.rvwrzw' m ' w!“ wmmvurz'” "7'“ I’m-'1’: W . ‘34 _ .._ Ef""..'1 ‘ | in ..y.+: . Vznfl. I “‘1‘. A” ““0": . _ 4- ' f In,“ ‘ ”I1 n. ’2: A. 12“] ' ‘_ : é'r :‘r 2‘1"- .. I ; 1' )1;ka . R1223} 1.3;” :3‘. " ' ' " ‘ ‘ ’ 1; . I ‘L. M ‘3 '0“? "z'. 2‘ o H..- L”: I“; in“ ’ ‘1 1‘ Inn ’ u "'f"( I I :32!“ zz‘u‘th \u‘ 1‘ . - 1- 1:! 31‘” {-3}. I I “ P 1 . {mu-3.? ii: KJLI'oI‘ (I’LL “an, 23.5; Ifi > ' " ““"jl '1‘. far 19‘ 1” g x F ‘ 21"» 3%.... M" "ta-"U W19“ {'mi” NV n‘r‘ 7r - w I ' “I‘ n". A u. :gt' . h‘ '0 f“: I," ' "' n ' t I . v ’.-Y"Ih' A. "‘ ' W t ~~ ‘ |‘ n: ‘1' m ~‘- INH' "3 J' «b . - o “' )‘A'. .‘0. I0 . Alt! . "‘lt’J 'o ‘13:“) :‘1‘4. .. . '. I v. It in - 2:? 1M ’. .' 2| .t an $53" “" 1W u;‘2 :'l‘\l *1 17.1}; -£ . .-‘_ "Y. m LIBRARY Michigan State University J L/ K._) C\ ’79 This is to certify that the dissertation entitled DNA BASED BIOSENSOR FOR THE DETECTION OF ESCHERICHIA COLI IN WATER SAMPLES presented by Maria I. Rodriguez- L6pez has been accepted towards fulfillment of the requirements for the Doctoral degree in Biosystems and Agricultural Engineering and Civil and Environmental Engineering ‘ é Major Professorial Signature /.1 f / 3/ 0 7 Date MSU is an affinnative-action, equal-opportunity employer PLACE IN RETURN BOX to remove thlS checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED wrth earlier due date if requested. DAIEDUE DAIEDUE DAIEDUE 6.07 p £1;ch Date-Due moo-p 1 DNA BASED BIOSENSOR FOR TIIE DETECTION OF ESCHERICHIA COLI IN WATER SAMPLES By Maria I Rodriguez-Lopez A DISSERTATION Submitted to Michigan State University in partial fulfillment ofthe requirements for the degree of DOCTOR OF PHILOSOPHY Departments of Biosystems and Agricultural Engineering and Civil and Environmental Engineering 2007 ABSTRACT DNA BASED BIOSENSOR FOR THE DETECTION OF ESCHERICHIA ('01.! IN WATER SAMPLES By Maria I Rodriguez-Lopez The principal goal of this research was to demonstrate the efficiency and capability of a model biosensor using Escherichia Cali DNA synthetic oligonucleotides for fast and accurate detection E. coli DNA. Molecular biology and cyclic voltammetry (CV) were combined to develop and test a model DNA-based biosensor. The hybridization capability of embedded DNA into polypyrrole (PPY) with complementary DNA samples was determined. The biosensor platform evaluated was a Platinum (Pt) electrode elctropolymerized with PPY. The recognition elements were oligonucleotides specific for [5-D glucuronidase. The biosensor was capable to generate distinctive CV signals for complementary and non-complementary DNA sequences. Cyclic voltammetry scanning between 0.0 and +0.70 V and 50 mV/s scanning rate were used to generate current vs. potential graphs. A range of DNA concentration of 10'6 g to 10'9 g was used to determine the hybridization signal recognition ofthe biosensor. Distinctive hybridization signals were obtained after 30 minutes hybridization time. The biosensor platform proved to be effective in the detection of complementary uidA 25 bp oligonucleotide and genomic DNA from E. coli K-12. The biosensor was successful in discriminating for cross hybridization using Salmonella ()phimurizun and C ampyluhucler jejuni. Genomic DNA isolated from natural waters demonstrated the capability of the biosensor to detect E. coli from environmental isolates. The total detection time took 40 minutes after sample preparation. DEDICATION To my mother Syra. Porque plantaste la semilla y aunque no viste su fruto aqui en la tierra. se que lo veras en el cielo. Feliz cumpleanos. To my father Vicente. Gracias por tu apoyo. tu amor y tu paciencia incondicional. iv ACKNOWLEDGMENTS I have so many people that I need to thank for this accomplishment that is hard to include everybody that contributed to my success. I would like to start with Dr. Evangelyn Alocilja. Excellent professional and on e of the best persons I have ever met. Vangie, thanks for your constant guidance, support, patience and believe that this project could be done, even when we started with nothing. Thanks for being there for my academic and personal life and thanks for all the prayers, I hope that you still pray for all of your student even after many years. Dr. Susan Masten, thank you for your support and for coming up with the idea of the dual degree. Also thanks you for all the great comments in the revision process. Dr James Tiedje. thank you for giving me my first opportunity in research as an undergraduate back in 1992. Also thank you for providing assistance and the use of the CME facilities. Dr. Dan Guyer. thanks for the input in the electrochemistry and the statistics part. Dr. Volodymyr Tarabara, thank you for your participation and valuable comments. To the Dr. Barbara O'Kelly and Dr. Percy Pierre. thanks for your constant support and overlooking our success as part as the SLOAN Program. Also many thanks to Dr. Greg Swain in the Chemistry department for the use oftheir facilities and the help ofyour graduate students in the electrochemistry portion of this project. To my friends and laboratory partners. Finny. Lisa, Zarini and Steve. Thanks for making the everyday so nice and listening to all my stories. Sorry for all the time that my stories distracted you. Lisa. I miss your desserts! To my undergraduate students Rahul and Emma, thanks for your hard work. To my friends at MSU, Camten, Lisveth, Asa, Karen and David, thanks for all the good times and your helping hand in the not so good times too. To my sisters Dagmar, Samari and Alina, thank you. To my family in Puerto Rico, thank you Dad for your constant love and inspiration. You are the best! Sergio, you are my fresh air, my companion and my love. Thank you for your patience in the last stage of this long race. Thanks for the final push. Te amo. And to my Lord Almighty, you showed me the way. please help me always follow your steps! Thank you! vi TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION - Hypothesis 1 Hypothesis 2 LITERATURE REVIEW _ _ - _ - _ - - ........... CHAPTER 1: WATER QUALITY REGULATIONS Section 1.1. Historical Background on Water Quality Regulations and Outcomes Osman—EX _- 7 Section 1.2. Fecal Coliforms - ............... 10 CHAPTER 2: CLASICAL METHODS OF ISOLATON, QUANTIFICATION AND IDENTIFICATION OF WATER PATHOGENS 13 Section 2.1 Classical Cultivation Techniques 14 Multiple-tube—fermentation technique (MTF) .......................................................... l4 Membrane Filtration Technique (MF) ...................................................................... l4 Section 2.2. Enzymatic approach 16 Section 2.3 Molecular Biology Approach - . ........ - 17 Immunoassays ........................................................................................................... l7 Nucleic Acids Approach ........................................................................................... 18 CHAPTER 3: HISTORICAL BACKGROUND ON CONDUCTIVE POLYMERS ................. - 24 Section 3.1. Historical Background - - 25 Section 3.2. Electrochemistry of polypyrrole 26 Stoichiometric Polymerization Reaction .................................................................. 27 Electroactivity and Conductivity .............................................................................. 30 Section 3.3. Practical Applications of Intrinsic Conductive Polymers (ICP) ....... 30 Section 3.4. Electrochemichal Techniques 32 Potentiometry ............................................................................................................ 32 Amperometry 32 Conductimetry ........................................................................................................... 33 Voltammetry ............................................................................................................. 33 Section 3.5. Cyclic Voltammetry - - - - _ ........ - - 34 Section 3.6 Instrumentation 39 Potentiostats .............................................................................................................. 39 vii CHAPTER 4. BIOSENSORS ....... - - . 42 Section 4.1 Principles of Biological Sensors - 43 Biological Sensor Specificity .................................................................................... 44 Section 4.2. Catalytic Sensors - -- 45 Section 4.3. DNA biosensors _ 46 DNA specific redox indicators .................................................................................. 48 DNA biosensors for the monitoring of environmental pollution .............................. 49 Platinum-PPY-DNA biosensors ............................................................................... 51 Chapter 5: Research Methods and Materials - . 55 Section 5.1. - 56 Fabrication of the Pt-PPY-uidA biosensor by incorporation of an uidA gene oligonucleotide into a conductive polymer-electrode biosensor system. ................ 56 Selection of DNA sequence for the detection of E. coli ........................................... 56 Incorporation of the uidA gene oligonucleotide onto the Pt-PPY electrode ............ 56 Physical characterization of the modified DNA-PPY electrode surface .................. 58 Section 5.2 - _- - 59 Functionalization and sensitivity analyses of the DNA biosensor using different concentrations of oligonucleotides. _ 59 Sensitivity analysis of the E. coli DNA biosensor .................................................... 59 Determination of CV and Hybridization Conditions ................................................ 60 Physical Characterization of the modified DNA-PPY electrode surface ................. 63 Statistical Analysis .................................................................................................... 63 Section 5.3. - -- -- -- -_ - - 65 Specificity of the biosensor using DNA from E. coli pure culture and from other common waterborne pathogenic microorganisms. - _ - 65 DNA extraction from pure cultures of reference strains ........................................... 65 Hybridization of pure DNA cultures and analysis using cyclic voltammetry. ......... 66 Statistical Analysis .................................................................................................... 67 Section 5. 4 - _ - - _ - - 68 Performance and stability of the biosensor in the presence of environmentally isolated total DNA from surface water samples- 68 Sample collection ...................................................................................................... 68 DNA extraction from water samples ........................................................................ 69 Statistical Analysis .................................................................................................... 69 CHAPTER 6. RESULTS AND DISCUSSION _- ........... - 70 Section 6.1 -- - __ - - - - .......... 71 Incorporation of a uidA gene oligonucleotide into a polypyrrole-coated platinum electrode biosensor system. __ ..... - 71 Electropolymerization of PPY .................................................................................. 71 Functionalization of the Pt-PPY-uidA biosensor ...................................................... 72 viii Physical Characterization ofthe modified DNA-PPY electrode surface ................. 74 Specificity of the uidA probe. ................................................................................... 75 Section 6.2. -- - - - _ 77 Functionality, selectivity and sensitivity of the DNA Biosensor using different values of key variables. 77 Delta Charge (AQ) analysis for the normalization of CV signals ............................. 81 Section 6.3 _ ........ -A - ...... - - 94 Specificity of the biosensor using DNA from E. coli pure culture and from other common waterborne pathogenic microorganisms. 94 Cyclic voltammetry analysis for the hybridization of E. coli K-12 genomic DNA and other common water pathogens. ............................................................................... 94 Delta Charge (AQ ) analysis for the normalization of CV signals ............................ 96 Section 6.4 - -_ -- _ - -- - -- - - - _- 103 Performance and stability of the biosensor in the presence of environmentally isolated total DNA from surface water samples. -- 103 Biological and chemical characterization of water samples. .................................. 103 CV profiles of total genomic DNA isolated from water samples. .......................... 105 Analysis of AQ values using ANOVA test. ............................................................ 106 CHAPTER 7. CONCLUSIONS - - - ...... ..... - 114 CONCLUSIONS - -- -- - - 115 CHAPTER 8. FUTURE RESEARCH - -- _ - _ _ _ - 117 CHAPTER 9. APPENDIX 119 CHAPTER 10. REFERENCES---.-- - - -- -- - ............ - - 152 ix LIST OF TABLES page Table 1.1 Total coliform and E. coli limits from US and International regulation agencies ........................................................................... 11 Table 4.1 Recent studies involving the use of DNA-PPY-biosensor and its different applications ........................................................................ 54 Table 6.1 Paramenters used for ANOVA analysis of 25 bp oligonucleotides... 84 Table 6.2 ANOVA analysis and significance values for 25 bp oligonucleotides 84 Table 6.3 Average A0 at different times and concentrations for 25bp oligonucleotides .............................................................................. 84 Table 6.4. Average delta charge (AQ) for all genomic DNA isolates .............. 98 Table 6.5 Comparison of AQ change percentage among signals and significance after Tukey-Kramer adjustment .......................................... 98 LIST OF FIGURES Figure 3.1 . Electrochemical cell configuration ....................................... Figure 3.2. Electrochemical polymerization of pyrrole. Adapted from Genies, et al.1983 ........................................................................... Figure 3.3 .Potential vs. Time in Cyclic Voltammetry .............................. Figure 3.4. Cyclic voltammograms of a reversible redox process ............... Figure 3.5. Irreversible (A) and quasi- reversible (B) cyclic voltammograms for redox processes ......................................................................... Figure 3.6. Schematic of a Potentiostat connected to a three cell electrochemical cell ........................................................................ Figure 4.1. Transmission electron microscopy (TEM) of a PPY film. The dark spaces within the film are the positively charged regions called polarons ....................................................................................... Figure 5.1. Electrochemical cell configuration ....................................... Figure 5.2. Functionalization of Pt-PPY electrode with the 25 bp uidA probe, followed by the hybridization of the functionalized working electrode with a complementary 25 bp oligonucleotide ......................................... Figure 6.1. Polymerization of 0.5M PPY/0.5 M KC1 onto Pt electrode. The resulting curves are cyclic voltammograms after 26(a) and 13 (b) cycles between 0.0 and 0.7 V at a scanning rate of 50mV/s ................................. Figure 6.2. Comparative CV electrodeposition for 111g of total DNA.Cyelic voltammograms after 26 cycle between 0.0 and 0.7 V at a scanning rate of 50mV/s for a) blank solution 0.1M glycine/0.1M NaCl, b) polymerization of PPY 0.05M/0.5 M KCI, complementary (c) and non complementary (d) uidA probe (lug total) in 0.1Mglycine/0.1MnaCl .................................... Figure 6.3. Subtractive CV of complementary and non—complementary oligonucleotides targeting E. coli uidA gene fragment. Potential range from 0.0 and 0.7 V, scanning rate of 50mV/s in 0.05 M PPY/0.5 M KC1; cyclic voltammograms after 26 cycles .......................................................... xi Page 27 29 35 36 38 40 51 57 62 71 73 74 Figure 6.4. Scanning electron microscopy (SEM) of bare Pt (A), and modified Pt surface (B) ................................................................... Figure 6.5. Subtractive CVs for different concentrations of Complementary and Non-Complementary Oligonucleotides. Hybridization temperature was 72°C ..................................................................................... Figure 6.6. Subtractive CV of all Complementary and Non-Complementary oligonucleotides at different concentrations .......................................... Figure 6.7. CVs for 1 pg of total complementary oligonucleotides at different hybridization times ............................................................ Figure 6.8. CVs for 100ng of complementary oligonucleotides at different hybridization times ......................................................................... Figure 6.9. Average AQ for lpg of Complementary and Non- Complementary probes at different hybridization times ........................... Figure 6.10. Average AQ of 100ng of Complementary and Non- complementary probes at different hybridization times ........................... Figure 6.11. Comparison of average AQ values with respect to oligonucleotide concentrations and hybridization times ........................... Figure 6.12. Subtractive CV Signals for 100ng of genomic DNA from pure cultures of common water pathogens after 30 minutes hybridization time ........................................................................................... Figure 6.13. CV Signals for 100ng of genomic DNA from pure cultures of common water pathogens after 30 minutes hybridization time ................... Figure 6.14. Comparison of average delta Q for 100ng of synthetic oligonucleotides ............................................................................ Figure 6.15 Comparison of average delta Q for 100ng of genomic DNA from bacterial pathogens Figure 6.16. Average E. coli CFU/100ml of water samples ........................ Figure 6.17. Turbidity measurements in NTU from water samples ............. Figure 6.18. Subtractive CVs for 100ng of genomic DNA from pure cultures of common water pathogens and genomic DNA from environmental water samples after 30 minutes hybridization time .......................................... xii 75 87 88 89 90 91 92 93 99 100 101 102 109 110 111 Figure 6.19. CV of hybridization with genomic DNA from water pathogens and total DNA from water samples ..................................................... 112 Figure 6.20. Comparison of average AQ for 100ng of synthetic oligonucleotides, genomic DNA from bacterial pathogens and water isolates ........................................................................................ 1 13 xiii INTRODUCTION Safe and clean water is a requirement for good public and environmental health. The detection of coliform bacteria by culture techniques, such as Escherichia coli (E. coli), is a standard test for water quality assessment. The presence of colifomts in any water source may indicate ineffective treatment, loss of chemical disinfectants, bacterial breakthrough, and intrusion of contaminated water into potable water supply or re-growth problems in the system. According to the Centers for Disease Control and Prevention (CDC), there is a rise in watcrbome disease outbreaks. Reports from 1999 and 2000 (Lee 2002) show a total of 39 of these outbreaks involving drinking water in 25 states, more than double the 17 outbreaks reported in 1997-98 (Barwiek 2000). Even though the cases of waterbome outbreaks have decreased from 39 to 31 in the latest CDC report from 2002, outbreak cases in surface recreational water has been on the rise with 65 cases (Yoder 2004). Because ofthe increase in such water outbreaks. rapid detection methods for E. coli in drinking and recreational waters can minimize the impact of these outbreaks. Current detection methods recommended by the America Water Works Association (AWWA) are very sensitive but require 24 to 48 hours for confirmation and are labor intensive (APHA 1998). The development of analytical devices for rapid detection and monitoring of chemical and biological species has led to the emergence of biosensors. The biosensor technology promises to offer new detection altematives for E. coli and other pathogenic bacteria. A biosensor uses a specific biological recognition agent. such as enzymes or nucleic acid in the form of DNA oligonucleotides. in close proximity to a transducer and converts the recognition event between the recognition agent and the target analyte into a measurable signal. Conductive polymers, such as polypyrrole (PPY) and polyaniline (PANI), are being extensively researched for their application in biosensors. To improve the signal transduction of conductive polymers. a highly conductive electrode is also used such as gold, platinum and glassy carbon. Electrochemical techniques, such as cyclic voltammetry, provide a way for the formation of the polymer film into a solid electrode such the conductive metal. Traditional molecular biology tools such as DNA oligonucleotides can be used as the dopant for the polymer film as well as the recognition agent for the biosensor. Combining two traditional techniques, such as cyclic voltammetry and DNA oligonucleotides, gives rise to the innovative creation of DNA based biosensors and a proliferation of research options in this emerging technology. The use of such electrochemical and molecular biology techniques can also add the benefit of obtaining a recognition signals independent of expensive labeling chemicals. The merging of those two classical techniques offers some important benefits that are indispensable for biosensors, rapid detection times, and specificity for a target in cost efiicient manner. The use of conductive polymers and cyclic voltammetry to detect a target biomolecule such as DNA has led to the development of multiple biosensors for clinical applications. Recent interest of biosensor development extends to public health and environmental applications. The focus of our interest is the use of polypyrrole (PPY) in developing an E. coli biosensor for water quality monitoring. At the moment. no one has proven the IQ effectiveness of a platinum (Pt)- -PPY-DNA-biosensor using environmental DNA samples. The principal goal of this research was to develop a highly specific, sensitive. real-time DNA-based biosensor for the potential detection of fecal coliforms in water. Molecular biology and chemical electro-deposition techniques were combined to develop and test a DNA —based biosensor. The biosensor platform evaluated was a Platinum (Pt) electrode electro-polymerized with polypyrrole (PPY). The recognition element was a 25 base pair (bp) oligonucleotide specific for Escherichia coli. Nucleic acids isolated from pure E. coli cultures and from water samples served as the analyte. The specific objectives ofthis project were: I) To incorporate a uidA gene oligonucleotide into a polypyrrole-coated platinum electrode biosensor system. 2) To determine the functionality and sensitivity of the DNA biosensor using different concentrations of oligonucleotides. 3) To test the specificity of the biosensor using DNA from E. coli pure culture and other common waterborne pathogenic microorganisms. 4) To determine the performance and stability of the biosensor in the presence of environmentally isolated total DNA from surface water samples. Hypothesis 1 A 25 bp oligonucleotide (uidA gene) will be absorbed success/ally into polari:ecl regions of the PPY/Um using cyclic voltammetry due to electrostatic interactions between the negatively charged DNA and positively charged I’I’l'polarons. Hypothesis 2 The hybridization of complementary uidA oligonucleotides from E. coli K—IZ will be distinguished from non-complementary sequences due to the change in current (I) afier application ofa potential (I ") in a real-time hybridization event. LITERATURE REVIEW CHAPTER 1: WATER QUALITY REGULATIONS Section I. 1. Historical Background on Water Quality Regulations and Outcomes Lack of safe drinking water is one of the main causes worldwide for the high mortality rates (Beaglehole 2004). A significant fraction of the worldwide population is in need of safe water supplies for consumption. The World Health Organization has estimated that over 2 million people die from a disease caused by contamination of drinking water and improper sanitation a year. Annually, there are at least 200 million cases reported of diarrhea caused by waterborne nature and at least 2.1 million deaths are reported due to similar circumstances (WHO 2004) . For the last three decades there has been a global effort to identify and study a variety of waterborne agents that includes Escherichia coli, C r}ptosporidimn, Legionella, I'ibrio cholerae. hepatitis E virus, To.