- Email: [email protected]
Quartz tuning fork as in situ sensor of bacterial biofilm
Sensors and Actuators B 210 (2015) 825–829
Contents lists available at ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Quartz tuning fork as in situ sensor of bacterial biofilm夽 Tomasz Piasecki a,∗ , Grzegorz Guła b , Karol Waszczuk a , Zuzanna Drulis-Kawa b , Teodor Gotszalk a a b
Faculty of Microsystem Electronics and Photonics, Wrocław University of Technology, Z. Janiszewskiego 11/17, 50-372 Wrocław, Poland Institute of Genetics and Microbiology, University of Wrocław, S. Przybyszewskiego 63/77, 51-148 Wrocław, Poland
a r t i c l e
i n f o
Article history: Available online 8 January 2015 Keywords: Pseudomonas aeruginosa Ringdown Resonant frequency Dissipation factor Viscosity Density
a b s t r a c t Piezoelectric quartz tuning forks (QTFs) were used as in situ sensors of biofilm formation. Modulated excitation technique allowed to measure the current generated by the QTF during the ring-down of the oscillations at conductive environments such as a liquid culture growth medium. New algorithm of the digital analysis of the ring-down signal was elaborated and used to determine the decay time constant of the vibrations and the damped vibration frequency. The dependence of these parameters on the time for QTF placed in liquid growth medium contained four strains of Pseudomonas aeruginosa was measured. It allowed to detect the biofilm formation at the QTF after few hours of incubation for all used strains at 106 cfu/ml initial concentration (2 h for PA01 strain) or 10 h for ATCC 27853 strain at 10 cfu/ml initial concentration. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The bacterial biofilm represents the most prevalent type of microbial growth. The extracellular matrix in which biofilm cells are embedded cause the resistance to antibiotics or immune defense [1,2] and good adhesion to the biotic and abiotic surfaces [3–5]. Microorganisms that produce extracellular matrix could colonize indwelling medical devices to form a biofilm. The deviceassociated infections result in substantial morbidity and mortality, and potentially represent a large increase the treatment cost [2,6]. There is a need for fast, preferably in situ method which would allow to determine whether the biofilm formation was prevented or to study the influence of the environment on the biofilm properties [7,8]. Such measurements may be performed using vibrating viscometers. They change their properties depending on the mass and rheology of the bacterial biofilm which adheres to the surface of the resonator [9,10]. The most common type of piezoelectric element used for biofilm detection is the quartz crystal microbalance (QCM) [11–13]. It is established method which is based by the observation
夽 Selected papers presented at EUROSENSORS 2014, the XXVIII edition of the conference series, Brescia, Italy, September 7–10, 2014. ∗ Corresponding author. Tel.: +48 713203223. E-mail address: [email protected] (T. Piasecki). http://dx.doi.org/10.1016/j.snb.2014.12.105 0925-4005/© 2015 Elsevier B.V. All rights reserved.
of the changes of resonant frequency during the incubation of the biofilm or additionally monitoring of the dissipation of the vibration energy in the QCM-D variant of that technique [14]. It vibrates in shear mode and typical resonant frequency is few MHz or higher. In this paper another type of resonator is used: the quartz tuning forks (QTFs). Commercially available QTFs are inexpensive and mass produced. Their resonant frequency is 215 Hz (about 32.8 kHz). They are commonly used as the time reference in real time clocks however they also find another applications. One of the alternative application of the QTF are the viscosity and mass density measurements. The properties of the fluid that surrounds the vibrating element change the resonant frequency and the damping of the vibrations [15–19]. Therefore it should be possible to detect the influence of viscous biofilm adhered to its surface. Commercial QTFs have the electrodes fabricated at the outer surface of the piezoelectric crystal. If QTF is immersed in conductive liquids, such as the liquid growth media used in the culture growth, the parasitic currents will flow when voltage is applied to the QTF. It renders common electrical methods unsuitable for determination of the vibrations of commercial QTF placed in conductive liquids. Special technique based on modulated excitation was elaborated. It allowed for direct measurement of the current generated by short-circuited, vibrating QTF during the ring-down phase of the vibrations [20]. Its further development and application in detection of biofilm formation is presented in this paper.
826
T. Piasecki et al. / Sensors and Actuators B 210 (2015) 825–829
QTF1
actuation
response
QTF2 Rt
... QTF6
400x
TRIG
OUT SYNC
DAQ Agilent U2531A
Agilent 33220A Fig. 1. The schematic diagram of the QTF oscillation ringdown measurement system.
2. Materials and methods 2.1. QTF oscillations ringdown measurement system Measurements of QTF oscillations ringdown were performed using computer controlled home-built system which utilized the measurement technique based on the modulated excitation of the QTF. In this method the vibrations of the resonator are excited by the sinusoidal voltage burst after which the resonator is short circuited and the current generated during the ring-down of the oscillations due to the direct piezoelectric effect is measured. The details of this technique were already published [20]. It allows to excite and measure the vibrations of the QTF covered with thin insulating layer despite the influence of the parasitic capacitance and the conductance of the liquid in which the QTF is immersed. The further development of that technique by the introduction of digital processing of the output signal is presented. The schematic diagram of the measurement system was shown in Fig. 1. It consisted of the Agilent 33220A function generator, home-built excitation and measurement electronics and Agilent U2531A DAQ module. The function generator provided the excitation and synchronization signals. The latter was used to synchronize all components of the measurement system with the burst and measurement phases. The home-built circuitry (Fig. 2) provided the excitation signal conditioning and splitting it into six channels which were switched electronically. It allowed to measure the vibrations of up to six QTFs placed in the six wells of 24-well titrate plate, switched sequentially. Current generated during the oscillation ring-down was measured using the transimpedance amplifier which output was additionally amplified 400 times. The effective conversion ratio of the system was kCVC = 0.4 V/nA. Output signal was sampled using Agilent U2531A 2MSa/s, 14 bit data acquisition module. The parameters of the excitation signal, channel switching and data acquisition
Fig. 2. The photograph of the QTF oscillation measurement system home-built actuation and readout electronics connected to a QTF holder on a titrate plate.
