FIELD DEPLOYABLE LOW COST MICROFLUIDIC SYSTEM FOR REAL-TIME
ISOTHERMAL DETECTION OF DRUG RESISTANT TUBERCULOSIS
by
Gregoire Francois Henri Seyrig

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Environmental Engineering
2012

ABSTRACT
FIELD DEPLOYABLE LOW COST MICROFLUIDIC SYSTEM FOR REAL-TIME
ISOTHERMAL DETECTION OF DRUG RESISTANT TUBERCULOSIS
by
Gregoire Francois Henri Seyrig

Tuberculosis (TB) caused by Mycobacterium tuberculosis is a global health issue that is
responsible for more than 1.6 million deaths annually; majority of which occur in developing
nations. In addition to routine TB, often inappropriate use of antibacterial drugs results in multidrug resistant (MDR) and extensively drug-resistant (XDR) TB strains, which are 500-600 times
more expensive to treat. Totally drug-resistant TB is also known. Developing nations often lack
the resources to diagnose MDR/XDR TB in time resulting in delayed treatment and further
infections. Diagnosis of MDR/XDR TB is challenging because it requires detection of many
single nucleotide polymorphisms (SNPs) imparting the resistance. Current strategies for the
diagnosis of MDR/XDR TB require expensive instruments costing ~$90,000 for equipment (e.g.,
®

GeneXpert ), $20-$50 per test, and well-trained technicians. None of these resources are
adequate in developing nations. Our group has developed an inexpensive genetic analysis
TM

platform, named Gene-Z

at a manufacturing cost of $500, which has the potential to carry out

real time isothermal amplification assays. It employs fluorescence-based loop-mediated
isothermal amplification (LAMP) in microfluidic chips for detection and quantification of
nucleic acids markers using an iPod Touch

TM

as control and visualization software. When fully

developed and validated, this platform could become a low-cost and simpler alternative to
®

TM

GeneXpert . Gene-Z

also has the ability to be operated using batteries and over time with

solar charger, making it the instrument of choice for limited resource settings. Using this
platform, the goal of this doctoral work was to develop a simple, sensitive, and inexpensive field
deployable microfluidic chip for detecting TB and demonstrate its potential for MDR TB using a
model SNP imparting resistance to isoniazid, the resistance that is currently measured using
®

GeneXpert

system. The following four objectives were set. (i) Reduce the detection time and

enhance the sensitivity for routine TB. (ii) Stabilize the molecular biology reagents to be stable
under field conditions for long periods of time, (iii) Develop a simplified low-cost microfluidic
chip design with sample in answer out capabilities, and (iv)Develop an isothermal approach for
detection of a model SNP to be used for detection of MDR TB using LAMP.
Based on optimization studies carried out using 11 fluorescent DNA-intercalating dyes
and Bst polymerase, a protocol was developed to detect 10 copies of M. tuberculosis genomic
DNA in less than 15 min. Using trehalose, a protocol was established to stabilize molecular
biology reagents for their storage at room temperature for up to one year with no or minimal loss
of activity. A polyester-based chip that did not contain any valves and allowed multiplex
detection without cross contamination between reaction chambers was also developed. An initial
proof-of-concept was demonstrated using an SNP associated with isoniazid resistance. Work
with patient sputum was outside the scope of this research because of the requirements
associated with a BSL3 facility. Overall, this work provides critical data necessary to detect TB,
MDR/XDR-TB using a simple platform suitable for limited resource settings.

Copyright by
GREGOIRE FRANCOIS HENRI SEYRIG
2012

This dissertation is dedicated to my family for their continuous support, including: my Father,
Mother, Sisters, my cousin Fabian Seyrig de Saussure, and my Godfather Olivier Favre, and the
woman of my life Krista C. Bennick for her unconditional love

v

ACKNOWLEDGMENTS

I will be the only one to be granted this degree; however this would have never been
possible without the help of many peoples. Krista has been my main source of inspiration and
this even before I met her, because she gave me the force of continuing this great journey even in
the most difficult moments. My parents and sisters have been of a great support during all these
years, they always believed in me and backed me up when I needed it the most. I am also very
grateful to my godfather Olivier and to my cousin Fabian for their precious support.
TM

Gene-Z

was created as part of a group within which every actor has played a critical

role to make this project possible. Professor Syed Hashsham recruited me to pursue my Ph.D. as
part of this fascinating project back in 2007. I am extremely grateful to him not only for hiring
me but also for being a great mentor and an amazing source of inspiration. I am also thankful to
Dr. Alison Cupples for letting me use her real-time thermocycler so many times, and to the other
members of my committee, Dr. James M. Tiedje, and Dr. Irene Xagoraraki, for all their precious
advices and for taking the time to evaluate my work. I would also like to warmly thank Drs.
Robert Stedtfeld and Dieter Tourlousse for their continuous help and collaboration, and all the
present and former members of Dr. Hashsham’s laboratory, particularly Tiffany, Maggie, Keara,
Amanda, Farhan and Vikram for their precious help, and working together. I also want to thank
Lori Larner, our administrative (and so much more) assistant for all her priceless help during
these years. In addition, it was a great honor to work for this degree in the Department of Civil
and Environmental Engineering, at Michigan State University, within the beautiful state of
Michigan, and with the support of wonderful people living here.

vi

Finally I would like to thank all my friends for believing in me and always been in my
back, even during the most difficult times, these include Dr. Matthieu Boyreau, Marie-Joe Bussy,
Pete & Katie Bennick, Antwaun Kent, Nikki Motson, Jamie Sherman, Nicolas & Amy
Pommereau, Dr Gerard Beck, “Junior”, and so many others.

vii

TABLE OF CONTENTS

LIST OF TABLES......................................................................................................................... xi!
LIST OF FIGURES ...................................................................................................................... xii!
LIST OF ABBREVIATIONS...................................................................................................... xiv!
CHAPTER 1. INTRODUCTION ................................................................................................... 1!
1.1. Genetic Markers for Wild-type TB and MDR/XDR-TB......................................................... 2!
1.2. Potential of LAMP for Detection of TB, MDR-TB, and XDR-TB ......................................... 5!
1.3. The Gene-ZTM Platform ......................................................................................................... 10!
1.4. Goal of this Study .................................................................................................................. 11!
1.4.1. Rationale........................................................................................................................... 11!
1.4.2. Specific aims .................................................................................................................... 12!
1.5. Developmental Team ............................................................................................................. 14!
1.6. Organization of this Dissertation ........................................................................................... 15!
CHAPTER 2. LOOP-MEDIATED ISOTHERMAL AMPLIFICATION AS A LOW COST AND
RUGGED TECHNIQUE FOR SCREENING OF MULTIPLE GENETIC MARKERS ............ 17!
2.1. Introduction............................................................................................................................ 17!
2.2. Isothermal DNA Amplification Techniques.......................................................................... 21!
2.3. LAMP Assay: A Simplified Description ............................................................................... 23!
2.3.1. Location of LAMP primers on the target strand .............................................................. 23!
2.3.2. Key mechanisms involved in LAMP reaction.................................................................. 26!
2.3.3. Parameters considered in designing LAMP primers ........................................................ 30!
2.3.4. Carrying out the LAMP reaction...................................................................................... 31!
2.3.5. Visualizing the amplification product .............................................................................. 32!
2.4. Performance and Validation Characteristics of LAMP ......................................................... 35!
2.4.1. Detection limit, time to positivity, and specificity in LAMP assays................................ 35!
2.4.2. The effect of inhibitors in LAMP assay ........................................................................... 37!
2.4.3. Ongoing improvements in LAMP assay .......................................................................... 42!
2.5. Development Needs for Waterborne Pathogens Screening and Future Directions of LAMP43!
2.5.1. On-chip sample processing............................................................................................... 43!
2.5.2. Miniaturization of LAMP for multiplexing, genotyping, and mutation detection ........... 47!
2.5.3. Lowering the cost of real-time imaging for high throughput screening........................... 48!
2.5.4. Validation studies with waterborne pathogens................................................................. 50!
CHAPTER 3. OPTIMIZATION OF REAL-TIME LOOP-MEDIATED ISOTHERMAL
AMPLIFICATION FOR FLUORESCENCE DETECTION ....................................................... 51!
viii

3.1. Introduction............................................................................................................................ 51!
3.2. Materials and Methods........................................................................................................... 53!
3.2.1. DNA targets...................................................................................................................... 53!
3.2.2. LAMP primers.................................................................................................................. 53!
3.2.3. Fluorescence-based LAMP .............................................................................................. 54!
3.2.4. Fluorescence-based LAMP in compact device ................................................................ 55!
3.3. Results and Discussion .......................................................................................................... 56!
3.3.1. Effect of dye type and concentration on threshold time................................................... 56!
3.3.2. Fluorescence intensity and enhancement during LAMP.................................................. 62!
3.3.3. Effect of Bst polymerase concentration on threshold time............................................... 65!
3.3.4. Evaluation of SYTO-82-based real-time LAMP for the detection of other pathogens.... 66!
3.3.5. Real-time LAMP with compact and inexpensive optics .................................................. 69!
3.4. Conclusions............................................................................................................................ 70!
CHAPTER 4. FREEZE-DRYING OF REAL-TIME PCR AND LAMP REAGENTS FOR
FIELD DEPLOYABLE MICROFLUIDIC CHIPS...................................................................... 72!
4.1. Introduction............................................................................................................................ 72!
4.2. Materials and Methods........................................................................................................... 73!
4.2.1. DNA targets...................................................................................................................... 73!
4.2.2. PCR primers ..................................................................................................................... 74!
4.2.3. LAMP primers.................................................................................................................. 74!
4.2.4. Real-time PCR.................................................................................................................. 75!
4.2.5. Real-time LAMP .............................................................................................................. 75!
4.2.6. Silicon microchips fabrication.......................................................................................... 76!
4.2.7. Real-time PCR in silicon microchips ............................................................................... 77!
4.2.8. Fabrication of COP chips ................................................................................................. 77!
4.2.9. Real-time LAMP in COP microchips............................................................................... 78!
4.2.10. Stabilization and conservation of real-time PCR reagents. ............................................ 79!
4.2.11. Stabilization of real-time LAMP reagents...................................................................... 79!
4.3. Results and Discussion .......................................................................................................... 80!
4.3.1. Effect of trehalose concentration...................................................................................... 80!
4.3.2. Effect of storage time ....................................................................................................... 83!
4.3.3. Effect of primers............................................................................................................... 85!
4.3.4. Freeze-drying PCR mixes in microfluidic chip................................................................ 86!
4.3.5. Freeze-drying real-time LAMP reagents in COP microfluidic chips............................... 89!
4.4. Conclusions............................................................................................................................ 91!
CHAPTER 5. DEVELOPMENT OF LOW-COST MICROFLUIDIC CHIPS FOR THE GENEZTM PLATFORM.......................................................................................................................... 93!
5.1. Introduction............................................................................................................................ 93!
5.2. Materials and Methods........................................................................................................... 94!
5.2.1. Chip design and fabrication.............................................................................................. 94!
5.2.2. Real-time LAMP .............................................................................................................. 98!
5.3. Results and Discussion .......................................................................................................... 98!
5.3.1. Chip design and fabrication.............................................................................................. 98!
5.3.2. Evaluation of possible cross-contamination................................................................... 100!
5.3.3. Reproducibility ............................................................................................................... 101!
ix

5.3.4. Quantification and sensitivity......................................................................................... 101!
5.3.5. Multiplex detection......................................................................................................... 105!
5.4. Conclusions.......................................................................................................................... 106!
CHAPTER 6. TOWARDS A FIELD-DEPLOYABLE AND INEXPENSIVE CHIP FOR SNP
DETECTION AND FUTURE PERPSECTIVES....................................................................... 107!
6.1. Introduction.......................................................................................................................... 107!
6.2. Materials and Methods......................................................................................................... 108!
6.2.1. DNA targets.................................................................................................................... 108!
6.2.2. LAMP primers................................................................................................................ 109!
6.2.3. Taq mutS-based real-time LAMP .................................................................................. 109!
6.3. Results and Discussion ........................................................................................................ 111!
6.4. Conclusions and perspectives .............................................................................................. 114!
APPENDICES ............................................................................................................................ 118!
APPENDIX 1.............................................................................................................................. 118!
APPENDIX 2.............................................................................................................................. 118!
APPENDIX 3.............................................................................................................................. 118
APPENDIX 5.............................................................................................................................. 118!
APPENDIX 6.............................................................................................................................. 118!
APPENDIX 7.............................................................................................................................. 118!
APPENDIX 8.............................................................................................................................. 118!
APPENDIX 9.............................................................................................................................. 118
APPENDIX 11............................................................................................................................ 118!
REFERENCES ........................................................................................................................... 147!

x

LIST OF TABLES

Table 1.1. Sensitivity and specificity of nucleic acid amplification-based tests commercially
available for clinical specimens.. .................................................................................................... 3!
Table 1.2. Genetic mutation associated with drug resistance in TB. ............................................ 5!
Table 1.3. Existing LAMP assays for tuberculosis. ...................................................................... 9!
Table 2.1. Overview of the etiology of waterborne disease outbreaks in the United States
between 1991 and 2002 ................................................................................................................ 19!
Table 2.2. Performances of LAMP for the detection of bacteria, parasites and viruses. ............. 39!
Table 3.1. Summary of the results obtained with the dyes used in this study.............................. 59!
Table 4.1. Effect of freeze-drying on PCR amplification efficiency ........................................... 87!
Table 6.1. Primers used in this study ......................................................................................... 109
Table A1.1. List of primers used in CHAPTER 3... .................................................................. 118
Table A5.1. Performances of existing fluorescence-based real-time LAMP assays compared with
results obtained using SYTO-82................................................................................................. 128
Table A8.1. List of primers used in CHAPTER 5 ..................................................................... 134
Table A1.1. Overview of the mutations in MDR-TB and XDR-TB.......................................... 144
!

xi

LIST OF FIGURES
Figure 1.1. (A) Replication of amplification among the 64 wells present on the polyester chip
TM
TM
using Gene-Z . (B) Digital photography of the Gene-Z
platform. ........................................ 8!
Figure 1.2. Developmental team and roles of various members.................................................. 15!
Figure 2.1. (A) The LAMP primer positions. For ease of explanation only the forward primers
and their position on the antisense strand of the template DNA are shown but the backward
primers act similarly on the sense strand. (B) Loop formation from a strand elongated from BIP.
The F1c tail is hybridized with the F1 region. .............................................................................. 25!
Figure 2.2. LAMP simplified: formation of the dumbbell shape structure and later structure.... 27!
Figure 2.3. Schematic representation of the LAMP products accumulation. .............................. 29!
Figure 2.4. Principle of LAMP detection using calcein............................................................... 33!
Figure 2.5. Real-time and direct detection of LAMP amplification product targeting G.
intestinalis genomic DNA............................................................................................................. 34!
Figure 2.6. (A) Images of 16 well COC chip with dilution series (5 ng, 0.5 ng, 0.05 ng, 0.005 ng,
and 0 ng) of DNA (stained with 1! Pico Green) with increasing exposure time. ........................ 46!
Figure 2.7. Signal intensity versus exposure time for the dilution series of DNA (5 ng, 0.5 ng,
and 0.05 ng) stained with Pico Green. .......................................................................................... 49!
Figure 3.1. Effect of orange fluorescent dyes on the Tt of real-time LAMP............................... 58!
Figure 3.2. Effect of green fluorescent dyes on the Tt of real-time LAMP................................. 61!
Figure 3.3. Comparison of fluorescence intensity (A) and SNR amplification profile (B) of dyes
of interest ...................................................................................................................................... 64!
Figure 3.4. Detection of several pathogens by real-time LAMP using SYTO-82....................... 67!
TM

Figure 3.5. Detection of M. tuberculosis with Gene-Z

device. .............................................. 70!

Figure 4.1. Microfluidic chips used in this study. Silica 3-well chip (A), and COP 8-well (B,
Note: The channels contain “sky blue spectrum” liquid food colorant (Ateco, Glen Cove, NY)
for better visualization). ................................................................................................................ 78!
Figure 4.2. Effect of trehalose concentration on freeze-dried PCR and LAMP reagents............ 82!

xii

Figure 4.3. Effect of storage time on freeze-dried (filled shapes) and non-freeze-dried (empty
shapes, reagents were stored at room temperature) real-time PCR mixes for various spiked target
(C. jejuni) copy numbers............................................................................................................... 84!
Figure 4.4. Amplification profiles of freeze-dried PCR and real-time LAMP mixes in
microfluidic chips. ........................................................................................................................ 90!
Figure 5.1. Schematic of the 64-well microfluidic chip. ............................................................. 95!
Figure 5.2. Schematic of the polyester chip embossing using a hydraulic press......................... 97!
Figure 5.3. Schematic of the chip lamination and filling............................................................. 99!
Figure 5.4. Fluorescence image of a single 15-well array after amplification to demonstrate lack
of cross-contamination between reaction wells in the absence of valves................................... 100!
Figure 5.5. Evaluation of inter- and intra-chip reproducibility.................................................. 103!
Figure 5.6. Evaluation of multiplex LAMP in the polyester microfluidic chip......................... 105!
Figure 6.1. Position of the FIP and BIP primers for the detection of SNP by LAMP............... 110!
Figure 6.2. LAMP amplification profile of the wild type primers for katG. ............................. 111
Figure A2.1. Comparison of fluorescence intensity (A) and SNR amplification profiles (B) of all
dyes. ............................................................................................................................................ 111!
Figure A3.1. Effect of Bst concentration on Tt of real-time LAMP using SYTO-81 or SYTO-82.
Target: 0.1 ng of M. tuberculosis genomic DNA. ...................................................................... 111!
Figure A4.1. Effect of the concentration of SYTO-81 on the LAMP monitoring in
the Gene-Z.. ................................................................................................................................ 111!
Figure A6.1. Qualitative comparison of the current results with existing results of fluorescencebased real-time LAMP................................................................................................................ 111!
Figure A7.1. Cutting patterns for the xurography...................................................................... 111!
Figure A9.1. Use of the hydraulic press..................................................................................... 111!
Figure A9.2. Pattern for membrane cutting (Membrane.GSD). ................................................ 111
Figure A9.3. Pattern for film cutting (Film.GSD). .................................................................... 111
Figure A9.4. Alignment of the membrane and the adhesive on the membrane......................... 111
Figure A10.1 Mutated target sequence (5’ to 3’) that was incorporated in a
pIDTSMART:ampicilin:blunt plasmid...................................................................................... 111
xiii

LIST OF ABBREVIATIONS

Abs/Em: absorption/emission
ATCC: American Type Culture Collection
BIP: backward inner primer
CCD: charged-couple device
CNC: computer numerical control
COC: cyclo olefin copolymer
COP: cyclo olefin polymer
CV: coefficient of variation
DAQ: data acquisition
ddH2O: double-distilled/double deionized water
DEP: dielectrophoresis
DEPC: diethylpyrocarbonate-treated
DNA: deoxyribonucleic acid
dNTP: deoxyribonucleotide
dsDNA: double-stranded deoxyribonucleic acid
ECN: estimated copy number
FDA: United States Food and Drug Administration
FIP: forward inner primer
HDA: helicase-dependant amplification
IMS: immunomagnetic separation
xiv

LAMP: loop-mediated isothermal amplification
LED: light-emitting diode
MDR: multi drug-resistant
min: minute
NASBA: nucleic acid sequence based amplification
NCBI: National Center for Biotechnology Information
PCR: polymerase chain reaction
PID: proportional-integral-derivative algorithm
PMB: paramagnetic beads
POC: point of care
PR: photo-resist
RCA: rolling circle amplification
Ref: reference
RNA: ribonucleic acid
RPA: replicase polymerase amplification
RT-PCR: reverse-transcript polymerase chain reaction
SDA: strand-displacement amplification
SMAP: SMart Amplification
SNP: single-nucleotide polymorphism
SNR: signal-to-noise ratio
TB: tuberculosis
TM: trademark
Tt: threshold time

xv

UV: ultra-violet
WHO: World Health Organization
Wi-Fi: Wireless Fidelity
XDR: extensive drug-resistant

xvi

CHAPTER 1. INTRODUCTION

Mycobacterium tuberculosis (TB) is a global health issue and multi-drugresistant/extensive drug-resistant TB (MDR/XDR-TB) is an emerging infectious disease. TB is a
leading killer despite its discovery by Koch more than 100 years ago. According to World Health
Organization (WHO), tuberculosis kills 1.68 million people every year (including 0.38 million
HIV positive patients), with 14 million cases (1). The same report also indicated that only 63%
of the TB cases are detected. This low detection rate has implications in the spread of the disease
as undiagnosed and untreated cases keep on feeding the epidemic (2). In 2006, it was estimated
that 625,000 lives could be saved per year if a 100% accurate single-visit test for TB was
available. It was also suggested that even if such a test for TB was only 85% sensitive and 97%
specific, it could still save 392,000 lives per year (3).
Inappropriate use of antibiotics is one of the major causes leading to the emergence of TB
strains resistant to first line of antibiotics (i.e., the drugs that are most commonly used to treat
TB) and sometimes second line of antibiotics (i.e., antibiotics that are used when the first line of
antibiotics do not work). TB strains that are resistant to two or more first line of antibiotics are
named MDR-TB. MDR-TB strains that are also resistant to second line of antibiotics are named
XDR-TB. According to the WHO, drug-resistant TB affects more than 500,000 people. At
present, rapid field tests that allow the detection of MDR/XDR-TB are scarce. One genetic of the
available assays for MDR-TB costs between $20-$50 based on the subsidy available
®

(GeneXpert

MTB/RIF, Cepheid, Sunnyvale, CA). Others use hybridization based approaches
®

for SNP detection that are equally expensive (e.g., InnoLIPA and Line Probe Assay). An ideal
detection approach for MDR/XDR-TB requires identifying multiple (10 to 50) single nucleotide
1

polymorphisms (SNPs) located over a number of genes and responsible for resistance to various
drugs. Detecting SNPs requires extensive primer design and performance validation. The success
rate for primer design targeting SNPs is rarely better than 5%. Therefore, development of a large
panel of SNPs is technically challenging and resource intensive. Detecting a large panel of SNPs
using field deployable point of care (POC) devices is currently not possible. It is one of the longterm goals of our group. As a first step, the focus of this research is on the SNP causing
isoniazid-resistance, which is one of the two resistances (i.e., with rifampin resistance) present in
MDR-TB.

1.1. Genetic Markers for Wild-type TB and MDR/XDR-TB
The gold standard for the detection of TB and drug-resistant TB is based on culture
techniques, which provide results in weeks to months causing delays in providing appropriate
treatment (2). Tests based on molecular amplification offer the advantage of rapid testing
because they can provide actionable results within hours. Genetic assays are based on genetic
markers specific to a given species or resistance genes of interest and correlation with drug
resistance is presumed but not fully demonstrated. However, several genetic markers are known
to perform well for the detection of wild-type TB and others are in use for MDR-TB. The 16S
rRNA gene and the insertion sequence IS6110, for example, are used by commercial tests for the
®

detection of TB, including the FDA-cleared Gen-Probe

AMPLIFIED Mycobacterium

tuberculosis Direct Test (Table 1.1). Other markers include the GyrA, rimmM, and Mpt83. The
®

Gen-Probe

AMPLIFIED Mycobacterium tuberculosis Direct Test offers a specificity of 92-

100% and a sensitivity of 97% in smear-positive pulmonary specimens targeting the 16S rRNA
gene. Roche AMPLICOR and BDProbe Tec

TM

tests have similar performance (4). In all cases,

2

however, the sensitivity is significantly reduced when using smear-negative/culture-positive
specimens.

Table 1.1. Sensitivity and specificity of nucleic acid amplification-based tests commercially
available for clinical specimens. AMTD: Gen-probe AMPLIFIED Mycobacterium tuberculosis
Direct Test. Adapted from (4).
Test

Genetic
marker

Roche Amplicor
BDProbe Tec
Gen-probe AMTD
GeneXpert
MTB/Rif
Gen-probe AMTD

16S rRNA
gene
IS6110 and
16S rRNA
gene
16S RNA
gene
TB and
rifampin
resistance
16S RNA
gene

Smear-positive pulmonary

Smear-negative pulmonary

Sensitivity
97

Specificity
>95

Sensitivity
40–73

Specificity
>95

90–100

92

33–100

83–97

92–100

>95

40–93

>95

100

98.3

48

N/A

92–100

>95

40–93

>95

The majority of MDR-TB strains harbor resistance to rifampin and isoniazid, two of the
best first line antibiotics for the treatment of TB. Because isolated resistance to rifampin is rare,
it is often used to indicate the presence of MDR-TB (5). Several tests focus on detecting
resistance to only rifampin, based on the assumption that rifampin resistance is likely to be
®

indicative of isoniazid resistance as well. For example, GeneXpert

MTB/RIF detects the

resistance to rifampin and TB simultaneously and is the only FDA-cleared test for the detection
of MDR-TB. It is based on TaqMan chemistry. A test that allows the detection of both
resistances simultaneously will be preferable to identify patients affected with a TB strain that is
only resistant to rifampin. A test that provides information about resistance to any other first line
3

of drugs (i.e., streptomycin, ethambutol, pyrazinamide) would be even more desirable to enable
the physician choose the most suitable drug.
An XDR-TB strain is an MDR-TB strain that is also resistant to any of the three existing
injectable second-line of antibiotics (i.e., kanamycin, amikacin or capreomycin) and any of the
fluoroquinolone drugs (5). A survey conducted among a network of 25 supranational
mycobacterial reference-laboratories showed that XDR-TB represents about 10% of the MDRTB cases (6). In another study conducted with HIV patients infected by XDR-TB, it was shown
that resistance was more likely to be transmitted than acquired, as none of the patient received
previous treatment for TB (7).
SNPs conferring resistance vary in their importance. For example, the SNPs present on
the rpoB gene of M. tuberculosis is responsible of more than 96% of the resistance to rifampin.
However, in order to efficiently detect resistance to rifampin or other antibiotic, it is necessary to
detect all mutations that are related to this specific resistance. As mentioned before, the number
of markers that are known to detect MDR/XDR-TB is large and this is one of the main reasons
for the non-availability of a rapid genetic test (Table 1.2). Until a multiplex and simple genetic
tests for SNPs becomes available, drug susceptibility testing (DST) will remain the only
alternative. DST is highly specific and sensitive because it is based on growth, but it requires
weeks to yield results as the growth rate is slow. Since the main outcome of a TB test is the
initiation of the appropriate treatment (2), it is crucial to provide an antibiotic-resistance profile
as early as possible. Tests that are based on molecular amplification can provide results within
hours and therefore, seem more appropriate for TB testing but they must provide information
equivalent to DST.

4

Table 1.2. Genetic mutation associated with drug resistance in TB. Adapted from (5)
Antibiotic

Gene

Percentage of Product of the gene
resistance to
this drug

Known mutations
(codon, nucleotide)

First line
Isoniazid
Isoniazid
Isoniazid
Isoniazid
Rifampin

katG
inhA
ahpC
kasA
rpoB

40-60
15-43
10
Unknown
>96

Catalase peroxidase
Reductase analog
Hydroperoxidase reductase
Carrier protein synthase
Subunit of RNA
polymerase

S315
C-15
G46

Pyrazinamide
Ethambutol
Streptomycin
Streptomycin

pncA
embB
rpsL
rrs

72-97
47-65
70
70

Pyrazinamidase
Arabinosyl transferase
Ribosomal protein S12
16S rRNA

M306, G406, Q497
K42, K88
A906, A510, C491

Second line
Ofloxacin
Capreomycin
Kanamycin
Fluoroquinolones

GyrA
tlyA
rrs
gyrA

87

Gyrase

D94G

90
75-94

16S rRNA
DNA gyrase subunit

A1401G
D94, A90

D516, S531, H526,
L522, L533, L511,
Q513

1.2. Potential of LAMP for Detection of TB, MDR-TB, and XDR-TB
One of the key developments in the last decade is in the area of isothermal nucleic acid
amplification approaches. Loop-mediated isothermal amplification (LAMP) is one such
technique employed for rapid isothermal detection of nucleic acids. It has significant potential
for field-based genetic testing of TB due to its simplicity and high yield of amplification product
(8). It is based on the use of polymerase with strand displacement activity (Bst polymerase large
fragment) and a set of four to six primers. A set of inner primers that both contain sequences
complementary to the sense and antisense strand of the target DNA initiate the amplification
process. These primers, named forward and backward inner primers (FIP and BIP respectively),
can form a stem-loop structures after being elongated by the polymerase. This structure results in
a self-priming event that ultimately results in an exponential growth of the amplification product
5

(8). A set of outer primers named F3 and B3, allows acceleration of the reaction by stranddisplacing the elongation product of FIP and BIP. The reaction can be further accelerated by the
use of an extra set of primers (named loop forward and loop backward or LF and LB) that can
hybridize in the loop structure that is formed by FIP and BIP. Because of the special design of
the primers and the high specific activity of the polymerase, LAMP produces large quantities of
amplified DNA. This high yield is accompanied by precipitation of pyrophosphate ions, which
results in increase in turbidity in a positive reaction mixture. For this reason, the product of the
reaction can be seen directly with naked eye at the end of the reaction without any postamplification modification. The increase in turbidity can also be measured in real-time, and the
time at which the turbidity increases above a certain background is related to the initial amount
of target DNA. LAMP can be performed in the presence of fluorescent DNA-intercalating dyes
®

(e.g., SYTO-82, SYBR

Green I), also allowing quantification. It is also not as inhibited by

substances generally inhibitory to PCR (e.g., humic substance, proteins, sputum components)
(9). Since its initial publication in 2000 (8), more than 600 studies are available targeting various
bacterial and viral agents important to global health (April 2012).
The possibility to detect TB by LAMP has been documented by numerous studies
targeting various genetic markers including the 16S rRNA gene and IS6110 (Table 1.3). The
limit of detection ranges from 1 to 100 copies. Studies are available with various sample types,
including sputum, pleural fluid, and blood, illustrating the potential of LAMP for clinical
samples. An assay based on real-time turbidimetry, for example, allowed the detection of 10
copies of M. africanum, M. bovis, M. microti, and M. tuberculosis which are members of the M.
tuberculosis complex (9). This assay targeting the 16S rRNA gene was tested on 200 sputum
samples from Nepalese patients and showed a sensitivity and a specificity of 100% (96 out of 96
6

culture positive-patients) and 94.2% (98 out of 104 culture-negative patients), respectively. This
is of relevance, as the detection of several markers is needed to recognize all the pathogenic
strains of TB, including non-members of the M. tuberculosis complex (e.g., M. avium). A
method that combines RT-LAMP and enzyme-linked immunosorbent assay (ELISA) was also
developed to detect directly the 16 rRNA of M. tuberculosis with a sensitivity of 1 RNA copy.
For this assay, the RNA was purified from sputum samples using a commercial kit (14). More
recently, it has been demonstrated that LAMP can also be used for SNP detection. Some of these
studies are based on a detection method that requires pre- and post-amplification step(s) such as
hybridization (18), digestion with restriction enzymes (19), or an enzymatic ligation, which are
not suitable for field deployment. In 2003, Iwasaki proposed a modified protocol of the LAMP
reaction for direct SNP detection (20). This protocol is based on the use of FIP and BIP primers
that both cover the SNP on the penultimate nucleotide of their 5’. The technique was tested for
genotyping a member of the cytochrome P450 family, and allowed discrimination of wild-types,
mutants, and heterozygotes from human genomic DNA and blood samples. LAMP was used for
in-situ typing of the L858R single-point mutation (21). This time, only the BIP primer contained
the mutation in its 5’ region (the exact position of the primer was not specified). The position of
the SNP in the primer may have an impact of the discrimination efficiency of mutations by
LAMP. For example, detection of SNPs was possible when using FIP or BIP primers that
contain the polymorphism on the ultimate nucleotide of their 5’ extremity (22). Elsewhere it was
shown that if the SNP was situated in the middle region of the FIP primer (and by deduction, of
the BIP primer too), it was not discriminated by LAMP. For these reasons, the authors used a
post-amplification method to type the SNPs. Thus it is likely that only some of the mutations will

7

be detectable by LAMP, depending on the possibility to design efficient primers in the SNPcontaining region of the target DNA.