\‘()plctstmt, Helicobacter pylori among many others. Human activity near water supplies is one of the major contributors to the emergence and spread of waterborne disease agents. The impact of human activity and use of water supplies with some other economical and social practices facilitate the spread of the diseases. Other contributing factors are the lack of water treatments. and the lack of regulation and enforcement in public health systems. Human demographics (overpopulation) and human behavior (contamination of water sources, lack of sanitary practices) as well as international travel and commerce, are major contributors to the spread of diseases in developing countries In the United States, the number of waterborne disease outbreaks has declined over the last two decades, in part due to the efforts to enforce regulation of public health systems by the US Environmental Protection Agency (EPA) in conjunction with public health entities, water supply utilities and local community citizens. The study and reporting of waterborne disease outbreaks is performed by the Centers for Disease Control and Prevention (CDC). The development and enforcement of the Clean Water Act in 1972 has significantly helped the spread and control of the diseases and probably has contributed to the reduction of outbreaks in the last 30 years. Created in 1972, the Federal Water Pollution Control Act has passed through several revisions and amendments in 1977, 1981, 1987, 1990 and most recently, in 2002. The Act established the guidelines for regulating discharges of pollutants into the waters of the United States. It gave EPA the authority to implement pollution control programs, such as setting wastewater standards for industry. The Clean Water Act also set water quality standards for all contaminants in surface waters including microbiological agents. Since the 1970‘s, the CDC, the EPA, and State agencies have developed and maintained a surveillance system for collecting and reporting data associated with disease outbreaks associated to drinking water and recreational water. Both Federal agencies, as well as the drinking water industry, are collaborating to better estimate the risk of waterborne disease, develop better methods for detecting new pathogens in drinking water, and identify human and animal sources of water contamination. Despite their declining occurrence, 127 waterborne disease outbreaks occurred during the last decade (Barwiek 2000). A total of 31 outbreaks caused by drinking water were estimated during 2001- 2002 (Blackburn 2004). These 31 outbreaks caused health problems to 1.020 person and resulted in 7 casualties (Blackburn 2004). The data obtained in the surveillance for waterborne diseases showed that the agent that caused the outbreak was identified in 77% of the cases. Seventy nine percent (79%) of the cases were associated with microbial pathogens that include Legionella species, Shigella, .S‘almonella species, norovirus, Giardia intestinalis, erptosporidium, Nalgeria jowleri, Escherichia coli 01572117, Campylobacterjejani and l'ersinia enterocolitica. Even though there is a decrease of waterborne disease outbreaks from drinking water supplies, the cases of outbreaks mediated by recreational waters has been in the rise in the last 30 years. A total of 65 outbreaks were detected, affecting a total of 2,563 persons and causing the death of 8 individuals (Yoder IS 2004). Ofthese cases, 46.2% involved gastroenteritis, of which 50% were caused by Cryptosporidium and 25% were caused by E. coli. All the fatalities (12%) were attributed by amebic meningoencephalitis caused by A'aegleriafou'leri. All the cases involving toxigenic E. coli were reported in fresh water venues. The increase in the number of outbreaks from recreational water could probably be due to the increase of surveillance and reports at both the local and state levels combined with true increases of waterbome disease agents. The increase on recreational water outbreaks might serve as a motivation for the development of fast sensitive biosensors for the detection of waterborne disease agents in all water supplies. The detection and removal of pathogenic bacteria from water sources is a task that requires a great deal of effort from specialized personnel and is of major economic concern for the community and government agencies in charge of the process. One of the most common pathogens disseminated in water sources is a group called enteric pathogens. The presence of such pathogens is due to the human activity near or at water sources that produces fecal contamination. When enteric pathogens such as Escherichia coli (E. coli) and E. coli 01572117 are present in water sources. generally they are present in very low concentrations. Therefore, very specific and sensitive methods are needed to successfully detect them, and usually the task is very time consuming. Another more diverse group (Coliforms) is more commonly used as an indicator of water contamination. The presence of coliforms in any water source may indicate ineffective treatment, loss of disinfectant, bacterial breakthrough (McFeters et al. 1986; Geldreich. BE. 1992), intrusion of contaminated water into potable water supply (Clark 1980), or re-growth problems (LeChevallier 1990) in the distribution system. Section 1.2. Fecal Coliforms The coliform group includes organisms from various genera and species that belong to the Enterobacteriaceae family. Various definitions exist for the coliform group. The most widely used definition is the one described in the Standard Methods for the Examination of Water and Waste Water (APHA 1998). The coliform group is described as aerobic and facultative anaerobic. Gram negative. non-spore-forming. rod shaped bacteria (APHA 1998). Member of this group are lactose fermenters and produce gas within 48 hrs at 35°C. The most common way to identify this group is by the multiple tube fermentation technique in which the gas formation and acid formation are confirmed. Another very commonly used technique is membrane filtration. in which the water source is filtered and the bacterial cells are collected onto membranes that are incubated onto an agar plate with Endo Agar that contains lactose. After a 24-hour incubation at 35°C, red colonies with a metallic green sheen can be observed. Another criterion used for the characterization of the colifom1 group is the production of the B-D- galactosidase enzyme. This test is a confirmatory step after lactose fermentation. There 10 are variations within intemational communities about the definition of coliform groups. The United States and Canada follow the same definition as appeared in the Standard Methods for the Examination of Water and Waste Water. The European countries vary slightly depending on the regulation group. The French Standardization Association (AFNOR 1990) includes in their definition of coliforms oxidase-negative and the ability to grow in the presence of bile salts within 48 hrs at 37°C. AFNOR also includes the thermotolerant coliforms (fecal coliforms), specifically E. call. which is capable of growing at 44°C and retains its fermentation properties at this temperature. All regulation agencies from different countries are commonly enforcing very stringent guidelines for the presence of coliform in water sources (Table 1.1). Table 1.1. Total coliform and E. coli limits from US and international regulation agencies Country Limits on Limits on Population Samples/month total E. coli measurements coliforms United States a 0/100ml(95‘.’/b) OIIOOml 1/10000 inhabitants (100%) Canada b 0/100ml 0/100ml <5000 4 samplesz’month (90%) (10096) 5000-9000 1/ 1000 inhabitants >9000 90+(1/10.000 inhabitants) World Health Org 0/100ml 0/1001111 c (950/0) (1000/o) a US Environmental Protection Agency (1990) b Ministere de la Sante (1996) c World Health Organization (1994) Because of the constant presence of E. coli in warm-blooded animals‘ intestinal flora. this organism is the best indicator of fecal contamination from humans. 'l'herefore the absolute absence of E. coli in drinking water is a standard guideline. There are three 11 principal detection methods approved by the US EPA: the multiple tubes fermentation test, the membrane filter technique and the presence-absence test (ONPG-MUG). AFNOR approved only the first two methods. All these methods exhibit limitations such as long periods of incubation time (24-48 hrs). interference from antagonist organisms. lack of specificity to coliform group, and low levels of detection for slow growing organisms. The total growing population can be enumerated by these methods (Amann et al. 1990). Many factors including metabolic, nutritional and environmental factors can contribute to a non-cultivable status or an active but non-cultivable states (Roszak and Colwell 1987; Colwell 2000; Joux 2000). The use ofbiosensors could diminish the issue ofdetection of non-cultivable microorganisms because ofthe lack of growth using traditional methods. CHAPTER 2: CLASICAL METHODS OF ISOLATON, QUANTIFICATION AND IDENTIFICATION OF WATER PATHOGENS l3 Section 2.1 Classical Cultivation Techniques llIultiple-tubefermentation technique (MTF) Multiple-tube-femtentation technique (MTF) is one of the oldest techniques used (for over 80 years) for water quality monitoring purposes. The method utilizes serial decimal dilutions from the original water samples followed by the inoculation of each dilution into a lactose or lauryl tryptose broth. After 48 hours of incubation at 35°C, the production of gas and acid formation along with cell suspension constitutes a positive result. A confirmatory test is then performed checking for the formation of gas after incubating in a brilliant green lactose bile broth for 48 hours at 35°C. The MTF result is presented in the form of the most probable number (MPN). MPN is a statistically semi- quantitative test that estimates the number of cells (APIIA 1998). Several factors can contribute to the effective detection of coliforms using MTF. The presence of naturally occurring bacteria in environmental water samples can interfere with the detection of colifomts (Seidler et al. 1981) and the culture media can interfere with cell growth (McFetcrs et al. 1982). The extensive time (48 hours) from presumptive test to confirmatory test is significant. Nevertheless. this technique is still the preferred one over other methods for highly turbid water samples. Membrane Filtration Technique (.IIF) Membrane Filtration Technique (MF) is one of the most widely accepted techniques all over the world. A fixed volume of the water sample is filtered through a 0.45 pm pore size that holds the bacterial cells. Then the filter is placed onto an agar plate with a selective medium and incubated for 24 hours at 35°C. The most widely used medium in 14 the US is m-Endo-type (APIIA 1998) while the most common medium in Europe is the Tertigol-TTC (AFNOR 1990). There have been studies that demonstrate that Endo Agar yield a higher number of colonies that can be considered as false positives or negatives (Grabow and du Preez 1979). Because of the presence of false positives and false negatives. coliform verification is recommended after incubation with these media. However one of the drawbacks of these techniques is the interference of background microorganisms from the water samples with coliform growth (Clark 1980). An effective water treatment process introduces many chemical, mechanical and environmental factors that affect the survival of these microorganisms, as shown by several researchers. In some cases the microbial cells may be damaged but not completely disrupted. In such cases the presence of viable but non-cultivable microorganisms represents a serious problem for accurate quantification and detection. Chlorination processes can damage cells and increase the sensitivity to culture medium salts, therefore interfering with growth (LeChevallier et al. 1983; McFeters et al. 1986). Another effect caused by chlorination to the cells is catalase enzymatic activity inhibition (Calabrese and Bissonnette 1990). The inability to synthesize catalase produces an accumulation of hydrogen peroxide that is toxic to the cell (Sartory 1995). Other processes like ozonation are equally disrupting to the cells (Adams et al. 1989). All these techniques are extremely effective for the elimination of bacterial cells from water sources but because of the disruption effects on the cells. they represent a challenge in the quantification process of viable but non-cultivable organisms. The MF technique offers an advantage over the MTF because large volumes of water sample can be processed. The increase in volume translates into a higher yield of organisms and higher sensitivity. Quantification of coliforms is very reliable using this technique. However, the process takes 24 hours, and confirmation requires up to 48 hours. Section 2. 2. Enzymatic approach The MTF and MF techniques are based on metabolic reactions that can also be determined by the detection of the enzyme to be targeted. In general, these enzymatic reactions are fast and provide great sensitivity. One of the most common enzymes used to detect the coliform group and E. coli is B-galactosidase, which breaks down lactose into galactose and glucose. The other most common E. coli enzyme is B-D- glucuronidase. This particular enzyme breaks down 13-D-glucupyranosiduronic derivates into alycone and D-glucuronionic acid. This enzyme has been known to be specific to E. coli (Kilian and Bulow 1976). The methods for the enzymatic detection are based on chromogenic (color) or fiuorogenic (fluorescence) measurements. At the enzyme- substrate fomiation, there is a cleavage of the chromogenic or fiuorogcnic molecule from the compound, releasing immediate fluorescence (Chrdst 1991). Presence/absence and enumeration techniques have been combined with enzymatic methods such as incorporation of one fluorogenic substrate into the MTF technique (Feng and Hartman 1982). In the presence of E. coli, the hydrolysis of MUGlu («l-methyllumbelliferyl-B-D- glucuronide) releases a fluorescent compound visible under UV light. This technique needs at least 18 hours of incubation for positive results. A variation of this technique 16 was tested using an additional substrate called ONPG (O-Nitrophenyl-B-D Galactopyranoside) (Edberg and Edberg 1988). This technique showed a sensitivity of one colony forming unit (CFU)/100 ml water sample (Rice et al. 1990) and required 24 hours of incubation time. This technique has been developed into several commercially available tests like Colilert (IDEXX Laboratories, Portland, ME), Colisure (Millipore, Bedford, MA) and ColiQuick (Hach. Loveland, CO). A more recent test, Quanti-Tray (IDEXX) is used in the MPN format for estimated quantification. In conclusion, these enzyme-substrate compound methods have overcome the high throughput quality of traditional methods but still require 24 hours for positive results. They are very reliable and highly specific, but generally are expensive. Other methods that do not require cultivation steps have been developed more recently (George et al, 2000). One technique is based on the fiuorogenic detection of 13-D- glucuronidase and B-galactosidase from freshwater samples in 30 minutes. l-lowever. the detection limits of the technique are above the regulation standards (20 C FU/ 100ml for fecal coliforms and 340 C FU/ 100ml for total coliforms). Although these techniques have improved the efforts required from the traditional culture techniques, they are still not able to reliably detect non-cultivable or injured microorganisms. Section 2.3 Molecular Biology Approach 1mmunoassa_1’s Immunoassays are based on the rapid detection of a biomolecule without the need of a long incubation period. One of the most common methods involves immunological detection of antigen-antibody complexes. The use of a monoclonal antibody was 17 developed using enzyme —linked immunosorbent assay (ELISA) for the detection of enterobacterial common antigen from water samples (Obst et a1, 1989). In 1989, ELISA techniques had a high detection limit (105cells/ml) and required 24-hour incubation periods. The technique also exhibited cross reactivity with cells from the Pseudomonas and Aeromonas groups. This cross-reactivity makes it unreliable for evaluation of water samples since it will most likely react with non-target natural microfiora (Obst 1989). The production of these monoclonal antibodies is also very complicated and not cost- effective. Nucleic Acids Approach Nucleic acid hybridization can be defined as the recognition of two complementary sequences between a probe and a target. The hybridization can occur between two DNA- DNA molecules or two DNA-RNA molecules. A successful hybridization depends on the specificity and the degree of homologous sequences between probe and target hybrid. Most hybridization methods can be performed without cultivation steps and include the direct extraction of the nucleic acid from the environmental sample. The use of nucleic acids also enables the specificity of the organisms to be at the class. genera. species or subspecies level. This discussion will be focused on the two most current methods that are available for the application of water monitoring. Polymerase Chain Reaction (PCR) uses the amplification of a target fragment using different cycles of replication. The end product is an exponentially amplified quantity of the initial target molecule. The PCR amplification requires a set of primers that are specific for the target molecule. It also requires a thermally stable replication enzyme known as Taq polymerase. which can increase the cost of the test significantly This technique has been widely applied in the 18 amplification of coliform group from water samples (Bej et al. 1990). Other waterborne pathogens have also been detected using PC R (Burtscher et al. 1999; Waage et al. 1999). The design of specific primers for the detection of the coliform group has presented a challenge over the years mainly because of the variation in the definitions of the coliform groups discussed earlier in this chapter. For example, the design of primers specific for the lacZ gene has been used based on the detection of the B-galactosidase enzyme by enzymatic methods. Several researchers have been able to identify coliforms with a concentration as low as l cell/100ml (Bej et al. 1990; Bej et al. 1991b; Fricker and Fricker 1996). But results revealed a cross hybridization with non-coliform associated bacteria, thus more specific primers are needed. Other genes have been also used as targets, such as the malB gene that codes for a transport protein in E. coli and some strains of Shigella and Salmonella (Bej et al. 1990). The gene that codes for the enzyme B-D- glucuronidase, known as uidA, is commonly used for identification of diverse aquatic strains of E. coli. The use of the uidA gene was proposed by Bej et a1 (1991). Other researchers have successfully used the uidA gene in combination with another region of the same gene, uidR (Iqbal et a1. 1997). These primers are specific for both E. coli and Shigella and present a better alternative of the PC R method than using the lacZ gene mentioned previously. The use of the uidA gene presents an advantage over other gene segments because it is E. coli specific and it also detects the pathogenic strain of E. coli 01572H7. This pathogenic strain cannot be detected using the enzymatic reaction for B-D-glucuronidase. Although it contains the uia’A gene it does not express the enzymatic product (Feng et al. 1991). The identification of the 16SrRNA molecule offers great specificity and is an altemative for identification of microbial strains. A set of primers 19 have been designed for the DNA sequence that codes for the V3 and V6 regions of l6SrRNA of E. coli and Shigella (Tsen et al. 1998). The detection of the hybridization signal was usually corroborated using radionucleotide labeled probes like 32P. The PCR amplification usually takes 2-3 hours to complete. but the samples must be exposed to radioactivity for at least 24 hours to develop measurable signals. Faster, radioisotope- independent techniques have been developed that use biotinylated (luck et al. 1996) or fiuorochromes-bound primers. PC R based methods are very sensitive but they lack the ability for quantification or to discriminate between viable and non-viable organisms. Various studies attempted the PCR-quantification application for water samples (Toranzos et al. 1993; Zachar et al. 1993), but no technique has been found to be a reliable method that substitutes for the traditional water quality methods. The development of the real time PCR is a promising altemative for the use of molecular techniques in a quantitative way. Real time PCR uses fiuorochrome signals along the amplification process. The monitoring of the fluorescent signal is studied during the exponential growth phase of the PC R cycle (Heid et al. 1996). The real time PC R has been used regularly and successfully in the detection of clinically important pathogens such as enterohemorragic E. coli and enterotoxigenic E. coli (Carroll 2001). Several real time amplification systems are available such as GenAmp (Applied Biosystems) and LightCycler (Roche, Manheim, CA). In general, most nucleic acid molecular biology techniques are extremely sensitive but some limitations are encountered when dealing with environmentally isolated samples. The extraction of nucleic acids is performed to the total cells present in the sample. and PCR cannot differentiate between viable and non-viable cells. Another drawback of 20 nucleic acid methods is the reaction inhibition by chemical components in the water samples such as colloid material and humic acids (Way et al. 1993; Straub et al. 1995). Besides PCR, other nucleic acid methods include in situ hybridization techniques. The design of nucleic acid probes that target 16SrRNA molecules are the most common for in situ hybridization processes (Olsen et al. 1986). The primary application is for classification of microorganisms in philogenetic studies (Amann et al. 1990). Besides the high specificity of this molecule it also confers a sensitivity quality because of larger number of copies mitochondrial RNA per cell than of genes. For example, there are four copies ofthe uidA gene per cell versus 103 to 105 ribosomes per cell. Fluorescent in situ hybridization (FISH) labeled probes are very popular and offer various advantages such as being radioisotope free, thus safer to the researcher, and requiring less time to process the signal. The most common fluorescent probes are fiuorescein. rhodamine (Del.ong 1993), Cy3 and Cy5 (Wessendorf and Brelje 1992; Ouverney and Fuhrman 1999). The major disadvantage of fluorochrome dyes is that they are expensive and require expensive scanning devices to detect the signal intensity. Beyond the absence/ presence of the fluorescence, the analysis of data for quantification process requires training in data management. Another disadvantage in the use of FISH for water quality detection is the non-availability of a specific probe that identifies the total coliform group. There are probes available for the Enterobacteriaceae family (Mittelman et al. 1997) for clinical pathogens. Another group of FISH probes were developed for the examination of wastewater samples (Loge et al. 1999). More FISH probes specific to E. coli have been developed (Poulsen et al. 1994; Regnault et al. 2000). These probes have been used for the detection of E. coli from various sources such as clinical samples. natural and sewage 21 waters as well as food samples. A set of these probes was also designed for drinking water (Delabre 2001). The developed probes for drinking water examination proved not to be efficient due to the low number of ribosome copies obtained after water treatment processes (Lebaron et al. 1997). An application of FISH using manual epifiuorescence enumeration was obtained using environmental water samples with a detection limit of 1 cell/ml in 7 hours (Ootsubo et a1. 2003). Very recently, the combination of whole cell hybridization with a direct viable count (DVC-FISH) enabled the detection and enumeration of highly diluted viable Enterobacteriaceae cells in one day (Baudart et al. 2005). Still, these techniques involve the use of very expensive fluorescent dyes and expensive scanning equipment for laser scanning cytometry (LSC, ScanRDI, Chemunex, Ivry sur Seine, France). However, latest results indicate the rapid application of such biological techniques to water quality monitoring. Another interesting application of molecular biology to water monitoring has been the development of a peptide nucleic acid (PNA) for E. coli detection using in situ hybridization. The PNA is a synthetic nucleic acid where the sugar backbone has been replaced with a peptide backbone. This infers more stability to the nucleic acid and facilitates the incorporation of PAN into different substrates. This substitution confers more resistance to salt concentrations, more binding capacity of the PNA to any substrate, and shorter periods of hybridization (Fricker and Fricker 1996; Prescott and F ricker 1999). A drawback of the use of PNA is that they are detected after two weeks of water collection, an indication that the technique does not differentiate between viable and non-viable microorganisms even after a longer period oftime. The inability to distinguish between viable and non—viable organisms has been also addressed for the use of FISH and rRNA for monitoring contamination ofdrinking water. This technique has established that rRNA content in water samples may not reflect the true growth status of the cell. Ribosomal RNA molecules have been detected after chlorination, heat deactivation (Sheridan et al. 1998) and UV irradiation (McKillip et al. 1998). These molecular techniques have contributed to the application of nucleic acid techniques to more cost efficient, time efficient biosensor technology for fast and accurate monitoring of water systems. 