were controlled using dedicated software at the personal computer. Typical measurement parameters were as follows: the excitation signal amplitude – 0.25 V and frequency at the resonance (around 28,500 Hz), burst length – 500 cycles, burst period – 35 ms. Such fast burst repetitions and the digital signal acquisition allowed to lower the noise in the output signal by averaging the results of consecutive oscillation ring-downs. Exemplary output signal from the measurement system was presented in Fig. 3. It was the result of the averaging of 256 waveforms acquired during the test measurement of physical saline at room temperature. The output voltage signal V was converted into the displacement x of the QTF prongs using (1): x=
V kCVC · ˛ · ω
(1)
where ω is the radial frequency of the vibrations and ˛ is the electromechanical coupling coefficient which was measured for the QTFs used in the experiments using SIOS SP-S 120 laser vibrometer as 16.8 C/m. The distortion at the start of the ring-down phase was caused by the saturation of amplification stage of the measurement system caused by the discharge of the parasitic capacitance of the QTF in liquid and that part of the signal was omitted in further analysis. The parameters evaluated from ringdown signal were: the signal offset y0 , the initial amplitude of the oscillations A, the time constant of oscillation decay tc and the frequency fr and phase shift ϕr of free, damped oscillations. First three were evaluated basing on the analysis of the average of the envelope of sampled signal and exponential decay fit to the difference between upper and lower envelopes (Fig. 3). Such approach cancelled the influence of any possible offset or distortion caused by slow changing
Fig. 3. Exemplary output signal from measurement system and its analysis (a): sampled system response, evaluated envelopes, the simulated ringdown signal after the complete analysis and the magnitude of the noise (b).
T. Piasecki et al. / Sensors and Actuators B 210 (2015) 825–829
827
Fig. 4. Mean values (lines) and standard deviation (bands) of the measured parameters during the incubation in MHB alone (reference) and in bacterial cultures of four P. aeruginosa strains: resonant frequency shift (a), normalized time constant (b) and changes of the dissipation factor (c).
signals superimposed on the ring-down signal. To determine the frequency of the vibrations fr and their phase ϕr the raw output signal was divided by the value of the previously determined exponential decay to normalize the amplitude and then fitted using four parameter sine fit algorithm [21]. From these parameters the quality factor Q and the dissipation factor D could be calculated: Q =
1 = · · fr D
(2)
To estimate the quality of the measurement system the simulated ringdown signal sr : sr (t) = y0 + A · e−tc/t · sin(2fr + ϕr )
(3)
where t is the time, and the magnitude of the noise nr :
nr (i) = xi − sr (ti )
(4)
where i is the number of the sample and ti is the time of the ith sample, were evaluated. The signal to noise ratio was defined as the ratio between the RMS values of the sr and nr . In case of the exemplary measurement presented in Fig. 3 the RMS values of signal and noise were 0.617 nm, 6.59 pm, respectively, and the SNR was 39.4 dB. 2.2. Biofilm culture preparation and incubation Two Pseudomonas aeruginosa isolates: P. aeruginosa ATCC 27853 and P. aeruginosa ATCC 15692 (PA01) purchased from the American Type Culture Collection (ATCC) were used as model strains of biofilm-forming organisms. P. aeruginosa clinical strains CF 708 and CF 217 were isolated from patients with cystic fibrosis (COST action BM1003). Bacteria were stored in Trypticase Soy Broth (Becton Dickinson and Company, Cockeysville, MD, USA) supplemented with 20% glycerol at −70 ◦ C. Pseudomonas strains were cultured on Mueller Hinton Agar (Becton Dickinson and Company, Cockeysville, MD, USA) for 18–24 h prior each experiment and subsequently suspended in physiological saline (PS) to an optical density equal to the McFarland No. 0.5 (approx. 108 cfu/ml). The starting culture of the P. aeruginosa strain was added to 2.25 ml of Mueller Hinton Broth (MHB) in the wells of 24-well titrate plate to obtain final density of 106 cfu/ml. If smaller concentrations were necessary then the starting culture was diluted 10, 100, 1000, 10,000, and 100,000 folds before adding to the MHB in the titrate plate wells to obtain final densities of 105 cfu/ml, 104 cfu/ml, 103 cfu/ml, 102 cfu/ml and 101 cfu/ml, respectively.
QTFs used in experiments were cleaned in isopropyl alcohol, dried and dip-coated with insulating layer of polyester varnish (PMR/F, Sigma Coatings) and cured for 6 h at 150 ◦ C. They were sterilized with 96% ethyl alcohol, mounted in sockets in the lid of the titrate plate which was part of the measurement system described above. During the experiment the culture was incubated at 37 ◦ C. 3. Results 3.1. Culture of four P. aeruginosa strains The vibration ringdown parameters of QTF incubated in the reference (MHB alone) and 106 cfu/ml P. aeruginosa cultures were measured sequentially for each of the six channels during the experiment. Typical values of initial amplitude of vibrations, frequency and ringdown time constant were about 4 nm, 28,500 Hz and 1 ms, respectively. There were discrepancies between values obtained for QTFs placed in the same environment therefore only the changes of resonant frequency and dissipation factor relative to the value at the 1st hour of incubation and the time constant normalized to the value at the 1st hour of incubation were presented. The experiments for each bacterial strain were repeated twice. Obtained values for each experiment were averaged in 15-min periods. The mean value and standard deviations of resonant frequency shift, the normalized time constant and dissipation factor changes were presented in Fig. 4a, b and c, respectively. Despite quite large standard deviation both resonant frequency and time constant rose monotonically and stabilized after 6 h of incubation for the pure MHB. Rapid change of these parameters at the first hour of the experiment were caused by gradual change of the temperature of the culture which was prepared at room temperature and then incubated at 37 ◦ C. The culture growth of each bacterial strains influenced all measured parameters. The resonant frequency and the time constant of vibrations decay were lowered and the dissipation factor rose. The parameters which were the most convenient for distinguishing between the culture and reference were the time constant of decay and the dissipation factor. The most apparent changes of parameters were observed for the model strains P. aeruginosa ATCC 27853 and ATCC 15692, for which the time constant decreased over 10% between 12 and 24 h of incubation. This result was consistent with the research conducted before, where the QTFs were used as the dry mass and the impedance sensors [22,23]. In that period of time thick layer of adhered biofilm covered the QTF surface. Similar results were
828
T. Piasecki et al. / Sensors and Actuators B 210 (2015) 825–829
Fig. 5. Values of measured parameters during the incubation of P. aeruginosa ATCC 27853 strain with initial bacteria concentrations varying from 101 to 106 cfu/ml: resonant frequency shift (a), normalized time constant (b) and changes of the dissipation factor (c).