Figure 1.1. (A) Replication of amplification among the 64 wells present on the polyester chip
TM

using Gene-Z

TM

. (B) Gene-Z

platform. For interpretation of the references to color in this

xand all other figures, the reader is referred to the electronic version of this dissertation

At present, there are no published studies related to isothermal typing of SNPs in M.
tuberculosis although efforts are underway at organizations such as Foundation for Innovative
New Diagnostics (FIND) focusing on SNP detection and sputum-based analysis. A method to
detect SNP isothermally will provide one of the most useful approaches in MDR/XDR TB
diagnosis. Once developed, the approach should be broadly useful in other areas where SNPbased genotyping is needed.

8

Table 1.3. Existing LAMP assays for tuberculosis. MT: M. tuberculosis; MB: M. bovis; MV: M.
avium; MI: M. intracellulare; MY: genus Mycobacterium; MA: M. africanum; MM: M. microti;
MC: M. canetti; MR: M. caprae; MP: M. pinnipedii; SP: sputum; SC: solid culture; LC: liquid
®

culture; PF: pleural fluid; BL: blood; SG1: SYBR

Green I. * 16S rRNA gene primers were

only Mycobacterium genus-specific. ** Amplification by reverse-transcript LAMP. Adapted
from (10).
Ref

MTB
species

(11)

(12)
(9)
(13)

(14)
(15)
(16)
(17)

Specimen
type

Genetic marker

Detection
method

MT, MB, Sputum
MV, MI, Solid
MY
cultures
Liquid
cultures
MT
Sputum

gyrB;
16S rRNA*

Endpoint
(SG1)

gyrB

MT, MB, Sputum
MA, MM
MT, MB Sputum
Solid culture
Pleural fluid
Blood
MT
Sputum
MT
Spiked
Sputum
MB
Sputum

16S rRNA gene

MT, MB, Sputum
MA, MM, Solid culture
MC, MR,
MP, SC

IS6110

16S rRNA**
16S rRNA gene
mpt83

9

Minimum
detected

Calcein/UV
fluorescence
Real-time
turbidimetry
Calcein/
manganese
chloride

40

N/A

60

10 copies
of DNA
1 pg of
DNA

Colorimetric
Real-time
turbidimetry
Endpoint
(SG1)

60
60

Endpoint
(SG1)

rimM

Assay
time
(min)
35 (solid
cultures)
60
(others)

90

60

60

5 copies
of DNA

1 copy
1 copy
of DNA
10 copies
DNA
1 copy of
DNA

TM

1.3. The Gene-Z

Platform

Existing equipment for molecular amplification are quite efficient but also very
®

expensive (e.g., The GeneXpert

costs $90,000). As part of the ongoing research within the

Environmental Genomics lab, our group has developed a low cost ($500) handheld gene analysis
TM

platform, named Gene-Z
TM

reactions. The Gene-Z

and that can performed simultaneously up to 64 independent

platform is entirely controlled wirelessly though a Wi-Fi module that
TM

can be connected to any wireless module such as an iPhone or an iPad. The Gene-Z

device

(Figure 1.1.A) can use microfluidic chips made of different materials including cyclo-olefin
polymer (COP) and polyester. Preliminary results with the COP chip indicate good replication
among the 64 wells of the chip. For example, there is no significant difference in threshold time
(Tt) between the 64 wells when the same number of copies is used (Figure 1.1.B). The Tt is the
time at which the amplification signal increases above its background value (see CHAPTER 2).
This Tt is exponentially related to the target copy number. Because Tt alone is used for
quantification, it is important that replication of the wells on the chip is good. While COP offers
the advantage of a higher transparency, polyester (another clear polymer) costs only 1/100

th

compared to COP on a per unit area basis. A new protocol to fabricate chips using 170 "m thin
polyester films was developed as part of this work (CHAPTER 5). Results indicated that the
lower transparency of polyester compared to COP, had no practical effect on the amplification
and/or the fluorescence detection. The polyester 64-well chip was prepared with an adaptation of
the protocol designed by Tourlousse (23). This high throughput Gene-Z platform offers the

10

potential to detect between 10 and 20 genetic markers simultaneously, depending on the number
of controls and replicates that are necessary.

1.4. Goal of this Study
The ultimate goal of this project was to develop an integrated detection system for TB
and MDR/XDR-TB that could be used at the POC in low resource settings. However, some of
the work presented here also includes targets other than TB because the work was completed as
part of developing the Gene-Z platform, which was also designed to detect many other bacterial
pathogens. Lower cost, ease of use, room-temperature stabilization, portability, rapid detection,
enhanced sensitivity, and specificity are some of the desired characteristics of the ultimate
system. Development of the device and the initial microfluidic chip has been previously
described by Stedtfeld and Tourlousse (23, 24). Optimization of the LAMP reaction for
enhanced detection limit and decreased time to result, stabilization of the reagents for long-term
storage without the need for refrigerator/freezer, development of an inexpensive and simpler
multiplex microfluidic chip for amplification, and demonstration of an isothermal method to
detect one SNP focusing on M. tuberculosis was part of this research.

1.4.1. Rationale
Several fluorescent DNA intercalating dyes have been used for LAMP but results vary
significantly from one study to another, making it difficult to establish which dye would be the
TM

most appropriate for use in the Gene-Z

system. An optimal dye for monitoring LAMP in the

microfluidic chip will be one that allows a rapid detection (i.e., low Tt), a low inhibition, and a
fluorescence enhancement upon amplification that is detectable using the low cost optical
11

TM

components of Gene-Z

. Therefore if would be valuable to compare several fluorescent dyes
TM

to evaluate which one performs best with LAMP and ultimately, with Gene-Z
TM

In order to perform LAMP using Gene-Z

.

in a POC setting, a method to stabilize the

reagents for field deployment is also needed. Freeze-drying is a well-known method to stabilize
molecular biology reagents for ambient temperature storage. This technique requires the use of a
carbohydrate (e.g., trehalose, sucrose) in order to protect sensitive molecules such as the
polymerase from oxidation or denaturation. However, little is known about the use of freezedrying to stabilize real-time PCR and LAMP reagents. A protocol for field deployment of
LAMP assays will be necessary for using Gene-Z.
The first chip that was used in the Gene-Z platform was made of COP and contained 128
micro-valves. The micro-valves were not only difficult to manufacture, but required the use of a
high pressure sealing system drastically increasing the complexity and cost of the device and
assay. Because the chip must be inexpensive, easy to fabricate, and used by health workers with
minimum training, it is critical to develop a chip that is easy to manufacture, made out of
inexpensive material, and easy to operate without training.

1.4.2. Specific aims

Specific Aim 1: Optimization of fluorescent dye selection for real-time LAMP
The optimization of real-time LAMP for fluorescence monitoring includes evaluating the
effect of various dye types and their concentration on assay performance (e.g., reaction time and
sensitivity). A number of DNA intercalating dyes with varying spectral ranges can be used for
12

®

LAMP (e.g., SYBR

Green, Pico Green, SYTO dyes, SYTOX Orange). Optimization of

polymerase concentration and pre-hybridization of the primers with target DNA may also be
helpful in developing a rapid and efficient protocol. This aim will be considered successful if
optimal dye and conditions can be determined for the set of dye selected.

Specific Aim 2: Development of a protocol to stabilize reagents for field deployment
As part of this aim, the effect of the trehalose concentration and primer design will be
evaluated on the performance of the freeze-dried reagents. This specific aim will be considered
successful if the reagents targeting at least 20 different signature sequences can be stored for up
to one year without significant loss of activity and sensitivity.

Specific Aim 3: Development of an inexpensive valve-less microfluidic chip for the Gene-Z
platform
This specific aim includes the determination of an inexpensive polymer that can be used
to manufacture chip by hot embossing, and the evaluation of possible cross-contamination
between reaction wells that are not enclosed from each other. This aim would be considered
successful if we can develop an inexpensive microfluidic chip, if all the reaction wells give
similar results in the Gene-Z, and if multiplex amplification is possible without crosscontamination.

Specific Aim 4: Development of a method to detect isothermally drug resistance in TB
This aim focuses on establishing a method to detect SNPs associated with MDR-TB. The
choice of mutation markers will be based on the occurrence with which they are associated to a
13

specific resistance. The method of Iwasaki (20) based on the presence of the SNPs on both FIP
and BIP will be first used. This approach is proposed by Eiken Chemicals Co., the entity
responsible for developing and commercializing the LAMP reagents and the primer-designing
software. Alternatively, the method of Iwasaki will be used in conjunction with the method
developed by Mitani and that use a specific DNA polymerase (Aac) and an additional proof
reading molecule (Taq mutS) to improve the discrimination of the SNP (25). In this case, Taq
mutS will be used with the commonly available Bst polymerase. This aim will be considered
accomplished if a method could be designed that discriminate between resistant and sensitive
genotype for a given SNP.

1.5. Developmental Team
TM

Development of the Gene-Z

platform was a concerted effort by the members of the
TM

Environmental Genomics team (Figure 1.2). Gene-Z

is composed of four main components

that were integrated in a same platform. My role in this project was focused on the molecular
biology (isothermal amplification of target DNA), optimization of the reaction conditions to
increase sensitivity and reduce time, development of method for isothermal SNP typing), freeze
drying, and development of an inexpensive, easy to use and disposable polyester microfluidic
chip.

14

Figure 1.2. Developmental Team and Roles of Various Members.

1.6. Organization of this Dissertation
The overall dissertation is organized as follows. Introduction to the problem is presented
in Chapter 1 (this chapter). In Chapter 2, characteristics and advantages of the LAMP reaction
are presented. Since this chapter was written for an edited book focusing on the potential of
LAMP for the multiplex detection of waterborne pathogens, it heavily refers to waterborne
pathogens. The factors related to LAMP, however, are all relevant for any pathogen. Chapter 3
presents material related to dye optimization as a draft manuscript. All work related to this aim is
complete and the draft manuscript is ready to be submitted. Chapter 4 presents data related to
freeze-drying (Aim 2), again as a draft manuscript. Some additional work related to optimization
of freeze-dried reagents in polyester chips is also presented. Chapter 5 presents work related to
15

the optimization and development of a polyester microfluidic chip for the Gene-Z platform.
Chapter 6 describe the work related to the development of a method based on LAMP to detect
SNP isothermally, with a proof of concept made using a mutation associated with isoniazid
resistance in TB.

16

CHAPTER 2. LOOP-MEDIATED ISOTHERMAL AMPLIFICATION AS A LOW COST
AND RUGGED TECHNIQUE FOR SCREENING OF MULTIPLE GENETIC
MARKERS

(This chapter was published as a book chapter related to waterborne pathogens under the title:
“Simple, Powerful, and Smart: using LAMP for low cost screening of multiple waterborne
pathogens”. Grégoire Seyrig, Farhan Ahmad, Robert D. Stedtfeld, Dieter M. Tourlousse, and
Syed A. Hashsham, (2010) in Environmental Microbiology: Current Technology and Water
Applications. Horizon Scientific Press, Norwich, UK.)

2.1. Introduction
Over the past half-century, treatment technologies and indicator-based monitoring have
served the water industry well in providing protection from waterborne pathogens. The former
will continue to serve as the primary barrier and is expected to include additional distributed
treatment technologies such as those being developed and implemented by various companies.
The utility of the latter to guard against the numerous waterborne disease outbreaks, however,
has been routinely questioned (26-28). Alternative strategies for providing water that is free from
microbial contaminants e.g., by directly measuring the pathogens or by using superior alternative
fecal indicators, are either cost prohibitive or yet to become available. Expensive pathogen
monitoring approaches cannot be put to use because in the developed nations, there is little
driving force to directly regulate the waterborne pathogens citing cost vs. risk as the determining
factor and in the developing nations funds are seldom available to even treat the water before
drinking, let alone measure the pathogens.
17

Although treatment must continue to serve as the first barrier against waterborne
pathogens, direct monitoring of pathogens should be the second barrier or an essential part of the
water treatment process. This is evident from the numerous outbreaks that occur in developed
countries and the huge loss of life that plagues the developing world. Between 1971 and 2002,
764 documented waterborne disease outbreaks were reported in the United States due to
contaminated drinking water, resulting in 575,457 cases of illness and 79 deaths. The annual
burden of waterborne diseases in the U.S. is much higher, with 38.9 million infections leading to
an estimated 19.5 million illnesses and resulting in productivity losses of the order of $20 billion
(29). Notable waterborne outbreaks include the Cryptosporidium outbreak in Milwaukee (30), E.
coli O157:H7 outbreak in Walkerton, Canada (31), and in Utah and New Mexico (32). Globally,
the impact of waterborne disease is much worse with more than 6,000 children dying each day
st

due to water-related illnesses. During the 21 century, the threat due to waterborne pathogens
will continue to be a major challenge due to increasing stresses on water resources, emergence of
pathogens with increased virulence and antibiotic resistance, and increased number of sensitive
populations. Hence, lowering the cost of screening tools for waterborne pathogens is a global and
urgent need.
Apart from reducing the cost, monitoring for multiple pathogens in a single assay is also
important for several other reasons. First, it is essential to monitor all known pathogens that
cause most of the waterborne outbreaks (Table 2.1). Second, some of the waterborne pathogens
also require genotyping (e.g., Escherichia coli and Cryptosporidium parvum). When genotyping
is not needed, the use of redundant markers for the same pathogen enhances reliability in making
calls for presence/absence (33). Third, as evident from Table 2.1, nearly 40% of the cases in
waterborne outbreaks are of unknown etiology. An intelligently designed multiplex approach
18

may help identify the potential candidates responsible for such outbreaks. Bringing some of the
recent advances in the area of genomics, microfluidics, and molecular biology may help device
such an intelligent approach.

Table 2.1. Overview of the etiology of waterborne disease outbreaks in the United States
between 1991 and 2002 (adopted from (29))
Etiological Agent
Cryptosporidium
Norovirus
Giardia
Salmonella non-Typhoid
E. coli O157:H7 and Campylobacter
Shigella
Campylobacter jejuni
E. coli O157:H7
V. cholera
Legionella
Plesiomonas shigelloides
Hepatitis A virus

Outbreak
14
12
25
3
1
9
7
11
2
6
1
2

12
2
14,264

0.1
0.0
100

1
77
33
1
207

Cryptosporidium Milwaukee 1993
Acute gastrointestinal illness
Chemical
Unidentified Small round-structured viruses
Grand total

Cases (%)
37.7
23.6
16
5.8
5.5
4.6
2.5
2.0
0.8
0.6
0.4
0.4

1
1
95

Campylobacter and Yersinia
Naegleria fowleri
Subtotal

Cases
5,371
3,361
2,283
833
781
663
360
288
114
80
60
56

403,000
16,036
577
70
433,497

Because a few microorganisms must be detected in a large volume of water, sample
concentration and target (or signal) amplification is equally important for monitoring of
waterborne pathogens. Strategies for target concentration have been discussed in Chapter 2.
Target amplification at the gene level has traditionally been accomplished using polymerase

19

chain reaction (PCR) (34,35) and reverse transcriptase PCR (RT-PCR), which require heating
and cooling cycles. If amplification by thermal cycling is replaced by isothermal approaches, it
eliminates the need for costly temperature cycling devices and may also result in shorter time to
complete the assay. Various isothermal amplification strategies include nucleic acid sequencebased amplification (36,37), strand displacement amplification (38), helicase-dependent
isothermal amplification (39), replicase-polymerase amplification (40) and LAMP (8). While all
of these methods have varying advantages/disadvantages, LAMP has a number of characteristics
that allow the assay to be rugged, sensitive, and specific and therefore suitable for the water
industry. These include: (i) generation of higher amounts of amplified DNA (500 ng/"L)
allowing simpler and inexpensive means for detection, (ii) no need for labeled probes and
requiring a single enzyme, and (iii) rapid, sensitive, and quantitative detection.
Since its first description in 2000 (8), over 500 studies have been described that use
LAMP including only a few but critical studies related to waterborne pathogens (41). These
initial studies on waterborne pathogens are critical because they have demonstrated the
superiority of LAMP compared to PCR when it comes to dealing with inhibitors, a problem that
is extremely important for waterborne pathogen detection. This chapter is devoted to introducing
LAMP as a low cost option to monitor waterborne pathogens directly in a multiplex manner.
After a brief description of other isothermal amplification techniques and the basic principle of
LAMP, major characteristics of this reaction that make it a suitable technology for the water
industry are described along with some of the aspects that need further development.

20

2.2. Isothermal DNA Amplification Techniques
Besides LAMP, which is the focus of this chapter and described in subsequent sections,
there are several other isothermal DNA amplification techniques including nucleic acid
sequence-based amplification (NASBA) (36,37), strand displacement amplification (38),
helicase-dependent isothermal amplification (39), recombinase polymerase amplification (40)
and rolling circle amplification. Application of NASBA to the detection of waterborne pathogens
is described in detail elsewhere (42). Key elements of other isothermal techniques are
summarized below.
Stand displacement amplification (SDA) is an isothermal amplification techniques based
on the action of a type II restriction endonuclease (e.g., HincII), an exonuclease deficient
polymerase that has a strand displacement activity (e.g., exo-Klenow) and two sets of primers.
One set (primers S1 and S2) contains a target binding region and a recognition site for HincII.
These primers target the sense and antisense strand of the template DNA and flank the region to
be amplified. The other set (primers B1 and B2) contain a target binding sequence located on the
5’ side of the region targeted by S1 and S2. In SDA, the primers form hybrids with the target
DNA after quick heating and cooling. After this initial hybridization step, the enzymes are added
to the reaction and extension begins. The extension due to one set of primers results in the
displacement of the strand extended by the other set of primers. The displaced single strands then
serves as templates for another round of hybridization, extension and displacement which further
results in a double-stranded form of the amplification product that contains the recognition site
for HincII at each extremity. This site is recognized and nicked by HincII, which results in a
DNA hybrid having a 3’ end that can be further extended by the polymerase. The repeating
action of nicking and elongation of each strand results in the exponential amplification, where

21

the amount of target sequence is doubled at each round of elongation (38). SDA can be sensitive
and fast. For example, Nadeau et al. described the amplification of 10 copies of DNA in 30 min
(43). SDA has been used for reverse transcription (44) and real-time quantitative monitoring to
detect various microorganisms including N. gonorrhoeae and C. trachomatis (45), and human
cytomegalovirus (46).
Helicase-dependent isothermal amplification (HDA) (39) and recombinase polymerase
amplification (RPA) (40) are two very similar methods that allow isothermal amplification of
DNA at 37°C. Both methods rely on the use of a pair of PCR-like primers, a polymerase, a
double-stranded DNA unwinding enzyme and a single-stranded DNA binding protein. The
reaction works in a manner similar to PCR with each primer being extended from the 3’ end after
annealing to each strand of the target DNA. The hybridization of the primers is facilitated by a
helicase (HDA) or a recombinase (RPA) that allows unwinding locally the target DNA, and by
single-strand DNA binding proteins to prevent the ejection of the primer by branch migration.
Both methods have been used in real-time to quantify DNA with fluorescent primer or DNA
intercalating dyes. A multiplex RPA assay is also described to detect sensitively (10 copies) four
different targets in the same mixture, using fluorescent probes of various colors (40). Recently,
HDA has been enhanced by fusing the polymerase with the helicase to avoid the use of several
enzymes in the same reaction, reducing the cost and complexity (47).
DNA can also be isothermally amplified using phi29 polymerase-based rolling circle
amplification (RCA). The technique is based on the hybridization of a specific primer to a
circular template DNA and its extension by the phi29 polymerase. After completing the
elongation of the whole template, the DNA polymerase reaches the primer-binding site again,
displaces the newly synthesized strand and continues the DNA synthesis along the circular

22

template for up to thousands of rounds. This results in large single-stranded DNA molecules
composed of repeated copies of the target sequence. The rate of amplification can be
significantly improved by using one or several additional primers that target the complementary
strand (i.e., the displaced DNA strand). The use of additional primers will results in a doublestranded DNA product of various sizes (48). The amplification efficiency of RCA is also related
to the size of the target sequence. For example, artificial sequences of 150 bp were amplified up
7

4

to 10 -fold using multiple primers, and a 8 kbp circular viral DNA was amplified 10 -fold (49).
RCA was used for different application including the detection of padlock probes, plasmids and
whole genome amplification of viral DNA (50).

2.3. LAMP Assay: A Simplified Description
LAMP is a simple assay with a complicated set of primers and has been described in
detail (8). The purpose of this section is not to repeat the above information but to provide a
simplified overview for the purpose of highlighting the salient features of LAMP. For more
detailed information, please refer to the original reference. Several parameters are important to
understand the mechanism of LAMP. These include location and purpose of various primer sets,
key mechanisms involved in copying of a certain region, strand displacement activity of Bst
polymerase, formation of secondary structure resulting in loops, and repetitive elongation of a
selected region resulting in the exponential amplification. These are briefly described below.

2.3.1. Location of LAMP primers on the target strand
LAMP assay critically needs at least two sets of primers named FIP/BIP (for forward
inner primer and backward inner primer) and F3/B3. Sometimes, a third primer set called Loop
23

F/Loop B is also used to enhance the efficiency and speed. Figure 2.1.A illustrates the location
of each of the forward primers (i.e., FIP, F3, and Loop F) of the three sets on the antisense
strand. A similar scenario can be envisioned for the backward or reverse primers on the sense
strand.
FIP is the forward primer of the first set of LAMP primers (FIP/BIP set) and has a special
design. It is two primers in one, to be used one at a time. The first part of FIP is a region named
F2 that is 18-24 nucleotides long and is complementary to the target sequence F2c. This part of
FIP is similar to a PCR forward primer. The second part of FIP is a tail (of F2) and consists of a
region named F1c, another 18-24 nucleotide long sequence that matches a region with the same
name on the strand being copied. F1c region is located approximately 20-40 nt away on the right
side of F2c. The two parts of FIP are connected by a short linker (TTTT). When Bst polymerase
acts at the 3’ end of FIP and starts copying F2c region and the target sequence ahead of it, the
copied strand results in the formation of a region complementary to F1c (hence called F1 for
obvious but somewhat confusing reason). This newly formed region F1 on the copied strand will
later help form the loop by hybridizing to F1c present as the tail of FIP.
F3 primer belongs to the second set of LAMP primers. It is also 18-22 nt long and located
to the left of the F2c region on the target strand. Displacement of the strand copied by FIP is one
of its main functions. As shown in Figure 2.1.A, when F3 is being copied in the 3’ direction to
the right, Bst polymerase will dislodge the strand made by FIP and hence produce additional
targets for copying. Since F3 region is on the original strand and not on the copies made by FIP,
F3 primer is capable of hybridizing only to the original strand.

24

Figure 2.1. (A) The LAMP primer positions. For ease of explanation only the forward primers
and their position on the antisense strand of the template DNA are shown but the backward
primers act similarly on the sense strand. (B) Loop formation from a strand elongated from BIP.
The F1c tail is hybridized with the F1 region.

25

Loop F primer is the third forward primer. It is located in the region in-between F2c and
F1c (Figure 2.1.A) When the strand is copied and F1c and F1 regions hybridize to each other
forming a loop (Figure 2.1.B), the location of Loop F primer ends up inside the loop. Hence the
name, Loop F. Loop F, however, acts in a manner similar to PCR primers and does not
participate in formation of a loop, which is solely due to the special design of FIP and strand
displacement by F3. Loop primer set is also optional. Its design is possible only in scenarios
where the region between F2c and F1c is reasonably long.

2.3.2. Key mechanisms involved in LAMP reaction
There are four key concepts that when put together explain the action of LAMP primers.
These are: (i) copying of DNA at the 3’ end of a primer (similar to PCR), (ii) strand
displacement at any location where Bst polymerase encounters a double-stranded structure, (iii)
formation of secondary structure, and loop due to the presence of F1c in FIP (or B1c in BIP), and
(iv) elongation of the product due to the interaction of the above three phenomena.
As for all polymerases, once the primers are hybridized, Bst polymerase starts copying at
the 3’end. For FIP, it happens, when F2 region of FIP hybridizes with its complementary
sequence (F2c) on the target strand. The whole strand gets copied and the copy contains the tail
(TTTT-F1c). Similarly, if F3 finds it complementary region on the target, it gets coped in the 3’
direction and the same is true for Loop F (or for that matter all the backward primers of the three

26

Figure 2.2. Lamp simplified: formation of the dumbbell shape structure and later structure.

The location of F3 implies that when it gets copied, the Bst polymerase will encounter the
copy that is made by FIP and displace it. Thus FIP has to act on the target strand already for F3
to replace the copied strand. The “enhanced” strand displacement activity of Bst polymerase is
27

important because at 65°C the double-stranded DNA is not denaturized. Therefore, the higher
strand displacement activity allows (i) the polymerase to synthesize DNA from a doublestranded template (that could be inhibited by the second strand), and (ii) to produce singlestranded DNA structures without the need of heating to a higher temperature. As a consequence,
the single-stranded products of FIP and BIP can be released more easily. The strand
displacement activity is also active in all other situations where a double-stranded structure is
encountered whether due to a doubles stranded DNA or due to secondary structure of the same
strand.
The Formation of loop occurs due to the presence of tail region (F1c) in FIP and the
corresponding F1 region on the strand copied by FIP. This is illustrated in Figure 2.1.B where
the copied strand contains F1c-TTTT-F2- NNN -LoopF- NNN -F1 followed by the rest of the
strand. F1c is free to hybridize to F1 forming a loop by everything that is in between. This first
loop formation does not result in a structure that can be copied from the end shown here (which
is 5’). The copied strand is acted upon by BIP from the other end, which results in a similar loop
creating a dumbbell shaped structure (Figure 2.2).

28

Figure 2.3. Schematic representation of the LAMP products accumulation.

For simplicity, the FIP/BIP primers are shown with tails and F3 is shown as straight lines
(B3 and Loop primers are not shown for simplicity but one can envision a similar reaction due to
those primers as well). As shown in Figure 2.2, FIP hybridized to its target strand, makes a copy,
F3 makes a copy of the target strand as well and as a result, FIP copy is dislodged. After
29

dislodging, FIP copy forms the loop at the left end. This copy made by FIP is acted upon by BIP
and B3 and as result a similar loop is formed at the right end. This is the key to forming the loop
at both ends. This is the famous dumbbell shaped structure, which is acted upon further to result
in elongation of the product. As shown, the loop regions have the F2 (and B2) regions and
therefore they can be further amplified by FIP (and BIP). Most of these subsequent
amplifications result in asymmetrical structures, one similar to a product observed before (which
is shown by a recycle line) and the other elongated. Thus, the primary dumbbell structure keeps
doubling in size. Because many amplicons of different sizes are produced (Figure 2.3), LAMP
product appears like a ladder in gel electrophoresis assay (as opposed to a single size product in
PCR). All these actions are obviously simultaneous which is nicely illustrated in a Flash
animation available at Eiken’s website: http://loopamp.eiken.co.jp/e/lamp/anim.html.

