Effect of static magnetic field on growth of Escherichia coli and relative response model of series piezoelectric quartz crystal Shufen Zhang, Wanzhi Wei,* Jinzhong Zhang, Youan Mao and Shujuan Liu THE LYST FULL PPER College of Chemistry and Chemical Engineering, Hunan University, Changsha 410082, P.R. China. E-mail: shufenzhang@hotmail.com; Fax: 0086 731 8824704 www.rsc.org/analyst Received 22nd October 2001, ccepted 4th December 2001 First published as an dvance rticle on the web 31st January 2002 The effect of magnetic field on the growth of bacteria was studied with the series piezoelectric quartz crystal (SPQC) sensing technique. The growth situations of Escherichia coli (E. coli) in the absence and presence of different intensities of static magnetic fields were examined and analyzed. The results showed that the growth of E. coli was inhibited due to the presence of magnetic fields. By fitting frequency shift (Df) versus time curves according to the frequency shift response equation of SPQC, the relationships between three kinetic growth parameters, i.e., the asymptote, the maximum specific growth rate m m and lag time l, and magnetic field intensity were established. Based on these results, a new response model containing the magnetic field intensity was derived as: Df = 167. 7-397B 4 2 46 { 1+ exp[ (.. e ) ( 442. + 1646. B- t) + 2]} The kinetic parameters of bacterial growth obtained from this model are close to those obtained from the logistics popular growth model, in which the concentration of the bacteria was determined by the traditional pour plate count method. Introduction Coliform bacteria are an important type of Gram-negative bacteria in human and animal intestines. Some types of coliform exhibit pathogenicity to the human body. Moreover, they have powerful progenitive ability. Therefore, it is necessary to study coliform bacteria for clinical medicine, food hygiene and environmental science. With increasing public interest in the possible impacts of magnetic fields on man and the environment, a large number of papers have appeared concerned with almost all biological systems including simple cells, 1,2 microorganisms, 3 5 lower animals, 6 and man. 7 Magnetic fields may enhance the stationary-phase-specific transcription activity of bacteria, 3 they may affect the phagocytic cells in blood and the activity of the rpos gene of E. coli. 4 lso, they could be used to protect against subsequent hypoxia insult. 8 Hence it is very important to study the effects of magnetic fields on living things. Traditional methods for detecting bacteria are the most probable number (MP), the pour plate count (PPC) 9 and turbidimetric methods. The MP and PPC techniques are relatively cumbersome, time-consuming and can not provide the real time information of growth situation. With respect to the turbidimetric technique, 10 because of the detection range of the spectrophotometer, the concentration range of the bacterial suspension that can be used is very limited. 11 He et al., 12 applied a series piezoelectric quartz crystal (SPQC) sensing technique to the rapid detection of bacteria, providing a new method for studying the growth situation of microorganisms. s a highly sensitive chemical and biological sensor, 13 the SPQC technique has been extensively used in various fields; such as rapid detection of antibiotics 14,15 and microorganisms, 16 detection of the biochemical oxygen demand in the environment, 17 and the study of the hemorheological characteristics of plasma and fermentation processes. 