23 CHAPTER 3: HISTORICAL BACKGROUND ON CONDUCTIVE POLYMERS 24 Section 3.1. Historical Background In 1910, Green and Woodhead reported the discovery of an aniline polymer that displayed an increased electrical conductivity after being treated with acetic acid. After these developments, scientists around the world worked to create organic substances (polymers) with metallic conductivity properties. In 1977 Shirakawa and Ikeda succeeded in polymerizing polyacetylene (PAc) in a film format (Shirakawa 1977). This was the first intrinsically conductive polymer (ICP) ever discussed in literature. Polypyrrole (PPY) is a polyheterocycline that has been extensively studied as a conductive polymer-forming film (Kanazawa 1979). One of its applications is the electrochemical deposition onto n-silicon for solar cell fabrication (Audebert 1985). Some researchers have used PPY films with a neurotransmitter as a mechanism for controlled relased of drugs into the brain (Zinger and Miller. 1984). An example of the multiple uses of conductive polymers is polyaniline (PANI), which has been used by Hitachi-Maxel for anti-static coating ofa 4 MB barium ferrite floppy disk (Friend 1993). The synthesis of PANI and PPY has been carried out on various substrates such as platinum (Pt), gold (Au), iron (Fe). aluminum (Al), stainless steel and carbon fibers, by both chemical and electrochemical methods. Electropolymerization is an effective technique for the deposition of polymer coatings onto various substrates (Su 1999). An example of these applications was demonstrated by the fabrication of an electrochemical (conductometric) biosensor using polyaniline molecules as its transducer by Dr. Evangeline Alocilja‘s laboratory group at Michigan State University. The lower limit of 25 its detection was determined to be 10I E. coli 01572117 colony-fonning units per ml (cfu/ml) in 6 minutes without use of reagents (Muhammad-Tahir and Alocilja 2002). Microbial testing in food is expected to increase due to food safety regulations (Alocilja and Radke 2003), and the electrochemical biosensor has great potential for commercialization for that application. Some other promising applications of biosensors are distributed among various fields, such as biotechnology, food and agriculture product processing, health care, medicine and environmental pollution monitoring. Section 3. 2. Electrochemistty of polypyrrole The oxidation of pyrrole to produce polypyrrole has an electrochemical stoichiometry different from more traditional electrochemical polymerization reactions. The bulk of the polymerization reaction takes place away from the electrode surfaces. The product is an electroactive film with conductive properties which is very stable and can be exposed to air. These free-standing films can be peeled off from a metal electrode and are easily manageable. The polymerization reaction proceeds via radical cation intermediates. The reaction is sensitive to the nucleophilic environment in the region near the electrode surface. limiting the choice of solvents and electrolyte. The electrolytic salt needs to be soluble. llalides are highly nucleophilic and easily oxidized. This interferes with the quality of the film production. Electrode-film preparation has been achieved best by using three cell electrodes (Figure 3.1). The nature of the working electrode is critical for the film preparation. The films 26 are produced by an oxidative process, so the electrode should not exhibit oxidative properties. Platinum or gold are excellent working electrodes. /—--»«Ag/AgC| ref. / P -— Carbon rod . ,N2/O. . 3 Av gas inlet [:L /// T ‘3‘ / ‘ 1 /‘Q/( \\/ / I,‘/\ u / / / .//,// C/ ,/f/‘ //I/ /v,/I / . . ‘ /,.—+V1ton o—nng Cu foil~ ~»\\ ’:‘ pf/ ~——working ' electrode 3‘ ET" “TEE Figure 3.1 . Electrochemical cell configuration. Figure courtesy of Dr. Greg Swain‘s Laboratory, Department of Chemistry. Michigan State University Stoichiometric Polymerization Reaction The polymer chains of polypyrrole consist of linked aromatic units. which are coupled (Figure 3.2). The coupling occurs at the carbon atoms. which are the most reactive toward addition and substitution reactions (Genies 1983). After the initial oxidation step, there is a coupling reaction, followed by a de-protonation and a one-electron oxidation in 27 order to regenerate the aromatic system. The initial coupling reaction involves two pyrrole monomers to produce dimmeric intermediates, and there is a steady-state coupling reaction, that involves the reaction of a pyrrole monomer with the oligomeric and polymeric intermediates. The original radical cation may undergo a radical coupling reaction with another radical to form a dimmer. or it may react as an electrophile and add to a neutral monomer. Since the coupling reaction must involve the coupling of two radical cations, polymerization reaction proceeds only when the potential is sufficiently high to oxidize the monomer (Genies 1983). At these potentials. the concentration of the neutral aromatic species is zero at the electrode and negligible in the region ofthe electrode. In the initial stages of the reaction, the charge consumption which accompanies the polymer formation is linearly dependent on the time and independent of the concentration of pyrrole for a constant potential electrolysis. Under steady state conditions. the coupling reaction must also occur between the radical cation of pyrrole and the radical cations of oligomers, since the dimmer. trimmer. and polymer are more easily oxidized than the monomer (Diaz 1986) They will also be present in the oxidized state and not the neutral form during the polymerization reaction. Under steady-state conditions. the current (1) depends on the rate of diffusion of pyrrole to the region of the electrode (Genies 1983). 28 $2 _a 8 .moEoO Set “32943 20:3 .3 cocastoEmEa EEEozooboo—u .m.m Ezwi : \ / _ u 5:83. 2:32.85 n F + c z of / \ + \ / u x / \ + u / \ .o~+.:«+ \ / u l u +.+=\ / u 3.:on .2332 BEES cozuo .353. +IN+ \ / u / \ 3...... I 2 21 ll .o + l 36 3.32332... / \ / \ Electroactivity and Conductivity The electroactivity of polypyrrole films (200-400A) was demonstrated to be chemically reversible and can be driven repeatedly without loss of electroactivity (Diaz 1980, 1981). Polypyrrole has a low oxidation potential; it is very sensitive to oxygen in the air. The rate of switching is limited by the mobility of the anion in and out ofthe film where the linear diffusion rates are 10'mcm2/sec (Genies 1983). As a result, the switching rates are very sensitive to the anion. The rate of oxidation is slightly faster than the rate of reduction. This rate dependency on the anion gives rise to very complicated cyclic voltammograms which represent the combined faradaic and capacitive currents (Diaz 1986) and whose forms are not fully understood. The anion influences the kinetics of the reaction and does not influence the switching potential. The switching potential of the film is sensitive to the presence of substituents along the polymer chain, affecting the stability ofthe film in air. The presence of any substituents in the nitrogen ofthe pyrrole unit makes the switching potential move towards the anodic side by about 0.6 V which is sufficient to make the film stable in air. This added stability greatly simplifies the handling and storage procedures for the films. Section 3.3. Practical Applications of Intrinsic Conductive Polymers (1C P) New areas of application have been developed for the use of 1C PS and its blends. These alternatives to conventional materials offer possible solutions to technical problems. One ofthe major uses for ICP‘s is the use of polyaniline in dispersion paints. Its use creates 30 an antistatic coating. It is also used for anticorrosion coating systems and has contributed to the creation of ultra thin (200nm) layers for coating materials such as printed circuit boards. The earliest reported ICP application was in free standing sensor devices. These sensors detected and measured levels of doping within the same material upon exposure to vapor- phase dopants. The early work of Shirakawa (Shirakawa 1977) and other researchers (Chiang 1977; Wnek 1980; Dury 1981; Guiseppi-Elie 1983) described the doping of polyacetylene by vapor iodine, bromide and Ast within tubes outfitted with conductive polymers. The sensor was a strip of un-doped polymer film with known physical dimensions suspended within the vapor stream of the dopant. The current within the probe was monitored using a probe to measure current. This simple sensor served as the basis for evolution of sensor applications using conductive polymers. Today various sensor configurations exist, contributing to various fields, such as electrochemistry. analytical chemistry. material science, biochemistry and biotechnology. Past and current research on ICP’s is as wide as the areas of its applications. These studies cover a wide range, from fundamental scientific research on redox mediation and electrocatalysis. sensor device configurations and design. to sensor application and commercialization (lvaska et a1. 1991, Hillman et al. 1987). Biological sensors are a group of sensors that employ a biologically active molecule as the recognition agent. Among the molecules employed for biosensors are enzymes, antibodies, DNA and RNA. Chemical and biological sensors using electroconductive polymers provide a powerful sensor technology. In order for the sensor to be effective it needs to be sensitive, selective and suppress the effect of cross reactivity with other 31 molecules. In addition. the polymer membrane may serve as support or matrix for the immobilized indicator molecule. Different recognilion-transduction formats are available for biosensors. Usually a hyphenated nomenclature demonstrates the type of molecule and transduced signal. For example, Pt-PPY-uiclA biosensors stand for Platinum- polypyrrole-target DNA molecule. As discussed in chapter 2, the uidA gene specific for E. coli was used in this study as the recognition agent. This nomenclature will be used for the biosensor that has been proposed in this study. Section 3. 4. Electrochemichal Techniques Potentiometry Potentiometry is the simplest form of sensing that uses electroconductive polymers. The polymer serves as a sensing membrane capable of reaching equilibrium with the contact solution. The measured signal is the result of changes in an open circuit potential (E) of the modified electrode versus a reference electrode. The modified electrode usually involves a conductive metal. such as gold or platinum. covered with the conductive polymer and the recognition molecule. The measurable changes can be the result of various events, such as shifts in the dopant's equilibrium in solution. ion-exchange processes with different ions in solution and redox equilibria within the metal electrode (Wang 2000). Amperometry Amperometry is the most common approach to electroconductive sensors. The measurable signal is derived from redox current resulting from a constant voltage maintained within the modified electrode. In most cases. the polymer plays a passive role for attachment or covalent bonding of the recognition agent. 32 C onductimetry C onductimetry is another well-known method for measuring signals from sensors. This technique provides 3 directly traceable signal with respect to a rate of change. The role of the electroconductive polymer is that of a transducer-active material with chemical amplification properties. The most commonly used transducer (polymer) (Sheppard et al. 1993) is casted on a planar micro-fabricated integrated electrode array of a defined cell constant. Conductivity measurements in aqueous solutions provide signals on a larger scale of approximately six orders of magnitude (Zaetsky et al. 1988’). A time of 10 minutes has been achieved for low detection limit using an antibody based conductimetric biosensor (Muhammad-Tahir and Alocilja 2003). Voltammetry Voltammetry involves a sweep of the electrode potential over a range associated with the redox reaction of the analyte. The measurable signal is derived from the change in a peak current associated with redox reactions (Wang 2000). The conductive polymer may serve as the catalyst. reducing the redox potential at which the analyte of interest is measured and therefore reducing the influence of background and interfering currents. It can also play a passive role such as providing covalent bonds, adsorption or anchorage for a redox mediator reaction. Electroconductive polymers also allow a fomt of indirect voltammetry of electro-inactive but ionic analytes. In these experiments. the analyte ion may induce a redox reaction in the polymer. but the polymer itself may not be efficiently 33 oxidized or reduced under the conditions of the test. Voltammetry offers the advantage that reference redox signals with standard controls may be measured to provide increased accuracy. Section 3. 5. Cyclic Voltammetry Cyclic Voltammetry is the most widely used for obtaining qualitative data about electrochemical processes. This technique gathers information on the kinetics of electron transfer and coupled chemical reactions or adsorption processes. The technique consists of scanning linearly the potential ofa working electrode using a triangular potential wave form. (See Figure3.3) 34 Cycle 1 Reverse E final Scan 7'. ‘/ '5 5 ‘5 '\ ‘3‘ Forward Scan E Initial Switching Potential Time Figure 3.3. Potential vs. Time in Cyclic Voltammetry. Adapted from (Wang and Jiang 2000) During the potential sweep, the potentiostat measures the output signal (current. I) resulting from the applied potential. These recordings result in cyclic voltammograms (CV). Figure 3.4 illustrates a characteristic cyclic voltammogram of a reversible redox couple. As the applied potential (E) is scanned along the solution. a cathodic current begins to increase until a maximum reduction peak is achieved. Then after the reduction is completed. the reverse scanning begins in the anodic zone where the oxidation is taking place. 35 Cathodic O—yR Forward scan Current T I 1 Reverse scan 0 o— R Anodic Potential Figure 3.4. Cyclic voltammograms ofa reversible redox process. The peaks of the cyclic voltammograms are due to the formation of a diffusion layer near the electrode surface (Wang 2000). The diffusion layer increases with the number of cycles that are performed during the voltammetric measurements. Thus. the change of amplitude ofthe peaks represents the change of the concentration gradient with time. The polymerization of PPY is a classic oxidation-reduction processs. but the entrapment of DNA into the polymer is not accompanied by an electron transfer. Thus it can be classified as an irreversible. or quasi-reversible, redox process. The scan rate of the process can be obtained using equation I. 36 0 11’2 15:,,=E°——1l 0.78—1n k1,, +ln M 1 an F D" RT [1 Equation 1 is a variation from the classical Nernst equation that determines the potential for a reversible process. All the parameters are defined as follows: Ep= potential axis E0=standard potential for a redox reaction R=universal gas constant (8.314] K" mol") T=Kelvin temperature D== diffusion coefficient (cmzs’l) nu: number of electrons transferred in an equation in the charge transfer step F= faraday constant (96.487 coulombs) k“: standard heterogeneous rate constant (cms'l) o=transfer coefficient The potential axis (E,,) occurs at potentials higher than Eu . The over potential is related to but independent of k0 and a . The current peaks 0,.) will appear less defined and will have lower current peaks than the completely reversible redox processes (See Figure 3.5). 37 Figure 3.5. Irreversible (A) and quasi- reversible (B) cyclic voltammograms for redox processes. Adapted from (Wang and Jiang 2000) The peak current (i,,) is given by the following equation: lp : (2.99x105}1((Zl’la)ll’2/ICDU2VLI2 [3] In equation 2. A is the electrode area (cmz). C is the concentration (mol cm'3) and D is the diffusion coefficient (cmzsl) n is the number of electrons transferred and v is the scan rate (Vs'l). According to this equation, the current peak is proportional to the bulk concentration but will be lower in its height (Wang and Iiang 2000). The current peak will be approximately 80% lower than the peak for the reversible redox process. In general, the voltammograms for a quasi-reversible system are more drawn-out and exhibit a large separation in peak potential compared to those ofa reversible system. (see Figure 3.5. curve B) Section 3.6 Instrumentation Potentiostats To generate the cyclic voltammograms, the current is measured inside an electrochemical cell containing three electrodes using a potentiostat (Figure 3.6). A potentiostat is a signal amplifier used to control voltage between two electrodes, a working electrode (WE) and a reference electrode (RE), to a constant value. The reference electrode maintains a constant voltage referring to the potential of the hydrogen electrode as a reference point. A silver wire covered with a silver chloride layer immersed in a chloride solution is one of the simplest reference electrode systems and is the choice for this project. As soon as current passes through this electrode, it is polarized, meaning that its potential (E) varies with current (I). To maintain a stable potential, no current should pass the reference electrode. A third electrode then is needed and is referred to as the counter electrode (CE). A current is forced between the working electrode and the counter electrode in order to keep the working electrode potential at a constant value with respect to the reference electrode. The potentiostat measures the potential difference between the working and the reference electrode without polarizing the reference electrode. The potentiostat compares the potential difference to a preset voltage and forces the current through the counter electrode towards the working electrode in order to counteract the difference between the preset voltage and existing working electrode potential. The potentiostat must have a bipolar operational amplifier (CPA) with two inputs: an inverting input and a non-inverting input. By introducing a voltage into the non-inverting input, it will produce an amplified voltage of the same sign and by 39 introducing the voltage into the inverting input, the result will be an amplified signal of opposite sign. To close the loop. the working electrode must be connected to the non- inverting input (+), the reference electrode to the inverting input (-) and the counter electrode to the output. Now the loop is closed and the working electrode is polarized to the difference between the reference inputs. The working electrode input is set to zero. To measure the current through the counter electrode, the system needs a resistor (R) in the counter electrode wiring, across which a voltage can be measured. proportional to the current flowing. The reference electrode is commonly protected by an input resistor (RS), preventing the potential amplifier from being destroyed by static high voltage shocks. — s l + _.°"A | " ! :7. RM Wa . a Figure 3.6. Schematic ofa Potentiostat connected to a three cell electrochemical cell. 40 The use ofa traditional electrochcmistry technique such as cyclic voltammetry. combined with molecular biology techniques such as DNA hybridization. is a non-traditional approach in the development of biosensor technology. The combination of conductive polymers like PPY with a platinum electrode is the working electrode platform of choice for this study. The embedding of a 25 bp oligonucleotide specific to the uidA gene that identifies E. coli species is the recognition agent of choice for this study. The recording of cyclic voltammograms is the method used for the analysis of hybridization with synthetic oligonucleotides and genomic DNA from common water pathogens. The application ofthe Pt-PPY-uidA biosensor has been designed for the rapid detection of E. coli from water samples. 41 CHAPTER 4. BIOSENSORS Section 4.1 Principles of Biological Sensors Unprecedented interest in the development of analytical devices for rapid detection and monitoring of chemical and biological species has led to the emergence of biosensors. The Biosensor technology promises new detection altematives for E. coli and other pathogenic bacteria. A biosensor uses specific biological recognition agents, such as enzymes or DNA oligonucleotides, in close proximity to a transducer and converts the recognition event between the recognition agent and the target analyte into a measurable signal. Recent developments in this technology have been applied to the detection of foodborne pathogens, with most of the architecture in the optical system. For example. Seo (Seo 1999) developed an integrated optic interferometer for detecting Salmonella tjphimurium with sensitivity of 105-107 colony forming units (cfu)/ml. A luminescence- based method could detect 102—103 cfu/ml of E. coli 01572117 and Salmonella tjphimurium in fresh produce (Mathew and Alocilja 2005). lVIicrofabrication has played an important role in the miniaturization of such devices. Micro electrochemical mechanical systems (MEMS) were used to detect whole cells of E. coli 01572117 using impedance measurements. (Radke 2005). Conductive polymers such as polypyrrole (PPY) and polyaniline (PANI) are being extensively researched for their application in biosensors. These types of materials exhibit interesting and promising electrical and optical properties only exhibited in inorganic materials. Both have a relatively high conductivity and good environmental stability (Kanga 1998). These conductive polymers differ from inorganic semiconductors 43 (i.e., silicon) in that they are molecular in nature (Duke 1980). The double strand DNA exhibits 1t- electron backbone configuration. facilitating faster electron transfer along the DNA chains (Kelley 1999) and therefore to the conductive polymer. Some potential applications of conductive polymers can be seen in diverse areas. such as electronics and photonics, as well as pharmaceuticals, food manufacturing, wastewater treatment and energy production. Biological Sensor Specificity Electroconductive polymer-based biosensors need to be target-specific. Specificity implies the response ofthe sensor to a specific analyte through the actions ofan indicator biomolecule. The polymer may serve as a transducer, but the response should display preference to the analyte. The sensor should also show a low response to non-target species. The methods for the preparation of these biosensors fall into one of the categories following this discussion. Physical adsorption is useful for conferring specificity to the biosensor (lvlalmros 1988). This method is most commonly used with proteins, where passive adsorption onto a polymer confers activity ofthe enzyme to the membrane. Another convenient method is physical occlusion during electropolymerization. In this case. the electropolymerization under oxidizing conditions provides a positively charged polymer. This phenomenon facilitates the occlusion and immobilization of the anions. The negative charges are required to balance the positive charges of the polymer backbone to maintain charge neutrality. These are provided by the biomolecules which usually have net negative charges (Shidmidzu 1987). 