2.3.3. Parameters considered in designing LAMP primers
A number of factors are important in designing efficient LAMP primers including the
distance between the primers, their melting temperature, and secondary structures. More
specifically, LAMP primers are designed with the following constraints: (i) length of primary
amplified sequence (the region between FIP and BIP) – between 120 to 160 bases, (ii) length of
FIP/BIP primers (sum of both F1-F2 or B1-B2 regions – about 36 bases, (iii) length of target
sequence between the F1-F2 (or B1-B2) regions – 20 to 40 bases, (iv) no more than 60 bases
end-to-end between F3 and F2 (or B3 and B2), (v) melting temperature (Tm) of 65 °C for F1 (or
B1) region and 60°C for F2 and F3 (or B2 and B3) regions, calculated using the nearest-neighbor
model, and (vi) GC content of primers between 40 to 65 %. Due consideration is given to

30

formation of primer dimers and secondary structures. Primers that have a free energy greater
than -4 kcal/mol at the 5’ or 3’ ends are eliminated.
All of the above parameters are automatically taken care of in PrimerExplorer

TM

, the

online software for LAMP primer design developed by Eiken Chemical Company. They may
also be adjusted manually. The newest version available for PrimerExplorer

TM

is V4. It can

design LAMP primers from sequences of at least ~200 bases in length and accepts files in plain
text, FASTA, GenBank or multiple alignments format but cannot read degenerated bases (e.g.,
R, Y). The program may present hundreds of primer sets if the initial sequence is long enough
(with an upper limit of 2000 bases). The information file from each set of primers can be
downloaded in Microsoft Excel. Loop primers are designed separately by uploading the
information file of a specific set of primers in PrimerExplorer

TM

. It is possible that the software

fails to generate loop primers especially, when the primers have to be designed from a short
sequence or, when the distance between F1 and F2 is too short. A description of how to use V4 is
available at: http://primerexplorer.jp/elamp4.0.0/index.html, which also has a manual describing
other details. The specificity of the primers can be checked against GenBank using BLAST (51).

2.3.4. Carrying out the LAMP reaction
Incubation of the reaction mixture (DNA, primer sets, Bst polymerase, dNTPs, dyes,
buffer etc.) is carried out at 62 - 65°C. To provide the incubation temperate, many options are
available ranging from simple water baths to thermal blocks, or to more expensive thermal
cyclers including the real-time thermal cyclers. A turbidimeter specifically designed for LAMP
real-time assay is also available (52). Typically, LAMP is carried-out using 1.6 µM each of FIP

31

and BIP primers, 0.2 µM each of F3 and B3 primers, 0.8 µM each of LF and LB primers, 800
mM betaine, 1.4 mM of each dNTPs, 20 mM Tris/HCl (pH 8.8), 10 mM (NH4)2SO4, 10 mM
KCl, 8 mM MgSO4, 8 mM Triton X-100, and 0.2 to 2.4 units/µL of Bst DNA polymerase-large
fragment (8). However, the concentration of each component, notably primers and salts, can be
adjusted to each primer set. A number of options are available for visualization as described
below.

2.3.5. Visualizing the amplification product
One of the reasons LAMP assay found its niche as a simple technique in the low cost
genetic diagnostics arena was the simplicity of visualizing the positive reaction. When the
reaction volume is large (at least 25 µL), the positive reaction can easily be ascertained without
any visual aid due to the presence of copious amounts of DNA as well as magnesium
pyrophosphate precipitate. Signals from the latter can be further enhanced by the addition of
calcein (Figure 2.4 and 2.5.A). For smaller reaction volumes, a fluorescence–based (gel
documentation system, real-time thermal cyclers) or turbidimetric device is needed. When using
a fluorescent device to track the amplified DNA, the fluorescent dyes routinely used in molecular
®

biology assays (e.g., SYBR

Green (53,54), ethidium bromide (55), YO-PRO-I (56), and

nameSYTO-9 (57) can be used. Other dyes such as EvaGreen, SYTO-80 to 85 or PicoGreen may
also be used and may have certain advantages (Seyrig et al., unpublished, Figure 2.5.B).

32

Figure 2.4. Principle of LAMP detection using calcein. The pyrophosphate ions produced during
2+

the amplification may deprive the calcein molecule from the manganous ion (Mn ) to which it
is initially combined. Calcein deprivation of manganous ion results in emission of fluorescence
of this molecule. The fluorescence of calcein may be increased when the molecule combines
with magnesium ions (Mg

2+)

present in the reaction mixture (58).

LAMP also allows tracking of the amplification reaction in real-time using turbidimetric
measurements because LAMP assay produces significant amount of magnesium pyrophosphate
4-

(and turbidity) formed due to the pyrophosphate (P2O7 ) released from the deoxyribonucleotide
triphosphates (dNTPs) and chelating with the magnesium ions present in the reaction mixture.
33

The magnesium pyrophosphate precipitates because its solubility product is exceeded. In the
absence of amplification, the solution remains clear (58). The turbidity associated with the
amplification of DNA is a phenomenon that is specific to LAMP, and is attributed to the higher
amplification efficiency of LAMP compared to PCR (54). It is generally not observed during
PCR.

Figure 2.5. Real-time and direct detection of LAMP amplification product targeting G.
intestinalis genomic DNA. (A) Direct detection of LAMP amplification product with fluorescent
metal indicator (calcein). (B) Real-time amplification curves from 0 (closed circles); 10 (open
circles); 100 (closed triangles); 1,000 (open triangles) and 10,000 copies of genome (closed
squares). The fluorescent dye is SYTO-82.

Secondary approaches to visualize the DNA or the turbidity may also be used. Calcein, a
fluorescent chelating agent is the most commonly used secondary agent to detect fluorescencebased end-point detection of the LAMP product. Calcein initially chelates manganese ions
2+

(Mn ) present in the solution and consequently its fluorescence remains quenched (Figure 2.4
and 2.5.A). During the amplification, the generated pyrophosphate ions will progressively
34

deprive the calcein molecules from the manganese ions, resulting in an emission of green
fluorescence of the calcein when irradiated under natural, ultra violet (UV) or blue light (Figure
2+

2.5.A). Interaction of free calcein with magnesium ions (Mg ) resulting in stronger
fluorescence is also reported (58).
Under unusual circumstances (e.g., to check false positive reactions or to confirm that the
amplification is indeed due to LAMP primers), the LAMP reaction can also be electrophoresed
on an agarose gel and visualized with double-stranded-DNA fluorescent intercalating agents
®

such as ethidium bromide or SYBR

Green I. Opening the positive reaction vials of LAMP,

however, is highly discouraged because of extremely problematic contamination issues due to
the high amount of amplicons generated. On a gel, the typical electrophoresis pattern of LAMP
product is not a single band obtained after PCR, but a ladder-like pattern. This is because LAMP
forms amplified products of various sizes consisting of alternately inverted repeats of the
targeted sequence on the same strand (59). Rarely, the LAMP product can also be digested with a
restriction enzyme prior to the electrophoresis, resulting in a single band on the gel that
corresponds to the lightest band of the undigested product. In such a case, a restriction site has to
be present in the target nucleic acid, or in the non-targeting region of the FIP and BIP primers
(8).
2.4. Performance and Validation Characteristics of LAMP
2.4.1. Detection limit, time to positivity, and specificity in LAMP assays
LAMP has been validated for a number of viruses, bacteria, and protozoa in a variety of
samples (Table 2.2). Viruses include both DNA and RNA viruses, of course with reverse
transcription. Excellent reviews mostly focusing on clinical infections are available on the
applications of LAMP to clinical specimens (60). LAMP validation studies with waterborne
35

pathogens are only beginning to emerge. Among the waterborne pathogens validated, main
organisms include Vibrio cholerae and Vibrio parahaemolitycus, M. tuberculosis, M. avium, and
M. intracellulare, E. coli O157:H7, Cryptosporidium spp., and noroviruses (Table 2.2).
Even though both PCR and LAMP can detect a single copy in a carefully carried out
assay, LAMP is emerging as a more sensitive technique. This claim is based on studies that yield
positive results with a lower number of copies compared to PCR. For V. cholerae and V.
parahaemolyticus, for example, 2 copies of target gene could be detected using LAMP while
PCR could detect only 20 copies under the same conditions (61,62). In another study, LAMP
was shown to be 100 times more sensitive than PCR and allowed detection of 3 copies B.
pertussis DNA in purified nasal swabs (63). Similarly, M. avium had a 10-fold higher sensitivity
with LAMP when compared to PCR (11). For Chikungunya RNA virus, real-time LAMP was
able to amplify 20 copies of the target but real-time PCR failed to detect less than 200 copies
(64). While these studies are only indicative of the trend and may be mainly due to the lower
inhibitory effect of LAMP (as described in the following section), it is an important observation
for waterborne pathogens screening where PCR inhibition is a problem.
With three primer sets, LAMP is also highly specific to its target. Most studies related to
specificity, however, are for clinical targets. For example, a turbidimetry-based real-time LAMP
assay to detect M. tuberculosis was evaluated against 104 culture-negative sputum samples. The
assay detected 100% of the culture positive samples and only 5.7% of the culture negative
samples (9). In another study, LAMP assay detected all the 227 Salmonella strains tested and
showed no amplification when bacteria species other than Salmonella were tested (65). In a
similar study, LAMP discriminated specifically Salmonella O4 group from more than 100 strains
of O4 and non-O4 Salmonella (66). LAMP targeting Salmonella also showed an important

36

specificity in the presence of 100 ng on non-Salmonella genomic DNA (67). The high specificity
of LAMP also has been shown for assays targeting several other organisms including L.
donovani (68), V. parahaemolyticus (69) and Cryptosporidium (70). This high specificity is the
basis for expecting that genotyping for relevant waterborne pathogens (e.g., for
Crypstosporidium spp. and E. coli) should be possible for the purpose of source tracking.

2.4.2. The effect of inhibitors in LAMP assay
A key strength of LAMP for microbial detection is its insensitivity to inhibitors that often
plague PCR-based assays (41). Because of this, and its high amplification yield, LAMP is often
more sensitive than PCR for ‘difficult to amplify’ samples. This is particularly important for
monitoring of waterborne pathogen due to the need to filter large volumes of water samples,
which also concentrates inhibitors that may not be easily removed during DNA extraction. In one
study, Sotiriadou and Karanis showed that PCR failed to detect Toxoplasma gondii in nearly half
of spiked water samples while LAMP yielded positive results in all samples (71). The same
group (70) also reported that LAMP was able to detect Cryptosporidium species in roughly onethird of PCR-negative fecal samples from various hosts, and verified detection specificity by
sequencing of the LAMP products. Superior tolerance of LAMP to culture medium, and
biological substances in serum, plasma, blood urine has also been reported (72,73), along with
better sensitivity of detection of hepatitis B virus in serum (74).
The robustness of LAMP also simplifies sample processing because sample purification
using immunomagnetic separation (IMS) is not needed (41). In this study, polyester ARAD
(named after the company) microfiber filtration was combined with direct DNA extraction and
LAMP for detection of Giardia duodenalis and Cryptosporidium spp. in drinking water. A
37

commercially available kit was used for DNA purification with minimal modification, and of the
32 µL purified DNA eluent, 12.5 µL was directly analyzed in a 25 µL LAMP reaction. The
ability to analyze the majority of DNA extracted from large water samples is important since it
improves assay sensitivity. Using this methodology, 100 (oo)cysts spiked in 1000 L water
samples (karst water, treated river water, and surface water) were consistently detected. Because
minimal assay optimization is required for LAMP, it may also be more easily adopted compared
to PCR. PCR often requires optimized DNA extraction or addition of enhancing substances to
the PCR mixture to relief inhibition, especially when IMS is skipped.
Aside from making sample processing easier, direct DNA extraction (i.e., without IMS)
is also beneficial when multiple pathogens need to be screened for simultaneously, which is
expensive and cumbersome using IMS since a different separation step is necessary for each
pathogen. With the increasing demand for such multiplexed monitoring technologies (75),
LAMP is even more attractive. Novel filtering techniques capable of concentrating multiple
types of waterborne pathogens (bacteria, viruses and protozoa) simultaneously (76-78) will also
play a role in such systems.

38

Table 2.2. Performances of LAMP for the detection of bacteria, parasites and viruses. N/A: information not available
Ref.

Target

Method

Assay time
(min)

(56)

Bacteria
Ammonia oxidizing
bacteria (amoA gene)

Real-time turbidimetry
Real-time (YO-PRO1 Fluo dye)

88

Real-time turbidimetry
Real-time (YO-PRO1 Fluo dye)

N/A
N/A

(65)
(79)
(80)
(61)

(62)

(12)
(63)
(81)
(82)

Salmonella

A.
End-point turbidimetry
actinomycetemcomitans
(purified samples)
amoA gene DNA
Real-time turbidimetry
Real-time (ABC fluorescent probe)

Limit of detection
(copy number)
2

1.0 x 10

1

1.0 x 10

0

2.0 x 10 (C.F.U.)
0

2.0 x 10 (C.F.U.)
1

60

1.0 x 10

N/A
N/A

1.0 x 10

35
70

2.9 x 10

40
60

2.0 x 10

V. cholerae
- pure culture
- human feces

Non quantitative real-time turbidimetry

M. tuberculosis
B. pertusis (purified
nasopharyngeal swabs)
E. coli O157:H7

End-point (calcein)
End-point (calcein, turbidimetry and
electrophoresis)
Cy3 labeled dCTP

40
60

Real-time turbidimetry

67
61
51

4

1.0 x 10

Non quantitative real-time turbidimetry

V. parahaemolitycus
- pure culture
- spiked shrimp

4

Parasites
P. malariae DNA
P. falciparum DNA
P. vivax DNA

39

N/A

0
0

1.4 x 10

0
0

2.0 x 10
N/A
0
2.4 x 10
N/A
1

1 x 10

2

1 x 10

2

1 x 10

Table 2.2 (cont’d)
(11)

(83)
(57)

M. tuberculosis
complex
M. avium
M. intracellulare
Cryptosporidium spp.
in water samples
Trypanosoma brucei
rhodesiense

(9)

M. tuberculosis

(84)

M. tuberculosis

(85)

DNA viruses
Hepatitis B Virus

(86)
(87)
(88)
(89)
(52)

Herpes simplex virus
type 1/2, Varicella
zoster virus
Herpes simplex virus
type 1 Herpes simplex
type 2
Human herpesvirus 6
White spot syndrome
virus
Lambda

®

End-point (SYBR Green)

End-point (electrophoresis and
turbidimetry)
End-point (non-quantitative real-time
with SYTO-9 dye, electrophoresis or
®
SYBR Green I addition)
Real-time turbidimetry

0

60
60
60

5.0 x 10

63

1.0 x 10

35

1

5.0 x 10

0

5.0 x 10

0

~1 cell per mL of blood

1

35

®

2.0 x 10

10 - 15

1.0 x 10

Real-time (SYBR Green I)
End-point electrophoresis

®

N/A

5.0 x 10

45

6.0 x 10

Real-time turbidimetry
Real-time turbidimetry

24
24

5.0 x 10

Real-time (SYBR Green, SYTO-80 85, SYTOX orange, Pico Green,
EvaGreen and Calceine)

1

1
1

2
3

1.0 x 10

1

2.5 x 10

2

Real-time turbidimetry

47

1.0 x 10

Real-time turbidimetry

45

2.0 x 10

40

3

Table 2.2 (cont’d)
(90)
(91)

RNA viruses*
Norovirus

1

End-point electrophoresis and
hybridization with Alexa Fluor 488 and
594 labeled probes
Real-time turbidimetry

60

Real-time turbidimetry

33

2.0 x 10
1 x 10

2

(64)

Severe acute
respiratory syndrome
coronavirus
Chikungunya virus

(55)

Rift Valley fever virus

Real-time (ethidium bromide)

18

(92)

Japanese encephalitis
virus
Norovirus

Real-time turbidimetry

27

1 x 10 (PFU)

End-point turbidimetry

60

0.1 - 1.0 x 10

(93)

* Amplification by RT-LAMP

41

19

1.0 x 10

1.0 x 10 (PFU)
1
1

0

1

2.4.3. Ongoing improvements in LAMP assay
LAMP assay shows considerable variability in Tt for the same copy number (Table 2.2).
For example, 10 copies of Rift-Valley fever virus was detected in ~10 min (55) while detection
of 10 copies of P. malariae took 60 min (82). However, when the same target and experimental
procedures are used with similar types of clinical samples, reproducible standard curves have
been observed (85). Thus, the potential exists for reproducible quantification of LAMP without
performing a standard curve with each test. At present, accurate quantification of copy number
using LAMP seems to be inferior to real-time PCR, at least in the lower range of 1-100 copies.
Some attempts have also been made to improve the sensitivity of LAMP. A heatdenaturation of the target prior to its addition into the LAMP tube was shown to increase the
sensitivity of the LAMP assay by two fold (8,57). Although this step was described by Notomi et
al. in one of the original LAMP publications (8), studies incorporating the initial denaturation
step are limited. Other attempts using alternative fluorescent dyes have succeeded in lowering
the time to less than 15 min at least for one target. Similar studies are needed for most
waterborne pathogens.
Even though LAMP is rather difficult to perform in a multiplex format, there are a few
examples where multiplexing has been attempted (94). In this study, two species of Babesia (B.
bovis and B. bigemina) were simultaneously assayed using a multiplex LAMP. Separation,
however, required digestion of the amplicons with EcoR1 followed by electrophoresis, which
obviously takes more time than carrying out the reaction separately and raises contamination
issues. Other examples of LAMP multiplexing include identification of three types of hepatitis A
virus using a set of seven primers (95). It is evident that multiplexing is possible but the
advantages of developing it must be evaluated against other factors.
42

It is also possible to carry out LAMP directly within the cells using Cy3 labeled
nucleotides during the amplification (81) and quantified by epifluorescence microscopy. Such an
approach may allow biomass quantification based on genes related to specific functions (e.g.,
virulence and marker genes of waterborne pathogens) and may allow the study of pathogens in
biofilms present in water distribution systems.

2.5. Development Needs for Waterborne Pathogens Screening and Future Directions of
LAMP
Sample concentration is a key step in waterborne pathogens screening because of the
particulate nature of microorganisms and the very dilute concentrations. The subject has been
summarized in detail elsewhere (75). After filtration, water concentrates need to be further
processed to yield DNA that is suitable for LAMP. To expedite and simplify monitoring of
waterborne pathogens, these steps should preferable be integrated with LAMP in a single
microfluidic chips.

2.5.1. On-chip sample processing
There are a number of sample processing techniques that could be integrated on the same
platform that is being used to carry out LAMP. LAMP has been shown to be sufficiently
resistant to inhibitors to enable analysis of crude DNA extracts (41,96). However, separation of
selected pathogens using IMS may still be desirable for certain applications, e.g., genotyping,
and this can be performed on-chip (97,98). The latter systems were similar to macroscale IMS
tools, and involved mixing of antibody-coated paramagnetic beads (PMBs) and sample on-chip,
followed by retention of the bead-cell conjugates using magnets and washing to remove sample

43

contaminants. To eliminate the need for mixing, which is difficult to achieve in microfluidic
systems, IMS using fluidized beds of immobilized beads in microchannels has also been
explored (99,100).
Dielectrophoresis (DEP), which utilizes non-uniform electric fields to separate cells
based on size, shape, and dielectric properties (101), is another popular technique for cell
separation and/or concentration in microchannels. Because it requires no cell labeling, DEP is
increasingly being explored for a variety of applications, including monitoring of water supplies
(102). Recently, DEP was used to concentrate and remove bacteria from water (103,104), and
was also coupled with on-chip DNA amplification using PCR (105).
Irrespective of whether cell enrichment is performed or not, cell lysis and DNA
purification are necessary steps in sample processing, and many approaches have been developed
to accomplish these in microscale systems (106). While these techniques have almost uniquely
been developed for clinical diagnostics starting from blood, urine or saliva samples, they could
also be adopted for DNA extraction from water concentrates. For cell lysis, all common
macroscale techniques have been miniaturized in chip-based formats. These include chemical
lysis, thermal lysis, electrical lysis (107-109) and mechanical lysis (110-113). Thermal lysis is
arguably the simplest to implement in chips where heaters are already in place for LAMP or
using a laser. The latter technique, known as optothermal lysis, takes advantage of the strong
absorbance of certain PMBs in the infrared region, and was demonstrated with both carboxyland antibody-coated PMBs (114,115). Electrical lysis is especially attractive when combined
with DEP for cell concentration (116) or electrical readout of DNA amplification. For difficult to
lyse cells, mechanical lysis by bead beating or sonication may be most effective, but requires
involved chips and control mechanisms.

44

As is the case for off-chip DNA extraction, solid-phase extraction (SPE) using silica
phases is also the most popular technique for DNA purification in microscale systems (117,118).
Recently, Mahalanabis et al. successfully extracted DNA from gram-positive bacteria using a
porous silica monolith in a cyclo olefin polymer chip to mechanically shear bacterial cell walls
by passage through the silica bed (119). Others have also investigated SPE in polymer chips
because of their low cost and ease of fabrication (117), utilizing e.g., immobilized silica beads
(120). Following a different approach, Witek et al. achieved effective DNA purification from
whole cell lysates in photo-activated polycarbonate microchannels, without the need for
chaotropic agents (121). While the latter system still involved some organics for DNA binding
and washing, completely aqueous systems for DNA purification in microdevices also exist, using
chitosan or amino-silanes as DNA binding phases (122,123).
A drawback of SPE is that large elution volumes are needed to recover the majority of
bound DNA due to limited hydrodynamic dispersion in microchannels, but using a mobile bed of
silica-coated PMBs to control DNA extraction by movement of an external magnet addresses
this issue (124). Still, the ability to obtain solid phases that have high DNA binding capacity but
are not prone to clogging remains challenging. Dispersed PMBs may be more promising in this
respect, and both silica-coated and charge-switchable PMBs have been used for DNA
purification in micro-devices (125,126).
Based on a more in-depth comparison of different techniques for microscale sample
processing and the increasing demand for multiplexed detection, SPE using dispersed PMBs,
preferably using charge-switchable beads so concentrated chaotropes are not needed, may be
most promising strategy for on-chip integration of DNA extraction and LAMP. Such a technique

45

would provide high DNA concentration factors, be able to handle relatively ‘dirty’ samples, and
can also be easily implemented in low cost polymer chips.

Figure 2.6. (A) Images of 16 well COC chip with dilution series (5 ng, 0.5 ng, 0.05 ng, 0.005 ng,
and 0 ng) of DNA (stained with 1! Pico Green) with increasing exposure time. Each dilution is
placed in triplicate. Negative controls (0 ng DNA with 1X Pico Green) are shown by dotted area.
(B) Schematic depicting the components of the fluorescent optical imaging system; 1: Blue LED;
2: collimating lens; 3: lens holder; 4: heat sink; 5; excitation filter; 6: COP chip; 7: emission
filter; 8: CCD lens; 9: CCD.

46

Finally, a novel DNA purification method recently developed by Marziali et al., although
more expensive, is attractive due to its ability to extract highly concentrated DNA free of
inhibitors from unpurified samples (127). This technique utilizes electric fields to focus and
concentrate DNA in agarose gels, with concentration factors up to 10,000 (128). Because the
focusing effect is unique to DNA, other molecules (such as proteins and humic acids) can be
actively eliminated from the DNA concentrate. While in its current version this technique
requires a large bench-scale instrument (www.borealgenomics.com), its simple mode of
operation (i.e., using electrodes, akin to DEP) offers potential for miniaturization and integration
with nucleic acid amplification on-chip. Some of these techniques, however, may be expensive
to integrate and will not necessarily result in the lowest cost option for waterborne pathogens
screening.

2.5.2. Miniaturization of LAMP for multiplexing, genotyping, and mutation detection
Several groups have focused on developing LAMP for lab-on-a-chip applications.
Researchers from Toshiba Corporation recently integrated LAMP on a DNA chip, and were able
to discriminate between 6 single-nucleotide polymorphisms (SNPs) related to rheumatoid
arthritis (18,129). In another group, a one-step LAMP assay was developed in a polyacrylamide
gel-based micro-chamber for detection of a min amount of DNA in less than an hour. Translating
LAMP assays on lab-on-a-chip has tremendous potential for high throughput detection of
microbes directly in the field, at moderate costs compared to the expensive equipment required
for PCR. In a miniaturized design, the LAMP macro-chip itself could consist of 100 to 500
wells, interconnected in a manner that allows multiple samples and genes to be analyzed. A 16well (albeit not interconnected) microchip with 500-700 "m diameter and 500 nL volume

47

fabricated in a cyclic olefin copolymer is shown in Figure 2.6.A. Currently, the cost of a 25 "L
LAMP assay is approximately $5. Thus the cost of a 100-well real-time LAMP assay using 500nl reactions will be ~$10, a reasonable price for testing say a dozen pathogens, each with
multiple marker genes. LAMP primers can be easily dispensed by a multichannel pipettor or
spotted by DNA spotter. Each well can represent a single real-time LAMP assay. As shown
below, the 500 "m diameter itself is too large considering the resolution of the imaging system
(say a CCD camera) can be as low as 3-8 "m. Thus the challenge for multiplexing at low cost is
ruggedness of the microfluidics and not imaging capabilities. Such a system can also be modified
for hybridization-based mutation detection.

2.5.3. Lowering the cost of real-time imaging for high throughput screening
Real-time LAMP assays are commonly performed with turbidity-meters and real-time
PCR machines. The need for expensive instrumentation ($20,000-$100,000) and highly trained
personnel, currently limits their widespread applications in the water industry. Therefore, a costeffective (e.g., $1,000-$5,000), rugged, and simple testing alternative with multiple pathogens
screening capability is highly needed. Such a system could be based on the principles employed
in gel documentation system and can be built using low cost components such as blue light
emitting diode (LED), excitation and emission filters, and monochrome charge coupled device
(CCD) camera (Figure 2.6.B). Such a system is more than sufficient to visualize LAMP, end
point or in real-time. A four-fold dilution series of DNA (10 ng/"L, 100 bp ladder) stained with
1! PicoGreen was imaged using a low cost CCD camera (Meade Corporation) with increasing
exposure times of 0.1 to 60 sec using a blue LED under appropriate excitation (470±10 nm) and
emission (520±10 nm) filters for the dye. The extracted signals for wells containing 5 ng to 0.05

48

ng of DNA shows increasing signal intensities with increasing exposure time and reaching a
plateau at higher exposure times (Figure 2.7). The current detection limit for this platform is
8

approximately 0.05 ng (4!10 copies) much lower than what will be available after LAMP
amplification (which is ~500 ng/"L DNA). This detection limit can further be improved by using
a brighter light source (e.g., 1-20 mW lasers) and better optics. What is unique is due to
improvements in CCD technology, lasers, and LEDs; development of such instruments is
possible at a cost that is significantly lower than the currently available systems.

Figure 2.7. Signal intensity versus exposure time for the dilution series of DNA (5 ng, 0.5 ng,
and 0.05 ng) stained with Pico Green.

49

2.5.4. Validation studies with waterborne pathogens
LAMP is an emerging technique for genetic screening and its initial applications have
focused on infectious agents affecting very large populations. Yet, its application to waterborne
pathogens, which causes productivity losses of the order of $20 billion in the U.S. and results in
the death of 6,000 children each day globally, is insignificant. LAMP validation studies needs to
be carried out for all waterborne pathogens using the currently available end point and real-time
set-ups. Affordable and rugged equipment capable of carrying out real-time LAMP assay in a
highly parallel manner must also be developed so that direct monitoring of waterborne pathogens
becomes a norm rather than a choice we cannot make.

50

CHAPTER 3. OPTIMIZATION OF REAL-TIME LOOP-MEDIATED ISOTHERMAL
AMPLIFICATION FOR FLUORESCENCE DETECTION

3.1. Introduction
LAMP is an emerging technique for rapid detection and quantification of nucleic acid
signature sequences (8) . Because the reaction is isothermal, is less affected by inhibitors, and
more cost effective than polymerase chain reaction (PCR), it is becoming a preferred alternative
for field deployable devices for nucleic acid amplification testing. LAMP-based assays are
already the method of choice for the detection of pathogens in low resource settings
(17,57,130,131). Real-time monitoring of LAMP can be performed by measuring the increase in
turbidity or the increase in fluorescence of double-stranded DNA (dsDNA) binding dyes in the
reaction mixture (132) . The time threshold (Tt – defined as the time at which the fluorescence or
turbidity crosses a predetermined cut-off signal/baseline) is directly related to the amount of
target sequence copies, which permits accurate quantification through the use of standard curves.
Fluorescent dye-based real-time LAMP has sensitivity similar to turbidity-based LAMP but is
generally faster (shorter Tt). This is illustrated by several studies showing that LAMP detection
time can be reduced by up to 35% when monitoring fluorescence rather than turbidity
(56,133,134). Fluorescence-based real-time LAMP can be performed in any conventional
microbiology laboratory equipped with a real-time PCR thermocycler. Additionally, high
amplification yields and the isothermal nature of LAMP make it well suited for integration with
simple POC diagnostics devices in limited-resource settings (23,24,135,136).
While the optimal conditions required for endpoint LAMP are well described
(8,58,60,135), a systematic examination of conditions specific to real-time fluorescence detection
51

have not been shown previously. Several dsDNA-intercalating dyes have been used for LAMP
but the reaction conditions and their performance vary significantly among studies.
®

SYBR Green I, SYTO-9 and YO-PRO-I have been used in various LAMP studies for pathogen
detection (e.g., S. enterica, Leptospira, hepatitis B virus) with good detection limit (10 copies of
target DNA) but with reaction times ranging from ten min to more than one hour. Reaction
conditions such as dye concentration, the use of a pre-hybridization step, sample type, primer
number (four to six), or incubation temperature varied significantly among studies. This
variability reduces the ability to establish a valuable dye comparison for real-time LAMP assays
(85,137,138). To the best of our knowledge, there is yet no information available that relates the
effect of dye type and concentration on LAMP performance. For quantitative PCR (qPCR),
which is a more established technology, it is known that certain dyes allow shorter detection time
and higher sensitivity. For example, SYTO-82 has been shown to not inhibit qPCR and allow a
®

50-fold better detection limit compared to SYBR Green I (139). Because LAMP and PCR both
use sensitive thermostable polymerases, we hypothesized that the type and concentration of
fluorescent dye may also affect the performance of LAMP.
In this study, we evaluated 11 dsDNA binding dyes, including several SYTO dyes and
®

SYBR Green I, for LAMP, with the goal of reducing detection time while maintaining
sensitivity.The effect of the concentration of each dye was evaluated with a LAMP assay for M.
tuberculosis and a constant concentration of target DNA. The sensitivity of LAMP was then
examined using the dye and concentration that resulted in the shortest detection time. Assay
sensitivity was evaluated with four pathogens including M. tuberculosis and S. enterica. Our
results show that the choice of dye and particularly its concentration may have a drastic effect on

52

the LAMP detection time. It was also demonstrated that a one minute pre-incubation of the
primers with the target DNA at 95°C prior to the addition of the polymerase (which would be
denatured at this temperature) (57) may be useful in improving assay sensitivity for certain
targets. This additional step results in a pre-hybridization of the primers with the target sequence.
The work reported here supports that the dye SYTO-81 may be more suitable for real-time
TM

detection in a platform with low cost optical settings (Gene-Z

). These results may significant

implications in the development of diagnostic device for the detection of infectious diseases at
the POC.

