I .12. .V .o 1.1. 1.; ~ .. O. t I? 3.? - h ..I o. O. V- 2.13... o 05. VI! I 3. l 7.... .V (1..... V V .c‘ .QVI .J I‘: v .0. ,V. .V.‘ J. .VV . V VI. . - V 0". .19. -..O . V _ V 0 . V. t . V. _ VV. I .. 3...?! :V... . .2 1.5789. 136...}. 313:9... VV. VEST. n IOA‘I‘VO.I‘. I. Cu. . numHUCI‘Jl I ouVQ...v.1V»: 3133:... 9o 3 . .c: .o . V I... 4-. V I V ‘V d; . . .OVCIOIO. ‘10.: V... -V .VVCIVVVO. sol... .t. V: t I. IVVI.V . I 1.. V 2:. 3.... \c vIQ . I .1 . .1 .V. I 3.3.1.... V. O... C. .9 . I p . . l V . V.‘ 1.. I._'.... J! r 0.. V . V .IIV . 7'. 0 IV... )VOI‘I. .V .9. 1“. Q. 5L: :QOVt‘I .03 III I .. III 3.1 a I. \ 1-. at .4\n.~ 33.2.4: I 1‘. 3.3... II. \ V I .. V .V . . I \511. 1.01.3.5; o): u V . . 3.0 . 0:5 Q . 7:. tr .1 Vi I D .4 . I . .V V . .. V..u.V...V.w:. .Vq. n... .V 1,. Uta-Q O5 '1‘ V .u'J. I. V A . I I ‘1. . .V .V. VQVWu-oV‘uAVQKo' .V INN-Vlvfi. . - .11... a . 3.. V V" o. “II-”vanqulé (“at or. ”1‘!” m...“ .0 ‘vé . .. ”a fin V V «II. V .1 - . — V in? .q. I» V.....b.V.V.VVVn.4I Jug»? Vnwm... VI. 2 . . MB... . I I. . a V . I a I. . V V .V . V. $13. .. hmVVVVVVVVHEVVVV... «VII: . . .V; “V.“ .1: V. .o c 1 . R V . . 1.5;. JVVVVVVVVVVVVVVVVV .VV _. .VVVVV I. V. ..+ .V V . .V . I} I. u I fit .I’Iu'..&.‘w C 1. a.“ . q 1“ . V . V. .VVV....V V .r .V. V. . V V I . V V. . I $1.2}. V, :3. 5 xts‘iuqaa \ .v V. Vega” V t ~I .ou’JaI‘ .Iv‘. .. in... V. .VVVVVVVVIVVVVVI V..." .V V . .V V . . . . o N 10 Q V o Ethan. ...VIV Va... not V n V ‘0. J . , .132." Q 5. V .t . I. in Q . in V V 0. .AV ‘ V 1.4! VVVI I...) )6..9.....II.VQ«..V1.IS “‘35.... .- Iua $3. I“ .3 V I \o...IVV.‘I..o a... . III...x ‘\ 3.“) k V ‘~I‘o.-c(d‘ I. ‘ 1o. l....t.:..x.\. .I. .V.._....S$I?af.).".1i:.tzc I28 . . )‘r.t‘.u‘ ova-1.6.: has» :3 a! V. VP...” .310. . .V V V V . V V V .1. V V a . V . I. I O. V .Q. .I. O‘Cftt V . .V...\. V9 V I: I... (I. .- 1If-QI-I . V I .II. VI.-V . . .VI . .I ‘Vo I .3 fl. .1. V .- V V V V... .V. o I -z . . V. 3...»... 31: o...'.«.~. o .. a. . V V V Y ~.V 1.. IV .a II II I . 2‘ VI 3 . V a: .I .V . . V V . 411 . )L . o..J.J. . V l.w1.. ‘ \l 1. .. ..£.. ..-. VI.~..V&.. .c.... I; .n..-.~V. . V $ 31... I‘VVOIIV oV v . . .V I. V. 1.3.9.53“. 2.... 2:: ...V..V...II..\... . I. .Q :9 .V . 'Q 0"...-. .V V 0.9.13.1}...3 V .1c 0.. I. . .u. .. 0. V. V . . .‘3 V 1. ‘1! ’4‘ ho 0. V V~V....V..I\f..v| V. . .3 V .V.‘ 2. V .VJV. . (2!... 153.32.. V I1. .a... V ..J .VOJI .VVIV. .0...\:. L . I... :.|V.31 3., IV‘V‘OV’JV ‘1‘... .i? 11... I IV. 3 .3 V... .n V :l .V. I\! .62! . IV. 9.. ..Q . I. IT? .- 03.....11. “I!“ .1 ‘ V V.::.“.\‘n|. . o 3 . J. .0... :3... 1... Q... on .. L3: .31.... .V: . V1 .12... . .V. VI. . V... (..V$ I 9...! 1..".‘9: o... $ .V‘Vfln:: Q .5. QIJI. I . .2i 0:9: 0. . L‘OOO .19. V. I"... I. . 0‘. c. A. :3. .0. Z. V V V :3‘3,‘V o... I 3.1-1 . . V V V V . . V V. v V.l.cI-n.00.\.. V .VV: . .V V V 1 V . VV . . . V V V . . V V V V V V . V . 1.-II1. V V. .V . V... 1. V V . . V ..'.o‘.. . .. V V V .V . V V . sIVVVVVInIQ . . V.... ..1 .30... o 3...... . IVVV V .' I004. son...) 3;. :1 .0... V. VII. 7. .V. .I. ..V I .u#9 n .V V V . 1 V . 1 V V. . V. O V V V . V . V . .V 1 V V V 1 . V V V V V V V V . .V VI .6....._ . VI- V . . 2! I .Q... . 1 .39 r .1...” V V . V V. I. o Jul-ona: I {I . u 0.... I9 I}... . onIVI. 93'. IV! . :05. V. 0...... I. V . V I V V... VIII‘I‘ a. ... .0...O :Vu: 1... V '4. :VI- I... . .20 .- n o... 9.1.02.0}. OV¢VII V .3? pic... 730...... V V- .. «I 3934313.... . V... II 7 _ .3 .V «I. 3. V IA 0- " .9...s‘ ‘- .§U.I..\3)Y.¢iv {I cog-"TIL..- I.“ ’IVV‘P. to 5.9.01. C. ‘VI .‘0’- ‘0: . I . . V uVL‘I' 135.com V2"); [Pa-.35., _ I r Vat} u...‘ ”.96.!!53i Vtii’s: .X: I O, V .VoEToI Us a. - 37'. AI . I.. o 1.. on... V . .01.. :0. .lov VIVA‘V: 'Y‘VI- .0. :0- V -. ul-l 3.... o o. ’21:}... to .V. 3.. A 091521.. c. . V VIV’VxI . 0.4 I?! I V . V . 1 b. V.... :0 O ..3 V $103; a . II. V .V I. 2.9 9.; . .0 . V .. co; . .:.|.3£. I ‘ ni‘i’tl.t,:.f.v a, ’.’.f of... lo IJIIK‘! :jg’gfioi ‘v-I‘Io I. 2“... {IVs-U {no .- 0'.-. ‘03} 02.32’i .V .VVZV;OVV IQVQQVVDDZV’IOISIEL‘I 1..-I59... 3V..- :9. .5...o... . 9.2 ,ctVl 31.1 o. 3". h in!” I; s V a...t0¢$.le,xou-I!V1~\fia’a 73?. I .3. 32:!3: 1".‘I II V‘\(‘ I. ‘3... a 0.3-. \J V... :30. {Va-121.4: VIA-V fl..-—ookr.o! 1.9.7 It...o;.1..3t. . I... V.... VVI.V$O..\9FI It"! .tla. U. .I V V 1. O V . .V .4 vocv...‘ :51 V. . VVVIV nt.a§"v 5.... IVI II. .140 O 14.0 A .VQoV'QIV-f .*_ .t. no ‘o‘IIVC . I. C‘ . S 1.2. ..o I. .V. I . V V. V V . I. . of V . II. . . 1..... .l V VOA.) I’1Xo 0-: I: .VIV.u..I.IIV. VIVI‘IAI V... _V . 0L 921....er ..II. 1.23? .‘ V1..-V. .0. I II... V... .V1 I V1.0 .II. 13...; 2.19... .V .. .QIIIIA 9.7. .V- V V I AIVOQI .. V... flu. ZV O. I. .0205. I V .. V . O J. u .3 '9'... . V. .0 V. 1.... ..1v~.|.Z - a .- .Q. V I. V. .IV 0.5. 5. =6 . I...!oo\ It ‘vV‘Jfin‘ 1"... r .I .V .. V V . \V‘..&‘v DVVQSI‘IVIE O? V V}. V . V V V .. \V I...A-I.1~£IzOl-K_\'V. V V VI¢.V‘VI-.UI.U..:.\‘CJ: Ut3§an \l V . V I n . z .0. . V\ V no.4... .V . V O V V a . ‘.. I a 0 V 'c V V . I V: I. . V. .0 ..V _. VII. .1 . . .. I V .V . . _ I.. .‘r.. .u o....nu .5 .- . I ..s V . o I I. I l . I V . V :59“ P .IIoV V.’IVIVI .. V VOOIVO .V! :- r¢ . .IoVVoi. a V .. I-VOVIVI; ... V V..... ... 1V. . . V O. . V. . V V . V ._V ‘ o. to. II‘ . I I V. .. III.) l OI '.II a I... VVVV . .. VVIVOIIVIVI IV... V V' o I .V. V . D. . . o o . I. . V .. V . I - Q‘: 0 . u A... AIID .1 I V . . . . 1- V II I. I. V I V . .V V .V. VII :0 o. . I V o .I.IV § O. I n o . I .-o . o 3 0 V . . o .I V. V» V. .. . V V . I... I V. . v Va}? o I. . a V... . . a. ..I 1...... 911' . .gV 1‘). II- IV‘IV. I 3 I V I I V. . V a 7.. V V. .‘..2 II: IIIIoIJI’tV‘C Ih‘cIQo‘V.‘ I p . .I .fiV I 0' . I . .a‘IVzu‘vIVDA V . . I 0.. .. . QIVIV‘II L?! p o .. . c u o I. V . V . I o I I .V . VI k I D‘ 0": . I. . V I. . . V V I . . ‘!.—VV ..voV'oV.o\ II .. . . . .V . . . . . I . . o .I. o .Vl‘ VII.\| '1 .. I o a ._ . d 0. I V. I . . ‘. 5.0V‘I1Io‘ V.-I . I. . . V& V. . . V O.\ 1..):PVVI1. . V. .a V. II. . V1. . 3 .. V V . . I ’0 .V - . VV1 ‘ I; o I Of-IIOI I‘ III 0.- .. 9 . I oo. V . t V V . V V. I V o . .0 s 1 I Iv‘u V (AV-IzooIV‘IbIoo. .w‘\. I . V I .. 1t V O T. 97' ~I ..V. v.0... I u. .VVV‘!III'I .\‘ III V . VV 9 . . . . .i I .V I V... o \ II. 2.... .V... I V... I .. u . . g V a .V o. . o. . VJ... . . o. . O V. V. 1o\.t _ V V o v _ II V. 4 . V V. _ .V . V. I . a. V. o I. 0 V . VI . . . o I. V . n J a V. a o I A .l v . a .. o .-.O.. 4: I VIII... .. I. I o I .V -V 9.... o . I I o. I . Ia . VVVv . Z. V V . _ _ . . I o :I. V V. o o . . v V w. .Ia O- V. . I O . V. V. V. I :11. V V . . o o I. .. —V o a I: . - I. I. o . a . V. c. V o . V I o I . I 3 I .V V V o .. . . . . u . . . V c . . u. . V o . . V. V V . I I w. . V: V .. ..V V-.. . . . .V V. .1 . . 1. ..V . .. I V I O . . V. I > I . I . n .0 g . .Io . C on > V .1 . . .I I . I . .. ‘. J . V .V. . u . . I I . . . o. . 3.... V .V .V r. .3. u . I . V V. ._—I. o. a I V . . v. . IV. .I II V. v. o b V... . V o I V. .0. a V. :q .. . .. .V; I . . — JVV V V l. V I. . a I .- V . I 1. I .. . . .. .u. I o | V .V .. . Iv . . . . V . In ‘ a . .V I .I u . K; . o . 1 V V o o o .. ¢ .0 a V I . V I. o .. V . .V. V . .0 II '1 .v... o oo- .I. . II .V . I IIV. I I .V. .. V ‘ I. V I .II .V. V I . . .I I I. o 9 a I . I . V. . ‘. . V . o . V . ,9. I o . V. V a o w . . . a II. I .I . I . - “V . I. . V V p 1. .I V . I. .I 9 .. I I. . It I o Q I I‘- V .. I n.7,. ‘- V I. . v T. V g l C .V a I V. - 0. on I I . LI .. A ~I v ch..QI- V’IVII. T I I I. .I _ . o . I V ..I .V .n . I I .o...\ I... .. . V I V . I . s . . V I .V. II.‘5’V VI.96 0325 3.2 SSDV .924 He 95 EBA—@825 oBmmoamme fiancee < momcmsueuzfi 3.8-3?— Taeh 5853.3 320 32338:: 68309 5 $326 cosmongfia figofiofl uBmEoEfio£ oEowBosm oat Rom 353533 553535.“ 38.358— Qmssa «SESOV $35828 3535550 flesh DESOV €3.88 £6qu 83856 A25 @3353 nets—5.5:— REFERENCES Bakheit, M.A., D. Torra, L.A. Palomino, O.M.M. Thekisoe, P.A. Mbati, J. Ongerth and P. Karanis, 2008. Sensitive and specific detection of Cryptosporidium species in PCR-negative samples by loop-mediated isothermal DNA amplification and confirmation of generated LAMP products by sequencing. Veterinary Parasitology, 158: 11-22. Bartholomeusz, D.A., R.W. Boutte and JD. Andrade, 2005. Xurography: rapid prototyping of microstructures using a cutting plotter. Journal of Microelectromechanical Systems, 14: 1364-1374. Becker, H. and C. Gartner, 2008. Polymer microfabrication technologies for microfluidic systems. Analytical and Bioanalytica] Chemistry, 390: 89-111. Chen, L.H., J. Yang, J. Yu, Z.J. Ya, L.L. Sun, Y. Shen and Q. Jin, 2005. VFDB: a reference database for bacterial virulence factors. Nucleic Acids Research, 33: D325-D328. Cutler, S.J., A.R. Fooks and W.H. van der Poel, 2010. Public health threat of new, reemerging, and neglected zoonoses in the industrialized world. Emerging Infectious Diseases 16: 1-7. Fang, X.E., Y.Y. Liu, J.L. Kong and X.Y. Jiang, 2010. Loop-mediated isothermal amplification integrated on microfluidic chips for point-of-care quantitative detection of pathogens. Analytical Chemistry, 82: 3002-3006. National Intelligence Council, 2000. The global infectious disease threat and its implications for the United States. Downloaded in July 2010 from http://www.dni.gov/nic/special_globalinfectious.htrnl. Gill, P. and A. Ghaemi, 2008. Nucleic acid isothermal amplification technologies - a review. Nucleosides Nucleotides & Nucleic Acids, 27: 224-243. Kiechle, EL. and C.A. Holland, 2009. Point-of-care testing and molecular diagnostics: miniaturization required. Clinics in Laboratory Medicine, 29: 555. Levy, SB. and B. Marshall, 2004. Antibacterial resistance worlwide: causes, challenges and responses. Nature Medicine 10: 8122-8129. Liang, S.Y., Y.H. Chan, K.T. Hsia, J.L. Lee, M.C. Kuo, K.Y. Hwa, C.W. Chan, T.Y. Chiang, J.S. Chen, F.T. Wu and DD. Ji, 2009. Development of loop-mediated 10 isotherrna] amplification assay for detection of Entamoeba histolytica. Journal of Clinical Microbiology, 47: 1892-1895. Lin, B. and M. Pop, 2009. ARDB - Antibiotic Resistance Genes Database. Nucleic Acids Research, 37: D443-D447. Lutz, 8., P. Weber, M. Focke, B. Faltin, J. Hoffmann, C. Muller, D. Mark, G. Roth, P. Munday, N. Amos, 0. Piepenburg, R. Zengerle and F. von Stetten, 2010. Microfluidic lab-on-a-foil for nucleic acid analysis based on isothermal recombinase polymerase amplification (RPA). Lab on a Chip, 10: 887-893. Morens, D.M., G.K. Folkers and A.S. Fauci, 2004. The challenge of emerging and re- emerging infectious diseases. Nature, 430: 242-249. Mori, Y., K. Nagamine, N. Tomita and T. Notomi, 2001. Detection of loop-mediated isotherrna] amplification reaction by turbidity derived fiom magnesium pyrophosphate formation. Biochemical and Biophysical Research Communications, 289: 150-154. Mori, Y. and T. Notomi, 2009. Loop-mediated isothermal amplification (LAMP): a rapid, accurate, and cost-effective diagnostic method for infectious diseases. Joumal of Infection and Chemotherapy, 15: 62-69. Paul, D., A. Pallandre, S. Miserere, J. Weber and J.L. Viovy, 2007. Lamination-based rapid prototyping of microfluidic devices using flexible thermoplastic substrates. Electrophoresis, 28: 1115-1122. Ramalingam, N., T.C. San, T.J. Kai, M.Y.M. Mak and H.Q. Gong, 2009. Microfluidic devices harboring unsealed reactors for real-time isothermal helicase-dependent amplification. Microfluidics and Nanofluidics, 7: 325-336. Varmus, H., R. Klausner, E. Zerhouni, T. Acharya, A.S. Daar and P.A. Singer, 2003. Grand challenges in global health. Science, 302: 398-399. Weigl, B.H., D.S. Boyle, T. de los Santos, R.B. Peck and M.S. Steele, 2009. Simplicity of use: a critical feature for widespread adoption of diagnostic technologies in low-resource settings. Expert Review of Medical Devices, 6: 461-464. Yager, P., G.J. Domingo and J. Gerdes, 2008. Point-of-care diagnostics for global health. Annual Review of Biomedical Engineering, 10: 107-144. 11 CHAPTER TWO VIRULENCE FACTOR ACTIVITY RELATIONSHIPS (VF AR): CHALLENGES AND DEVELOPIVLENT APPROACHES Tourlousse, Dieter M.; Stedtfeld, Robert D.; Baushke, Samuel W.; Wick, Lukas M.; Hashsham, Syed A. (2007) Water Environment Research 79(3):246-59. INTRODUCTION Virulence factor activity relationships (VF AR) is a predictive approach proposed by the National Research Council's (NRC's) Committee on Drinking Water Contaminants (Washington, DC.) to classify and rank waterborne pathogens (NRC, 1999). Classification and prioritization of chemical and microbial contaminants is necessary to develop a contaminant candidate list (CCL), which serves as the basis for regulation (U .8. EPA, 2005). In essence, VFAR is loosely based on the concept of quantitative structure activity relationships, which have been used for many years to predict health effects of chemical substances and aid in their regulation (Cronin et al., 2003). As such, VFAR is expected to integrate the growing knowledgebase in the area of genomics (and other levels of cellular organization, depending on their availability) of pathogenic microorganisms with clinical information, and combine this with parameters related to the host and environment, to predict potential health risks associated with waterborne microorganisms. Because of the availability of information, it is anticipated that VFAR will first be developed at the genomic level. Later, it will be extended to include information from transcriptomics, proteomics, and phenomics. 12 To begin the process, the NRC Committee suggested that all pathogenic microorganisms may be described in terms of descriptor variables and response or outcome variables. Descriptor variables are those that may prove useful in predicting the response of the microorganisms. Examples of descriptor variables related to the microorganisms include toxins, adhesion and invasion factors, and protein secretion pathways. Response variables are virulence, potency, and persistence. Measures of response variables include minimum infectious dose (or other measures of dose-response relationships), mortality, morbidity, survival time in the environment, and disinfection kinetics. VF AR is then the mathematical model or algorithm that links descriptor variables to response variables in a quantitative and predictive manner. It was suggested that, for the purpose of developing VF AR, the definition of virulence might even be expanded to include clinical virulence (severity of the disease), pathogenicity (the ability to cause disease), and persistence (survival time) of the microorganism in the environment. The committee substantiated the feasibility of the VF AR concept by stating the following: o VFAR is scientifically valid and robustly applicable; 0 VFAR will be able to extend a known relationship between a virulence attribute and human disease to a situation in which the attribute is found in a new or unexpected circumstance or in microorganisms that have not been heretofore recognized as potentially pathogenic; o VFAR will account for the likelihood that adverse human outcomes (i.e., disease) will continue to be discovered in association with the action or 13 presence of a microbial contaminant, microbial gene, or gene product in the clinical setting; 0 VFAR is congruent with the direction that other government, private, and public agencies are taking; and 0 the development of VFAR will satisfy the requirements that development of a CCL be systematic, scientifically sound, transparent, and involve broad public participation. Although the above aspects of VFAR were highlighted to support its feasibility, they may also serve as guidelines during its development. A number of exploratory projects have been initiated by the US. Environmental Protection Agency (Washington, DC.) to identify data availability and gaps for the development of VF AR. Evidently, VFAR needs to capture the widest possible range of molecular and clinical information related to waterborne pathogens, encompassing bacteria, viruses, and protozoa, and synthesize this information to enable a more systematic development of a microbial CCL. It will most likely require extensive new analysis, research, and development. The ultimate goal of VFAR is to develop predictive capabilities for waterborne pathogens by linking response to descriptor variables. Because of the inherent complexity and multifunctional nature of virulence factors, the American Academy of Microbiology (Washington, DC.) report suggested that VFAR, when developed, must be used with considerable caution (Cangelosi et al., 2004). 