispensed to a 96-well late and fluorescence intensity was determined by the plate reader. The second study involved the precise assay of HTA produced according to the ultrasound power applied, related to Psp. In this case, TA concentration and irradiation time were set to 2 mM and 60 s, respectively. The experiment was conducted on a group of 20 samples table from a single dilute solution and irradiated sequentially. This group was dispensed to 4 series of 5 samples sequentially irradiated at a Psp of 0, 2, 4, 6 and 8 dBm. The fluorescence intensity of each sample was determined by spectrofluorimetry. During sample treatment, approximately twenty spectral readings were taken and later used to assess the cavitation phenomenon in the medium.
3. Results 2.2.2. Measurement instruments Fluorescence measurements were made using a micro-plate reader (Applied Biosystems, CytofluorÒ Series 4000) for the first study and a spectrofluorimeter (Kontron InstrumentsÒ, SFM 25) for the second study. The selected excitation filter on the micro-plate reader was centred on 313 nm (±27.5 nm), whereas the emission filter was centred on 460 nm (±20 nm). In order to achieve the best possible signal to noise ratio, 96-well glass micro-plates, black with flat bottoms, were used (WhatmannÒ, black glass bottom Clearview). The reader’s specific gain was adjusted to 80. This instrument is able to perform 96 fluorescence measurements in less than 1 min, with a good degree of sensitivity. Measurements were expressed in arbitrary units (AU) and the measurement uncertainty was of 1 AU. The SFM 25 spectrofluorimeter operates in the 200–800 nm range, with a 1 nm accuracy, thanks to a source made up of a 150 W high-pressure Xenon lamp and excitation and emission monochromators with a minimum bandwidth of 5 nm. More precise than the micro-plate reader, this instrument was used to characterize the HTA solution, with the accurate measurement of the excitation kex and emission kem wavelengths. Once the kem and kex wavelengths have been determined, the instrument’s calibration function adjusts the photo-multiplier voltage, matching the fluorescence of a reference solution (2 lM HTA) to the displayed 100% value. A second function called ‘‘auto-blank”; is run against a control solution. The residual fluorescence of a control solution (2 mM TA) is set to 0% and the instrument computes the correction factor F required to readjust the measurement range. The relative fluorescence (RF) obtained is deduced from the following relationship: RF = F * (Measured fluorescence control fluorescence). Thus, all measurements are standardized and comparable.
3.1. TA concentration and irradiation time The influence of TA concentration on the fluorescence of the medium, irradiated at a fixed ultrasound power, was ascertained for three different irradiation times using a micro-plate reader. Measurements are shown on two diagrams: one according to TA concentration (Fig. 3a) and the other according to irradiation time (Fig. 3b). Within the tested TA concentration range, for each irradiation time, the measured fluorescence values followed a normal distribution, as shown in Table 1. The relationship between fluorescence and irradiation time was estimated by means of a regression line with the following equation: Fluorescence (AU) = 0.2 t (s) + 8.4 with a coefficient of determination, r2 = 0.85. 3.2. Characterization of cavitation Cavitation signal spectra were recorded during the treatment of each sample. These records gave information on the activity of the cavitation phenomenon. In particular, the spectral contour rises with the Psp increase. This broadband noise is characteristic of the presence of the inertial cavitation phenomenon. Amounts of hydroxyl radicals were estimated by measuring the production of HTA within the reaction volume by spectrofluorimetry. The measurement results (Fig. 4) show that RF, or the HTA content within each analysed sample, increases with Psp. At 0 dBm, the mean RF was of circa 5.5% (±2.0). It took on successive values of 30.7% (±3.1), 55.5% (±5.9), 65.5% (±2.8) and 30.7% (±3.1) for respective set points of 2, 4, 6 and 8 dBm. The existing relationship between HTA fluorescence [RF (%)] and the power set point
342
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
a
60 s 30 s
35
80
30 25
RF (%)
Fluorescence (UA)
100
120 s
40
20 15 10
60 40 20
5 0
0
5
10
15
0
20
TA concentration (mM)
b
2
4
6
8
10
Setpoint (dBm)
40
Fig. 4. HTA fluorescence variation versus Psp (dBm): set of four successive irradiation series.