3.2. Materials and Methods
3.2.1. DNA targets
Genomic

DNA

from

M.

tuberculosis

(ATCC

25177),

methicillin-resistant

Staphylococcus aureus (MRSA, ATCC 700699D-5), Giardia intestinalis (ATCC 30888D) and S.
enterica subsp. enterica serovar thyphimurium (ATCC 700720D) was obtained from the
American

Type

Culture

Collection

(ATCC,

Manassas,

VA)

and

resuspended

in

diethylpyrocarbonate-treated, nuclease-free sterile water (Fischer Scientific, Pittsburgh, PA).

3.2.2. LAMP primers
A set of six specific LAMP primers (FIP, BIP, F3, B3, LF and LB) (APPENDIX 1) was
used for each target. Primer sets for G. intestinalis, S. aureus, and S. enterica were designed
using

Primer

Explorer

version

4

(Eiken

Chemicals

Co.,

LTD,

Tokyo,

Japan,

http://primerexplorer.jp/e). Primers for S. aureus were designed to target the mecA gene from a
consensus of 15 sequences that were aligned with Bioedit Sequence Alignment Editor (Ibis
53

Biosciences, Carlsbad, CA). Primers for S. enterica were designed from a consensus of three
sequences of the fljB gene. For G. intestinalis only one sequence was used to design a primer set
targeting the !-giardin gene. The specificity of G. intestinalis, S. enterica and S. aureus primer
sets was checked against the GenBank database using NCBI BLAST (140). For M. tuberculosis,
primers described in the literature were used to target the 16S rRNA gene (9). All primers were
obtained from Integrated DNA Technologies (IDT, Coralville, IA).

3.2.3. Fluorescence-based LAMP
LAMP reactions were performed in a volume of 20 µL consisting of 1.6 µM each of FIP
and BIP primers, 0.2 µM each of F3 and B3 primers, 0.8 µM each of LF and LB primers, 0.8 M
betaine (Sigma, St Louis, MO), 1.4 mM of each dNTP (Invitrogen Corporation, Carlsbad, CA),
20 mM Tris-HCl (pH 8.8), 10 mM (NH4)2SO4, 10 mM KCl, 8 mM MgSO4, 8 mM Triton X100, 0.2 to 2.4 units/µL of Bst DNA polymerase, large fragment (New England Biolabs Inc.,
Ipswich, MA) and various concentrations of a dye: SYTO-80, SYTO-81, SYTO-82, SYTO-83,
®

SYTO-84, SYTO-85 or SYTOX Orange, SYBR Green I, and PicoGreen (Invitrogen
Corporation), EvaGreen (Biotium Inc., Hayward, CA) or calcein (“Fluorescent detection
reagent”, EIKEN Chemical Co., Ltd.). Depending on the color range of their emission maxima,
®

the dyes are referred to here as either green (for SYBR Green I, PicoGreen, EvaGreen and
calcein) or orange (for SYTOX-orange and all SYTO dyes). Samples were loaded in 200-µL
PCR tubes (VWR International, West Chester, PA) and incubated at 64°C for 36 min in a
Chromo4TM Real-time PCR detector (Bio-Rad Laboratories, Hercules, CA). Fluorescence
intensity was measured every minute with the channel 1 (for green dyes) or with Channel 2 (for
54

®

orange dyes) of the thermal cycler using the calibration settings for SYBR Green I (green dyes)
or Cy3 (orange dyes), furnished by the manufacturer. For pre-hybridization of the target DNA
with the primers, samples were incubated for 1 min at 95°C with a polymerase-free LAMP
reaction mixture, and cooled at 4°C prior to adding the polymerase. Fluorescence data and Tt
values were collected with OpticonTM software v3.1 (Bio-Rad) and transferred to Microsoft
Excel where they were converted into SNRs as follow: SNR = (X-µ)/", where, X is the
fluorescence signal (arbitrary units, a.u.), µ the baseline signal and ", the standard deviation of
the baseline signal.

3.2.4. Fluorescence-based LAMP in compact device
Fluorescence-based LAMP was also performed in a portable genetic diagnostics device,
TM

termed Gene-Z

TM

(84) to determine which dye was better suited. The Gene-Z

device is

characterized by a simple, inexpensive, and compact mechanism for real time detection of
multiple reactions with volumes less than or equal to 2 µL. Briefly, the optical components
consist of a single photodiode to capture emission signals that are captured via 64 individual
750-µm diameter optical fibers, one for each reaction well. Time-lapse initiation of 64 different
light-emitting diodes (LEDs) (with a wavelength to excite fluorescence in each reaction well)
allows differentiation of fluorescent signals. A single 12.5-mm diameter long-pass glass
emission-filter is placed between the photodiode and the optical fibers to block the light used for
excitation. A disposable polyester chip is used for equal sample displacement into each of the 64
reaction wells. Two separate chip designs were used during the course of these experiments with
reaction volumes of 2 µL and 0.5 µL. In detail, the chip contains four separate sample loading
55

channels, each with 16 reaction wells. The samples were loaded into each of the individual
TM

channels of the microfluidic chip. The microfluidic chip was then incubated in the Gene-Z

device for 30 min at 63ºC. During incubation, fluorescence data from each reaction well is
measured every 12.8 s. Readings are streamed wirelessly (using a WiFi module inside the
device) to an iPod Touch where data is sorted and can be emailed for further processing (84).For
each plot, data were smothered by averaging 19 data points. The background signal, that was
arbitrarily chosen to be the average signal of the 17 first data point, was removed from every data
point. Velocity plots were constructed by deriving (slope based on six data points) the
amplification curves.
Experiments tested with the Gene-Z device included comparing real-time amplification
®

plots for experiments performed with SYBR Green I dye and a concentration gradient of
®

SYTO-81. Tests with SYBR Green I were performed with 470 nm excitation blue LEDs, and a
525 nm emission filter, and tests with SYTO-81 were perfromed with 525 nm green LED and a
570 nm emission filter.

3.3. Results and Discussion
3.3.1. Effect of dye type and concentration on threshold time
The Tt of fluorescence-based LAMP was compared for various concentrations of orange
and green dyes, using the BioRad Chromo4

TM

real-time thermocycler and a constant amount of

M. tuberculosis genomic DNA (0.1 ng or approximately 2

4

10 genomes). For each dye, a

threshold baseline was defined as the average florescence signal of the five first cycles plus five

56

times the standard deviation of this signal. An optimal dye concentration, defined as the concentration that resulted in the shortest Tt, was determined for each dye (Figure 3.1) although
several of the dyes were working efficiently over a broad range of concentrations.
Amplification curves were observed for all orange dyes (Figure 3.1). SYTO-82, SYTO84, and SYTOX orange resulted in the shortest Tt (p < 0.05) and SYTO-81 had the widest
working range of concentrations. SYTO-82 at 2 µM and SYTO-84 at 1 µM, and SYTOX orange
at 0.1 µM resulted in the shortest Tt (< 7.2min) of all orange dyes tested. Higher concentrations
of SYTOX orange resulted in inhibition of amplification (data not shown). Besides SYTO-80,
which resulted in a Tt of 11.8 min at its optimal concentration, the SYTO dyes gave a positive
result in 7 to 9 min, with an optimal concentration range of 1 to 2 µM. All SYTO dyes had a
working concentration range of 0.1 to 10 µM. Within this range, Tt was only moderately affected
by the concentration. Above 10 µM, amplification was observed for SYTO-80, 81 and 84, but
not for the other SYTO dyes. In fact, the Tt of SYTO-80 and 84 were doubled when the
concentration was increased from 10 to 20 µM. In contrast, there was almost no difference in Tt
over all the concentrations tested for SYTO-81. At its optimum concentration (2 µM), the Tt of
SYTO-81 was 7 min compared to 9 min when used at 20 µM revealing a wider working
concentration range for this dye than the others.

57

Figure 3.1. Effect of orange fluorescent dyes on the Tt of real-time LAMP. Reactions were
performed with 0.1- 20 µM of SYTO-80-85 or SYTOX orange. When no amplification occurred,
no point is shown. Target: 0.1 ng of M. tuberculosis genomic DNA. The error bars represent the
standard error from triplicate reactions.
58

Table 3.1. Summary of the results obtained with the dyes used in this study.
Fluorescent dye
Green dyes
®
SYBR Green I
PicoGreen
EvaGreen
Calcein

Abs./Em.* (nm)

Optimal
concentration

497/520

0.25 X

7.39 ± 0.14

502/523
500/530
495/515

1X
0.1 X
0.1 X

8.02 ± 0.01
8.06 ± 0.14
13.2 ± 0.12

Tt (min)

Orange dyes
SYTO-80
531/545
1 µM
11.8 ± 0.22
SYTO-81
530/544
2 µM
7.77 ± 0.04
SYTO-82
541/560
2 µM
6.77 ± 0.10
SYTO-83
543/559
1 µM
9.51 ± 0.37
SYTO-84
567/582
1 µM
6.97 ± 0.13
SYTO-85
567/583
2 µM
8.37 ± 0.11
SYTOX orange
547/570
0.1 µM
7.17 ± 0.05
*Absorption wavelength/Emission wavelength. The real-time thermocycler was set on channel 1
or 2 to measure the green or orange dyes respectively

®

Amplification curves were also observed for all green dyes, with SYBR Green I
resulting in the shortest Tt. The molar concentrations of several dyes were not provided by the
manufacturers. Hence they are expressed here as dilution factors of the standard working
®

concentration of 1 X rather than µM (Figure 3.2). SYBR Green I at 0.25 X resulted in the
shortest Tt (7.4 min) among all the green dyes tested. However, this dye had a very narrow
®

working range. No amplification occurred for SYBR Green I concentration outside the 0.1 - 0.5
®

X range, confirming an earlier report (139). The same optimum SYBR Green I concentration as

59

in the present study was described; therefore the mechanism of inhibition by this dye may be
®

similar for LAMP and PCR. The inhibitory effect of SYBR Green I on LAMP has been noted
previously (141), although information about optimal concentration is lacking. EvaGreen had a
similar response with an optimal concentration of 0.1 X corresponding to a Tt of about 8 min and
a working range of 0.05 to 0.5 X. EvaGreen has already been successfully used in LAMP at a
concentration of 0.5 X (142). It has also been shown to inhibit LAMP, although the dye
concentration of the assay was not given (133). In another study, the Tt of EvaGreen-based
LAMP doubled when the concentration was increased from 0.1 to 0.25 X. Real-time monitoring
®

of LAMP using PicoGreen had a Tt similar to SYBR Green I and EvaGreen-based LAMP
(approximately 8 min) at an optimal PicoGreen concentration of 1.0 X, but it had a wider
working range of 0.05 to 5.0 X with positive amplification curves at all concentrations.
PicoGreen (1X) has been used in a LAMP assay to detect foot-and-mouth disease virus (10
detectable copies) in about 22 min (143). Calcein, a common dye for visual endpoint detection of
LAMP (58), also allowed the detection of LAMP in real-time. However, the Tt was almost twice
®

as long as that of LAMP using SYBR Green I. In contrast to other dyes, calcein does not bind to
DNA. It fluoresces when deprived of calcium ions in solution. Free calcium ions decrease
because of their association with pyrophosphate ions released from dNTPs during DNA
®

amplification. For the target tested, all other green dyes (SYBR

Green I, PicoGreen, or

EvaGreen) were superior to calcein on the basis of Tt measured by a commercial real-time PCR
machine. Ordinarily this benefit of fluorescent dye and lower Tt will be negated because the very

60

purpose of using calcein is to be able to see the end point by naked eye. However, when
combined with field deployable low-cost genetic analysis systems, some fluorescent dyes may be
better compared to calcein.

Figure 3.2. Effect of green fluorescent dyes on the Tt of real-time LAMP. Reactions were
performed with 0.01 to 5.0 X of Calcein, EvaGreen, PicoGreen or SYBR

®

Green I. When no

amplification occurred, no point is shown on the graph. Target: 0.1 ng of M. tuberculosis
genomic DNA. The error bars represent the standard error from triplicate reactions.

Our results suggest that fluorescence-based real-time LAMP yield shorter Tt than
turbidity-based real-time LAMP. Except for calcein and SYTO-80, all dyes resulted in detectable
amplification of 0.1 ng of M. tuberculosis after 7 to 9 min (Table 3.1) This is almost twice as
61

fast as the results of Pandey et al., who used the same primer set in a turbidity-based real-time
LAMP assay, and detected 0.1 ng of M. tuberculosis in about 15 min (9). Pandey used a lower
concentration of inner primers than the concentration that is generally used in LAMP. However,
we were not able to see any difference in Tt between the concentration of inner primers used by
Pandey (1.2 µM) and the regular one (1.6 µM) (data not shown). Similarly, Chen et al. showed
that SYTO-9-based real-time LAMP was twice as fast compared to turbidity-based real-time
LAMP in detecting Vibrio parahaemolyticus (133). Likewise, Aoi et al. obtained a Tt for YOPRO1-based real-time LAMP that was 20 min shorter than a turbidity-based real-time LAMP.

3.3.2. Fluorescence intensity and enhancement during LAMP
The fluorescence intensity of the dyes during LAMP was analyzed by plotting raw
fluorescence data (Figure 3.3.A) and SNR (Figure 3.3.B) as a function of amplification time.
Absolute fluorescent intensity and noise play an important role in the choice of detector. After
completion of the LAMP reaction (indicated by plateauing of the signal intensity), SYTO-82 at
its optimum concentration had the highest fluorescent signal among all the dyes. The
fluorescence intensity of SYTO-82 increased about 1000-fold over the noise as a result of the
amplification process. The fluorescence intensity of SYTO-81 at optimum concentration had a
similar profile but the increase in SNR was only 100-fold at the end of the reaction. Furthermore,
SYTO-81 has a slight concentration-dependent inhibitory effect on LAMP. Higher
concentrations of SYTO-81 (up to 20 µM) were evaluated and resulted in a fluorescent intensity
and enhancement that were even higher than that of SYTO-82 but at the cost of a slight increase
in Tt (APPENDIX 2). Other SYTO dyes had fluorescence intensities 5 to 20 times lower than

62

SYTO-82, and fluorescence enhancement ranging from 10- to 100-fold. With a fluorescence of
about 0.01 a.u. after amplification, SYTOX Orange, was 100-fold less bright than SYTO-82. The
fluorescence of SYTOX Orange increased to approximately 70-fold after amplification. Among
the green dyes, calcein was the brightest with a fluorescence ranging between 0.4 and 0.6 a.u.,
when amplification occurred. However, calcein also had the highest baseline signal (0.1 a.u.),
®

hence the low SNR of 25, the lowest among all dyes tested. EvaGreen and SYBR Green I
showed low fluorescence enhancement, similar to calcein, and were almost 100-fold less bright.
®

PicoGreen was twice as bright as SYBR Green I and had 2-fold higher enhancement in
fluorescence.
Overall, SYTO-81 and SYTO-82 resulted in the highest fluorescence signal and
enhancement. However, based on their fluorescence responses, all dyes tested were suitable for
fluorescence-based real-time LAMP using the Chromo4TM thermocycler. Similar results are
expected if other conventional real-time thermocyclers (e.g., ABI Prism, Rotor-Gene) are used as
long as they have comparable sensitivities and dynamic range as the Chromo4TM. Much research
has focused on developing rapid and inexpensive tests to amplify nucleic acids in real-time using
miniaturized systems (144). Developers of such systems including our own group aim to
integrate small and low-cost optical setups (84), which can vary significantly in terms of
sensitivity and dynamic range. These parameters should be taken into consideration when
choosing an appropriate fluorescent dye for real-time LAMP because, as our results suggest,
fluorescent DNA dyes have varying brightness, resulting Tt, and final SNR after amplification.

63

Figure 3.3. Comparison of fluorescence intensity (A) and SNR amplification profile (B) of dyes
of interest. Target: 0.1 ng of M. tuberculosis genomic DNA. Data points represent the average
signal of three amplifications.
64

For these reasons, dyes with high brightness and high fluorescent enhancement such as SYTO-81
and SYTO-82 or calcein may be more appropriate for simple optical setups. This may change
with the availability of better photodiodes and improvements in light collection systems.

3.3.3. Effect of Bst polymerase concentration on threshold time
The possibility to reduce Tt was also investigated over a range of Bst polymerase
concentrations using 0.1 ng of M. tuberculosis and SYTO-82 or SYTO-81 dye (APPENDIX 3).

For both SYTO-81 and SYTO-82, an optimal polymerase concentration was defined as the
concentration that allowed the shortest Tt, and was determined to be 0.64 - 1.28 units/µL. Below
this concentration of Bst polymerase, Tt increased, suggesting a lack of sufficient polymerase in
the reaction mixture. Above this concentration range, Tt increased again, most likely due to
inhibitory effect similar to what had previously been observed for polymerases used in PCR (35).
Because no significant difference (p > 0.05) was observed between 0.64 and 1.28 units/µL, 0.64
units/µL was chosen as a suitable Bst polymerase concentration for our assays.
A Bst polymerase concentration of 0.64 units/µL will cause a Tt reduction of
approximately 14% (~1 min) when compared to the conventionally used concentration of 0.32
units/µL (56,57,85,133,137,138,145-148). However, these results were obtained using purified
genomic DNA, and little is known about the effect of higher concentrations of Bst for clinical
samples. Use of higher concentrations of polymerase has been described for PCR and generally

65

allows faster amplification (149), but may also result in loss of target specificity (35). That is not
to say that the same will happen for LAMP. Even though it is a polymerase-based assay, it may
not lose specificity because specificity in LAMP is superior due to the use of four to six primers
(8). A more systematic study to evaluate the effect of increasing Bst polymerase concentration on
specificity of LAMP may be necessary to fully characterize the effect.

3.3.4. Evaluation of SYTO-82-based real-time LAMP for the detection of other pathogens
SYTO-82-based real-time LAMP was evaluated with several primer sets targeting four
human pathogens (M. tuberculosis, S. aureus, S. enterica, and G. intestinalis) with serial
4

dilutions of the genomic DNA (10-10 copies) for each target individually (Figure 3.4) An
additional pre-hybridization step was used with the goal of reducing detection time and
improving sensitivity. In contrast to the limit of detection, the limit of quantification is the
minimum numbers of copies of target than can be significantly quantified using a standard curve.
For all organisms, linear correlations confirmed the potential of real-time LAMP for
quantification of target DNA. Assays made with preheated templates occasionally had shorter Tt
and were generally more sensitive than assays with non-preheated templates. When G.
intestinalis DNA (10 copies) was present in the preheated samples, the Tt was 4.5 min (45%)
shorter compared to non-preheated samples (p < 0.05). In contrast, the Tt was not statistically
improved when using a preheating step for M. tuberculosis, S. aureus or S. enterica. However,
an increased LAMP sensitivity was observed for all assays with preheated templates. For M.
tuberculosis, pre-heating the templates allowed detection of 10 copies of DNA, which was 20–
fold higher than the sensitivity obtained for non-preheated samples. Similarly, for S. aureus,
66

denaturation of the target resulted in an assay 10-fold more sensitive (10 copies). In addition, the
limit of detection was the same as the limit of quantification for all assays with preheated
templates.

Figure 3.4. Detection of several pathogens by real-time LAMP using SYTO-82. The effect of a
pre-hybridization of the primers with the target DNA on the sensitivity and Tt of real-time
LAMP is also presented. Close circles: pre-hybridization; open circles: no pre-hybridization.

Preheating the template prior to Bst addition may considerably improve the limits of
detection and quantification of real-time LAMP. In some cases, it may also accelerate the
reaction by up to 45%. This phenomenon has been described earlier (55,135) and might be due to
67

the denaturation step allowing a faster and more efficient hybridization of the LAMP primers
with the template DNA at the beginning of the reaction. However, a heat denaturation of the
template requires an additional step and therefore adds time, as the polymerase needs to be added
after the preheating step to avoid its denaturation. Whether this preheating step is beneficial in an
assay scheme will depend on the requirement and conditions, including desired sensitivity and
rapidity.

Amplification reactions performed in this study typically show faster Tt compared to
previously published results with real-time LAMP (Appendices 2 and 5). This higlights the
ability to detect specifically and quantitatively minute amounts of target nucleic acids in a short
period of time. Ten copies of at least four different organisms (M. tuberculosis, S. enterica, G.
intestinalis and S. aureus) were detected in less than 21 min of incubation. Similar results were
achieved by Peyreffite et al. (2008) who were able to detect 10 copies of Rift Valley fever virus
by real-time reverse transcript LAMP in about 18 min, using ethidium bromide (55). However,
this dye is a known mutagen and hence not suitable for use in user-friendly field deployable
devices. The dye YO-PRO1 was also used in real-time LAMP to detect up to 10 copies of P.
jirovecii in less than 20 min (148). However, the target sequence was incorporated in a plasmid
and it is possible that the reaction time would have been different if genomic DNA had been
used. Rapid (< 12 min) detection of 10 copies of the fungus B. cinerea DNA by fluorescencebased real-time LAMP was described but no information was provided about the fluorescent dye
and the concentration that was used (147).

68

3.3.5. Real-time LAMP with compact and inexpensive optics
One of the key motivations for optimizing the reaction and systematically testing
fluoresent dyes was to find a combination that would allow the use of simple, low cost, and offthe-shelf optical components, without sacrificing the sensitivity, robustness, and limit of
detection obtained with commercially available real-time platforms. Only the combined use of
LAMP, with high amplification yields, and a high-concentration of non-inhibiting intercalating
dye permited detection of an amplification profile with such components (84). Tests were
®

performed with both SYBR Green I and a varying concentration of SYTO-81 to demonstrate
the utility of this dye over more conventional dyes for detection with this simple optical setup.
Results showed amplification curves when using SYTO-81 at 20 µM. In contrast no
®

amplification was possible using SYBR Green I (Figure 3.5) These results highlight that
®

TM

SYTO-81 is a better fit than SYBR Green I for the Gene-Z

device. This is likely due to the

possibility to use this dye at a high concentration, allowing readings even when using a device
with low-cost optical settings. This was further confirmed by using SYTO-81 at various
TM

concentration (2, 5, 10, and 20 µM) in the Gene-Z

). It was possible to detect an amplification

for every dye concentrations tested. However, when using SYTO-81 at 20 µM, the Tt was about
1 min shorter and less variable than when using any of the other concentration.

69

TM

Figure 3.5. Detection of M. tuberculosis with Gene-Z
TM

of M. tuberculosis genomic DNA in the Gene-Z

device. Amplification plots of 0.1 ng

device using SYBR green I (A) and SYTO-

81 (B). Resulting velocity plots (slope of 6 points) for SYBR green I (C) and SYTO-81 (D). A
polyester microfluidic chip of 64 wells of 2 µL was used for this experiment.

3.4. Conclusions
A number of strategies were used to increase the performance of real-time LAMP. The
evaluation of several fluorescent dyes revealed that SYTO-81 and 82 resulted in the fastest
amplification time, and the highest fluorescence signal and enhancement. In some cases,
preheating the template at 95°C for 1 min prior to amplification allowed a reduction in
amplification time and improvement in the limit of detection and the limit of quantification.
70

When the Bst polymerase concentration was doubled, the amplification time was slightly
reduced (by approximately 1 min) in comparison to the concentration generally used for LAMP
(0.16 units/µL). Using an optimized protocol (SYTO-82, 1-min heat denaturation of template,
and 0.32 units/µL of Bst), it was possible to amplify up to 10 copies of M. tuberculosis in less
than 15 min and 10 copies of S. enterica in less than 20 min. Our results that are typically faster
than previous published data (APPENDIX 5 and APPENDIX 6) also emphasize how the
optimization of the LAMP protocol may have a significant impact on the result in terms of time
and sensitivity. Finally our results demonstrate that SYTO-81 can be used to detect LAMP using
low-cost optical settings. The current findings are significant for developing real-time LAMP
assays and hand-held isothermal genetic analysis systems for low resource settings.

71

CHAPTER 4. FREEZE-DRYING OF REAL-TIME PCR AND LAMP REAGENTS FOR
FIELD DEPLOYABLE MICROFLUIDIC CHIPS

This chapter will be submitted for publication as: “Freeze-drying of real-time PCR and LAMP
mixtures: application to microfluidic chips”. By Grégoire Seyrig, Dieter M. Tourlousse, Robert
D. Stedtfeld, Onnop Srivannavit, Vikram S. Harichandran, Erdogan Gulari, James M. Tiedje, and
Syed A. Hashsham. Please note that initial work with freeze drying was carried out with silicon
wafer chips and PCR reagents. Because the data is useful for other applications, it is included.
The main device and system are now mostly using polyester chips.

4.1. Introduction
Freeze-drying is a well-established technique for stabilization of sensitive molecular
biology reagents including polymerases, antibodies, and various other enzymes. It is a critical
step in stabilizing reagent mixtures used for amplification of nucleic acids in the field. Klatzer et
al. stored freeze-dried PCR reagents at room temperature for up to a year with no loss of activity
(150). Similarly, freeze-dried TaqMan-based real-time PCR reagents have been stored at roomtemperature for up to six months with no loss of activity (151).
To avoid the denaturation of polymerase while stored at room-temperature in a dry
format, the amplification mixture must be stabilized using a non-reactive sugar (e.g., trehalose
(152)). Trehalose is a non-reducing disaccharide known to protect proteins and enzymes from
denaturation and/or oxidation while stored in dried state (152,153). Trehalose is also known to
enhance PCR efficiency of GC-rich targets (154). For freeze-drying, trehalose is generally used
at concentrations ranging from 5 to 15% (w/v)(150-152). However, studies documenting its
72

optimization are rare. In addition, even though freeze-drying has been used for well-established
nucleic acids amplification techniques (PCR, NASBA, and RPA), it is yet to be shown for the
reagents used in LAMP. LAMP having one of the greatest potentials for resource-limited settings
with no need for expensive equipment and complex sample processing can have the maximum
benefit from freeze-drying.
This study demonstrates the use of freeze-drying as a method to stabilize reagents used in
real-time PCR and LAMP, and also documents its performance on microfluidic chips. For both
amplification methods, an optimal concentration of trehalose has been evaluated based on
amplification efficiency of gene targets with varying GC content and copy number. Results
indicate that the efficiency of freeze-dried LAMP and PCR amplification mixture is not
influenced by primer characteristics and the freeze-dried mixtures can be stored for up to one
year with no loss of activity. Use of freeze-drying method in microfluidic chips made out of
either silicon or cyclo-olefin polymer (COP) had no influence on the efficiency.

4.2. Materials and Methods
4.2.1. DNA targets
DNA targets. Genomic DNA from 12 microbial pathogens was used as the target for the
PCRs. Cultures of Bacillus cereus, C. parvum, Helicobacter pylori, Legionella pneumophila
subsp. pneumophila, Listeria monocytogenes, V. cholerae, V. parahaemolyticus and Yersinia
enterocolitica subsp. enterocolitica type strains were obtained from the American Type Culture
Collection (ATCC, Manassas, VA), and grown according to the protocol provided. DNA was
extracted using a Promega Wizard DNA extraction kit (Promega, Madison, WI) (155). For
Campylobacter jejuni subsp. jejuni, Escherichia coli O157:H7, Staphylococcus aureus subsp.

73

aureus (deposited as Staphylococcus aureus Rosenbach) and S. enterica subsp. enterica serovar
typhimurium (deposited as S. typhimurium) only genomic DNA was directly obtained from the
ATCC.

4.2.2. PCR primers
Several sets of primers were designed to target virulence and marker genes from the 12
microbial pathogens [4]. The primers were designed using Primer Express (Applied Biosystems,
Foster City, CA) from aligned consensus. Alignments were originally made using Kodon
software (Applied Maths, Austin, TX). The specificity of the primers was checked against the
GeneBank database using NCBI BLAST. When available, primers described in the literature for
successful real-time PCR were also used (155). Overall, 33 primer pairs were used in this study
with amplicon length varying from 68 to 211 bases.

4.2.3. LAMP primers
Several sets of LAMP primers were designed to target virulence marker genes from V.
cholerae. The primers were designed using Primer Explorer (Eiken, Tokyo, Japan). The same
alignments that were used to design the PCR primers were used to design the LAMP primers.
Each primer set was composed of 3 pairs of primers, consisting of F3, B3, forward and backward
inner primers (FIP and BIP, respectively), and Loop-F and Loop-B primers.

74

4.2.4. Real-time PCR
Real-time PCR master mixes were prepared using SYBR

®

Green PCR core reagents

(Applied Biosystems) and consisted of 3 mM magnesium chloride, 1! SYBR PCR buffer, 800
µM deoxynucleoside triphosphate blend (with deoxy-uracyl triphosphate), 200 nM forward and
reverse primers, and 0.025 units/"L Amplitaq Gold® DNA polymerase. D-(+)-trehalose
dihydrate (Sigma-Aldrich, St. Louis, MO) was added to the master mixes at a final concentration
ranging from 0 to 25% (wt/vol). Reagents and samples were loaded into 384-well clear optical
reader plates (Applied Biosystems, Foster City, CA). Thermal cycling and fluorescence
monitoring were carried out in an ABI Prism 7900 HT sequence detection system (Applied
Biosystems). For all experiments, 3 "L of appropriately diluted genomic DNA was added to 7
"L of real-time PCR mix using a Biomek® 2000 laboratory automation workstation (Beckman
Coulter, Fullerton, CA). Negative controls were prepared with diethylpyrocarbonate (DEPC)treated and nuclease-free sterile water (Fischer Scientific, Pittsburgh, PA) instead of DNA. All
reactions were performed in triplicate. After initial polymerase activation at 95˚C for 10 min, 40
cycles of the following program was used for the PCR amplification: denaturation at 95˚C for 10
s, annealing at 58˚C for 45 s and elongation at 72˚C for 45 s.