18 However, to our knowledge, few reports about the application of the SPQC sensing technique to the effects of biomagnetism have been published. In this work, the SPQC sensing technique combined with the growth situation of bacteria was used to study the effect of a static magnetic field on the growth of E. coli. Experimental Reagents Beef extract, yeast extract and peptone were supplied by the Shanghai Biochemicals Company (China). ll chemicals used were analytical reagent grade purchased from Shanghai Chemical Regents Company (China). Doubly distilled and sterilized water was used throughout. The culture medium for E. coli was as follows: beef extract, 3 g; peptone, 10 g; yeast extract, 6 g; sodium chloride, 5 g; distilled water, 1000 ml. The ph value of medium was adjusted to 7.4 by phosphate buffer solution. The medium was sterilized by autoclaving at 121 C for 30 min. pparatus The experimental set-up employed is shown in Fig. 1. The SPQC sensor is constructed with an T-cut quartz crystal with a conductivity cell in series, in which a sensitive frequency shift response to conductivity of the solution can be reached. The quartz crystal used in this work is an T-cut 9 M Hz crystal with 12.5 mm diameter which is produced by state-run o. 707 factory (Beijing, China). The electrode consisted of two pieces of platinum plate; the cell constant was 1.04 cm. Model SS3341 universal frequency counter at a resolution of 1 Hz (Shijiazhuang, China) was employed to measure the frequency output. thermostatic water-jacket was used to control the DOI: 10.1039/b109617f nalyst, 2002, 127, 373 377 373 This journal is The Royal Society of Chemistry 2002
culture temperature at 37 C. Ring magnet (inner diameter 7.5 cm, outer diameter 15 cm, thickness 3.5 cm) was purchased from the center of magnet produced (Changsha, China), of which the maximum magnetic field intensity was 0.6 T. The magnetic field intensity acting on the detection system was measured with a CT 5 Hoare Gaussmeter. Bacteria The E. coli used in the experiments was E. coli DH 5a, which was obtained from the College of Life Science of Hunan ormal University (Changsha, China). Four loopfulls of an E. coli culture from an agar slant were used to inoculate 250 ml of culture medium in 500 ml sterilized conical flasks. It was incubated for 16 h at 37 C in an incubator, then preserved in a 4 C refrigerator and was used within 1 week. The culture gives an approximate concentration of 7.6 3 10 7 cells ml 21 by the PPC method. Procedures In the detection of the growth curve, 5 ml of stock culture E. coli solution and 10 ml of medium were added to a 20 ml detection cell and mixed thoroughly. Then the conductivity electrode was immersed in this. The detection cell was stuffed with a rubber plug to isolate any air and placed in a thermostatic water-jacket at 37 C. ring magnet was placed under the water-jacket to make the magnetic line of force pass through the detection cell vertically. Then, the growth of the bacterium was monitored by SPQC sensor over a certain time interval. The detection data was collected every 10 min, which was measured through a frequency counter. Results and discussion Theory Metabolizing bacteria can transform uncharged or weakly charged substrates into highly charged products, such as proteins (which decompose into amino acids), lipids (to acetate) and carbohydrates (to lactate). These charged products cause an alternation of the conductivity of the culture medium. Thus, the growth characteristic of the bacteria can be reflected through the conductivity variation of the detection system. However, the traditional conduction method is not sufficiently sensitive when a high background conductivity exists in the culture medium. The traditional piezoelectric quartz crystal (PQC) sensor can respond to not only the mass changes at the surface of the electrode, but also to various physical properties of the solution, such as conductivity, permittivity, viscosity and density. 19 n SPQC sensor was constructed by connecting an T-cut PQC and a conductivity electrode in series. Because the crystal locates either in vacuums or in the gas-phase, which are out of contact with the solution, the conditions of the crystal are independent of the properties of the solution and the frequency of the SPQC depends only on the permittivity and conductivity of the solution. s the permittivity hardly changes with conductivity, the SPQC sensor has a sensitive frequency shift (Df) response to the conductivity in the electrolyte solutions. 20 The frequency shift is linearly related to the change in conductivity over some ranges. 