35
Fluorescence (UA)
0
30 25
4.2. TA concentration and irradiation time
20
The influence of TA content in the irradiated medium on HTA fluorescence was assessed for three irradiation times (30, 60 and 120 s). First of all, the fluorescence levels in non-irradiated media, with or without TA, were of between 7 AU and 9 AU. These values are equivalent to those of irradiated media without TA (Fig. 3a). In this case, the fluorescence readings correspond to the background noise, which is highly dependent upon instrument gain and on light diffusion by the measurement plate and medium in the considered well. This noise is independent of the concentration of TA used, ranging from 0 to 18 mM. Changes in fluorescence after irradiation can therefore be assumed to be exclusively caused by the appearance of OH radical due to the ultrasound cavitation phenomenon. As soon as they are formed, the OH radicals are highly unstable and the TA concentration must be sufficiently high to maintain an optimum interaction probability between TA and OH. It is therefore necessary to define a TA concentration that renders HTA fluorescence readings independent of this parameter. It must not, however, be too high. Excessive concentrations lead to the formation of molecular aggregates that, due to their steric footprint, could reduce the probability of specific interaction between the radicals and their potential binding sites on TA. This ‘‘quenching” effect could cause an abnormal drop in the resulting HTA fluorescence values [8]. For the selected TA concentrations (0.4– 18 mM), the measurements made failed to show these limits. Indeed, the more or less long irradiation (30, 60 and 120 s) of these media induced relatively stable and normally distributed fluorescence levels (Table 1) with time. Irradiation time influences fluorescence, which is estimated by the regression line whose equation is as follows:
15 10 5 0
0
30
60
90
120
Irradiation time t (s) Fig. 3. (a) Influence of the TA concentration (0; 0.4; 1; 1.5; 2; 3; 4; 5; 6; 7.5; 9; 10; 12; 15; 18 mM) and irradiation time on the measured fluorescence levels. (b) Relation between the measured fluorescence levels and the irradiation time. Table 1 Statistical results. Irradiation time (s)
Number of measures
Fluorescence mean value (UA)
Standard Deviation
Shapiro– Wilk test W
pvalue
30 14 4.9 1.5 0.91 0.17 60 14 13.7 2.5 0.98 0.96 120 14 23.5 4.4 0.93 0.33 Independence of the concentration in TA: means and standard deviations for each series of measures Study of the adjustment quality for the relation fluorescence (t) Residues 42 0 3.2 0.97 0.35 analysis
[Psp (dBm)] is expressed by the following 2nd degree polynomial regression curve:
RF ¼ 0:66ðPsp 4Þ2 þ 9:14P sp þ 16:05 ðr2 ¼ 0:98Þ: 4. Discussion 4.1. Methodology With the ulterior goal of studying the effects induced by cavitation in the cellular environment, samples were created using PBS, an isotonic and non-toxic physiological medium comprised mainly of sodium chloride (137 mM) and potassium phosphate (9 mM). Comparative tests of this solution with water showed no significant differences in relative fluorescence readings.
FluorescenceðAUÞ ¼ 0:2tðsÞ þ 8:4 The intercept constant (8.4 AU) represents the estimated mean fluorescence level of non-irradiated media. This value matches previously made observations. Consequently, the production of OH radicals is proportional to medium irradiation time, thus demonstrating that the transient cavitation phenomenon is initiated and maintained in the irradiated medium. Residue analysis shows random distribution of values around the estimated curve (Table 1 and Fig. 3b). The measurements are therefore independent. These results are in conformity with the random fluctuations of the ‘‘cavitation level” induced by the acoustic field applied to the measurement cell. Fluctuation amplitude, on the other hand, did not remain constant
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
with irradiation time. A probable explanation stems from changes in physical parameters over time. In particular, increasing irradiation times lead to corresponding increases in irradiated medium temperature. Thus, preliminary measurements performed with our configuration showed a 12 °C temperature increase, measured by temperature sensor (Physitemp, thermocouple microprobe IT23) placed in the measurement cell and offset from the acoustic focal point by 4 mm. The temperature increase measured is mainly due to an absorption phenomenon of polyethylene structure of measurement cell. Furthermore, this temperature increase could cause an increase in static pressure in the sealed measurement cell. Temperature and static pressure are central physical parameters for the development of the cavitation process and the inability to guarantee their stability within an ultrasound field could therefore lead to a very high degree of variability in its effects, including OH production. To conclude, the TA concentrations of between 0.4 and 18 mM in the irradiated media did not cause any significant deviations in the measured fluorescence levels. These levels were, however, dependent upon irradiation time and increased in a linear fashion with this parameter. To accurately assay hydroxyl radicals generated by the focused ultrasound system, the following results (Fig. 3a) show that a TA concentration of 2 mM is sufficient. A 60 s irradiation time is a good compromise for limiting the induced thermal effects, while guaranteeing significant HTA production (Fig. 3b). 