ISSN 1061-9348, Journal of Analytical Chemistry, 2016, Vol. 71, No. 7, pp. 676–684. © Pleiades Publishing, Ltd., 2016.

ARTICLES

Rapid Diagnosis of Mycobacterium Tuberculosis with Electrical Impedance Spectroscopy in Suspensions Using Interdigitated Microelectrode1 A. R. Pourmira,*, A. R. Bahrmandb,**, S. H. Ettefagh Farc, A. R. Hadizadeh Tasbitib, and Sh. Yarib aDepartment

of Biophysics, Institute of Biochemistry and Biophysics (IBB), University of Tehran Tehran, Iran b Tuberculosis Department, Pasteur Institute of Iran Tehran 13164, Iran c Young Researchers and Elites Club, North Tehran Branch, Islamic Azad University Tehran, Iran *e-mail: [email protected] **e-mail: [email protected] Received May 9, 2015; in final form, July 26, 2015

Abstract—A Mycobacterium tuberculosis (MTB) bacilli are still widely spreading and have to be diagnosed fast and efficiently. Therefore, a new simple and rapid method was proposed to detect MTB by the impedance properties of MTB suspensions using interdigitated microelectrodes. As a result, MTB suspensions in deionized (DI) water with different cell concentrations generated different electrical impedance spectral responses. Whereas MTB suspensions in 0.9 wt. % NaCl solution did not produce any significant differences in the impedance spectra in response to different cell concentrations. In DI water suspensions, the impedance at 1 kHz decreased with increasing cell concentrations. The impedance of MTB suspension in DI water has been discussed; it was found to be resulted from the cell wall charges and release of ions from the cells. There was a linear relationship between the impedance and logarithmic value of the cell concentration in the cell concentration range of 102 to 108 cfu/mL, which can be expressed by the regression equation of Z (kΩ) = –456lnN (cfu/mL) + 9717 with R2 = 0.99. Detection limit was calculated as 10 4 cfu/mL, which is comparable with many label-free immunosensors for detecting pathogenic bacteria reported in the literature. This work demonstrated that MTB concentration can be determined through measuring the impedance of MTB suspensions in DI water. This new detection mechanism can be an alternative for current impedance methods available for detecting bacterial cells. Keywords: electrical impedance spectroscopy (EIS), interdigitated microelectrode, mycobacterium tuberculosis bacilli, bacteria detection DOI: 10.1134/S1061934816050099

Mycobacterium tuberculosis is a dangerous pathogenic bacterium, leading to tuberculosis (TB) [1]. Tuberculosis is a collective name which refers to the bacterial infection caused by the members of MTB complex which infect lungs (pulmonary) as well as kidneys, lymph nodes, bones, and joints (extra-pulmonary) [2]. Currently, about one-third of the human population is infected with TB worldwide and it is a major public health problem that may emerge as a complication along with acquired immune deficiency syndrome (AIDS). Emergence of drug-resistant bacteria species and AIDS epidemic has been considered as the main cause for the widespread reoccurrence of TB. To effectively control tuberculosis, it is crucial to 1 The article is published in the original.

perform rapid diagnosis and treatment, because the bacilli can be transmitted through the infected respiratory tract of each patient to an average of 12–15 people per year [1]. Principally, MTB cell wall includes a number of antigens (glycolipids or proteins), which may be used in preliminary serologic diagnosis [3]. When MTB cells grow in a liquid medium without detergent, they form tight bundles or cords, which consist of the bacilli in which the orientation of the long axis of each cell is parallel to the long axis of the cord. Glycolipid was first called cord factor by Bloch and later was identified as trehalose dimycolate [4–8]. It was shown that trehalose dimycolate spontaneously forms a crystalline monolayer on hydrophobic surfaces that are more rigid and stable than those formed by any other bio-