14 Following the NRC report, the CCL Classification Work group of the National Drinking Water Advisory Council (NDWAC) (Washington, DC.) suggested prioritizing waterborne pathogens by evaluating them for five attributes-potency, severity, prevalence, persistence or mobility, and magnitude (N DWAC, 2004). According to this report, potency, or the amount of contaminant required to cause illness, could be scored between I and 1 1, with the most potent pathogens being those that are known to cause a high rate of morbidity and water-related disease in healthy individuals (score 11), and the least potent being those for whom genetic sequences are available in searchable databases but with no documented virulence genes (score 1). Severity, or the seriousness of the health effects, could be scored with the summation of responses (yes = 1, no - O) to a series of 12 carefully constructed questions based on morbidity and mortality, location and intensity of the infectious processes, extent of infection, time lost to illness, requirement of medical intervention, and associated chronic manifestation or disabilities. The total score (0 to 12) would then represent the severity of the adverse health effects caused by the microorganism. Prevalence, or the frequency of the microbial contaminant occurring in drinking water, could be scored with the most prevalent pathogens being those that are detected in drinking water (given a score of 7), and the least prevalent being those that are not detected in drinking water and have a narrow host range, limited primarily to humans (given a score of 1). Persistence/ mobility describe the potential for amplification under ambient conditions, sedimentation and absorption characteristics, and survival characteristics. It could be scored between 2 to 5, with the highest scoring pathogens being those that are stable for weeks or longer and are able to amplify or are protected from symbiotic relationships, and the lowest scoring being those that typically 15 die rapidly in water (i.e., in days). Magnitude is the contaminant concentration relative to the level that causes a perceived health effect. It could be scored between 0 and 6, with the highest scoring pathogens being those that have caused numerous recently documented waterborne disease outbreaks in the United States or other developed countries, and the lowest scoring being those that have never caused a waterborne disease outbreak anywhere or its biological properties diminish its ability to do so. The NDWAC report suggests selecting pathogens for the CCL using a "prototype classification" algorithm based on the above attributes. This classification scheme represents a significant first step in making the CCL process transparent and systematic and provides guidelines to extract usable and quantifiable data from an extremely complex and dynamic research area. Such classification of pathogenic attributes can also serve as the basis for systematic VFAR development. The VF AR models should enable the prediction and comparison of attributes, such as potency, severity, and persistence, based on comprehensive analysis of descriptor variables. Prevalence and magnitude can then be accessed through sampling and monitoring of water and outbreak samples. The primary step in the VFAR development process is the collection of information related to descriptor and response variables. Unfortunately, this information is scattered in various searchable and non-searchable databases and journal articles, mostly in a form that necessitates further processing. It is also becoming evident that environment and host-related factors are as critical as pathogen-related descriptors in predicting response variables and therefore must be considered among the descriptor variables. A VFAR algorithm must also account for the quality and availability of data, and biological 16 variability. Specifying probability distributions for the input descriptor variables and carrying these uncertainties through to the predicted outcomes can accomplish this. This paper summarizes some of the challenges that must be addressed during the collection, interpretation, and measurement of information related to VFAR, with a focus on genetic descriptors. However, similar care may need to be taken for descriptors at higher levels of cellular organization. This paper also proposes three critical elements of the VFAR development and validation process. The first is the construction of a comprehensive VFAR database housing all information related to descriptor and response variables. The second is the development of mathematical and statistical models to synthesize this information and relate response to descriptor variables. A Bayesian approach is discussed as an example of the type of modeling tools more suitable for VFAR development. The third element is a high-throughput monitoring strategy to identify the prevalence of genetic descriptors of virulence in drinking water supplies and sources. On-chip polymerase chain reaction (PCR) is presented as an economical means to monitor hundreds to thousands of genetic descriptors at a specificity and sensitivity level that is relevant to waterborne pathogens. CHALLENGES IN DEVELOPING VFAR The NRC report defined VFAR in very broad terms, perhaps to allow multiple interpretations and comprehensive development of the concept; this itself poses a challenge. Interpreted broadly, VFAR must synthesize virulence-associated information at all levels of cellular organization, for all classes of waterborne pathogens 17 (encompassing bacteria, protozoa, and viruses), and combine this with relevant host- and environment-related factors. It might even be proposed that such a comprehensive analysis would contribute to our understanding of the emergence of new waterborne pathogens. In narrower interpretations, VFAR should at least be able to prioritize selected waterborne pathogens through comparison, with an initial emphasis on comparative genomics and descriptors of human health. As evident from Figure 2.], this paper employs the former interpretation. It illustrates that a broad set of descriptor variables (associated with the microorganism, host, and environment) determines the response (virulence, potency, and persistence). The predicted response has an associated uncertainty that is the result of biological variability and uncertainties associated with descriptor and response variables. Response variables then serve as the basis for ranking pathogens within the CCL and assigning new pathogens to the CCL for regulatory purposes. 18 :28 :m :85on Be :oumtomm6\m_mofi mg 5 mowmfim hoists, 3:38: :30an 05 :o woman can A00 05 :53, gamete: mo wfixfi: :5 400 05 8 Eomofie: Bo: wfi:w_mm< .8332? 8:88: :5 .835er :m cam—Bet? HomwoBB :5 35880:: me 3:33 8:8 coho 33683 :e 26: 38860.5 30:23? Ea 85368: $0583 moBMES 020:8: “3:8: 8 30:80:35 9:: .76: .Emmqmwcoocomfi 05 3 e832 moist? SE53: 86:8me : .35.: “SE, a we 3:28:53 2: no 533%.. 339230 fin ”mu—DOE \ .. floaofifiaq :oczfibmfi 38>» - 33093th «GOP—«mob .5ng | $5238 .633 Eco—55:35— - hezmzzogz: H 33859.3 E258; - . m :owofimm ~.-.----------.---~ N :owofimm 33% 0:25:— - 528%.: 28:00 - :8 .m 35365: fl HMO: w 82:23.: _ cowofimm 3388:...— 2: o . \ moi—SatogEB DEE—8.5 anal / I 35% 8:852 2853—2 - $93.: $25 - E82»: SEES—cu - $55.? 533353 :0: - 55298 05% EOE—swom - 2: o 2:89? 5:053 £32m - I I 322% :o_mm>=_ ecu 5323.. - coco—E > mEon - 8.82.90 uiunninouinunuu A 2m_z< cm: 35% commando. $8530. ovnovomxx oammom/w _ v _ magma: 3mmmm>< 8382 ocmmmmn?‘ magmas omm— S ommnmm>< 333% F moaamoma: 338% m Samoa: , _ moooc 8:8 comofia no.8 8m «S E E .35 888 I 88. 8 + 88. 8 + 88.8 + e a s 3: $8 I 88.8 + 888 + 88.8 + 8m 2 NM 85 88822:» x 888 I 828 + 88.8 + 888 + Q 8 S 88 «as 8m8 + 88.8 + 88.8 + 88.8 + 8 8 cm 8 «Re 6.8 I 888 + 888 + 88.8 + 2 2 2 88 § 888 I 5.8 + 888 + 88.8 + a a 2 8 8. 5.8 + 88.8 + 88.8 + 8.88 + 8 on 8 8: 825858858 .8 E8 I 88 + 88.8 + 828 + a a a an 88.8 I 88.8 + 88.8 + 888 + x. 2 mm «as :88 I 88.8 + 88.8 + 828 + M: M: M: 88% E8 I 88.8 + 88.8 + 88.8 + 2 2 2 s? x x x x N om cm 88 2:828 .s 58 I 828 + 88.8 + 8W8 + S S 2 88 :5 888 I 888 I 888 I 888 I t t 2 8 :8 3.38.50 3 as: 73 .m_on§m mono 8 850% 2a 338 3% wEbSo 98 an: 38 a 3 83m mm 3608 och .mofiaoouoa 50a 05 was 52 05 886E onmEB Br 98 £05583 5mm 05 was 5mm 2: 38me a .058 E moxon 05 mo $555: 23 .MZm :3on wfimflzofi 3 mEEooom d8 8 Seton 80¢ .cotOm 3m $8 3on 6.2—ow .933 be =a .8.“ A3 DU..\.. 23 $5 film A3 mm Nd EDGE 032 mm_oc-9-_mcm_m cozbmt m>_u_mon_ 8F 2 F 3 we no mo ed r @ov 23:3 u+u 2 8 om 8 cm cm 82 W . \_ _ I ‘I \ O 0 0|.— _ l O K, O O Gilli. L I O . \ O I A T'IO O ., \, G 9|m i To 2 \ ‘37 0‘, .2 OIF H I ~\ ,\ x . w V \ C 0.. . \ E 11)J \ 15% O CO H N L 8 O I O 6 O ‘“‘~~o—\kgj \ \ I ‘.\\‘ 0“ a O 0 HJ 633 mmroaH—oou—o—o-o—u—a—————o—— \\ .\ . \ C 0 $0 0 O O A .0 V 3 Q I I I 74 Success rate of probe design as a function of AG In order to optimize the probe selection criteria, whether probes with positive or negative signals could be identified in silico based on evaluation of their theoretical (thermodynamic) properties was evaluated. Based on previous studies, it was anticipated that probe-target binding affinity (quantified as AG) would be the most valuable parameter for this screening. Only a weak linear relationship (r2 = 0.34) was observed between AG and the natural logarithm of signal intensity (data not shown). This variability in signal intensity for probes with comparable AG precluded evaluation of probe design criteria in terms of signal intensity based on AG. To determine the relationship between AG and probe behavior, hybridization patterns were interpreted in terms of positive/negative signals, irrespective of their strength. For this analysis, all 791 target probes were sorted according to their AG and binned, and the percentage of positive probes in each bin calculated (Figure 3.3). A drastic decrease in the probe design success level was observed for probes with a AG less negative than —1 7 kcal/mol and a %GC less than 34.4 (as estimated using the linear relationship between %GC content and AG). A similar analysis with increasing SNR thresholds revealed analogous trends, but the results were shified toward more negative AG (data not shown). The factors contributing to the unevenness in probe signal for a given target may also explain the lack of hybridization signal for probes with a highly negative AG. In accordance with our observations, Reyes-Lopez et al. (2003) previously reported that the proportion of predicted signals observed experimentally increased for 9-mer probes with increasingly negative AG. 75 G+C content (%) 7o 60 so 40 3o 20 100 — m J 3 9 9° " - 0. cu 80 — .2 ‘ a.» 'g 70 - - Q. E 60 ~ - d) E a) 50 ~ _ 0. 40 J l L l -24 -22 -20 -18 -16 -14 0 AG dupiex (kcal/mole) FIGURE 3.3 Success of probe design as a function of AG. Each symbol indicates the percentage of positive probes for bins of 40 probes, and the error bars indicate the standard deviations for three replicates. The percentage of positive probes is plotted at the median AG of each bin. The %GC was derived from the linear relationship between AG and %GC. Based on this analysis, a AG threshold of —17 kcal/mol could be a valuable threshold for future oligonucleotide probe design. Although this criterion was demonstrated only for 18-mers in this study, a similar approach to assess the efi‘ect of AG on the success level of probe design could be adopted for longer probes. Also, this threshold is in accordance with the findings of Loy et al. (2005), who proposed —16 kcal/mol for 18- mers targeting the 16S rRNA gene. This suggestion was based on the observation that the majority of probes displaying positive signals (including cross-hybridization signals) were attributed to probe-target duplexes with a AG more negative than -16 kcal/mol. It should be noted that small deviations from these AG design thresholds may be observed 76 for hybridizations performed under different experimental conditions and/or for AG estimates obtained using other nearest-neighbor model parameters. Finally, the AG criterion was derived from hybridization patterns with high-abundance and low- complexity target mixtures. For target mixtures having very different compositions in terms of target sequence abundance and diversity, these rules should be applied with caution. Multiplex amplification A split multiplex PCR assay was developed to amplify all VMG amplicons. Of the 720 probes displaying positive signals afier hybridization of monoplex PCR amplicons, 673 (93.5%) yielded positive signals after hybridization of multiplex PCR amplicons (Table 3.1). For 28 of the 47 probe sets, all probes yielded positive signals. In general, the probes that displayed negative signals with the multiplex PCR amplicons also yielded low signals with the monoplex PCR amplicons (data not shown). For the cth gene of V. cholerae, only 2 of 20 probes yielded positive hybridization signals after hybridization of amplicons generated by multiplex PCR. This was attributed to poor amplification in the multiplex PCR rather than to low probe hybridization efficiency. With monoplex PCR, an amplification product of the expected length was observed, verifying that the designed primers successfully amplified the cth gene of V. cholerae ATCC 39315. In addition, all probes targeting the cth gene yielded positive signals after hybridization of the (:th gene amplicon generated by monoplex PCR. The 27 targeted probes (3.9%, excluding probes targeting the cth gene) displaying negative signals were distributed among multiple VMG amplicons, and hence redesign or further optimization of the multiplex 77 PCR assays was expected to be ineffective. In addition, the low levels of the signals of these probes with the monoplex PCR amplicons suggested that their sensitivity may be limited; therefore, these probes were masked during further analysis. Performance of the VMG biochip with spiked water samples The performance of the VMG biochip probe sets was further tested with samples containing pathogen DNA spiked into DNA extracted from three different water samples: tap water, tertiary effluent from a wastewater treatment plant, and river water. For pathogens spiked at a relative abundance level of 0. 1%, the median PFs for 46 targeted VMG amplicons were 0.79, 0.87, and 0.75 for tap water, tertiary effluent, and river water, respectively (Figure 3.4). The median PFs for the 67 other probe sets were 0.05 for tap water, 0.09 for tertiary effluent, and 0.17 for river water (Figure 3.4). By applying a PF threshold of 0.5, the numbers of spiked VMG amplicons assigned as present were 39 for tap water (85%), 40 for tertiary effluent (87%), and 39 for river water (85%). For the nontarget probe sets, the PF was always less than 0.5, and consequently none of the corresponding genes were identified as present in the water samples. Probe sets targeting S. aureus VMG amplicons indicated the presence of this pathogen in all three spiked water samples, although the VMG amplicons assigned as present were in disagreement among samples. The VMG amplicons of C. perfiingens could not be detected in any of the spiked water samples. 