21 Especially, when a high background conductivity exists in the detection system, the sensitivity and accuracy are higher than those obtained with a low conductivity background. 22 Therefore, the SPQC sensing technique can be used to study the growth of bacteria in different situations. The change of conductivity (DG) and the number of bacteria were determined in order to establish the relationship between them. Experiments showed that a linear relationship exists between DG and the logarithm of the relative population size. DG = k ln 1 (1) 0 where k 1 is a coefficient, both DG and k 1 have dimensions ms, 0 is the initial bacterial number and is the bacterial number at time t.ccording to ref. 21, the following linear relationship between DG and the frequency shift response of the SPQC sensor is valid over a certain range. Ωf = k 2 ΩG (2) where Df is defined as the difference between f 1 and f 2 (Df = f 1 f 2 ). f 1 is the initial frequency and f 2 is the frequency at time t; k 2 is a coefficient with dimensions Hz ms 21. combination of eqns. (1) and (2) gives the following expression: Df = k1k2ln = K ln (3) 0 0 where K is defined as the frequency shift response coefficient with dimensions Hz. typical bacterial growth curve is well described by the logistics popular growth model according to ref. 23: ln = 4m (4) 0 m { 1+ exp[ ( l - t) + 2]} where the asymptote (with no dimensions) is the relative maximum value of ln/ 0 ; the maximum specific growth rate, m m in h 21 is defined as the tangent at the inflection point; the lag time, l in h 21 is defined as the x-axis intercept of this tangent. Each of these parameters has a specific biological meaning, and each happens to change with a change in the growth situation of the microorganism. From eqns. (3) and (4), the frequency shift response equation can be obtained as: Df = K 4m (5) m { 1+ exp[ ( l - t) + 2]} This equation can be used to describe the growth process of bacteria monitored by the SPQC sensing technique. Typical Df response curve at normal growth Fig. 1 Experimental set-up diagram of the SPQC system. 1, Frequency counter; 2, TTL-IC oscillator; 3, quartz crystal; 4, rubber plug; 5, thermostatic water outlet; 6, conductivity electrodes; 7, detection cell; 8, detection solution; 9, thermostatic water-jacket; 10, ring magnet; 11, thermostatic water inlet. typical response curve of frequency shift versus time is shown in Fig. 2a. It can be seen that Df does not change noticeably during the initial ~ 4.5 h, to form the first plateau. Then the frequency shift increases continuously for about 3 h, and then reaches a stable level to form the second plateau. The above response curve can be expected from the point of the bacterial 374 nalyst, 2002, 127, 373 377
growth theory. Since bacterial growth lies in the lag phase, the growth of bacteria is very slow over the initial time and little nutrient is transformed into highly charged products. Hence, the growth of E. coli leads to little change of frequency and the first plateau appears. When the growth of E. coli enters a logarithmic growth period, the growth of the bacteria consumes large amounts of nutrient leading to a quick increase in conductivity, which produces an increase of the frequency. Since there is a limit of the amount of nutrients and other factors, the growth of E. coli reaches saturation phase and the bacteria grow slowly or even do not grow any more. s a result, Df does not change noticeably any more and the second plateau appears. The above frequency shift response curves indicated that the shape of Df is a sigmoidal curve just like the theoretical growth curve of bacteria described by the logistics model. The first plateau phase shown in Fig. 2 corresponds to the lag phase of the growth of bacteria, the quickly increasing phase to the exponential growth phase and the second plateau phase to the bacterial growth saturation phase. The frequency shift response of the medium without bacteria was examined and the response curve is shown in Fig. 2b. During the monitoring under our experimental conditions, it was found that the frequency shift response had no obvious