676

RAPID DIAGNOSIS OF MYCOBACTERIUM TUBERCULOSIS

logic amphiphiles. The monolayer is a two-dimensional crystal of regular linear arrays with hydrophilic (trehalose) and hydrophobic (mycolate) domains [3, 7, 8]. MTB cell wall is highly dynamic and strongly responds to environmental changes through either deviate ions or conforming constituent macromolecules. Electrical potential of MTB cell wall is the result of exposed phosphate and amino groups and is sensitive to the surrounding pH. Thus, effective dissociation constants for surface charged groups depends on environmental properties and defines local hydrophobicity and electric potential [1, 9]. Impedance technique (as one of the principle electrical/electrochemical transductions) has turned into a fertile area for developing interdisciplinary methods which are applied in a wide range of biological and biomedical detections. This issue has occurred due to a number of cases including: i—electrical properties of the biological entities and/or biological reactions motivate attention to impedance techniques; ii— impedance is one of the most promising techniques for developing label-free, real-time, and non-invasive methods for biological detection; and iii—impedance, as an electro analytical technique, can easily interface with miniaturized devices, such as biosensors and biochips, to meet the growing need for offering an analytical footprint which is considerably smaller than the laboratory-based instruments [10]. Impedance technique has a long history in application in the field of microbiology as a means for detecting and/or quantifying pathogenic bacteria and has recently received increasing attention from researchers in different fields. The electrical nature of bacterial cells and their electrophysiology are essential factors for developing impedance methods in order to detect bacterial cells. Identical to all biological cells, bacterial cells consist of adjacent of structure of materials that have very different electrical properties [11]. Biosensors are specific and highly sensitive; therefore, they can detect a broad spectrum of analysts present in sputum, salvia, serum, and urine with requiring minimum sample preparation [11]. There are various signature-charged groups in bacterial cell wall which may associate or dissociate upon changes in the pH or ionic strength of the suspending fluid and define certain affinity when bacteria approaches charged surfaces of another bacterium or a substratum. The involved surface electrostatic inter-actions may cause particular changes in the conformation of the charged molecules in favor of repulsion or induce adhesion to surface. Moreover, changes in the associated peptidoglycan layer with the surface of bacterial cell wall may change the permeability to solvent, solutes, and ions. An electric double layer is formed when a charged surface comes into contact with an electrolyte solution. In this situation, space distribution from charges, which mainly consists of counter-ions comJOURNAL OF ANALYTICAL CHEMISTRY

Vol. 71

677

ing from the electro neutral solution, will become equilibrated with the surface charges [9, 12]. Based on the electrophysiological and impedance properties of biological cells, four major mechanisms that have been reported for the detection/quantification of bacterial cells using impedance measurements are available: (i) using impedance microbiology, as a technique based on the measurement of change in electric impedance in a medium or a reactant solution [13, 14]. Impedance changes are mainly produced by the release of ions from cells [15]. Applications of this detection mechanism for detecting bacterial cell from classic impedance microbiology to novel on-chip impedance microbiology have been reviewed by Silley and Forsythe [13], Wawerla et al. [14], and Yang and Bashir [16]. (ii) Using the insulating properties of the cell membrane. Because of their highly insulated cell membrane, the cells attached to an electrode surface effectively reduce the electrode area, to which the current reaches and hence increases the interface impedance. Most of the impedance biosensors for bacterial detection are based on this principle, in which the antibodies that are specific for the target bacterial cells are immobilized on an electrode surface; then, they serve to facilitate the attachment of cells to the electrode surface and provide selectivity to the sensor. These impedance biosensors evidently require cell attachment/adherence to electrode surfaces in order to obtain certain coverage for generating detectable impedance signals. To obtain impedance signals, these sensors may or may not use a redox probe in detection systems [17–20] and they have been developed for the detection of various bacterial cells with the detection limits of 102 to 108 cfu/mL and detection times of 30 min to several hours [21]. In the present work, for the detection of MTB cells based on method (i), a rapid impedance measurement was proposed using interdigitated microelectrodes. Electrical impedance spectroscopic responses of MTB suspensions were found to be highly dependent on the buffers used in the detection systems. Conductivity of bacterial suspensions has been reported for studying the electrical properties of bacterial cell surface and the related cell surface interfacial physiology [22, 23]; but it has not yet been reported to quantify the concentration of bacterial cells in the suspensions. In this work, MTB suspensions in DI water and 0.9 wt % NaCl solution (to achieve the advantages this method, we used 0.9 wt % NaCl solution that use in traditional diagnosis methods) were studied over a wide range of frequencies and it was found that MTB suspensions in DI water with different cell concentrations could result in different electrical impedance spectral responses. In a certain frequency range, impedance of the cell suspension was related to the cell concentration in the suspension, which could provide an alternative for quantifying bacterial cells in a label-free, inexpensive, and very simple approach.