78 Unspiked VMGs Spiked VMGs I o o o i l i . . | Drinking water | [ L I limo | | o oo } L 'L . | Tertiary effluent i 00 O O : o o o F—J—J } . | River water 0 0|— l o o co T 0.0 0.2 0.4 0.6 0.8 1 .0 Positive fraction FIGURE 3.4 Performance of the VMG biochip with DNA from multiple water sources supplemented with pathogen DNA. DNA (10 pg) of each pathogen was spiked into 10 ng of DNA from different water sources, yielding a relative abundance for each pathogen of 0.1%. For spiked samples, 673 probes (46 VMG amplicons) were analyzed. For raw samples, 2,034 probes (67 VMG amplicons) were analyzed. The box plots indicate the distribution of the PF for target and non-target probe sets. The dashed line at a PF of 0.5 indicates the selected threshold for presence/absence calls. The boundaries of the boxes indicate the 25th and 75m percentiles, and the whiskers indicate the 10m and 90th percentiles. The median is given as a solid line, the mean is shown as a dashed line, and outlying data points are shown as open symbols. When pathogen DNA was spiked at a relative abundance of 0.01% into DNA extracted fi'om river water, the presence/absence calls for the VMG amplicons varied for a given pathogen (Table 3.1). A similar observation was made previously by Wilson et al. (2002b) and was explained by the variable sensitivity among probes. These researchers observed that for pathogens spiked at a fractional abundance of 0.025% (500 fg of pathogen DNA in 2 ng of DNA fi'om air), only five out of eight probe sets for Francisella tularensis and three out of eight probe sets for Y. pestis were assigned as positive. For 79 pathogen DNA spiked into tertiary effluent DNA at a relative abundance level of 0.001% none of the target genes could be detected (data not shown). Hence, the detection limit of the VMG biochip was 0.1% to 0.01%, depending on the pathogen and VMG. This detection limit is within the range of sensitivity previously reported for microbial diagnostic arrays targeting either the 16S rRNA gene (Maynard et al., 2005; Palmer et al., 2006; Sanguin et al., 2006a) or multiple VMGs (Wilson et al., 2002b). However, advances that enhance the sensitivity are required to achieve reliable detection at abundance levels approaching the minimum infectious dose in environmental samples. These advances include improved experimental techniques for target gene enrichment and/or microarray signal enhancement and optimized data analysis algorithms. The PF threshold applied in this study was less stringent than the cutoff values (at least 0.8) used by previous researchers (Wilson et al., 2002b; DeSantis et al., 2005). Both target gene sequence diversity and abundance aflect the selection of an optimal threshold for the PF. Uncharacterized sequence diversity within the target gene may lead to reduced PFs due to the increased potential for mismatching probe-target duplexes. Low target gene abundance may also yield decreased PF 3 due to variability in probe sensitivity. In both cases, a lower threshold PF needs to be used to reduce false-negative calls. Although the concept of redundant probe sets for enhanced reliability in presence/absence calls has been exploited widely, the effect of target gene abundance on PF 5 is less well documented. As demonstrated in this study and by Wilson et al. (2002b), lower PFs are expected with decreasing target gene amounts, and the extent of the decrease may vary among probe sets. 80 The use of mismatch probes is a common strategy to identify cross-hybridization signals on microarrays and to enhance the reliability of presence/absence calls. In this study, excellent specificity was observed without inclusion of mismatch probes, even for environmental samples. A combination of various factors, including the use of redundant probe sets (with an average of 15 probes per VMG amplicon afier experimental screening), the use of short probes with increased discriminatory power compared to long probes, reduction of sample complexity by target gene enrichment using multiplex PCR, and use of the melting curve approach, contributed to the elimination of false-positive calls. Probe selectivity as a function of AG A selective probe should yield high hybridization signals with target sequences and low signals with nontarget sequences. However, due to the biochemical nature of the probe-target hybridization process, these two criteria cannot be optimized independently. This translates into a trade-off between probe specificity and sensitivity that must be considered. Based on our analysis shown in Figure 3.3, AG was selected as the probe design parameter quantifying the trade-off between probe specificity and sensitivity. Probe sensitivity was derived fi'om the hybridization patterns of the target probes, while probe specificity was quantified based on the hybridization patterns of the nontarget probes. The hybridization results with the raw tap water sample were omitted fiom this analysis due the minimal amount of nontarget probes with positive signals. The hybridization results for all three water samples spiked with pathogen DNA at a relative abundance of 0.1% were included in the analysis. 81 Both target and nontarget probes were sorted according to their AG and binned, and the percentage of positive probes computed for each bin. When the percentage of positive probes was plotted as a function of AG, two distinct regions were apparent (Figure 3.5b), and this effect was independent of the window size used for probe binning (data not shown). In region 1 (AG more negative than —l9.3 kcal/mol), 93.8:i:2.8% of the target probes displayed positive signals, independent of the AG. The percentage of nontarget probes with positive signals was influenced more by AG and increased from 14.7% for probes with a AG of —19.9 kcal/mol to 42.7% for probes with a AG of —23.2 chmol (average increase, 6.8% per AG). In region 2 (AG less negative than —19.3 kcal/mol), the percentage of positive target probes decreased rapidly, and only 31.3% of the probes yielded positive signals for probes with 3 AG of —1 5.6 kcal/mol (average decrease, — 16.1% per AG). The percentage of nontarget probes with positive signal was significantly lower in this region (average, 9.2:t3.9%). As evident from Figure 3.5b, an increase in the percentage of positive target probes was always associated with an increase in the percentage of positive nontarget probes. Probe selectivity, reflecting the ability of a probe to detect intended target sequences and exclude nontarget sequences, quantifies this trade-off and is the parameter that needs to be maximized. Probe selectivity, calculated by determining the difference in the percentages of positive probes for target and nontarget probes as a function of AG, was estimated based on the linear regression lines in Figure 3.5b. The highest selectivity (~80%) was observed for probes with a AG of ——19.3 kcal/mol and a %GC of 47.2 (Figure 3.5a). Thus, probes with a AG of —19.3 kcal/mol provide the best trade-off between 82 sensitivity and specificity. Probes with 3 AG deviating fi'om this optimum displayed lower selectivity. The selectivity was more than 70% for probes with a AG between —18.6 and —21 .1 kcal/mol, which correspond to a %GC content between 42.3 and 56.1. Interestingly, the decrease in selectivity in region 2 was approximately two-fold greater than that in region 1 (—6.6% per AG for region 1 and —14.0% per AG for region 2). It should be noted that for DNA samples containing more complex nontarget sequences, the increase in the percentage of positive nontarget probes (i.e., cross-hybridization) for increasingly negative AG may be more pronounced, while for more abundant target sequences the decrease in the percentage of positive target probes for decreasingly negative AG may be suppressed. 83 (a) G+C content (%) 7o 60 50 40 3o 20 100 __ I I I I 7 I q 2x3 I? .2 4...: U 1’ Q) m (b) Region 1 Region 2 100 ~ ~ I L 80 .b O 20 Percent positive probes AG 0 (kcaI/mole) duplex FIGURE 3.5 Derivation of optimal probe design criteria. (a) Selectivity, expressed as the difference in the percentage of positive probes for target and nontarget probes, as a function of AG. (b) Percentage of positive probes as a function of AG for target and nontarget probes (0 and 0, respectively). Each symbol indicates the percentage of positive probes for bins of 40 target and 80 nontarget probes, with the error bars representing the standard deviation for samples and replicates (n = 6 for target probes and n = 4 for nontarget probes). The %GC was derived from the linear relationship between AG and %GC. 84 A number of studies have demonstrated the effectiveness of AG as a theoretical probe selection parameter for both long and short probes and RNA or DNA targets (Luebke et al., 2003; Matveeva et al., 2003; Reyes-Lopez et al., 2003; Taroncher-Oldenburg et al., 2003; Rhee et al., 2004; Tiquia et al., 2004; He et al., 2005; Loy et al., 2005; Liebich et al., 2006; Antwerpen et al., 2007). In contrast, Pozhitkov et al. (2006) recently suggested that all theoretical (thermodynamics-based) screening of oligonucleotide probes should be omitted due to poor correlations between AG (and AG for intra- and intermolecular self-structures) and experimentally observed signal strengths for rRNA targets. Similarly, only a weak linear correlation was observed between probe signal intensity and AG (data not shown). This indicates that hybridization signals cannot be accurately predicted with current thermodynamics models derived from hybridizations in solution. However, thermodynamic parameters may still be indicative of probe behavior by reflecting the probability that a probe will yield signal intensity higher than a given threshold. Based on the effect of AG on probe selectivity, an optimal AG of -19.3 kcal/mol is suggested for 18-mer oligonucleotide probes under the described hybridization conditions. These probes should yield the highest confidence in presence/absence calls for a given number of probes per target gene. For probes with either more or less negative AG with lower selectivity, more probes are needed to attain high confidence in presence/absence calls. In general, probes with a more negative AG displayed better sensitivity but poorer specificity, which is interestingly analogous to the effect of length. In general, long probes (50- to 70-mers) are more sensitive than short probes (15- to 30-mers) but display lower specificity. 85 Cost and flexibility in microarray synthesis are often recognized as two of the major limiting factors for the adaptation of microarray technology for diagnostic purposes (Hashsham et al., 2004; Call, 2005). In this study, a maskless light-directed in situ microarray synthesis technology developed'at the University of Michigan was employed. This technology employs a digital micromirror device to generate preprogrammed light patterns on the chip surface, triggering deprotection of the 5’-hydroxyl group in conventional phosphoramidite monomers. Synthesis of oligonucleotides using this chemistry provides high fidelity and a stepwise yield and also allows synthesis of probes that are up 100 nucleotides long. In addition, synthesis of new chips is low cost, rapid, and flexible and involves simply uploading a list of probe sequences in the optical unit. Other advantages of in situ probe synthesis include high spot uniformity and probe molecule density. In addition, continuous recycling of the hybridization solution in the microfluidic chips used in this study provides increased signal uniformity within a spot and increased reproducibility of the hybridization signals. The higher cost per chip compared to conventional glass slides is the major limitation of this platform. The ability to rapidly synthesize biochips with updated and reiterated probe sets is considered critical for diagnostic purposes as new gene sequence information is appearing almost daily and this information should be incorporated in the probe selection exercise and data analysis as soon as it becomes available. CONCLUSIONS In summary, we developed and evaluated a coupled format of multiplex PCR and DNA biochip for simultaneous detection of 12 bacterial pathogens in water. Because of 86 the use of redundant probe sets targeting multiple VMGs, false-positives eliminated. Pathogens could be detected at a relative abundance of 0.1 to 0.01%, depending on the pathogen. Analysis of the hybridization patterns also showed that probes with a AG of - 19.3 kcal/mole provided the best trade-off between sensitivity and specificity. In future studies, the VMG biochip will be applied to additional environmental samples, and the presence/absence calls will be verified independently using real-time PCR. 87 REFERENCES Al-Khaldi, S.F., K.M. Myers, A. Rasooly and V. Chizhikov, 2004. Genotyping of Clostridium perfiingens toxins using multiple oligonucleotide microarray hybridization. Molecular and Cellular Probes, 18: 359-367. Antwerpen, M.H., M. Schellhase, E. Ehrentreich-Forster, F. Bier, W. Witte and U. Nubel, 2007. DNA microarray for detection of antibiotic resistance determinants in Bacillus anthracis and closely related Bacillus cereus. Molecular and Cellular Probes, 21: 152-160. Blake, RD. and S.G. Delcourt, 1996. Thermodynamic effects of formamide on DNA stability. Nucleic Acids Research, 24: 2095-2103. Bodrossy, L. and A. Sessitsch, 2004. Oligonucleotide microarrays in microbial diagnostics. Current Opinion in Microbiology, 7: 245-254. Bruun, G.M., R. Wernersson, A.S. Juncker, H. Willenbrock and H.B. Nielsen, 2007. Improving comparability between microarray probe signals by thermodynamic intensity correction. Nucleic Acids Research, 35: 8. Call, D.R., 2005. Challenges and opportunities for pathogen detection using DNA microarrays. Critical Reviews in Microbiology, 31: 91-99. Call, D.R., M.K. Borucki and F.J. Loge, 2003. Detection of bacterial pathogens in environmental samples using DNA microarrays. Journal of Microbiological Methods, 53: 235-243. Call, D.R., F.J. Brockman and D.P. Chandler, 2001. Detecting and genotyping Escherichia coli 0157 : H7 using multiplexed PCR and nucleic acid microarrays. International Journal of Food Microbiology, 67: 71-80. Chandler, D.P., G.J. Newton, J.A. Small and D.S. Daly, 2003. Sequence versus structure for the direct detection of 16S rRNA 'on planar oligonucleotide microarrays. Applied and Environmental Microbiology, 69: 2950-2958. Chizhikov, V., A. Rasooly, K. Chumakov and D.D. Levy, 2001. Microarray analysis of microbial virulence factors. Applied and Environmental Microbiology, 67: 325 8- 3263. Denef, V.J., J. Park, J.L.M. Rodrigues, T.V. Tsoi, S.A. Hashsham and J.M. Tiedje, 2003. Validation of a more sensitive method for using spotted oligonucleotide 88 DNA microarrays for functional genomics studies on bacterial communities. Environmental Microbiology, 5: 933-943. DeSantis, T.Z., C.E. Stone, S.R. Murray, J.P. Moberg and G.L. Andersen, 2005. Rapid quantification and taxonomic classification of environmental DNA from both prokaryotic and eukaryotic origins using a microarray. FEMS Microbiology Letters, 245: 271-278. Dimitrov, RA. and M. Zuker, 2004. Prediction of hybridization and melting for double-stranded nucleic acids. Biophysical Journal, 87: 215-226. El Fantroussi, S., H. Urakawa, A.E. Bernhard, J.J. Kelly, P.A. Noble, H. Smidt, G.M. Yershov and D.A. Stahl, 2003. Direct profiling of environmental microbial populations by thermal dissociation analysis of native rRNAs hybridized to oligonucleotide microarrays. Applied and Environmental Microbiology, 69: 23 77- 2382. Eom, H.S., B.H. Hwang, D.H. Kim, I.B. Lee, Y.H. Kim and HJ. Cha, 2007. Multiple detection of food-borne pathogenic bacteria using a novel l6S rDNA-based oligonucleotide signature chip. Biosensors & Bioelectronics, 22: 845-853. Franke-Whittle, I.H., S.H. Klammer and H. Insam, 2005. Design and application of an oligonucleotide microarray for the investigation of compost microbial communities. Journal of Microbiological Methods, 62: 37-56. Gao, X.L., E. Gulari and X.C. Zhou, 2004. In situ synthesis of oligonucleotide microarrays. Biopolymers, 73: 579-596. Gao, X.L., E. LeProust, H. Zhang, 0. Srivannavit, E. Gulari, P.L. Yu, C. Nishiguchi, Q. Xiang and X.C. Zhou, 2001. A flexible light-directed DNA chip synthesis gated by deprotection using solution photogenerated acids. Nucleic Acids Research, 29: 4744-4750. Gonzalez, S.F., M.J. Krug, M.E. Nielsen, Y. Santos and D.R. Call, 2004. Simultaneous detection of marine fish pathogens by using multiplex PCR and a DNA microarray. Journal of Clinical Microbiology, 42: 1414-1419. Hashsham, S.A., L.M. Wick, J.M. Rouillard, E. Gulari and J.M. Tiedje, 2004. Potential of DNA microarrays for developing parallel detection tools (PDTs) for microorganisms relevant to biodefense and related research needs. Biosensors & Bioelectronics, 20: 668-683. 