288432 Sensitive Piezoelectric Cantilever Sensors for DNA, Parasites and Toxins

Thursday, November 1, 2012: 10:36 AM
408 (Convention Center )
Kishan Rijal, Sen Xu and Raj Mutharasan, Department of Chemical and Biological Engineering, Drexel University, Philadelphia, PA

Abstract. Macro-scale piezoelectric self-sensing and self excited cantilever sensors that exhibit high-order modes near ~ 1 MHz are shown to exhibit sub-femtogram sensitivity in both ssDNA immobilization and hybridization assays in flow format.  We show in this presentation that the sensors are able to (1) detect the presence of single nucleotide polymorphism at 1 fM, (2) conduct polymerase chain extension reaction on the sensor, and (3) detect stx2 gene derived from the pathogen E. coli O157:H7 at 400 cells, (4) detect waterborne parasites such as Cryptosporidium parvum and Giardia lamblia, and toxins such as microcystins. Examination of sensor response in presence of non-target proteins and extraneous material is also characterized.

Introduction. Currently, detection of DNA and RNA oligomers in complex matrix containing extraneous nucleic acid and non-target proteins is of great interest in medical diagnostics.  Also of interest is the detection of waterborne parasites and toxins. These applications are as varied as identification of pathogen to tissue dependent microRNA (miRNA) expression. The latter has important medical applications in diagnosis of cancer and degenerative neurological disorders.  Resonant-mode cantilever sensors have been shown to be highly sensitive in liquid environments and give real-time measurement of DNA hybridization in situ under continuously flowing liquid conditions.  We examine application of piezoelectric-excited millimeter-sized cantilever (PEMC) sensors for detecting both natural and synthetic ssDNA and RNA.  We also examined detection sensitivity for the waterborne parasites in natural river water and in waters spiked with non-target proteins.

Experimental. Piezoelectric cantilever sensors were fabricated as described earlier1.  These sensors have been characterized by various means to show sensitivity at femtogram levels2-6.  A typical detection experiment begins with securing the sensor in a custom isothermal flow cell at a flow rate of ~ 0.5 mL/min and 32 °C.  After initial steady-state in resonant frequency is reached the PEMC gold surface was functionalized with thiolated DNA probe and then remaining gold sites were back-filled with 6-mercaptohexanol (MCH) to prevent non-specific adsorption and optimize secondary structure of the functionalized layer for target hybridization.  Subsequent to surface functionalization, the circulating buffer was changed to sample containing target nucleic acid (e.g. stx2 gene derived from E. coli O157:H7, others).  In some experiments, after the target hybridization step, the four dNTP and Taq polymerase was introduced to observe and measure increase in sensor response as the probe ssDNA was extended along the target strand. Experiments with target sequence containing SNP in the middle or at the end of the probe sequence was carried out to determine fidelity in sensor response.

For detecting parasites, source water (tap, river) was spiked with Cryptosporidium parvum and Giardia lamblia at of 5 to 10,000 (oo)cysts/mL and microcystins LR in the range of 1 pg/mL 1o 1 ng/mL in both tap water and then exposed to PEMC sensors functionalized with the corresponding IgG antibodies and installed in a flow cell for in-liquid measurement

Results and Discussion. Typically, ~ 2 mm2 gold layer is used as the PEMC functionalized surface area.  Based on the average diameter of the double helix (~ 2 nm), we estimate that a maximum ~ 1012 ssDNA probes (~ 7.4 ng for a 7 KDa ssDNA) can be chemisorbed, assuming a defect free Au-surface.   The following are salient results:  [1] E. coli O157:H7 at various concentrations were mixed with ground beef mix was used as test natural sample.  A short heat-and-chill procedure, and mechanical shearing method was developed to extract genomic DNA, and various concentrations representing from 400 to 5000 cells were prepared, and was exposed to sensors prepared with an immobilized ssDNA strand (designed with Primer criteria) for detecting stx2 gene.  Control experiments with a wild strain of E. coli gave no response while the pathogen have over 400 Hz response for the 400 cell sample, and over 1600 Hz shift for sample containing 4,500 cells.  [2] A 288-nt ssDNA was first hybridized to the sensor immobilized with an 18-mer ssDNA probe which caused a shift down of 1300 Hz for a target of 3 aM.  The position of hybridization left ssDNA containing 103 nucleotides.  Subsequent addition of the four dNTPs and Taq polymerase cased a further shift of 1800 Hz indicating the increased in mass due to polymerase reaction.  Polymerase reaction was confirmed by lack of hybridization and hybridization with a 20-mer that was complementary to the exposed 103-nt strand.  [3]. Hybridization experiments with SNP located in the middle of a 10-mer gave no response, while it being present in the middle of a 20-mer and larger gave a small response of a few hundred Hz.  Subsequent exposure to complementary strand gave full hybridization response.  [4]. Detection of C. parvum was done both in PBS and in 25% milk background at 1 mL/min. The limit of detection was 5 oocysts in 25% milk. Following the detection step, a second IgM antibody specific to Cryptosporidium was exposed to the sensor for confirming detection. The response correlated with C. parvum concentration tested, and served as a method for reducing false positive sensor response. [5] Detection of G. lamblia was carried out at 0.5 mL/min to 5.0 mL/min in a similar manner. Detection of as few as 10 cysts is feasible in 30 min.  Higher flow rates gave a higher sensor response magnitude and provided improved signal-to-noise ratio. The specificity of the sensor was challenged with non-target cells (E. coli O157:H7 and C. parvum). Detection was confirmed with environmental scanning electron microscopy. Successful detection of G. lamblia cysts spiked tap water and Schuylkill River (Philadelphia, PA) water showed the feasibility of direct assay of source water samples without a pre-treatment step. One liter samples at 1 cyst/mL or 10 cysts/mL were successfully analyzed.

Conclusion. In summary, we showed PEMC sensors are extremely sensitive and selective, yet robust enough for rapid detection of waterborne parasites. Presence of non-target strands does not interfere with hybridization. Acknowledgement.  The authors acknowledge financial support through EPA STAR Grant R833007 and NSF CBET-0828987.



1.         D. Maraldo, K. Rijal, G. Campbell and R. Mutharasan, Anal. Chem., 2007, 79, 2762-2770.

2.         G. A. Campbell and R. Mutharasan, Anal. Chem., 2007, 79, 1145-1152.

3.         D. Maraldo, F. U. Garcia and R. Mutharasan, Anal. Chem., 2007, 79, 7683-7690.

4.         D. Maraldo and R. Mutharasan, Anal. Chem., 2007, 79, 7636-7643.

5.         D. Maraldo and R. Mutharasan, Sens. Actuator B-Chem., 2008, 132, 140-148.

6.         K. Rijal and R. Mutharasan, Langmuir, 2007, 23, 6856-6863.

Extended Abstract: File Not Uploaded