Anal. Chem. 2005, 77, 7242-7245
Improvement of Immunodetection of Bacterial Spore Antigen by Ultrasonic Cavitation Kathryn A. J. Borthwick,† Tracey E. Love,‡ Martin B. McDonnell,‡ and W. Terence Coakley*,†
Cardiff School of Biosciences, Cardiff University, Cardiff CF10 3TL, Wales, U.K., and Dstl, Porton Down, Salisbury, SP4 OJQ, U.K.
Ultrasonic cavitation was employed to enhance sensitivity of bacterial spore immunoassay detection, specifically, enzyme-linked immunosorbent assay (ELISA) and resonant mirror (RM) sensing. Bacillus spore suspensions were exposed to high-power ultrasound in a tubular sonicator operated at 267 kHz in both batch and flow modes. The sonicator was designed to deliver high output power and is in a form that can be cooled efficiently to avoid thermal denaturation of antigen. The 30-s batch and cooled flow (0.3 mL/min) sonication achieved an ∼20fold increase in ELISA sensitivity compared to unsonicated spores by ELISA. RM sensing of sonicated spores achieved detection sensitivity of ∼106 spores/mL, whereas unsonicated spores were undetectable at the highest concentration tested. Improvements in detection were associated with antigen released from the spores. Equilibrium temperature increase in the tubular sonicator was limited to 14 K after 30 min and was maintained for 6 h with cooling and flow (0.3 mL/min). The work described here demonstrates the utility of the tubular sonicator for the improvement in the sensitivity of the detection of spores and its suitability as an in-line component of a rapid detection system. There is a requirement for rapid, sensitive identification of pathogenic bacteria for detection and diagnostic purposes.1 The infectious doses of many bacterial pathogens are in the range 101104 cells. Thus, many of the standard immunoassay techniques are not sensitive enough for diagnosis, while other approaches, such as traditional culture methods, are time-consuming, delaying the time for effective treatment to be administered.2,3 Current immunoassays detect the capture of whole cells or spores through the use of an appropriate antibody. The sensitivity of these assays could be improved through the prior disruption of the organism to release more readily detectable proteins and cell fragments. There are many methods by which disruption may be achieved including methods such as freeze thawing, the use of lysozymes, bead milling, or nebulization. The work described here investigates the use of ultrasonic cavitation. Integration of an ultrasonic * Corresponding author. E-mail:
[email protected]. Fax: ++44 29 2087 4305. † Cardiff University. ‡ Dstl. (1) Fox, A.; and 17 other attendees. J. Microbiol. Methods 2002, 51, 247-254. (2) Peruski, L. F.; Harwood Peruski, A. Biotechniques 2003, 35, 840-846. (3) Ivnitski, D.; Abdel-Hamid, I.; Atanasov, P.; Wilkins, E. Biosens. Bioelectron. 1999, 14, 599-624.
7242 Analytical Chemistry, Vol. 77, No. 22, November 15, 2005
immunoenhancing device with minimal temperature effects could produce substantial improvements in sensitivity of existing immunoassay-based techniques for Bacillus species.4-11 Ultrasonic cavitation is a rapid, reagentless method of mechanically stressing cells. When high-powered ultrasound is applied to an aqueous sample, gas nuclei or cavities in the sound field expand and collapse, generating localized high stresses. Temperature increase accompanies the deposition of ultrasonic energy in a sample. The highly ordered structure of Bacillus species spores provides strong resistance to mechanical and physical stresses, partly due to the dehydrated interior and protective proteinaceous coat.12,13 It has, for example, been reported that significant breakage of spores required over 1 h using a standard probe sonicator.14 Sonication has been successfully employed for Bacillus spore detection by PCR in shorter times using modified acoustic approaches.15-20 Thermal denaturation of DNA during prolonged sonication is not an issue for PCR since thermal denaturation of DNA is a preliminary step in the assay and DNA can also renature at lower temperatures. Immunodetection of intact antigen, however, requires that temperature increase be limited. (4) Zourob, M.; Mohr, S.; Treves Brown, B. J.; Fielden, P. R.; McDonnell, M. B.; Goddard, N. J. Anal. Chem. 2005, 77, 232-242. (5) Kozel, T. R.; Murphy, W. J.; Brandt, S.; Blazar, B. R.; Lovchik, J. A.; Thorkildson, P.; Percival, A.; Lyons, C. R. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 5042-5047. (6) Longchamp, P.; Leighton, T. Lett. Appl. Microbiol. 2000, 31, 242-246. (7) Speight, S. E.; Hallis, B. A.; Bennett, A. M.; Benbough, J. E. J. Aerosol Sci. 1997, 28, 483-492. (8) Quinlan, J. J.; Foegeding, P. M. J. Rapid Methods Autom. Microbiol. 1998, 6, 1-16. (9) Bake, M. R.; Weimer, B. C. Appl. Environ. Microb. 1997, 63, 1643-1646. (10) Bruno, J. G.; Yu, H. Appl. Environ. Microb. 1996, 62, 3474-3476. (11) Gatto-Menking, D. L.; Yu, H.; Bruno, J. G.; Goode, M. T.; Miller, M.; Zulich, A. W. Biosens. Bioelectron. 1995, 10, 501-507. (12) Driks, A.; Setlow. P. In Prokaryotic Development; Brun, Y. V., Shimkets, L. J., Eds.; ASM Press: Washington, DC, 2000; pp 191-212. (13) Driks, A. Trends Microbiol. 2002, 10, 251-254. (14) Berger, J. A.; Marr, A. G. J. Gen. Microbiol. 1960, 22, 147-157. (15) Fyske, E. M.; Olsen, J. S.; Skogan, G. J. Microbiol. Methods 2003, 55, 1-10. (16) Kuske, C. R.; Banton, K. L.; Adorada, D. L.; Stark, P. C.; Hill, K. K.; Jackson, P. J. Appl. Environ. Microb. 1998, 64, 2463-2472. (17) Belgrader, P.; Hansford, D.; Kovacs, G. T. A.; Venkateswaran, K.; Mariella, R., Jr.; Milanovich, F.; Nasarbadi, S.; Okuzumi, M.; Pourahmadi, F.; Northrup, M. A. Anal. Chem. 1999, 71, 4232-4236. (18) Chandler, D. P.; Brown, J.; Bruckner-Lea, C. J.; Olson, L.; Posakony, G. J.; Stults, J. R.; Valentine, N. B.; Bond, L. J. Anal. Chem. 2001, 73, 37843789. (19) Taylor, M. T.; Belgrader, P.; Furman, B. J.; Pourahmadi, F.; Kovacs, G. T. A.; Northup, M. A. Anal. Chem. 2001, 73, 492-496. (20) Johns, M.; Harrington, L.; Titball, R. W.; Leslie, D. L. Lett. Appl. Microbiol. 1994, 18, 236-238. 10.1021/ac050576c CCC: $30.25
© 2005 American Chemical Society Published on Web 10/08/2005
Figure 1. Schematic of tubular sonicator. Adapted from Borthwick et al.24
Immunodetection of Bacillus subtilis var. niger (BG) spores has been investigated here in a novel sonication device, developed around a tubular piezoceramic transducer operated at 267 kHz with an integrated cooling system. MATERIALS AND METHODS. Sonicator. Details of the 267-kHz tubular sonication device (Figure 1) have recently been described.21 A tubular piezoceramic transducer was coupled to the central 0.6-mL sample volume using a stainless steel tube. Power was supplied either by (i) a wave generator (Agilent model HP 33120A) providing an input voltage to an amplifier (model 240L, ENI, Rochester, NY) or (ii) a more compact combined frequency generator and amplifier (LF 256, T & C Power Conversion Inc.). Samples were located in the axial region of the cylinder where the acoustic field was focused. Power deposition in the system, estimated from the product of thermal capacity and initial heating rates, was 36 W.21 Selection of the sonication frequency for optimal cavitation was guided by a resonance modeling program as described by Borthwick et al.21 and determined by aural and visual observations as well as by measurements of erythrocyte disruption. The sonicator was housed in a die-cast steel box with heat sinks and fans for cooling. The 0.6-mL batch samples and 5-mL volumes in a flow arrangement were sonicated. Flow was provided by an Ismatec BV-GE peristaltic pump, and tubing was attached to either end of the device by a stainless steel adapter. Spore Suspension. B. subtilis var. niger (previously Bacillus globigii or BG) spore stock suspensions were suspended in sterile distilled water at 1010 or 1011 cfu/mL (Health Protection Agency, U.K.). Any vegetative cells or debris were removed from stock suspensions by washing the spores at 10000g for 10 min and resuspension in twice the original volume of ice-cold sterile distilled water. This was repeated 10 times. Spores were finally suspended in sterile distilled water at 109 cfu/mL prior to sonication. The presence of >95% phase bright spores was confirmed using phase contrast microscopy. Sample Sonication. The 0.6-mL aliquot of BG spores at ∼109 cfu/mL in sterile distilled water was sonicated in batch mode at 267 kHz in the tubular system with and without cooling. The 5-mL spore samples were sonicated in flow mode (0.3 mL/min) with cooling. Batch sonicated samples were removed using a minipastette (Alpha Laboratories) and stored in 1.5-mL microcentrifuge tubes. Samples sonicated with flow were collected directly into 15-mL centrifuge tubes. (21) Borthwick, K. A. J.; Coakley, W. T.; McDonnell, M. B.; Nowotny, H.; Benes, E.; Gro ¨schl, M. J. Microbiol. Methods 2005, 60, 207-216.