89 He, Z.L., L.Y. Wu, X.Y. Li, M.W. Fields and J.Z. Zhou, 2005. Empirical establishment of oligonucleotide probe design criteria. Applied and Environmental Microbiology, 71: 3753-3760. Held, G.A., G. Grinstein and Y. Tu, 2003. Modeling of DNA microarray data by using physical properties of hybridization. Proceedings of the National Academy of Sciences of the United States of America, 100: 7575-7580. Keramas, G., D.D. Bang, M. Lund, M. Madsen, S.E. Rasmussen, H. Bunkenborg, P. Telleman and C.B.V. Christensen, 2003. Development of a sensitive DNA microarray suitable for rapid detection of Campylobacter spp. Molecular and Cellular Probes, 17: 187-196. Kostic, T., A. Weilharter, A. Sessitsch and L. Bodrossy, 2005. High-sensitivity, polymerase chain reaction-free detection of microorganisms and their functional genes using 70-mer oligonucleotide diagnostic microarray. Analytical Biochemistry, 346: 333-335. Lane, 8., J. Evermann, F. Loge and D.R. Call, 2004. Amplicon secondary structure prevents target hybridization to oligonucleotide microarrays. Biosensors & Bioelectronics, 20: 728-735. Li, E.S.Y., J.K.K. Ng, J.H. Wu and W.T. Liu, 2004. Evaluating single-base-pair discriminating capability of planar oligonucleotide microchips using a non- equilibrium dissociation approach. Environmental Microbiology, 6: 1197-1202. Liebich, J., C.W. Schadt, S.C. Chong, Z.L. He, S.K. Rhee and J.Z. Zhou, 2006. Improvement of oligonucleotide probe design criteria for functional gene microarrays in environmental applications. Applied and Environmental Microbiology, 72: 1688-1691. Lin, B.C., K.M. Blaney, A.P. Malanoski, A.G. Ligler, J.M. Schnur, D. Metzgar, K.L. Russell and D.A. Stenger, 2007. Using a resequencing microarray as a multiple respiratory pathogen detection assay. Journal of Clinical Microbiology, 45: 443- 452. Lin, B.C., G.J. Vora, D. Thach, E. Walter, D. Metzgar, C. Tibbetts and D.A. Stenger, 2004. Use of oligonucleotide microarrays for rapid detection and serotyping of acute respiratory disease-associated adenoviruses. Journal of Clinical Microbiology, 42: 3232-3239. Liu, W.T., H.L. Guo and J.H. Wu, 2007. Effects of target length on the hybridization efficiency and specificity of rRNA-based oligonucleotide microarrays. Applied and Environmental Microbiology, 73: 73-82. 90 Liu, W.T., A.D. Mirzabekov and D.A. Stahl, 2001. Optimization of an oligonucleotide microchip for microbial identification studies: a non-equilibrium dissociation approach. Environmental Microbiology, 3: 619-629. Loy, A. and L. Bodrossy, 2006. Highly parallel microbial diagnostics using oligonucleotide microarrays. Clinica Chimica Acta, 363: 106-119. Loy, A., K. Kusel, A. Lehner, H.L. Drake and M. Wagner, 2004. Microarray and functional gene analyses of sulfate-reducing prokaryotes in low-sulfate, acidic fens reveal cooccurrence of recognized genera and novel lineages. Applied and Environmental Microbiology, 70: 6998-7009. Loy, A., A. Lehner, N. Lee, J. Adamczyk, H. Meier, J. Ernst, K.H. Schleifer and M. Wagner, 2002. Oligonucleotide microarray for 16S rRNA gene-based detection of all recognized lineages of sulfate-reducing prokaryotes in the environment. Applied and Environmental Microbiology, 68: 5064-5081 . Loy, A., C. Schulz, S. Lucker, A. Schopfer-Wendels, K. Stoecker, C. Baranyi, A. Lehner and M. Wagner, 2005. 16S rRNA gene-based oligonucleotide microarray for environmental monitoring of the betaproteobacterial order "Rhodocyclales". Applied and Environmental Microbiology, 71: 1373-1386. Luebke, K.J., R.P. Balog and H.R. Garner, 2003. Prioritized selection of oligodeoxyribonucleotide probes for efficient hybridization to RNA transcripts. Nucleic Acids Research, 31: 750-758. Markham, N .R. and M. Zuker, 2005. DINAMelt web server for nucleic acid melting prediction. Nucleic Acids Research, 33: W577—W58l. Markoulatos, P., N. Siafakas and M. Moncany, 2002. Multiplex polymerase chain reaction: A practical approach. Journal of Clinical Laboratory Analysis, 16: 47- 51. Matveeva, O.V., S.A. Shabalina, V.A. Nemtsov, A.D. Tsodikov, R.F. Gesteland and J.F. Atkins, 2003. Thermodynamic calculations and statistical correlations for oligo-probes design. Nucleic Acids Research, 31: 4211-4217. Maynard, C., F. Berthiaume, K. Lemarchand, J. Hare], P. Payment, P. Bayardelle, L. Masson and R. Brousseau, 2005. Waterbome pathogen detection by use of oligonucleotide-based microarrays. Applied and Environmental Microbiology, 71: 8548-8557. 91 Palmer, C., E.M. Bik, M.B. Eisen, P.B. Eckburg, T.R. Sana, P.K. Wolber, D.A. Relman and PD. Brown, 2006. Rapid quantitative profiling of complex microbial populations. Nucleic Acids Research, 34. Panicker, G., D.R. Call, M.J. Krug and A.K. Bej, 2004. Detection of pathogenic Vibrio spp. in shellfish by using multiplex PCR and DNA microarrays. Applied and Environmental Microbiology, 70: 7436-7444. Pozhitkov, A., P.A. Noble, T. Domazet-Loso, A.W. Nolte, R. Sonnenberg, P. Staehler, M. Beier and D. Tautz, 2006. Tests of rRNA hybridization to microarrays suggest that hybridization characteristics of oligonucleotide probes for species discrimination cannot be predicted. Nucleic Acids Research, 34. Ratushna, V.G., J.W. Weller and CJ. Gibas, 2005. Secondary structure in the target as a confounding factor in synthetic oligomer microarray design. BMC Genomics, 6. Reyes-Lopez, M., A. Mendez-Tenorio, R. Maldonaldo-Rodriguez, M.J. Doktycz, J.T. Fleming and KL. Beattie, 2003. Fingerprinting of prokaryotic 16S rRNA genes using oligodeoxyribonucleotide microarrays and virtual hybridization. Nucleic Acids Research, 31: 779-789. Rhee, S.K., X.D. Liu, L.Y. Wu, S.C. Chang, X.F. Wan and J.Z. Zhou, 2004. Detection of genes involved in biodegradation and biotransformation in microbial communities by using 50—mer oligonucleotide microarrays. Applied and Environmental Microbiology, 70: 4303-4317. Sanguin, H., A. Herrera, C. Oger-Desfeux, A. Dechesne, P. Simonet, E. Navarro, T.M. Vogel, Y. Moenne-Loecoz, X. Nesme and G.L. Grundmann, 2006a. Development and validation of a prototype 16S rRNA-based taxonomic microarray for Alphaproteobacteria. Environmental Microbiology, 8: 289-307. Sanguin, H., B. Remenant, A. Dechesne, J. Thioulouse, T.M. Vogel, X. Nesme, Y. Moenne-Loccoz and G.L. Grundmann, 2006b. Potential of a 168 rRNA-based taxonomic microarray for analyzing the rhizosphere effects of maize on A grobacterium spp. and bacterial communities. Applied and Environmental Microbiology, 72: 4302-4312. SantaLucia, J., 1998. A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics. Proceedings of the National Academy of Sciences of the United States of America, 95: 1460-1465. Sergeev, N., M. Distler, S. Courtney, S.F. Al-Khaldi, D. Volokhov, V. Chizhikov and A. Rasooly, 2004. Multipathogen oligonucleotide microarray for environmental and biodefense applications. Biosensors & Bioelectronics, 20: 684-698. 92 Siripong, 8., JJ. Kelly, D.A. Stahl and BE. Rittmann, 2006. Impact of prehybridization PCR amplification on microarray detection of niuifying bacteria in wastewater treatment plant samples. Environmental Microbiology, 8: 1564- 1574. Stedtfeld, R.D., L.M. Wick, S.W. Baushke, D.M. Tourlousse, A.D. Herzog, Y.M. Xia, J.M. Rouillard, J.A. Klappenbach, J.R. Cole, E. Gulari, J.M. Tiedje and S.A. Hashsham, 2007. Influence of dangling ends and surface-proximal tails of targets on probe-target duplex formation in 16S rRNA gene-based diagnostic arrays. Applied and Environmental Microbiology, 73: 380-3 89. Taroncher-Oldenburg, G., E.M. Griner, C.A. Francis and BB. Ward, 2003. Oligonucleotide microarray for the study of functional gene diversity in the nitrogen cycle in the environment. Applied and Environmental Microbiology, 69: 1159-1171. Tiquia, S.M., L.Y. Wu, S.C. Chong, S. Passovets, D. Xu, Y. Xu and J.Z. Zhou, 2004. Evaluation of 50-mer oligonucleotide arrays for detecting microbial populations in environmental samples. Biotechniques, 36: 664-+. Tourlousse, D.M., R.D. Stedtfeld, S.W. Baushke, L.M. Wick and S.A. Hashsham, 2007. Virulence factor activity relationships: Challenges and development approaches. Water Environment Research, 79: 246-259. Urakawa, H., P.A. Noble, S. El Fantroussi, J.J. Kelly and D.A. Stahl, 2002. Single- base-pair discrimination of terminal mismatches by using oligonucleotide microarrays and neural network analyses. Applied and Environmental Microbiology, 68: 235-244. Volokhov, D., A. Rasooly, K. Chumakov and V. Chizhikov, 2002. Identification of Listeria species by microarray-based assay. Journal of Clinical Microbiology, 40: 4720-4728. Vora, G.J., C.E. Meador, M.M. Bird, C.A. Bopp, J.D. Andreadis and D.A. Stenger, 2005. Microarray-based detection of genetic heterogeneity, antimicrobial resistance, and the viable but nonculturable state in human pathogenic Vibrio spp. Proceedings of the National Academy of Sciences of the United States of America, 102: 19109-19114. Warsen, A.E., M.J. Krug, S. LaFrentz, D.R. Stanek, F.J. Loge and D.R. Call, 2004. Simultaneous discrimination between 15 fish pathogens by using 168 ribosomal DNA PCR and DNA microarrays. Applied and Environmental Microbiology, 70: 4216-4221. 93 Wick, L.M., J.M. Rouillard, T.S. Whittam, E. Gulari, J.M. Tiedje and S.A. Hashsham, 2006. On-chip non-equilibrium dissociation curves and dissociation rate constants as methods to assess specificity of oligonucleotide probes. Nucleic Acids Research, 34. Wilson, K.H., W.J. Wilson, J.L. Radosevich, T.Z. DeSantis, V.S. Viswanathan, T.A. Kuczmarski and G.L. Andersen, 2002a. High-density microarray of small- subunit ribosomal DNA probes. Applied and Environmental Microbiology, 68: 253 5-2541. Wilson, W.J., C.L. Strout, T.Z. DeSantis, J.L. Stilwell, A.V. Carrano and G.L. Andersen, 2002b. Sequence-specific identification of 18 pathogenic microorganisms using microarray technology. Molecular and Cellular Probes, 16: 1 1 9-127. Wu, L.Y., D.K. Thompson, G.S. Li, R.A. Hurt, J.M. Tiedje and J.Z. Zhou, 2001 . Development and evaluation of functional gene arrays for detection of selected genes in the environment. Applied and Environmental Microbiology, 67: 5780- 5790. Zhang, L., T. Hurek and B. Reinhold-Hurek, 2005. Position of the fluorescent label is a crucial factor determining signal intensity in microarray hybridizations. Nucleic Acids Research, 33. 94 CHAPTER FOUR REAL TIME FLUOROGENIC LOOP-MEDIATED ISOTHERMAL AMPLIFICATION IN POLYMER CHIPS USING AN INEXPENSIVE AND COMPACT MULTIPLEXED OPTICAL SENSOR INTRODUCTION In developed nations, nucleic acid amplification testing (NAAT) is rapidly replacing conventional techniques for diagnosis of infectious diseases due to its superior speed, sensitivity and specificity. Presently, NAAT is however too costly and complex to be employed in developing countries, where the burden of infectious diseases is most severe (El Ekiaby et al., 2010). A pressing need therefore exists for more affordable NAAT devices that can be used in low-resource settings and by minimally trained personnel at point of care (V armus et al., 2003; Weigl et al., 2009). Low cost microfluidics and lab- on-a-chip (Auroux et al., 2004; Zhang et al., 2006; Chen et al., 2007) along with compact and low-cost optical sensors for assay readout (Kuswandi et al., 2007) are the main technologies enabling development of such devices. Devices for NAAT can furthermore be simplified by using isothermal methods for nucleic acid amplification rather than the common polymerase chain reaction (PCR), which requires thermal cycling. Several such techniques have been developed in recent years, including helicase-dependent amplification (Vincent et al., 2004), recombinase polymerase reaction G’iepenburg et al., 2006), and loop-mediated isothermal amplification (Notomi et al., 2000; LAMP). The latter technique in particular (i.e., 95 LAMP) has attracted tremendous attention for simple and robust NAAT (Mori and Notomi, 2009). Aside from its isothermal character, appealing features of LAMP include: excellent specificity due to the use of four to six specific primers and relatively high amplification temperature, need for only a single enzyme, and superior tolerance to substances that typically inhibit PCR (Seyrig et al., 2010a). Another salient attribute of LAMP is the vast amount of DNA generated during amplification (Notomi et al., 2000). This permits direct visualization of positive amplification based on turbidity formed as a result of pyrophosphate precipitation, either visually at the end of the reaction or in real time using a temperature-controlled turbidity sensor (Mori et al., 2004). The high amplification yield also simplifies fluorescence detection due to increased signal strength in the presence of DNA binding dyes, which can be further enhanced using a non- inhibiting dye at high concentration (Seyrig et al., 2010b). The latter is crucial with the prospect of reducing the cost and size of NAAT devices since these are mostly determined by the requirements of the optical module. Today, LAMP is most often performed in conventional reaction tubes. Integration of the technique in microfluidic devices is arguably the next logical step to further enhance its usability and versatility (Mori and Notomi, 2009). Compared to tubes, microfluidic chips employ a much smaller assay volume, which reduces reagent consumption and associated cost, and is also beneficial for multiplexed detection by assay parallelization. Several studies previously reported miniaturization of LAMP in poly(methyl acrylate) microchips (Hataoka et al., 2004) and polyacrylirnade-based micro-chambers (Lam et al., 2008). Toward low-cost and compact instruments, Fang et al. (2010) described a 96 poly(dimethylsiloxane)-glass multiwell chip and integrated turbidity sensor. The latter consisted of high-intensity red light emitting diode (LED), optical fibers and a phototransistor, which is comparable to the sensor developed by Lee et al. ( 2008). Compared to turbidity, fluorescence sensing may be better suited for high throughput chips since the former typically requires a sufficiently long optical path and co-linear positioning of the light source and detector, which may complicate chip design. In this regard, in the studies by Fang et al. (2010) and Lee et al. (2008) monitoring of only a single reaction well was demonstrated. Furthermore, the time to positive detection may also be shorter for fluorogenic LAMP (Aoi et al., 2006). Compact and robust devices for LAMP that allow multiplexed NAAT by parallel sensing of a multitude of reaction wells in a microfluidic format are currently lacking. In this work, we focus on the evaluation of two key components of such an instrument: disposable amplification chips and a temperature-controlled photodiode-based optical sensor. The optical module was drastically simplified by taking advantage of the high fluorescence signal generated using SYTO 81 dye at elevated concentration. To eliminate mechanical movement, which is common in multiplexed photodiode-based sensors (Dishinger and Kennedy, 2008), the optical module consists of an array individually LEDs (one per reaction well) and a single photodiode. To read the different wells, the LEDs are lit in a time-staggered fashion (i.e., lit one at a time) and the signal fi'om the photodiode then assigned to the wells based on the illmnination pattern (Ren et al., 2009). To prevent optical cross-talk, which may be problematic in polymer chips (Irawan et al., 2005), the chips were fabricated as thin film shell microstructures, via a modified hot 97 embossing technique using a moldable counter tool. Using a novel primer set for the diarrhea] pathogen Shigella, the system was demonstrated to perform comparably to an expensive and bulky real time PCR instrument. EXPERIMENTAL SECTION Chip design and fabrication The chips consisted of an array of seven reaction wells with a volume of approximately 2.5 uL each. The wells had a V-shaped layout, consisting of inlet channel for sample dispensing, an outlet channel for air venting, and sensing well. During incubation, the chips were placed vertically in order for air bubbles, when formed, to move away from the sensing region under the influence of gravity. The chips were fabricated as shell microstructures out of 180 um ZeonorFilm® (ZF14-l 80; Zeon Chemicals, Louisville, KY) by hot embossing using a thermoplastic counter tool (acrylonitrile butadiene styrene, ABS from K—mac Plastics). The embossing mold was fabricated via stereolithography (SLA) out of high resolution Somos® NanoToolTM (FineLine Prototyping; Raleigh, NC). The mold was thermally cured and coated with a thin layer of nickel (SLArmorTM) to facilitate de-embossing. A landing shoulder was placed around the length of the mold to prevent excessive lateral flow of the ABS (Yao and Kuduva-Rarnan-Thanumoorthy, 2008), and the mold contained features for eight separate chips to expedite the fabrication process (Figure 4. lb). 