No. 7

2016

678

POURMIR et al.

Mycobacterium tuberculosis bacili suspension

Glass cover The IME

Impedance analyzer Water at 25°C

Fig. 1. Experimental set-up of the impedance measurement with the IME in MTB suspension.

EXPERIMENTAL Chemicals. 0.9 wt % NaCl solution (justly) was purchased from Pasteur Institute of Iran. 0.9 wt % NaCl solution of pH 7.4 was made by dissolving in 100 mL DI water (simulated solution in human body) according to the instruction. The conductivity of the 0.9 wt % NaCl solution was ~47 mS/cm, while that of DI water was ~3 μS/cm, as measured by a pH/conductivity meter (207452 CDMZE, Radiometer Copenhagen, Denmark). Bacteria culture and media. MTB culture which was purchased from Pasteur Institute of Iran was on Lowenstein–Jensen medium. The culture was grown in fresh egg at 37°C for 4–8 W [24]. Cells were plated by centrifuging (Eppendorf, Westbury, NY) at 3000 g for 5 min and resuspended in sterilized DI water or 0.9 wt % NaCl solutions. They were then washed 3 times in order to get rid of ionic residues from the growth medium; afterward, they were serially (1 : 10) diluted with DI water or 0.9 wt % NaCl solution to the desirable concentrations for further experiments. The cell numbers were determined by makeup Mac Farland solution including appropriate dilutions and prepared by diluting cells in DI water to desirable concentrations. For fluorescent imaging, MTB cells were stained with Auramine O and Rhodamine B stains (golden yellow) molecular probes for visualization purposes under a fluorescence microscope. All the stained bacterial cells were centrifuged and washed with DI water or 0.9 wt % NaCl solution 3–5 times to remove excess dye molecules. Electrical impedance spectroscopy and devices. The device for EIS measurements consisted of an array of

interdigitated microelectrodes (IME), as shown in Fig. 1. The gold IME was fabricated on a flat glass substrate with total of 100 pairs of finger electrodes. The finger electrodes were 20 μm in width and had 20 μm spaces between them (purchased from Tiny Layer Laboratory, University of Tehran). IME was cleaned with ethanol (70%) and DI water and dried with a stream of nitrogen. Impedance measurements were performed using a Solarton impedance analyzer device, Interface Solarton SI 1287 model, made in the UK with the ZView/ZPlot software. For measurements, IME was placed into the chamber and covered with a glass cover. One of the two microband array electrodes was connected to the working and sense probes and the other was connected to the reference and counter electrodes on the impedance analyzer. EIS measurements were carried out in the frequency range of 1 Hz to 100 kHz. Bode (impedance and phase versus frequency) curves were also recorded. Impedance at fixed frequency was measured using the capacitance–potential (C/E) program at 1 kHz with the amplitude of ±50 mV. Impedance data were recorded every minute. All the tests were performed at room temperature. Simulation was performed using the ZView program. Sixteen points of data from each measured spectrum were automatically selected by the software as the input to an equivalent circuit to generate a fitting spectrum. Fluorescence imaging. Fluorescent images were taken using an Olympus BX51 fluorescence microscope (Japan) with an Olympus DP70 camera (Central Laboratory, Faculty of Hygiene, Tehran University of Medical Sciences).