Total and Viable Cell Counts. A serial 10-fold dilution of the samples was performed. For total counts, spores were visualized under phase contrast microscopy and counted in a hemocytometer. For viable counts, 100 µL of diluted sample was plated onto LB agar plates, in triplicate. Plates were incubated overnight at 37 °C, and the colonies were counted. Electron Microscopy. The ultrastructure of selected spore samples was examined by transmission electron microscopy (TEM). Spores were fixed using 5% (v/v) gluteraldehyde in piperazine-1,4-bis(2-ethanesulfonic acid) buffer, embedded, and cut into thin sections. The sections were floated onto plastic-coated nickel grids and stained with lead citrate and uranyl acetate. The samples were viewed using a transmission electron microscope. Protein Determination. Protein concentration was determined in duplicate or triplicate by protocols based on Lowry’s method 22 (reagents from Sigma). Direct ELISA. Sonicated and unsonicated BG samples were coated, as follows, onto Immulon 2, 96-well microtiter plates in serial dilution. The samples were diluted (from the initial 109 spores/mL) in carbonate/bicarbonate buffer, pH 9.5 (Sigma); 100 µL of diluted sample was added to appropriate wells and the resultant mixture incubated overnight at 4 °C. The plates were washed 3 times with PBS with 0.05% polyethylenesorbitan monolaurate (Tween 20) (PBST), once with PBS, and then blocked with 1% milk powder (Marvel) in PBST for 1 h at room temperature. Polyclonal rabbit antibody raised against whole BG spores, R-BG (Dstl, UK), was added at a concentration of 10 µg/mL to the appropriate wells and left to incubate for 1 h. The plates were washed as previously described, and then an R-rabbit HRP conjugate (Sigma) was added to appropriate wells. This was left to incubate for 1 h, and then the plate washed 4 times with PBST and once with PBS. ABTS development buffer was added, and the plates were left to incubate for 20 min while color developed. Absorbance was measured at 405 nm. Each dilution of a sample was tested in duplicate. RM Sensing. Detection of analyte using the resonant mirror (RM) and details of the instrument have been described by Cush et al.23 Polyclonal R-BG was immobilized onto the surface of a low molecular weight (T70) carboxymethylated dextran surfacecoated cuvette (Labsystems, Affinity Sensors, Cambridge, U.K.). The sensor surface was washed 3 times with PBST and then activated using 0.2 M 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide-0.05 M N-hydroxysulfosuccinimide sodium salt for 5 min. The surface was again washed 3 times with PBST and 3 times with 10 mM acetate buffer, pH 4.5. The 100 µg/mL R-BG was added in 10 mM acetate buffer pH 4.5 and incubated for 15 min. The surface was then washed 3 times with PBST. Unreacted groups were blocked with 1 M ethanolamine pH 8.5, and the surface was washed 3 times with PBST. Selected samples were added at appropriate concentrations and incubated for 10 min; the surface was washed 3 times with PBST, to remove bound antibody; the surface was regenerated with 20 mM potassium hydroxide and washed 3 times with PBST. Each dilution of a sample was tested in triplicate. Changes in resonance position are measured in arc seconds, where 1 arc s is 1/3600 of 1° and the angle formed by a full circle is 360°. (22) Bollag, D. M.; Rozycki, M. D.; Edelstein, S. J. In Protein Methods, 2nd ed.; Bollag, D. M., Ed.; Wiley-Liss: New York, 1996; pp 57-81. (23) Cush, R.; Cronin, J. M.; Stewart, W. J.; Maule, C. H.; Molloy, J.; Goddard, N. J. Biosens. Bioelectron. 1993, 8, 347-354.