4.3. Characterization of cavitation Inertial cavitation levels were assessed by HTA production in the reaction medium, measured by spectrofluorimetry. The influence of acoustic power on HTA production was assessed at between 5 and 30 W, corresponding to an excitation power range of between 0 and 8 dBm. Analysis of the extent of the measurement range leads to the following two observations: (i) When the set point is equal to or less than 0 dBm, the fluorescence readings are very low, or at least undetectable by spectrofluorimetry within the pre-calibrated measurement range. They reflect a limitation in the production of OH radicals, with an absence or non-persistence of the inertial cavitation phenomenon. This conclusion is supported by the spectral recordings which, on the whole, presented a characteristic line at half-frequency and the absence of broadband noise corresponding to presence of non-inertial cavitation phenomenon. For these settings, the appearance of inertial cavitation remains an exceptional, undoubtedly very brief phenomenon, difficult to detect with the instruments used, hence with a low tendency to produce OH radicals. At 0 dBm, spectra displayed fluctuant shapes with the random presence of broadband noise attached to inertial cavitation phenomenon, thus confirming the difficulties in initiating the phenomenon at this setting. (ii) The studied range of settings did not exceed 8 dBm as higher excitation power levels tend to damage the measurement cell at the ultrasound wave impact point. Indeed, the acoustic power is such that the cavitation phenomenon occurs in degassed water at this interface. As the acoustic power transmission conditions are no longer met, the fluorescence readings at such settings were deliberately omitted. In the considered acoustic power range, RF readings increased with Psp (dBm), showing an increase in activity due to the cavitation phenomenon in the medium according to Psp. These results are supported by the mean spectral readings showing an increase in spectral contours with Psp. The relationship RF (Psp)
343
was estimated by means of a 2nd degree polynomial regression curve, with a very high correlation (coefficient of determination: r2 = 0.98). Residue distribution is normal, with a standard deviation of 3.9 (Shapiro–Wilk test: W = 0.95; pvalue = 0.50; number of values: 20). The polynomial equation possesses a relatively low dominant coefficient: 0.66. This factor remains highly relevant (pvalue = 0.0002), resulting in a low degree of convex curvature of the curve in the selected irradiation range. Curve orientation is closely linked to the slope of the carrier line: 9.14. Acoustic power is the main parameter from which OH production may be estimated. It is worth reminding that the data are the result of successive irradiation operations conducted under very similar external temperature and pressure conditions, thus leading to relatively reproducible results. A more complex model could be developed, taking into account Psp, temperature and static pressure inside the measurement cell. In practice, however, specific sensors are difficult to integrate into the sealed cell and, furthermore, would disrupt the ultrasound field. In all cases, the inertial cavitation process starts at 0 dBm (acoustic power: 5 W) with a quasi-similar production of OH (RF 5%). The relative fluorescence (image of the level of inertial cavitation) is, at first approximation, proportional to the logarithm of acoustic power or peak acoustic pressure within the measurement cell of the experimental set up (Fig. 2). 5. Conclusion Several authors have demonstrated that the analysis of the water sonolysis reaction, occurring in a medium exposed to the ultrasound cavitation phenomenon, constitutes an effective and interesting means of indicating the intensity of this phenomenon. TA dosimetry is a simple and suitable method for characterizing the inertial cavitation level of a focused ultrasound system with sufficient detection efficiency based on the fluorescence phenomenon. It has been shown that the production of OH is, at first approximation, proportional to the logarithm of the acoustic power applied to the medium and that the proportionality factor is governed by experimental conditions. It should be remembered, however, that this method employs a chemical dosimeter and as such, it may be only the ideal method for estimation of inertial cavitation production. These results constitute a prior step in the study of new therapeutic approaches involving cavitation, such as sonochemotherapy or ultrasound-mediated transfection. Acknowledgement This work was supported by the ‘‘Association Française contre les Myopathies (AFM)” Grant #9594. References [1] J.R. Mc Lean, A.J. Mortimer, A cavitation and free radical dosimeter for ultrasound, Ultrasound Med. Biol. 14 (1988) 59–64. [2] X. Fang, G. Mark, C. von Sonntag, OH radical formation by ultrasound in aqueous solutions – part I: the chemistry underlying the terephthalate dosimeter, Ultrason. Sonochem. 3 (1996) 57–63. [3] T.J. Mason, J.P. Lorimer, D.M. Bates, Y. Zhao, Dosimetry in sonochemistry: the use of aqueous terephthalate ions as a fluorescence monitor, Ultrason. Sonochem. 1 (1994) 91–95. [4] M. Digby, F.A. Duck, E.J. Lenz, G.J. Price, Technical note: measurement of collapse cavitation in ultrasound fields, Brit. J. Radiol. 68 (1995) 1244–1248. [5] G. Mark, A. Tauber, R. Laupert, H.P. Schuchmann, D. Schultz, A. Mues, C. von Sonntag, PH-radical formation by ultrasound in aqueous solution – part II: terephthalate and Fricke dosimetry and the influence of various conditions on the sonolytic yield, Ultrason. Sonochem. 5 (1998) 41–52. [6] G.J. Price, F.A. Duck, M. Digby, W. Holland, T. Berryman, Measurement of radical production as a result of cavitation in medical ultrasound fields, Ultrason. Sonochem. 4 (1997) 165–171.
344
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
[7] Y. Hu, Z. Zhang, C. Yang, Measurement of hydroxyl radical production in ultrasonic aqueous solutions by a novel chemiluminescence method, Ultrason. Sonochem. 15 (2008) 665–672. [8] G.J. Price, E.J. Lenz, The use of dosimeters to measure radical production in aqueous sonochemical systems, Ultrasonics 31 (1993) 451–456.
[9] W.A. Armstrong, R.A. Facey, D.W. Grant, W.G. Humphreys, A tissue-equivalent chemical dosimeter sensitive to 1 rad, Can. J. Chem. 41 (1963) 1575–1577. [10] R.W. Matthews, The radiation chemistry of the terephthalate dosimeter, Radiat. Res. 83 (1980) 27–41. [11] F. Davidson, Ultrasonic Power Balances, Preston Ed. Spinger-Verlag, 1991.