44 Besides these preparation methods. there are two general categories of electrochemical biosensors depending on the nature of the biological recognition process. The first recognition process includes bio-catalytic devices that involve the use of enzymes. cells or tissues as immobilized recognition agents. The second group is called affinity sensors which are based on the affinity of two bio-molecules. such as antibodies with membrane receptors or nucleic acids with their homologous sequences. Section 4.2. Catalytic Sensors Enzymes are proteins that catalyze chemical reactions in biological systems. The enzymes (E) are usually very specific to their respective substrate (S). making them highly selective molecules. Electrodes can be coupled to a layer of enzymes to monitor a wide variety of substrates of clinical. environmental and food safety importance. A successful enzyme-based biosensor depends on the immobilization of the enzyme layer. Direct contact between the enzyme and the sensing surface is needed along with enzyme stability. The simplest method to entrap the enzyme on the electrode is with a dialysis membrane, but conductive polymer films are an alternative for entrapment. The response of the enzyme electrode depends on the kinetics of the enzyme-substrate reaction (equation 3). k1 E+S(_)ES—"3—>E+P k2 The substrate (S) combines with the enzyme (E) to form an intermediate complex (ES) which breaks downs to form products (P) thus liberating the enzyme. At a fixed 45 concentration, the rate of the reaction (v) is given by the Michaelis ——Menten equation. (equation 4) K..+IS] “1 where Km is the Michaelis-Menten constant and V”. is the maximum rate of the reaction. The term Km corresponds to the substrate concentration for which the rate is equal to half of 1",". For the design of enzyme based sensors, it is desirable to have a high value of 1'”, and a low value of Km. More sensitive devices can be design by coupling two enzymatic reactions in a chain (Yang et al. 1991). One of the most famous and studied enzyme based sensors is the glucose sensor. Developed in 1967 (Updike 1967), it is based on the enzymatic reaction where. in the presence of oxygen. glucose oxidase liberates gluconic acid and peroxide. 'I ‘ '1).‘(l¢. - - Glucose + 0, f "m” " "9 >glucomcac1cl + ]{.,01 At the present time. there are multiple varieties of biosensors commercially available for public use. The glucose sensor has been successfully developed and commercialized to the extent that there are dozens of products available for diabetes monitoring approved by the American Diabetes Association (ADA). Other common type of catalytic biosensor is the ethanol sensor for various uses (Malinaukas 1978). Section 4.3. DNA biosensors The incorporation of nucleic acids into electrochemical transducers is one of the newest and most promising technologies in biosensor development. DNA complementary base 46 pairing or hybridization offers considerable promise for obtaining sequence-specific information. In the past 10 years, a considerable number of DNA-based biosensors have been designed based on electrochemical transducers. These DNA-based biosensors are proliferating due to their simplicity, accuracy and cost-effectiveness. Different platforms have been designed ranging from the simplest to the most elaborate ones. The first studies on DNA capture and differentiation from a single or double strand were performed by Palecek (Palecek 1960). More recently, the oxidation of DNA has been measured by adsorption- stripping voltammetry by inducing electrostatic forces of the analyte onto the electrode surface (Palecek 1988). The purine bases of DNA (adenine and guanine) can be oxidized onto any type of electrode surface including carbon, indium tin oxide (1T0) and polymers such as PPY and PANI (Singhal 1997). The first studies using DNA and electrochemical sensors to detect hybridization were performed by Milan and Mikkelsen (Millan 1993). The earliest DNA biosensors used direct DNA electrochemistry, which is based on the redox of DNA onto different electrode types such as mercury. gold. carbon or platinum. This technique offers the advantage of not requiring labeling for the detection of DNA (Singhal 1997; Jelen 2002; Wang 2002). It also extreme target sensitivity (down to fentomoles) and can be adapted to several types of electrodes (Palecek 1988; Ozkan 2002). Two of the disadvantages of this technique are that it can have a high background signal, and the biosensors are usually single use. The recent development of peptide nucleic acids (PNA) may solve the hybridization background situation (Kerman 2003). 47 There have also been developments in indirect DNA electrochemistry that involve the use of an electrochemical intermediate to link the DNA probe or the target DNA. 111 1997 an intermediary synthetic molecule was linked to a 13mer DNA oligonucleotide and to a PPY platinum modified electrode, but the formation of the complex induced a decrease in current due to the bulky conformational changes along the polymer backbone (Korri- Youssoufi 1997). Other intermediates like ruthenium complexes have been used to interact with the DNA bases capable of oxidation. Measurable xidation indicates the presence of a DNA hybrid (Yang 1991). The same technique has been used in conjunction with PCR to detect gene expression in tumor cells (Armistead 2002). Most of these techniques for indirect DNA electrochemistry do not require a labeling step and they can serve for the detection of multiple targets within the same electrode. One of the major disadvantages of this technique is the synthesis of complicated intermediate complexes. Also, they usually are single use biosensors. DNA specific redox indicators The use of DNA bound to reporter molecules is one of the major advances in DNA biosensors technology. The technique is analogous to the use of fluorescence in biotechnology studies. The hybridization event triggers the electrochemical response of the redox indicator molecule. In 1994, one such device was developed to detect a mutation of one gene that codes for the cystic fibrosis (Millan 1993). The detection limit was demonstrated to be in the femtomol range. These studies used a Co(bpy)33i DNA marker combined with cyclic voltammetry on a carbon paste electrode. Since then. other redox markers have been 48 used, such as Co (phen);3+ using chrono-potentiometry on a carbon paste (Wang. J 1996). Studies using pulse voltammetry have also been used with ferrocenyl naphthalene compounds on gold electrodes (Takenaka 2000). More recently. the use of magnetic beads instead of a solid electrode as a planar surface for the hybridization techniques and electrochemistry studies was demonstrated (Palecek et al. 2002). This study used an osmium-based molecule (Os.bipy) linked to a DNA molecule and enzyme-linked immunoassay. One of the major disadvantages of these techniques is the need for some type of chemical labeling, which makes the process less cost-effective. Nanotechnology has also contributed to the design of new biosensor platforms. The use of nanoparticle labels with different redox potentials has enabled the incorporation of DNA particles onto various surfaces. One example is the use of colloidal gold particles with probes that hybridize with the target. The particles then are separated magnetically and subjected to a second hybridization with a nanoparticle-labeled reporter (Wang 2003). The complexity of most of these systems is one of the major drawbacks in using them. The assays involve too many steps, and there has been some trouble with the reliability of the surface structures used. 0n the other hand. they are extremely sensitive. detecting in a femtomol (10's) to zeptomol (102') range. Another advantage of the techniques is that different nanoparticlcs can be used to target different analytes in a single sample (Wang 2001). DNA biosensors/or the monitoring ofenvironmental pollution In most recent years, biosensors have been developed that use electrochemistry and DNA in water quality monitoring. Most of them measure the effect of chemical toxicants and 49 its effect on the DNA instead of detecting bacterial pathogens. A disposable DNA biosensor was developed by immobilization of double strand (ds)-calf thymus DNA on the surface of a carbon screen-printed electrode (SPE). The oxidation signal of the guanine base was obtained by square wave voltammetry as the analytical signal. The presence of toxic compounds in wastewater samples was confirmed by their effect on guanine oxidation (Lucarelli et al. 2002; Lucarelli et al. 2004). Other biosensors were developed using an artificial plasmid that included a regulatory protein restriction factor (Xle) inserted in E. coli plasmid with luciferase activity. This regulatory DNA sequence is involved in the degradation pathway of BTEX (benzene. toluene. ethylbenzene, and xylene) of I’semlomomis species. The detection of E. coli cells with the plasmid was observed after incubation of whole cells in different BTEX concentrations. The biosensor used bioluminescence to detect the luciferease (qu) gene in the E. coli plasmid (Kim et al. 2005). A similar design based on the bioluminescence of the lux gene was developed to detect metabolic and catabolic strains capable of degrading chlorinated solvents in ground water samples (Bhattacharyya et al. 2005). Other more complicated systems for assessing water quality involve cellular analysis and notification of antigen risks and yields (CANARY) biosensor. This system uses fractional analysis of biological warfare agents. This biosensor is based on a lymphocyte cell for cellular analysis and analytes present in the samples. It involves the dissociation and binding kinetics of analytes present in the solution or in the environmental samples (Moms and Sadana 2005). It does not involve the use of DNA or conductive polymers but is an example of the development of biosensors for pathogens from environmental SOUTCCS. 50 Platinum-PPY-DNA biosensors The electrical conductivity of PPY has been demonstrated to be in the range of 10'3 to 103 chn'1 (Diaz et al. 1986). Electrical conduction in PPY is the result of electron movement within delocalized orbitals and positive charge defects known as polarons (Devreux 1987). These positive charges are located every three or four pyrrole monomers along the polymer backbone and is the place where negatively charged dopants (DNA is this case) are deposited (Satoh 1986). Figure 4.1 shows a transmission electron microscopy (TEM) of the positively charged gaps that correspond to the polaron region (Pande 1998). The DNA can form a bond with PPY based on the interchanging of dopant molecules within PPY and negatively charged biomolecules such as DNA (Boyle 1990). Hydrogen bonding to phosphate oxygen in the DNA backbone can enhance binding to DNA. PPY will provide the hydrogen bonds through its nitrogen atoms. Figure 4.1. Transmission electron microscopy (TEM) of a PPY film. The dark spaces within the film are the positively charged regions called polarons. (Used with permission of Pande et a1, 1998). 51 Several researchers have started to work with DNA-PPY sensors to test the kinetics of adsorption of DNA onto conductive polymers (Table 4.1). Using free standing PPY films exposed to radioactive-labeled 32F double strand DNA (Minehan et al. 1994; Pande 1998) demonstrated the adsorption kinetics of DNA-PPY films. These studies showed that DNA uptake exhibited t”2 dependence. These results were for adsorption of DNA into a PPY without the use of CV. Other researchers have used ITO coated electrodes with PPY and polyvinyl sulfonate films (PVS) to characterize the immobilization of calf thymus DNA on the prepared films (Gambhir 2001). A label-free Pt-PPY-DNA biosensor was designed for hybridization purposes using a 27 bp oligonucleotide that codes for a human gene (Thompson 2003). In addition to this, the biosensor used a layer of poly (2,5-dithienylpyrrole) modified with a phosphoric acid residue to promote ion exchanges with the DNA. Cyclic voltammograms recorded the hybridization process, from which Thompson concluded that hybridization events were clearly distinguished because of the addition of chloride ion exchanges. One of the most significant works published using label-free direct DNA adsorption to PPY was done by Wang‘s research team (1999). This work used PPY and homologous DNA oligonucleotides on a glassy carbon electrode to demonstrate the DNA doping onto conductive films. The use of homologous DNA oligonucleotides effectively demonstrated the doping effect of DNA onto the PPY films. but homologous sequences are not naturally present in samples from living organisms. The purpose of our project was to take advantage of the DNA adsorption onto polymer films in real time by the application of voltage during electro-deposition. The focus of our interest was the use of polypyrrole in developing a novel E. coli biosensor for water 52 quality monitoring. The effectiveness of Pt-PPY-DNA platform using environmental DNA samples has not been reported. This project studied the specificity and stability of the Pt-DNA-PPY biosensor. Unlike other DNA-conductive polymer designs (Korri- Youssoufi 1997), the E. coli DNA biosensor for this dissertation research used genomic DNA extracted from natural environments as its target biomolecule, instead ofa synthetic DNA oligonucleotide. A 25 base pair (bp) probe from the uidA gene from E. coli K-12 was tested. Two set of synthetic 25 bp complementary and non-complementary oligonucleotides were used as positive and negative controls. 53 Table. 4.1. Recent studies involving the use of DNA-PPY-biosensor and its applications. Type of DNA Working Signal Measurement Reference Electrode/ Platform 25 bp oligo PPY-Precursor-Oligo Cyclic Korri et al. 1997 voltammogram pBr322 PPY-free films Conductivity Pande et al. 1988 (4-point probe) 20 bp oligo GC-PPY-DNA Cyclic Wang et al. 1999 homologous voltammogram hybridization event dsCalf thymus PPY-PVS-ITO Conductivity Gambhir et al. (4-point probe) 2001 ds PPY-DNA-Pt-PVDF Ion transport Misoska et al. 2001 Salmon sperm 27 bp oligos Pt-PPY-pTPTC3-P03- Ion exchange of Thompsons et al. Hz-oligo cations in solution 2003 54 CHAPTER 5: RESEARCH METHODS AND MATERIALS Section 5.]. Fabrication of the Pt-PPY-uidA biosensor by incorporation of an uidA gene oligonucleotide into a conductive polymer-electrode biosensor system. Selection ofDNA sequence/or the detection ofE. coli The gene that encodes for the enzyme B-D- glucuronidase, known as uidA, was selected for the identification of diverse aquatic strains of E. coli. This particular sequence has been used as a standard target for the identification of E. coli strains. The sequence for the uidA gene was obtained from the public data base GenBank (accession no. Ml4641). One 25 bp oligonucleotide from E. coli K-12 uidA gene, positions 1640 to 1805 was synthesized. The synthesis of the oligonucleotide was carried out at the Genomics Technology Support Facility at Michigan State University with the sequence 5'- CGTTATACGGAACGCTCCAGCGTTT-3' (25 bp uidA probe). Two other oligonucleotides were synthesized to be used as complementary target (5'- AAACGCTGGAGCGTTCCGTATAACG-3') and as non-complementary target (5'- GCAATATGCCTTGCGAGGTCGCAAA-3'). Incorporation ofthe uidA gene oligormcleoticle onto the Pt-I’I’I'electrocle The biosensor design used is a modification of the DNA-based oligonucleotide- functionalized PPY used by Korri-Youssouffi (Korri-Youssoufi 1997). A three electrode cell (Figure 5.1) comprising a Pt working electrode (3 mm diameter), a Ag/AgCl (3 M NaCl) reference electrode, and a carbon rod counter electrode were placed against a 56 copper plate and connected to a Potentiostat Versastat Model 11 (Princeton Applied Research). V,/—— ' Agv‘AgCl ,,_Carbon \9 Reference I Counter 3% electrode l electrode Nitrogen \_V) _ — as _//>.»\ ':_i1 1:1 - ’.\ .. / .g M. , .1 \lk/ mlct N‘s. lr- _ Viton O Ring .. _. -- 1ZTPt electrode . 13:7 Figure 5.1. Electrochemical cell configuration. Figure courtesy of Dr. Greg Swain’s laboratory. Department of C hemistry. Michigan State University. The electrochemical cell had a total volume of 2 ml that consisted of 0.05 M distilled pyrrole (Aldrich) and 2ul of 500ug/ml (a total of lug) oligonucleotide probe into a solution of 1M KCI used as the electrolyte. The electro-polymerization was successful and measured by a continuous cyclic voltammetry scanning for 26 cycles between 0.0 and +0.70V at a scan rate of 50mV/s. Electro-polymerization of 0.05M PPY was achieved using 1M KCI as electrolyte following previous protocols (Wang 1999). After 57 electro-polymerization. the modified surface was rinsed with sterilized water. Measurement of background signals was performed by cyclic voltammetry with a blank electrolyte solution of 0. l M glycine/0.1M NaCl. Physical characterizatirm ofthe n'todi/ied DNA -PP I" electrode surface Scanning Electron Illicroscopy Scanning electron microscopy (SEM) images of the unmodified bare platinum electrode were taken using a Hitachi S-4700 11 Field Emission Scanning Electron Microscope. The relatively smooth surface of a 600nm X 600 nm region of the bare platinum surface was contrasted by obtaining a SEM picture of the same area of the modified platinum electrode after electrodeposition with a 0.05M solution of polypyrrole conductive polymer. The images were used to characterize the formation of the polymer and the structure and roughness ofthe modified electrode surface. 58 Section 5.2 Functionalization and sensitivity analyses of the DNA biosensor using different concentrations of oligonucleotides. Sensitivity analysis ofthe E. coli D.\"A biosensor Hybridization experiments were carried out with 25 bp complementary and non- complementary to the uidA gene oligonucleotides. The hybridization solution consisted of 2 ml of 0.1M glycine/0.1M NaCl and was also used as a blank solution for measurement of background signal. A working potential of +0.7V was applied for 15 seconds and allowed to decay for 60 seconds prior to the addition (spiking) of the non- complementary sequence. Application of a 0.0 V to +0.7V potential range was performed to allow a proper electrodeposition of the pyrrole without overoxidation of the modified electrode. Five cycles were recorded and then followed by the spiking of complementary 25 bp oligonucleotide. Functionality analysis was performed using a standard concentration for target and non-target DNA of 10'6g of total DNA. Figure 5.2 is a graphical representation of the polymerization of the PPY onto the Pt electrode followed by its functionalization with the uidA probe. The electrochemical response was measured using a Princeton Applied Research potentiostat/galvanostat Model Versastat II to generate cyclic voltammograms (CV). The performance of the biosensor under the standard DNA concentration was evaluated using voltammograms of current (I) vs. potential (E) vs. Ag/AgCl. Subtractive voltammograms were generated, taking into consideration the background signal and the signal during 59 electro-deposition and hybridization events. All cyclic voltammograms shown in this study are the average of 3 replications. Determination of CV and Hybridization Conditions Additional experiments with synthetic oligonucleotides were performed with different electrolyte solutions (0.1M. 0.2M and 0.25 M of NaCl) to verify the ionic strength for the hybridization process that produced a characteristic deep purple film. Also hybridization temperatures of 76°C, 64°C and room temperature were tested to find the most effective hybridization temperature. These two hybridization temperatures are the annealing temperatures of the 25 bp probes used in this study based on their base pair composition. The use of higher temperatures would have separated the complementary strands and completely disabled any hybridization process. Hybridization incubation times were also perfomted at 15, 30. 60 and 180 minutes to identify the optimum hybridization time. Due to the nature of the DNA hybridization kinetics. no hybridization was measured for less than 15 minutes of incubation time. Hybridization times longer than 180 minutes were not used since the objective of the study was fast performance of the biosensor design. A variation of the electrolyte solution was used in the same manner as described before for the preparation of the Pt-PPY-Oligonucleotide biosensor. The same three-electrode cell (Figure 5.1), comprised of a Pt working electrode (3 mm diameter). a Ag/AgCl (3 M NaCl) reference electrode and a carbon rod counter electrode. was placed against a copper plate and connected to a Potentiostat Versastat Model 11 (Princeton Applied Research). The electrochemical cell had a total of 2 ml volume consisting of 0.05 M 60 distilled pyrrole (Aldrich) and 2g] of 500ug/ml for a total of lug oligonucleotide probe in the solution or a molar concentration of 6.15 X 10'5 M. The electro-polymerization was achieved by a continuous cyclic voltammetry scanning between 0.0 and +0.70V at a scan rate of 50mV/s. The potentiostat was run for 26 cycles. Following electro- polymerization, the modified surface was rinsed with sterile water. Measurements of background signals were performed by cyclic voltammetry with 2 ml of blank electrolyte solution (0.25M NaCl). Functionality analysis was performed using a standard concentration for target and non- target DNA of 10'6g of total DNA (6.15 X 10'5 M). Further analyses were performed using 10'7 to 10'9 g of total DNA (6.15 x 10'6 to 10'7 M). A typical value for total genomic DNA in one cell of E. coli is 17 femtograms. Therefore the approximately amount of genomic DNA used in the study was equivalent to a range of 107 to 105 cells. The electrochemical response was measured using the procedure previously discussed in this section. Voltammograms of current (I) vs. potential (E) were evaluated to see the performance of the biosensor under the standard DNA concentration. Subtractive voltammograms were generated taking in consideration the background signal measured during electro deposition and hybridization events. All cyclic voltammograms graphically represented in this study are the average 3 replications. 61 Become—3533 .3 mm EcoEoEEoo a 5:5 ovoboflo wife? voE—Eozoca of. Seneca; 2: a 8323 .Boa <2: .3 no. as as. 0882. Zea co Seazaéeoci .2 saw: a 22:6 .953 553 .858_m-& .5 ouoboofi nod—323:3“. 3553.— :_:,._ 3.2.: 1:.— L z \ z z + z I \ see 895 53 x 5:363»: z z .. \ 303 <3: o >o .5 m: 33:30.. .6 nausea? 