98 To emboss the chips, a sheet of ZeonorFilm® was sandwiched between the ABS tool and embossing mold, and preheated to 150°C in a heated press (Carver press model 4386; Carver, Wabash, IN). A pressure of 1000 kg was then applied for 5 min (Figure 4.1a), after which the system was cooled to 105 °C while maintaining the embossing pressure. After further cooling at room temperature for l min, the embossing mold was removed, leaving the chip in the ABS tool, as illustrated schematically in Figure 4. la. The chips were then removed from the ABS tool and cleaned with 1% Liqui-Nox (Alconox), rinsed with copies amounts of distilled water, soaked in isopropanol, and air-dried in a 65 °C oven. To obtain enclosed microchannels, the chips were laminated with high performance optical film (MicroAmp® Optical Adhesive Film; Applied Biosystems; Foster City, CA). Prior to lamination, 1 mm sample dispensing ports and cross-shaped air vents were patterned into the film using a commercial-grade knife plotter (Bartholomeusz et al., 2005; CrafiROBO model CC33OL-20; Graphtec). The patterned film was visually aligned onto the chips, and briefly pressed at 1000 kg (at room temperature) using a piece of silicon rubber (70 A, McMaster-Carr) to obtain a uniform and hermetic seal. To prevent damaging its protruding features, the chip was placed back in the sacrificial ABS tool for bonding. 99 .208 2: mo fimcfl 05 988m 52:05 wficafl £5, 208 @5885 on“ we managed SV .39 How—Boo 03338 a mam: mmoooa wfimmonao we: 05 we 2383an A3 .339 coax—ea SE £5 2: we 959.389 .2— ..8 .5502 fie :30:— coEsonm wfitcmq 383$ QED 3v .03 mm< a SE AGO 100 Temperature-controlled optical module The temperature-controlled optical module is shown schematically in Figure 4.2, and consists of the three main following components: (1) a non-transparent chip holder, (2) an optical sensor, and (3) a heater. The optics is comprised of seven individually addressable green LEDs (RL3-G4518; Super Bright LEDs, St. Louis, MO), seven 1 mm core polymer optical fibers (IF C U1000; Industrial Fiber Optics, Tempe, AZ — razor cut and diamond polished), a blue/ green enhanced photodiode with built-in 500 M9 photovoltaic amplifier (ODA-6WB-500M; OptoDiode Corp., Newbury Park, CA), and a colored glass 590 nm longpass filter (N T54-658; Edmund Optics, Barrington, NJ). The LEDs were placed directly below the reaction wells to improve illumination efficiency in the absence of lenses. Also, the optical fibers were butted against the walls of reaction wells to collect sufficient amounts of light, and oriented at a 90° angle with respect to the LEDs to prevent excessive LED light fi'om entering the fiber. The fibers were butt-coupled to the photodiode and the longpass filter placed between the fiber ends and the photodiode to block the majority of LED light coupled into the fiber. The heater consisted of a Kapton heater (Omega) attached to a thin sheet of aluminum. A T-type thermocouple (Omega) was placed in an additional channel in the chip holder to monitor the temperature of the chip. To assemble the setup, an ABS chip holder was fabricated using an additional SLA mold containing optical fiber alignment channels, following the fabrication technique outline above. Subsequently, 1 mm through holes were drilled through the chip holder below the center of the reaction wells to act as pinholes and prevent LED light fiom 101 reaching adjacent reaction wells. Additionally, 3 mm blind holes were drilled through the bottom of the chip holder in which the LEDs were inserted. The optical fibers were then inserted in the chip holder and fix in place using a small amount of glue. Because the LEDs and optical fibers were integrated as part of the chip holder, the system was highly robust and less prone to misalignment. To couple the fibers to the photodiode, a fixture was fabricated out of black Huntsman RenShape 7820 resin via SLA (Figure 4.2). After placing the chip into the holder, the aluminum top cover with heater was fastened on top using clamps and the entire system covered from ambient light. Electronic controls LabView was used for system control and data acquisition (National Instruments). To this end, all hardware was connected to a personal computer through a Multifunction USB Data Acquisition card (NI USB-6009). To control the temperature of the chip, a feedback mechanism involving (PID) and pulse width modulation (PWM) was implemented. Signals fi'om the thermocouple were amplified and conditioned using a LabView 9211 thermocouple input module. The analog output generated from the PID controller was used to adjust the duty cycle of the PWM driver (DRV102T, Texas Instruments). For all experiments, the temperature of the chip was set to 63°C. 102 .co_oo E coca—among v.8 noumtommmwhmmofi £5 E woman: .onomwoconm 05 8 £03 05 Bob Em: wcmwcom whim Bongo 05 at? 683839 05 main Rama 353825 05 36.508 23 mama 05 mo wetowmmb wocowwfim -08: .3 38 Ba £33 c0382 3:232: .958 Egon—tonne 2: 5 35.8938 «5.3.5.. 2: he eta—5:5 Na. ”my—DUE SHE =3:an when: 3230 c960 con—mom ., \\ /____ 322. a5 All: %oaeofi mam: 5* ., >3 a»; 103 To control the LEDs, a high-speed CMOS logic 4-to-16 line decoder/demultiplexer (CD74HCT4514; Texas Instruments) was used. High logic output fi'om the decoder signaled LED drivers (STLAOl; STMicroelectronics) to provide a constant current of 130 mA to the LEDs. For each measurement, the average of 100 readings collected during a 200 ms time period was used. An analog switch was used to power down the photodiode between measurements as this was observed to improve signal stability. For all experiments, recording of the signals was started when the chip reached 60°C, which took approximately 5 min from room temperature. Loop-mediated isothermal amplification The system was evaluated using a novel primer set for the ipaH gene of the diarrheal pathogen Shigella. The primers were designed using Primer Explorer V4, and Blast search used to verify the specificity of the primers. The sequences of the primers were as follows (written in the 5’-3’ direction): F3: TGGCTGGAAAAACTCAGTGC; B3: TCTGACTITATCCCGGGCAA; FIP; CATGTGAGCGCGACACGGTCGCAGTCTITCGCTGTTGCT; BIP: AGGCATCAGAAGGCCITITCGACCAGAATTTCGAGGCGGAAC; LF: TCACAGCTCTCAGTGGCATCAG; and LB: CTCTCCCTGGGCAGGGAAAT. Reaction mixtures contained 1.6 uM each of F IP and BIP primer, 200 nM each of F3 and B3 primer, 800 mM betaine (Sigma Aldrich), 1.4 mM of each dNTP (Invitrogen), 20 mM Tris-HCl (pH 8.8), 10 mM (NI-I4)2804, 10 mM KCl, 8 mM MgSO4, 8 mM Triton X- 100, 0.64 units/uL of Bst DNA polymerase (New England Biolabs) and 20 uM of SYTO 81 (Invitrogen), unless otherwise stated. Genomic DNA fiom Shigellaflexneri 2a str. 104 245 7T was obtained from the American Type Culture Collection (Manassas, VA) and used for all experiments. Selected experiments were also run in parallel in a commercial real time PCR instrument fiom Bio-Rad Laboratories for comparison. To calculate the tirne-to-positivity (TTP), raw signal intensities were baseline- corrected by subtracting the average signal obtained in the first five minutes. The TTP was then defined as the time at which the baseline-corrected signal reached an arbitrary cut-off of 100 mV. RESULTS AND DISCUSSION The research described herein is part of a project aimed at developing a low-cost, easy-to—use and compact device for LAMP in disposable microfluidic chips. Such devices have widely recognized potential to improve human health globally, by enabling rapid and accurate diagnosis of infectious diseases to be performed in low-resource settings and by minimally trained personnel (Yager et al., 2006). For the purpose of this study, a simple multiwell chip was used to circumvent challenges associated with distribution of a sample across multiple reaction wells, with the goal of evaluating the thin film chips in conjunction with the LED-photodiode based multiplexed optical sensor. Microfluidic chips The amplification chips were fabricated as thin film shell microstructures to prevent optical crosstalk between neighboring reaction wells (discussed below). To fabricate the chips, a moldable counter tool was used rather than a matching hard counter mold to 105 eliminate alignment issues and also reduces mold cost. Also, by placing a landing shoulder around the chip features in the mold, pressure build up was improved by preventing excessive lateral flow of the softened counter tool (Y ao and Kuduva-Raman- Thanumoorthy, 2008). This design was crucial as the glass transition temperature of ABS is substantially lower than that of COP, as a result of which initial optimization of the embossing conditions using molds lacking a landing shoulder proved difficult. Using the modified mold, fabrication was highly robust and not drastically influenced by embossing pressure and/or temperature dwelling time. The here presented technique is comparable to rubber-assisted microforrning, a recently developed method in which thermostable rubber serves as the counter tool (N agarajan and Yao, 2009). A benefit of the latter technique may be that polymers other than COP can also be readily used while for the method used in this study more optimization may be needed to find a suitable moldable counter tool material to ensure physical and chemical compatibility. However, using a moldable counter tool provides the benefit that sharp features can be fabricated that may not be attainable using rubber-assisted microforming. This study also demonstrates the utility of molds fabricated via SLA for hot embossing and rapid prototyping of polymer chips. Compared to other mold fabrication techniques, SLA is much cheaper and also provides the ability to fabricate complex features with few design constraints. For certain applications, the relatively high surface roughness of SLA molds may be problematic, e. g., for assays requiring carefully controlled fluid flow or when assays are performed at elevated temperature due to the propensity of air bubble formation. However, neither was critical in this study and we 106 have extensively utilized SLA molds for rapid and inexpensive prototyping of more complex chips. Optical module The main premise of this study was that simple optics would suffice for real time monitoring of LAMP due to its high amplification yield. Furthermore, our group previously reported that SYTO 81 is only marginally inhibiting at concentrations up to 20 uM (Seyrig et al., 2010b), which drastically enhanced signal strength. As expected, this effect was also observed in the here described chips (Figure 4.3). While a dye concentration of 2 uM was sufficient for detection, a concentration of 20 uM was used for all subsequent experiments, recognizing the small shift in 'ITP. Importantly, the optical module described here is much cheaper and simpler than that typically used for fluorescence sensing in microfluidic chips (Kuswandi et al., 2007). More specifically, a simple colored glass emission filter with an emission spectrum shifted to the right of the maximal emission peak of SYTO 81, along with the orthogonal positioning of the LEDs with respect to the fibers, was sufficient to drop the optical background signal well below the baseline signal prior to amplification. Also, no lenses were necessary since the optical fibers were butt-coupled to the wells and photodiode, which reduces cost and also provides flexibility in terms of positioning of the chip with respect to the photodiode. 107 l" o A E. 1.5 . 0? 8 é: 1.0 ‘ 7'“ 0.5 l a m U) 0.0 . . - 39M 0 5 10 15 20 25 Time (min) FIGURE 4.3 Enhancement of the signal intensity at elevated SYTO 81 concentrations. Compared to previous work in which a similar optical module was used for fluorogenic LAMP with calcein dye (T omita et al., 2008), the use of SYTO 81 presents a significant improvement as the former typically results in delayed amplification compared to SYTO 81 (Seyrig et al., 2010b). Optical cross-talk and reproducibility Optical cross-talk between closely spaced reaction wells can occur in polymer chips due to LED light or emitted fluorescence light propagating through the chip substrate as a result of its waveguiding properties (Irawan et al., 2005). To evaluate whether the shell- structured chip design effectively prevented this, the following experiment was performed. The central well of the chip was filled with water and the four surrounding wells with amplification reagents and DNA. As shown in Figure 4.4, no increase in signal 108 was measurable in the central well, indicating that optical crosstalk did not affect the accuracy of the measurements. 2:) > Neighboring well > Neighboring well > Central well (water) > Neighboring well > Neighboring well > “3+0 l Neighboring wells le-l * % 1e-2 . H '5’.“ ‘i a . x“ l “O 1e-3 l I Central well 1e-4 - \ 1e-5 . T 1 0 5 10 15 20 Time (min) FIGURE 4.4 Lack of optical crosstalk between neighboring reaction wells. The left panel illustrates the experimental design and the right panel shows the amplification plots. Data are fiom a single representative experiment. Images in this thesis/dissertation are presented in color. Evaluation of the intra- and inter-chip variability revealed that amplification and detection was comparable across wells and chips. While the raw signal intensities varied several-fold between different wells within a single chip, probably due to alignment differences in this manually assembled system, normalization of the amplification plots showed that all amplification curves displayed comparable shapes and positions (Figure 4.5). In terms of TIP, the coefficient of variation (CV) for the seven wells in a single chip was roughly 4%, with the average TTP not being significantly different for the three chips. The CV was however, as expected, higher than the CV observed using the commercial real time PCR instrument, which was less than 1%. A smaller CV was also 109 reported using a commercial real time turbidity meter (Mori et al., 2004), but comparison with the data here reported is difficult since the latter measures absorbance and was demonstrated using a different primer and high target gene concentration. The data presented in Figure 4.5 also shows that TTP obtained using the chips and simple sensor may be slightly smaller than that in the commercial real time PCR machine. This observation is interesting as the latter contains a more expensive optical module with better sensitivity, but by optimization of the fluorescent dye comparable results can be generated using much simplified optics. l4 .- 12 T .7: T T g 10 9-4 l—t F4 8 .l 6 Chipl Chip2 Chip3 BioRad FIGURE 4.5 Intra-chip and inter-chip variability. Data represent the mean and standard deviation of the TTP for each chip and that of a commercial real time PCR machine. Quantification To further evaluate the system, a traditional standard curve was generated for genome copy numbers ranging fiom ~10 to 106 copies per channel. Good linearity was observed between the starting copy number and TTP with a correlation coefficient of 0.95 (Figure 110 4.6), indicating that the developed system is well suited for target gene quantification. The limit of detection was ~100 c0pies per reaction wells, which is similar to the sensitivity achieved in conventional tubes using a commercial real time PCR instrument. (a) deltaRn O 0.0 - , A . . 