Contents lists available at ScienceDirect
Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultsonch
Assay of hydroxyl radicals generated by focused ultrasound L. Villeneuve a,*, L. Alberti c, J.-P. Steghens d, J.-M. Lancelin e, J.-L. Mestas b a
Hospices Civils de Lyon, Pôle d’Information Médicale, Evaluation et Recherche, Lyon F-69003, France Inserm, U556, Lyon, F-69003, France; Université de Lyon, Lyon, F-69003, France INSERM, U590, Centre Léon Bérard, Université Lyon 1, Lyon F-69008, France d Hospices Civils de Lyon, Laboratoire de Biochimie et de Biologie, Lyon F-69003, France e CNRS, UMR 5180, Université de Lyon, Lyon F-69003, France b c
a r t i c l e
i n f o
Article history: Received 29 July 2008 Received in revised form 15 September 2008 Accepted 16 September 2008 Available online 10 October 2008 PACS: 43.35 Keywords: Cavitation Ultrasound Sonolysis Terephthalic acid Hydroxyl radicals Fluorescence
a b s t r a c t Water sonolysis leads to the formation of hydroxyl radicals (OH). Various techniques are used to detect the OH production and thus to assess the level of ultrasound-mediated cavitation generated in vitro. In this study, we used terephthalic acid (TA) as an OH trap. This method is based on the fluorescent properties of hydroxyterephthalic acid (HTA) formed by the reaction of TA with OH and used as an indicator of the degree of inertial cavitation caused. The experimental system is comprised mainly of a focused piezoelectric ultrasound transmitter and a measurement cell containing 1X PBS/TA diluted solution. In the first part, we aimed to characterize the most appropriate experimental conditions (TA dosimeter solution, irradiation time) in order to optimize the resulting HTA fluorescence values. Then, we could determine that the HTA production increased with the level of the cavitation phenomenon caused by the acoustic power from which OH production may be estimated. Ó 2008 Elsevier B.V. All rights reserved.
1. Introduction The presence of ultrasound-induced inertial cavitation manifests itself by very intense local physicochemical effects: microstreaming, sonoluminescence, brutal increase in local temperature, formation radical-containing chemical species [1–4]. The energy levels achieved are such that they can lead to the homolytic cleavage of covalent bonds between the oxygen and hydrogen atoms comprising the water molecules present. Water sonolysis leads to the formation of reactive oxygen species (ROS) such as hydroxyl radicals (OH) and hydrogen peroxide (H2O2) [2] according to the following equations:
H2 O ! H þ OH H þ O2 ! HO2 HO2 þ HO2 ! H2 O2 þ O2 OH þ OH ! H2 O2 H þ H2 O ! H2 OH * Corresponding author. Tel.: +33 472 681 944; fax: +33 472 681 931. E-mail address: [email protected] (L. Villeneuve). 1350-4177/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ultsonch.2008.09.007
Currently, various techniques are used to detect the production of ROS and thus to assess the level of ultrasound-mediated cavitation generated in vitro within a reaction volume [5]. In addition to electron spin resonance recognized as the reference technique for assaying OH produced by ultrasound-mediated cavitation [3,6], a chemiluminescent method is known for its enhanced sensitivity and lower detection limits [7]. A number of other techniques are less sensitive, but easier to implement. Thus, as iodometry, based on the Weissler reaction [8], can be used to detect hydrogen peroxide (H2O2), or whereas the Fricke dosimeter is based on the oxidation of Fe2+ ions by OH, HO2 and H2O2 compounds present in the medium [2,8]. These techniques are used to detect compounds generated by reactions involving OH radicals. They are called indirect methods and tend to lose their sensitivity when the irradiated medium contains other organic compounds with which the radicals can interact. Beyond their very high chemical reactivity, OH radicals are characterised by their very short life span (circa 109 s). Their detection therefore requires that they be captured as soon as possible after their formation. In this respect, the method presented in this study, using terephthalic acid (TA) as a hydroxyl radical trap, is an interesting alternative. Initially developed by
340
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
a electrical power applied to the focused ultrasound source of between 40 mW and 200 W, corresponding to a focal point pressure of between 0.07 and 6.32 Mpa (Fig. 2). A composite ceramic dish with a resonant frequency of 1 MHz, mounted on a very high acoustic impedance medium (backing) acts as a transducer dedicated to reception of the acoustic signal produced by the cavitation phenomenon in the propagation medium. This transducer, with a diameter of 100 mm and a radius of curvature of 100 mm, is placed perpendicular to the sound axis of the ultrasound transmitter and positioned such that its centre of curvature is aligned with the centre of the measurement cell. The received signal is fed through a wide-band isolation transformer, amplified (NF Electronic InstrumentsÒ BX31) and then sent to a spectrum sweep-analyzer (Hewlett PackardÒ ESA L 1500 A). The examined frequency range is 0.1–7.1 MHz, the frequency resolution bandwidth 10 kHz and the duration of analysis 50 ms. Ten successive power spectra are averaged to reduce noise. A polyethylene pipette (CMLÒ; bulb dimensions: diameter = 9 mm; length = 20 mm), mechanically integral to the excitation source, is used as a measurement cell and contains the dosimeter solution (1.4 ml) to be exposed to the ultrasound. The posterior part of the measurement cell is sealed to isolate its contents from the outside medium. It is then directed along the sound axis and centred on the ultrasound transmitter’s centre of curvature (focal point). Ultrasound wave attenuation by the measurement cell wall (thickness: 0.3 mm) is less than 0.1 dB.