change. Typical Df response curve in the presence of a static magnetic field Fig. 3b shows a typical response curve for the frequency shift that indicates the effect of a 0.2 T magnetic field on a medium Fig. 2 a, Typical frequency shift response curve during the normal growth of E. coli b, Typical frequency shift response curve of medium. without bacteria. It can be seen that the frequency shift decreases continuously for the initial ~ 13 h. fter that, the Df response remains almost constant. This phenomenon basically agrees with that in ref. 24 where the physical and chemical properties of magnetized water were investigated. The conductivity of an electrolyte solution is related to the ionic concentration, the number of ionic charge and the ionic mobility. 25 So the frequency shift of the sensor also depends on these factors. During monitoring, because no biological process happened, the ionic concentration and the number of the ionic charge did not change. Thus, the decrease of the frequency shift may be caused by the effect of the magnetic field on the ionic mobility. fter that, the Df response remains almost constant, which may indicate that the magnetization of the magnetic field on the medium achieved saturation. For the medium containing bacteria, the magnetic field has an effect not only on the medium but also on the growth of the bacteria. Fig. 3a shows a typical response curve, which reflects the effect of a 0.2 T magnetic field on the medium and the E. coli. To examine the effect of a magnetic field on bacteria, we subtracted the response value in Fig. 3b from the corresponding value in Fig. 3a and obtained the response curve, which only reflected the effect of the magnetic field on E. coli, as shown in Fig. 3c. It can be observed that Df has no obvious change in the initial ~ 7.5 h, then the frequency increases about 960 Hz and finally reaches a stable level. In comparison with Fig. 2a, Fig. 3c displays similar change trends. Both are sigmoidal curves and can be roughly divided into three phases, i.e., the first plateau phase, the quickly increasing phase and the second plateau phase. However, one can also examine the difference between them from three aspects. Firstly, the time that the first plateau lasted is different. In the presence of the 0.2 T magnetic field, the time the phase lasts is longer than that in the normal case (about 3 h). Secondly, the signal size of the response is different. The frequency shift between the two plateaus is different. Df is about 1200 Hz in a normal growth case, but in the presence of a 0.2 T magnetic field, the frequency shift is ca. 960 Hz. Thirdly, the rate of change of the response signal in the quickly increasing phase is different. It changes by about 1200 Hz in only 3 h in the normal growth case. However, it changes by only about 960 Hz in 6.5 h in the presence of the 0.2 T magnetic field. Since the Df response curve corresponds to the theoretical bacterial growth curve, the above differences indicate that the lag time is prolonged, and the growth rate and the maximum number of bacteria are reduced due to the effect of the magnetic field. The reason for these observations may be that the activities of some of the enzymes in E. coli were inhibited by the magnetic field. Estimation of kinetic parameters of bacterial growth and establishment of a new model Fig. 3 a, The typical response curve of culture system affected by 0.2 T magnetic field. b, The typical response curve of medium affected by 0.2 T magnetic field. c, The typical response curve of E. coli affected by 0.2 T magnetic field. In order to establish the relationship between the frequency shift response and the number of bacteria, we detected the relative number of bacteria using the PPC method and obtained the value of K as 167.7 Hz. By taking, m m and l in eqn. (5) as estimation parameters, we use the nonlinear fitting program embedded in Sigmaplot Scientific Graphing Software Version 2.0 to fit the Df response. The relative sum of the residual squares (q r ) is used as the criterion to reflect the validity of fitting: q r = n  1 ( Df - Df ) n  1 fit 2 exp ( Df ) 2 exp (6) nalyst, 2002, 127, 373 377 375