the peak position on the logarithm of the initial cfu/ml was linear as presented in Fig. 6. The peak shifted by 27.6 min each time the initial concentration changed twofold. This value is in agreement with the doubling time reported for the P. aeruginosa strains [24]. 4. Conclusions
Fig. 6. The dependence of the position of the negative peak of the time constant observed for the P. aeruginosa ATCC 27853 strain culture on the initial concentration of the colony forming unit.
observed for clinical strains but they occurred later for CF708 strain or were much slower for CF217 because CF isolates exhibit slower growth rate than reference ones. Apart from the gradual changes of parameters after 12 h of incubation or more also the negative peaks at resonant frequency and time constant and positive peaks at the loss factor were observed for each of the P. aeruginosa strains between 2 and 6 h of incubation. They were the most apparent for the ATCC 27853 and CF217 strains. The explanation of such behaviour requires further investigation. One of the hypothesis is that observed changes may be caused by the initial strong adhesion of bacterial cells to the surface of the QTF with subsequent formation of the extracellular matrix. Adhesion of the cells may initially lower the resonant frequency while the formation of more mechanically compliant extracellular matrix reduces the influence of the adhered cells on the QTF resonance parameters. 3.2. Culture of diluted ATCC 27853 strain Second experiment was conducted to verify the dependence of observed changes of QTF vibration parameters caused by the biofilm formation on the initial concentration of colony forming units in the culture. Results obtained at the first 12 h of incubation of P. aeruginosa ATCC 27853 strain initial concentrations varying from 101 to 106 cfu/ml were shown in Fig. 5. Result confirmed the appearance of peaks at all measured parameters, even at the start concentrations as low as 10 cfu/ml. The peaks were shifted in time depending on the initial concentration of colony forming units in the culture. The dependence of
Authors presented the method which allows to measure the vibration parameters of the commercial quartz tuning fork (QTF) with insulating coating placed in the liquid bacteria growth medium, which is highly conductive liquid. Method utilizes modulated actuation and the digital processing of the ring-down signal which allowed to determine the resonant frequency and the time constant of the vibration decay precisely with estimated resolution of single Hertz and 1%, respectively. QTFs were successfully used as in situ sensors of biofilm formation of four P. aeruginosa strains. The formation of the biofilm influenced measured resonant frequency and the time constant of vibration decay and calculated dissipation factor. Comparison between the reference (pure growth medium) and the experiment allows to detect the presence of biofilm formation as soon as after about 2 h of incubation for PA01 strain. Lower initial concentrations increase the detection time. The biofilm formation of 10 cfu/ml initial concentration ATCC 27853 strain was detected after 10 h. Presented method is convenient for in situ and on-line observation of the biofilm development and may be used for example in testing of the coatings or substances which should prevent the biofilm formation. Acknowledgements The research presented in this paper was supported by the project “Academy of Development as the key to strengthen human resources of the Polish economy”, co-financed by the European Union through the European Social Fund, Polish National Science Centre Grant, No. 2011/01/N/NZ7/00156 and the Wroclaw University of Technology Statutory Grant. The authors acknowledge COST (European Cooperation in the field of Scientific and Technical Research) action BM1003, “Microbial cell surface determinants of virulence as targets for new therapeutics in Cystic Fibrosis”. References [1] T. Mah, G.A. O’Toole, Mechanisms of biofilm resistance to antimicrobial agents, Trends Microbiol. 9 (2001) 34–39.