4.2.5. Real-time LAMP
LAMP reactions consisted of 1.6 µM each of FIP and BIP primers, 0.2 µM each of F3
and B3 primers, 0.8 µM each of LF and LB primers, 0.8 M betaine (Sigma, St Louis, MO), 1.4
mM of each dNTP (Invitrogen Corporation, Carlsbad, CA), 20 mM Tris/HCl (pH 8.8), 10 mM
(NH4)2SO4, 10 mM KCl, 8 mM MgSO4, 8 mM Triton X-100, 0.2 to 2.4 units/mL of Bst DNA
polymerase, large fragment (New England BioLabs Inc., Ipswich, MA) and 20 µM SYTO-81
75

fluorescent dye (Invitrogen). D-(+)-trehalose dihydrate was added to the master mixes at a final
concentration ranging from 0 to 25% (wt/vol). For all experiment 1 "L of appropriately diluted
genomic DNA was added to 19 "L in 0.2-mL PCR tubes (VWR International, West Chester,
PA) and incubated at 63°C for 36 min in a Chromo4 Real-time PCR detector (Bio- with Channel
2 of the thermal cycler.

4.2.6. Silicon microchips fabrication
A layer of photoresist (PR 1813) was applied by spin-coating at 2000 rpm for 30 sec on a
silicon wafer. The wafer was soft-baked at 105 °C on a hot plate for 1 min and was then exposed
2

to UV radiation (404.7 nm, 10 mJ/cm , 30 sec) to define the channel pattern. After removal of
the activated photoresist by developer solution (MF319, Shipley, MA) for 1 min, the wafer was
hard baked at 115 °C on a hot plate for 2 min. The pattern was transferred to the silicon wafer by
deep reactive ion etching to obtain approximately 210 "m deep channel. The photoresist was
then removed by a resist stripper (PRS2000, Shipley, MA). The new layer of photoresist
(AZ9260, Clariant, NJ) was applied at the backside of the wafer by spin-coating at 2000 rpm for
30 sec. The wafer was soft-baked at 110 °C on the hot plate for 2 min 30 sec and exposed to UV
2

radiation (404.7 nm, 10 mJ/cm , 90 sec) to define the inlet/outlet holes. The activated photoresist
was then removed by a resist developer (AZ400K, Clariant, NJ, H2O, 1:3) for 2 min. The wafer
was then loaded into the deep reactive ion etcher to create the through inlet/outlet holes. The
silicon wafer, containing etched structures, was cleaned. A 0.2-micron thick thermal oxide was
then grown on the wafer. Finally, the silicon wafer was anodically bonded with the glass wafer
(Figure 4.1).

76

4.2.7. Real-time PCR in silicon microchips
On-chip real-time PCR was carried out in an OSA™ platform (IMSTAR S.A., Paris,
France) associated to an Eclipse 50i fluoro microscope (Nikon Instruments Inc., Melville, NY).
The on-chip real-time PCR mix was prepared as described above, with 15% trehalose. In all
cases, 1.5 "L of genomic DNA was added to 3.5 "L of PCR mix. After an initial polymerase
activation at 95˚C for 10 min, 30 cycles of the following program were used for the PCR
amplification: denaturation at 95˚C for 15 s, annealing at 58˚C for 45 s and elongation at 72˚C
for 45 s.

4.2.8. Fabrication of COP chips
The cyclo olefin polymer (COP) chip used for real-time LAMP consisted of an array of
eight reaction wells (2.5 µL each) and were manufactured out of 180 µm ZeonorFilm# (ZF14180, Zeon Chemicals, Louisville, KY) by hot embossing as previously described (23). Briefly,
the ZeonorFilm# was sandwiched in a heated (150°C) press (Carver, Wabash, IN) between an
embossing mold (FineLine Prototyping, Raleigh, NC) and an acrylonitrile butadiene styrene
(ABS) counter mold (K-mac Plastics, Grand Rapid, MI) at a pressure of 1000 kg for 5 min. The
chips were then laminated with MicroAmp® Optical Adhesive Film (Applied Biosystems;
Foster City, CA) in which 1 mm sample-dispensing-ports and cross-shaped air-vents were
previously patterned using a CraftROBOT Pro (model CE5000-40-CRP, Graphtec America Inc.,
Santa Ana, CA). After dispensing the LAMP reagents (including the target DNA) in the wells,
the chip was sealed by lamination with MicroAmp® Optical Adhesive Film (Figure 4.1.B.)

77

4.2.9. Real-time LAMP in COP microchips
Real-time LAMP was performed in a custom multiplex hand-held gene analyzer. Briefly,
the chip was incubated at 63°C with a Kapton heater controlled via LabVIEW (National
Instrument, Austin, TX) using a proportional-integral-derivative algorithm (PID) and a pulse
width modulator (PWM) as described elsewhere (24). Fluorophores were excited using 532 nm
emitting diodes (Super Bright LEDs Inc. St. Louis MO). Optical fibers (1 mm diameter,
Industrial Fiber Optics, Tempe, AZ) were used to transmit the fluorescence from each reaction
wells to an ODA-6WB-500M photodiode (OptoDiode Corp., Newbury, CA). Voltage data
measured by the photodiode were acquired by a LabVIEW USB 6009 data acquisition (DAQ)
system.

Figure 4.1. Microfluidic chips used in this study. Silicon 3-well chip (A), and COP 8-well (B,
Note: The channels contain “sky blue spectrum” liquid food colorant (Ateco, Glen Cove, NY)
for better visualization).

78

4.2.10. Stabilization and conservation of real-time PCR reagents.
Real-time PCR mixes were freeze-dried in 15-mL centrifuge tubes (Denville Scientific
Inc., Metuchen, NJ) according to the following protocol. The real-time PCR mixture was divided
into aliquots, which were rapidly frozen by submersion in liquid nitrogen for 1 min. The aliquots
were then freeze-dried for 16 h in a Freezemobile 12ES (SP Industries, Gardiner, NY). The
freeze-dried real-time PCR mixes were then stored at room temperature, in the dark, in a
Nalgene® transparent polycarbonate vacuum-desiccator (Thermo Fisher Scientific Inc.,
Waltham, MA) for periods ranging from 1 to 365 d. Each vial contained sufficient master mix
for 30 real-time PCR reactions and the content was reconstituted by addition target-containing
ddH2O up to a final volume of 280 "L. As a control, some of the aliquots were not freeze-dried
and were conserved at room temperature in the dark.
For the silicon microfluidic PCR chips, reagents (4 "L) were loaded into the
microchannels, the chip frozen at -80˚C, freeze-dried overnight, and stored in the darkness at
room temperature, in a vacuum desiccator for 1 to 30 d. To avoid thawing of the frozen master
mix during transfer of the microfluidic chips from the freezer to the freeze-drier, the chips were
placed on an aluminum block during freezing, transfer and freeze-drying as previously described
(156). At the time of use, the reagents were reconstituted by addition of ddH2O up to a final
volume of 4 µL.

4.2.11. Stabilization of real-time LAMP reagents
Real-time LAMP reagents were freeze-dried in 15-mL centrifuge tubes using a similar
protocol as described above. Briefly, the tubes containing 110 "L of LAMP mix were submerged
in liquid nitrogen for one min and then freeze-dried for 16 h. Each vial contained sufficient
79

reagents for 5 real-time LAMP reactions and the content was reconstituted by addition of 95 "L
of water. The rehydrated reagents were then separated in 19 "L aliquots. To each aliquot, 1 "L
of water, containing 0.1 ng of the DNA target, or no target, was added.
For the Zeonor$ microfluidic LAMP chips, the surface of each reaction well was first
modified by photo-polymerization of acryl polymers using a previously described method (157).
Each reaction well was loaded with 2.5 "L of a photochemical modification solution, containing
10% acrylamide, 1% acetone and 0.1% benzophenone (Sigma), placed under a high performance
UV transilluminator (UVP LLC, Upland, CA) for 20 min, rinsed three times with ddH2O and
dry at 50°C for 10 min. Reagents (2.5 "L) were then loaded into each well, the chip frozen at 80˚C, placed in a freeze-drier overnight, and stored for 1 d in the darkness, at room temperature,
in a vacuum desiccators. To avoid thawing of the frozen master mix during transfer of the
microfluidic chips from the freezer to the freeze-drier, the chips were kept in the sacrificial ABS
counter mold that was placed on an aluminum block during freezing, transfer and freeze-drying
(156). After storage, the chip was laminated with MicroAmp® Optical Adhesive Film in which 1
mm sample-dispensing-ports and cross-shaped air-vent was previously patterned using the same
protocol as described above. The reagents were then reconstituted by addition of a final volume
of up to 2.5 "L of ddH2O containing the target DNA and the chip was then sealed with optical
adhesive.

4.3. Results and Discussion
4.3.1. Effect of trehalose concentration
The effect of trehalose concentration on the stability of freeze-dried real-time PCR
®

reagents was evaluated with a SYBR

Green-based real-time assay targeting C. jejuni. The
80

estimated copy number determined using freeze-dried reagents was evaluated as a function of the
trehalose concentration for several spiked copy numbers (Figure 4.2.A) For trehalose
concentrations of 5%, 10%, and 15%, the estimated copy numbers were in good agreement with
the spiked copy number. At 20% trehalose, the estimated copy number decreased very slightly
but significantly (P < 0.0001) for the assays that were spiked with 100 or 1000 copies. This trend
was not observed for the assay targeting 10 copies of target since the freeze-dried mix containing
20% trehalose yielded similar result than the ones containing 10 or 15%. In the absence of
trehalose, no amplification was observed. At a trehalose concentration of 25%, no amplification
was observed, which may have been due to inhibition (data not shown). The effect of trehalose
on PCR efficiency was also evaluated on non-freeze-dried control reagents. The spiked copy
number was consistent with the spiked copy number for trehalose concentrations %20%. Above
this concentration the estimated copy number decreased significantly (data not shown).
The effect of trehalose concentration was evaluated on real-time LAMP with a set of
primers targeting V. cholerae. The Tt to detect a constant amount (0.1 ng) of V. cholerae
genomic DNA was evaluated as a function of the trehalose concentration for freeze-dried and
non-freeze-dried real-time LAMP reagents (Figure 4.2.B). Amplification occurred only for
samples that were freeze-dried in the presence of trehalose. The Tt of samples that were freezedried with 2.5 and 5 % trehalose ranged between 19 and 22 min, and were not significantly
different from each other due to a large variability (p = 0.2). In contrast, samples that were
freeze-dried with 10% trehalose resulted in amplification with the shortest Tt. Reagents that were
freeze-dried with higher concentrations of trehalose (15% and more) resulted in amplification
with either higher Tt values or fully inhibited amplification for trehalose concentration of 20% or

81

more. For control reagents that were not freeze-dried, the Tt increased linearly with the trehalose
2

concentration (R = 96%) revealing a concentration-dependent inhibiting effect on LAMP. It
was not possible to obtain any amplification for trehalose concentrations equal or higher than
20%. Simply drying (rather than freeze-drying) the PCR reagents in the same carbohydrate
matrix led to loss of enzyme activity after reconstitution (data not shown).

Figure 4.2. Effect of trehalose on freeze-dried reagents. (A) Effect on the estimated copy
number of freeze-dried real-time PCR mixes for various target (C. jejuni) spiked copy number:
10 copies (circles), 100 copies (triangles) and 1,000 copies (squares). The data shown are mean
values for six replicates reactions and the standard errors bars representing the intra experiment
variation are shown but are generally smaller than the symbols. ECN: estimated copy number.
(B) Effect of trehalose concentration on the Tt of freeze-dried (squares) and non-freeze-dried
(triangles) real-time LAMP reagents. The data shown are mean values for six replicates reactions
and the standard errors bars represent the intra experiment variation.

82

These results demonstrate that trehalose is essential for preventing real-time PCR or realtime LAMP mix activity loss during drying. Trehalose is most likely involved in the stabilization
of the DNA polymerase (152,153,158) although the effect of trehalose was not specifically tested
on the enzyme but on the whole PCR master mix. The results also suggest the need to use
trehalose at a concentration ranging between 10 and 15% for PCR and 10% for LAMP. This
trend is in agreement with the results of Colaco et al., who reported an optimum trehalose
concentration of 15% for the stabilization and storage of a modified T7 DNA- polymerase (152).
Other studies reported trehalose concentrations between 5 and 10% for the freeze-drying-based
stabilization of PCR and Taqman-based real-time PCR (150,151). However, the trehalose
concentration was not optimized in those studies. Despite the fact that the actual results
suggested that trehalose did not have a significant inhibitory effect on the PCR when its
concentration was equal or under 15%, enhancement of PCR as previously described by Spiess
et al. (154), was not observed. In contrast, LAMP was inhibited by trehalose on a concentrationdependent manner, and more particularly for concentrations above 5 %. This inhibiting effect
was even more important for samples that were freeze-dried which might be due to the presence
of too much glycerol in the mixture. Many polymerases, including Bst, are provided in pure
glycerol that is used as a cryoprotectant. In the absence of sufficient amounts of trehalose, it is
possible that the glycerol prevents the reagents from completely freezing, altering the freezedrying process.

4.3.2. Effect of storage time
The effect of storage time on the efficiency of freeze-dried reagents was evaluated with
real-time PCR mixes (containing primers for C. jejuni) that were stored for various periods of

83

time. The estimated copy number was measured as a function of storage time for different spiked
copy numbers (Figure 4.3). The estimated copy number obtained with control (i.e., not freezedried) real-time PCR mixes decreased rapidly with prolonged storage, and no amplification was
observed after 90 days of storage. In contrast, the estimated copy number measured using freezedried real-time PCR mixes remained stable up to one year for all spiked copy numbers. The
small difference in estimated copy number at different storage times was attributed to the
variability of the target dilution, as the curves from freeze-dried and controls followed similar
patterns in terms of variability.

Figure 4.3. Effect of storage time on freeze-dried (filled shapes) and non-freeze-dried (empty
shapes, reagents were stored at room temperature) real-time PCR mixes for various spiked target
(C. jejuni) copy numbers: 10 copies (circles), 100 copies (triangles) and 1,000 copies (squares).
ECN: estimated copy number. The data shown are mean values for six replicates reactions
(experiment was conducted twice with triplicates).
84

Overall, these results demonstrate that freeze-dried real-time PCR mixes can be stored at
room temperature (~20°C) for a period of up to one year without loss of activity. After this
length of storage time at room temperature, the freeze-dried real-time PCR reagents were as
effective as if they were fresh. For this reason, it would seem that reagents could be stored at
room temperature for even longer periods. However, it is possible that the stability of real-time
PCR mixes may be affected if they are stored at higher temperatures. Klatzer et al. (1998)
showed that freeze-dried PCR reagents stored at 37°C were only stable for 3 months. In contrast,
those that were stored at room temperature were stable for a year. This effect of storage
temperature should be taken in consideration if the freeze-dried reagents have to be stored in
unusually warmer conditions than in the laboratory. Although the effect of storage time on the
efficiency of freeze-dried LAMP reagent is under investigation and will be reported separately.
Considering that Bst polymerase is less sensitive to inhibitory materials and is more robust, we
expect that LAMP reagents may survive storage after freeze drying equally well, if not better
than PCR and real-time PCR reagents.

4.3.3. Effect of primers
The effect of primer compositions on the efficiency of freeze dry reagent was evaluated
using real-time PCR. For this manner, assays targeting various organisms were freeze-dried and
stored for 15 days in the same conditions as described above. The minimum copy detected was
measured as a function of the PCR efficiency (155) for freeze-dried and non-freeze-dried
reagents prepared the day of the freeze-dried mixture rehydration (Table 4.1). The efficiency of
non-freeze-dried real-time PCR mixes varied between 78.8 and 111 % for the different primer

85

sets, with an average of 88.9 % and a standard error of 1.32 %. The efficiency of freeze-dried
real-time PCR mixes varied between 79.0 and 103 %, with an average of 89.9 % and a standard
error of 1.06 %; this was not significantly different than the non-freeze-dried reagents. In
addition, irrespective of GC content or amplicon length, freeze-drying had so significant effect
on PCR efficiency.
These results confirmed the hypothesis that the nature of the PCR primers should not
have a significant effect on the stability of freeze-dried PCR reagents. This was somewhat
expected since many companies ship synthesized oligonucleotides that have been dried. Based
on the assumption that freeze-drying has the same consequence as drying on the
oligonucleotides, we expected the same absence of inhibiting effect while freeze-drying PCR
reagent.

4.3.4. Freeze-drying PCR mixes in microfluidic chip
Real-time PCR mixes were freeze-dried in silicon microchips. After rehydration with
water containing the DNA template or not, the chips were sealed and thermocycled on the
IMSTAR platform where the fluorescence intensity was measured as a function of the cycle
number (Figure 4.4.A). An amplification curve was observed for both freeze-dried and nonfreeze-dried PCR mixes. These amplification curves of freeze-dried and non-freeze-dried
reagents had similar patterns and positions with cycle threshold values of 10.0 and 10.8,
respectively. No amplification curve was observed for the negative control (with no template
DNA).

86

Table 4.1. Effect of freeze-drying on PCR amplification efficiency (FD: freeze-dried; NFD: non-freeze-dried).
Organism (ATCC No.)

Primer pair name

Amplicon
GC %

Efficiency (%)
FD / Control

Bacillus_sp_cerA
Bacillus_sp_clo

Amplicon
length
(bp)
101
101

B. cereus (10987)

38
40

95.5 / 93.2 +
92.3 / 110.7 -

C. jejuni subsp. jejuni
(700699)

Campylobacter_jejuni_hipO

129

34.0

103.2 / 84.4 +

Campylobacter_jejuni_cdtB
Campylobacter_jejuni_cdtC
Campylobacter_jejuni_gyrA_Fukushima et al
Campylobacter_jejuni_racR
Campylobacter_jujuni_mapA_Inglis et al

73
120
192
84
94

49.0
35.0
38.6
40.0
35.4

84.2 / 84.2 =
89.5 / 80.7 +
96.1 / 81.5 +
100.0 / 88.6 +
83.3 / 85.0 -

C. parvum (PRA-67D)

Cryptosporidium_parvum_CP23GI,GII
Cryptosporidium_parvum_hsp70GI,GII

90
76

48.0
46.0

98.1 / 87.1 +
97.3 / 102.4 -

E. coli O157:H7 (19585)

Escherichia_coli_stx1_Jinneman et al
Escherichia_coli_uidA

128
145

44.6
57.0

88.3 / 86.5 87.4 / 88.1 -

H. pylori (700392)

Helicobacter_pylori_ureA
Helicobacter_pylori_virD4

134
143

42.0
37.0

97.8 / 92.1 +
80.1 / 88.3 -

L. pneumophila subsp.
pneumophila
(33152)

Legionella pneumophila_lepB

128

41.0

96.9 / 94.0 +

Legionella_pneumophila_mip

70

49.0

88.6 / 97.9 -

L. monocytogenes (3931D)

Listeria monocytogenes_actA
Listeria monocytogenes_hlyA_prfA_Rudi et al
Listeria monocytogenes_inlA

211
112
93
72

39.0
37.2
41.0
51.0

83.6 / 78.8 +
79.0 / 79.9 90.8 / 84.1 +
85.4 / 82.9 +

Listeria monocytogenes_mpl
87

Table 4.1 (cont’d)
S. enterica subsp. enterica
(700720D)

Salmonella_fimA
Salmonella_fljB
Salmonella_invA_Hoorfar et al

76
116
119

59.0
43.0
43.8

83.5 / 97.5 92.1 / 103.3 86.6 / 80.4 +

S. aureus subsp. aureus
(700699)

Staphylococcus aureus_seC
Staphylococcus aureus_tsst1

131
88

29.0
36.0

91.4 / 83.3 +
88.5 / 85.6 +

V. cholerae (43996)

Vibrio_ cholera_ctxA
Vibrio_ cholera_ctxB
Vibrio_ cholera_ctpA
Vibrio_ cholera_zot

133
112
74
68

45.0
36
53
53

84.7 / 86.3 87.2 / 84.7 +
90.8 / 83.5 +
93.5 / 91.7 +

V. parahaemolyticus (55075)

Vibrio parahaemolyticus_omp_U
Vibrio parahaemolyticus_toxR

75
71

52
49

91.1 / 93.7 87.2 / 91.0 -

Y. enterocolitica subsp.
enterocolitica (55075)

Yersinia enterocolitica_ystsA

76

46.0

83.4 / 92.0 -

88

These results demonstrate that real-time PCR mixes can also be freeze-dried in
microfluidic chips. This is of particular interest since such chips may be used in portable devices
for detecting microbes in the field (159). In 2004, the group of Paul Yager showed that
functional proteins might be stabilized on a dry format in polydimethyl siloxane (PDMS)
microfluidic devices. However in this case the proteins were dried at 55°C in the presence of
trehalose or dextran. Trough modeling studies, these authors demonstrated that reconstitution of
the dried protein matrix was easily controlled, thus facilitating long-term storage, portability, and
ease of use (159).

4.3.5. Freeze-drying real-time LAMP reagents in COP microfluidic chips
The possibility to amplify DNA from on-chip freeze-dried LAMP reagents was
investigated using a primer set targeting V. cholerae and a COP microfluidic chip. It was
possible to obtain an amplification curves from the freeze-dried reagents in the photo-chemically
modified COP chip (Figure 4.4.B). This amplification curve had similar characteristics (e.g., Tt,
amplitude) than the one obtained from fresh non-freeze-dried reagents. In contrast it was not
possible to obtain any amplification from reagents that were freeze-dried in microfluidic chips
that were not photo-chemically modified. However, for many channel, it was not possible to
rehydrate the freeze-dried mixture because one of the outlet of the chamber was clogged with the
freeze-dried mix.

89

Figure 4.4. Amplification profiles of freeze-dried PCR and real-time LAMP mixes in
microfluidic chip. (A) Real-time PCR in silicon microchips. The curves were flattered by
removing the background fluorescence (average fluorescence of the 8 first cycles) for each assay.
The threshold fluorescence was defined as the average background fluorescence of the 8 first
cycles plus 5 times the standard deviation for this average. Freeze-dried reagents (closed circles),
not freeze-dried control reagents (closed squares), freeze-dried reagents, no target (open circles).
(B) Real-time LAMP in Zeonor® chips. The curves were flattered by removing the background
fluorescence (average fluorescence of the 6 first min) independently for each assay. These curves
are four replicates and represent the amplification profile of reaction chambers containing freezedried reagents only. It was possible to rehydrate only half of the chambers.

These results illustrate the performance of freeze-dried LAMP reagents in COP
microfluidic chips. However, the surface of these chips requires photo-chemical modification
prior the freeze-drying, which is likely due to the hydrophobic nature of the COP surface. This
phenomenon was already described by Furuberg et al. who dried NASBA reagents in cyclo-

90

olefin copolymer (COC). These authors described the impossibility to recover functional
amplification reagents from chip that were not coated with a hydrophilic layer, likely because of
the adsorption of the enzyme on the hydrophobic walls of the wells (160). Because, of the poor
efficiency to rehydrate freeze-dried mixtures in the COP microfluidic chip, and the high
efficiency to rehydrate the freeze-dried reagents in tubes, we believe that for microfluidic
applications, it would be more simple to keep the reagents freeze-dried in a tube. Similar
problems should be expected for chips that are made with other materials (e.g., polyester)
because clogging is due to the foaming of the reagents in the 100-500 !m capillaries.

4.4. Conclusions
The present study demonstrated that real-time PCR and LAMP reagents could be freezedried and stored at room temperature for a period up to one year with no loss of performance.
The technique requires the use of trehalose sugar at concentrations ranging between 10 and 15%
to protect the reagents (most likely the polymerase) from degradation while stored in a dry state.
The nature of the primers does not seem to affect the efficiency of freeze-dried reagent although
this was only tested on PCR reagents. The technique can also be applied to store nucleic acids
amplification reagents at room temperature in microfluidic channels made of various material
such as silica or COP. This is of particular interest since the need for microfluidic-based tests for
POC diagnostic of infectious diseases and detection of microbes in the environment is increasing
as fast as the technologies are appearing. Storing PCR or LAMP reagents on a freeze-dry state
also facilitates the easiness to use of the amplification techniques as the reagents only require to
be rehydrated with target-containing water. This avoids the sequential addition of each PCR

91

reagent by multiple pipetting steps, which can reduce the precision and the easiness of the
experiment.

92

CHAPTER 5. DEVELOPMENT OF LOW-COST MICROFLUIDIC CHIPS FOR THE
TM

GENE-Z

PLATFORM

Figure 5.3, 5.4, 5.5, and 5.6 adapted from Stedtfeld, R., Tourlousse, D., Seyrig, G.,
Stedtfeld, T., Kronlein, M., Price, S., Ahmad, F., Gulari, E., Tiedje, J. and Hashsham, S. (2012)
Gene-Z: A Device for Point of Care Genetic Testing using a Smartphone. Lab on a Chip, 12,
1454-1462

(79)

with

authorization

from

the

Royal

Society

of

Chemistry.

http://pubs.rsc.org/en/content/articlelanding/2012/LC/C2LC21226A

5.1. Introduction
TM

The Gene-Z

platform is a fully functional and packaged POC device that was

developed by our group for the rapid (<30min), quantitative and multiplex detection of nucleic
acids signature sequences with high sensitivity and specificity (78). The platform is made of four
main components that are (i) a quantitative isothermal nucleic acids amplification method, (ii) a
64-well microfluidic chip, (iii) a compact an inexpensive fluorescence detection device, and (iv)
an iPod Touch application to control the system wirelessly. The isothermal nucleic acid
amplification and its development were described in detail in previous chapters of this
dissertation, and the development of the fluorescence detection device, the iPod application, and
the microfluidic chip were previously thoroughly described by other members of our group
(23,24,84).
The initial chip is an array of 64 reaction wells that was prepared by hot embossing and
lamination. The chip allows amplifying several pathogens with a high sensitivity, in a short
amount of time (~30 min) (23). However, this chip has two big limitations. First, it uses 128

93

microvalves to enclose independently each reaction well after the chip is filled with the sample.
Not only these micro-valves require three levels of lamination, they also require a very strong
and complex pressing system inside the Gene-Z device to close the micro-valves efficiently.
Second. the base of the chip is fabricated out of cyclo-olefin polymer (COP), which was initially
chosen because of its high transparency, good moldability, low water absorption (<0.01%), and
good chemical resistance. Despite its good characteristics for the fabrication and use of
microfluidic chips, COP is an expensive (~$4 per chip) material that is also very difficult to
purchase, as orders may take up to several months. Because the microfluidic chip for the Gene-Z
has to be inexpensive (less than $1) and easy to fabricate, it is critical to develop a valve-less
chip using a less expensive polymer.
This chapter describes the development of an inexpensive microfluidic chip for use in the
Gene-Z platform, and that is easy to use and fabricate. This new chip is valve-less and was made
out of inexpensive polyester. This new chip was successfully evaluated for the detection of up to
four target genes by real-time multiplex LAMP in the Gene-Z platform.

5.2. Materials and Methods
5.2.1. Chip design and fabrication
The new microfluidic chip was made based on the previous design developed in our
group (23) with the main differences that (i) the is valve-less, and (ii) it is made of polyester, an
common polymer that is roughly 35-fold less expensive than COP. Briefly, the chip is 85 x 65
mm and consists of four arrays of 16 LAMP reaction wells of about 1 µL each. In each array,
every reaction well is connected by a 250 x 250 µm micro-channel to an air-venting port on one
side, and to a sample distribution channel on the other side. The sample distribution channel is

94

connected to a 1.5-mm loading port. The air-venting ports of each array are connected by a
depression where a hydrophobic membrane can be placed during fabrication (Figure 5.1). The
filling volume of each array is about 40 µL, of which about 60% remains in the micro-channel or
below the hydrophobic membrane.

Figure 5.1. Schematic of the 64-well microfluidic chip. This new chip design allows parallel
analysis of up to four samples for up to 16 different target genes.

The features of the chip (e.g. microfluidic channels, reaction wells) were prepared by hot
embossing of a 127-µm thick moisture-resistant polyester film (McMaster Carr, Princeton, NJ )
with using an aluminum mold, and a reusable 6.35-mm thick FDA-compliant silicone rubber
(durometer 60A, McMaster Carr) as counter-tool (Figure 5.2). The mold was made in aluminum
alloy 7075-T651 by computer numerical control (CNC) assisted machining (Proto Labs, Inc.,
95

Mapple Plain, MN). A landing shoulder was placed at each side of the mold to keep the rubber
from flowing laterally too excessively. The hot embossing is performed by sandwiching the
polyester film between the aluminum mold and the rubber counter-tool, and by placing the
sandwich between the heated plates of a hydraulic lab press (Model 4386, Carver Inc., Wabash,
IN). The time the system reached a temperature of 218°C, a pressure of 230 kg was applied; the
pressure was then elevated to 5,500 kg and kept as such for one min. The system was then cooled
down to 93°C by turning on the hydraulic cooling system of the press. Finally, the pressure was
released, the sandwich was removed from the heated platens, and the chip was detached from the
mold.
To enclose the microchannels and the reaction wells, the chip was laminated with a layer
®

of optical film (MicroAmp Optical Adhesive Film, Life Technologies, Carlsbad, CA). Prior to
lamination, the film was patterned with 1.5-mm sample filling ports (one for each array, 4 total)
and 1-mm air-venting ports (16 per array, 64 total) by xurography using a commercial-grade
knife plotter (CraftROBOT Pro CE-CRP, Graphtec, Irvine, CA) with a 45°, 0.9 mm blade, and a
cutting quality of 3. For cutting, the optical film was placed on a carrying sheet, liner facing up,
and the blade z-axis was manually adjusted so that it cuts through the whole film. After cutting,
the unnecessary materials (i.e., the parts of the film that have been cut) were removed using a
transfer tape (JVCC DC-1114 Double Coated Film Tape, FindTape.com, Skillman, NJ), leaving
through holes in the film. Hydrophobic membranes were placed between the optical film and the
embossed polyester sheet, at the position of the depression that was designed for this matter. The
hydrophobic membranes were also patterned by xurography (see APPENDIX 7 for all the
cutting patterns). The patterned film was aligned on top of the chip manually (Figure 5.3.A). To
obtain an even seal, the chip was sandwiched between a layer of FDA-compliant silicone rubber
96

(positioned on the side of the micro-channel features) and a steel sheet (on the side of the optical
film), and the sandwich is placed in the press with a pressure of 1,800 kg, for 30 s (room
temperature). Each four sides of the chip were sealed using a manual impulse bag sealer (Omcan
Food Machinery, Niagara Falls, NY) to insure a hermetic sealing of the chip. After filling the
chip with the LAMP reagents, it was laminated with an un-patterned layer of optical tape to seal
the air-vents and the sample filling ports.
The filling of the chip is made using a conventional pipette with a 200-µL tip, the pointed
end of which fits snugly the filling port. Alternatively, a 2-mm thick polycarbonate filling band
with through holes (Figure 5.3.B) were aligned and attached to the chip at the level of the filling
port. This alternative method requires to place the filling band on top of the chip, to align it with
the sample filling ports, and to stick it using some JVCC DC-1114 Double Coated Film Tape
This extra step also requires that the filling band is detached from the chip prior sealing the chip
with the second layer of optical tape.