0 5 10 15 20 25 Time (min) R2 = 0.95 TTP (min) 8 :: UIO\\]OO\O 2 3 i 5 Log(copy number) FIGURE 4.6 Target gene quantification. (a) Amplification curves for different target copy numbers and (b) standard curve. The amplification plots are from a single representative experiment and the symbols in the standard curve represent the average and standard error of three replicates. These results also demonstrate that native COP is compatible with LAMP, although a minor level of inhibition may occur at low target copy numbers, as is apparent from the slight non-linearity of the calibration plot at the lowest copy number yielding positive 111 amplification. However, this should not impact quantification potential since other curve fits (e.g., a quadratic) can readily be used to account for this effect (data not shown). This observation is important since surface treatment, as was described in other studies using COP and the related polymer cyclo olefin copolymer chips for isothermal nucleic acid amplification (Furuberg et al., 2008; Lutz et al., 2010), adds an additional level of complexicity and cost to the chip fabrication process. Furthermore, initial evaluation of two different treatment techniques (UV —ozone and photografting) did not improve amplification efficiency at low target copy numbers (data not shown). CONCLUSIONS In summary, two key components of a low-cost and compact device for microfluidic NAAT using LAMP were developed: disposable amplification chips and a multichannel fluorescence sensor. The optics was drastically simplified by taking advantage of the high amplification yield of LAMP and a non-inhibiting DNA-binding (SYTO 81) at elevated concentration. Optical crosstalk was eliminated due to the use of shell-structured chips, which were fabricated using inexpensive molds via hot embossing using a moldable counter tool. Combined with film lamination and xurography, this technique provides an attractive means for rapid and inexpensive prototyping of microfluidic' chips. The developed system is well suited for integration in low-cost and compact devices. The total component cost for the seven-channel sensor was less than $100, with the majority of the cost stemming fi'om the pre-amplified photodiode. Furthermore, due to the low-cost of the LEDs and optical fibers, moderate firrther multiplexing of the sensor 112 would not increase the cost by more than a few dollars. The sensor can also be build very compact due to the use of polymer optical fibers rather than lenses, which provides flexibility in positioning of the photodiode with respect to the chip. In its current setup, the size is mainly determined by the size of the LEDs but using surface Mount LEDs could reduce this. Toward firrther development, the main step is to modify the chips so that a single pipetting step is sufficient for loading of a multitude of reaction channels. In addition, the chip and sensor throughput will be increased to 64, to enable analysis of multiple samples for several genetic markers in a single instrument run. 113 REFERENCES Aoi, Y., M. Hosogai and S. Tsuneda, 2006. Real-time quantitative LAMP (loop- mediated isothermal amplification of DNA) as a simple method for monitoring ammonia-oxidizing bacteria. Journal of Biotechnology, 125: 484-491. Auroux, P.A., Y. Koc, A. deMello, A. Manz and P.J.R. Day, 2004. Miniaturised nucleic acid analysis. Lab on a Chip, 4: 534-546. Bartholomeusz, D.A., R.W. Boutte and JD. Andrade, 2005. Xurography: Rapid prototyping of microstructures using a cutting plotter. Journal of Microelectromechanical Systems, 14: 1364-1374. Chen, L., A. Manz and P.J.R. Day, 2007. Total nucleic acid analysis integrated on microfluidic devices. Lab on a Chip, 7: 1413-1423. Dishinger, J.F. and RT. Kennedy, 2008. Multiplexed detection and applications for separations on parallel microchips. Electrophoresis, 29: 3296-3305. El Ekiaby, M., N. Lelie and J.P. Allain, 2010. Nucleic acid testing (NAT) in high prevalence-low resource settings. Biologicals, 38: 59-64. Fang, X.E., Y.Y. Liu, J.L. Kong and X.Y. Jiang, 2010. Loop-Mediated Isothermal Amplification Integrated on Microfluidic Chips for Point-of-Care Quantitative Detection of Pathogens. Analytical Chemistry, 82: 3002-3006. Furuberg, L., M. Mielnik, A. Gulliksen, L. Solli, LR. Johansen, J. Voite], T. Baier, L. Riegger and F. Karlsen, 2008. RNA amplification chip with parallel microchannels and droplet positioning using capillary valves. Microsystem Technologies-Micro-and Nanosystems-Information Storage and Processing Systems, 14: 673-681. Hataoka, Y., L.H. Zhang, Y. Mori, N. Tomita, T. Notomi and Y. Baba, 2004. Analysis of specific gene by integration of isothermal amplification and electrophoresis on poly(methyl methacrylate) microchips. Analytical Chemistry, 76: 3689-3693. Irawan, R., S. Tjin, P. Yager and D. Zhang, 2005. Cross-talk problem on a fluorescence multi-channel microfluidic chip system. Biomedical Microdevices, 7: 205-211. Kuswandi, B., Nuriman, J. Huskens and W. Verboom, 2007. Optical sensing systems for microfluidic devices: A review. Analytica Chimica Acta, 601: 141-155. 114 Lam, L., S. Sakakihara, K. Ishizuka, S. Takeuchi, H.F. Arata, H. Fujita and H. Noji, 2008. Loop-mediated isothermal amplification of a single DNA molecule in polyacrylarnide gel-based microchamber. Biomedical Microdevices, 10: 539-546. Lee, S.Y., J.G. Huang, T.L. Chuang, J.C. Sheu, Y.K. Chuang, M. Hall, D.R. Meldrum, C.N. Lee and C.W. Lin, 2008. Compact optical diagnostic device for isothermal nucleic acids amplification. Sensors and Actuators B-Chemical, 133: 493-501. Lutz, S., P. Weber, M. Focke, B. Faltin, J. Hoffman, C. Muller, D. Mark, G. Roth, P. Munday, N. Armes, O. Piepenburg, R. Zengerle and F. von Stetten, 2010. Microfluidic lab-on-a-foil for nucleic acid analysis based on isothermal recombinase polymerase amplification (RPA). Lab on a Chip, 10: 887-893. Mori, Y., M. Kitao, N. Tomita and T. Notomi, 2004. Real-time turbidimetry of LAMP reaction for quantifying template DNA. Journal of Biochemical and Biophysical Methods, 59: 145-157. Mori, Y. and T. Notomi, 2009. Loop-mediated isothermal amplification (LAMP): a rapid, accurate, and cost-effective diagnostic method for infectious diseases. Journal of Infection and Chemotherapy, 15: 62-69. Nagarajan, P. and D.G. Yao, 2009. Rubber-assisted micro forming of polymer thin films. Microsystem Technologies-Micro-and Nanosystems-Information Storage and Processing Systems, 15: 251-257. Notomi, T., H. Okayama, H. Masubuchi, T. Yonekawa, K. Watanabe, N. Amino and T. Hase, 2000. Loop-mediated isothermal amplification of DNA. Nucleic Acids Research, 28. Piepenburg, 0., CH. Williams, D.L. Stemple and NA. Armes, 2006. DNA detection using recombination proteins. PLOS Biology, 4: 1115-1121. Ren, K.N., Q.L. Liang, X. Mu, G.A. Luo and Y.M. Wang, 2009. Miniaturized high throughput detection system for capillary array electrophoresis on chip with integrated light emitting diode array as addressed ring-shaped light source. Lab on a Chip, 9: 733-736. Seyrig, G., F. Ahmad, R.D. Stedtfeld, D.M. Tourlousse and S.A. Hashsham, 2010a Simple, powerful, and smart: Using LAMP for low cost screening of multiple waterborne pathogens. In Environmental Microbiology: Current Technology and Water Applications. Caister Academic Press. 115 Seyrig, G., K. Towerey, R.D. Stedtfeld, T.M. Stedtfeld, D. M. Tourlousse, A.M. Cupples, J.M. Tiedje and S. A. Hashsham, 2010b. Evaluation of alternative fluorescent DNA dyes for faster real-time loop-mediated isothermal amplification. In preparation. Tomita, N., Y. Mori, H. Kanda and T. Notomi, 2008. Loop-mediated isothermal amplification (LAMP) of gene sequences and simple visual detection of products. Nature Protocols, 3: 877-882. Varmus, H., R. Klausner, E. Zerhouni, T. Acharya, A.S. Daar and P.A. Singer, 2003. Grand challenges in global health. Science, 302: 398-399. Vincent, M., Y. Xu and H.M. Kong, 2004. Helicase-dependent isothermal DNA amplification. EMBO Reports, 5: 795-800. Weigl, B.H., D.S. Boyle, T. de los Santos, R.B. Peck and M.S. Steele, 2009. Simplicity of use: a critical feature for widespread adoption of diagnostic technologies in low-resource settings. Expert Review of Medical Devices, 6: 461-464. Yager, P., T. Edwards, E. Fu, K. Helton, K. Nelson, M.R. Tam and B.H. Weigl, 2006. Microfluidic diagnostic technologies for global public health. Nature, 442: 412-418. Yao, D.G. and R. Kuduva-Raman-Thanumoorthy, 2008. An enlarged process window for hot embossing. Journal of Micromechanics and Microengineering, 18. Zhang, C.S., J.L. Xu, W.L. Ma and W.L. Zheng, 2006. PCR microfluidic devices for DNA amplification. Biotechnology Advances, 24: 243-284. 116 CHAPTER FIVE A SINIPLE 64-CHANNEL MICROFLUIDIC CHIP FOR LOOP- MEDIATED ISOTHERMAL AMPLIFICATION INTRODUCTION Devices for nucleic acid amplification tests (NAAT) that are capable of screening for multiple pathogens or genetic markers simultaneously are urgently needed to address the global burden of infectious diseases (Yager et al., 2008; Weigl et al., 2009). Multiplexed detection is necessary to accurately identify the etiologic agent(s) responsible for diseases that can be caused by multiple pathogens, such as infectious diarrhea (Antikainen et al., 2009), respiratory infections (Kerdsin et al., 2010), and for antibiotic resistance typing emanating fiom a number of different protein-coding genes (Martineau et al., 2000). Multiplex PCR and microarrays are two main platforms for multiplexed detection; while both are powerful in conventional laboratory settings they are not easily integrated in low-cost and compact devices. This is mainly due to the need for sensitive optics and complex reagent handling equipment. A more promising method for multiplexed detection involves parallelization of the assays in microfluidic chips. A number of assays using polymerase chain reaction (PCR) have been developed by miniaturization of common multiwell plates (Matsubara et al., 2005; Morrison et al., 2006) or using a microfluidic network for sample distribution (Liu et al., 2003). More recently, isothermal techniques nucleic acid amplification have also been implemented in microfluidic chips (Hataoka et al., 2004; F uruberg et al., 2008; Lam 117 et al., 2008; Ramalingam et al., 2009b; Fang et al., 2010; Lutz et al., 2010). In many cases, however, sample dispensing and chip sealing represent the two main challenges that were not adequately addressed. Most chips required either many manual steps or bulky peripheral equipment for loading of the sample and sealing of the chips. Therefore, a need still exists for more user-fiiendly and robust microfluidic chips for multiplexed NAAT. In this work, a polymer microfluidic chip for multiplexed NAAT using loop-mediated isothermal amplification (LAMP) is presented. The chip consists of four distinct arrays of 15+1 reaction channels containing dried primers, allowing parallel analysis of four samples for a multitude of genetic targets. The chip is relatively simple to use as it requires only a single pipetting step to dispense the sample in a distribution channel that routes the sample to the different reaction wells. After dispensing, the sample inlet port of the chip is sealed with tape. Basic development for such array-type chips and sample propagation involving capillary action (F uruberg et al., 2008; Ramalingam et al., 2009a) or centrifirgation (Lutz et al., 2010) are not novel. In our design, pressure generated by depression of the plunger of the pipette is used to induce fluid flow. Another distinguishing feature of our chips is the use of robust pinch-type valves for sealing of the reaction wells. To fabricate these valves, a novel technique involving film lamination and xurography (Bartholomeusz et al., 2005) was developed. Finally, to demonstrate the usability of the chip, three major pathogens responsible for infectious diarrhea (Campylobacterjejuni, Salmonella and Vibrio cholerae) were detected using endpoint fluorogenic LAMP. 118 EXPERIMENTAL SECTION Chip design and operation The chip is roughly the size of credit card and contains four arrays of 15+] reaction wells (Figure 5.1). For each array, the 15 reaction wells are connected to a sample distribution channel (250 in x 250 um) that delivers the sample to different reaction wells from a single inlet port. The other four channels are connected to an additional sample distribution channel and inlet port, with the purpose of accommodating negative assay controls. The volume of sample loaded in the chip is ~30 uL per array, of which roughly half remains in the distribution channel. The chip is loaded using a conventional pipettor, requiring only a single pipetting step per sample. Since capillary action is not sufficient to fill the chips due to the hydrophobic nature of untreated COP and the presence of various features that may act as capillary valves, pressure is provided by depressing the plunger of the pipettor to push the sample through the sample distribution channel. Hence, the inlet port is designed to fit snugly around the pointed end of a 200 uL pipette tip (Figure 5.2b). ll9 Air vents Sample 1 Microvalves Sample 2 Reaction wells l Sample 3 Sample 4 Control FIGURE 5.1 Schematic of the 64-channel microfluidic chip. The chip enables parallel analysis of four samples for up to 15 different gene targets per sample. Images in this thesis/dissertation are presented in color. 120 As the sample moves through the distribution channel, sequential filling of the reaction wells occur within a matter of seconds (Figure 5.2a). In the process, air inside the chip is purged out through air vents placed upstream of each reaction well. Gas- permeable hydrophobic membrane (GVHP 09050, Millipore) covering the air vents prevents liquid from exiting the chip due to the pressure generated with the pipettor. After dispensing of the sample, the inlet ports are easily sealed with tape. The chip is then placed in a dedicated cartridge in which microvalves placed up- and downstream of each reaction well are closed under the action of an array of plungers, as described below. Because closing of the valves leads to fluid movement due to the non-zero dead volume of the valves, it was necessary to allow for backflow of a small portion of the sample without wetting the surface needed for sealing. Therefore, the inlets ports were placed in a cavity lower than the level of the surface for lamination (Figure 5.2b). Chip fabrication For rapid and inexpensive prototyping of the chips, lamination (Paul et al., 2007) and xurography (Bartholomeusz et al., 2005) were used. The latter is a simple prototyping technique that employs computer-controlled cutting plotter to generate structures in thin polymer films. High performance optical film with pressure sensitive adhesive (MicroAmp® Optical Adhesive Film; Applied Biosystems) was used for all layers to prevent inhibition of LAMP. A low-cost commercial grade cutting plotter (CraftROBO Pro model CC33OL-20; Graphtec) equipped with a 60°, 0.9 mm blade and set to a cutting quality of 3 was used for xurography. 121 (a) l"- '~*-- --'* ’2... :‘t ---- J .1" “"3 T m .| fl: 3" .fl‘: s... L' m "- ‘-‘ ‘u‘ “" Lil .. r3“: *. L- ---._ --‘ 1:" 3"?" -' - r-fih—k .24 L- Coda-9v .- ‘ :‘M‘u: t—f‘ ... ‘_ >—--:- “5.3! l - .. ... .. . .-_- "T“: -;.-'_‘ . afim .-~.J. ll ~ - i ll -." ‘5‘": 2“ Lung”? ‘31 ll 1 :t'fifi: “7‘ u cg.