62 Physical Characterization ofthe modified DNA-PP Y electrode surface Scanning Electron lltlicroscopy Scanning electron microscopy (SEM) images of modified Pt-DNA-PPY working electrodes were taken using a Hitachi S-4700 11 Field Emission Scanning Electron Microscope. The methodology is the same described in Section 5.1. The relatively rough surface of a 600 nm x 600 nm region of the modified Pt-DNA-PPY working electrode surface was contrasted against a SEM picture of the same area of the bare platinum film. Incorporation of the uidA probe into the pyrrole subunits was achieved after electrodeposition with a 0.05M solution of polypyrrole conductive polymer. The images were used to characterize the formation of the polymer and structure and roughness of the modified electrode surface. Statistical Analysis The data for analysis consisted ofthe collection oftwo technical replicates (cycles 13 and 26). The cyclic voltammograms of Potential (E) vs. Current (I) were recorded for each technical replicate. For statistical significance, each experiment was perfomied in triplicates (biological replicates) of both the background signal and the sample signal. Each cycle consisted of 383 data points. The maximum and minimum current points were recorded. In order to analyze the 383 data points. cyclic voltammograms were normalized after the calculation of the change in charge by the integration of the Ell components. The use of an integral value usually correlates to the loss of some data. but due to the extensive number of data points this loss was irrelevant, and Delta charge values were successfully used for the normalization and analysis. Delta charge values 63 were generated and recorded by the Power CV software (Princeton Applied Research, TN). Analysis of Variance (ANOVA) was performed using the Delta charge values and analyzed for verification of statistical significance between technical samples. biological samples at a constant probe and target concentration at room temperature and 15 minutes of hybridization time. ANOVA was performed using SAS System software. The model consisted of analysis ofestimated means for the delta charge value measured at a 95% of confidence level (p=0.05). The effect of interactions between conditions. technical and biological replicates were analyzed using the Mixed Procedure model. A similar statistical model was followed for the optimization of hybridization conditions, such as of hybridization temperature. concentration and time. 64 Section 5.3. Specificity of the biosensor using DNA from E. coli pure culture and from other common waterborne pathogenic microorganisms. DNA extraction from pure cultures of'reference strains Bacterial cultures were grown overnight at 37°C using Luria-Bertani (LB) broth. Total genomic DNA from E. coli K-12 (positive control) and other microorganisms, such as Salmonella thyphimurium (cross hybridization control), and Camp)'lobacter jejuini (negative control), were extracted from the bacterial cells using QiaAmp isolation system (Qiagen, Inc.). One ml of bacterial cells in suspension was placed in a 1.5 ml microcentrifuge tube and centrifuged for 5 min at 7500 rpm, forming a cell pellet. A volume of 180 111 of ATL Buffer was added to the bacterial pellet for lysis of the bacterial cell. Twenty ml of Proteinase K were added to help the disruption of the cell wall. The solution was mixed by vortex and incubated at 56°C for 3 hours with occasional vortex during the incubation time. After spin down of the tube, 200g] of precipitation buffer (Buffer AL) was added and mixed by pulse-vortexing for 15 seconds. then incubated at 70°C for 10 min. To begin precipitation of the DNA, then 200ml of 100% ethanol was added to each tube, mixed by pulse-vonexing for 15 seconds, applied to a QlAamp spin column and centrifuged at 8000rpm for 1 min. DNA purification continued with the addition of 500ml of Buffer AW 1 (composition of buffer not provided by manufacturer), and followed by a l min centrifugation at 8000rpm. A second purification step followed by adding 500 ml of buffer AWZ and centrifuged at 14.000 rpm for 3 minutes. The previous step was performed twice to eliminate any ethanol carryover. The elution of DNA from the column was achieved by adding 200ul of molecular biology graded water 65 (Sigma) followed by incubation for 5 minutes. This step was performed twice to increase DNA yield. Quantification of total genomic DNA was achieved by measuring its absorbance at 260nm using a UV spectrophotometer. DNA purity was verified by using 260/280 nm ratios and by gel electrophoresis using 1% agar and run for 1 hour after applying 100 Volts in a mini gel electrophoresis system (Embi Tec. San Diego. CA). Hybridization of pure DNA cultures and analysis using cyclic voltammetry. The same three-electrode cell (Figure 5.1) comprising a Pt working electrode (3 mm diameter), a Ag/AgCl (3 M NaCl) reference electrode and a carbon rod counter electrode were placed against a copper plate and connected to a Potentiostat Versastat Model 11 (Princeton Applied Research). The electrochemical cell had a volume of 2 ml consisting of 0.05 M distilled pyrrole (Aldrich) and 2111 of 500ug/ml (for a total of lug) of oligonucleotide probe. or a molar concentration of 6.15 X 10'5 M. The electro- polymerization was achieved by a continuous cyclic voltammetry scanning between 0.0 and +0.70V, with a scan rate of 50mV/s, for 26 cycles. Following electro- polymerization, the modified surface was rinsed with sterilized water. Measurements of background signals were performed by cyclic voltammetry with 2 ml of a blank electrolyte solution (0.25M NaCl). After determining the optimum DNA concentration and hybridization time with the synthetic oligonucleotides. all the hybridization events were measured using target genomic DNA at a standard concentration of 10'7g of total DNA (6.15 X 10'6 M or 100ng). All genomic DNA from E. coli K-12 (positive control). Salmonella thyphimurium (cross hybridization control). and Campylobacterjeiuini (negative control) 66 were boiled for 10 minutes and immediately incubated in ice water for another ten minutes to separate the double strands for hybridization events. Three 0.25M NaCl/6.15 X 10'6 M DNA solutions were prepared for each of the genomic DNA’s. Each was placed inside the electrochemical cell for 30 minutes to hybridize with the Pt-PPY-uidA biosensor. The hybridization was achieved by a continuous cyclic voltammetry scanning between 0.0 and +0.70V at a scan rate of 50mV/s for 26 cycles. The electrochemical response was measured using a Princeton Applied Research potentiostatl’galvanostat Model Versastat II by generating cyclic voltammograms. Voltammograms of current (I) vs. potential (V/Ag/AgCl) were evaluated. Subtractive voltammograms were generated taking into consideration the background signal and the signal during hybridization events. All cyclic voltammograms shown in this study are the average of 3 replications. Statistical Analysis A variation of the Analysis of Variance (ANOVA) was performed using the delta charge values and analyzed for verification of statistical significance between biological samples at a constant target concentration of (6.15 xIOT’M) and at standard hybridization time (30 minutes) and a standard hybridization temperature (room temperature). The model consisted of analysis of estimated means for the delta charge value measured to a 95% of confidence level (P<0.05). The effect of interactions between conditions. technical and biological replicates were analyzed using the Mixed Procedure Model. 67 Section 5. 4 Performance and stability of the biosensor in the presence of environmentally isolated total DNA from surface water samples. Sample collection Water samples were collected from the Red Cedar river. located at the Farm Lane bridge on the Michigan State University campus in East Lansing. Michigan. Multiple samples were collected from the river. Each sample consisted of 500 ml of water collected in sterile Whirl-Pack containers (Nasco, USA) at a 30 cm depth. Each container was attached to the end of a rope with a 30 cm mark and then submerged in the river up to the mark. The samples were processed according to the protocols recommended in the standard methods for the examination of water and wastewater (APIIA 1998). Triplicates of the water samples then were taken to the laboratory for further processing within 1 hour of collection. The water samples were also tested for turbidity as an indicator of organic matter. The samples were tested by the Michigan State University Department of Public Safety using the protocols recommended in the standard methods for the examination of water and wastewater (APIIA 1998). The results were expressed as colony forming units (CFU) /100ml of water. The results were used to determine the amount of cells per sample in these studies. The bacterial counts provided by the MSUDPS were used by MSU to determine the total body contact (TBC). This TBC nomenclature corresponded to the term used by the Michigan Department of Environmental Quality that established a limit of 300 CFU/100ml or lower as safe for human use of recreational waters. All quantitative results were expressed as C ELI/100ml which is the standard unit for quantification ofbacterial cells in water (APHA. 1998). 68 DNA extraction from water samples Microbial cells were recovered by the membrane filtration technique. Five hundred ml of water was filtered through a SterifilE aseptic system using a 0.22 pm pore size Durapore® membrane filter (Millipore). The membranes were incubated at 37°C for 3 hours in 10 ml of sterile Triptic Soy Broth. This 3 hours incubation period constitutes an enrichment step that is commonly perfomIed when isolating viable cells from environmental water samples (Jenkins et al. 2005). The enrichment step ensured that the majority of the DNA was extracted from viable organisms instead of non-viable ones. The samples were processed as described above for the isolation and quantification of total genomic DNA. Statistical A nalysis ANOVA was performed using the SAS System. The Model consisted of analysis of estimated means for the delta charge value measured to a 95% of confidence level (P< 0.05). The effect of interactions between conditions. technical and biological replicates was analyzed using the Mixed Procedure model. Biological replicates were increased (9 samples) to improve the statistical significance due to the environmental variability ofthe samples. 69 CHAPTER 6. RESULTS AND DISCUSSION 70 Section 6. I Incorporation of a uidA gene oligonucleotide into a polypyrrole-coated platinum electrode biosensor system. E lectropol ymerization ofPP 1’ Successful electropolymerization ofthe polypyrrole was achieved using 2 ml ofa 0.05M PPY/0.5 M KCl solution. Figure 6.1 shows typical cyclic voltammograms of the PPY electropolymerization onto Pt. The current for 26 cycles (a) was higher than that for 13 cycles (b) indicating the successful deposition of polypyrrole on the Pt. All CV were recorded using a potential between 0.0 and 0.7 V at a scanning rate of 50 mV/s. The sharp decrease in current after 0.6V indicates the over-oxidation of the pyrrole onto the platinum surface. 100: 250b,"? ‘~~ 3 ‘ “. _.__- u o T' T "if“ ' A . . 0 :2 '50“ 3 . U .Ioo~ -50- I I I 0.1 0.2 0.3 0.4 0.6 0.0 0.7 Potential (V) Figure 6.1. Polymerization of 0.05M PPY/0.5 M KC1 onto Pt electrode. The resulting curves are cyclic voltammograms after 26 (a) and 13 (b) cycles between 0.0 and 0.7 V at a scanning rate of 50mV/s. 71 F unctionalization ofthe Pt-PPY-uidA biosensor. The preparation of the modified Pt-PPY-uiu’A biosensor was achieved by electrodeposition of 0.05M PPY with 1 pg of 25 bp uidA probe. The same amount of total DNA was used for both complementary oligonucleotide specific for E. coli uidA gene and non-complementary oligonucleotide. After PPY-DNA electrodeposition, the application of potential was suspended for 15 minutes. During that time. the spiking of non-complementary oligonucleotide was followed by spiking with complementary target oligonucleotides. The cyclic voltammograms (CV) for a blank solution. electrodeposition process and hybridization with complementary and non-complementary oligonucleotides are demonstrated in Figure 6.2. The electrolyte solution over bare platinum shows a current peak of 307 uA and the background during the electrodeposition shows a current peak of 185 uA. There is an observable change in the current after hybridization of the complementary oligonucleotide (quA) and for non- complementary oligo (120pA). The drop in current after each hybridization event is distinguishable from one another. This drop in voltage corresponds to the over-oxidation of the polypyrrole film after the 0.6V value. This is a typical behavior of polypyrrole at that Potential range. The reduction of the current vs. potential range also is an indication of the interaction of the species with the working electrode surface. Therefore the most reduced current vs. potential was observed for the uidA probe which forms total hybridization with the aid-"1 probe attached to the film. There was not a distinguishable difference between the genomic complementary and complimentary DNA with the use of this hybridization solution; therefore another hybridization solution such as 0.5M KCI that yielded better results was taken in consideration for our studies. Another observable pattern in the CV curves is the difference in area under the curve that is an indication of 72 the current vs. voltage change after different target and non-complementary oligonucleotides have interacted with the modified surface. The use of a different hybridization solution such as 0.1M glycine/ 0.1 NaCl demonstrated the difference in CV patterns. Subtractive cyclic voltammograms of hybridization signals from the background (Figure 6.3) generated using 0.05M PPY/ 0.5M KCl show a difference in the hybridization process between complementary and non-complementary sequences of DNA using the concentration of lpg/pl of synthetic probes. There is a difference in the current peaks for complementary sequence at 46 pA as well as for the non- complementary sequence at 27 pA at the potential of 567 mV(Rodriguez and Alocilja 2005). These results show a distinction in hybridization versus non-hybridization signals with this DNA concentration. The formation of a hybrid due to the recognition of the probe by the complementary sequence means a successful transfer of electrons along the dsDNA chain to the conductive PPY. This explains a higher current output signal than that obtained for a non-complementary reaction. 3:00 I I (.1) mm 51 2".3 CI « .',| A '1 3:1‘ 1". - , 2'30 ‘ 31° ,tli' (in) 13mg n H 1' . {7“ ‘ C. a' ' 'tll:|_)14l:'tlld‘ 1 "I L] ‘1) / :I/l/{Irlf - _ y : ~12“! $572.11 12m: A 1L] lJ 3 M I; r- L) 3 JJ ‘ I:;...'.f ‘__.:"' ' If, 0 l,_____-_ _._ ,___~ _-_- .5 .7, $27.14!“ ~ ' ‘ I#‘-- _'-__" ' "_.A I".— C}: J—-—-—-—- -— —-- .— — - - e - v v . 113th 301] 313'.) .1le I] 552:] EIHZI .7013 :j—zijujl I PotentialppV) Figure 6.2. Comparative CV electrodeposition for lug of total DNACyclic voltammograms after 26 cycles between 0.0 and 700 mV at a scanning rate of SOmV/S for (a)blank solution 0.1M glycine/0.1M NaCl, (b)polymerization of PPY 0.051‘le"0.5 M KC], and complementary (c)and non-complementary (d) uidA probe (lug total) in 0.1Mglycine/O.1M NaCl 73 80 Cljl'llplél'l'lE‘l'llal‘f L" fr” \\ non-co ft'lplétr'léhlaty ' I p fi ‘\ K 40 1 ”_o 1-, y \\ ‘. r' P. \ . "" ‘ fit /’—_-—I I 'fl‘}. \ -2: :0 "/p 5 9‘ t — _ ‘— ‘— .- .. _ ~_ -_.——=’o.'o -" H. I I L- Vb —~-_ -- ‘ 3 30 i ": l k. | I 3 \ L) - ‘ ~ ‘ :10 \ i 430 t . 01 02 03 Dd 0‘3 08 Cl? Potential (V) Figure 6.3. Subtractive CV of complementary and non-complementary oligonucleotides targeting E. coli uidA gene fragment. Potential range from 0.0 and 0.7 V, scanning rate of 50mV/s in 0.05 M PPY/0.5 M KCl; cyclic voltammograms after 26 cycles. Physical C haracterizution of the modified DEM-PPY electrode surface Scanning Electron illicroscopy The scanning electron microscopy (SEM) images of the bare (A) and modified (B) Pt electrodes are shown in Figure 6.4. The bare Pt (A) has a smoother surface than the PPY coated Pt surface (B), indicating adequate modification of the working electrode. Figure 6.4 B shows the modified Pt-PPY-uidA surface. The dark regions on the modified surface are the positively charged polarons where the DNA probes are doped within the PPY that were demonstrated in 1998 by Pande. llydrogen bonds between the PPY and the oxygen molecules from the phosphate group in the backbone of the DNA chain allow 74 this embedding. The resolution of the SEM image (100 nm) is not high enough to clearly show the DNA structure. - 5 "‘11“. "T.- ;. .42.".2. .1 35“; -3 ‘ M _Q‘: '“s 1" '3 $3" 1'?" ‘ ;st'~i .'\ l‘ ‘L 5'6. - 71:" ”'52? n :‘ . _‘i‘ 335’. - :2 $5 3* . ‘Ifu Vin"? ‘iru‘l . v‘ . l 5“ If. 714*? Z I, ' .1} Mn _-w".' . 5‘43»? . f ‘3: .‘ k‘l ; 7 7‘" 5""! ‘5;$H‘k“wa.“‘l~\.t '"h‘n‘vsl‘m Figure 6.4. Scanning electron microscopy (SEM) of bare Pt (A), and modified Pt surface (B) Specificity of the uidA probe. The specificity of the biorecognition was demonstrated with the use of a synthetic 25bp oligonucleotide specific for the E. coli uidA gene sequence. The probe sequence has been utilized to detect E. coli strains from environmental samples and its specificity has been established by several researchers (Feng et. a1, 1991). Specificin was demonstrated after distinctive signals of complementary sequence yielded a higher current signal than that obtained for a non-complementary sequence. The synthesis of the functionalized electrode and hybridization events took place for a total period of less than two hours. However, the functionalization of the electrode, which is the bulk of the work, can be done offline. Hybridization can be performed in less than 30 minutes. For example. when manufactured in large quantities, the functionalized electrodes can be stored and 75 used for hybridization purposes in a time frame of 10-20 minutes. To obtain results in 40 minutes for detection of the hybridization signal is a significant reduction in time from current detection or culture techniques (24 to 48 hrs). This improvement in real time detection represents an incredible reduction oftime. critical in a bioterror attack situation or when rapid water quality monitoring is needed. A ssDNA biosensor was successfully designed and fabricated. The modified working electrode Pt-PPY-uidA probe was functionalized using electrodepositon techniques. A total DNA amount of lug was sufficient to detect hybridization events. The hybridization event was clearly distinguishable from non-complementary sequence using CV techniques. The hybridization event was detected with a short period of incubation (15 min). This demonstrates the great potential of the DNA-based biosensor as a viable tool for rapid biosensor response and its possible use in water quality and other events where rapid detection might be needed. 76 Section 6.2. Functionality, selectivity and sensitivity of the DNA Biosensor using different values of key variables. F unctionalization ofE. coli biosensor using different ionic strength electrolytes. Polymerization experiments were carried out using different concentrations of NaCl as electrolyte. An appropriate electrolytic concentration that could both enable the production of good quality PPY films and not affect the hybridization time was essential in this study. The use of low ionic strength electrolyte solutions has proven not to be the best option to produce high quality PPY films (Wang 1999). The ionic strength of 0.25M NaCl has been reported to be adequate for hybridization at relatively low temperatures close to 25°C (Piunno et a1. 1999). It has been demonstrated that DNA adsorption onto PPY increases with increasing ionic strength. The optimum adsorption range was found to be from 0.1M to 0.3M (Saoudi et al. 1997). The effect of several ionic strength solutions in the immobilization of DNA oligonucleotides in sensor surfaces was also demonstrated (Watterson et al. 2002). A low ionic strength of 0.25M corresponded to a high immobilization density of probes into the biosensor. This low ionic strength might be translated into less negative charges surrounding the DNA molecules. Therefore. there could be a decrease in electrostatic repulsion between DNA molecules. This decrease in repulsion between molecules may create greater embedding capabilities to the DNA. 77 We were able to determine the ionic strength in which the electrodeposition procedure produced a characteristic deep purple polymer film and a distinctive cyclic voltammogram. After several cyclic voltammograms using 0.1M, 0.2M and 0.25 M NaCl, the 0.25M concentration was the most effective in the generation of cyclic voltammograms that reflected the polymerization of pyrrole and the incorporation of the uidA gene into the polymer film. Lower concentrations of NaCl failed to produce a typical CV. After determination of the adequate ionic strength for the generation ofCV profiles, the effect of the hybridization temperature in the generation of CV signals was determined. The DNA oligonucleotides were incubated up to the melting temperature of the probe (Tm) used, calculated to be 72°C. No difference in CV profiles was observed at 0.25 M ionic strength between 72°C, 64°C and 23°C (room temperature, RT). Succeeding incubation periods were performed at room temperature. The application of a voltage to the DNA molecules might contribute to a faster migration towards the positive polaron region in the surface of the biosensor and therefore the effect of temperature might have been reduced due to the application of the voltage. Probe concentration effects were also taken into consideration for the functionalization of the biosensor. After incubation of the 10'6 g to 10'9g probe /0.25M NaCl with the functionalized PPY film. we obtained the cyclic voltammograms that can be seen in Figure 6.5. (page 87). In this figure the CV for different concentrations of the total oligonucleotides ranging from 10'6 to 10 ’9 g after subtraction of the background signal can be compared. Background signals were registered in the same current range as those for non-complementary signals. These results demonstrated that non-complementary targets did not bind to the uidA probe embedded in the PPY film. All C V s displayed a 78 decreased current at potentials beyond 0.6 V. The drop in the current beyond this potential corresponded to the over oxidation of the PPY film (Wang 1999). Previous evidence has demonstrated that at these positive potential ranges, there is a loss of the 7t- electron network and film conductivity (Mostany and Scharilker 1997) causing the over oxidation peaks. The DNA oligonucleotide did not undergo any oxidation at the potential range used. The reduction potential of individual nitrogenous bases was determined to be in the range of 1.2 to 1.7 V, a much higher potential than the one used for this study (Steenken,et al . 1997). The peak observed around the 0.1V potential may have been an effect of the background subtraction and did not correspond to any redox activity by the DNA. We can then observe a small difference with the lng total DNA, followed by the 1 ug total DNA. The biggest difference in the current after subtraction from background was obtained with the 100 ng of total DNA. After these results it was concluded that the best probe quantity to determine the biggest difference in current from background signal corresponded to theIOO ng total DNA. The current range seemed to decrease with the decrease in oligonucleotides quantity with the exception of the lng amount. The 1 ng CV profile was very close to the background signal, suggesting that this concentration was too low for an accurate identification of the hybridization event. The macro scale of the biosensor could have caused this concentration limitation. This detection limit is probably not as low as desired for a commercial biosensor. The increase in the detection limit could be solved by a reduction of the Pt-PPY-uidA biosensor scale and an increase in the surface area. The use of microelectrodes and microelectronic devices could be a potential future scope in solving the detection limit. The use of nano scale electrochemistry could also be of advantage to this detection limit factor. Lower 79 detection limits (in the range of fentomols) have been obtained in other DNA based biosensors at a much lower scale and with the use of nanoparticles to increase the polymerization area (Wang 2003). Lowering the target concentration recognition to the fentogram level will result in the detection limit to go down to 1 cell. This estimation might be possible because the total DNA composition of a single E. coli cell is approximately 17 fg. The use of carbon nanotubules attached to magnetic beads for an increased surface area can lower the concentration of DNA needed as target a 1.000 fold (Wang, 2004). The increased surface area also produces an increase in voltage that could be recorded in the CV by the biosensor. The CV signals using 1 pg of complementary sequence were significantly different from the background and the non-complementary ones as well as different from the CV using a 100 ng sample. The decrease in current after hybridization with complementary oligonucleotides has been reported previously (Korri-Youssoufi 1997). This phenomenon may be the result of the increased charge density generated by the formation of the double strand DNA. Figure 6.6 shows voltammograms of complementary signals and their corresponding non-complementary signals. All of these signals are distinctively different from background signals. Background signals were comparable in dimensions to those obtained for non-complementary targets. Background signals have not been shown for simplicity purposes since they overlap with the non-complementary signals. The analysis also demonstrated great variability within background signals. Therefore. subtracting the background signals from the actual signals introduced variability to the resulting CVs. Analysis of actual CV signals was performed without subtracting the background to reduce the variability. 80 The CV signals for lug complementary and non-complementary probes with incubation times of 30, 60, and 180 minutes, without subtraction from background signals. are shown in Figure 6.7. At this oligonucleotides concentration, there was a significant difference from the background signal and no difference for different hybridization times. Figure 6.8 shows the effect of hybridization time with a 100 ng concentration of complementary and non-complementary probes against the background signal. The difference was significant only after 30 minutes of hybridization time. These results are confirmed by statistical analysis at 95% confidence that can be compared in Table 6.3 and will be discussed in the following section. The higher current range demonstrated by background and non-complementary probe solutions might correspond to the doping of both the Cl' anion and the negatively charged DNA into available polaron sites that are not occupied by the 25bp probe. This induces a flow of electron transfer along the PPY film, resulting in CV profiles with higher current output. Hybridization events and the formation of a double strand (ds) after hybridization might cause an obstruction of the n- electrons from the dsDNA to the PPY resulting in CVs with a reduced current range output. According to DNA adsorption kinetics studies, 85% of DNA used was adsorbed into PPY after 10 minutes. and total equilibrium of adsorption kinetics were achieved in less than 45 minutes (Saoudi et a1. 