Armstrong et al. and used in radiation chemistry [9–10], this method is based on the fluorescent properties of hydroxyterephthalic acid (HTA), formed by the reaction of TA with OH. Indeed, contrary to TA, that has no fluorescent properties its hydroxylation by product (HTA) emits a fluorescent signal that can be observed at 424 nm when excited by a light source at a wavelength of 323 nm [8]. The present study consisted in assessing the production of OH radicals based on the assay of HTA, used as an indicator of the degree of inertial cavitation caused by a focused ultrasound system within a reaction volume. It was conducted in two steps. The first consisted in assessing the influence of TA concentration and irradiation time on the fluorescence of the treated medium; the second step focused on precisely characterising the degree of cavitation according to the acoustic power of the ultrasound source used. 2. Materials and methods 2.1. Experimental system The experimental system is comprised mainly of a focused ultrasound transmitter and a measurement cell containing the reaction medium (Fig. 1). Both elements are contained in a tank filled with degassed water (bath temperature: 20–25 °C; dissolved oxygen content: 3 ± 0.5 mg/L, measured by an oxymeter (Oxi 340 WTWÒ). The ultrasound transmitter is a focused piezoelectric device (Channel Industries Inc.Ò, C 5500) with a diameter of 100 mm, a radius of curvature of 100 mm and a resonant frequency of 473.7 kHz. The focal volume, defined by an acoustic pressure greater than 50% of maximum pressure, is 35 mm deep and 4 mm in diameter. The use of a sine wave generator (Hewlett-PackardÒ 33120 A; setting range: 22–22 dBm), followed by an attenuator (16 dB) and an amplifier (AdeceÒ, 200 W, 50 dB) generates
2.2. Dosimetry 2.2.1. Chemical solutions A reference solution containing 2 lM HTA in a 1X PBS (phosphate-buffered saline) solution was prepared fresh prior to each
Spectrum analyser
20 dB
IEEE Bus
Acoustic receiver Silicon Transducer Sine generator Amplifier
Focal volume
473,7 kHz
PC controller
Degassed water
Focal volume
Measurement cell
Fig. 1. Experimental system.
Focalized transducer
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
Power
10
Pressure
100
10
1
1
0.1 -15
-12
-9
-6
-3
0
3
6
9
12
15
Pressure (MPa)
Acoustic Power (W)
1000
0.1 18
Psp (dBm) Fig. 2. Influence of the power set point (Psp) on the acoustic power [11] and peak pressure measured on the focal point (hydrophone GEC Marconi Y-33-7611).
fluorescence measurement campaign, to check and calibrate the measurement instruments. The HTA was taken from a 20 mM stock solution prepared according to the method described by Mason et al. [3]. This stock solution was stored at +4 °C in the dark; its pH was systematically checked prior to use (pH = 7.3 ± 0.1; Radiometer CopenhagenÒ, PHM 210 standard pH meter). The reference solution excitation and emission spectra were measured by spectrofluorimetry. The maximum excitation and emission wavelengths were, respectively, kex = 323 nm and kem = 424 nm. A 20 mM TA stock solution (Acros OrganicsÒ) was prepared according to the method of Mason et al. [3] and stored at +4 °C in the dark for several weeks. Prior to use, its pH (7.3 ± 0.1) was systematically checked.
341
2.2.3. Experimental protocol The established protocol consisted in irradiating samples of 1X PBS/TA solution diluted to a fixed ultrasound power, for a determined time. Experiments were scheduled and test media were pooled before irradiating them successively. After treatment, samples were shielded from surrounding light to avoid any risk of untimely excitation, then frozen. The freezing step serves to block any chemical reactions that could taint subsequent fluorescence measurements. This step has no influence on fluorescence measurements and removes measurement variability caused by varying times between irradiation and fluorescence measurement. It also allows a large number of irradiated samples to be collected before conducting a measurement campaign. The first study focused on the influence of TA concentration and irradiation time on fluorescence. Three series of media with concentrations ranging from 0 to 18 mM TA were irradiated for 30, 60 and 120 s, respectively. Ultrasound power was set to 30 W, which corresponds to an electrical power level setpoint (Psp) of 8 dBm. The media were dispensed to a 96-well late and fluorescence intensity was determined by the plate reader. The second study involved the precise assay of HTA produced according to the ultrasound power applied, related to Psp. In this case, TA concentration and irradiation time were set to 2 mM and 60 s, respectively. The experiment was conducted on a group of 20 samples table from a single dilute solution and irradiated sequentially. This group was dispensed to 4 series of 5 samples sequentially irradiated at a Psp of 0, 2, 4, 6 and 8 dBm. The fluorescence intensity of each sample was determined by spectrofluorimetry. During sample treatment, approximately twenty spectral readings were taken and later used to assess the cavitation phenomenon in the medium.