T. Piasecki et al. / Sensors and Actuators B 210 (2015) 825–829 [2] H.-C. Flemming, J. Wingender, The biofilm matrix, Nat. Rev. Microbiol. 8 (2010) 623–633. [3] C. Arciola, D. Campoccia, P. Speziale, L. Montanaro, J. Costerton, Biofilm formation in Staphylococcus implant infections: a review of molecular mechanisms and implications for biofilm-resistant materials, Biomaterials 33 (2012) 5967–5982. [4] H. Vlamakis, Y. Chai, P. Beauregard, R. Losick, R. Kolter, Sticking together: building a biofilm the Bacillus subtilis way, Nat. Rev. Microbiol. 11 (2013) 157–168. [5] O. Petrova, K. Sauer, Sticky situations: key components that control bacterial surface attachment, J. Bacteriol. 194 (2012) 2413–2425. [6] U. Romling, C. Balsalobre, Biofilm infections, their resilience to therapy and innovative treatment strategies, J. Intern. Med. 272 (2012) 541–561. [7] V. Kalia, Quorum sensing inhibitors: an overview, Biotechnol. Adv. 31 (2013) 224–245. [8] P. Stewart, Mini-review: convection around biofilms, Biofouling 28 (2012) 187–198. [9] W. Jones, M. Sutton, L. McKittrick, P. Stewart, Chemical and antimicrobial treatments change the viscoelastic properties of bacterial biofilms, Biofouling 27 (2011) 207–215. [10] L. Pavlovsky, J. Younger, M. Solomon, In situ rheology of Staphylococcus epidermidis bacterial biofilms, Soft Matter 9 (2013) 122–131. [11] A.L. Schofield, T.R. Rudd, D.S. Martin, D.G. Fernig, C. Edwards, Real-time monitoring of the development and stability of biofilms of Streptococcus mutans using the quartz crystal microbalance with dissipation monitoring, Biosens. Bioelectron. 23 (2007) 407–413. [12] I.M. Marcus, M. Herzberg, S.L. Walker, V. Freger, Pseudomonas aeruginosa attachment on QCM-D sensors: the role of cell and surface hydrophobicities, Langmuir 28 (15) (2012) 6396–6402. [13] Z. Ali, W. O’Hare, B. Theaker, Detection of bacterial contaminated milk by means of a quartz crystal microbalance based electronic nose, J. Therm. Anal. Calorim. 71 (2003) 355–362. [14] A. Janshoff, C. Steinem (Eds.), Piezoelectric Sensors, Springer Series on Chemical Sensors and Biosensors, Springer, Berlin, Heidelberg, New York, 2006. [15] T. Retsina, S.M. Richardson, W.A. Wakeham, The theory of a vibrating-rod viscometer, Appl. Sci. Res. 43 (1987) 325–346. [16] Y.-H. Cho, A. Pisano, R. Howe, Viscous damping model for laterally oscillating microstructures, J. Microelectromech. Syst. 3 (1994) 81–87. [17] K. Kapoor, V. Kanawade, V. Shukla, S. Patil, A new tuning fork-based instrument for oscillatory shear rheology of nano-confined liquids, Rev. Sci. Instrum. 84 (2013), 025101-025101-7. [18] A. Chałupniak, K. Waszczuk, K. Hałubek-Głuchowska, T. Piasecki, T. Gotszalk, J. Rybka, Application of quartz tuning forks for detection of endotoxins and Gramnegative bacterial cells by monitoring of limulus amebocyte lysate coagulation, Biosens. Bioelectron. 58 (2014) 132–137. [19] K. Waszczuk, T. Piasecki, K. Nitsch, T. Gotszalk, Application of piezoelectric tuning forks in liquid viscosity and density measurements, Sens. Actuators B: Chem. 160 (2011) 517–523. [20] K. Waszczuk, T. Piasecki, K. Nitsch, T. Gotszalk, Modulation of excitation signal as a method for measurement of the mechanical properties of miniature piezoelectric resonant devices in a viscous and conductive environment, Meas. Sci. Technol. 24 (2013) 085304. [21] IEEE Standard for Terminology and Test Methods for Analog-to-Digital Converters, IEEE Std 1241-2000. ´ atkowski, ˛ J. Olszewski, W. Herwich, Z. Drulis[22] K. Waszczuk, G. Guła, M. Swi Kawa, J. Gutowicz, T. Gotszalk, Evaluation of Pseudomonas aeruginosa biofilm
829
formation using piezoelectric tuning fork mass sensors, Sens. Actuators B: Chem. 170 (2012) 7–12. [23] T. Piasecki, G. Guła, K. Nitsch, K. Waszczuk, Z. Drulis-Kawa, T. Gotszalk, Evaluation of Pseudomonas aeruginosa biofilm formation using quartz tuning forks as impedance sensors, Sens. Actuators B: Chem. 189 (2013) 60–65. [24] B. Gottenbos, H.C. van der Mei, H.J. Busscher, Initial adhesion and surface growth of Staphylococcus epidermidis and Pseudomonas aeruginosa on biomedical polymers, J. Biomed. Mater. Res. 50 (2000) 208–214.
Biographies Tomasz Piasecki was born in Wrocław, Poland. He has electronic (MSc in 2000) and microelectronic (PhD in 2005) background. He is currently a lecturer at the Faculty of Microsystem Electronics and Photonics at the Wrocław University of Technology. His main interests are the electric methods in materials and devices characterization. He specializes in the impedance spectroscopy measurement, analysis and new methods development but also is interested in other types of electric measurement techniques. ´ ´ aska, ˛ Sl Poland. He graduated from the FacGrzegorz Guła was born in 1984 in Sroda ulty of Biology Science at University of Wrocław in 2008, receiving the MSc degree in biology with a specialization in microbiology. He received the PhD degree of biology from University of Wrocław, Faculty of Biology Science in 2012. He is a researcher of the Department of Pathogen’s Biology and Immunology in Institute of Genetics and Microbiology at University of Wrocław, Poland. Currently he is working on application of new methods of observation of bacterial biofilm formation. ´ aski, ˛ Karol Waszczuk was born in 1984 in Wodzisław Sl Poland. He graduated from the Faculty of Microsystems Electronics and Photonics of Wrocław University of Technology in 2007, receiving the MSc degree in electronics and telecommunications. Since then, he has been working at Research and Development Department for EBS Inkjet Systems Poland. He is pursuing his PhD degree at Wrocław University of Technology, working on biochemical and microbiological applications of resonant mass sensors. Zuzanna Drulis-Kawa was born in Wrocław, Poland. She received PhD degree of biology, University of Wrocław in 2000. She received first and second clinical microbiology degrees at Medical University of Wrocław in 1997 and 2000, respectively. She has been honoured with bioMerieux-PTM Research Award (2009), the prize of Prof. K. Bassalik Contest (2006) for her scientific work and Team Award of the Minister of Science and Higher Education for teaching achievements in 2013. She is Head of the Department of Pathogen Biology and Immunology, Institute of Genetics and Microbiology, University of Wrocław, Poland. She has authored over 60 scientific publications. Teodor Gotszalk was born in Wrocław, Poland. He received the MSc degree from the Faculties of Electronics and of Electrical Engineering of Wrocław University of Technology in 1989 and 1991, respectively. In 1996, he received the PhD degree from the Institute of Electronic Technology of the Wrocław University of Technology. He has been honored with Siemens Research Award (2000) and the prize of Polish Science Foundation FNP (1997) for his scientific work. He is the head of the Division of Micro- and Nanostructures Metrology at the Faculty of Microsystems Electronics and Photonics of Wrocław University of Technology. He has authored over 100 scientific publications.