2000‘). These results, along with the results from statistical analysis. supported the decision to use 30 minutes for hybridization times. Delta Charge (A Q) anulysisfbr the normalization QfCIx'signals. Electrochemical analyses using cyclic voltammetry for DNA hybridization studies do not exhibit the typical CV graphs with evident cathodic and anodic peaks from reversible 81 redox reactions. DNA does not undergo a redox reaction at the potential range used for these studies. Therefore, a more sensitive analysis of the CV was obtained using delta charge value (AQ), which represents the integral of current across the selected set of points with respect to time. The AQ value was expressed in mili-Coulombs (mC) and was chosen to normalize the area under the curve that represents the totality of the 383 data points obtained in every CV. It is also an analytical tool that pemiits comparison of the change in the current as a result of the hybridization process. Besides the subtractive CVs, analysis of variance (ANOVA) of the delta charge value was perfromed to detemtine the statistical significance of the different experimental conditions. Table 6.1 summarizes the parameters used for the statistical analysis with 95% confidence using ANOVA. The parameters tested were the melting temperatures (Tm) (72°C, 64°C and 23°C), two cycles and three experimental replicates. For the hybridization analysis using 1 pg concentration, the only significant difference was for the background signal against its corresponding oligonucleotide signal. All CV signals were demonstrated to be significantly different from their corresponding background using a type 3 test of fixed effects in the ANOVA analysis (P=0.042). These results were summarized in Table 6.2. The P values were used to determine statistical significant differences in this study between all the parameters used. The F values in this statistical analysis reflected a variance in the signal due to noise or background therefore they should only be used with 2 degrees of differences. Basically an F value higher that 4.3 was expected. In a statistically significant difference ANOVA, a larger the F value usually results in a smaller P value. That relationship was demonstrated in Table 6.2. The different hybridization temperatures did not affect the hybridization event using lpg of total oligonucleotides. Table 6.3 presents the average value of AQ in mC for different concentrations of probes (lpg and 100ng) at different hybridization times (30, 60, and 180 minutes). The percentage of change for the AQ value varied from one concentration to the other as well as for hybridization times. It was observed that for both probe concentrations, the highest change in AQ was for the complementary sequence after 30 minutes incubation time in relation to the non-complementary probe after the same incubation period. There was a 46% AQ difference between the two probes at 100ng and 30 minute hybridization period. The 46% difference could represent a reference value point to be used to discriminate a significantly different signal from an insignificant one among complementary vs. non- complementary probes. The highest AQ values after 30 minutes of hybridization time versus 60 and 180 minutes might be due to a longer exposure of the PPY surface to the hybridization solutions, the C l' anions might become incorporated in the polymer causing a less quality signal due to doping. That is probably why the signal quality obtained after 30 minutes was a better one than the ones obtained after longer doping periods. No signals were recorded before 30 minutes to determine if a good quality signal would have been obtained in a shorter period of hybridization time. 83 Table 6.1. Paramenters used for ANOVA analysis of 25 bp oligonucleotides Class Levels Parameters Signal type 3 Background Complementary Non-Complementary Concentration I 1 pg Temperature 3 64°C, 72°C, 23°C Replications 3 1, 2, 3 Cycles 2 13. 26 Table 6.2 ANOVA analysis and significance values for 25 bp oligonucleotides Effect Num DF* Den DF F Value P Value Signal 2 26 l 1.26 0.0003 Background 1 4 0.73 0.4416 Signal*Backg 2 26 6.80 0.042 *DF=degrees of freedom. Conditions used was actual signal vs. background Table 6.3. Average A Q at different times and concentrations for 25bp oligonucleotides Signal Type Hybridization Average AQ Average AQ "/0 of AQ Time (minutes) (mC) (mC) change lpg 100ng Complementary 180 -40.46 $5.52 -24.88 $7.12 38% 30 -54.65 $5.52 -63.62 $7.12 14% 60 -43.31 $5.52 -26.59 $7.12 38% Non 180 -50.91 $5.52 -41.46 $7.12 18% Complementary 30 -50.70 $5.52 -117.33$7.l2 57% 60 -42.78 $5.52 -37.84 $7.12 12% 84 The highest AQ value obtained was the one for the non-complementary signal (-1 17.33 $7.12 mC) after 30 minutes of hybridization period. This value was especially important since it yielded a close value to background signal (-1 18.22 $13.23 mC). The high values for both the background and the non-complementary signal corresponded to the doping of the PPY polaron regions that were not occupied by the 25 bp probe. The doping of the Cl' anion interacted with the conductivity of the PPY making it more electroactive. These results can be better observed in Figure 6.9 where average AQ values are compared for lpg of complementary and non-complementary oligonucleotides after 30, 60 and 180 minutes of hybridization time. There was not a statistically significant difference in AQ values at any ofthe hybridization times for this particular concentration. The same studies were performed using a concentration of 100ng of complementary and non-complementary oligonucleotides during 30, 60 and 180 minutes of hybridization time. Figure 6.10 shows the AQ mean values for 100ng of complementary and non- complementary probes. In this case, ANOVA analysis confirmed the statistically significant difference between hybridization times at this particular concentration after 30 minutes of hybridization time. After 30 minutes of hybridization, the change in charge value was —63.62 $ 7.12 mC for the complementary target. This represented a 60% decrease in AQ value after 60 and 180 minutes of hybridization time. Comparative AQ values from both concentration and hybridization times are summarized in Figure 6.11. The hybridization of different concentrations of the complementary oligonucleotide affected the electroactivity of the PPY film and a change in charge was observed in the range of —63.62 $7.12 mC for 100ng of complementary probe vs. —54.65 $5.52 mC for lpg complementary probes after 30 minutes of hybridization time. This represented only 85 a 14 % increase in the complementary oligonucleotide hybridization signals from lpg to 100ng. In contrast, the value of AQ decreased 38% for lower concentrations after longer periods of hybridization times (60 and 180 minutes). Note the high value (-1 17.33 $7.12 mC) for non-complementary probe signal after 30 minutes of hybridization. This value was very similar from the background value of —-118 mC. This corresponded to a 46% change in AQ value for the 100ng complementary probe after 30 minutes of hybridization time. The background AQ value was not included in the graph for simplification reasons. The background signal overlaps with the non-complementary signals creating a busy graph. The most statistically significant different value corresponded to the one obtained after 30 minutes of hybridization time. The negative value of the integral corresponded to the net negative charge of the DNA probes and therefore was observed at the anodic portion of the CV. 86 .Uomn 8.3 25.8an2 cogfivtn»: .mbgofloscowzo meQEoEEoUéoZ new bquEoEEoU .8 3285525“. EobbE 5.. m>U 03685.5 66 8:3... .3255“: _.o no go no so no No to o mo-wom N- mo-moo.~- EgcoEoEEoo cozl m3 >em.coEo.anol ocoow foEoEoEEool ac. ignoEoEEool mo-wom.7 mo-woo. F- oo-woo.m- oo+woo.o (V) lawn!) wo-woo.m mo-woo.w mo-wom.w mo-woo.m mo-mom.N 87 . mcozmhcoocg .coeobfi 8 menace—03:03? b5:oEoEEoU-=oZ ecu bacoEoEEoU :m .6 >0 ”Stoabnzm 6.0 oSwE 9. season. 90 no 0.0 no to no No to o _ . momoom- momomd- mcoop boucoanEoo :02 o 953 EmacoEoEEoo .-.. m3 b5:oEo.anOcoz no. 9; beacoanEoo -i momoow- m5 EmacoEoEEoo .52.... m5 b2:oEo.ano 9» women..- momooe- .H...m.m.m.¥.”my_ . . . 8.33. m .3. 3.... . . u oo+mood v.” w oomooh memo? mowome . momood mo-mom.w 88 .85: 5:36:23 fiesta a movsofloscowzo ESSEuEEoo _82 do w: _ .8 m>U 8.0 23E 5 32.88. w.o 5.0 0.0 md to ad No pd o 8-8m. P- ucaoamxomm If. omh m2. tfimcoanEoolT 3.89 F- om... m2. bflcoanEool 8:. m3 bmucoEQQEool mo-woo.w- mo-moo.©- . mo.woo.v. momood- oo+mood (v) wanna mo-woo.N .8-m8.¢ . ...ll....l.l..l...l...l...l 4 . : ._ . ., ¥ _,..____ : 8-83 8.25.» vo-woo. —. 89 ad 5 .o .88: cocmflucnxc Eobb6 6 326203533 CaEuEu—QEOU .5 wag: .3 m>U dd “Ear. 00 mo 2 3.888 to m o Nd Yo 8: 952 888.9950 .1 8 ._. 98F b35835 It. 2:99.08 m>< dd 0.5%... .8355. 2.: sou-flute»: on- 295829.00 coz I ESEmEEonVI , 8. 8| V A m at. .w 0 D. W m on- m m b . ONI o..- 91 moE: cosafiflff 2.29:6 8 $305 bacoEoEEgéoZ can EngoiEoU? wag— go OQ owfiu>< .c _ .0 223; .8352. as: Sagas»: EmwcmEmano .52 D baucmEmano D +—————4 8558 of SEES o0 moSEE cm L h L OE- ONT 00F- ow- 00. CV. (ow) o cusp ofimahv .mufic cosfi€tn3 28 2235538 9280—03533 9 89%“: 53> mos—9 O< 0920.3 mo :ScanEoU ._ _.o Sawf— 0:- .8555. 2:: 8:353»... M . or- bflcwanE8 9.59 D . oo..- EmacmEoEEoo :0: 959 aflcwEwano Q; I , m . 8. w. bchEmano :0: Q; I w B m 0 . 09 \w: D W H a . om- ow; och . p .. . ‘ ‘ O 93 Section 6.3 Specificity of the biosensor using DNA from E. coli pure culture and from other common waterborne pathogenic microorganisms. Cyclic voltammetry analysis for the l2}-‘l7fl(ll3£lll()n ofE. coli K-12 genomic DNA and other common water pathogens. Cyclic voltammograms were recorded to test the specificity of the hybridization for a total amount of 100 ng of E. coli K-12 genomic DNA with the functionalized Pt-PPY- uidA biosensor. A total amount of 100ng corresponds to approximately 104 cells. Cross hybridization with other enterogenic species was also tested using the same concentrations of 100ng for Salmonella 0yflzimurium and (.211an'lohucterjejuni genomic DNA. The optimum hybridization conditions for generating distinctive CVs were previously determined using a 25 bp complementary and non-complementary oligonucleotide specific for the uidA that identifies E. coli species. These results were discussed as part of section 6.2 in this chapter. From that discussion we concluded that a concentration of 100ng of total DNA and 30 minutes of hybridization time at room temperature were the conditions that yield hybridization events that best distinguished complementary from non-complementary species. The specificity of the uidA gene was determined using genomic DNA from E. coli K-12. Total genomic DNA from S. ophimurimn and C. jejuni were used to determine cross hybridization of the uidA 25 bp probe with other enteric pathogens. For the rest of the discussion, we will refer to C. jejuni as a negative sample since it was used as a negative control due to its lack of the uidA gene. Subtractive cyclic voltammograms for the genomic DNA of pure strains were 94 generated using 100ng E. coli genomic DNA (see Figure 6.12). E. coli CV profiles showed to be very close to the CV curve from the synthetic probe. This is an indication that hybridization between the probe and the E. coli genomic target has been ddetected by the biosensor. The recognition of the genomic E. coli DNA target was within the AQ value obtained from the 25 bp complementary probe. The CV profiles for S. (ht'phimurium and C. jejuni genomic DNA was observed in significantly different current areas (AQ) than the one from E. coli. The distinctive separate hybridization CV curves of S. (hyphimurium, C. jejuni and E. coli demonstrated the specificity of the uidA probe to g E. coli. No cross hybridization reaction was determined by the CV curves from Salmonella and Campy-’lobacter genomic DNA. As previously determined from statistical analyses, the variability of the background signals was taken out of the CV by using the actual CV curve without subtraction. A significant difference could be seen after 30 minutes of hybridization time. These results are confirmed by statistical analysis at 95%confidence that can be compared in Table 6.4. Table 6.4 also contains AQ data pertinent to results from water samples to be discussed in the following section. Figure 6.13 shows the CV profiles after 30 minutes hybridization of 100ng of E. coli K- 12 , Salmonella and C anmylobaeter genomic DNA. We can observe the different current area (AQ) of the CV by type of target DNA. Each CV is statistically different current range from each other making it possible to identify them in unknown samples. 95 Delta Charge (AQ ) analysis for the normalization ofCV signals A more in depth analysis of the CV profiles was obtained using delta charge value (AQ), which represents the integral of current across the selected set of points with respect to time. We compared the values of AQ for the different genomic DNA extracted from pure cultures against complementary, non-complementary and background CV signals. The average AQ value for the background solution of 0.25M NaCl was in the range of -1 18 at 10.25 mC (see Figure 6.14). The non-complementary CV signal was very close to background with a value of —1 17.33 :l: 12.24 mC. The AQ value for complementary oligonucleotides was -63.62 $10.81 mC. There was a 46% difference in charge from complementary oligonucleotide to the background signal. This 46% change in AQ is statistically significantly according to the ANOVA analysis. This value can be used as a reference point for a distinctive CV signal from an unknown sample. Observable changes in AQ values were obtained after CV analysis from E. coli vs. background signal. The AQ for 100ng of E.coli genomic DNA was -49.64 :t 0.65 mC. This value is 58 % lower than the background signal. This reduction is comparable to the 60% decrease in charge using 60 and 180 minutes of hybridization time vs. 30 minutes hybridization time discussed in the previous section. Therefore a pattern of A0 changes was observed using complementary and non-complementary oligonucleotides as well as background signals ranging from 46% to 60 % charge decrease. A similar trend can be observed by comparing the AQ value of Salmrmella (-28.96 :t 1.01 mC) to the value of E. coli resulting in a 42% charge decrease and a 54% charge decrease from complementary oligonucleotides. The pattern continues for the AQ values of negative genomic DNA (-18.37 :t 0.44 mC), which is 37% lower than the genomic DNA of 96 Salmonella and 71% lower than the complementary oligonucleotide values (see Figure 6.15 and Table 6.5). The 37% AQ decrease was the lowest percentage change obtained among all genomic DNA samples that was significantly different. This 37% change could be determined as a threshold value to differentiate what is significant from what is not significantly different. The intrinsic difference in charge that results from the 25 bp complementary probe and the one from genomic E. coli DNA can be explained in terms of the electrochemistry of PPY by dopant size. The effect of DNA size on current interactions with PPY was demonstrated using 20 bp oligonucleotides and ds calf thymus DNA by square wave voltammetry (Jiang and Wang 2001). The steric interaction of a short DNA oligonucleotide could be expected to be more direct that the interaction of ds genomic DNA. This can be translated into a higher electroactive response from shorter ss oligonucleotides. In contrast, genomic DNA has secondary and tertiary structures and more steric impediments in its interaction with the PPY film, thus reducing the redox activity of the PPY network. This could be translated into a reduction ofcurrent response and a lower charge value. The genomic DNA entrapment could also result in less interaction with electrolytes in the solution, affecting the charge exchange with the PPY film. The specificity of the Pt—PPY-uia’A biosensor has been demonstrated using a total of 100ng of genomic DNA from E. coli K12 as a positive control. Salmonella typhimurimn as a cross-hybridization control and C ampylohacter jejuni as a negative control. The use of the AQ value to normalize the 383 data point obtained in each CV was a good indicator of the percent difference between genomic strains. The lowest AQ percent 97 difference from the genomic DNA strains was obtained from Salmonella against negative strains (37%). A 46% AQ difference was obtained from complementary oligonucleotides vs. non-complementary uia’A oligonucleotides. Both percent values should be considered as threshold values when the target molecules are short DNA oligonucleotides (46%) and when the target values are total genomic DNA (37%). Table 6.4. Average delta charge (AQ) for all genomic DNA isolates Signal Average delta Q (mC) Pr > |t| Non Complementary -1 17.33 i 12.24 <.0001 Complementary -63.62 i712 <.0001 E. Coli 4964 $0.65 <.0001 Water -40.57 i: 4.64 <.0001 Salmonella -28.96 i 1.01 <.0001 Negative -18.37 i 0.44 <.0001 Table6.5. Comparison of AQ change percentage among signals and significance after Tukey-Kramer adjustment Signal Compared Signal Percentage Statistically Adjusted P difference in Difference A0 (9%) (P<0.05) E. Colt Non Complementary 58 Yes 0.0134 Complementary 22 No 0.7866 Water samples 18 No 0.4091 Salmonella 42 Yes 0.0001 Negative 63 Yes 0.0009 Negative Non Complementary 84 Yes <.0001 Complementary 71 Yes 0.0049 Salmonella 3 7 Yes <.0001 Water samples 55 Yes 0.0013 Non Complementary 46 Yes 0.0357 Complementary Water samples 65 Yes 0.0001 Salmonella 75 Yes <.0001 Salmonella Water samples 27 No 0.1867 Complementary 54 Yes 0.0438 Water Samples Complementary 36 No 0.3965 98 6:5 :ozmflvtn»; 3558 cm 8cm mcowoéaq .295 525.8 Co 85:30 Ban :5: <20 BEosw t8 952 .8 233m >0 028855 .N_.o Rama SEE-8.. mo 2 co no to no .3 to o . . l . _ . 8-83- mm 2&0 m>ummoziomm o 88 3 8 953 .. mcoop <20 o_Eocmm m__ocoE_mw ,. 8-80..- 382220 2888 =8 .m . t . _ . .t 8.28.? . s . m - x z 3...}!!! A...“ H w I x, .u. a». , w 4.? s l\ . . UV r oo+wooo a _ : 8-89m mo-woo.p 99 as: 8:368: 835:. cm 5cm mcowofiaq 863 5.555 Co 35:20 obi Eoc <75 288% go wag— ..8 £33m >U .26 oSwE o6 .>:£c3oa no to no i V on... acooptanEoEEoolll camp acoow 25,302+ coop a=ocoEEwlxl on» 953 __oo.mlol on... Nd _..o vo-wooN- woman;- vo-moo. —- mo-moo.m- oo+moo.o (v) wanna mag-mood Vo-mooé 2?on— vo-moo.N 100 .muncoflozcowzo 235?? .3 w: cc— .8 O 876 9.0229250 585956 .3 .0 2:3... ove- Bcoo: 833.3523 .33 ONP- OOP- ow- (Owl!) Inna 101 .mcowocfia 3:203 E0: <20 BEE-aw? w: 2: 88 O 33... owseu.»<_o :oflSgEo-v .26 2...»: 3.50: 00.953 «099—. .0 (20 SEQ-am _SOh on- l-n-l Oct 102 |~~l O ‘? bloom ON- 1-4 or- 025002 m=®coE=~m :00 .m Section 6.4 Performance and stability of the biosensor in the presence of environmentally isolated total DNA from surface water samples. Biological and chemical charactertation of water samples. Multiple water samples were collected from the Red Cedar River that crosses the Michigan State University Campus. Because of its convenient location and its recreational uses, multiple samples were collected from the river to test the performance and stability of the Pt-PPY-uia'A biosensor. Weekly monitoring of the river's microbial populations was carried out by the Michigan State University’s Office of the University Physician as part of an agreement with the state of Michigan to monitor total body count for recreational waters. Water sampling was carried out in parallel with the weekly monitoring carried out by MSU officials in order to use their data as support for our studies. Quantification ofE. coli cellsfrom water samples. E. coli cells were grown using the Membrane Filtration technique recommended by the EPA (USEPA 1986, 2002) and enclosed in the Standard Methods for the Examination of Water and Wastewater (APHA, 1998). Figure 6.16 shows the average E. coli CFU/ 100ml obtained at the Farm Lane sampling site of the river. Temporal and spatial variability was observed in these samples. This spatial and temporal variability is characteristic of most environmental samples. E. coli C F U/100 counts were as low as 73 CPU/100ml (sampling event 3) and as high as 2.500 CFU/100ml for sampling event 8. Typically, low levels of E. coli cells are encountered during periods of low precipitation 103 and high CF U counts are obtained after rainfall. Swimming in water that does not meet the recreational standards does not mean illness. It increases the risk for the public of being exposed to pathogenic microorganisms including E. coli (C heung 1990). A count of 300 or lower meets Michigan’s Total Body Contact (TBC) Standard(MDF.Q 1997). Results that fall between 300 and 1.000 meet Michigan’s Partial Body Contact (PBC) Standard. If the results exceed 1.000 contacts. the river water must be avoided (MSU 20041 Turbidity Water quality was also monitored and recorded using turbidity measurements. Figure 6.17 presents the average nephelometric turbidity units (NTU) obtained after each sampling. The turbidity ranges from 2 to 8.61 NTU. There is a correlation with turbidity and microbial growth that can be observed in figures 16 and 17. The low turbidity measurement corresponded to low CFU counts and high turbidity measurements corresponded to high CFU results. This correlation between microbial growth and turbidity measurements is important to determine the quality of the water. Suspended and colloidal matter that includes clay, slit, organic and, inorganic matter. as well as plankton and microbial cells, are the major causes for turbidity. Another important parameter that can be used for water quality assessment is the total organic content (TOC). This parameter measures the variety of organic compounds at different oxidation states (APHA 1998). Because of the availability of these compounds as carbon and energy sources for natural water microbiota. it is an excellent parameter for the estimation of microbial water quality. The average TOC measurements for the Red 104 Cedar River has been estimated as 10-15 ppm (parts per million) for periods of no precipitation and as high as 20 ppm for periods of high precipitation. This data was obtained from personal communication with Shawn McElmurry, a Ph.D. graduate student of Dr. Thomas Voice from the Environmental Water Chemistry laboratory at the Department of Civil and Environmental Engineering at MSU. C V profiles of total genomic DNA isolatedfrom water samples. Because of the presence of organic and inorganic compounds in natural waters, raw water samples were not used to perform CV studies. The presence of numerous organic and inorganic chemical species can serve as counterions and can cause interference with the PPY film and the CV hybridization for the detection of the target genomic DNA. Isolation and purification of total genomic DNA was performed in order to obtain high quality target material and reduce the interference of other naturally present anions from the water sample. Extraction was perfomied according to the protocol discussed in Chapter 5. Figure 6.18 shows the subtractive cyclic voltammograms for 100ng of the pure strains genomic DNA as well as the total genomic DNA isolated from water samples. The distinction of a separate hybridization CV curve from Salmonella, Campylobacter and E. coli, demonstrated the specificity of the uia’A probe to E. coli. Figure 6.19 shows the CV profiles without background subtraction for 100ng of total DNA of the same pathogens as well as the results from total DNA isolated from water samples. A sequence in delta charge value decreases as a function of hybridization events with the different target molecules. Therefore. we were able to successfully use 105 the CV to identify the hybridization signal of 100ng of total DNA isolated from water after 30 minutes of hybridization. The water samples signal can be observed in a similar current area than the ones observed for genomic E. coli DNA. This demonstrates the specificity of the probe using samples isolated from the environment. which contains E. coli cells and a mixture of unknown water microfiora. Statistical analyses with 95% confidence supported this observation. Analysis of AQ values using ANOVA test. The average AQ values for all the target molecules were reported earlier in Table 6.4. A characteristic decrease in charge was shown in a descending order starting with non- complementary oligonucleotides, followed by the complementary oligonucleotides, then by E. coli genomic DNA, water samples, Salmonella and the negative strain. The average AQ value obtained from hybridization CV using total DNA of water samples was —40.57 i 4.64 mC. This AQ value resulted in a 65% difference from background signals and non-complementary oligonucleotides. These differences are reflected in Table 6.5 and Figure 6.20 and are in agreement with statistical analyses that confirm a statistically significant difference from those two target samples. More interesting is that water samples reflected only an 18% difference from the E. coli genomic DNA. This 18% decrease in charge signal from E. coli genomic DNA reflected. in combination with statistical analysis of the AQ values, a minimum difference that was not statistically different from one another. Statistically significant results were obtained for water samples against all the target samples except for the case of Salmonella (P=0.1867). This 106 small interaction between the two signals might be an effect of background DNA from unknown species or by the steric effects of the total DNA of high molecular weight isolated from the water sample. The statistical analysis using the integral values of E vs I proved to be of accurate in the effort to demonstrate the presence of E. coli cells from both positive controls and environmental water samples. Despite the disadvantage of losing data by integration calculations, the normalized data reflected an accurate and reliable method to determine differences among samples and Delta charge values were successfully used for the normalization and analysis process. In general, these results reflect a great potential for the use of the Pt-PPY-uidA as a model type biosensor for the rapid detection of E. coli from water sources. Besides the successful distinctive recognition of hybridization events from different DNA targets. the detection time was reduced from 24 hours (using traditional methods) to a total of 5 hours. The 5 hours included the sampling collection with a 3 hours enrichment step followed by DNA extractions. The actual detection time included 30 minutes of hybridization and the CV could be obtained after ten minutes of Potential (E) application using cyclic voltammetry. The real-time detection period for the Pt-PPY-uidA biosensor was 40 minutes. The development of nucleic acids based biosensors for the rapid detection of pathogens in water samples is still in its initial steps. There is no doubt about the need for the development of rapid detection methods different from the current ones used for the quantification and detection of pathogens from environmental samples (Rompre et a1. 