3. Results 2.2.2. Measurement instruments Fluorescence measurements were made using a micro-plate reader (Applied Biosystems, CytofluorÒ Series 4000) for the first study and a spectrofluorimeter (Kontron InstrumentsÒ, SFM 25) for the second study. The selected excitation filter on the micro-plate reader was centred on 313 nm (±27.5 nm), whereas the emission filter was centred on 460 nm (±20 nm). In order to achieve the best possible signal to noise ratio, 96-well glass micro-plates, black with flat bottoms, were used (WhatmannÒ, black glass bottom Clearview). The reader’s specific gain was adjusted to 80. This instrument is able to perform 96 fluorescence measurements in less than 1 min, with a good degree of sensitivity. Measurements were expressed in arbitrary units (AU) and the measurement uncertainty was of 1 AU. The SFM 25 spectrofluorimeter operates in the 200–800 nm range, with a 1 nm accuracy, thanks to a source made up of a 150 W high-pressure Xenon lamp and excitation and emission monochromators with a minimum bandwidth of 5 nm. More precise than the micro-plate reader, this instrument was used to characterize the HTA solution, with the accurate measurement of the excitation kex and emission kem wavelengths. Once the kem and kex wavelengths have been determined, the instrument’s calibration function adjusts the photo-multiplier voltage, matching the fluorescence of a reference solution (2 lM HTA) to the displayed 100% value. A second function called ‘‘auto-blank”; is run against a control solution. The residual fluorescence of a control solution (2 mM TA) is set to 0% and the instrument computes the correction factor F required to readjust the measurement range. The relative fluorescence (RF) obtained is deduced from the following relationship: RF = F * (Measured fluorescence control fluorescence). Thus, all measurements are standardized and comparable.
3.1. TA concentration and irradiation time The influence of TA concentration on the fluorescence of the medium, irradiated at a fixed ultrasound power, was ascertained for three different irradiation times using a micro-plate reader. Measurements are shown on two diagrams: one according to TA concentration (Fig. 3a) and the other according to irradiation time (Fig. 3b). Within the tested TA concentration range, for each irradiation time, the measured fluorescence values followed a normal distribution, as shown in Table 1. The relationship between fluorescence and irradiation time was estimated by means of a regression line with the following equation: Fluorescence (AU) = 0.2 t (s) + 8.4 with a coefficient of determination, r2 = 0.85. 3.2. Characterization of cavitation Cavitation signal spectra were recorded during the treatment of each sample. These records gave information on the activity of the cavitation phenomenon. In particular, the spectral contour rises with the Psp increase. This broadband noise is characteristic of the presence of the inertial cavitation phenomenon. Amounts of hydroxyl radicals were estimated by measuring the production of HTA within the reaction volume by spectrofluorimetry. The measurement results (Fig. 4) show that RF, or the HTA content within each analysed sample, increases with Psp. At 0 dBm, the mean RF was of circa 5.5% (±2.0). It took on successive values of 30.7% (±3.1), 55.5% (±5.9), 65.5% (±2.8) and 30.7% (±3.1) for respective set points of 2, 4, 6 and 8 dBm. The existing relationship between HTA fluorescence [RF (%)] and the power set point
342
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
a
60 s 30 s
35
80
30 25
RF (%)
Fluorescence (UA)
100
120 s
40
20 15 10
60 40 20
5 0
0
5
10
15
0
20
TA concentration (mM)
b
2
4
6
8
10
Setpoint (dBm)
40
Fig. 4. HTA fluorescence variation versus Psp (dBm): set of four successive irradiation series.