-- .I l . l g H g; : 5.. I . l r. . ._w I g i. ‘1. ll t u , :-- -.' J“ \ II .j‘ - J .'l . i . . g ‘9‘ " . . - ' fi - t~ls t~3s Pipette tip Pressure from pipettor \ / Cavity ‘ Empty sample Inlet port distribution channel Sample liquid due to backflow \ Spilled sample liquid FIGURE 5.2 Microfluidic operation of the chip. (a) Time-lapsed images showing filling of the reaction wells as the sample passes through the distribution channel. (b) Schematic of the inlet port design. Images in this thesis/dissertation are presented in color. The chip contains three layers, as illustrated in Figure 5.3: Layer 1 — a cyclo olefin polymer sheet (COP; ZeonorFilm®, Zeon Chemicals) with embossed channel and well features, Layer 2 - a first layer of adhesive optical film with 256 through-holes for routing of the sample through the valve seats, 64 through-holes for air venting and 5 122 through-holes for sample dispensing, and Layer 3 — a second layer of adhesive optical film with 69 through-holes for air venting, 5 through-holes for sample dispensing, and 128 valve features. Layer 1 was fabricated as detailed in Chapter 3 via a modified hot embossing technique using a mold with landing shoulder fabricated by stereolithography and a moldable counter too]. To fabricate Layer 2, an adhesive film is placed on backing paper with the transparent film face-up, and Pattern 1 (see Appendix for the different patterns and process schemes) cut using a blade with its position in the holder adjusted so that the release liner remains intact. After cutting, the unnecessary material is then easily weeded out using transfer tape, leaving through-holes in the transparent film. To fabricate Layer 3, an adhesive film is placed on backing paper with the release liner face-up, and cutting of the necessary patterns performed in two steps. First, Pattern 2a, which generates the through-holes for air venting, is cut using a blade with its position adjusted so that all layers are pierced. Second, Pattern 2b, which generates both the valve seats and alignment guides for the hydrophobic membranes, is cut using a blade with its position adjusted so that it cuts only through the release liner. The unwanted material of the valve features is then removed using transfer tape, taking care not to detach the release liner from the adhesive layer. 123 Sample Hydrophobic membrane $“‘ Valve feature ‘ ‘ «‘5‘ Layer 3 Through-holes Layer 1 FIGURE 5.3 Rendering of the multilayer structure of the microfluidic chip. The dashed red line indicates the flow path of the sample liquid through the different layers during filling of the chip. For clarity, only a portion the chip and the sample flow for a single well are shown. Images in this thesis/dissertation are presented in color. In the next step, the valve seats are formed by submerging the entire adhesive film in isopropanol for 10 to 15 min, which dissolves the adhesive not protected by release liner (Figure 5.4). The film is then cleaned using paper to remove any remaining adhesive; this step is crucial since any left-over adhesive may clog the through-holes in Layer 1. Subsequently, the features cut for alignment of the membranes are removed and the membranes put in place. 124 After releasing the liner from the patterned films, the chip is assembled by visually aligning them on top of Layer 1. The different structures are then firmly bonded together in a hydraulic press, using a piece of rubber to attain a uniform bond. Cutting knife in holder (b) (3) Cutting lines I -: ' III I ' . '--‘ .-:- -:=:.:*-'-t.-'sr-.r.~-":’.I« * -- Release liner Adhesive () Isopropanol FIGURE 5.4 Methodology for fabrication of the valve seats. (a) The position of the blade in its holder is adjusted so that only the release liner is cut when placed face-up on backing paper, as illustrated in (b). The unnecessary material is then removed using transfer tape and the adhesive film placed in isopropanol (c), which dissolves the exposed portion of the adhesive layer ((1). Microfluidic valves A schematic of the structure and working principle of the valves is shown in Figure 5.5a. The valves consist of a discontinued microchannel that routes the sample through the valve seats formed in Layer 2. When a plunger depresses Layer 2, the through-hole in Layer 1 is covered and the valve closed. This is demonstrated in Figure 5.5b, which shows a photograph of three reaction wells with closed valves. Since the plunger is fabricated out of clear plastic, the underlying channel with closed through-hole is visible. 125 Optical film Adhesive (a) I E J ‘* ‘p‘ Cid—lilfiyerg’ n: ' ' Layer2 {— Layer 1 Discontinued microchannel Plunger Rubber sheet (b) To air vents Sealed through-holes in Layer 2 Transparent plungers Reaction wells Rubber sheet From distribution channel FIGURE 5.5 Illustration of the pinch-type microvalves. (a). Schematic of the structure and working principle of the microvalves. (b). Close-up photograph of three reaction wells with closed microvalves. Images in this thesis/dissertation are presented in color. The position of the reaction wells has been enhanced for clarity. Experimental setup The experimental setup for evaluation of the chips is shown schematically in Figure 5.6. It consists of a heated chip holder, a fixture for closing the microvalves, and 126 imaging system for endpoint detection of fluorogenic LAMP. As chip holder, one of the formed pieces of ABS that served as sacrificial counter tool for hot embossing was used. To heat the chip, a strip heater was placed below the chip holder and temperature controlled using proportional—integral—derivative (PID) control and pulse width modulation (PWM), implemented using LabView (National Instruments) as discussed in more detail in Chapter 4. To close the microvalves, a transparent lid with pre-aligned plungers was pressed down onto the chip using two toggle clamps. The imaging system consisted of a portable 50 mW green laser (Laserglow Technologies; Tomato, Canada), an engineered diffuser (Thorlabs, Newton, NJ), an 555 nm bandpass filter (Chroma Technologies, Bellows Falls, VT) and a monochrome CCD camera (Deep Sky Irnager Pro; Meade Instruments Corporation, Irvine, CA). I Camera | Laser Lens Diffiiser 2:: Emission filter Transparent lid with plungers \ Chip holder Toggle clamp Heating strips FIGURE 5.6 Schematic of the experimental setup for evaluation of the chip by endpoint fluorogenic LAMP. 127 Loop-mediated isothermal amplification The chip was evaluated using six LAMP primers sets for three major diarrheal pathogens, C. jejuni, Salmonella and V. cholerae (Table 5.1). The primers were designed using Primer Explorer V4 and Blast search used to verify the specificity of the primers. The amount of primers dried in each reaction well was that needed to yield a final concentration of 1.6 uM each of FIP and BIP primer, 800 nM each of LB and PB primer, and 200 nM each of F3 and B3 primer after dissolution. TABLE 5.1 Primers used in this study. Pathogen Primer sequences (5’-3’) Gene C. jejuni FP: GCAAGACAATA'ITATTGATCGC 61.041 40 RP: GCAAGACAATATTATI‘GATCGC FIP: ACAGCACCGCCACCTATAGTAGAAGCTTI'TTTAAACTAGGGC BIP: AGGCAGCAGAACTTACGCATTGAGTTTGAAAAAACATTCTACCTCT LF: CTAGCTGCTACTACAGAACCAC LB: CATCAAGCTTCACAAGGAAA mapA FP: CTAAAAAT’I‘CTCAATGCAGTTCT RP: ACCGCATTAAAATTCACATC F IP: ACAACATTGAATTCCAACATCGCTATGTGAAAGTCCTGGTGGT BIP: CACTTTAGACACTGGTATTGCTTTGACAAATAACT'ITITCCCTTI‘AGC LF: TGTATAAAAGCCCT'ITAATCTTTGC'ITCA LB: ATGAT‘TI‘AAAAGAAGRGCAAAAGGTT Salmonella FP: CGGCCCGATI'ITCTCTGG inv A RP: CGGCAATAGCGTCACCTT F IP: GCGCAGCATCCGCATCAATAATATGGTATGCCCGGTAAACAG BIP: GAACGGCGAAGCGTACTGGACATCGCACCGTCAAAGGAA LF: CCTTCAAATCGGCATCRATACTCAT LB: AAGGGAAAGCCAGCTFTACG 128 Table 5.1 continued hilA V. cholerae toxR rtxA F P: GGATCAGGTTCAATCCGAG RP: TGTACAATATTATCATTMCCATCGG F IP: TTCGTAATGGTCACCGGCAGAGTCTGCATTACTCTATCGTG BIP: AACTGCCGCAGTGTTAAGGATATCATCTGCCCGGAGAT LF: GCGCATACTGCGATAATCCCTT LB: CTTGAGCTCATGGATCAATTACGCC FP: CGAGTGGAAACGGTTGAAGA RP: AGGGGAAGTAAGACCGCTAT FIP: GCACACTGCTTGAYTCTGCGTACGAAAGCGAAGCTGCTCAT BIP: AGCCACTGTAGTGAACACACCGTCGATTCCCCAAGTT'I‘GGAG LF: ACAGATTCTGGCTGAGAGATGTC LB: CAGCCAGCCAATGTTGTGAC FP: ACCGTACTTGGTCTAACCGT RP: TTGCGTGTAGACATCTTCGG F IP: CCTGTCTCAACGCGTGAACTGATTGGCGATGTGACCT'ITGATG BIP: TTAGTCCGTAAAGGCAAAGTGGGCGATGTGTCCGCTCAATGCG LF: TGTTCGCGGCACCAGCA LB: GATATTACTCTGCAAGGTGCTGG a From Yamazaki et al. (2008) The reaction mixture loaded in the chip contained 800 mM betaine (Sigma Aldrich), 1.4 mM of each dNTP (Invitrogen), 20 mM Tris-HCl (pH 8.8), 10 mM (NH4)2804, 10 mM KCl, 8 mM MgSO4, 8 mM Triton mm, 0.64 units/uL of Bst DNA polymerase (New England Biolabs) and 20 uM of SYTO 81 (Invitrogen). Genomic DNA fi'om C. jejuni strain NCTC 11168, Salmonella enterica strain LT2, and V. cholerae strain 01 biovar eltor str. N16961 was obtained from the American Type Culture Collection (Manassas, VA). 129 RESULTS AND DISCUSSION Chip design and fabrication The microfluidic chip is capable of analyzing four samples for up to 15 different genetic markers per sample. The chip is well suited for adoptation in low-resource settings as only a simple pipettor is needed for sample dispensing and sealing of the chips merely requires covering the sample inlet port with tape. Evidently, it is recognized that sample processing also needs to be addressed to provide a system with sample-in answer- out capabilities. Propagation of the sample through the distribution channel relies only upon the pressure generated by the pipettor. Compared to capillary-driven flow, which is a common technique for low-cost microfluidics, this provides the benefit that design of the chip is simplified. This is because capillary-driven flow requires careful analysis of the flow behavior in expanding features that may be present in the chip (Zhang et al., 2005). Others have relied upon centrifugation for sample distribution in microfluidic chips for NAAT (Lutz et al., 2010), which requires additional instrumentation. The valves implemented in the chip are “normally-open” valves, which is important fi'om the perspective of the user. It allows the user to directly load the sample, after which closing of the valves is accomplished in a dedicated chip cartridge. Furthermore, due to the shell- structure of the chips, alignment is much simplified, which is important considering the need for correct positioning of the plungers with respect to the chip. 130 In terms of fabrication, the developed method for locally removing adhesive is, to our knowledge, novel. This technique is highly versatile since a cutting plotter is used to define the features that need to be removed. A drawback of the use of optical film is its limited flexibility, as a result of which substantial pressure is necessary to activate the valves. Using a polydimethylsiloxane layer may be more appealing in this respect, but bonding of the latter to other polymeric substrates is challenging (Tang and Lee, 2010), which would complicate the fabrication process. Optical isolation of assay chambers in the microfluidic chip To evaluate the efficacy of the valves, it was evaluated whether fluidic cross-talk occurred between adjacent reaction wells. To this end, two different experiments were performed. In the first experiment, primers were dried in alternating reaction channels and the chip loaded with LAMP reaction mixture containing target DNA (Figure 5.7). In the second experiment, primers along with target DNA was dried down in alternating reaction channels, with the other channels only containing primer (Figure 5.7) and the chip loaded with solely reaction mixture. 131 Sample with DNA (40 pg/pL) Sample without DNA No primer . Primer plus DNA aner Primer FIGURE 5.7 Fluidic isolation of the reaction wells. The content dried in the reaction wells is shown for the lower two wells, but this pattern applies to the entire chip. Images were taken after 40 min of amplification in the presence of 20 M SYTO 81 dye. Both experiments clearly demonstrated that no DNA (primers or amplicons) migrated between adjacent reaction channels, as no fluorescence was visible in the wells lacking primer in the first experiment and the wells lacking DNA in the second experiment. This suggests that the valves effectively isolated the reaction wells and each reaction well could be considered as an independent assay not influenced by the neighboring wells. 132 Sample commsition V. cholerae V. cholerae Salmonella Salmonella C. jejuni C. jejuni ASSEW layout Amplification control (n=3) V. cholerae (toxR) Salmonella (invA) Salmonella (hilA) C. jejuni (cj0414) C. jejuni (mapA) FIGURE 5.8 Multiplexed detection using the microfluidic chips. The samples contained 2 pg/uL of DNA from the indicated pathogens, which amounts to roughly 500 genomic copies per reaction well. The position of the different primer sets is indicated. Images were taken after 40 min of amplification in the presence of 20 uM SYTO 81 dye. Furthermore, inspection of the chips after amplification verified that amplification predominantly occurred in the region enclosed by the valve pairs, demonstrating that the valves are effective in isolating the reaction wells. Sealing of the reaction wells also prevented formation of air bubbles, which always occurred in unsealed reaction wells due to the surface roughness of the molds. Demonstration of parallel detection of multiple pathogens To illustrate the utility of the chips for multiplexed pathogen detection, six different LAMP primer sets for three major diarrheal pathogens (C. jejuni, Salmonella and V. cholerae) were dispensed in duplicate in the chips along with three positive amplification controls (Figure 5.8). Three binary genomic DNA mixtures were then loaded in the chip 133 along with reaction mixture and subjected to LAMP for 40 min, after which the fluorescence was imaged. For all pathogen mixtures, the pattern of positive/negative wells corresponded to those expected, which indicates that the chips performed as expected (Figure 5.8). It also suggests that the newly designed primers are specific for the genomic DNA tested, but additional testing is necessary with other bacteria and stool samples. CONCLUSIONS In summary, a robust and user-friendly polymer microfluidic chip for multiplexed LAMP was developed. The chip contains four arrays of 15+] interconnected reaction channels, allowing analysis of multiple samples for several genetic targets. To load the sample and seal the chip, only a common pipettor and tape is required. This makes the chip very attractive for adoptation in resource-limited settings. Effective fluidic isolation of the reaction wells was accomplished using pinch-type microvalve pairs, which offer the possibility to further increase the density of the reaction wells since fluidic cross-talk between closely spaced wells is not a concern. In terms of fabrication, a novel method for monolithic integration of hundreds of pinch-type valves in a few simple step using a cutting plotter and lamination was presented. Toward further development, the chip will be combined with a modified version of the optical module described in Chapter 4 to provide a low-cost and compact device for LAMP in disposable microfluidic chips with pre-dispensed and stabilized reagents. After 134 validation of the primer sets with field samples, this device could be used for detection of bacteria responsible for the majority of infectious agents responsible for diarrhea. 135 APPENDIX Through-holes for sample routing (600 pm) mough'hOICS fin n o o , . . 3 . a o o o a o forairventing coo 0. out» no on. o. no. : (750m) :2: I: 2:: I: :2: I: :1: I. I 0 O O O Through-holes for sample dispensing (1 mm) Plot Pattern 1 Weed out unwanted material Film with through-holes FIGURE Al. Schematic of Pattern 1 and process scheme. 136 I\ r~ I -‘\ /\ \ I“. '\ '\ /\ \-/‘ . . . O (.1 \I'.) ., \, )(.‘ (\ \,-‘\ <1 (,’\I /\ ,-_/ \ /\,/. ;‘ I\ -\ ,‘ 1‘ l‘ 9 C l O ( r \.—' ) .; \. ’\/ (\- \/'\4 (.J \z"’ t‘\ /- /\ '\ r‘./ I‘ r .‘\ /\ \./‘ a o a o ‘ . \-‘ ’ -, \,/\. (t \ “k '» / (,'\I —\ .. /\ \ r~.r~ /- -\.,'\ I\ ,~ /~ . o . o (,, (’1‘ l ‘, \, l\. (\. \\\\\ C .‘ \’)\; ’\ /~ /\ -\ x-./- w /\ '\ \ ‘-" o o a o ( . \.' 1 , \,'\x (\ \, '~I («s C,'\' ‘\ z- 1 ~ 1' /- -‘ . ‘I\ ,\ -- 1‘ . . . a (./ \. l c )\. (\1 a K; ( l \,)\I ‘ ,~ /\ , I- - .~‘I\ r, /~ . . . a (-1 \. I ( \ )\.' (\l t \I (,/ \,)\I \ r- /\ I /‘~ -\ .-x I, A I ( . ( ) C ' C ) \1 \. \ \. \J \l .1 \- " O 0 I 0 (,‘ I.‘ 1‘ (I I ),.. (,‘ ~‘(\ (" A)r\ -1 \. I \ \/ \/ \ I -2 \. ‘1 O 0 O O ( \ (__l ,’ . ,’ (,, ,H.‘ ,‘ .— ,\ O . . 0 -1 \. ~ \, \. \/ \ ~I (.1 \.)\/ ~‘ [\‘/‘ / '\)/\ ,\ 1‘ )'o\ " _-, IN 9 o . ° (‘1 \.' t- \l \n (‘1 .- s! (.1 \,)‘t ( \ ,'. 1‘ ,r‘ ,\‘/. ( ‘\ \“I\ ,-~ -\ 1‘ . ° .1 \. . ‘ / x \/ \I -."\z (,1 \.)\I . O <-‘ ,2 ’3 l" '- ‘/~ (fl ,x‘;‘ " ’__ I‘ t/ \. . J \ xx \I ‘-'\2 (,1 \.’\/ . 9 o o ( , Q /\ ,- ‘lx ( -\ \j'\ (,.‘ ,-_),, ‘1 . . 1 x \/ \I ., \I ‘1 \, \/ o o . ° (_‘ (\ /\ I ,' \ ( ‘\ (x "4‘ ,3 [\‘/r ‘1 . * x J \l ‘\l s‘, \_, r, o o o 0 ( ‘ ,» /\ ( /- ("\ ’\'\ 1‘. /\ " ‘1 x. z \s \I a 'K1 k . \./ \_ I o 0 ° // Valve features Through-holes for air venting (750 pm) Alignment gurde for membranes coo.» <30: (Kb _. Q® «we: a C): <39 "D— Plot Pattern 2a Plot Pattern 2b Weed. out “11W ted material Soak film in Film with adhesive isoopropanol locally removed (valve seats) /—/=- M Weed out unwanted Film with alignment guide Place membrane material for membrane Film with valve seats exposed FIGURE A2. Schematics of Pattern 2a and 2b and process scheme. Images in this thesis/dissertation are presented in color. 