Contents lists available at ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Quartz tuning fork as in situ sensor of bacterial biofilm夽 Tomasz Piasecki a,∗ , Grzegorz Guła b , Karol Waszczuk a , Zuzanna Drulis-Kawa b , Teodor Gotszalk a a b
Faculty of Microsystem Electronics and Photonics, Wrocław University of Technology, Z. Janiszewskiego 11/17, 50-372 Wrocław, Poland Institute of Genetics and Microbiology, University of Wrocław, S. Przybyszewskiego 63/77, 51-148 Wrocław, Poland
a r t i c l e
i n f o
Article history: Available online 8 January 2015 Keywords: Pseudomonas aeruginosa Ringdown Resonant frequency Dissipation factor Viscosity Density
a b s t r a c t Piezoelectric quartz tuning forks (QTFs) were used as in situ sensors of biofilm formation. Modulated excitation technique allowed to measure the current generated by the QTF during the ring-down of the oscillations at conductive environments such as a liquid culture growth medium. New algorithm of the digital analysis of the ring-down signal was elaborated and used to determine the decay time constant of the vibrations and the damped vibration frequency. The dependence of these parameters on the time for QTF placed in liquid growth medium contained four strains of Pseudomonas aeruginosa was measured. It allowed to detect the biofilm formation at the QTF after few hours of incubation for all used strains at 106 cfu/ml initial concentration (2 h for PA01 strain) or 10 h for ATCC 27853 strain at 10 cfu/ml initial concentration. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The bacterial biofilm represents the most prevalent type of microbial growth. The extracellular matrix in which biofilm cells are embedded cause the resistance to antibiotics or immune defense [1,2] and good adhesion to the biotic and abiotic surfaces [3–5]. Microorganisms that produce extracellular matrix could colonize indwelling medical devices to form a biofilm. The deviceassociated infections result in substantial morbidity and mortality, and potentially represent a large increase the treatment cost [2,6]. There is a need for fast, preferably in situ method which would allow to determine whether the biofilm formation was prevented or to study the influence of the environment on the biofilm properties [7,8]. Such measurements may be performed using vibrating viscometers. They change their properties depending on the mass and rheology of the bacterial biofilm which adheres to the surface of the resonator [9,10]. The most common type of piezoelectric element used for biofilm detection is the quartz crystal microbalance (QCM) [11–13]. It is established method which is based by the observation
夽 Selected papers presented at EUROSENSORS 2014, the XXVIII edition of the conference series, Brescia, Italy, September 7–10, 2014. ∗ Corresponding author. Tel.: +48 713203223. E-mail address: [email protected] (T. Piasecki). http://dx.doi.org/10.1016/j.snb.2014.12.105 0925-4005/© 2015 Elsevier B.V. All rights reserved.
of the changes of resonant frequency during the incubation of the biofilm or additionally monitoring of the dissipation of the vibration energy in the QCM-D variant of that technique [14]. It vibrates in shear mode and typical resonant frequency is few MHz or higher. In this paper another type of resonator is used: the quartz tuning forks (QTFs). Commercially available QTFs are inexpensive and mass produced. Their resonant frequency is 215 Hz (about 32.8 kHz). They are commonly used as the time reference in real time clocks however they also find another applications. One of the alternative application of the QTF are the viscosity and mass density measurements. The properties of the fluid that surrounds the vibrating element change the resonant frequency and the damping of the vibrations [15–19]. Therefore it should be possible to detect the influence of viscous biofilm adhered to its surface. Commercial QTFs have the electrodes fabricated at the outer surface of the piezoelectric crystal. If QTF is immersed in conductive liquids, such as the liquid growth media used in the culture growth, the parasitic currents will flow when voltage is applied to the QTF. It renders common electrical methods unsuitable for determination of the vibrations of commercial QTF placed in conductive liquids. Special technique based on modulated excitation was elaborated. It allowed for direct measurement of the current generated by short-circuited, vibrating QTF during the ring-down phase of the vibrations [20]. Its further development and application in detection of biofilm formation is presented in this paper.
826
T. Piasecki et al. / Sensors and Actuators B 210 (2015) 825–829
QTF1
actuation
response
QTF2 Rt
... QTF6
400x
TRIG
OUT SYNC
DAQ Agilent U2531A
Agilent 33220A Fig. 1. The schematic diagram of the QTF oscillation ringdown measurement system.
2. Materials and methods 2.1. QTF oscillations ringdown measurement system Measurements of QTF oscillations ringdown were performed using computer controlled home-built system which utilized the measurement technique based on the modulated excitation of the QTF. In this method the vibrations of the resonator are excited by the sinusoidal voltage burst after which the resonator is short circuited and the current generated during the ring-down of the oscillations due to the direct piezoelectric effect is measured. The details of this technique were already published [20]. It allows to excite and measure the vibrations of the QTF covered with thin insulating layer despite the influence of the parasitic capacitance and the conductance of the liquid in which the QTF is immersed. The further development of that technique by the introduction of digital processing of the output signal is presented. The schematic diagram of the measurement system was shown in Fig. 1. It consisted of the Agilent 33220A function generator, home-built excitation and measurement electronics and Agilent U2531A DAQ module. The function generator provided the excitation and synchronization signals. The latter was used to synchronize all components of the measurement system with the burst and measurement phases. The home-built circuitry (Fig. 2) provided the excitation signal conditioning and splitting it into six channels which were switched electronically. It allowed to measure the vibrations of up to six QTFs placed in the six wells of 24-well titrate plate, switched sequentially. Current generated during the oscillation ring-down was measured using the transimpedance amplifier which output was additionally amplified 400 times. The effective conversion ratio of the system was kCVC = 0.4 V/nA. Output signal was sampled using Agilent U2531A 2MSa/s, 14 bit data acquisition module. The parameters of the excitation signal, channel switching and data acquisition
Fig. 2. The photograph of the QTF oscillation measurement system home-built actuation and readout electronics connected to a QTF holder on a titrate plate.