2002). The existence of powerful and sensitive DNA based techniques such as PCR and 107 real time PCR are still very attractive for the scientific community because of its sensitivity and accuracy. The amplification of DNA from environmental samples using PCR offers the advantage that it can detect as low as 1cell/ 100ml of water (Bej et al. 1991a). But this technique still has the disadvantage of not being able to detect the viability state of cells from which the DNA is amplified. That is the main reason for the incorporation ofa 3 hours enrichment step is used after collection of water samples. The enrichment of water samples can contribute to the growth and recovery of viable but non- culturable microorganisms (VBNC) (Roszak and Colwell 1987; Baudart et al. 2005). By taking in account the importance of the enrichment step with the use of nucleic acid based biosensors, the Pt-PPY-uidA-biosensor generates signals mainly derived from complementary sequences of VBNC microorganisms in 40 minutes vs. two hours using PCR, and 7 hours using FISH (Ootsubo et al. 2003). A total of 100ng of genomic DNA was extracted from the Red Cedar River water. We obtained significantly different CVs after 30 minutes of hybridization time using the Pt- PPY-uidA biosensor. Statistical analyses with 95% confidence demonstrated that AQ values were good tools for the detection of E. coli cells from environmental water samples. Water samples showed only 18% difference in AQ value from the ones obtained for E. coli genomic DNA. The potential use of the Pt-PPY-uidA sensors was demonstrated in a total of 5 hours from sample collection to results or only in 40 minutes after sample processing. 108 moEEmw 821 m H .muEEam 35$? _:_co_5.._u :3 Q own3>< .c_.o Eswi .oom ooow ooom twoou swswemo (08.1.) zunoo £909 MOI. comm 000m 109 m .moEEam 833 So... DFZ E mEquSmaoE 565::- .N- _ .0 2:3; 835mm 35. .38 8a n v m N F to tow) saun Autumn; or 110 .oE: cos—«Even»: 335:. on coca 336mm 833 3:25.235 :5: U o>zoflfism .w _ .o 23?. E .2228 mo-mom.- m.o 5.0 0.0 md v.0 m.o Nd to o l: , - ,- , . - - . mo-moo.- ocoow 25¢qu x 9.53 2.96.5.5 . 953.235 .u 953 __oo.m-o . mag-mome- - mo-wooe- O W n t, oo-mon- m tail-mil} l O, h” , 00.. w 2. oo+woo.o . .. 8.83. l. ,- mag-moo; mo-mom.P Ill .8353 .295 Eo¢ <20 .89 van mcowofiwm 533 E9: cozmfiEB»; .8 >0 .Ed 232“. 2.5.3.... No 9o 00 to no No to 0 «38¢ om... acoo Pbflcflcflgoo lT Egomrfiaemm 35;... 59%? 25%qu 8p :8. 9.88.8... on» 8? :8 m... «$8.: 358.7 258.9 0 H n . 8+m8.o m m 8183 «$8.. 3&0? voflood 112 moan—oi . 8:; was mcowozaa 3882 :5: <20 u_Eocow .moEpofloscoE—o 285:? go mace 5.. OG owns; go :cmcaanU .c~.o Eswi 3:8: 8.953 398 .o 53 .88 ovw- . ONT 00F- .0? (am) o 93:90 00. H .2, ON. 0280va w__mcoE_mm wmaEmm $63 :00 .w ESCmEmEEoo EmucmEmano coz I I l ‘ . -i|l-|-- . .|!l -ll'i- , 1 pl . .l .. I r3111! . - I r!- I - o 113 CHAPTER 7. CONCLUSIONS 114 Conclusions A ssDNA biosensor was successfully designed and fabricated. The modified working electrode Pt-PPY-uidA was functionalized using electrodepositon techniques. A DNA concentration of lug was sufficient to detect hybridization events. The hybridization event was statistically distinguishable from non-complementary sequence using CV techniques. Hybridization event was detected with a short period of incubation. This demonstrates the great potential of the DNA based biosensor as a viable tool for rapid biosensor response and its possible use in water quality and other events where rapid detection might be needed. We have successfully determined the conditions that yield the most distinguishable hybridization signals. This conditions, 100ng of target molecule and 30 minutes hybridization time, were the most appropriate for the creation of statistically different CV curves by the Pt-PPY-uidA biosensor using 25 bp complementary and non- complementary oligonucleotides. We determined the use of 0.25M NaCl as an appropriate hybridization solution for the creation of distinctive C Vs. We were able to observe that the use of a total of 100ng of uidA complementary oligonucleotides was the concentration that yields the most distinguishable signals for the creation of statistically significant different CVs. We also determined the most appropriate hybridization time to be 30 minutes at room temperature to obtain significantly different CVs. 115 The specificity of the Pt—PPY-uidA biosensor has been demonstrated using a total of 100ng of genomic DNA from E. coli-K12 as a positive control. Salmonella (_whimuriwn as a cross-hybridization control and C ampylobucter je/zmi as a negative control. The use of the AQ value to normalize the 383 data points obtained in each CV was a good indicator of the "/0 difference between genomic strains. The lowest AQ % from the genomic DNA strains was obtained from .S‘ulmunella against negative strains (37%). A 46% AQ difference was obtained from complementary oligonucleotides vs. non- complementary uidA oligonucleotides. Both % values should be considered as threshold values when the target molecules are short DNA oligonucleotides (46%) and when the target values are total genomic DNA (37%). Statistical analyses with 95% confidence demonstrated that AQ values were good tools for the detection of E. coli cells from environmental water samples. Water samples showed only a 18% difference in AQ value from the ones obtained for E. coli genomic DNA. The potential use of the Pt-PPY-ulclA biosensor was demonstrated and it was effective in obtaining results in 40 minutes after sample preparation. 116 CHAPTER 8. FUTURE RESEARCH 117 FUTURE RESEARCH Future research approaches should focus around the following suggestions: 1) The Pt-PPY-uidA biosensor should be constructed in a lower scale that enables lower detections limits. 2) The Pt-PPY-biosensor should be developed using other oligonucleotides for the detection 0 more diverse range of pathogens form environmental sources. 3) Studies using 16$rRNA with a specific probe for the uidA section of the RNA molecule should be used for both specificity and lower detection limits. 4) Design studies involving micro fluidics and multiple electrochemistry channels outputs should be performed in order to perform the detection of multiple pathogenic targets in a single biosensor. Investigating the mechanisms responsible for the electrochemical properties of ssDNA vs. dsDNA and the effect of short sequences oligonucleotides vs. genomic DNA should be performed in order to exploit these properties for the developing of more sensitive biosensors. 118 CHAPTER 9. APPENDIX 119 l ANOVA Analysis of AQ values using SAS System 1 mierogram at different times without background Cycle13 The Mixed Procedure Table 9.1 Type 3 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F signal 1 12 0.93 0.3538 Inc 2 12 7.77 0.0068 . signal*lnc 2 12 4.44 0.0360 [ l ! Table 9.2 Least Squares Means 6 Effect signal Inc Estimate Standard DF t Value Pr > . Error 111 signal*1nc comp T180 -40.4600 2.5242 12 -l6.03 <.0001 signal*1nc comp T30 -54.6533 2.5242 12 -21.65 <.0001 signal*lnc comp T60 -43.3133 2.5242 12 -l7.16 <.0001 signal*lnc noncomp T180 -50.9100 2.5242 12 -20.17 <.0001 signal*1nc noncomp T30 -50.7000 2.5242 12 -20.09 <.0001 signal*lnc noncomp T60 -42.7800 2.5242 12 -l 6.95 <.0001 Table 9.3 Tests of Effect Slices Effect Inc Num DF Den DF F Value Pr > F signal*lnc T180 1 12 8.57 0.0127 signal*lnc T30 1 12 1.23 0.2898 signal*lnc T60 1 12 0.02 0.8837 l microgram at different times without background ANOVA Analysis of AQ values using SAS System Cycle 26 The Mixed Procedure Table 9.4 Type 3 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F signal 1 12 0.20 0.6655 Inc 2 12 9.76 0.0032 signal*lnc 2 12 6.24 0.0139 Table 9.5 Least Squares Means Effect signal Inc Estimate Standard DF t Value Pr > Error |t| signa1*lnc comp T180 -31.2033 1.7870 12 -17.46 <.0001 signal*1nc comp T30 -43.4300 1.7870 2 -24.30 <.0001 signal*lnc comp T60 -36.6800 1.7870 2 -20.53 <.0001 signal*Inc noncomp T180 -39. 1400 1.7870 2 -21.90 <.0001 signal*lnc noncomp T30 -40.5000 1.7870 2 -22.66 <.0001 signa1*lnc noncomp T60 -33.6133 1.7870 2 -18.81 <.0001 Table 9.6 Tests of Effect Slices Effect Inc Num DF Den DF F Value Pr > F signal*lnc T180 1 12 9.86 0.0085 signal*lnc T30 l 12 1.34 0.2689 signal*lnc T60 1 12 1.47 0.2483 121 ANOVA Analysis of AQ values using SAS System 100ng at different times without background Cyclel3 The Mixed Procedure Table 9.7 Type 3 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F signal 1 12 21.79 0.0005 Inc 2 12 43.79 <.0001 signal*lnc 2 12 5.26 0.0228 l l Table 9.8 Least Squares Means Effect signal Inc Estimate Standard DF t Value Pr > Error |t| signal*lnc comp T180 -24.8867 7.1295 12 -3.49 0.0045 signal*lnc comp T30 -63.6200 7.1295 12 -8.92 <.0001 signal*lnc comp T60 —26.5967 7.1295 12 -3.73 0.0029 signal*1nc noncomp T180 -4 l .4600 7.1295 12 -5.82 <.0001 signal*1nc noncomp T30 -117.33 7.1295 12 -16.46 <.0001 signal‘lnc noncomp T60 -37.8400 7.1295 12 -S.31 0.0002 Table 9.9 Tests of Effect Slices Effect Inc Num DF Den DF F Value Pr > F signal*lnc T180 1 12 2.70 0.1262 signal*lnc T30 1 12 28.38 0.0002 signal*lnc T60 1 12 1.24 0.2866 ANOVA Analysis of AQ values using SAS System 100ng at different times without background Cycle26 The Mixed Procedure Table 9.10 Type 3 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F signal 1 12 8.30 0.0138 Inc 2 12 25.78 <.0001 signal*1nc 2 12 0.64 0.5421 Table 9.11 Least Squares Means Effect signal Inc Estimate Standard DF t Value Pr > Error |t| signal*1nc comp T180 -17.6433 6.5966 12 -2.67 0.0202 signal*1nc comp T30 -55.3167 6.5966 12 -8.39 <.0001 signal*1nc comp T60 -22.9267 6.5966 12 -3.48 0.0046 signal*1nc noncomp T180 -32. 1933 6.5966 12 -4.88 0.0004 signal*1nc noncomp T30 -78.7567 6.5966 12 -1 1.94 <.0001 signal*lnc noncomp T60 -31.4800 6.5966 12 -4.77 0.0005 Table 9.12 Tests of Effect Slices Effect Inc Num DF Den DF F Value Pr > F signal*1nc T180 1 12 2.43 0.1448 signal*lnc T30 1 12 6.31 0.0273 signal*1nc T60 1 12 0.84 0.3773 ANOVA Analysis of AQ values using SAS System 100 ng T30 all samples Cycle 13 Table 9.13 Type 3 Tests of Fixed Effects Effect Num DF DenDF F Value Pr > F signal 5 12 9.70 0.0007 1 Table 9.14 Least Square Means Effect signal Estimate Standard DF t Value Pr > |t| Error signal Comp -63 .6200 12.0996 12 -5 .26 0.0002 l signal E. coli -53.5233 12.0996 12 -4.42 0.0008 signal Negative -18.3733 12.0996 12 -1.52 0.1548 signal Non Comp -117.33 12.0996 12 -9.70 <.0001 signal Salmonella -28.9633 12.0996 12 -2.39 0.0339 signal Water -19.5733 12.0996 12 -1.62 0.1317 Table 9.15 Differences of Least Square Means Effect Signal vs. Signal Estimate Standard DF T Pr > |t| Error value Signal Comp E. coli -10.0967 17.1115 12 -0.59 0.5661 Signal Comp Negative -45.2467 17.11 15 12 -2.64 0.0214 Signal Comp Non Compl 53.7100 17.1115 12 3.14 0.0086 Signal Comp Salmonella -34.6567 17.1115 12 -2.03 0.0657 Signal Comp Water -44.0467 17.1 1 15 12 -2.57 0.0244 Signal E. coli Negative -35.1500 17.1115 12 -2.05 0.0624 Signal E. coli Non Compl -63.8067 17.1115 12 3.73 0.0029 Signal E. coli Salmonella -24.5600 17.1115 12 -1.44 0.1768 Signal E. coli Water -33.9500 17.1115 12 -1.98 0.0706 Signal Negative Non Compl 98.9567 17.1115 12 5.78 <.0001 Signal Negative Salmonella 10.5900 17.1115 12 0.62 0.5476 Signal Negative Water 1.2000 17.1115 12 0.07 0.9452 Signal Non Compl Salmonella -88.3667 17.1115 12 -5.16 0.0002 Signal Non Compl Water -97.7567 17.1115 12 -5.71 <.0001 Signal Salmonella Water -9.3900 17.1115 12 -0.55 0.5932 ANOVA Analysis of A0 values using SAS System 100 ng T30 all samples Cycle 26 Table 9.16 Type 3 Tests of Fixed Effects Effect Num DF DenDF F Value Pr > F signal 5 12 6.29 0.0043 Table 9.17 Least Square Means Effect signal Estimate Standard DF t Value Pr > |t| Error signal Comp -55.3167 10.2361 12 -5.40 0.0002 signal E. coli -43.2300 10.2361 12 -4.22 0.0012 signal Negative -14.0033 10.2361 12 -1.37 0.1964 signal Non Comp -78.7567 10.2361 12 -7.69 <.0001 signal Salmonella -23.5333 10.2361 12 -2.30 0.0403 signal Water -15.1200 10.2361 12 -1.48 0.1654 Table 9.18 Differences of Least Square Means Effect Signal vs. Signal Estimate Standard DF T Pr > |t| Error value Signal Comp E. coli -12.0867 14.4760 12 -0.83 0.4201 Signal Comp Negative -4l.3133 14.4760 12 -2.85 0.0145 Signal Comp Non Compl 23.4400 14.4760 12 1.62 0.1314 Signal Comp Salmonella -31.7833 14.4760 12 -2.20 0.0485 Signal Comp Water -40.1967 14.4760 12 -2.78 0.0168 Signal E. coli Negative -29.2267 14.4760 12 -2.02 0.0664 Signal E. coli Non Compl 35.5267 14.4760 12 2.45 0.0304 Signal E. coli Salmonella -19.6967 14.4760 12 -1.36 0.1986 Signal E. coli Water -28.1 100 14.4760 12 - l .94 0.0760 Signal Negative Non Compl 64.7533 14.4760 12 4.47 0.0008 Signal Negative Salmonella 9.5300 14.4760 12 0.66 0.5228 Signal Negative Water 1.1 167 14.4760 12 0.08 0.9398 Signal Non Compl Salmonella -55.2233 14.4760 12 -3.81 0.0025 Signal Non Compl Water -63.6367 14.4760 12 -4.40 0.0009 Signal Salmonella Water -8.4133 14.4760 12 -0.58 0.5719 _- ANOVA Analysis of AQ values using SAS System 10 ng T30 all samples Cycle 13 Table 9.19 Type 3 Tests of Fixed Effects Effect Num DF DenDF F Value Pr > F signal 3 8 1.03 0.4284 Table 9.20 Least Square Means Effect signal Estimate Standard DF t Value Pr > |t| Error signal E. coli -25.2700 227.31 8 -0.11 0.9142 signal Negative -487.90 227.31 8 -2.15 0.0641 signal Salmonella -32.5033 227.31 8 -0.14 0.8898 signal Water -20.0700 227.31 8 -0.09 0.9318 Table 9.21 Differences of Least Square Means Effect Signal vs. Signal Estimate Standard DF T Pr > |t| Error value Signal E. coli Negative 462.63 321.46 8 1.44 0.1881 Signal E. coli Salmonella 7.2333 321.46 8 0.02 0.9826 Signal E. coli Water -5.2000 321.46 8 -0.02 0.9875 Signal Negative Salmonella -455.39 321.46 8 -1.42 0.1943 Signal Negative Water -467.83 321.46 8 -1.46 0.1837 Signal Salmonella Water -12.4333 321.46 8 -0.04 0.9701 ANOVA Analysis of AQ values using SAS System 10 ng T30 all samples Cycle 26 Table 9.22 Type 3 Tests of Fixed Effects Effect Num DF DenDF F Value Pr > F signal 3 8 1.03 0.4281 Table 9.23 Least Square Means Effect signal Estimate Standard DF t Value Pr > |t| Error signal E. coli -20.8133 214.28 8 -0.10 0.9250 signal Negative -455.99 214.28 8 -2. l 3 0.0660 signal Salmonella -25.1667 214.28 8 -0.12 0.9094 signal Water -15.0733 214.28 8 -0.07 0.9456 Table 9.24 Differences of Least Square Means Effect Signal vs. Signal Estimate Standard DF T Pr > |t| Error value Signal E. coli Negative 435.18 303.03 8 1.44 0.1889 Signal E. coli Salmonella 4.3533 303.03 8 0.01 0.9889 Signal E. coli Water -5.7400 303.03 8 -0.02 0.9854 Signal Negative Salmonella -430.82 303.03 8 -1.42 0.1929 Signal Negative Water -440.92 303 .03 8 -1.46 0.1838 Signal Salmonella Water -10.0933 303 .03 8 -0.03 0.9742 127 ANOVA Analysis of AQ values using SAS System 1 microgram at different times with background Cycle26 The Mixed Procedure . i- v Q --a‘ ._p- Table 9.25 Type 3 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F signal 1 12 3.84 0.0736 Inc 2 12 0.26 0.7754 signal*1nc 2 12 1.85 0.1987 Table 9.26 Least Smiares Means Effect signal Inc Estimate Standard DF t Value Pr > Error |t| signal*lnc comp T180 25.0633 3.5389 12 7.08 <.0001 signal*lnc comp T0 24.3833 3.5389 12 6.89 <.0001 signal*1nc comp T60 17.2267 3.5389 12 4.87 <.0001 signal*1nc noncomp T180 27.8400 3.5389 12 7.87 <.0001 signal*lnc noncomp T30 25.1500 3.5389 12 7.1 1 <.0001 signal*1nc noncomp T60 30.6733 3.5389 12 8.67 <.0001 Table 9.27 Tests of Effect Slices Effect Inc Num DF Den DF F Value Pr > F signal*1nc T180 1 12 0.31 0.5892 signa1*1nc T30 1 12 0.02 0.8802 signa1*lnc T60 1 12 7.22 0.0198 ANOVA Analysis of AQ values using SAS System 100 ng at different times with background Cycle 13 The Mixed Procedure Table 9.28 Type 3 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F signal 1 12 0.37 0.5521 Inc 2 12 0.44 0.6519 signal*1nc 2 12 3.40 0.0675 Table 9.29 Least Squares Means Effect signal Inc Estimate Standard DF t Value Pr > Error |t| signal*lnc comp T180 29.1933 7.6079 12 3 .84 <.0024 signal*1nc comp T30 54.5800 7.6079 12 7.17 <.0001 signa1*Inc comp T60 33 .9000 7.6079 12 4.46 <.0008 signal*lnc noncomp T180 40.2333 7.6079 12 5.29 <.0002 signal‘lnc noncomp T30 28.2367 7.6079 12 3 .71 <.0030 signal*1nc noncomp T60 37.8033 7.6079 12 4.97 <.0003 Table 9.30 Tests of Effect Slices Effect Inc Num DF Den DF F Value Pr > F signal*lnc T180 12 1.05 0.3251 signal*1nc T30 12 5.99 0.0307 signal*1nc T60 12 0.13 0.7231 1.. ANOVA Analysis of AQ values using SAS System 100 ng at different times with background Cycle 26 The Mixed Procedure Table 9.31 Type 3 Tests of Fixed Effects Effect Num DF DenDF F Value Pr > F signal 1 12 0.00 0.9609 Inc 2 12 0.74 0.4976 signal*1nc 2 12 1.37 0.2920 Table 9.31 Least Squares Means Effect signal Inc Estimate Standard DF t Value Pr > Error |t| signal*lnc comp T180 17.8600 5.0301 12 3.55 0.0040 signal*lnc comp T30 29.6967 5.0301 12 5.90 <.0001 signal*1nc comp T60 17.4233 5.0301 12 3.46 0.0047 signal*1nc noncomp T180 23.9833 5.0301 12 4.77 0.0005 signal*1nc noncomp T30 20.0733 5.0301 12 3.99 0.0018 signal‘lnc noncomp T60 20.3067 5.0301 12 4.04 0.0016 Table 9.33 Tests of Effect Slices Effect Inc NumDF Den DF F Value Pr > F signa1*1nc T180 12 0.74 0.4062 signal*lnc T30 12 1.83 0.2011 signal*1nc T60 12 0.16 0.6924 130 ANOVA Analysis of AQ values using SAS System 100 ng at T30 without background Cycle 13 The Mixed Procedure Table 9.34 Type 3 Tests of Fixed Effects Effect Num DF DenDF F Value Pr > F signal 5 12 1.05 0.4328 .p .‘. Table 9.35 Least Square Means Effect signal Estimate Standard DF t Value Pr > 2‘ Error |t| T‘ signal Comp 54.5800 12.6092 12 4.33 0.0010 1 ‘ signal E. coli 44.7433 12.6092 1.. 3.55 0.0040 signal Negative 21.8467 12.6092 12 1.73 0.1088 signal Non Comp 28.2367 12.6092 12 2.24 0.0448 signal Salmonella 32.2867 12.6092 12 2.56 0.0250 signal Water 23.3767 12.6092 12 1.85 0.0885 Sample Table 9.36 Differences of Least Square Means Signal vs. Signal Estimate Standard DF T value Pr > Itl Effect Error Signal Comp E. coli 9.8367 17.8321 12 0.55 0.5913 Signal Comp Negative 32.7333 17.8321 12 1.84 0.0913 Signal Comp Non Compl 26.3433 17.8321 12 1.48 0.1654 Signal Comp Salmonella 22.2933 17.8321 12 1.25 0.2351 Signal Comp Water 31.2033 17.8321 12 1.75 0.1056 Signal E. coli Negative 22.8967 17.8321 12 1.28 0.2234 Signal E. coli Non Compl 16.5067 17.8321 12 0.93 0.3729 Signal E. coli Salmonella 12.4567 17.8321 12 0.70 0.4981 Signal E. coli Water 21.3667 17.8321 12 1.20 0.2540 Signal Negative Non Compl «6.3900 17.8321 12 036 0.7263 Signal Negative Salmonella -10.4400 17.8321 12 -0.59 0.5691 Signal Negative Water -1.5300 17.8321 12 -0.09 0.9330 Signal Non Compl Salmonella -4.0500 17.8321 12 -0.23 0.8242 Signal Non Compl Water 4.8600 17.8321 12 0.27 0.7898 Signal Salmonella Water 8.9100 17.8321 12 0.50 0.6263 131 III-ArthFlL I.. «fl-III... [w 88.? mm. _ mm a m 887. -._ A .a 2...; -._ .3 =5 “E :52 .35 285 8:... 8 an; m as: 8.8 2...:- _o8.v 8.0 a 8:2: 88.8- 858 we? 88v 4;. _N 88.4 88.8- 33 .82.. 580 8.8- _N 8:: 88.8- 4% .28.. 88v 8.8- _N 28.2 8.: _- 85882 385 58v 8. .4- a 88.8 ES _- omz _éwa 88v m _ .2- a m _ 8.8 88.8- B .8? Sta _2_ A i 28> . E 82:35 3.25:”.— ..awa 8.8. 2.32 mega—am .231— bmd 03:. 9:539..— tofiz 2:. 6399.39 «2:. .333 5.3 :2. 5a.. :5 ass:— 2 280 .62—32 SEQ-.2180? a. ”Em: «nu-<9 .3 EoEfiit< 53.8mm mOZ< 132 888 5:82 $.33 888 2: _m 28. : : _ 8.8 2:8 :3 28;. 88.8 .2522 -283 3.8.8 2.2” _m 88.2 88.3” 2:8 4% .222 $28 5.53. -283 288 3.8 _m 82.4 882 _ :3 35- 28; 88.8 3:52 -343 28.8 8.2“- ...H 82.2 8:8- :58 9:882 28; 88.8 .255. .53 58v 8.0 _m 88.2 a $2:- 3<3 2:882 28; 88v .253. .93 88v 2.:- a 38.2 888- 3? uaa <75 3:850 wits—2: mot—8am =< a: wag—58 8.5 35— Nvd «Ea-_- Ape—SUV 146 h i: . It .. e. .ma- .9. v.3 :3 SE mace. mesa: .m- new ES: cE mace. mmem- Sue- Ex :3 SE mace. mmev- 6%. e. . v ES... eE mace. 2cm- 2e- .3 :3 eE mace. e . m? E- n: :3 cE mace. 32. .me- mew :3 e2 mace. R. . m- .e- 2% ES.» eE mace. 2.3a mt- mew E<3 eE maec. came- eem- .e E5; eE mace. 3.3. 08. .2 ES.» eE mace. 3.3- 8.- cc E<3 eE mace. maem- e. . e- v.2 SS» eE mace. 8.2. S- Sm :3 eE mace. ecem- .3- 3 :3 eE mace. mesm- Nem- m3 ES... eE mace. ~08. SA. «.3 E35 eE mace. .03. e3- new E55 cE mace. ec.- e.ee.- 3m eeeoaoz cE mace. e... .m. .- 3m 3882 eE mace. 3?- eme- n: 3852 9E mace. 3 .- Nee- eem escuacz eE mace. name- as... nee meoeaoz eE mace. 8.3- $3. eme 9:8an SE maec. m.ac.- a. .c.- cam mmeae cE mace. $5 $5 @3225 AUEV ad :aok wanna—U 3°1— Jaon— «E05530 aw:— _u—_M_m he mug—Ah. 05:. acmagfitah: Eozah—ao—sny 35:. 8.355»: 320:5 .a moi—flaw .533 6:: <2: 3:350 mic—:2: moi—cam =< .3 wag—58 as: Bad mvfi 035,—. :5:ch I47 mm. .- m9.- mem ES.. emE m:. 2.2- m?- N. . m ES: cmE ma. .2.- mm- mem ES... eE me. me...- eé..- v.2 ES... cmE ma. 3mm- Ee- mma ES... cmE ma. 8. . m- e..E- .Ea ES.. emE m... E...- mEm- mcm ES... emE mac. m . .2- 2..- mmm ES... cmE mac. Ease.- mem- nee ES... cmE mac. a. .8. mm..- ..mm E<>. emE mac. 3:. m..e- 3m ES... emE mac. ace-m- S- e: ES.. emE mac. Eve.- eev 0mm ES... emE maee. anem- mmm- mmm ES... cmE mace. 3.:- eem- In ES... cmE mace. EEm .- .mm- 92 ES.. emE mace. WE.- E. . m- e. . m ES... cmE mace. .2..- v.3- Mme ES... emE mace. AS: 43.. 92:55. AuEv 04 saun— ucuhsmv .33 JNUA— ~=o.—.—:U aw;— _¢:u_w ac Ugh-_- OEFF acme-«Nm—an—h: Camus-Susanamv 82E 823.25.: .53.... «a moan—am .333 can <75 2:850 wit-:2: 3.5.—am =< .3 w:¢3.:¢ «:5 in: N96 033—. 