Figure 5.2. Schematic of the polyester chip embossing using a hydraulic press.

97

5.2.2. Real-time LAMP
The evaluation of the chip was made using four sets of LAMP primers targeting eaeA and
stx2 genes for E. coli, and vickR and mecA genes for S. aureus (APPENDIX 8). Primers were
dried in each reaction well in order to yield a working concentration of 1.6 µM each of FIP and
BIP, 0.2 µM each of F3 and B3, and 0.8 µM each of LF and LB primers. Each array of the chip
is loaded with a mixture containing the target DNA (E. coli ATCC 19585, or S. aureus ATCC
700699D-5), 800 mM betaine, 1.4 mM of each dNTPs, 20 mM Tris/HCl (pH 8.8), 10 mM
(NH4)2SO4, 10 mM KCl, 8 mM MgSO4, 8 mM Triton X-100, 0.64 units/µL of Bst DNA
polymerase-large fragment, 20 µM SYTO-81, 0.2% Pluronic F-68 (Life Technologies), 1 mg/mL
of bovine serum albumin (BSA; New England Biolabs, Beverly, MA).

5.3. Results and Discussion
5.3.1. Chip design and fabrication
The new microfluidic chip allowed the analysis of four samples, and up to 16 different
nucleic acids signature sequences. The chip was much easier to fabricate than the previous
design and is made of polyester, which is roughly 35-fold less expensive than COP, and also
much easier to purchase. Because the chip does not contain any valves, only three layers were
needed for the fabrication (the embossed polyester sheet, the laminating patterned film, and the
sealing film,) versus four layers for the previous valve-containing COP chip. Moreover, only one
layer of patterned film had to be precisely aligned, which further facilitated the fabrication
process. Additionally, the absence of valve allowed saving another fabrication step to remove
locally the adhesive from the optical film. Because the new chip design include a depression for

98

the hydrophobic membrane, this critical component was easier to incorporate during the
fabrication of the chip, accelerating again the manufacturing process.
A

B

Figure 5.3. (A) Schematic of the chip lamination. Figure and legend reproduced from reference
(84) . (B) Schematics of the chip filling using a detachable filling band.

99

The fabrication of the chip took roughly 20 min and was performed by at least two other
members of the lab with minimum training. A standard operating protocol was designed for this
purpose and is now used in our lab (APPENDIX 9).

5.3.2. Evaluation of possible cross-contamination

Figure 5.4. Fluorescence image of a single 15-well array after amplification to demonstrate lack
of cross-contamination between reaction wells in the absence of valves. For this experiment,
5

primers for the eaeA gene of E. coli were dehydrated in 12 wells, with 1.7 " 10 genome copies
of E. coli added to alternating wells (wells marked with ‘-’ and ‘+’ contain primers only and
primers plus DNA, respectively; ‘*’ indicates wells without dehydrated DNA and primers) (84).

Since the new chip design did not contain any valves, cross contamination was possible
between connected wells. Cross contamination may be due to the migration of the amplified
product, the target DNA, or the primers, from a reaction well to another. Because the chip was
design to have relatively long channel paths between the reaction wells, we hypothesized that
cross contamination was not likely to happen. This hypothesis was tested by injecting primerand DNA-free LAMP reagents in a chip that contained (i) dried primers with target DNA, (ii)
only dried primer, and (iii) DNA- and primer-free wells. After one hour of incubation in the
100

Gene-Z, an amplification signal was observed only in the reaction wells that contained both
primers and target DNA (Figure 5.4) One of the primer- and DNA- free reaction-well showed a
non-negligible signal but this signal was not considered as amplification because it stayed
constant during the whole incubation (data not shown). These results confirm the absence of
cross-contamination between reaction wells, which enable to circumvent the use of micro-valves
in microfluidic chips (161,162). In order to reduce further the size of valve-less microfluidic
chips, more investigation related to the effect of inter-well distance and micro-channel diameter
on cross-contamination, would be helpful.
5.3.3. Reproducibility
In order to reach a reliable and sensitive quantification in the Gene-Z, it is critical to
evaluate inter- and intra-chip variability. This was tested using as set of primers (for stx2 of E.
coli) that was previously dried in each well of three different chips (for a total of 60 replicates
per chip). Each array of every chip was filled with LAMP reagents containing a constant amount
5

of target DNA (1.7 x 10 genomes per µL). Based on the measurement of the Tt, we observed
high intra-chip reproducibility (p<0.05) for each of the three chips tested, with coefficients of
variation (CV) of 5.2%, 5.2% and 6.9% respectively (Figure 5.5.A and B). There was no
significant difference (p<0.05) in Tt between the three chips tested (8.6 +/- 0.7 min and CV of
7.8%), revealing a good inter-chip reproducibility.

5.3.4. Quantification and sensitivity
The sensitivity and quantification ability of the chip in the Gene-Z platform was
evaluated using a set of primers targeting the eaeA gene of E. coli and a10-fold dilution series of

101

E. coli genomic DNA (Figure 5.5.C and D). Because one chip only contains four arrays (i.e.,
only four different samples may be evaluated per chip), a second chip was used to evaluate a
wider range of target concentrations. Using these two chips, it was possible to detect as low as 13
copies per reaction well with a Tt of about 21 min. However at this target concentration, only 3
out of 15 channels showed amplification in the chip. In contrast it was not possible to detect this
amount of target using a conventional thermocycler, likely due to a lack of replicates (three).
Above this target concentration, amplification happened in all wells, as well as in the all the
tubes in the conventional thermocycler. In both cases, Tt seemed to decrease exponentially with
2

the increasing target copy number (R = 0.9778) showing good potential for quantification. It is
expected that a lower limit of detection may be achieved using a different set of primers, as
previously reported (CHAPTER 2).
These results suggest that the Gene-Z platform has an sensitivity that is comparable to
conventional real-time thermocyclers, which is in contradiction with previously published results
where sensitivity was better in a conventional thermocycler (163). The improved sensitivity of
our microfluidic chip may be due to the use of a surfactant (Pluronic F-68) in addition to the
conventional surface passivation agent (BSA), to reduce inhibitory effects associated with high
surface-to-volume ratio within the microfluidic channels. The difference in sensitivity may also
be due a variability of the primer efficiency. Extra experiments performed in our lab showed that
some primer sets gave better results in the Gene-Z while other primer sets performed better in a
conventional thermocycler (data not shown). Another possible explanation to this difference of
performance might be the primer concentration.

102

Figure 5.5. Evaluation of inter- and intra-chip reproducibility. Three separate chips were
prepared with primers for the stx2 gene of E. coli dehydrated in all reaction wells. For the
5

experiment, samples contained LAMP reagents supplemented with 1.7 " 10 genome copies per
well, and the assays was monitored in real-time in the Gene-Z. (A) Amplification plots for all
reaction wells in a single representative chip. (B) Mean, standard deviation of the Tt for all
reaction wells in each of the four arrays in three separate chips. (C) Amplification plots for a 10fold serially diluted E. coli genomic DNA sample using the eaeA gene primer set obtained using
the Gene-Z. (D) Corresponding standard curve as determined using the Gene-Z (filled symbols)
and the real-time PCR instrument using tubes (empty symbols). Data represent the mean and
standard deviation of 3 determinations for the commercial real-time PCR instrument and 3 to 15
determinations for the Gene-Z (84).
103

Since the primers are dried only in the reaction wells of the chip, it is possible that their
concentration varies a little from the theoretical/expected one, possibly due to a slight diffusion
of the primers in the channels neighboring the reaction well.
5.3.5. Multiplex detection

Figure 5.6. Evaluation of multiplex LAMP in the polyester microfluidic chip. (A) Fluorescence
image of two arrays of the chip after 60 min incubation. The layout of the dried primers in the
104

arrays in indicated. (B,C and D) Real-time amplification plots for all wells of the three arrays
filled with (A) S. aureus, (B) E. coli, and (C) both templates DNA. No amplification was
observed in the array with no template (84).
Multiplex LAMP amplification was evaluated with primers for stx2 and eaeA genes of E.
coli, and primers for mecA and vicK genes of S. aureus, that were dried in triplicates in each
4

array of a chip. Sample containing either S. aureus genomic DNA (2.0 x 10 copies), E. coli (3.1
4

x 10 copies), both target, or no-template DNA were loaded in each one of the four arrays.
Amplification occurs only in the expected reaction well (Figure 5.6).

This results suggest that it is possible to detect up to four different targets with no cross
reactivity. This is very likely due to the fact that each primers set, even if in a same array, is
locally isolated in its respective reaction wells, because the cross-talk between the reaction wells
is forbidden by the microchannel length, diameter, and surface structure (161).

5.4. Conclusions
An inexpensive, easy-to-fabricate, and user-friendly polyester 64-well microfluidic chip
was developed. Sufficient distance between each reaction well allows acceptable fluidic isolation
to circumvent the fabrication of micro-valves to prevent cross contamination. This chip allowed
the simultaneous analysis of two genes each of E. coli and S. aureus by multiplex real-time
LAMP. This chip was also successfully evaluated for the detection of M. tuberculosis in a
previous chapter (see 3.3.5). This chip is well suited to perform LAMP in the Gene-Z platform.
Next steps include the development of a user-friendly and secure loading port that allows the
one-step filling of the chip with potentially dangerous samples (e.g., HIV contaminated blood).
105

106

CHAPTER 6. TOWARDS A FIELD-DEPLOYABLE AND INEXPENSIVE CHIP FOR
SNP DETECTION AND FUTURE PERPSECTIVES

6.1. Introduction
One of the main problems in detecting MDR/XDR-TB is that the majority of the
phenotypes associated with drug-resistance in TB are caused by SNPs. At least 18 mutations
associated to first-line drugs, and at least 3 mutations associated with second-line drugs (Table
6.1) are known. In order to establish a valuable drug-resistance profile with the Gene-Z platform,
it is therefore critical to establish a method to detect at least two dozen SNPs. The microfluidic
chip that was developed for the Gene-Z is composed of 64 reaction wells (0), which theoretically
enables the detection of up to 30 mutations in duplicate with some negative and positive controls.
Consequently, this chip should allow the level of multiplexing that is needed for detecting
mutations associated with MDR/XRD-TB. However, the use of the Gene-Z platforms also
requires an efficient method to amplify SNPs isothermally and in one step.
Several methods have been developed to detect SNPs (164,165) although only a few
allow direct isothermal amplification. For example, Iwasaki and collaborators developed a
method to detect SNPs by a modified real-time LAMP, and were able to detect mutations of the
cytochrome P450 gene directly, in less than one hour (20). However, this method requires that
the SNP is present at the penultimate position of the 3’ end of both FIB and BIP primers, which
might not always be possible in terms of primer design. Alternatively, a method that is similar to
LAMP was used in conjunction with a protein of the mismatch repair pathways named Taq
mutS. This method named smart amplification process version 2 (or SMAP 2) allows a wider
range of acceptable positions for the mutation in the primers because the Taq mutS suppresses
107

completely the elongation of mismatch-containing primers, thus preventing amplification. For
this reason, only one primer is needed to target the polymorphism that does not necessarily have
to be on the penultimate position of the primer. Therefore the mutation neither has to be on both
primers, nor has it to be on a specific position of the primer. Although this method has been used
to detect several SNPs, it requires the use of a specific DNA polymerase (Aac) that is more
expensive than Bst polymerase that is generally used in LAMP.
Here we propose a new method to detect SNP isothermally in one step by combining the
advantages of LAMP and SMAP 2. The detection of the SNP is based on (i) the use of a set of
LAMP inner primers (FIP and BIP) that both contain the SNP at the penultimate position of their
respective 3’ ends, and (ii) Taq mutS to suppress the amplification of mismatched primers. The
proof of concept of the method was evaluated with the codon S315 of the katG gene. Mutations
of this codon are associated with 40-60% of the resistance to isoniazid (166), one of the first-line
drugs against TB.

6.2. Materials and Methods
6.2.1. DNA targets
Genomic DNA from wild type M. tuberculosis (ATCC 25177), was obtained from the
American

Type

Culture

Collection

(ATCC,

Manassas,

VA)

and

resuspended

in

diethylpyrocarbonate-treated, nuclease-free sterile water (Fischer Scientific, Pittsburgh, PA). The
wild type katG DNA contain the codon S315, which has the sequence AGC and code for a serine
amino acid. For the mutant DNA, the 200 bp sequence that was used for the primer design was
synthesized and incorporated industrially pIDTSMART:ampicilin:blunt plasmid (IDT,
Coralville, IA). This sequence contained the mutated codon S315T, which has the sequence

108

AGG (instead of AGC) and code for an threonine (T) rather than a serine (S) (see Appendix 10)
. Both targets (the wild type DNA and the synthetic mutated DNA) were diluted to a
4

concentration of 10 copies/µL using the conversion factor of 670 g/mole of base pairs.

6.2.2. LAMP primers
A set of six specific LAMP primers was designed (FIP, BIP, F3, B3, LF and LB)
targeting the wild type codon S315 of the katG gene (Table 6.1). The F3, B3, LF, LB and FIP
primers were designed using Primer Explorer version 4 (Eiken Chemical Co., LTD, Tokyo,
Japan, http://primerexplorer.jp/e) and the BIP primer was designed manually. FIP and BIP were
designed so that both they contains the polymorphism in the penultimate position of their
respective 3’ ends (Figure 6.1) The specificity of the primer sets was checked against the
GenBank database using NCBI BLAST (140).

Table 6.1. Primers used in this study
Target
M. tuberculosis
katG wild type S315
(AGC = serine)

Primer name
F3-KatG S315-1
B3-KatG S315-1
LF-KatG S315-1
LB-KatG S315-1
FIP-KatG S315 wt-1
BIP-KatG S315 wt-1

Sequence (5'-3')
AGGCTGCTCCGCTGGA
AGCAGGGCTCTTCGTCAG
TGCCATACGAGCTCTTCCA
ACGAACACCCCGACGAAAT
GCTGGTGATCGCGTCCTTACCGCAGATGGGCTTGGGC
AGCGGCATCGAGGTCGTATGGTCGTAGCCGTACAGGATCTC

6.2.3. Taq mutS-based real-time LAMP
LAMP reactions were performed in a volume of 20 µL consisting of 1.6 µM each of FIP
and BIP primers, 0.2 µM each of F3 and B3 primers, 0.8 µM each of LF and LB primers, 0.8 M
109

betaine (Sigma, St Louis, MO), 1.4 mM of each dNTP (Invitrogen Corporation, Carlsbad, CA),
20 mM Tris-HCl (pH 8.8), 10 mM (NH4)2SO4, 10 mM KCl, 8 mM MgSO4, 8 mM Triton X100, 0.2 to 2.4 units/µL of Bst DNA polymerase, large fragment (New England Biolabs Inc.,
Ipswich, MA), 0.04 µg/uL of mismatch repair initiation protein Taq mutS (Affymetrix, Santa
Clara, CA) and 2 µM SYTO-81. Samples were loaded in 200-µL PCR tubes (VWR
International, West Chester, PA) and incubated at 64°C for 40-60 min in a Chromo4TM Real-time
PCR detector (Bio-Rad Laboratories, Hercules, CA). Fluorescence intensity was measured every
minute with the Channel 2 of the thermal cycler, using the pre-calibration settings for Cy3,
furnished by the manufacturer. Fluorescence data and Tt values were collected with OpticonTM
software v3.1 (Bio-Rad) and transferred to Microsoft Excel.

Figure 6.1. Schematic of the position of FIP & BIP primers for the detection of SNP by LAMP.

110

6.3. Results and Discussion
The possibility to detect isoniazid resistance in M. tuberculosis was evaluated on the
S315 polymorphism of the katG gene. This polymorphism was chosen because it counts for 4060% of all the mutations associated to isoniazid resistance (5). A primer set was designed to
detect the wild type genotype and was evaluated using the wild type target sequence (purified
genomic DNA,) and a mutation-containing synthetic sequence (G->C on the codon 315). This
experience was based on the hypothesis that a imperfect match would result in a delayed or
absent amplification. Alternatively, the same experience was performed in the presence of Taq
mutS, a mismatch repair initiation protein that is known to increase the discrimination of
mismatch hybridized primers (25)

Figure 6.2. LAMP amplification profile of the wild type primers for katG. In the presence of
Taq mutS (right panel) only the wild type DNA triggers an amplification. In the absence of Taq
mutS, both targets result in amplifications, but the mutant amplification time is about 10 min
longer than for the wild type.

111

In the absence of mutS, it was possible to observe an amplification signal for the wild
type target after roughly 20 min (Figure 6.2, left panel). An amplification signal was also
observed using the mutated target, but after a period of about 30 min. These results are in
contrast with those of Iwasaki who, using a primer set that was similarly designed (i.e., SNP at
the penultimate position of the 3’ end of both FIP and BIP) observed an amplification only for
the reaction that contain the perfect target (20). However, this study only evaluated the detection
of one polymorphism and no information was provided about the optimization of the primers. In
addition, the study used a different concentration of betaine depending on whether the assay was
for the wild type (1.0M) or for the mutant DNA (0.8M). It is possible that the effect observed in
this study was due to the differential concentration in betaine. Betaine is a well-known PCR
enhancer that facilitates the strands separation and decrease the melting temperature of dsDNA
(154). For a better SNP discrimination, it would be useful to evaluate the effect of betaine
concentration with the premise of finding a constant concentration that works for both wild type
and mutation assays. The effect of DMSO concentration could also be evaluated because this
molecule is known to increase the specificity of the base pairing (167,168). But the concentration
of these enhancers may have to be optimized for every new assay because the melting
(hybridization) temperature of LAMP primers is drastically affected by the length and GC
content. However, our results suggest that SNP discrimination is still possible based on a large
difference in amplification time between the wild type and the mutant. Detecting SNPs by this
method would require a reaction with the wild type primers and a reaction with the mutation
primers simultaneously. A decision based on which primer set results in the fastest amplification
can then be made. However, this would require that both primer sets have the same amplification
efficiency. This might not always be the case. For example, the substitution of a pyrimidine for a

112

purine might affect the melting temperature by up to several degrees of temperature. For
example, using an online Tm calculator of New England Biolabs (172) the simple substitution of
a G to T (in the FIP primer that was used in this study), resulted in a Tm drop of 2 ºC (from 77ºC
to 75ºC). Such a temperature difference may have an impact on the primers efficiency and
therefore on the amplification time.
In the presence of Taq mutS, the wild type target amplified after about 30 min, and the
amplification of the mutant seemed to be completely suppressed. The fact that the wild type
target amplified only after 30 min (rather than 20 min, in the absence of Taq mutS) suggests that
mutS may also have a slight inhibiting effect on the Bst polymerase has it was already reported
(25). This was somewhat expected since the function of this protein is to increase DNA
proofreading by binding to mismatch bases and preventing their elongation by the polymerase.
This proofreading function might increase the time for the amplification to be detected.
According to these results, the addition of Taq mutS may allow discriminating SNPs by real-time
LAMP. Our results also confirm that the function of Taq mutS is not only specific to the aac
polymerase used in SMAP, but also to Bst polymerase. It can therefore be used to improve the
specificity of LAMP for the detection of SNPs. This is in contradiction with the results of Mitani
was not able to suppress the amplification of the mutant target using Taq MutS in LAMP (25).
However, we were not able to reproduce these results using a Taq mutS purchased from another
company (Wako Chemicals), and unfortunately, Affymetrix discontinued their Taq mutS
product. The variable results obtained using the two different Taq mutS might be due to a batchto-batch variability, to bad transportation conditions, or to the method of fabrication and/or
purification.

113

6.4. Conclusions and perspectives
A new method was developed to detect a SNPs associated with isoniazid resistance in
TB. The method is based on the use of a set of inner primers that both contain the SNP on the
penultimate position of the 3’ end. The genotyping is based on the significant difference in Tt
between the wild type and the mutant sequence. Adding the mismatch repair initiation protein
Taq mutS, which allowed suppressing the amplification of any incorrect sequence, further
optimized the method. Based on the principle that a significant difference in amplification is
required to discriminate between mutant and wild type genotypes, both methods could be
theoretically used to type SNPs.
These results have implication for the detection of drug resistance in TB using the 64well microfluidic chip that was developed for Gene-Z. The SNP genotyping will imply the
presence of primer sets for both wild type and mutant in the same chip, but in different reaction
wells. Based on this principle, the primer set resulting in the shortest Tt will define the genotype.
However, this system might be inefficient for diagnosing patients that are infected by strains of
both genotypes, unless the two genotypes occur at the same level. The addition of Taq mutS
could circumvent this issue, as it would suppress the amplification of any primer that is not
perfectly hybridized to its target. In this case, the observation of an amplification for both wild
type and mutant primer will be only associated with the presence of both genotype in the sample,
no matter the occurrence.
One of the main challenges with this new SNP-typing method is the primer design.
Because the SNP position directs the primer localization and sequence, it reduces the flexibility
for the primer design, when compared to regular LAMP. It is therefore expected that some SNPs
will not be detectable by this method. A possible alternative would be to use LAMP primers that
114

have select positions modified with locked nucleic acids (LNA) (169). LNA monomers are
known to have a higher affinity for complementary nucleic acids sequences. This is due to the
fact that LNA monomers contain a modified ribose moiety that can form a covalent bound
between complementary bases only. Because of this, the difference in Tm resulting from the
hybridization of LNA probes with a perfect or a non-perfect target is larger than when using
regular unmodified probes. This high difference in Tm might represent an advantage for LAMP
because the primers that will not hybridize perfectly to the target are likely to stay denaturized at
the LAMP working temperature.
The possibility to use nucleic acid amplification methods has a great potential for the
diagnostics of tuberculosis and other infectious diseases. However, because of their high cost and
level of complexity, these methods are generally confined in fully equipped microbiology
laboratories, and are performed by highly trained personnel. For these reasons, there is a real
need for the development of inexpensive and user-friendly platforms that can be used at the POC
or directly in the field to detect microbes. In the case of tuberculosis, not only several organisms
can be responsible of the disease, but also several mutations need to be detected to establish an
accurate antibiotic resistance profile. Therefore it is critical to screen for multiple signature
sequences and develop a test with multiplex amplification. Another challenge in the diagnostics
of tuberculosis, is to develop a method that allows the detection polymorphisms due to single
base substitution (i.e. SNPs) as drug resistance in tuberculosis is essentially conferred by such
mutations (166). To solve these challenges our group developed the Gene-Z platform, an
inexpensive (<$500) system that allows multiplex amplification with a result in less than 30 min.
This nucleic acid amplification platform is fully functional and can be used by unskilled
personnel.
115

As part of using the Gene-Z platform in field settings for TB diagnostics, several critical
tasks have been accomplished. First, a method to perform fluorescence-based real-time LAMP in
microfluidic chips was optimized. This protocol was used with success to detect M. tuberculosis
in less than 15 min using the Gene-Z platform. Second, an easy method to freeze-dry molecular
biology reagents was developed. This method allowed storing molecular biology reagents for up
to one year with no loss of activity. The use of freeze-dried reagents also represents a good step
towards the ease of use because the reagents only need to be rehydrated with a target-containing
sample. Third, an inexpensive valve-less polyester microfluidic chip that is easy to manufacture
and that is now used in the Gene-Z platform was developed. Finally, a proof-of-concept for a
new method to detect an MDR-associated SNP using LAMP was presented. This method was
successfully used to detect a SNP associated to isoniazid resistance in tuberculosis.
In order to implement Gene-Z at the POC, the next step would be to begin testing the
chip and the device using clinical specimens, first for routine TB, and then for MDR/XDR TB.
This requires the development of a system that allows an untrained used to fill safely the chip
with a sample. It is therefore crucial to create a safe and efficient sample receiving system for the
chip. Another future challenge will be to develop a strategy to purify nucleic acids from complex
samples. For examples, sputum samples that are typically used in TB diagnostics are highly
heterogeneous, and may have very low counts of M. tuberculosis. A method developed by
Akonni Biosystems (Frederick, MD) used the bind-wash-elute method in a single pipette tip
(named TruTip

TM

) to purify TB nucleic acids in four minutes and with a recovery of more than

90%. Due to its simplicity and rapidity, this method has good potential to be used in conjunction
with the Gene-Z to detect TB at the POC. Because inhibitors present in blood and serum less
impact LAMP, it is likely that some pathogens can be detected directly without any sample

116

preparation. However, the cost of TruTip

TM

($5) may need to be incorporated in the overall cost

of the assay. It is worth noting that even with one SNP, the assay developed here is equivalent in
terms of target measured and much lower in cost compared to GeneXpert. Hence, it should now
be validated using clinical specimens in a BSL3 facility.
Finally, the number of SNPs to be detected must be increased to at least a dozen before
!

the overall assay can be treated as better than GeneXpert . This requires a more efficient way to
design primers for SNPs than what was available for LAMP as well as an alternative
enzyme/mechanism for Taq mutS. One such alternative is locked nucleic acid (LNA), which has
been extensively used in PCR-based SNP detection (169-171). Future studies should focus on
exploring additional alternative approaches for isothermal SNP detection. Finally, stability and
shelf-life of reagents in extremely hot climates (e.g., closer to 45ºC) needs to be evaluated.
Overall, if the assay with clinical specimens is successful (i.e., has a competitive performance in
terms of sensitivity and specificity), measuring MDR/XDR TB using a $500 device at a cost of
$2-$5 per test that can detect at least two dozen SNPs will be a significant contribution to the
field of TB diagnostics and control.

117

APPENDICES

118

APPENDIX 1.

Table A1.1 List of primers used in CHAPTER 3.
Target

Primer name

Sequence (5’-3’)

G .intestinalis
Betagiardin gene

Giardia_F3
Giardia_B3
Giardia_FIP
Giardia_BIP
Giardia_LF
Giardia_LB
MTB-F3
MTB-B3
MTB-FIP
MTB-BIP
MTB-FLP
MTB-BLP
FLjB_F3
FljB_B3
FljB_FIP
FljB_BIP
FljB_LF
FljB_LB
MecA_F3
MecA_B3
MecA_FIP
MecA_BIP
MecA_LF
MecA_LB

GATCTCGAGACGGGCATT G
CCTGGATAGCGCGGATCT
AGCCCTCTGCGACCTTCTCGCACGGAGAACGCAGAAAGG
GTTAGCGCTGCCACGACAGACTGCTCGTTGACGCACTTC
TGAGCTGGTCGTACATCTTCTT
GCGCTCACAAACACGAAGC
CTGGCTCAGGACGAACG
GCTCATCCCACACCGC
CACCCACGTGTTACTCATGCAAGTCGAACGGAAAGGTCT
TCGGGATAAGCCTGGACCACAAGACATGCATCCCGT
GTTCGCCACTCGAGTATCTCCG
GAAACTGGGTCTAATACCGG
CCCGTAACGCTAACGACG
CTGCGCCAGGACTTTCAC
GAACCGCCAGTTCACGCACACTCCATTGCGCAGACCAC
ACTCCCAGTCTGACCTCGACTCAGTCTGGCCGGATACACG
CGTTCAGCGCGCCTTCA
CATCCAGGCTGAAATCACCC
AAAAAACGAGTAGATGCTCAA
TGGCCAATTCCACATTGT
TCCCAATCTAACTTCCACATACCATAAAACAAACTACGGTAACATTGA
CATAGCGTCATTATTCCAGGAATGCCGGTCTAAAATTTTACCACGT
TTTAACAAAATTAAATTGAACGTTGCGA
AGAAAGACCAAAGCATACATATTGAAAA

M .tuberculosis 16S
rRNA gene

S. enterica
FljB gene

S. aureus mecA gene

119

APPENDIX 2.

Figure A2.1. Comparison of fluorescence intensity (A) and SNR amplification profiles (B) of all
dyes. Comparison of fluorescence intensity (A and C) and SNR amplification profile (B and D)
of SYTO-80, SYTO-81, SYTO-82, SYTO-83, SYTO-84, SYTO-85 and SYTOX Orange (A and
B), and calcein, EvaGreen, SYBR Green I, and PicoGreen (C and B). All dyes were used at their
optimum concentration. Target: 0.1 ng of M. tuberculosis genomic DNA.

120

Figure A2.1 (cont’d)

121

Figure A2.1 (cont’d)

122

Figure A2.1 (cont’d)

123

APPENDIX 3.

Figure A3.1. Effect of Bst concentration on Tt of real-time LAMP using SYTO-81 or SYTO-82. Target: 0.1 ng of M. tuberculosis
genomic DNA. The error bars represent the standard error from triplicate values.

124

APPENDIX 4.