35
Fluorescence (UA)
0
30 25
4.2. TA concentration and irradiation time
20
The influence of TA content in the irradiated medium on HTA fluorescence was assessed for three irradiation times (30, 60 and 120 s). First of all, the fluorescence levels in non-irradiated media, with or without TA, were of between 7 AU and 9 AU. These values are equivalent to those of irradiated media without TA (Fig. 3a). In this case, the fluorescence readings correspond to the background noise, which is highly dependent upon instrument gain and on light diffusion by the measurement plate and medium in the considered well. This noise is independent of the concentration of TA used, ranging from 0 to 18 mM. Changes in fluorescence after irradiation can therefore be assumed to be exclusively caused by the appearance of OH radical due to the ultrasound cavitation phenomenon. As soon as they are formed, the OH radicals are highly unstable and the TA concentration must be sufficiently high to maintain an optimum interaction probability between TA and OH. It is therefore necessary to define a TA concentration that renders HTA fluorescence readings independent of this parameter. It must not, however, be too high. Excessive concentrations lead to the formation of molecular aggregates that, due to their steric footprint, could reduce the probability of specific interaction between the radicals and their potential binding sites on TA. This ‘‘quenching” effect could cause an abnormal drop in the resulting HTA fluorescence values [8]. For the selected TA concentrations (0.4– 18 mM), the measurements made failed to show these limits. Indeed, the more or less long irradiation (30, 60 and 120 s) of these media induced relatively stable and normally distributed fluorescence levels (Table 1) with time. Irradiation time influences fluorescence, which is estimated by the regression line whose equation is as follows:
15 10 5 0
0
30
60
90
120
Irradiation time t (s) Fig. 3. (a) Influence of the TA concentration (0; 0.4; 1; 1.5; 2; 3; 4; 5; 6; 7.5; 9; 10; 12; 15; 18 mM) and irradiation time on the measured fluorescence levels. (b) Relation between the measured fluorescence levels and the irradiation time. Table 1 Statistical results. Irradiation time (s)
Number of measures
Fluorescence mean value (UA)
Standard Deviation
Shapiro– Wilk test W
pvalue
30 14 4.9 1.5 0.91 0.17 60 14 13.7 2.5 0.98 0.96 120 14 23.5 4.4 0.93 0.33 Independence of the concentration in TA: means and standard deviations for each series of measures Study of the adjustment quality for the relation fluorescence (t) Residues 42 0 3.2 0.97 0.35 analysis
[Psp (dBm)] is expressed by the following 2nd degree polynomial regression curve:
RF ¼ 0:66ðPsp 4Þ2 þ 9:14P sp þ 16:05 ðr2 ¼ 0:98Þ: 4. Discussion 4.1. Methodology With the ulterior goal of studying the effects induced by cavitation in the cellular environment, samples were created using PBS, an isotonic and non-toxic physiological medium comprised mainly of sodium chloride (137 mM) and potassium phosphate (9 mM). Comparative tests of this solution with water showed no significant differences in relative fluorescence readings.
FluorescenceðAUÞ ¼ 0:2tðsÞ þ 8:4 The intercept constant (8.4 AU) represents the estimated mean fluorescence level of non-irradiated media. This value matches previously made observations. Consequently, the production of OH radicals is proportional to medium irradiation time, thus demonstrating that the transient cavitation phenomenon is initiated and maintained in the irradiated medium. Residue analysis shows random distribution of values around the estimated curve (Table 1 and Fig. 3b). The measurements are therefore independent. These results are in conformity with the random fluctuations of the ‘‘cavitation level” induced by the acoustic field applied to the measurement cell. Fluctuation amplitude, on the other hand, did not remain constant
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
with irradiation time. A probable explanation stems from changes in physical parameters over time. In particular, increasing irradiation times lead to corresponding increases in irradiated medium temperature. Thus, preliminary measurements performed with our configuration showed a 12 °C temperature increase, measured by temperature sensor (Physitemp, thermocouple microprobe IT23) placed in the measurement cell and offset from the acoustic focal point by 4 mm. The temperature increase measured is mainly due to an absorption phenomenon of polyethylene structure of measurement cell. Furthermore, this temperature increase could cause an increase in static pressure in the sealed measurement cell. Temperature and static pressure are central physical parameters for the development of the cavitation process and the inability to guarantee their stability within an ultrasound field could therefore lead to a very high degree of variability in its effects, including OH production. To conclude, the TA concentrations of between 0.4 and 18 mM in the irradiated media did not cause any significant deviations in the measured fluorescence levels. These levels were, however, dependent upon irradiation time and increased in a linear fashion with this parameter. To accurately assay hydroxyl radicals generated by the focused ultrasound system, the following results (Fig. 3a) show that a TA concentration of 2 mM is sufficient. A 60 s irradiation time is a good compromise for limiting the induced thermal effects, while guaranteeing significant HTA production (Fig. 3b). 