were controlled using dedicated software at the personal computer. Typical measurement parameters were as follows: the excitation signal amplitude – 0.25 V and frequency at the resonance (around 28,500 Hz), burst length – 500 cycles, burst period – 35 ms. Such fast burst repetitions and the digital signal acquisition allowed to lower the noise in the output signal by averaging the results of consecutive oscillation ring-downs. Exemplary output signal from the measurement system was presented in Fig. 3. It was the result of the averaging of 256 waveforms acquired during the test measurement of physical saline at room temperature. The output voltage signal V was converted into the displacement x of the QTF prongs using (1): x=
V kCVC · ˛ · ω
(1)
where ω is the radial frequency of the vibrations and ˛ is the electromechanical coupling coefficient which was measured for the QTFs used in the experiments using SIOS SP-S 120 laser vibrometer as 16.8 C/m. The distortion at the start of the ring-down phase was caused by the saturation of amplification stage of the measurement system caused by the discharge of the parasitic capacitance of the QTF in liquid and that part of the signal was omitted in further analysis. The parameters evaluated from ringdown signal were: the signal offset y0 , the initial amplitude of the oscillations A, the time constant of oscillation decay tc and the frequency fr and phase shift ϕr of free, damped oscillations. First three were evaluated basing on the analysis of the average of the envelope of sampled signal and exponential decay fit to the difference between upper and lower envelopes (Fig. 3). Such approach cancelled the influence of any possible offset or distortion caused by slow changing
Fig. 3. Exemplary output signal from measurement system and its analysis (a): sampled system response, evaluated envelopes, the simulated ringdown signal after the complete analysis and the magnitude of the noise (b).
T. Piasecki et al. / Sensors and Actuators B 210 (2015) 825–829
827
Fig. 4. Mean values (lines) and standard deviation (bands) of the measured parameters during the incubation in MHB alone (reference) and in bacterial cultures of four P. aeruginosa strains: resonant frequency shift (a), normalized time constant (b) and changes of the dissipation factor (c).
signals superimposed on the ring-down signal. To determine the frequency of the vibrations fr and their phase ϕr the raw output signal was divided by the value of the previously determined exponential decay to normalize the amplitude and then fitted using four parameter sine fit algorithm [21]. From these parameters the quality factor Q and the dissipation factor D could be calculated: Q =
1 = · · fr D
(2)
To estimate the quality of the measurement system the simulated ringdown signal sr : sr (t) = y0 + A · e−tc/t · sin(2fr + ϕr )
(3)
where t is the time, and the magnitude of the noise nr :
nr (i) = xi − sr (ti )
(4)
where i is the number of the sample and ti is the time of the ith sample, were evaluated. The signal to noise ratio was defined as the ratio between the RMS values of the sr and nr . In case of the exemplary measurement presented in Fig. 3 the RMS values of signal and noise were 0.617 nm, 6.59 pm, respectively, and the SNR was 39.4 dB. 2.2. Biofilm culture preparation and incubation Two Pseudomonas aeruginosa isolates: P. aeruginosa ATCC 27853 and P. aeruginosa ATCC 15692 (PA01) purchased from the American Type Culture Collection (ATCC) were used as model strains of biofilm-forming organisms. P. aeruginosa clinical strains CF 708 and CF 217 were isolated from patients with cystic fibrosis (COST action BM1003). Bacteria were stored in Trypticase Soy Broth (Becton Dickinson and Company, Cockeysville, MD, USA) supplemented with 20% glycerol at −70 ◦ C. Pseudomonas strains were cultured on Mueller Hinton Agar (Becton Dickinson and Company, Cockeysville, MD, USA) for 18–24 h prior each experiment and subsequently suspended in physiological saline (PS) to an optical density equal to the McFarland No. 0.5 (approx. 108 cfu/ml). The starting culture of the P. aeruginosa strain was added to 2.25 ml of Mueller Hinton Broth (MHB) in the wells of 24-well titrate plate to obtain final density of 106 cfu/ml. If smaller concentrations were necessary then the starting culture was diluted 10, 100, 1000, 10,000, and 100,000 folds before adding to the MHB in the titrate plate wells to obtain final densities of 105 cfu/ml, 104 cfu/ml, 103 cfu/ml, 102 cfu/ml and 101 cfu/ml, respectively.
QTFs used in experiments were cleaned in isopropyl alcohol, dried and dip-coated with insulating layer of polyester varnish (PMR/F, Sigma Coatings) and cured for 6 h at 150 ◦ C. They were sterilized with 96% ethyl alcohol, mounted in sockets in the lid of the titrate plate which was part of the measurement system described above. During the experiment the culture was incubated at 37 ◦ C. 3. Results 3.1. Culture of four P. aeruginosa strains The vibration ringdown parameters of QTF incubated in the reference (MHB alone) and 106 cfu/ml P. aeruginosa cultures were measured sequentially for each of the six channels during the experiment. Typical values of initial amplitude of vibrations, frequency and ringdown time constant were about 4 nm, 28,500 Hz and 1 ms, respectively. There were discrepancies between values obtained for QTFs placed in the same environment therefore only the changes of resonant frequency and dissipation factor relative to the value at the 1st hour of incubation and the time constant normalized to the value at the 1st hour of incubation were presented. The experiments for each bacterial strain were repeated twice. Obtained values for each experiment were averaged in 15-min periods. The mean value and standard deviations of resonant frequency shift, the normalized time constant and dissipation factor changes were presented in Fig. 4a, b and c, respectively. Despite quite large standard deviation both resonant frequency and time constant rose monotonically and stabilized after 6 h of incubation for the pure MHB. Rapid change of these parameters at the first hour of the experiment were caused by gradual change of the temperature of the culture which was prepared at room temperature and then incubated at 37 ◦ C. The culture growth of each bacterial strains influenced all measured parameters. The resonant frequency and the time constant of vibrations decay were lowered and the dissipation factor rose. The parameters which were the most convenient for distinguishing between the culture and reference were the time constant of decay and the dissipation factor. The most apparent changes of parameters were observed for the model strains P. aeruginosa ATCC 27853 and ATCC 15692, for which the time constant decreased over 10% between 12 and 24 h of incubation. This result was consistent with the research conducted before, where the QTFs were used as the dry mass and the impedance sensors [22,23]. In that period of time thick layer of adhered biofilm covered the QTF surface. Similar results were
828
T. Piasecki et al. / Sensors and Actuators B 210 (2015) 825–829
Fig. 5. Values of measured parameters during the incubation of P. aeruginosa ATCC 27853 strain with initial bacteria concentrations varying from 101 to 106 cfu/ml: resonant frequency shift (a), normalized time constant (b) and changes of the dissipation factor (c).