Figure A4.1. Effect of the concentration of SYTO-81 on the LAMP monitoring in the Gene-Z.
Amplification plots (A), resulting velocity plots (B), and resulting Tt (C). Tt was calculated as
the first point of the velocity curve that goes above an arbitrary threshold value of 0.001. A
polyester microfluidic chip of 64 wells of 0.5 µL was used for this experiment and the target was
10 ng of M. tuberculosis genomic DNA.

125

Figure A4.1 (cont’d)

126

Figure A4.1 (cont’d)

127

Figure A4.1 (cont’d)

128

APPENDIX 5.

Table A5.1. Performances of existing fluorescence-based real-time LAMP assays compared with
results obtained using SYTO-82. For the literature data, when available, the Tt and target copy
number values were directly plotted. When not available, values were extracted from regression
analysis (standard curves) or amplification plots using a grid made with Microsoft Power Point.
For this purpose, the published plot was copied and cut in a Power Point slide. For the target,
when only the mass concentration was given, it was converted in copy number using the genome
size, and the conversion factor of 670 g/mole of base pairs.

129

Table A5.1 (Cont’d)
Dye (concentration)

Target

Lowest # of
copies detected

Real-time LAMP
®
SYBR Green I (N/A)

Leptospira

YO-PRO1 (N/A)
YO-PRO1 (250 ng/mL)
SYTO-82 (2.0 µM)

S. enterica
P. jirovecii
S. aureus

SYBR Green I (1 X)
SYTO-82 (2.0 µM)
SYTO-82 (2.0 µM)
SYTO-82 (2.0 µM)
YO-PRO1 (N/A)
SYTO-9 (0.4 µM)
Ethidium bromide (250 fg/µL)

B. xylophilus*
M. tuberculosis
S.thyphymurium
G. Intestinalis
Amonia oxidizing bacteria**

SYBR Green I (1 X)

Hepatitis B virus

Unknown dye

a

Approx. Tt (min)

Reference

1
a
2
10
10

30

(138)

29
18
20

(65)
(148)
Current Study

10
10
10
10
10
a
47
100

41
9
12
17
75

(137)
Current Study
Current Study
Current Study
(56)

45
28

(133)
(145)

37

(85)

B. cinerea

100
b
136

12

(147)

EvaGreen (0.5 X)

P. ramorum

138

28

(142)

YO-PRO1 (N/A)
SYTO-9 (3.34 µM)

B. melitensis
T. brucei rhodesiense

272
N/A

19
10

(146)
(57)

®

®

V. parahaemolyticus*
Lambda DNA virus

c

d

Real-time RT-LAMP
Ethidium bromide (5 fg/mL)
RVFV
10
18
(55)
PicoGreen (1/1000 dilution)
Foot and mouth disease virus 10
22
(143)
*
**
All assays were performed using the 6 LAMP primers (F3, B3, FIP, BIP, LF and LB) except (5 primers) and (4 primers).
a
b,c,d
number of cells;
copy number calculated from mass concentration assuming a genome size of 38, 65, and 3.3 Mb respectively,
and a conversion factor of 670 g/mol of base

130

APPENDIX 6.

Figure A6.1 Qualitative comparison of the current results with existing results of fluorescencebased real-time LAMP. (A) to (D): real-time LAMP assay using SYTO-82 (2 µM), targeting G.
intestinalis, M. tuberculosis, S. aureus, and S. enterica respectively (current study). References
for the figures are as follow. [2] Njiru ZK, Mikosza ASJ, Armstrong T, Enyaru JC, Ndung'u JM,
et al. 2008 Loop-Mediated Isothermal Amplification (LAMP) Method for Rapid Detection of
Trypanosoma

brucei

rhodesiense.

PLoS

Negl

Trop

Dis

2(2):

e147.

doi:10.1371/journal.pntd.0000147; [7] Aoi Y, Hosogai M, Tsuneda S (2006) Real-time
quantitative LAMP (loop-mediated isothermal amplification of DNA) as a simple method for
monitoring ammonia-oxidizing bacteria. J Biotechnol 125: 484-91; [8] Chen SY, Ge BL (2010)
Development of a toxR-based loop-mediated isothermal amplification assay for detecting Vibrio
parahaemolyticus. BMC Microbiol 10. Available: http://www.biomedcentral.com/14712180/10/4. Accessed: October 2010; [9] Han FF, Ge BL (2010) Quantitative detection of Vibrio
vulnificus in raw oysters by real-time loop-mediated isothermal amplification. International J
Food Microbiol 142: 60-6; [16] Cai T, Lou GQ, Yang J, Xu D, Meng ZH (2008) Development
and evaluation of real-time loop-mediated isothermal amplification for hepatitis B virus dna
quantification: A new tool for hbv management. J Clin Virol 41: 270-6; [17] Kikuchi T, Aikawa
T, Oeda Y, Karim N, Kanzaki N (2009) A rapid and precise diagnostic method for detecting the
pinewood nematode Bursaphelenchus xylophilus by loop-mediated isothermal amplification.
Phytopathology 99: 1365-9; [18] Lin XA, Chen Y, Lu YY, Yan JY, Yan J (2009) Application of
a loop-mediated isothermal amplification method for the detection of pathogenic Leptospira.
Diag. Microbiol Infect Dis 63: 237-42; [24] Tomlinson J, Barker I, Boonham N (2007) Faster,

131

simpler, more-specific methods for improved molecular detection of Phytophthora ramorum in
the field. Appl Environ Microbiol 73: 4040-7; [28] Nagamine K, Hase T, Notomi T (2002)
Accelerated reaction by loop-mediated isothermal amplification using loop primers. Mol Cell
Probes 16:223-9; [29] Ohtsuki R, Kawamoto K, Kato Y, Shah MM, Ezaki T, Makino SI (2008)
Rapid detection of Brucella spp. by the loop-mediated isothermal amplification method. J App
Microbiol 104: 1815-23; [30] Tomlinson JA, Dickinson MJ, Boonham N (2010) Detection of
Botrytis cinerea by loop-mediated isothermal amplification. Lett Appl Microbiol 51: 650-7; [31]
Uemura N, Makimura K, Onozaki M, Otsuka Y, Shibuya Y, Yazaki H, et al. (2008)
Development of a loop-mediated isothermal amplification method for diagnosing Pneumocystis
pneumonia. J MedMicrobiol 57: 0-7, [33] Peyrefitte CN, Boubis L, Coudrier D, Bouloy M,
Grandadam M, Tolou HJ, Plumet S (2008) Real-time reverse-transcription loop-mediated
isothermal amplification for rapid detection of rift valley fever virus. J Clin Microbiol 46: 36539; [34] Hara-Kudo Y, Yoshino M, Kojima T, Ikedo M (2005) Loop-mediated isothermal
amplification for the rapid detection of Salmonella. FEMS Microbiol Lett 253:155-61.

132

133

APPENDIX 7.

Figure A7.1 Cutting patterns for the xurography. (A) patterned film, and (B) membranes cutting
pattern

134

APPENDIX 8.
Table A8.1. List of primers used in CHAPTER 5
Target

Primer name

Sequence (5’-3’)

S. aureus mecA gene

MecA_F3
MecA_B3
MecA_FIP
MecA_BIP
MecA_LF
MecA_LB

AAAAAACGAGTAGATGCTCAA
TGGCCAATTCCACATTGT
TCCCAATCTAACTTCCACATACCATAAAACAAACTACGGTAACATTGA
CATAGCGTCATTATTCCAGGAATGCCGGTCTAAAATTTTACCACGT
TTTAACAAAATTAAATTGAACGTTGCGA
AGAAAGACCAAAGCATACATATTGAAAA

S. aureus vickR gene

vickR_F3
vickR_B3
vickR_FIP
vickR_BIP
vickR_LF
vickR_LB

GTCAAGAAATTGGAGAAATTCG
ACCAATTACCTTTTTATCGACTT
GGACAGAACTATCATTCGCTTTTTGGACCAAATTATTATTGCGACGAC
GCACTATCACTAGGACAATCAAACGATATTATATACCCAGACACGGT
ATTGATTAGACTACGGTTAGACTGC
TTAAAAGATTATGGCGGTGGTAAGG

E. coli eaeA gene

eaeA_F3
eaeA_B3
eaeA_FIP
eaeA_BIP
eaeA_LF
eaeA_LB

AGCTCTAACAATGTACAGCT
AGTTGCAGTTCCTGAAACA
GTCTTATCCGCCGTAAAGTCCGCCGTTCTGTCGAATGGTC
CTAAAGCGGATAACGCCGATACCCAGGGACATTAGCCTGAG
CCCAACCTGGTCGACAACTT
ATTACTTATACCGCGACGGTGAA

E. coli stx2 gene

stx2_F3
stx2_B3
stx2_FIP
stx2_BIP
stx2_LF
stx2_LB

GAGATATCGACCCCTCTTG
AATCTGAAAAACGGTAGAAAGT
TCCACAGCAAAATAACTGCCCAACATATATCTCAGGGGACCA
GATGTCTATCAGGCGCGTTTTGCCGTATTAACGAACCCGG
TGTGGTTAATAACAGACACCGATG
ACCATCTTCGTCTGATTATTGAGC

135

APPENDIX 9.
Standard operating protocol (SOP) for the fabrication of a polyester chip
th

Version 1, by Gregoire Seyrig, January 25 2012

Material:

One Carver Hydraulic Press (with water cooling system): Model 3851, Carver, Inc, Wabash, IN
One 6” x 6” x 0.25” FDA-compliant Silicon Rubber (Durometer of 60A), McMaster Carr,
Elmhurst IL
Two 6” x 6” x 0.035” Economy-grade Stainless Steel Sheet (type 430), McMaster Carr
One 5” x 5” x 0.005” Clear Polyester Film, McMaster Carr
One Craft Robot Pro plotter, Graphtec America, Inc., Irvine, CA
One Craft Robot Pro Carrier sheet A3, Graphtec America, Inc., Irvine, CA
Two Optical films: MicroAmp optical adhesive Film (PN: 4311971), Life Technologies,
Carlsbad, CA
One Durapore Membrane Filter, Millipore Coproration, Billerica, MA
Double sided tape JVCC DC-1114, Findtape.com, Skillman, NJ
One Aluminum mold, Proto Labs, Mapple Plain, MN
Scissors

Step 1: Press

136

Figure A9.1. Use of the hydraulic press.

Cut a piece of polyester (about 5” x 5”) using the scissors
Turn-on the carver press and set the temperature at 425ºF for each heated platen
During heating (Figure A9.1):
-

place the mold at the center of a steel plate,

-

place the polyester sheet on top of the mold,

-

place the rubber on top of the polyester sheet,

-

place another steel plate on top of the rubber,

-

place the whole “sandwich” between the two platen of the press,

-

Apply 500 pounds of pressure and wait for temperature to rise.

Once at temperature (425ºF):
-

apply 12,000 pounds of pressure,

-

turn the cooling system on (e.g., turn the water on).

Wait for system to cool down to about 200ºF.

137

Release pressure (turn the release pressure valve to anti-clockwise).
Remove the “sandwich” from the press, remove the steel plates, the rubber and detach the chip
from the mold manually.
Make sure the integrity of the chip is ok (no visible breakpoints or holes).

Step 2: Preparation of patterned hydrophobic membrane

Turn the plotter one.
Place the membrane on the carrier sheet (at the spot that is indicated on the carrier sheet).
Place a protective blue disc (furnished with the membranes) on top of the membrane.
Fix the all preparation by using a layer of adhesive film.
Place the carrier sheet in the plotter with the membrane facing up.
Lock the carrier sheet (pull “levier” in the back) and press “enter” on the plotter.

Turn the software on and load the membrane pattern file named Membrane.GSD (Figure 1.2)
Press “cut”, a little confirmation window will open.
Press “enter” and another conformation window will open.
Press “enter” and the plotter will begin to cut.

138

Figure A9.2. Pattern for membrane cutting (Membrane.GSD)

Once the plotter has completed its task:
-

pull “levier” down on the plotter and remove the carrier sheet,

-

detach the optical tape and the blue disc from the carrier sheet,

-

remove the parts of membrane that are not wanted using a sharp blade,

-

remove the membranes one by one using a sharp blade, and place them in a petri dish for

later use.

Step 3. Preparation of the patterned adhesive film

Turn the plotter on.
Place the optical film, liner facing up, on the carrier sheet (at the indicated spot).
139

Place the carrier sheet in the plotter.
Lock the carrier sheet (pull “levier” in the back) and press “enter” on the plotter.

Turn the software on and load the adhesive pattern file named Film.GSD (Figure A9.3).
Press “cut”, a little confirmation window will open.
Press “enter” and another conformation window will open.
Press “enter” and the plotter will begin to cut.

Once the plotter has completed its task:
-

pull the “levier” down on the plotter and remove the carrier sheet,

-

using a double-side tape, remove the cutoff parts,

-

detach the adhesive film from the carrier sheet.

Using an air duster, remove any cutoff parts that stayed trapped on the blade

140

Figure A9.3. Pattern for film cutting (Film.GSD)

Step 4: Preparation of the chip

Place the patterned hydrophobic membranes at the appropriate spots on the chip.
Tape the patterned adhesive on the chip.
Make sure it is well aligned with the membrane and the chip (Figure A9.4)

141

Figure A9.4. Alignment of the membrane and the adhesive on the membrane

Step 5: Finalization of the chip

Place the chip, structure facing up, on a steel plate
Place the rubber on top of it
Place another steep plate on top of the rubber
Place in the hydraulic press (at room temperature) and apply 5,000 pounds of pressure for ten
seconds
Release pressure and remove the chip
Using a photo-cutter, cut the edges of the chip

142

Seal the four edges of the chip using a plastic bag sealer (line sealer)

References:

Stedtfeld, R.D. et al., 2012, Lab On a Chip (accepted)
Tourlousse D.M.. 2010, Michigan State University, PhD thesis

143

APPENDIX 10.
Figure A10.1. Mutated target sequence (5’ to 3’) that was incorporated in a
pIDTSMART:ampicilin:blunt plasmid. The position of the mutation is indicated in bold and is
underlined. The mutation S315R corresponds from a amino acid change from serine to threonine,
which confers to M. tuberculosis the resistance to the first line drug isoniazid.

ACGTGGAGGCGATCACACCGCAGACGTTGATCAACATCCGGCCGGTGGTCGCCGCG
ATCAAGGAGTTCTTCGGCACCAGCCAGCTGAGCCAATTCATGTACCAGAACAACCC
GCTGTCGGGGTTGACCTACAAGCGCCGACTGTCGGCGCCGGGGCCCGGCGGTCTGT
CACGTGAGCGTGCCGGGCTGGAGGTCCGCGACGTGCACCCGTCGCACTACGGCCGG
ATGTGCCCGATCGAAACCCCAGGCTGCTCCGCTGGAGCAGATGGGCTTGGGCTGGA
AGAGCTCGTATGGCACCGGAACCGGTAAGGACGCGATCACCACCGGCATCGAGGTC
GTATGGACGAACACCCCGACGAAATGGGACAACAGTTTCCTCGAGATCCTGTACGG
CTACGAGTGGGAGCTGACGAAGAGCCCTGCTGGC

144

APPENDIX 11.
Table A11.1. Overview of the mutations in MDR-TB and XDR-TB. RMP, rifampin; INH,
isoniazid; FQL, fluoroquinolone; AMI&CP, aminoglycosides and cyclic peptides; EMB,
ethambutol; STR, streptomycin. Coverage by commercial hybridization-based assays: mutations
covered by

a

INNO-LiPA Rif. TB (LiPA) assay,

b

®

GenoType

CombiChipTM Mycobacteria Drug-resistance Detection DNA chip; and
MTBDRsl assay. Table adapted from Tourlousse, 2010 (23).

145

c

MTBDRplus assay,
d

®

GenoType

Drug

Gene/region

Codon

RMP

rpoB

D516

Mutation
a,b,c
a
D516V
, D516Y

S531

S531L

H526

H526Y

Others

S522L , L533P ,L511P, Q513L , F514L

katG
oxyR-aphC
Intergenic region

S315
G-46
C-39

inhA promotor

C-15

S315T1
G-46A
C-39T
b
C-15T

T-8A

T-8A , T-8C

A-16

A-16

D94

D94N , D94Y , D94H , D94G , D94A

A90

A90V

gyrB

S91
N538

S91
N538D

rrs

A1401

A1401G

C1402

C1402T , C1402A

G1484
A1408

INH

FQL

AMI/CP

gyrA

a,b,c

c

, S531W

a,b,c

c

, H526D

a,b,c

c

b,c

b

c

c

b

c

, H526R , H526L , H526P
c

c

, S315T2 , S315N , S315I

b

b

d

d

d

d

d

d

d

d

d

d

EMB

embB

M306
Other

G1484T
A1408G
d
M306I1 , M306I2, M306I3, M306V, M3016L
G406D, G406S, G406R

STR

rpsL

K43
K88
C491
A906
A513

K43T
K88Q, K88R
C491T
A906C
A513C, A513T

rrs

146

c

REFERENCES

147

REFERENCES

1.

World Health Organization. (2010). Global tuberculosis control: WHO Report 2010,
Geneva, Switzerland.

2.

McNerney, R. and Daley, P. (2011) Towards a point-of-care test for active tuberculosis:
obstacles and opportunities. Nature Reviews Microbiology, 9, 204-213.

3.

Keeler, E., Perkins, M.D., Small, P., Hanson, C., Reed, S., Cunningham, J., Aledort, J.E.,
Hillborne, L., Rafael, M.E., Girosi, F. et al. (2006) Reducing the global burden of
tuberculosis: the contribution of improved diagnostics. Nature, 444 Suppl 1, 49-57.

4.

Palomino, J.C. (2009) Molecular detection, identification and drug resistance detection in
Mycobacterium tuberculosis. FEMS Immunology and Medical Microbiology, 56, 103111.

5.

LoBue, P. (2009) Extensively drug-resistant tuberculosis. Current Opinion in Infectious
Diseases, 22, 167-173.

6.

Center for Disease Control and Prevention (CDC), C.f.D.C.a.P. (2006) Emergence of
Mycobacterium tuberculosis with extensive resistance to second-line drugs-worldwide,
2000-2004. MMWR Morbidity and Mortality Weakly Report, 55, 301-305.

7.

Gandhi, N.R., Moll, A., Sturm, A.W., Pawinski, R., Govender, T., Lalloo, U., Zeller, K.,
Andrews, J. and Friedland, G. (2006) Extensively drug-resistant tuberculosis as a cause
of death in patients co-infected with tuberculosis and HIV in a rural area of South Africa.
Lancet, 368, 1575-1580.

8.

Notomi, T., Okayama, H., Masubuchi, H., Yonekawa, T., Watanabe, K., Amino, N. and
Hase, T. (2000) Loop-mediated isothermal amplification of DNA. Nucleic Acids
Research, 28, e63.

9.

Pandey, B., Poudel, A., Yoda, T., Tamaru, A., Oda, N., Fukushima, Y., Lekhak, B., Risal,
B., Acharya, B., Sapkota, B. et al. (2008) Development of an in-house loop-mediated
isothermal amplification (LAMP) assay for detection of Mycobacterium tuberculosis and
evaluation in sputum samples of Nepalese patients. Journal of Medical Microbiology, 57,
439-443.

148

10.

Neonakis, I.K., Spandidos, D.A. and Petinaki, E. (2011) Use of loop-mediated isothermal
amplification of DNA for the rapid detection of Mycobacterium tuberculosis in clinical
specimens. European journal of clinical microbiology & infectious diseases, 1-6.

11.

Iwamoto, T., Sonobe, T. and Hayashi, K. (2003) Loop-mediated isothermal amplification
for direct detection of Mycobacterium tuberculosis complex, M. avium, and M.
intracellulare in sputum samples. Journal of Clinical Microbiology, 41, 2616-2622.

12.

Boehme, C.C., Nabeta, P., Henostroza, G., Raqib, R., Rahim, Z., Gerhardt, M., Sanga, E.,
Hoelscher, M., Notomi, T., Hase, T. et al. (2007) Operational feasibility of using loopmediated isothermal amplification for diagnosis of pulmonary tuberculosis in microscopy
centers of developing countries. Journal of Clinical Microbiology, 45, 1936-1940.

13.

Zhu, R.Y., Zhang, K.X., Zhao, M.Q., Liu, Y.H., Xu, Y.Y., Ju, C.M., Li, B. and Chen,
J.D. (2009) Use of visual loop-mediated isotheral amplification of rimM sequence for
rapid detection of Mycobacterium tuberculosis and Mycobacterium bovis. Journal of
Microbiological Methods, 78, 339-343.

14.

Lee, M.F., Chen, Y.H. and Peng, C.F. (2009) Evaluation of reverse transcription loopmediated isothermal amplification in conjunction with ELISA-hybridization assay for
molecular detection of Mycobacterium tuberculosis. Journal of Microbiological Methods,
76, 174-180.

15.

Fukasawa, T., Oda, N., Wada, Y., Tamaru, A., Fukushima, Y., Nakajima, C. and Suzuki,
Y. (2010) A Novel Method for the Purification of DNA by Capturing Nucleic Acid and
Magnesium Complexes on Non-Woven Fabric Filters under Alkaline Conditions for the
Gene Diagnosis of Tuberculosis by Loop-Mediated Isothermal Amplification (LAMP).
Japanese Journal of Infectious Diseases, 63, 246-250.

16.

Zhang, J., Zhang, G.H., Yang, L., Huang, R., Zhang, Y., Jia, K., Yuan, W. and Li, S.J.
(2011) Development of a loop-mediated isothermal amplification assay for the detection
of Mycobacterium bovis. Veterinary Journal, 187, 393-396.

17.

Aryan, E., Makvandi, M., Farajzadeh, A., Huygen, K., Bifani, P., Mousavi, S.-L., Fateh,
A., Jelodar, A., Gouya, M.-M. and Romano, M. (2010) A novel and more sensitive loopmediated isothermal amplification assay targeting IS6110 for detection of
Mycobacterium tuberculosis complex. Microbiological Research, 165, 211-220.

18.

Nakamura, N., Ito, K., Takahashi, M., Hashimoto, K., Kawamoto, M., Yamanaka, M.,
Taniguchi, A., Kamatani, N. and Gemma, N. (2007) Detection of six single-nucleotide
polymorphisms associated with rheumatoid arthritis by a loop-mediated isothermal
amplification method and an electrochemical DNA chip. Analytical Chemistry, 79, 94849493.

149

19.

Higashimoto, Y., Ihira, M., Ohta, A., Inoue, S., Usui, C., Asano, Y. and Yoshikawa, T.
(2008) Discriminating between varicella-zoster virus vaccine and wild-type strains by
loop-mediated isothermal amplification. Journal of Clinical Microbiology, 46, 26652670.

20.

Iwasaki, M., Yonekawa, T., Otsuka, K., Suzuki, W., Nagamine, K., Hase, T., Tatsumi,
K.-i., Horigome, T., Notomi, T. and Kanda, H. (2003) Validation of the Loop-Mediated
Isothermal Amplification Method for Single Nucleotide Polymorphism Genotyping with
Whole Blood. Genome Letters, 2, 119-126.

21.

Ikeda, S., Takabe, K., Inagaki, M., Funakoshi, N. and Suzuki, K. (2007) Detection of
gene point mutation in paraffin sections using in situ loop-mediated isothermal
amplification. Pathology International, 57, 594-599.

22.

Fukuta, S., Mizukami, Y., Ishida, A. and Kanbe, M. (2006) Development of loopmediated isothermal amplification (LAMP)-based SNP markers for shelf-life in melon
(Cucumis melo L.). Journal of Applied Genetics, 47, 303-308.

23.

Tourlousse, D.M. (2010), Development of low-cost and easy-to-use tools for multiplex
pathogen detection, Michigan State University, East Lansing.

24.

Stedtfeld, R.D. (2009), Development and validation of a multiplex hand-held gene
analyzer, Michigan State University, East Lansing.

25.

Mitani, Y., Lezhava, A., Kawai, Y., Kikuchi, T., Oguchi-Katayama, A., Kogo, Y., Itoh,
M., Miyagi, T., Takakura, H., Hoshi, K. et al. (2007) Rapid SNP diagnostics using
asymmetric isothermal amplification and a new mismatch-suppression technology.
Nature Methods, 4, 257-262.

26.

Ahmed, W., Huygens, F., Goonetilleke, A. and Gardner, T. (2008) Real-time PCR
detection of pathogenic microorganisms in roof-harvested rainwater in Southeast
Queensland, Australia. Applied and Environmental Microbiology, 74, 5490-5496.

27.

Harwood, V.J., Levine, A.D., Scott, T.M., Chivukula, V., Lukasik, J., Farrah, S.R. and
Rose, J.B. (2005) Validity of the indicator organism paradigm for pathogen reduction in
reclaimed water and public health protection. Applied and Environmental Microbiology,
71, 3163-3170.

28.

Tourlousse, D.M., Ahmad, F., Stedtfeld, R.D., Seyrig, G., Duran, M., Alm, E.W. and
Hashsham, S.A. (2008) Detection and Occurrence of Indicator Organisms and Pathogens.
Water Environment Research, 80, 898-928.

150

29.

Reynolds, K., Mena, K. and Gerba, C. (2008) Risk of waterborne illness via drinking
water in the United States. Reviews of Environmental contamination & Toxicology, 192,
117-158.

30.

Mac Kenzie, W.R., Hoxie, N.J., Proctor, M.E., Gradus, M.S., Blair, K.A., Peterson, D.E.,
Kazmierczak, J.J., Addiss, D.G., Fox, K.R., Rose, J.B. et al. (1994) A massive outbreak
in Milwaukee of cryptosporidium infection transmitted through the public water supply.
New England Journal of Medicine, 331, 161-167.

31.

Williams, R.C., Isaacs, S., Decou, M.L., Richardson, E.A., Buffett, M.C., Slinger, R.W.,
Brodsky, M.H., Ciebin, B.W., Ellis, A. and Hockin, J. (2000) Illness outbreak associated
with Escherichia coli O157:H7 in Genoa salami. E. coli O157:H7 Working Group.
Canadian Medical Association Journal, 162, 1409-1413.

32.

Grant, J., Wendelboe, A.M., Wendel, A., Jepson, B., Torres, P., Smelser, C. and Rolfs,
R.T. (2008) Spinach-associated Escherichia coli O157:H7 outbreak, Utah and New
Mexico, 2006. Emerging Infectious Diseases, 14, 1633-1636.

33.

Miller, S.M., Tourlousse, D.M., Stedtfeld, R.D., Baushke, S.W., Herzog, A.B., Wick,
L.M., Rouillard, J.M., Gulari, E., Tiedje, J.M. and Hashsham, S.A. (2008) In situsynthesized virulence and marker gene biochip for detection of bacterial pathogens in
water. Applied and Environmental Microbiology, 74, 2200-2209.

34.

Saiki, R.K., Scharf, S., Faloona, F., Mullis, K.B., Horn, G.T., Erlich, H.A. and Arnheim,
N. (1985) Enzymatic amplification of beta-globin genomic sequences and restriction site
analysis for diagnosis of sickle cell anemia. Science, 230, 1350-1354.

35.

Saiki, R.K., Gelfand, D.H., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullis, K.B.
and Erlich, H.A. (1988) Primer-directed enzymatic amplification of DNA with a
thermostable DNA polymerase. Science, 239, 487-491.

36.

Compton, J. (1991) Nucleic acid sequence-based amplification. Nature, 350, 91-92.

37.

Kievits, T., van Gemen, B., van Strijp, D., Schukkink, R., Dircks, M., Adriaanse, H.,
Malek, L., Sooknanan, R. and Lens, P. (1991) NASBA isothermal enzymatic in vitro
nucleic acid amplification optimized for the diagnosis of HIV-1 infection. Journal of
Virological Methods, 35, 273-286.

38.

Walker, G.T. (1993) Empirical aspects of strand displacement amplification. PCRMethods and Applications, 3, 1-6.

39.

Vincent, M., Xu, Y. and Kong, H. (2004) Helicase-dependent isothermal DNA
amplification. EMBO Reports, 5, 795-800.
151

40.

Piepenburg, O., Williams, C.H., Stemple, D.L. and Armes, N.A. (2006) DNA detection
using recombination proteins. PLoS Biology, 4, e204.

41.

Plutzer, J., Torokne, A. and Karanis, P. (2010) Combination of ARAD microfibre
filtration and LAMP methodology for simple, rapid and cost-effective detection of human
pathogenic Giardia duodenalis and Cryptosporidium spp. in drinking water. Letters in
Applied Microbiology, 50, 82-88.

42.

Kumar, P., Sen, M.K., Chauhan, D.S., Katoch, V.M., Singh, S. and Prasad, H.K. (2010)
Assessment of the N-PCR assay in diagnosis of pleural tuberculosis: detection of M.
tuberculosis in pleural fluid and sputum collected in tandem. PLoS One, 5, e10220.

43.

Nadeau, J.G., Pitner, J.B., Linn, C.P., Schram, J.L., Dean, C.H. and Nycz, C.M. (1999)
Real-time, sequence-specific detection of nucleic acids during strand displacement
amplification. Analytical Biochemistry, 276, 177-187.

44.

Hellyer, T.J., DesJardin, L.E., Teixeira, L., Perkins, M.D., Cave, M.D. and Eisenach,
K.D. (1999) Detection of viable Mycobacterium tuberculosis by reverse transcriptasestrand displacement amplification of mRNA. Journal of Clinical Microbiology, 37, 518523.

45.

Fontana, C., Favaro, M., Cicchetti, O., Minelli, S., Pistoia, E.S. and Favalli, C. (2005)
Performance of strand displacement amplification assay in the detection of Chlamydia
trachomatis and Neisseria gonorrhoeae. Japanese Journal of Infectious Diseases, 58,
283-288.

46.

Chen, Q.H., Bian, Z.H., Chen, M., Hua, X., Yao, C.Y., Xia, H., Kuang, H., Zhang, X.,
Huang, J.F., Cai, G.R. et al. (2009) Real-time monitoring of the strand displacement
amplification (SDA) of human cytomegalovirus by a new SDA-piezoelectric DNA sensor
system. Biosensors & Bioelectronics, 24, 3412-3418.