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 71

No. 7

2016

RAPID DIAGNOSIS OF MYCOBACTERIUM TUBERCULOSIS

RESULTS AND DISCUSSION EIS of MTB suspensions in DI water and 0.9 wt % NaCl solution. Figure 2 presents the Bode impedance spectra of MTB suspensions in DI water (a) and in 0.9 wt % NaCl solution (b) along with their equivalent circuits and fitting spectra. For MTB suspension in DI water, the measured spectrum (Fig. 2a, blank dots) was a typical Bode plot for a system, in which polarization was due to kinetic processes. Based on the general electronic equivalent model of an electrochemical cell [25] and behavior of the IME [26], an equivalent circuit, consisting of ohmic resistance (Rs) of the solution between two electrodes and double layer capacitance (Cdl), electron-transfer resistance (Ret) around each electrode, was considered for the simulation of the measured spectra. Among these electrical elements, Rs represents the properties of the bulk solution. The other two elements, Cdl and Ret, represent the dielectric and insulating features at the electrode/electrolyte interface. For the simulation, 16 data points on the impedance measured spectrum were automatically selected using the software and applied as the input to the equivalent circuit to generate a fitting impedance spectrum (Fig. 2, solid line). The agreement between the measured data and fitting spectra indicated that the equivalent circuits provided a feasible, if not unique, model for describing the impedance characteristics of MTB suspensions in DI water. Using this simulation, the values of Cdl, Ret, and Rs were 76.8 μF, 777.68 and 4.32 kΩ, respectively, with the mean error for modulus impedance of 0.5%. For MTB suspensions in 0.9 wt % NaCl solution, the impedance spectrum (Fig. 2b, blank dots) demonstrated two domains: i.e. a double layer region in low frequency range from 1 to approximately 600 Hz and a resistive region in the frequency range of ~600 to 100 kHz. The electric impedance behavior of the cell suspension in 0.9 wt % NaCl solution can be represented by the equivalent circuit of the IME system in the aqueous solutions which have been previously reported [26–29]. In this circuit model, two identical double layer capacitances (Cdl) of each set of the IME were serially connected to the medium resistance (Rs). Cdl dominated the impedance in the low frequency range (double layer region), whereas Rs dominated it in high frequency range (resistive region). By simulation, the values of Cdl and Rs were 1.46 μF and 717.06 kΩ, respectively, with the mean error for modulus impedance of 4.0%. Detailed reasons for the difference in the impedance spectral behavior between the cell suspensions in DI water and 0.9 wt % NaCl need to be further discussed. Characteristics of the spectrum along with the equivalent circuit in Fig. 2a suggested that some electrochemical reactions occurred on the IME electrodes in the cell suspension in DI water. The electrontransfer resistance (Ret) was the parameter which indicated the electrochemical reactions of the electrochemically active species in the suspension, implying that the cells JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 71

679

may release some electrochemically active composites to the DI water. However, in the cell suspension in 0.9 wt % NaCl, the impedance spectrum did not demonstrate any characteristics relating to electrochemically active parameters, which implied that the cells may not release active electrochemically active species into 0.9 wt % NaCl. EIS of bacterial suspensions with different concentrations. Figure 3 shows the Bode impedance spectra of MTB suspensions in DI water (a) and 0.9 wt % NaCl solution (b) with different cell concentrations from 102 to 108 cfu/mL. As it can be observed, the impedance spectra of MTB suspensions in DI water responded differently from cell concentrations in the frequency range of 100 Hz to 10 kHz, whereas impedance spectra of MTB suspensions in 0.9 wt % NaCl showed no significant difference at any frequency. For cell suspensions in DI water, impedance decreased with the increasing cell concentrations in the frequency range of 100 Hz to 10 kHz, which would allow estimating the cell concentration in the DI water suspension using the impedance value at a fixed frequency. As the best representative frequency, 1 kHz was identified as the test frequency for investigating the relationship between impedance value and cell concentration in DI water suspensions [1]. Figure 4 demonstrates typical impedance responses at 1 kHz to the samples containing different bacterial concentrations in impedance measurements. Impedance of each sample was measured after the cells were suspended in DI water for ~1 h. Impedance value at 1 kHz increased as the concentration of cells in the suspension was decreased. When bacterial concentration decreased from 108 to 107, 106, 105, 10 4 cfu/mL impedance of the suspension significantly increased from 1 to 1.2, 2, 3.7, and 4.5 kΩ. When MTB cell concentrations were lower than 10 4 cfu/mL, impedance values of the suspensions were not significantly different from each other or from DI water. By fitting the impedance spectra of cell suspensions in DI water in Fig. 3 to the equivalent circuit model in Fig. 2a, it was found that the values of almost all electrical elements of the suspensions, Ret, and Rs changed when the cell concentration changed, except that Cdl was left relatively stable. This simulation result was consistent with the spectra in Fig. 3. As reported in the previous studies, Cdl usually appears on the spectra in the double layer region which is always in the low frequency range (<10 Hz) and usually contributes to the impedance value in a frequency-dependent manner ( Z = − j 2πfC , where f is frequency); in this region, impedance decreases with frequency. As shown in Fig. 3, impedances on the spectra in low frequency range (<10 Hz) were not significantly different from other cell concentrations. Frequency ranges from 100 Hz to 10 kHz on these impedance spectra were resistive regions which reflected the impedance contributed by all resistive components in the detection