the peak position on the logarithm of the initial cfu/ml was linear as presented in Fig. 6. The peak shifted by 27.6 min each time the initial concentration changed twofold. This value is in agreement with the doubling time reported for the P. aeruginosa strains [24]. 4. Conclusions
Fig. 6. The dependence of the position of the negative peak of the time constant observed for the P. aeruginosa ATCC 27853 strain culture on the initial concentration of the colony forming unit.
observed for clinical strains but they occurred later for CF708 strain or were much slower for CF217 because CF isolates exhibit slower growth rate than reference ones. Apart from the gradual changes of parameters after 12 h of incubation or more also the negative peaks at resonant frequency and time constant and positive peaks at the loss factor were observed for each of the P. aeruginosa strains between 2 and 6 h of incubation. They were the most apparent for the ATCC 27853 and CF217 strains. The explanation of such behaviour requires further investigation. One of the hypothesis is that observed changes may be caused by the initial strong adhesion of bacterial cells to the surface of the QTF with subsequent formation of the extracellular matrix. Adhesion of the cells may initially lower the resonant frequency while the formation of more mechanically compliant extracellular matrix reduces the influence of the adhered cells on the QTF resonance parameters. 3.2. Culture of diluted ATCC 27853 strain Second experiment was conducted to verify the dependence of observed changes of QTF vibration parameters caused by the biofilm formation on the initial concentration of colony forming units in the culture. Results obtained at the first 12 h of incubation of P. aeruginosa ATCC 27853 strain initial concentrations varying from 101 to 106 cfu/ml were shown in Fig. 5. Result confirmed the appearance of peaks at all measured parameters, even at the start concentrations as low as 10 cfu/ml. The peaks were shifted in time depending on the initial concentration of colony forming units in the culture. The dependence of
Authors presented the method which allows to measure the vibration parameters of the commercial quartz tuning fork (QTF) with insulating coating placed in the liquid bacteria growth medium, which is highly conductive liquid. Method utilizes modulated actuation and the digital processing of the ring-down signal which allowed to determine the resonant frequency and the time constant of the vibration decay precisely with estimated resolution of single Hertz and 1%, respectively. QTFs were successfully used as in situ sensors of biofilm formation of four P. aeruginosa strains. The formation of the biofilm influenced measured resonant frequency and the time constant of vibration decay and calculated dissipation factor. Comparison between the reference (pure growth medium) and the experiment allows to detect the presence of biofilm formation as soon as after about 2 h of incubation for PA01 strain. Lower initial concentrations increase the detection time. The biofilm formation of 10 cfu/ml initial concentration ATCC 27853 strain was detected after 10 h. Presented method is convenient for in situ and on-line observation of the biofilm development and may be used for example in testing of the coatings or substances which should prevent the biofilm formation. Acknowledgements The research presented in this paper was supported by the project “Academy of Development as the key to strengthen human resources of the Polish economy”, co-financed by the European Union through the European Social Fund, Polish National Science Centre Grant, No. 2011/01/N/NZ7/00156 and the Wroclaw University of Technology Statutory Grant. The authors acknowledge COST (European Cooperation in the field of Scientific and Technical Research) action BM1003, “Microbial cell surface determinants of virulence as targets for new therapeutics in Cystic Fibrosis”. References [1] T. Mah, G.A. O’Toole, Mechanisms of biofilm resistance to antimicrobial agents, Trends Microbiol. 9 (2001) 34–39.