Analytical Chemistry, Vol. 77, No. 22, November 15, 2005
7243
Figure 3. Protein release from 109 cfu/mL BG spore suspensions sonicated for various times measured using the Lowry method. Error bars represent 1 standard error of the mean (SEM); number of data (n) at each time are 0 s, n ) 10 (2 samples in duplicate and 2 samples in triplicate); 30 s, n ) 6 (3 samples in duplicate); 60 s, n ) 8 (4 samples in duplicate); 120 s, n ) 6 (3 samples in duplicate); and 300 s, n ) 6 (3 samples in duplicate).
Figure 2. TEM analysis of BG spores (a) before and (b) after 300-s sonication. Arrows on (b) indicate reduction in spore coat and increased debris. Scale bars represent 1 µm.
Temperature Monitoring. Sample temperature in the tubular sonicator was measured using a 4.7-kΩ calibrated thermistor bead positioned in an approximately central location. Readings were taken before and after batch sonication and at selected times during continuous sonication for up to 6 h. RESULTS AND DISCUSSION Selection of Sonication Frequency. The one-dimensional modeling program used predicted a strong resonance around 270 kHz in the system. Stereomicroscopic observation around the predicted resonance clearly demonstrated that strong acoustic cavitation was occurring at a frequency of 267 kHz; the characteristic bubble activity was accompanied by the noise associated with acoustic cavitation. Erythrocyte suspensions were exposed to acoustic cavitation at the operationally selected frequency of 267 kHz in the system achieving over 99% cell disruption in 3 s, confirming occurrence of strong acoustic cavitation at the resonance. Loss of Viability and Alteration to Spore Structure. Viable counts of BG spores sonicated in the tubular sonicator showed no significant decrease in viability after any exposure time tested, suggesting the spores were not broken open. TEM analysis of untreated BG spores revealed the presence of a dense outer spore coat and a small amount of debris (Figure 2a). After 300-s sonication, the integrity of the spore coat was reduced and the amount of debris in the sample had increased due to disruption of the spore coat (Figure 2b). There was no evidence for breakage of the BG spores (Figure 2b), in agreement with the viable count data. Protein release from the spores, measured using the Lowry method, increased with increasing exposure times (Figure 3), reaching a plateau value after 120 s. The detection of protein 7244 Analytical Chemistry, Vol. 77, No. 22, November 15, 2005
release from sonicated spores is consistent with electron microscopic observations of spore coat removal. Detection of Spore Antigen by Immunoassay. The threshold absorbance for antigen detection by enzyme-linked immunosorbent assay (ELISA) was defined here as twice the background absorbance level. This absorbance corresponded to the signal from 107 cfu/mL unsonicated BG spores ([S]unson thresh). The ultrasoundinduced gain in antigen detection was defined as 107/([S]son thresh), where [S]son thresh is the sonicated spore concentration giving the threshold absorbance level. Batch, uncooled, sonication of BG spores for times up to 300 s all showed at least a 10-fold increase in direct ELISA response (Figure 4a). Assay sensitivity was highest after 30 s, achieving an ∼20-fold increase in sensitivity (Figure 4a). It is noted that as microbial disruption by ultrasound is independent of cell concentration25 similar gains in sensitivity can be expected when concentrations lower than 109/mL are employed. Since the sensitivity gain is ×20 and the ELISA detection threshold sensitivity signal has been operationally defined above as that associated with detection of a 107/mL concentration of unsonicated spores, it follows that a clear signal would be detected from 107/20, or 5 × 105 spores/mL. Increases in ELISA sensitivity similar to those for 30-s batch sonication were achieved when a cooling system and flow-through sample processing were applied. The highest response was achieved at a flow rate of 0.3 mL/min (Figure 4a), equivalent to a residence time of 120-s batch sonication. This flow rate was used in all further flow experiments. The increase in measurable protein (Figure 3) but decrease in ELISA absorbance (Figure 4a) with longer batch sonication times suggests antigen denaturation or fragmentation is occurring either due to heating or to the sonication itself. Thermistor monitoring of the temperature in the sample chamber in the absence of air cooling shows increases of 3, 10, and 22 K after 30-, 120-, and 300-s batch sonication.21 The temperature increase was 15 K after 20-min batch sonication with cooling and remained at that equilibrium level for the following 40 min. Batch samples sonicated with and without cooling for 300 s showed similar ELISA responses, suggesting that thermal denaturation is not responsible for the reduction in ELISA absorbance over longer exposure times. (24) Borthwick, K. A. J.; Love, T.; McDonnell, M.; Coakley, W. T. Proc. of The 5th World Congress on Ultrasonics 2003; pp 1157-1160. (25) James, C. J.; Coakley, W. T.; Hughes, D. E. Biotechnol. Bioeng. 1972, 14, 33.