No. 7

2016

680

POURMIR et al.

(a)

(b)

106

106 105 Impedance, Ω

Impedance, Ω

105

104

102

Rs

101

Cdl

Ret

Ret 102 100

103

Cdl

Cdl

103

104

102 103 Frequency, Hz

104

101 100

105

101

Rs

Cdl 102 103 Frequency, Hz

104

105

Fig. 2. Impedance spectra of MTB suspensions in DI water (a) and in 0.9 wt % NaCl solution (b) along with their fitting spectra and the equivalent circuits. Frequency range of 1 Hz–100 kHz, amplitude of ±50 mV, MTB concentration of 1 × 10 6 cfu/mL; (o) – measured data, (—) – fitting data.

(a)

(b)

105

105

0.9 wt % NaCl

104

Impedance, Ω

Impedance, Ω

DI water

103 108 cells/mL

108 cells/mL

104

103

1 kHz 102 102 100

101

102 103 Frequency, Hz

104

105

100

101

102 103 Frequency, Hz

104

105

Fig. 3. Impedance spectra of MTB suspensions in DI water (a) and 0.9 wt % NaCl solution (b) with the cell concentrations of 102 to 108 cfu/mL along with water and NaCl as controls. Frequency range of 1 Hz–100 kHz and amplitude of ±50 mV.

system and, in this case, included Ret and Rs. At 1 kHz, impedance value was believed to reflect the combination of Ret and Rs components. These results indicated that alteration in cell concentration could alter the suspension impedance by changing Ret and Rs components in the detection system. Therefore, the cell concentration in the DI water suspensions can be determined by measuring the total impedance of the suspensions at 1 kHz regardless of whether changes were from the Ret or Rs components. However, the

impedance of cell suspensions in 0.9 wt % NaCl solution did not cause any difference at any tested frequency in response to different cell concentrations; thus, it could not be used as an indicator for cell concentrations in 0.9 wt % NaCl solutions. Discussion on the impedance properties of cell suspensions in DI water. Observation of decreases in the impedance of the cell suspensions with increasing cell concentrations indicated that cell suspensions with high concentrations were more conductive than those

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 71

No. 7

2016

RAPID DIAGNOSIS OF MYCOBACTERIUM TUBERCULOSIS

In some studies, conductivity of the cell suspension has been considered to vary linearly with the bacterial cell volume fraction [22, 29]. Figure 5 shows images of the MTB cell distribution in the test chamber at different concentrations, representing cell densities in the suspensions at certain cell concentrations. Here, it can be expected that the conductivity of the solution should proportionally vary with the number of the cells in the solution or the cell concentration when total volume is fixed. Another source of MTB cells for altering the conductivity of DI water suspension is ion release from MTB cells [29]. Quantifying bacterial concentration in DI water suspension by impedance. Under the experimental conditions of this study, the plot of the impedance values as a function of the bacterial concentrations is shown in Fig. 6. A linear relationship was found between the impedance and logarithmic values of the cell concentration in the range of 10 4 to 108 cfu/mL. Impedance decreased as a result of increasing cell concentration. The linear regression equation was Z (kΩ) = –456 lnN (cfu/mL) + 9717 with R2 = 0.99 and detection limit was calculated as 10 4 cfu/mL. Descriptions of linear regression equation and detecJOURNAL OF ANALYTICAL CHEMISTRY