2...:ch 148 Raw Data of All Samples Including Genomic DNA and Water Samples at Different Concentrations and 30 minutes Hybridization Time Table 9.43 Concentration Hybridization Type Cycle High Low AQ (mC) Temperature of Signal Current Current °C Peak (pA) Peak (pA) 100ng RT EC 13 73.3 -98.5 ~50.7l 100ng RT EC 26 71.3 -93.7 -39.41 100ng RT background 13 71 ~101.9 -95.94 100ng RT background 26 67.5 -9 I .1 -62.03 100ng RT EC 13 62.3 -90 49.75 100ng RT EC 26 60.5 -84.6 -38.7 100ng RT background 13 83.8 -113.5 -117.9 100ng RT background 26 79.6 -98.9 -80.09 100ng RT EC 13 70.7 -86.6 «48.46 100ng RT EC 26 68.9 -82.9 -38.94 lOng RT background 13 88.4 -1 19.1 -121.1 lOng RT background 26 85.1 - 106.2 -84.22 lOng RT EC 13 79.7 -100.9 -54.62 lOng RT EC 26 77.6 -97.2 -43.48 lOng RT background 13 98.5 -128.9 -127.5 lOng RT background 26 95.5 -1 16.7 -87.88 lOng RT water 13 89.9 -1 13.9 -57.92 lOng RT water 26 88 -1 10.1 -52.04 lOng RT background 13 102.7 -l30.9 425.4 lOng RT background 26 99.5 -118.3 -85.55 lOng RT salmonella 13 94.6 -116.9 -60.91 lOng RT salmonella 26 92.3 -1 12.5 49.07 lOng RT background 13 92.7 -120.3 -116.6 lOng RT background 26 89.3 -107.9 -78.74 lOng RT negative 13 84.7 -108.4 -S4.46 lOng RT negative 26 82.1 -103.9 42.6 lOng RT background 13 109.5 -l40.1 -151.5 lOng RT background 26 106.6 -128.7 -l27.l lOng RT water 13 99.6 -1 13.8 -61 .82 lOng RT water 26 97.5 ~109.5 -77.57 lOng RT background 13 163.7 -178.8 ~151.3 lOng RT background 26 160.5 -l67.S -11() lOng RT water 13 155.9 - 1 66.2 -81.44 lOng RT water 26 153.8 -161.6 -66.22 149 (Cont’d) Table 9.43 Raw Data of All Samples Including Genomic DNA and Water Samples at Different Concentrations and 30 minutes Hybridization Time Concentration Hybridization Type Cycle High Low AQ (mC) Temperature of Signal Current Current °C Peak (pA) Peak (pA) lOng RT background 13 94.9 -124.5 -1 16.1 10ng RT background 26 91.4 -1 12.6 -80.62 lOng RT EC 13 86.1 -110.3 -53.95 lOng RT EC 26 83.9 -105.3 -41.55 lOng RT background 13 87.1 -116.2 ~12 1.7 lOng RT background 26 84.5 -104.4 -88.8 lOng RT water 13 79.5 -98.9 -53.6 lOng RT water 26 77.6 -94.8 -42.48 lOng RT background 13 104.9 -134.5 -171 lOng RT background 26 101.4 -122.7 -131 lOng RT salmonella 13 91.6 -106.3 ~57.28 lOng RT salmonella 26 89.5 -103.3 -45.66 lOng RT background 13 96.8 -125.3 -168.9 lOng RT background 26 93.7 -113.4 -116.6 lOng RT negative 13 87.9 -100.3 -54.18 lOng RT negative 26 85.7 -96.6 -42.56 lOng RT background 13 84 -1 14.1 -125.9 lOng RT background 26 80.7 -101.1 -89.92 lOng RT water 13 72.5 -91.5 -56.8 10ng RT water 26 69.7 -88.1 -47.18 lOng RT background 13 109.1 -143.8 -176.6 lOng RT background 26 104.6 -133.8 -148.5 lOng RT water 13 94.4 -109 -55.73 lOng RT water 26 90.8 -104.2 -41.83 10ng RT background 13 102.6 -129.1 -1 18 lOng RT background 26 99.3 -1 17.4 -81.41 lOng RT EC 13 94.2 -115 -56.03 lOng RT EC 26 91.7 -110.5 44.04 lOng RT background 13 104.1 -130.2 442.6 lOng RT background 26 101.1 -117.6 -103.7 lOng RT water 13 95.9 -110.2 -57.13 lOng RT water 26 93.7 -106.2 45.59 10ng RT background 13 97.7 -119.2 -123.6 lOng RT background 26 95 -108 -84.93 lOng RT salmonella 13 90.7 -108.2 -57.55 lOng RT salmonella 26 88.8 -104.3 «16.42 lOng RT background 13 101.1 -130.9 449.5 lOng RT background 26 99.1 -120 -113.3 150 (Cont’d) Table 9.43 Raw Data of All Samples Including Genomic DNA and Water Samples at Different Concentrations and 30 minutes Hybridization Time Concentration Hybridization Type Cycle High Low AQ (mC) Temperature of Signal Current Current °C Peak (uA) Peak (uA) lOng RT negative 13 94.3 -1 10.9 64.83 lOng RT negative 26 91.9 -106.3 -50.45. lOng RT background 13 100.7 -131.9 -155.5 lOng RT background 26 98.3 -121.4 -125.3 lOng RT water 13 85.5 -103.5 -54.35 lOng RT water 26 83.1 -99.2 -43 .09 lOng RT background 13 100.7 -131.9 -155.5 lOng RT background 26 98.3 -12 l .4 -l25.3 lOng RT water 13 92.4 -105.2 -59.17 lOng RT water 26 90.1 -101 -46.35 151 CHAPTER 10. REFERENCES Adams, J. C., M. S. Lytle, et al. (1989). "Comparison of methods for enumeration of selected coliforms exposed to ozone." Applied And Environmental Microbiology 55(1): 33-35. AFNOR (1990). Eaux-methodes d essais. Recuel de Normes Francaises. Paris, La Defense. Alocilja, E. C. and S. M. Radke (2003). "Market analysis ofbiosensors for food safety." Biosensors and Bioelectronics 18(5-6): 841-846. Amann, R. 1., L. Krumholz, et a1. (1990). "Fluorescent-01igonucleotide probing of whole cells for determinative, phylogenetic, and environmental studies in microbiology." Journal Of Bacteriology 172(2): 762-770. APHA (1998). Standard Methods for the Examination of Water and Waste Water. Washington, DC, AWWA,WEF. Armistead, P. M. T., H. Holden (2002). "Electrochemical Detection of Gene Expression in Tumor Samples: Overexpression of Rak Nuclear Tyrosine Kinase." Bioconiugate Chemistry 13(2): 172-176. Audebert, P., Bidan, G., (1985). "Polyhalopyrroles: Electrochemical synthesis and some characteristics." J. Electroanal. Chem. 190: 129-139. Barwick, R. S., D. A. Levy, G. F. Craun, M.J. Beach, and R. L. Calderon. (2000). "Surveillance for Waterbome-Disease Outbreaks - United States. 1997-1998." MMWR 49(SSO4): 1-35. Baudart, J., A. Olaizola, et a1. (2005). "Assessment of a new technique combining a viability test, whole-cell hybridization and laser-scanning cytometry for the direct counting of viable Enterobacteriaceae cells in drinking water." FEMS Microbiology Letters 243(2): 405-409- Beaglehole, R., A. Irwin and. T. Prentice (2004). The World Health Report 2004. Geneva, Switzerland, World Health Organization: 1-96. Bej, A. K., M. H. Mahbubani, et al. (1991a). "Polymerase chain reaction-gene probe detection of microorganisms by using filter-concentrated samples." Applied And Environmental Microbiology 57(12): 3529-3534. Bej, A. K., S. C. McCarty, et al. (1991b). "Detection of coliform bacteria and Escherichia coli by multiplex polymerase chain reaction: comparison with defined substrate 153 and plating methods for water quality monitoring." Applied And Environmental Microbiology 57(8): 2429-2432. Bej, A. K., R. J. Steffan, et al. (1990). "Detection of coliform bacteria in water by polymerase chain reaction and gene probes." Applied And Environmental Microbiology 56(2): 307-314. Bhattacharyya, J ., D. Read, et al. (2005). "Biosensor-based diagnostics of contaminated groundwater: assessment and remediation strategy." Environmental Pollution 134(3): 485-492. Blackburn , B. Y. J., Craun GF, Hill V, Levy DA, Chen N, Lee SH, Calderon RL, Beach MJ. (2004). Surveillance for WAterbom-Disease Outbreaks Associated with Drinking Water-United States 2001-2002. Atlanta, GA, Public Health Prevention Service, Epidemiology Program Office, CDC, USA: 23-45. Boyle, A., Genies, E., Fouletier,M. (1990). Electroanal. Chem. 279: 179. Burtscher, C., P. A. Fall, et a1. (1999). "Detection of Salmonella spp. and Listeria monocytogenes in suspended organic waste by nucleic acid extraction and PC R." Applied And Environmental Microbiology 65(5): 2235-2237. Calabrese, J . P. and G. K. Bissonnette (1990). "Improved membrane filtration method incorporating catalase and sodium pyruvate for detection of chlorine-stressed coliform bacteria." Applied And Environmental Microbiology 56(1 1): 3558-3564. Carroll, L. B, Adams, J.K., Sullivan, M., Besser, J.M., Bartkus, J.M (2001). A real-time SYBR Green PCR method for the detection of enterotoxigenic E. coli. ASM General Meeting, Orlando, USA. Cheung, W. H. S., K.C.K. Chang, R.P.S. Hung, and J.W.L. Kleevans (1990). "health effects of beach water pollution in Hong Kong." Epidem. Infect. 105: 139-162. Chiang, C. k., C. R. Fincher, Jr., Y.W. Park, A.J. Heeger, H. Shirakawa, EJ. Louis. S.C. Gau, and AG. MacDiarmid (1977). "Electrical conductivity on dOped polyacetylene." Phys. Rev. Len-39(17): 1098. Chrost, R. J. (1991). Microbial Enzymes in Aquatic Environments. New York. Springer- Verlag. Clark, J. A. (1980). "The influence of increasing numbers of nonindicator organisms upon the detection of indicator organisms by the membrane filter and presence- absence tests." CAN. J. MICROBIOL. 26(7): 827-832. Colwell, R. R., Grimes, DJ. (2000). Nonculturable microorganisms in the environment. Washington, DC, ASM Press. 154 Davey, J. M., Ralph, S. F., Too, C. O. and Wallace,G. G. (1999). "Synthesis, characterisation and ion transport studies on polypyrrole/p0lyvinylphosphate conducting polymer materials." Synthetic Metals 99(3): 191-199. Delabre, K., V. Dile, M.R. De Roubin, D. Gatel, F. Poty and J. Cavard (2001). New analytical tools for distribution system surveillance. Proceedings of AWWA- Annual Conference, Washington, DC, American Water Works Association. DeLong, E. F. (1993). Single-cell identification using fluorescently labeled, ribosomal RNA-specific probes. Handbook of Methods in Aquatic Microbial Ecology. B. F. S. P.F. Kemp, E.B. Sherr and J .J. Cole. Boca Raton, FL, Lewis Publishers: 285— 294. i Devreux, F. G., F., Nechtschein, M., Villeret,B. (1987). "ESR investigation of polarons P1 and bipolarons in conducting polymers:: the case of polypyrrole." Synth. Met. 18: . 89-94. Diaz, A. F., and Castillo, 1.1. (1980). J. Chem. Soc. Chem. Comm. 397. 1' Diaz, A. F., and J. Bargon (1986). Electrochemical Synthesis of Conductive Polymers. Handbook of Conducting Polymers. T. A. Skotheim. New York, Marcell Dekker, Inc. 1: 727. Diaz, A. F., Castillo, J.I. , Logan, J.A. and W.Y. Lee (1981). "Electrochemistry of conducting polypyrrole films." J. Electroanal. Chem. 129: l 15. Duke, C. B., Schein, LB (1980). "Organic solids: is energy-based theory enough?" Phys. Today 33: 42-48. Dury, M. E. (1981). Synthesis and Characterization of polyacetylene, (C H)x and its derivates. Department of chemistry, University of Pennsylvania. Edberg, S. C. and M. M. Edberg (1988). "A defined substrate technology for the fig enumeration of microbial indicators of environmental pollution." The Yale Journal Of Biology And Medicine 61(5): 389-399. iii F eng, P., R. Lum, et al. (1991). "Identification of uidA gene sequences in beta-D- glucuronidase-negative Escherichia coli." Applied And Environmental Microbiology 57(1): 320-323- Feng, P. C. and P. A. Hartman (1982). "Fluorogenic assays for immediate confirmation of Escherichia coli." Applied And Environmental Microbiology 43(6): 1320- 1329. 155 Fricker, E. J. and C. R. Fricker (1996). "Use of two presence/absence systems for the detection of E. coli and coliforms from water." Water Research 30(9): 2226-2228. Friend, R. H. (1993). Rapra Review Report. Conductive Polymer 11. 6: 23. Gambhir, A., Gerard M, Jain SK, Malhotra BD. (2001). "Characterization of DNA immobilized on electrochemically prepared conducting polypyrrole-polyvinyl sulfonate films." Appl Biochem Biotechnol. 96(1-3): 303-309. Geldreich. BB, K. R. F., J.A. Goodrich, E.W. Rice, R.M. Clark and D.L. Swerdlow (1992). "Searching for a water supply connection in the Cabool, Missouri disease outbreak of Escherichia coli 01572H7." Water Res. 26: 1 127-1 137. Genies, E. M., G. Bidan and AF. Diaz (1983). J. Electrochem. Soc. 149: 101. Grabow, W. O. and M. du Preez (1979). "Comparison of m-Endo LES, MacConkey. and Teepol media for membrane filtration counting of total coliform bacteria in water." Apflied And Environmental Microbiology 38(3): 351-358. Guiseppi-Elie, A. (1983). Synthesis and characterization of polyacetylene: 1. Stabilbity of doped polyacetylene, 2. Surface chemistry of polyacetylene. Material Science and ENgineering, MIT. Heid, C. A., J. Stevens, et al. (1996). "Real time quantitative PCR." Genome Research 6(10): 986-994. Iqbal, 8., J. Robinson, et al. (1997). "Efficiency of the polymerase chain reaction amplification of the fluid gene for detection of Escherichia coli in contaminated water." Letters in Applied Microbiology 24(6): 498-502. Jelen, F ., Yosypchuk B, Kourilova A, Novotny L, Palecek E. (2002). "Label-free determination of picogram quantities of DNA by stripping voltammetry with solid copper amalgam or mercury electrodes in the presence of copper." Anal Chem. 74(18): 4788-4793. J iang, M. and J. Wang (2001). "Recognition and detection of oligonucleotides in the presence of chromosomal DNA based on entrapment within conducting-polymer networks." Journal of Electroanalytical C hemistfl 500(1-2): 584-589. Joux, F., Lebaron, P. (2000). "Use of fluorescent probes to ases physiological functions of bacteria at single-cell level." Microb. Infect. 2: 1523-1537. J uck, D., J. Ingram, et a1. (1996). "Nested PCR protocol for the rapid detection of Escherichia coli in potable water." Canadian Journal of h-licrobiology 42(8): 862- 866. 156 Kanazawa, K. K., Diaz, A.F., Geiss. R.H., Gill, W.D., Kwak, J.F., Logan, J.A., Rabolt. J.F., Street, GB (1979). "Organic Metals: polypyrrole, a stable synthetic metallic polymer." J. Chem. Soc. Chem. Commun: 854-855. Kanga, E. T., K. G. Neoha and K. L. Tanb (1998). "Polyaniline: A polymer with many interesting intrinsic redox states." Progress in Polymer Science 23(2): 277-324. Kelley, S. O., and J. K. Barton (1999). "Electron Transfer Between Bases in Double Helical DNA." Science 283(5400): 375-381. Kerman, K. O., Dilsat; Kara, Pinar; Erdem, Arzum; Meric, Burcu; Nielsen, Peter E.; 02302, Mehmet (2003). "Label-free bioelectronic detection of point mutation by using peptide nucleic acid probes." Electroanalysis 15(7): 667-670. . Kilian, M. and P. Bulow (1976). "Rapid diagnosis of Enterobacteriaceae. 1. Detection of , bacterial glycosidases." Acta Pathologica Et Microbiologica Scandinavica. A! Section B, Microbiology 848(5): 245-251. Kim, M. N., H. H. Park, et a1. (2005). "Construction and comparison of Escherichia coli 1 ‘ ’ whole-cell biosensors capable of detecting aromatic compounds." Journal of Microbiological Methods 60(2): 235-245. Korri-Youssoufi, H., F. Garnier, P. Srivastava, P. Godillot, and A. Yassar. (1997). "Toward Bioelectronics: Specific DNA recognition based on an Oligonucleotide- Functionalized Polypyrrole." J. Am. Chem. Soc 119: 7388-7389. Lebaron, P., P. Catala, et a1. (1997). "A new sensitive, whole-cell hybridization technique for detection of bacteria involving a biotinylated oligonucleotide probe targeting rRNA and tyramide signal amplification." Applied and Environmental Microbiology 63(8): 3274-3278. LeChevallier, M. W., S. C. Cameron, et a1. (1983). "New medium for improved recovery of coliform bacteria from drinking water." Applied And Environmental Microbiology 45(2): 484-492. LeChevallier, W. (1990). "Coliform bacteria in drinking water: a review." J. AWWA 82: 74-86. 31 Lee. S. H., D. A. Levy, G. F. Craun, M. J. B, and R. L. Calderon, (2002). "Surveillance for Waterbome-Disease Outbreaks -United States, 1999-2000." MMW R 51(8808): 1-28. Loge, F. J., R. W. Emerick, et al. (1999). "Development of a fluorescent 16S rRNA oligonucleotide probe specific to the family Enterobacteriaceae." Water Environment Research 71(1): 75-83. 157 Lucarelli, F ., A. Kicela, et a1. (2002). "Electrochemical DNA biosensor for analysis of wastewater samples." Bioelectrochemistry 58(1): 1 13-118. Lucarelli, F ., G. Marrazza, et al. (2004). "Carbon and gold electrodes as electrochemical transducers for DNA hybridisation sensors." Biosensors and Bioelectronics 19(6): 515-530. Malinaukas, A., and J. Kulys (1978). "Alcohol, lactate and glutamate sensors based on oxidoreductases with regeneration of nicotinamide adenine dinucleotide." Anal. Chim. Acta 98: 31. Malmros, M. K., J. Gulbinski III andW.B. Gibbs (1988). "A-semi conductive polymer film sensor for glucose." Biosensors 3: 71. Mathew, F. P. and E. C. Alocilja (2005). "Porous silicon-based biosensor for pathogen detection." Biosensors and Bioelectronics 20(8): 1656-1661. McFeters, G. A., S. C. Cameron, et al. (1982). "Influence of diluents, media, and membrane filters on detection fo injured waterborne coliform bacteria." Applied And Environmental Microbiology 43(1): 97-103. McFeters, G. A., J. S. Kippin, et a1. (1986). "Injured coliforms in drinking water." Applied And Environmental Microbiology 51(1): 1-5. McKillip, J. L., L.-A. Jaykus, et a1. (1998). "rRNA stability in heat-killed and UV- irradiated enterotoxigenic Staphylococcus aureus and Escherichia coli 015721-17." Applied and Environmental Microbiology 64(11): 4264-4268. MDEQ (1997). A Strategic Environemntal Quality Monitoring Program for Michigan Surface Waters. Lansing, Michigan, Michigan Department of Environmental Quality. Millan, K. a. M. S. (1993). "Sequence-selective biosensor for DNA based on _.1 electroactive hybridization indicators." Anal Chem. 65(17): 2317-2323. Minehan, D. S., K. A. Marx, et al. (1994). "Kinetics of DNA binding to electrically conducting polypyrrole films." Macromolecules 27(3): 777-783. Mittelman, M. W., M. Habash, et a1. (1997). "Rapid detection of Enterobacteriaceae in urine by fluorescent 16S rRNA in situ hybridization on membrane filters." Journal of Microbiological Methods 30(2): 153-160. Morris, B. A. and A. Sadana (2005). "A fractal analysis of pathogen detection by biosensors." Biophysical Chemistry 113(1): 67-81. 158 Mostany, J. and B. R. Scharifker (1997). "Impedance spectroscopy of undoped. doped and overoxidized polypyrrole films." Synthetic Metals 87(3): 179-185. MSU (2004). Michigan State University Report for MSU Watershed Management Project. Muhammad-Tahir, Z. and E. C. Alocilja (2003). "A conductometric biosensor for biosecurity." Biosensors and Bioelectronics 18(5-6): 813-819. Muhammad-Tahir, Z. a. and E. C. Alocilja (2002). A Disposable Membrane Strip Immunosensor. First IEEE International Conference on Sensors - IEEE Sensors 2002, Jun 12-14 2002, Orlando, FL, United States, Institute of Electrical and Electronics Engineers Inc. Obst, U., I. Hubner, M. Wecker and D. Bitter-Suerrnann (1989). "Immunological method using monoclonal antibodies to detect Enterobacteriaceae in drinking water." Agua 38: 136—142. Olsen, G. J., D. J. Lane, et al. (1986). "Microbial ecology and evolution: a ribosomal RNA approach." Annual Review Of Microbiology 40: 337-365. Ootsubo, M., T. Shimizu, et a1. (2003). "Seven-hour fluorescence in situ hybridization technique for enumeration of Enterobacteriaceae in food and environmental water sample." J Appl Microbiol 95(6): 1 182-1 190. Ouvemey, C. C. and J. A. F uhrman (1999). "Combined microautoradiography-I6S rRNA probe technique for determination of radioisotope uptake by specific microbial cell types in situ." Applied And Environmental Microbiology 65(4): 1746-1752. Ozkan, D., Erdem A, Kara P, Kerrnan K, Meric B, Hassmann J, 02302 M. (2002). "Allele-specific genotype detection of factor V Leiden mutation from polymerase chain reaction amplicons based on label-free electrochemical genosensor." Anal Chem. 74(23): 5931-5936. Palecek, E. (1960). "Oscillographic polarography of highly polymerized deoxyribonucleic acid." Nature 188: 656—657. Palecek, E. (1988). "Adsorptive transfer stripping voltammetry: determination of nanogram quantities of DNA immobilized at the electrode surface." Anal Biochem. 170(2): 421-431. Palecek, E., M. Fojta, et al. (2002). "New approaches in the development of DNA sensors: hybridization and electrochemical detection of DNA and RNA at two different surfaces." Bioelectrochemistry 56(1-2): 85-90. 159 “‘23 .I- _ Pande, R., Ruben GC, Lim JO, Tripathy S, Marx KA. (1998). " DNA bound to polypyrrole films: high resolution imaging, DNA kinetics and internal migration." Biomaterials 19: 1657- 1667. Piunno, P. A. E., J. Watterson, et a1. (1999). "Considerations for the quantitative transduction of hybridization of immobilized DNA." Analytica Chimica Acta 400(1-3): 73-89. Poulsen, L. K., F. Lan, et a1. (1994). "Spatial distribution of Escherichia coli in the mouse large intestine inferred from rRNA in situ hybridization." Infection and Immunity 62(11): 5191-5194. Prescott, A. M. and C. R. Fricker (1999). "Use of PNA oligonucleotides for the in situ detection of Escherichia coli in water." Molecular and Cellular Probes 13(4): 261- 268. Radke, S. M., and EC. Alocilja (2005). "A high density microelectrode array biosensor for detection of E. coli OlS7:H7." Biosens Bioelectron 20(8): 1662-1667. Regnault, B., S. Manin-Delautre, et a1. (2000). "Oligonucleotide probe for the visualization of Escherichiacoli/Escherichia fergusonii cells by in situ hybridizationzspecificity and potential applications." Research in Microbiology 151(7): 521-533- Rice, E. W., M. J. Allen, et a1. (1990). "Efficacy of [beta]-glucuronidase assay for identification of Escherichia coli by the defined-substrate technology." APPL. ENVIRON. MICROBIOL. 56(5): 1203-1205. Rodriguez, M. l. and E. C. Alocilja (2005). "Embedded DNA-Polypyrrole Biosensor for Rapid Detection of Escherichia Coli." Sensors Journal, IEEE 5(4): 733-736. Rompre, A., P. Servais, et a1. (2002). "Detection and enumeration of coliforms in drinking water: current methods and emerging approaches." Journal of Microbiological Methods 49(1): 31-54. Roszak. D. B. and R. R. Colwell (1987). "Survival strategies of bacteria in the natural environment." Microbiological Reviews 51(3): 365-3 79. Saoudi, B., C. Despas, et a1. (2000). "Study of DNA adsorption on polypyrrole: Interest of dielectric monitoring." Sensors and Actuators, B: Chemical 862(1): 35-42. Saoudi, B., N. Jammul, et al. (1997). "DNA adsorption onto conducting polypyrrole." Synthetic Metals 87(2): 97-103. Sartory, D. P. (1995). "Improved recovery of chlorine-stressed coliforms with pyruvate supplemented media." Water Science and Technology 31(5-6): 255-258. 160 Satoh, M., Kaneto K, Yoshino, K. (1986). "Dependences of electrical and mechanical properties of conducting polypyrrole films on conditions of electrochemical polymerization in an aqueous medium." Synth. Met. 14: 289-296. Seidler, R. J., T. M. Evans, et a1. (1981). "LIMITATIONS OF STANDARD COLIFORM ENUMERATION TECHNIQUES." Journal of the American Water Works Association 73(10): 538-542. Seo, K. H., B. R., Hartman NF, Campbell DP (1999). "Development of a rapid response biosensor for detection of Salmonella typhimurium." Journal of Food Protection 62(5): 431-437. Sheridan, G. E. C., C. 1. Masters, et a1. (1998). "Detection of mRNA by reverse transcription-PCR as indicator of viability in Escherichia coli cells." Applied and Environmental Microbiology 64(4): 1313-1318. Shidmidzu, T. (1987). "Functionalized conducting polymers for development of new polymeric reagents." React. Polym 6: 221. Shirakawa, H., E. J. Louis, AG. MacDiarmid, C.K. Chiang, and A.J. Heeger (1977). "Synthesis of electrically conductive organic polymers: Halogen derivates of polyacetylene, (CH)x." J .Chem. Soc., Chem. Commun: 578. Singhal, P., Kuhr WG. (1997). "Ultrasensitive voltammetric detection of underivatized oligonucleotides and DNA." Anal Chem. 69(23): 4828-32. Straub, T. M., 1. L. Pepper, et a1. (1995). "Removal of PCR inhibiting substances in sewage sludge amended soil." Water Science and Technology 31(5-6): 31 1-315. Su. W., and J. O. lroh (1999). "Electropolymerization of pyrrole on steel substrate in the presence of oxalic acid and amines." Electrochimica Acta 44(13): 2173-2184. Takenaka, S., K. Yamashita,M. Takagi,Y. Uto,H. Kondo (2000). "DNA Sensing on a DNA Probe-Modified Electrode Using F errocenylnaphthalene Diimide as the Electrochemically Active Ligand." Anal. Chem. 72: 1334-1341. Thompson, L. 1., Janusz Kowalik, Mira Josowicz, and J iri Janata (2003). "Label-Free DNA Hybridization Probe Based on Conducting Polymer. " J. Am. Chem. Soc 125: 324-325. Toranzos, G. A., A. J. Alvarez. et al. (1993). "Application of the polymerase chain reaction technique to the detection of pathogens in water." Health-Related Water Microbiology 1992; Water Science and Technology 27(3-4): 207-210. 161 Tsen, H. Y., C. K. Lin, et a1. (1998). "Development and use of 16S rRNA gene targeted PCR primers for the identification of Escherichia coli cells in water." Journal Of Applied Microbiology 85(3): 554-560. Updike, S. J., and Hicks, GP. (1967). "The enzyme electrode." Nature 214(986-988). USEPA (1986). Ambient Water Quality Criteria for Bacteria. Washington. DC, United States Environmental Protection Agency. USEPA (2002). Implementation Guidance and for Ambient Water Quality Contamination for Bacteria. Washington, DC., United States Environmental Protection Agency. Waage, A. S., T. Vardund, et a1. (1999). "Detection of low numbers of pathogenic Yersinia enterocolitica in environmental water and sewage samples by nested polymerase chain reaction." Journal Of Applied Microbiology 87(6): 814-821. Wang, J. (2000). Analytical Electrochemistyy. New York, Wiley-VCH. Wang, J ., D. Xu, D.K. Kawde, A.N.and Polsky. (2001). "Metal nanoparicle-based electrochemical stripping potentiometric detection of DNA hybridization." Anal. Chem. 73: 5576-5581. Wang, J. and M. Jiang (2000). "Toward genolelectronics: Nucleic acid doped conducting polymers." Langmuir 16(5): 2269-2274. Wang. J., Kawde AN. (2002). "Amplified label-free electrical detection of DNA hybridization." Analyst 127(3): 383-386. Wang, J., M. Jiang, A. Fortes and B. Mukherjee. (1999). "New lobe-free DNA recognition based on doping necleic acid probes within conducting polymer films." Anal. Chim. Acta 402: 7-12. Wang. J., Polsky, R., Merkoci, A., and Turner, KL. (2003). "Electroactive beads for ultrasensitivve DNA detection." Langmuir 19: 989-991. Wang. J. X. C., G. Rivas,H. Shiraishi. (1996). Anal. Chim.Acta 326: 141. Watterson, J., P. A. E. Piunno, et a1. (2002). "Practical physical aspects of interfacial nucleic acid oligomer hybridisation for biosensor design." Analytica Chimica Acta 469(1): 115-127. Way, J. S., K. L. Josephson, et al. (1993). "Specific detection of Salmonella spp. by multiplex polymerase chain reaction." Applied And Environmental Microbiology 59(5): 1473-1479. 162 *“.m ' m l A“ Wessendorf, M. W. and T. C. Brelje (1992). "Which fluorophore is brightest? A comparison of the staining obtained using fluorescein, tetramethylrhodamine. lissamine rhodamine, Texas red, and cyanine 3.18." Histochemistry 98(2): 81-85. WHO (2004). Water, sanitation and hygiene, World Health Organization. Wnek, G. E. (1980). Synthesis and properties of electrically conducting polymers. Polymer Science and Engineering, University of Massachusetts. Yang, X., G. Johansson, D. Pfeiffer and, F. Scheller (1991). "Enzyme electrodes for ADP/ATP with enhanced sensitivity due to chemical amplification and intermediate accumulation." Electroanalysis 3: 659-663. (2004). Surveillance for waterbome-disease outbreaks associated with recreational water--United States, 2001—2002. Atlanta, GA, Public Health 4 Prevention Service, Epidemiology Program Office, CDC, USA: 1-22. " Yoder JS, B. B., Craun GF, Hill V, Levy DA, Chen N, Lee SH, Calderon RL, Beach MJ. P Zachar, V., R. A. Thomas, et a1. (1993). "Absolute quantification of target DNA: a simple ' " competitive PCR for efficient analysis of multiple samples." Nucleic Acids Research 21(8): 2017-2018. 163 1111111111111111