4.3. Characterization of cavitation Inertial cavitation levels were assessed by HTA production in the reaction medium, measured by spectrofluorimetry. The influence of acoustic power on HTA production was assessed at between 5 and 30 W, corresponding to an excitation power range of between 0 and 8 dBm. Analysis of the extent of the measurement range leads to the following two observations: (i) When the set point is equal to or less than 0 dBm, the fluorescence readings are very low, or at least undetectable by spectrofluorimetry within the pre-calibrated measurement range. They reflect a limitation in the production of OH radicals, with an absence or non-persistence of the inertial cavitation phenomenon. This conclusion is supported by the spectral recordings which, on the whole, presented a characteristic line at half-frequency and the absence of broadband noise corresponding to presence of non-inertial cavitation phenomenon. For these settings, the appearance of inertial cavitation remains an exceptional, undoubtedly very brief phenomenon, difficult to detect with the instruments used, hence with a low tendency to produce OH radicals. At 0 dBm, spectra displayed fluctuant shapes with the random presence of broadband noise attached to inertial cavitation phenomenon, thus confirming the difficulties in initiating the phenomenon at this setting. (ii) The studied range of settings did not exceed 8 dBm as higher excitation power levels tend to damage the measurement cell at the ultrasound wave impact point. Indeed, the acoustic power is such that the cavitation phenomenon occurs in degassed water at this interface. As the acoustic power transmission conditions are no longer met, the fluorescence readings at such settings were deliberately omitted. In the considered acoustic power range, RF readings increased with Psp (dBm), showing an increase in activity due to the cavitation phenomenon in the medium according to Psp. These results are supported by the mean spectral readings showing an increase in spectral contours with Psp. The relationship RF (Psp)
343
was estimated by means of a 2nd degree polynomial regression curve, with a very high correlation (coefficient of determination: r2 = 0.98). Residue distribution is normal, with a standard deviation of 3.9 (Shapiro–Wilk test: W = 0.95; pvalue = 0.50; number of values: 20). The polynomial equation possesses a relatively low dominant coefficient: 0.66. This factor remains highly relevant (pvalue = 0.0002), resulting in a low degree of convex curvature of the curve in the selected irradiation range. Curve orientation is closely linked to the slope of the carrier line: 9.14. Acoustic power is the main parameter from which OH production may be estimated. It is worth reminding that the data are the result of successive irradiation operations conducted under very similar external temperature and pressure conditions, thus leading to relatively reproducible results. A more complex model could be developed, taking into account Psp, temperature and static pressure inside the measurement cell. In practice, however, specific sensors are difficult to integrate into the sealed cell and, furthermore, would disrupt the ultrasound field. In all cases, the inertial cavitation process starts at 0 dBm (acoustic power: 5 W) with a quasi-similar production of OH (RF 5%). The relative fluorescence (image of the level of inertial cavitation) is, at first approximation, proportional to the logarithm of acoustic power or peak acoustic pressure within the measurement cell of the experimental set up (Fig. 2). 5. Conclusion Several authors have demonstrated that the analysis of the water sonolysis reaction, occurring in a medium exposed to the ultrasound cavitation phenomenon, constitutes an effective and interesting means of indicating the intensity of this phenomenon. TA dosimetry is a simple and suitable method for characterizing the inertial cavitation level of a focused ultrasound system with sufficient detection efficiency based on the fluorescence phenomenon. It has been shown that the production of OH is, at first approximation, proportional to the logarithm of the acoustic power applied to the medium and that the proportionality factor is governed by experimental conditions. It should be remembered, however, that this method employs a chemical dosimeter and as such, it may be only the ideal method for estimation of inertial cavitation production. These results constitute a prior step in the study of new therapeutic approaches involving cavitation, such as sonochemotherapy or ultrasound-mediated transfection. Acknowledgement This work was supported by the ‘‘Association Française contre les Myopathies (AFM)” Grant #9594. References [1] J.R. Mc Lean, A.J. Mortimer, A cavitation and free radical dosimeter for ultrasound, Ultrasound Med. Biol. 14 (1988) 59–64. [2] X. Fang, G. Mark, C. von Sonntag, OH radical formation by ultrasound in aqueous solutions – part I: the chemistry underlying the terephthalate dosimeter, Ultrason. Sonochem. 3 (1996) 57–63. [3] T.J. Mason, J.P. Lorimer, D.M. Bates, Y. Zhao, Dosimetry in sonochemistry: the use of aqueous terephthalate ions as a fluorescence monitor, Ultrason. Sonochem. 1 (1994) 91–95. [4] M. Digby, F.A. Duck, E.J. Lenz, G.J. Price, Technical note: measurement of collapse cavitation in ultrasound fields, Brit. J. Radiol. 68 (1995) 1244–1248. [5] G. Mark, A. Tauber, R. Laupert, H.P. Schuchmann, D. Schultz, A. Mues, C. von Sonntag, PH-radical formation by ultrasound in aqueous solution – part II: terephthalate and Fricke dosimetry and the influence of various conditions on the sonolytic yield, Ultrason. Sonochem. 5 (1998) 41–52. [6] G.J. Price, F.A. Duck, M. Digby, W. Holland, T. Berryman, Measurement of radical production as a result of cavitation in medical ultrasound fields, Ultrason. Sonochem. 4 (1997) 165–171.
344
L. Villeneuve et al. / Ultrasonics Sonochemistry 16 (2009) 339–344
[7] Y. Hu, Z. Zhang, C. Yang, Measurement of hydroxyl radical production in ultrasonic aqueous solutions by a novel chemiluminescence method, Ultrason. Sonochem. 15 (2008) 665–672. [8] G.J. Price, E.J. Lenz, The use of dosimeters to measure radical production in aqueous sonochemical systems, Ultrasonics 31 (1993) 451–456.
[9] W.A. Armstrong, R.A. Facey, D.W. Grant, W.G. Humphreys, A tissue-equivalent chemical dosimeter sensitive to 1 rad, Can. J. Chem. 41 (1963) 1575–1577. [10] R.W. Matthews, The radiation chemistry of the terephthalate dosimeter, Radiat. Res. 83 (1980) 27–41. [11] F. Davidson, Ultrasonic Power Balances, Preston Ed. Spinger-Verlag, 1991.