Vol. 71

105

Impedance, Ω

with lower concentrations, reflecting that MTB cells contributed to conductive compositions of the suspension. Also, two possible sources were found for the MTB cells to alter the impedance of DI water suspensions; one was the electrical nature of MTB cell surfaces. The MTB cell walls contain various acidic groups such as carboxyl, phosphate, and amino groups, which may be present at high concentrations [30]. Generally, they contain higher concentrations of anionic groups than those of cationic groups, which results in a negative cell wall charge at neutral pH. This charge is compensated for by counter-ions that penetrate into the porous cell well and to a minor extent by co-ions which are expelled from it, thereby conferring electrostatic charge to the cell periphery [22, 23]. Charge density of MTB cell wall can be as high as 0.2‒0.6 C/m2. In the present experiments, conductivity of the DI water was in the range as low as about 2– 4 μS/cm to up to about 12–18 μS/cm. When MTB cells were suspended in such low conductive DI water and reached sufficient concentration (10 4 cfu/mL in the present experiments), they could alter the conductivity of the suspension because of their cell wall charges. In contrast, in 0.9 wt % NaCl solution, the MTB cell wall charges were not sufficient to alter the conductivity of the bulk solution due to the high background conductivity of 0.9 wt % NaCl (~47 mS/cm). Therefore based on power hydrophobicity of MTB cell wall and tendency to organization of stable and rigid agglomeration e.g. cord factor, ion release from MTB in solutions e. g. 0.9 wt % NaCl, is a problem. But DI water the hydrophobicity of MTB decreases and the most of ions release which affects the impedance of suspension and it.

681

2 3 5 104 10 10 DI water 10 10 6

104

107 108

103

102

0

20

40 60 Time, min

80

Fig. 4. Typical impedance responses to the samples with different concentrations of cells when the samples were measured sequentially in the impedance detection chamber; the arrows indicate the time when a sample was introduced into the detection chamber. The numbers in the figure indicate the concentrations of MTB cells in the samples, cfu/mL.

tion limit based on kind of different bacteria given in table. These results indicated that the cell concentrations in DI water suspensions can be determined from the impedance of the MTB suspensions, which can be an alternative approach for quantifying bacterial cells in suspensions other than impedance microbiology and impedance biosensors for bacteria detection. This method did not require any labels or amplification steps and its detection limit was comparable with many other label-free immunosensors in terms of detecting pathogenic bacteria using different transducer techniques, including smear microscopy (AFB), PCR, and ELISA. In this study, a new, simple, and rapid impedance method was demonstrated for detecting MTB cells by the impedance properties of bacterial cell suspensions using IME, which was beyond the mechanisms currently used for detection of MTB cells. In addition, this method is safe, fast and inexpensive. Thus, these are the major reasons for the selectivity of this approach. It was found that MTB cell suspensions in DI water with different cell concentrations can result in different electrical impedance spectral responses, whereas cell suspensions in 0.9 wt % NaCl solutions could not produce any significant differences in impedance spectra in response to different cell concentrations. The impedance value at a fixed frequency (1 kHz) of the cell suspension was investigated in order to quantify cell concentration in the suspensions in DI water, the result of which showed that impedance decreased with the increasing cell concentrations in the suspensions. Impedance of MTB cell suspensions in DI water was discussed and it was found to be resulted from the cell wall charges and the release of ions from the cells. A linear relationship was also

No. 7

2016

682

POURMIR et al.