137 REFERENCES Antikainen, J., E. Tarkka, K. Haukka, A. Siitonen, M. Vaara and J. Kirveskari, 2009. New 16-plex PCR method for rapid detection of diarrheagenic Escherichia coli directly from stool samples. European Journal of Clinical Microbiology & Infectious Diseases, 28: 899-908. Bartholomeusz, D.A., R.W. Boutte and JD. Andrade, 2005. Xurography: Rapid prototyping of microstructures using a cutting plotter. Journal of Microelectromechanical Systems, 14: 1364-1374. Fang, X.E., Y.Y. Liu, J.L. Kong and X.Y. Jiang, 2010. Loop-Mediated Isothermal Amplification Integrated on Microfluidic Chips for Point-of-Care Quantitative Detection of Pathogens. Analytical Chemistry, 82: 3002-3006. Furuberg, L., M. Mielnik, A. Gulliksen, L. Solli, LR. Johansen, J. Voite], T. Baier, L. Riegger and F. Karlsen, 2008. RNA amplification chip with parallel microchannels and droplet positioning using capillary valves. Microsystem Technologies-Micro-and Nanosystems—Information Storage and Processing Systems, 14: 673-681. Hataoka, Y., L.H. Zhang, Y. Mori, N. Tomita, T. Notomi and Y. Baba, 2004. Analysis of specific gene by integration of isothermal amplification and electrophoresis on poly(methyl methacrylate) microchips. Analytical Chemistry, 76: 3689-3693. Kerdsin, A., R. Uchida, C. Verathamjamrus, P. Puangpatra, K. Kawakami, P. Puntanaku], S. Lochindarat, T. Bunnag, P. Sawanpanyalert, S. Dejsirilert and K. Oishi, 2010. Development of Triplex SYBR Green Real-Time PCR for Detecting Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella spp. without Extraction of DNA. Japanese Journal of Infectious Diseases, 63: 1 73-1 80. Lam, L., S. Sakakihara, K. Ishizuka, S. Takeuchi, H.F. Arata, H. Fujita and H. Noji, 2008. Loop-mediated isothermal amplification of a single DNA molecule in polyacrylarnide gel—based microchamber. Biomedical Microdevices, 10: 539-546. Liu, J., C. Hansen and SR. Quake, 2003. Solving the "world-to-chip" interface problem with a microfluidic matrix. Analytical Chemistry, 75: 4718-4723. Lutz, S., P. Weber, M. Focke, B. Faltin, J. Hoffmann, C. Muller, D. Mark, G. Roth, P. Munday, N. Armes, O. Piepenburg, R. Zengerle and F. von Stetten, 2010. Microfluidic lab-on-a-foil for nucleic acid analysis based on isothermal recombinase polymerase amplification (RPA). Lab on a Chip, 10: 887-893. 138 Martineau, F., F.J. Picard, L. Grenier, P.H. Roy, M. Ouellette, M.G. Bergeron and E. Trial, 2000. Multiplex PCR assays for the detection of clinically relevant antibiotic resistance genes in staphylococci isolated from patients infected after cardiac surgery. Journal of Antimicrobial Chemotherapy, 46: 527-533. Matsubara, Y., K. Kerman, M. Kobayashi, S. Yamamura, Y. Morita and E. Tamiya, 2005. Microchamber array based DNA quantification and specific sequence detection from a single copy via PCR in nanoliter volumes. Biosensors & Bioelectronics, 20: 1482-1490. Morrison, T., J. Hurley, J. Garcia, K. Yoder, A. Katz, D. Roberts, J. Cho, T. Kanigan, S.E. Ilyin, D. Horowitz, J.M. Dixon and C.J.H. Brenan, 2006. Nanoliter high throughput quantitative PCR. Nucleic Acids Research, 34. Paul, D., A. Pallandre, S. Miserere, J. Weber and J.L. Viovy, 2007. Lamination-based rapid prototyping of microfluidic devices using flexible thermoplastic substrates. Electrophoresis, 28: 1 115-1 122. Ramalingam, N., H.B. Liu, C.C. Dai, Y. Jiang, H. Wang, Q.H. Wang, K.M. Hui and H.Q. Gong, 2009a. Real-time PCR array chip with capillary-driven sample loading and reactor sealing for point-of-care applications. Biomedical Microdevices, 11: 1007-1020. Ramalingam, N., T.C. San, T.J. Kai, M.Y.M. Mak and H.Q. Gong, 2009b. Microfluidic devices harboring unsealed reactors for real-time isothermal helicase-dependent amplification. Microfluidics and Nanofluidics, 7: 325-3 36. Tang, L. and N.Y. Lee, 2010. A facile route for irreversible bonding of plastic-PDMS hybrid microdevices at room temperature. Lab on a Chip, 10: 1274-1280. Weigl, B.H., D.S. Boyle, T. de los Santos, R.D. Peck and M.S. Steele, 2009. Simplicity of use: a critical feature for widespread adoption of diagnostic technologies in low-resource settings. Expert Review of Medical Devices, 6: 461-464. Yager, P., G.J. Domingo and J. Gerdes, 2008. Point-of-care diagnostics for global health. Annual Review of Biomedical Engineering, 10: 107-144. Yamazaki, W., M. Taguchi, M. Ishibashi, M. Kitazato, M. Nukina, N. Misawa and K. Inoue, 2008. Development and evaluation of a loop-mediated isothermal amplification assay for rapid and simple detection of Campylobacterjejuni and Campylobacter coli. Journal of Medical Microbiology, 57: 444-451. 139 Zhang, L., T. Hurek and B. Reinhold-Hurek, 2005. Position of the fluorescent label is a crucial factor determining signal intensity in nricroarray hybridizations. Nucleic Acids Research, 33. 140 CHAPTER SIX MULTIPLEXED DETECTION OF DIARRHEAL BACTERIA USING LOOP-MEDIATED ISOTHERMAL AMPLIFICATION IN A LOW-COST AND PORTABLE MICROFLUIDIC DEVICE INTRODUCTION Infectious diarrheal disease is the second leading cause of mortality and morbidity worldwide (Kosek et al., 2003; Black et al., 2010). More than 1.3 million children die each year due to diarrhea, the vast majority in developing nations (Black et al., 2010). Rapid and accurate identification of the responsible agents is crucial for effective management of the disease and also to prevent its spread in the community. However, the only available method in impoverished countries involves culturing, which takes 3-5 days and also requires considerable expertise. Because infectious diarrhea can be caused by several pathogens, including protozoa, viruses and bacteria, multiplex detection techniques are necessary for accurate diagnosis. Several such methods based on nucleic acid amplification tests (NAAT) using multiplex PCR (Wang, 2008; Antikainen et al., 2009; O'Leary et al., 2009) and rrricroarrays (You et al., 2008; Suo et al., 2010) have been developed, and some of these are commercially available, such as the EntericBio multiplex PCR system fiom Serosep Ltd. (Limerick, Ireland). Because these assays require bulky and expensive instrumentation they are, however, not typically available in developing nations. Therefore, a need exists for more affordable and easy-to-use devices for NAAT for diagnosis of diarrheal pathogens to help 141 health care providers prescribe more effective treatments and also provide valuable surveillance and epidemiological data. The goal of this study was to apply a low-cost and compact device for loop-mediated isothermal amplification (LAMP) for rapid and accurate detection of six major diarrheal bacteria: Campylobacterjejuni, Salmonella, enterotogenic Escherichia coli (ETEC), Vibrio cholerae, and Yersinia enterocolitica. The instrument integrates the two key components of this device developed in previous work integrated in a fully packaged device: a 64-channel polymer microfluidic chip (described in Chapter 5) and a LED/photodiode-based optical sensor (described in Chapter 4). The pathogens targeted in this study are from the high priority list prepared by Bill and Melinda Gates Foundation. One of the aims of the Foundation is ‘to develop and deliver innovative tools and approaches to help prevent, diagnose, and treat enteric and diarrhea] diseases around the world.’ The approach presented here addresses this need by providing a low-cost, battery- operated and easy-to-use device for multiplexed detection of diarrhea] pathogens using NAAT. DEVICE INTEGRATION Toward integration of the device in a fully packaged and functional prototype, several tasks had to be completed. First, the electronics for control and data acquisition needed to be transferred to a custom designed printed circuit board (PCB), with all software being contained in a single microcontroller. Second, a housing needed to be designed to contain the chip holder with embedded heater, optical components (LEDs, optical fibers and 142 photodiode) and PCB. Third, a Wi-Fi data communication protocol needed to be established between the microcontroller and iPod Touch. Due to the breath of these tasks, a concerted effort from several students and collaborators was necessary, and the prototype presented here is based on this combined endeavor. Embedded control and data acquisition Embedding of the system involved design of a custom PCB and programming of a rrricrocontroller. As described in the Ph.D. dissertation of my colleague Dr. Robert Stedtfeld, at the core of the instrument is a 32-bit ARM 7 microcontroller (LPC2378, NXP Semiconductors) for temperature control, triggering of the LEDs, acquiring timed signal readings from the photodiode and sensing the data to a Wi-Fi module (Mini- Microprocessor Wi-Fi Core Module Rabbit Module), which then sends the data to the iPod Touch (Figure 6.1). To control the chip temperature, the microcontroller receives temperature readings from a thermocouple placed on the chip holder and adjusts the voltage applied to the heater using proportional—integral—derivative (PID) control and pulse width modulation (PWM). The microcontroller also trigger the LEDs in a predetermined pattern using four de-multiplexers connected to 64 surface mount 50 mA LED drivers. Finally, it also records the photodiode readings and sends it to a Wi-Fi module through a serial communication port. The Wi-Fi module then automatically routes the data in real time to the iPod Touch for plotting and storage. To start the instrument, the microcontroller receives a string from the iPod Touch that contains information about the requested assay time and temperature. 143 Photodiode LEDs I :ver '--- 4to 16 line Relay (I Duvet . — _ .- . decoder <1: Driver .. .. _ - . l :Analog Input Digital Output Battery ; ARM microcontroller :Analog Output 'Analog Input + t H ' ' Strip heater and WiFi module , PWM DC D C —+—'——\ I thermocouple converter FIGURE 6.1 Functional block diagram of the modules of the embedded instrument Kindly provided by Robert D. Stedtfeld. Images in this thesis/dissertation are presented in color. GeneZ software A specific application (GeneZ) was developed for the iPod Touch by Scott Price, another team members involved in this project. The software allows the user to set the reaction temperature and time (Figure 6.2), and start the assay. During the instrument run, data readings from the photodiode are continuously sent from the instrument to the iPod Touch, where the data is sorted and plotted for each reaction well. At the end of the assay, the software automatically calculates the tirne-to-positivity (TTP) for each reaction well, calculates averages for replicated assays and estimates the target gene copy number using a predefined calibration plot. 144 lost 11) lost 11) Test Type Tcmociaxure Durations ‘ Temperature Durations Deneurre Denature Number of Cycles Annealtng Extension Results Number of Cycles FIGURE 6.2 Screen shots GeneZ application for the iPod Touch. Images in this thesis/dissertation are presented in color. Housing design and assembly A first prototype of the instrument was completed as part of this study (Figure 6.33). The housing measures 15 X 10 X 5 cm and is fabricated via stereolithography out of black Huntsman RenShape 7820 resin (F ineLine Prototyping; Raleigh, NC). The housing consists of two parts that can be separated to facilitate assembly of the system. The key functional component of the instrument is the chip holder with integrated heater, LEDs and optical fibers. Compared to the system evaluated in Chapter 4, the layout of the optics needed to be modified to permit sensing of a higher density chip (Figure 6.4). More specifically, rather than coupling the fibers to the wells at a 90° angle, the fibers needed to transect the chip holder and be oriented at a 135° angle. To improve 145 the coupling efficiency between the fibers to the reaction wells, one of the sides of the wells has a draft angle of 45°. Also, due to space constraints, a 750 um fiber was used rather than the 1 mm fiber used in Chapter 4. To assemble the system, a chip holder was fabricated out of black acrylonitrile butadiene styrene (ABS) by hot embossing with the mold used for fabrication of the chips. Flexible electrical resistance heating tape strip heater (TEMPCO Electric Heater Corporation; Wood Dale, IL) was embedded in the holder during embossing to achieve effective heating of the chip holder (Figure 6.4). Using this type of heating element rather than a more conventional thin film heater was necessary for the LEDs and optical fibers to be inserted in the holder, and proved highly effective and versatile. After fabrication of the holder with embedded heaters, holes were manually drilled for the fibers and LEDs. The latter were mounted on a custom-designed PCB board, as shown in Figure 6.3b, with spacers to raise the LEDs from the board. Prior to assembly of the housing, the fibers were glued in place and their ends bundled and coupled to the photodiode using a simple fixture. The chip holder was then placed on the PCB board with LEDs, with the fibers being routed in the ‘wings’ in the housing. This was arguably the most challenging step during assembly of the prototype due the limited space between the bottom of the chip holder and PCB. 146 (a) (b) PllOIOLllOLIC ll “ - .3 . ,. -. . ()pllC'dl lIlICl'S ' ‘ I -'>(‘lll])l1ulrlcl‘ ,2 l lcatcr lends mncouplc FIGURE 6.3 Photographs of the developed instrument. (a) The device along with the microfluidic chip and iPod Touch. (b) The disassembled prototype showing the various components in the device. Images in this thesis/dissertation are presented in color. 147 . - > Strip heater / / / // / Optical fibers LEDs FIGURE 6.4 Rendering of the chip holder with embedded heater, LEDs and optical fibers. The inset shows the positioning of the LEDs and optical fibers with respect to the wells. Images in this thesis/dissertation are presented in color. Microfiuidic chips The microfluidics chip layout, operating principle and the method for fabrication are described in detail in Chapter 4, and only a brief summary is provided below. The chips are the size of a credit card and contain four arrays of 15+1 reaction wells (1 pL per well), enabling parallel analysis of four samples for up to 15 pathogens or genetic markers. Each anay of 15 reaction wells is connected to a sample inlet port in which the user dispenses a sample using a common pipettor. The distribution channel routes the sample to the different reaction wells in which primers were pre-dispensed and dried during chip fabrication. After dispensing, the chip is sealed and placed in the instrument. An additional layer of PCR tape is then placed on the chip to prevent evaporation through the hydrophobic membranes that serve as air vents. Upon closing of the lid, an array of 148 pre-aligned plungers closes microvalves that are placed up and down-stream of each reaction well. PRIMER SETS AND LAMP CONDITIONS Fourteen different LAMP primers for established genetic markers of each pathogen were designed using PrimerExplorer4 or retrieved from the literature (Table 6.1). For each pathogen, reference genomic DNA was obtained fiom the American Type Culture Collection (Manassas, VA). The LAMP primers were dispensed in the chips prior to assembly of the different layers in an amount to yield a final concentration of 1.6 uM each of F IP and BIP primer, 800 nM each of LF and LB primer, 200 nM each of F3 and B3 primer after dissolution in the reaction well. The LAMP mixture loaded in the chips consisted of 800 mM betaine (Sigma Aldrich), 1.4 mM of each dNTP (Invitrogen), 20 mM Tris-HCl (pH 8.8), 10 mM (NI-I4)2804, 10 mM KC], 8 mM MgSO4, 8 mM Triton X- 100, 0.64 units/uL of Bst polymerase (New England Biolabs), 20 uM of SYTO 81 dye (Invitrogen) and target DNA (1 uL per 25 uL LAMP mixture). 149 HO085. 5:. SE 598 E5 E5 885 “.3 Ho 30 @053 5~m=§~0h A585 :3 a “0:50 00:55:00 5.0 03.? 151 H<©<<