Figure 4. Detection of BG spore antigen by (a) direct ELISA; batch sample sonication without cooling for 0 ([), 30 (9), and 300 s (2), and 0.3 mL/min flow sonication with cooling (×). Error bars represent 1 SEM; for 0 s, n ) 14 (7 samples in duplicate), and for all other times n ) 6 (3 samples in duplicate). The dashed horizontal line (- - ) represents the threshold of detection. (b) RM sensor; sample sonicated with cooling for 120 s (9) and unsonicated sample ([). Error bars represent 1 SEM; for all samples n ) 6 (2 samples in triplicate). NB: Small error bars may be obscured by large symbol size.
Therefore, in batch mode, short sonication times (30 s) with minimal temperature rises (3 K) were sufficient to give the greatest increases in ELISA detection. With cooling and a 0.3 mL/ min flow rate applied, a 14 K temperature rise was observed and maintained over a 6-h sonication time.21 Under these flow conditions (giving a sample residence time of 120 s in the sonicator), optimal increases in ELISA detection were achieved over long total sonication times (e.g., 16 min 40 s for a 5-mL sample volume), for which temperature rises were within protein-tolerable ranges. This result illustrates that the tubular sonicator is capable of sonicating samples over long time periods, in flow, while maintaining a temperature that would not induce thermal denaturation for immunological detection methods. This property of the tubular system relates to the efficient and focused deposition of energy into the axial region where the sample is located.24 The thermal mass of the assembly is also greater than that normally associated with the sample region of conventional 20-kHz cell disruptors while the surface area is greater, leading to efficient application of air cooling. These properties contrast with conventional sonication while at the same time the physical size of the tubular transducer, coupling layer, and sample region is much less than that of a conventional sonicator. A direct ELISA was performed on the whole sample, supernatant, and resuspended pellet of unsonicated (Figure 5a) and 30-s batch sonicated (Figure 5b) BG spores. The results showed an increase in ELISA absorbance of the sample supernatant similar to that found in the whole samples. In contrast, the resuspended pellet showed little or no increase in absorbance after sonication. Improvements in the ELISA detection sensitivity after sonication
Figure 5. ELISA on whole sample ([), supernatant (9), and resuspended pellet (2) of diluted 109 cfu/mL BG spore samples: (a) unsonicated control and (b) 30-s batch sonicated sample. Error bars represent 1 SEM; for all samples n ) 4 (2 samples in duplicate). The dashed horizontal line (- - -) represents the threshold of detection. NB: Small error bars may be obscured by large symbol size.
are, therefore, associated with an increase in soluble antigens and fragments available in suspension to bind to the antibody. This suggests sonicated BG spore samples containing small antigenic fragments would be detectable using the RM sensor, whereas unsonicated whole spores are not detectable due to their large size. RM detection of analyte after sonication of BG spores for various times showed an improvement in assay sensitivity to a detection limit of ∼106 spores/mL. Figure 4b shows the response for a 120-s batch sonicated sample as well as an unsonicated control. Increasing sonication time and use of flow sonication did not affect the detection limit achieved, although there was a slightly higher response at lower concentrations, suggesting a higher number of detectable fragments may be available for binding. CONCLUSIONS The 267-kHz tubular sonicator increased the sensitivity of immunoassay-based detection of BG spore antigen by release of soluble antigens and possibly fragments from the spore surface rather than by spore breakage. A cooling system ensured that protein denaturation did not occur due to excessive temperature rises during sonication over extended operating times. This novel tubular sonicator has the potential for development to a flowthrough, in-line system that can sonicate small or large sample volumes within a rapid time frame to disrupt bacterial cells for increased sensitivity of immunoassay-based detection. ACKNOWLEDGMENT The UK Ministry of Defence funded this work. Thanks to S. Smith (Dstl, UK) for the electron microscopy images. Received for review April 5, 2005. Accepted September 1, 2005. AC050576C Analytical Chemistry, Vol. 77, No. 22, November 15, 2005
7245