108 cfu/mL

107 cfu/mL Electrodes

Electrodes

Gap Gap

106 cfu/mL

105 cfu/mL

Electrodes

Electrodes

Gap

Gap

104 cfu/mL

Gap Electrodes

20 μm

Fig. 5. Linear relationship between the logarithmic values of the concentration of MTB and the impedance measured at 1 kHz. Error bars are standard deviations of 3–6 measurements.

observed between the impedance and log value of the cell concentration in the range of 10 4 to 108 cfu/mL, which can be expressed by the regression equation of Z (kΩ) =–456lnN (cfu/mL) + 9717 with R2 = 0.99. Detection limit was 10 4 cfu/mL that is comparable

with many other label-free immunosensors for detecting pathogenic bacteria using different transducer techniques of using for detection of MTB. Advances in microfabrication have paved the way for the miniaturization of many traditional detection platforms into

Characterization of linear regression equation and detection limit according to type of bacteria Linear regression equation Z (kΩ)–N

R2

Detection limit, cfu/mL

S. typhimurium

Z (kΩ)= –2.0logN + 5.23

0.98

3 × 10 6

[10]

Mycobacterium tuberculosis

Z (kΩ)= –456lnN + 9717

0.99

1 × 10 4

This work

Bacteria

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 71

Reference

No. 7

2016

RAPID DIAGNOSIS OF MYCOBACTERIUM TUBERCULOSIS

DI water Impedance at 1 kHz, kΩ

6 5 4 3 2

y = –456.4lnx + 9717 R2 = 0.99

1 108 107 106 105 104 103 102 101 100 Bacterial concentration, cfu/mL

Fig. 6. Representative densities of the MTB cells in DI water with concentrations from 108 to 10 4 cfu/mL in a thin layer right above the IME. Dots are the cells stained with Auramine O and Rhodamine B. The thin layer of the solution is about 100 μm. All the images were taken with a focus on the surface of the electrodes. Dots that are out of focus are the cells in the suspensions. The bacterial concentrations were 108, 107, 106, 105, and 10 4 cfu/mL.

microdevices/chips. To further improve the detection limit of this method, the use of microdevices which can concentrate the low number of bacterial cells into an ultramicrodetection chamber would be very useful. The assay reported in this study is simple and easy to perform within 1–2 h of sample collection. Although this method has some limitation such as number of bacilli in fluids (samples), we believe that can be easily done by the personnel with minimal technical training. This type of easy low cost and fast access to results can encourage people to start the diagnostic test. Additionally, to our knowledge, this is the first study on a rapid test based on detection of these bacilli by EIS method. Based on the present study, this test is useful in diagnose of TB as screening method (for negative smear) in the field. ACKNOWLEDGMENTS Authors would like to acknowledge Department of Biophysics, Institute of Biochemistry and Biophysics (IBB), University of Tehran, Iran, and Pastor Institute of Iran for providing experimental facilitates. We thank the researchers in TB Department for their technical support. REFERENCES 1. Zhou, L., He, X., He, D., Wang, K., and Qin, D., J. Immun. Res., 2011, 193963. doi 10.1155/2011/193963 JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 71