where Df fit and Df exp represent the fitted and experimentally obtained Df values, respectively, and n is the number of response signal points. Fig. 4 shows the experimental response curve in Fig. 2 in comparison with the relevant fitting curve of eqn. (5). Fig. 5 shows the experimental response curve and the relevant fitting curve in the presence of a 0.4 T magnetic field. The experimental response in Fig. 5 is obtained just like that in Fig. 3c and the following experimental curves are also obtained like that. It can be seen that the fitted curves are in agreement with the experimental response curves. The small value of q r (see Table 1) shows that the Df versus time curve is suitable to reflect the growth situation of the bacteria, and the growth of the bacteria in the two cases can be described by this equation. Fig. 4 The response curve and the relative fitting curve obtained from two methods for bacterial normal growth process. :, Change in bacterial concentration with time which was obtained with the PPC method. 5, Typical response curve of Df to the bacterial growth situation. Line a, Fitting curve of bacterial concentration based on the Logistics model. Line b, Fitting curve of Df t based on the proposed model. In order to further study the effect of magnetic fields, the growth situation of E. coli was also investigated in the presence of 0.05, 0.1, and 0.6 T magnetic fields. The results fitted according to eqn. (5) are shown in Table 1. It can be seen that the three parameters of bacterial growth are affected by the magnetic fields. The lag time (l) is prolonged, the maximum specific growth rate (m m ) and the asymptote () are reduced. Since the three parameters represent the growth characteristics of the bacteria, any change in the three parameters indicates that the growth of the bacteria is inhibited due to the impact of the magnetic field. The inhibition is more serious with increasing magnetic field intensity. To study the effect of the magnetic field on the growth parameters in detail, the curves of, m m and l versus the intensity of the magnetic field (B) were plotted, see Fig. 6. From Fig. 6, it can apparently be seen that the three kinetic growth parameters vary regularly with an increase of B. and l can be linearly expressed by B. However, an exponential relationship exists between m m and B. The regression equations were obtained as follows: = 7.25 2 7.11B (r = 0.998) (7) m m = 2.46e 2397B (q r = 1.7 3 10 23 ) (8) l = 4.42 + 16.45B (r = 0.999) (9) From eqns. (5), (7), (8) and (9), the following equation can be easily obtained: Df = 167. 7-397B 4 2 46 { 1+ exp[ (. ) (10). e ( 442. + 1646. B- t) + 2]} This equation reflects the relationship between the intensity of the magnetic field and the frequency shift response, and can be considered as a magnetic field inhibition response model. The value of B can be estimated using the measured Df according to eqn. (10). Moreover, eqn. (10) can be used to predict the Df response with known B. The model can be applied only in relatively low magnetic fields. When a high magnetic field exists, the bacteria will be inhibited and not grow at all and m m will close to an infinitely Fig. 5 The response curve and the relative fitting curve obtained from two methods for affected by 0.4 T magnetic field. :, Change in bacterial concentration with time which was obtained with the PPC method. 5, Typical response curve of Df to the bacterial growth situation. Line a, Fitting curve of bacterial concentration based on the logistics model. Line b, Fitting curve of Df t based on the proposed model. Table 1 Parameters obtained by fitting Df according to eqn. (5) (K = 167.7) Magnetic field/t m m l q r 0 7.26 2.47 4.46 7.89 3 10 24 0.05 7.02 1.99 5.22 1.86 3 10 24 0.1 6.48 1.64 5.89 2.34 3 10 24 0.2 5.72 1.03 7.93 3.5 3 10 23 0.4 4.41 0.58 10.92 9.19 3 10 24 0.6 3.02 0.20 14.30 2.26 3 10 24 Fig. 6 a, Relationship between the asymptote () and intensity of magnetic field (B). b, Relationship between the maximum specific growth rate (m m ) and intensity of magnetic field (B). c, Relationship between the lag time (l) and intensity of magnetic field (B). 376 nalyst, 2002, 127, 373 377