T. Piasecki et al. / Sensors and Actuators B 210 (2015) 825–829 [2] H.-C. Flemming, J. Wingender, The biofilm matrix, Nat. Rev. Microbiol. 8 (2010) 623–633. [3] C. Arciola, D. Campoccia, P. Speziale, L. Montanaro, J. Costerton, Biofilm formation in Staphylococcus implant infections: a review of molecular mechanisms and implications for biofilm-resistant materials, Biomaterials 33 (2012) 5967–5982. [4] H. Vlamakis, Y. Chai, P. Beauregard, R. Losick, R. Kolter, Sticking together: building a biofilm the Bacillus subtilis way, Nat. Rev. Microbiol. 11 (2013) 157–168. [5] O. Petrova, K. Sauer, Sticky situations: key components that control bacterial surface attachment, J. Bacteriol. 194 (2012) 2413–2425. [6] U. Romling, C. Balsalobre, Biofilm infections, their resilience to therapy and innovative treatment strategies, J. Intern. Med. 272 (2012) 541–561. [7] V. Kalia, Quorum sensing inhibitors: an overview, Biotechnol. Adv. 31 (2013) 224–245. [8] P. Stewart, Mini-review: convection around biofilms, Biofouling 28 (2012) 187–198. [9] W. Jones, M. Sutton, L. McKittrick, P. Stewart, Chemical and antimicrobial treatments change the viscoelastic properties of bacterial biofilms, Biofouling 27 (2011) 207–215. [10] L. Pavlovsky, J. Younger, M. Solomon, In situ rheology of Staphylococcus epidermidis bacterial biofilms, Soft Matter 9 (2013) 122–131. [11] A.L. Schofield, T.R. Rudd, D.S. Martin, D.G. Fernig, C. Edwards, Real-time monitoring of the development and stability of biofilms of Streptococcus mutans using the quartz crystal microbalance with dissipation monitoring, Biosens. Bioelectron. 23 (2007) 407–413. [12] I.M. Marcus, M. Herzberg, S.L. Walker, V. Freger, Pseudomonas aeruginosa attachment on QCM-D sensors: the role of cell and surface hydrophobicities, Langmuir 28 (15) (2012) 6396–6402. [13] Z. Ali, W. O’Hare, B. Theaker, Detection of bacterial contaminated milk by means of a quartz crystal microbalance based electronic nose, J. Therm. Anal. Calorim. 71 (2003) 355–362. [14] A. Janshoff, C. Steinem (Eds.), Piezoelectric Sensors, Springer Series on Chemical Sensors and Biosensors, Springer, Berlin, Heidelberg, New York, 2006. [15] T. Retsina, S.M. Richardson, W.A. Wakeham, The theory of a vibrating-rod viscometer, Appl. Sci. Res. 43 (1987) 325–346. [16] Y.-H. Cho, A. Pisano, R. Howe, Viscous damping model for laterally oscillating microstructures, J. Microelectromech. Syst. 3 (1994) 81–87. [17] K. Kapoor, V. Kanawade, V. Shukla, S. Patil, A new tuning fork-based instrument for oscillatory shear rheology of nano-confined liquids, Rev. Sci. Instrum. 84 (2013), 025101-025101-7. [18] A. Chałupniak, K. Waszczuk, K. Hałubek-Głuchowska, T. Piasecki, T. Gotszalk, J. Rybka, Application of quartz tuning forks for detection of endotoxins and Gramnegative bacterial cells by monitoring of limulus amebocyte lysate coagulation, Biosens. Bioelectron. 58 (2014) 132–137. [19] K. Waszczuk, T. Piasecki, K. Nitsch, T. Gotszalk, Application of piezoelectric tuning forks in liquid viscosity and density measurements, Sens. Actuators B: Chem. 160 (2011) 517–523. [20] K. Waszczuk, T. Piasecki, K. Nitsch, T. Gotszalk, Modulation of excitation signal as a method for measurement of the mechanical properties of miniature piezoelectric resonant devices in a viscous and conductive environment, Meas. Sci. Technol. 24 (2013) 085304. [21] IEEE Standard for Terminology and Test Methods for Analog-to-Digital Converters, IEEE Std 1241-2000. ´ atkowski, ˛ J. Olszewski, W. Herwich, Z. Drulis[22] K. Waszczuk, G. Guła, M. Swi Kawa, J. Gutowicz, T. Gotszalk, Evaluation of Pseudomonas aeruginosa biofilm
829
formation using piezoelectric tuning fork mass sensors, Sens. Actuators B: Chem. 170 (2012) 7–12. [23] T. Piasecki, G. Guła, K. Nitsch, K. Waszczuk, Z. Drulis-Kawa, T. Gotszalk, Evaluation of Pseudomonas aeruginosa biofilm formation using quartz tuning forks as impedance sensors, Sens. Actuators B: Chem. 189 (2013) 60–65. [24] B. Gottenbos, H.C. van der Mei, H.J. Busscher, Initial adhesion and surface growth of Staphylococcus epidermidis and Pseudomonas aeruginosa on biomedical polymers, J. Biomed. Mater. Res. 50 (2000) 208–214.
Biographies Tomasz Piasecki was born in Wrocław, Poland. He has electronic (MSc in 2000) and microelectronic (PhD in 2005) background. He is currently a lecturer at the Faculty of Microsystem Electronics and Photonics at the Wrocław University of Technology. His main interests are the electric methods in materials and devices characterization. He specializes in the impedance spectroscopy measurement, analysis and new methods development but also is interested in other types of electric measurement techniques. ´ ´ aska, ˛ Sl Poland. He graduated from the FacGrzegorz Guła was born in 1984 in Sroda ulty of Biology Science at University of Wrocław in 2008, receiving the MSc degree in biology with a specialization in microbiology. He received the PhD degree of biology from University of Wrocław, Faculty of Biology Science in 2012. He is a researcher of the Department of Pathogen’s Biology and Immunology in Institute of Genetics and Microbiology at University of Wrocław, Poland. Currently he is working on application of new methods of observation of bacterial biofilm formation. ´ aski, ˛ Karol Waszczuk was born in 1984 in Wodzisław Sl Poland. He graduated from the Faculty of Microsystems Electronics and Photonics of Wrocław University of Technology in 2007, receiving the MSc degree in electronics and telecommunications. Since then, he has been working at Research and Development Department for EBS Inkjet Systems Poland. He is pursuing his PhD degree at Wrocław University of Technology, working on biochemical and microbiological applications of resonant mass sensors. Zuzanna Drulis-Kawa was born in Wrocław, Poland. She received PhD degree of biology, University of Wrocław in 2000. She received first and second clinical microbiology degrees at Medical University of Wrocław in 1997 and 2000, respectively. She has been honoured with bioMerieux-PTM Research Award (2009), the prize of Prof. K. Bassalik Contest (2006) for her scientific work and Team Award of the Minister of Science and Higher Education for teaching achievements in 2013. She is Head of the Department of Pathogen Biology and Immunology, Institute of Genetics and Microbiology, University of Wrocław, Poland. She has authored over 60 scientific publications. Teodor Gotszalk was born in Wrocław, Poland. He received the MSc degree from the Faculties of Electronics and of Electrical Engineering of Wrocław University of Technology in 1989 and 1991, respectively. In 1996, he received the PhD degree from the Institute of Electronic Technology of the Wrocław University of Technology. He has been honored with Siemens Research Award (2000) and the prize of Polish Science Foundation FNP (1997) for his scientific work. He is the head of the Division of Micro- and Nanostructures Metrology at the Faculty of Microsystems Electronics and Photonics of Wrocław University of Technology. He has authored over 100 scientific publications.