683

2. Gilckman, M.S. and Jacobs, W.R., Cell, 2001, vol. 104, no. 4, p. 477. 3. Baren, V., Havelkova, M., kaustova, J., Dvorska, L., and Pavlik, I., Vet. Med. (Prague, Czech Repub.), 2006, vol. 51, no. 7, p. 365. 4. Villeneuv, M., Kawai, M., Watanabe, M., Aoyagi, Y., Hitotsuyanagi, Y., Takeya, K., Gouda, H., Hirono, S., Minikin, D.E., and Nakahara, H., Biochim. Biophys. Acta, 2007, vol. 1768, no. 7, p. 1717. 5. Barkan, D., Liu, Z., Sacchettini, J.C., and Glickman, M.S., Chem. Biol., 2009, vol. 16, no. 5, p. 499. 6. Glickman, M.S., Cox, J.S., and Jacobs, W.R., Mol. Cell, 2000, vol. 5, no. 4, p. 717. 7. Davidsen, J., Rosenkrands, T., Christensen, D., Vangala, A., Kirby, D., Perrie, Y., Agger, E. M., and Andersen, P., Biochim. Biophys. Acta, 2005, vol. 1718, nos. 1–2, p. 22. 8. Dulayymi, J.R.A., Baird, M.S., Maza-Iglesias, M., Beken, S.V., and Grootem, J., Tetrahedron Lett., 2009, vol. 50, no. 26, p. 3702. 9. Dziubakiewicz, E., Hrynkiewicz, K., Walzyk, M., and Buszewski, B., Colloids Surf., B, 2013, vol. 104, no. 1, p. 122. 10. Yang, L., Talanta, 2008, vol. 74, no. 5, p. 1621. 11. Pethig, R., and Markx, G.H., Trends Biotechnol., 1997, vol. 15, no. 10, p. 426. 12. Poortinga, A.T., Bos, R., Norde, W., and Busscher, H.J., Surf. Sci. Rep., 2002, vol. 47, no. 1, p. 1. 13. Silley, P., and Forsythe, S., J. App. Bacteriol., 1996, vol. 80, no. 1, p. 233. 14. Wawerla, M., Stolle, A., Schalch, B., and Eisgruber, H., J. Food. Prot., 1999, vol. 62, no. 12, p. 1488. 15. Owicki, J. and Parce, J., Biosens. Bioelectron., 1992, vol. 7, no. 4, p. 255. 16. Yang, L. and Bashir, R., Biotech. Adv., 2008, vol. 26, no. 2, p. 135. 17. Ruan, C., Yang, L., and Li, Y., Anal. Chem., 2002, vol. 47, no. 18, p. 4814. 18. Yang, L., Li, Y., and Erf, G.F., Anal. Chem., 2004, vol. 76, no. 4, p. 1107. 19. Radke, S.M. and Alocilja, E.C., Biosens. Bioelectron., 2005, vol. 20, no. 8, p.1656. 20. Suehiro, J., Ohtsubo, A., Hatano, T., and Hara, M., Sens. Actuators, B, 2006, vol. 119, no. 1, p. 319. 21. Cheng, X., Liu, Y., Irimia, D., Demirci, U., Yang, L., Zamir, L., Rodriguez, W.R., Toner, M., and Bashir, R., Lab. Chip, 2007, vol. 7, no. 6, p. 746. 22. van der Wal, A., Minor, M., Norde, W., Zehnder, A., and Lyklema, J., Colloids Surf., B, 1997, vol. 9, nos. 1– 2, p. 81. 23. Wilson, W.W., Wade, M.M., Holman, S.C., and Champlin, F.R., J. Microbiol. Methods, 2001, vol. 43, no. 3, p. 153.

No. 7

2016

684

POURMIR et al.

24. Hadizadeh Tasbiti, A.R., Bahrmand, A.R., Shokrgozar, M.A., Chanaie, M., Fateh, A., Karimi, A., and Yari, Sh., J. Clin. Lab. Anal., 2014, vol. 28, no. 1, p. 70. 25. Bard, A.J. and Faulkner, L.R., Electrochemical Methods: Fundamentals and Applications, New York: Wiley, 1980, 2nd ed. 26. Van Gerwen, P., Laureyn, W., Laureys, W., Huyberechts, G., Op De Beek, M., Baert, K., Suls, J., Sansen, W., Jacobs, P., Herman, L., and Mertens, R., Sens. Actuators, B, 1998, vol. 49, no. 1 p. 73.

27. Laureyn, W., Nelis, D., van Gerwen, P., Baert, K., Hermans, L., Magnee, R., Pireaux, J.J., and Maes, G., Sens. Actuators, B, 2000, vol. 68, no. 1, p. 360. 28. Yang, L., Li, Y., Griffis, C.L., and Johnson, M.G., Sens. Actuators, B, 2004, no. 1, p. 1139. 29. Zou, Z., Kai, J., Rust, M.J., Han, J., and Ahn, C.H., Sens. Actuators, A, 2007, vol. 136, no. 2, p. 518. 30. Carstensen, E.L. and Marquis, R.E., Biophys. J., 1969, vol. 9, no. 4, p. 536.

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 71

No. 7

2016