small value, while l will reach an infinitely large value. The relationship described in eqns. (7) (9) will not be valid. Verifying the validity of the model To check the proposed model, the bacterial growth parameters obtained from eqns. (7) (9) were compared with those from the logistic popular growth model, in which the concentration of the bacteria was determined by the conventional PPC method. Typical fitting and experimental curves for the bacterial concentration in the absence and in the presence of a 0.4 T magnetic field are presented in Figs. 4 and 5, respectively. The fitting results of the three parameters obtained from the two models are shown in Table 2. The results show that the parameters of bacterial growth are close to those obtained from the proposed equation, which indicates that the proposed model can accurately reflect the bacterial growth process. Moreover, the advanced proposed model reflects the effect of a magnetic field on the bacterial growth. and the model can be used to predict the growth trend of bacteria under a given magnetic field intensity. lthough the biomagnetic mechanisms need further investigation, this work provides a new method of controlling the growth of bacteria and for studying the biological effects of magnetic fields. cknowledgements This work was supported by the ational atural Science Foundation of China. Table 2 Bacterial growth parameters obtained from two models m m l Magnetic field/t Proposed Logistics Proposed Logistics Proposed Logistics 0 7.25 7.28 2.46 2.41 4.42 4.46 0.05 6.89 6.92 2.02 1.92 5.24 5.18 0.1 6.53 6.46 1.65 1.64 6.06 5.97 0.2 5.82 5.75 1.11 1.06 7.71 7.85 0.4 4.40 4.36 0.50 0.56 11 10.28 0.6 2.96 3.05 0.22 0.24 14.29 14.12 References 1 P. Bochv, Bl. Bechev and M. Magrisso, Bioelectrochem. Bioenerg., 1992, 27, 45. 2 B. Kula and M. Drozdz, Bioelectrochem. Bioenerg., 1996, 39, 21. 3 K. Tsuchiya, K. Okuno, T. no, K. Tanaka, H. Takahashi and M. Shoda, Bioelectrochem. Bioenerg., 1999, 48, 383. 4 H. Shin-ichiro, I. Yoshimasa, O. Kazumasa,. Takashi and S. Maloto, Bioelectrochemistry, 2001, 53, 149. 5 K. Tsuchiya, K. akamura, K. Okuno, T. no and M. Shoda, J. Ferment. Bioeng., 1996, 81, 344. 6 T. Koana, M. Ilehata and M. alagawa, Bioelectrochem. Bioenerg., 1995, 36, 95. 7 M. Motta, Y. Halk,. Gandhari and C. Chen, Bioelectrochem. Bioenerg., 1998, 47, 297. 8. L. Di Carlo, J. M. Mullins and T.. Litovitz, Bioelectrochemistry, 2000, 52, 9. 9 M. L. Speak, Compendium Of Methods For The Microbiological Examination Of Foods, merican Public Health ssociation, Washington, DC, 1976. 10 J. Z. Zhang, W. Z. Wei, Y.. Mao, W. Lan and M. J. Chen, nal. Sci., 2000, 16, 1265. 11 L. L. Bao, X. G. Qu, H. M. Chen, X. L. Su, S. Z. Yao and W. Z. Wei, Mikrochim. cta, 1999, 132, 61. 12 F. J. He, W. H. Zhu, Q. Geng, L. H. ie and S. Z. Yao, nal. Lett., 1994, 27, 655. 13 R. L. Bunde, E. J. Jarvi and J. Rosentreter, Talanta, 1998, 46, 1223. 14 H. W. Tan, L. Deng, L. H. ie and S. Z. Yao, nalyst, 1997, 122, 179. 15 H. W. Tan, R. H. Wang, S. H. Wang, W. Z. Wei and S. Z. Yao, nal. Lett., 1998, 31, 949. 16 J. Z. Zhang, L. L. Bao, S. Z. Yao and W. Z. Wei, Microchem. J., 1999, 62, 405. 17 S. H. Si, J. Y. Xu, L. H. ie and S. Z. Yao, J. Biochem. Biophys. Methods, 1996, 31, 135. 18 J. Z. Zhang, Y. T. Xie, X. Y. Dai and W. Z. Wei, J. Microbiol. Methods, 2000, 44, 105. 19 S. Z. Yao and T.. Zhou, nal. Chim. cta, 1988, 212, 61. 20 D. Z. Shen, Z. Y. Li, L. H. ie and S. Z. Yao, nal. Chim. cta, 1993, 280, 209. 21 D. Z. Shen, W. H. Zhu, L.. Hua and S. Z. Yao, nal. Chim. cta, 1993, 276, 87. 22 S. Y. Yao and Z. H. Mo, nal. Chim. cta, 1987, 193, 97. 23 M. H. Zwietering, I. Jongenburger, F. M. Rombacts and K. V. Riet, ppl. Environ. Microbiol., 1990, 56, 1875. 24 W. Shao, Z. Xiong, H. Z. Li and H. Zhou, Microbiol. ewslett., 2000, 27, 112; (in Chinese). 25 S. Li, Q. Zeng and P. Zhang, Physical Chemistry, Higher Education Press, Beijing, 1985, pp. 236 238, (in Chinese). nalyst, 2002, 127, 373 377 377