47.

Motre, A., Li, Y. and Kong, H. (2008) Enhancing helicase-dependent amplification by
fusing the helicase with the DNA polymerase. Gene, 420, 17-22.

48.

Nelson, J., Cai, Y., Giesler, T., Farchaus, J., Sundaram, S., Ortiz-Rivera, M., Hosta, L.,
Hewitt, P., Mamone, J., Palaniappan, C. et al. (2002) TempliPhi, phi29 DNA polymerase
based rolling circle amplification of templates for DNA sequencing. Biotechniques,
Suppl, 44-47.

49.

Rector, A., Tachezy, R. and Van Ranst, M. (2004) A sequence-independent strategy for
detection and cloning of circular DNA virus genomes by using multiply primed rollingcircle amplification. Journal of Virology, 78, 4993-4998.

152

50.

Johne, R., Müller, H., Rector, A., van Ranst, M. and Stevens, H. (2009) Rolling-circle
amplification of viral DNA genomes using phi29 polymerase. Trends in Microbiology,
17, 205-211.

51.

Li, X., Zhang, S., Zhang, H., Zhang, L., Tao, H., Yu, J., Zheng, W., Liu, C., Lü, D.,
Xiang, R. et al. (2009) A loop-mediated isothermal amplification method targets the
phoP gene for the detection of Salmonella in food samples. International Journal of Food
Microbiology, 133, 252-258.

52.

Mori, Y., Kitao, M., Tomita, N. and Notomi, T. (2004) Real-time turbidimetry of LAMP
reaction for quantifying template DNA. Journal of Biochemical and Biophysical
Methods, 59, 145-157.

53.

Yoshida, A., Nagashima, S., Ansai, T., Tachibana, M., Kato, H., Watari, H., Notomi, T.
and Takehara, T. (2005) Loop-mediated isothermal amplification method for rapid
detection of the periodontopathic bacteria Porphyromoas ginvivalis, Tannerella forsythia,
and Treponema denticola. Journal of Clinical Microbiology, 43, 2418-2424.

54.

Mori, Y., Nagamine, K., Tomita, N. and Notomi, T. (2001) Detection of loop-mediated
isothermal amplification reaction by turbidity derived from magnesium pyrophosphate
formation. Biochemical and Biophysical Research Communications, 289, 150-154.

55.

Peyrefitte, C.N., Boubis, L., Coudrier, D., Bouloy, M., Grandadam, M., Tolou, H.J. and
Plumet, S. (2008) Real-time reverse-transcription loop-mediated isothermal amplification
for rapid detection of Rift Valley fever virus. Journal of Clinical Microbiology, 46, 36533659.

56.

Aoi, Y., Hosogai, M. and Tsuneda, S. (2006) Real-time quantitative LAMP (loopmediated isothermal amplification of DNA) as a simple method for monitoring ammoniaoxidizing bacteria. Journal of Biotechnology, 125, 484-491.

57.

Njiru, Z.K., Mikosza, A.S.J., Armstrong, T., Enyaru, J.C., Ndung'u, J.M. and Thompson,
A.R.C. (2008) Loop-Mediated Isothermal Amplification (LAMP) Method for Rapid
Detection of Trypanosoma brucei rhodesiense. PLoS Neglected Tropical Diseases, 2.

58.

Tomita, N., Mori, Y., Kanda, H. and Notomi, T. (2008) Loop-mediated isothermal
amplification (LAMP) of gene sequences and simple visual detection of products. Nature
Protocols, 3, 877-882.

59.

Parida, M., Sannarangaiah, S., Dash, P.K., Rao, P.V. and Morita, K. (2008) Loop
mediated isothermal amplification (LAMP): a new generation of innovative gene
amplification technique; perspectives in clinical diagnosis of infectious diseases. Reviews
of Medical Virology, 18, 407-421.

153

60.

Mori, Y. and Notomi, T. (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.

61.

Yamazaki, W., Seto, K., Taguchi, M., Ishibashi, M. and Inoue, K. (2008) Sensitive and
rapid detection of cholera toxin-producing Vibrio cholerae using a loop-mediated
isothermal amplification. BMC Microbiology, 8, 94.

62.

Yamazaki, W., Ishibashi, M., Kawahara, R. and Inoue, K. (2008) Development of a loopmediated isothermal amplification assay for sensitive and rapid detection of Vibrio
parahaemolyticus. BMC Microbiology, 8, 163.

63.

Kamachi, K., Toyoizumi-Ajisaka, H., Toda, K., Soeung, S.C., Sarath, S., Nareth, Y.,
Horiuchi, Y., Kojima, K., Takahashi, M. and Arakawa, Y. (2006) Development and
evaluation of a loop-mediated isothermal amplification method for rapid diagnosis of
Bordetella pertussis infection. Journal of Clinical Microbiology, 44, 1899-1902.

64.

Parida, M.M., Santhosh, S.R., Dash, P.K., Tripathi, N.K., Lakshmi, V., Mamidi, N.,
Shrivastva, A., Gupta, N., Saxena, P., Babu, J.P. et al. (2007) Rapid and real-time
detection of Chikungunya virus by reverse transcription loop-mediated isothermal
amplification assay. Journal of Clinical Microbiology, 45, 351-357.

65.

Hara-Kudo, Y., Yoshino, M., Kojima, T. and Ikedo, M. (2005) Loop-mediated isothermal
amplification for the rapid detection of Salmonella. Fems Microbiology Letters, 253, 155161.

66.

Okamura, M., Ohba, Y., Kikuchi, S., Takehara, K., Ikedo, M., Kojima, T. and Nakamura,
M. (2009) Rapid, sensitive, and specific detection of the O4 group of Salmonella enterica
by loop-mediated isothermal amplification. Avian Diseases, 53, 216-221.

67.

Xie, S., Yang, J., Chen, C., Zhang, X., Wang, Q. and Zhang, C. (2008) Study on
biosorption kinetics and thermodynamics of uranium by Citrobacter freudii. Journal of
Environmental Radioactivity, 126-133.

68.

Takagi, H., Itoh, M., Islam, M., Razzaque, A., Ekram, A., Hashighuchi, Y., Noiri, E. and
Kimura, E. (2009) Sensitive, specific, and rapid detection of Leishmania donovani DNA
by loop-mediated isothermal amplification. American Journal of Tropical Diseases and
Hygiene, 81, 578-582.

69.

Nemoto, J., Sugawara, C., Akahane, K., Hashimoto, K., Kojima, T., Ikedo, M., Konuma,
H. and Hara-Kudo, Y. (2009) Rapid and specific detection of the thermostable direct
hemolysin gene in Vibrio parahaemolyticus by loop-mediated isothermal amplification.
Journal of Food Protection, 72, 748-754.

154

70.

Bakheit, M., Torra, D., Palomino, L., Thekisoe, O., Mbati, P., Ongerth, J. and Karanis, P.
(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 Parasitolology, 158, 11-22.

71.

Sotiriadou, I. and Karanis, P. (2008) Evaluation of loop-mediated isothermal
amplification for detection of Toxoplasma gondii in water samples and comparative
findings by polymerase chain reaction and immunofluorescence test (IFT). Diagnostic
Microbiology and Infectious Disease, 62, 357-365.

72.

Poon, L., Wong, B., Ma, E., Chan, K., Chow, L., Abeyewickreme, W., Tangpukdee, N.,
Yuen, K., Guan, Y., Looareesuwan, S. et al. (2006) Sensitive and inexpensive molecular
test for falciparum malaria: detecting Plasmodium falciparum DNA directly from heattreated blood by loop-mediated isothermal amplification. Clinical Chemistry, 52, 303306.

73.

Kaneko, H., Kawana, T., Fukushima, E. and Suzutani, T. (2007) Tolerance of loopmediated isothermal amplification to a culture medium and biological substances.
Journal of Biochemical and Biophysical Methods, 70, 499-501.

74.

Moslemi, E., Shahhosseiny, M.H., Javadi, G., Praivar, K., Sattari, T.N. and Amini, H.K.
(2009) Loop mediated isothermal amplification (LAMP) for rapid detection of HBV in
Iran. African Journal of Microbiology Research, 3, 439-445.

75.

Seki, M., Poulsen, K., Haubek, D. and Kilian, M. (2008) Novel loop-mediated isothermal
amplification method for detection of the JP2 clone of Aggregatibacter
actinomycetemcomitans in subgingival plaque. Journal of Clinical Microbiology, 46,
1113-1115.

76.

Tani, H., Teramura, T., Adachi, K., Tsuneda, S., Kurata, S., Nakamura, K., Kanagawa, T.
and Noda, N. (2007) Technique for quantitative detection of specific DNA sequences
using alternately binding quenching probe competitive assay combined with loopmediated isothermal amplification. Analytical Chemistry, 79, 5608-5613.

77.

Maruyama, F., Kenzaka, T., Yamaguchi, N., Tani, K. and Nasu, M. (2003) Detection of
bacteria carrying the stx2 gene by in situ loop-mediated isothermal amplification. Applied
and Environmental Microbiology, 69, 5023-5028.

78.

Han, E.T., Watanabe, R., Sattabongkot, J., Khuntirat, B., Sirichaisinthop, J., Iriko, H., Jin,
L., Takeo, S. and Tsuboi, T. (2007) Detection of four Plasmodium species by genus- and
species-specific loop-mediated isothermal amplification for clinical diagnosis. Journal of
Clinical Microbiology, 45, 2521-2528.

155

79.

Karanis, P., Thekisoe, O., Kiouptsi, K., Ongerth, J., Igarashi, I. and Inoue, N. (2007)
Development and preliminary evaluation of a loop-mediated isothermal amplification
procedure for sensitive detection of Cryptosporidium oocysts in fecal and water samples.
Applied and Environmental Microbiology, 73, 5660-5662.

80.

Stedtfeld, R., Tourlousse, D., Seyrig, G., Stedtfeld, T., Kronlein, M., Price, S., Ahmad,
F., Gulari, E., Tiedje, J. and Hashsham, S. (2012) Gene-Z: A Device for Point of Care
Genetic Testing using a Smartphone. Lab on a Chip, 12, 1454-1462

81.

Cai, T., Lou, G.Q., Yang, J., Xu, D. and Meng, Z.H. (2008) Development and evaluation
of real-time loop-mediated isothermal amplification for hepatitis B virus DNA
quantification: A new tool for HBV management. Journal of Clinical Virology, 41, 270276.

82.

Kaneko, H., Iida, T., Aoki, K., Ohno, S. and Suzutani, T. (2005) Sensitive and rapid
detection of herpes simplex virus and varicella-zoster virus DNA by loop-mediated
isothermal amplification. Journal of Clinical Microbiology, 43, 3290-3296.

83.

Kimura, H., Ihira, M., Enomoto, Y., Kawada, J., Ito, Y., Morishima, T., Yoshikawa, T.
and Asano, Y. (2005) Rapid detection of herpes simplex virus DNA in cerebrospinal
fluid: comparison between loop-mediated isothermal amplification and real-time PCR.
Medical Microbiology and Immunology, 194, 181-185.

84.

Ihira, M., Yoshikawa, T., Enomoto, Y., Akimoto, S., Ohashi, M., Suga, S., Nishimura,
N., Ozaki, T., Nishiyama, Y., Notomi, T. et al. (2004) Rapid diagnosis of human
herpesvirus 6 infection by a novel DNA amplification method, loop-mediated isothermal
amplification. Journal of Clinical Microbiology, 42, 140-145.

85.

Mekata, T., Sudhakaran, R., Kono, T., Supamattaya, K., Linh, N.T., Sakai, M. and Itami,
T. (2009) Real-time quantitative loop-mediated isothermal amplification as a simple
method for detecting white spot syndrome virus. Letters in Applied Microbiology, 48, 2532.

86.

Fukuda, S., Takao, S., Kuwayama, M., Shimazu, Y. and Miyazaki, K. (2006) Rapid
detection of norovirus from fecal specimens by real-time reverse transcription-loopmediated isothermal amplification assay. Journal of Clinical Microbiology, 44, 13761381.

87.

Hong, T.C., Mai, Q.L., Cuong, D.V., Parida, M., Minekawa, H., Notomi, T., Hasebe, F.
and Morita, K. (2004) Development and evaluation of a novel loop-mediated isothermal
amplification method for rapid detection of severe acute respiratory syndrome
coronavirus. Journal of Clinical Microbiology, 42, 1956-1961.

156

88.

Toriniwa, H. and Komiya, T. (2006) Rapid detection and quantification of Japanese
encephalitis virus by real-time reverse transcription loop-mediated isothermal
amplification. Microbiology and Immunology, 50, 379-387.

89.

Yoda, T., Suzuki, Y., Yamazaki, K., Sakon, N., Kanki, M., Kase, T., Takahashi, K. and
Inoue, K. (2009) Application of a modified loop-mediated isothermal amplification kit
for detecting norovirus genogroups I and II. Journal of Medical Virology, 81, 2072-2078.

90.

Straub, T.M. and Chandler, D.P. (2003) Towards a unified system for detecting
waterborne pathogens. Journal of Microbiological Methods, 53, 185-197.

91.

Morales-Morales, H.A., Vidal, G., Olszewski, J., Rock, C.M., Dasgupta, D., Oshima,
K.H. and Smith, G.B. (2003) Optimization of a reusable hollow-fiber ultrafilter for
simultaneous concentration of enteric bacteria, protozoa, and viruses from water. Applied
and Environmental Microbiology, 69, 4098-4102.

92.

Polaczyk, A.L., Narayanan, J., Cromeans, T.L., Hahn, D., Roberts, J.M., Amburgey, J.E.
and Hill, V.R. (2008) Ultrafiltration-based techniques for rapid and simultaneous
concentration of multiple microbe classes from 100-L tap water samples. Journal of
Microbiological Methods, 73, 92-99.

93.

Smith, C.M. and Hill, V.R. (2009) Dead-End Hollow-Fiber Ultrafiltration for Recovery
of Diverse Microbes from Water. Applied and Environmental Microbiology, 75, 52845289.

94.

Iseki, H., Alhassan, A., Ohta, N., Thekisoe, O.M., Yokoyama, N., Inoue, N., Nambota,
A., Yasuda, J. and Igarashi, I. (2007) Development of a multiplex loop-mediated
isothermal amplification (mLAMP) method for the simultaneous detection of bovine
Babesia parasites. Journal of Microbiological Methods, 71, 281-287.

95.

Yoneyama, T., Kiyohara, T., Shimasaki, N., Kobayashi, G., Ota, Y., Notomi, T., Totsuka,
A. and Wakita, T. (2007) Rapid and real-time detection of hepatitis A virus by reverse
transcription loop-mediated isothermal amplification assay. Journal of Virological
Methods, 145, 162-168.

96.

Karanis, P. and Ongerth, J. (2009) LAMP - a powerful and flexible tool for monitoring
microbial pathogens. Trends in Parasitology, 25, 498-499.

97.

Liu, R.H., Yang, J.N., Lenigk, R., Bonanno, J. and Grodzinski, P. (2004) Self-contained,
fully integrated biochip for sample preparation, polymerase chain reaction amplification,
and DNA microarray detection. Analytical Chemistry, 76, 1824-1831.

157

98.

Lien, K.Y., Lee, W.C., Lei, H.Y. and Lee, G.B. (2007) Integrated reverse transcription
polymerase chain reaction systems for virus detection. Biosensors & Bioelectronics, 22,
1739-1748.

99.

Furdui, V.I. and Harrison, D.J. (2004) Immunomagnetic T cell capture from blood for
PCR analysis using microfluidic systems. Lab on a Chip, 4, 614-618.

100.

Beyor, N., Seo, T.S., Liu, P. and Mathies, R.A. (2008) Immunomagnetic bead-based cell
concentration microdevice for dilute pathogen detection. Biomedical Microdevices, 10,
909-917.

101.

Voldman, J. (2006) Electrical forces for microscale cell manipulation. Annual Review of
Biomedical Engineering, 8, 425-454.

102.

Meighan, M.M., Staton, S.J.R. and Hayes, M.A. (2009) Bioanalytical separations using
electric field gradient techniques. Electrophoresis, 30, 852-865.

103.

Lapizco-Encinas, B.H., Davalos, R.V., Simmons, B.A., Cummings, E.B. and
Fintschenko, Y. (2005) An insulator-based (electrodeless) dielectrophoretic concentrator
for microbes in water. Journal of Microbiological Methods, 62, 317-326.

104.

Wu, V.C.H. and Wu, C.C. (2008) Filtration and removal of foodborne pathogens using
dielectrophoretic phenomena. Journal of Rapid Methods and Automation in
Microbiology, 16, 62-72.

105.

Bhattacharya, S., Salamat, S., Morisette, D., Banada, P., Akin, D., Liu, Y.S., Bhunia,
A.K., Ladisch, M. and Bashir, R. (2008) PCR-based detection in a micro-fabricated
platform. Lab on a Chip, 8, 1130-1136.

106.

Kim, J., Johnson, M., Hill, P. and Gale, B.K. (2009) Microfluidic sample preparation: cell
lysis and nucleic acid purification. Integrative Biology, 1, 574-586.

107.

Lu, H., Schmidt, M.A. and Jensen, K.F. (2005) A microfluidic electroporation device for
cell lysis. Lab on a Chip, 5, 23-29.

108.

Wang, H.Y., Bhunia, A.K. and Lu, C. (2006) A microfluidic flow-through device for
high throughput electrical lysis of bacterial cells based on continuous dc voltage.
Biosensors & Bioelectronics, 22, 582-588.

109.

Park, K., Akin, D. and Bashir, R. (2007) Electrical capture and lysis of vaccinia virus
particles using silicon nano-scale probe array. Biomedical Microdevices, 9, 877-883.

158

110.

Taylor, M.T., Belgrader, P., Furman, B.J., Pourahmadi, F., Kovacs, G.T.A. and Northrup,
M.A. (2001) Lysing bacterial spores by sonication through a flexible interface in a
microfluidic system. Analytical Chemistry, 73, 492-496.

111.

Di Carlo, D., Jeong, K.H. and Lee, L.P. (2003) Reagentless mechanical cell lysis by
nanoscale barbs in microchannels for sample preparation. Lab on a Chip, 3, 287-291.

112.

Kim, J., Jang, S.H., Jia, G.Y., Zoval, J.V., Da Silva, N.A. and Madou, M.J. (2004) Cell
lysis on a microfluidic CD (compact disc). Lab on a Chip, 4, 516-522.

113.

Kido, H., Micic, M., Smith, D., Zoval, J., Norton, J. and Madou, M. (2007) A novel,
compact disk-like centrifugal microfluidics system for cell lysis and sample
homogenization. Colloids and Surfaces B-Biointerfaces, 58, 44-51.

114.

Cho, Y.K., Lee, J.G., Park, J.M., Lee, B.S., Lee, Y. and Ko, C. (2007) One-step pathogen
specific DNA extraction from whole blood on a centrifugal microfluidic device. Lab on a
Chip, 7, 565-573.

115.

Lee, J.G., Cheong, K.H., Huh, N., Kim, S., Choi, J.W. and Ko, C. (2006) Microchipbased one step DNA extraction and real-time PCR in one chamber for rapid pathogen
identification. Lab on a Chip, 6, 886-895.

116.

de la Rosa, C., Tilley, P.A., Fox, J.D. and Kaler, K. (2008) Microfluidic Device for
Dielectrophoresis Manipulation and Electrodisruption of Respiratory Pathogen
Bordetella pertussis. IEEE Transactions on Biomedical Engineering, 55, 2426-2432.

117.

Price, C.W., Leslie, D.C. and Landers, J.P. (2009) Nucleic acid extraction techniques and
application to the microchip. Lab on a Chip, 9, 2484-2494.

118.

Wen, J., Legendre, L.A., Bienvenue, J.M. and Landers, J.P. (2008) Purification of nucleic
acids in microfluidic devices. Analytical Chemistry, 80, 6472-6479.

119.

Mahalanabis, M., Al-Muayad, H., Kulinski, M.D., Altman, D. and Klapperich, C.M.
(2009) Cell lysis and DNA extraction of gram-positive and gram-negative bacteria from
whole blood in a disposable microfluidic chip. Lab on a Chip, 9, 2811-2817.

120.

Chung, Y.C., Jan, M.S., Lin, Y.C., Lin, J.H., Cheng, W.C. and Fan, C.Y. (2004)
Microfluidic chip for high efficiency DNA extraction. Lab on a Chip, 4, 141-147.

121.

Witek, M.A., Llopis, S.D., Wheatley, A., McCarley, R.L. and Soper, S.A. (2006)
Purification and preconcentration of genomic DNA from whole cell lysates using
photoactivated polycarbonate (PPC) microfluidic chips. Nucleic Acids Research, 34, e74.

159

122.

Cao, W.D., Easley, C.J., Ferrance, J.P. and Landers, J.P. (2006) Chitosan as a polymer
for pH-induced DNA capture in a totally aqueous system. Analytical Chemistry, 78,
7222-7228.

123.

Nakagawa, T., Tanaka, T., Niwa, D., Osaka, T., Takeyama, H. and Matsunaga, T. (2005)
Fabrication of amino silane-coated microchip for DNA extraction from whole blood.
Journal of Biotechnology, 116, 105-111.

124.

Legendre, L.A., Morris, C.J., Bienvenue, J.M., Barron, A., McClure, R. and Landers, J.P.
(2008) Toward a simplified microfluidic device for ultra-fast genetic analysis with
sample in/answer out capability: application to T-cell lymphoma diagnosis. Journal of the
Association for Laboratory Automation, 13, 351-360.

125.

Lien, K.Y., Liu, C.J., Kuo, P.L. and Lee, G.B. (2009) Microfluidic System for Detection
of alpha-Thalassemia-1 Deletion Using Saliva Samples. Analytical Chemistry, 81, 45024509.

126.

Lien, K.Y., Liu, C.J., Lin, Y.C., Kuo, P.L. and Lee, G.B. (2009) Extraction of genomic
DNA and detection of single nucleotide polymorphism genotyping utilizing an integrated
magnetic bead-based microfluidic platform. Microfluidics and Nanofluidics, 6, 539-555.

127.

Marzialli, A., Pel, J., Bizzotto, D. and Whitehead, L.A. (2005) Novel electrophoresis
mechanism based on synchronous alternating drag perturbation. Electrophoresis, 26, 8290.

128.

Pel, J., Broemeling, D., Mai, L., Poon, H.L., Tropini, G., Warren, R.L., Holt, R.A. and
Marziali, A. (2009) Nonlinear electrophoretic response yields a unique parameter for
separation of biomolecules. Proceedings of the National Academy of Sciences of the
United States of America, 106, 14796-14801.

129.

Nakamura, N., Ito, K., Takahashi, M., Hongo, S., Hashimoto, K., Kawamoto, M.,
Taniguchi, A., Kamatani, N. and Gemma, N. (2009) Clinical verification of a
combination technology of a loop-mediated isothermal amplification method and an
electrochemical DNA chip for personalized medicine. Clinical Biochemistry, 42, 11581161.

130.

Lucchi, N.W., Demas, A., Narayanan, J., Sumari, D., Kabanywanyi, A., Kachur, S.P.,
Barnwell, J.W. and Udhayakumar, V. (2010) Real-Time Fluorescence Loop Mediated
Isothermal Amplification for the Diagnosis of Malaria. PLoS One, 5, 7.

131.

Grab, D.J., Nikolskaia, O.V., Inoue, N., Thekisoe, O.M., Morrison, L.J., Gibson, W. and
Dumler, J.S. (2011) Using Detergent to Enhance Detection Sensitivity of African
Trypanosomes in Human CSF and Blood by Loop-Mediated Isothermal Amplification
(LAMP). PLoS Neglected Tropical Diseases, 5, e1249.
160

132.

Gill, P. and Ghaemi, A. (2008) Nucleic acid isothermal amplification technologies: a
review. Nucleosides Nucleotides Nucleic Acids, 27, 224-243.

133.

Chen, S.Y. and Ge, B.L. (2010) Development of a toxR-based loop-mediated isothermal
amplification assay for detecting Vibrio parahaemolyticus. BMC Microbiology, 10, 41

134.

Han, F.F. and Ge, B.L. (2010) Quantitative detection of Vibrio vulnificus in raw oysters
by real-time loop-mediated isothermal amplification. International Journal of Food
Microbiology, 142, 60-66.

135.

Seyrig, G., Ahmad, F., Stedtfeld, R.D., Tourlousse, D.M. and Hashsham, S.A. (2010),
Environmental Microbiology: Current Technology and Water Applications. Horizon
Scientific Press, Norwich, UK.

136.

Ahmad, F., Seyrig, G., Tourlousse, D.M., Stedtfeld, R.D., Tiedje, J.M. and Hashsham,
S.A. (2011) A CCD-based fluorescence imaging system for real-time loop-mediated
isothermal amplification-based rapid and sensitive detection of waterborne pathogens on
microchips. Biomedical Microdevices.

137.

Kikuchi, T., Aikawa, T., Oeda, Y., Karim, N. and Kanzaki, N. (2009) A Rapid and
Precise Diagnostic Method for Detecting the Pinewood Nematode Bursaphelenchus
xylophilus by Loop-Mediated Isothermal Amplification. Phytopathology, 99, 1365-1369.

138.

Lin, X.A., Chen, Y., Lu, Y.Y., Yan, J.Y. and Yan, J. (2009) Application of a loopmediated isothermal amplification method for the detection of pathogenic Leptospira.
Diagnostic Microbiology and Infectious Disease, 63, 237-242.

139.

Gudnason, H., Dufva, M., Bang, D.D. and Wolff, A. (2007) Comparison of multiple
DNA dyes for real-time PCR: effects of dye concentration and sequence composition on
DNA amplification and melting temperature. Nucleic Acids Research, 35, e127.

140.

Stedtfeld, R.D., Baushke, S.W., Tourlousse, D.M., Miller, S.M., Stedtfeld, T.M., Gulari,
E., Tiedje, J.M. and Hashsham, S.A. (2008) Development and experimental validation of
a predictive threshold cycle equation for quantification of virulence and marker genes by
high-throughput nanoliter-volume PCR on the OpenArray platform. Applied and
Environmental Microbiology, 74, 3831-3838.

141.

Lam, L., Sakakihara, S., Ishizuka, K., Takeuchi, S., Arata, H.F., Fujita, H. and Noji, H.
(2008) Loop-mediated isothermal amplification of a single DNA molecule in
polyacrylamide gel-based microchamber. Biomedical Microdevices, 10, 539-546.

161

142.

Tomlinson, J., Barker, I. and Boonham, N. (2007) Faster, simpler, more-specific methods
for improved molecular detection of Phytophthora ramorum in the field. Applied and
Environmental Microbiology, 73, 4040-4047.

143.

Dukes, J.P., King, D.P. and Alexandersen, S. (2006) Novel reverse transcription loopmediated isothermal amplification for rapid detection of foot-and-mouth disease virus.
Archives of Virology, 151, 1093-1106.

144.

Zhang, Y. and Ozdemir, P. (2009) Microfluidic DNA amplification-A review. Analytica
Chimica Acta, 638, 115-125.

145.

Nagamine, K., Hase, T. and Notomi, T. (2002) Accelerated reaction by loop-mediated
isothermal amplification using loop primers. Molecular and Cellular Probes, 16, 223229.

146.

Ohtsuki, R., Kawamoto, K., Kato, Y., Shah, M.M., Ezaki, T. and Makino, S.I. (2008)
Rapid detection of Brucella spp. by the loop-mediated isothermal amplification method.
Journal of Applied Microbiology, 104, 1815-1823.

147.

Tomlinson, J.A., Dickinson, M.J. and Boonham, N. (2010) Detection of Botrytis cinerea
by loop-mediated isothermal amplification. Letters in Applied Microbiology, 51, 650657.

148.

Uemura, N., Makimura, K., Onozaki, M., Otsuka, Y., Shibuya, Y., Yazaki, H., Kikuchi,
Y., Abe, S. and Kudoh, S. (2008) Development of a loop-mediated isothermal
amplification method for diagnosing Pneumocystis pneumonia. Journal of Medical
Microbiology, 57, 50-57.

149.

Mullis, K.B., Ferrâe, F. and Gibbs, R. (1994) The Polymerase chain reaction. Birkhauser,
Boston.

150.

Klatser, P., Kuijper, S., van Ingen, C. and Kolk, A. (1998) Stabilized, freeze-dried PCR
mix for detection of mycobacteria. Journal of Clinical Microbiology, 36, 1798-1800.

151.

Tomlinson, J., Boonham, N., Hughes, K., Griffen, R. and Barker, I. (2005) On-site DNA
extraction and real-time PCR for detection of Phytophthora ramorum in the field.
Applied and Environmental Microbiology, 71, 6702-6710.

152.

Colaço, C., Sen, S., Thangavelu, M., Pinder, S. and Roser, B. (1992) Extraordinary
stability of enzymes dried in trehalose: simplified molecular biology. Biotechnology (N
Y), 10, 1007-1011.

162

153.

Crowe, J., Crowe, L., Oliver, A., Tsvetkova, N., Wolkers, W. and Tablin, F. (2001) The
trehalose myth revisited: Introduction to a symposium on stabilization of cells in the dry
state. Cryobiology, 43, 89-105.

154.

Spiess, A., Mueller, N. and Ivell, R. (2004) Trehalose is a potent PCR enhancer: lowering
of DNA melting temperature and thermal stabilization of Taq polymerase by the
disaccharide trehalose. Clinical Chemistry, 50, 1256-1259.

155.

Stedtfeld, R., Baushke, S., Tourlousse, D., Miller, S., Stedtfeld, T., Gulari, E., Tiedje, J.
and Hashsham, S. (2008) Development and experimental validation of a predictive
threshold cycle equation for quantification of virulence and marker genes by highthroughput nanoliter-volume PCR on the OpenArray platform. Applied and
Environmental Microbiology, 74, 3831-3838.

156.

von Graberg, S. and Gieseler, H. (2006), Proc. CPPR Freeze-Drying of Pharmaceuticals
and Biologicals, Garmisch-Partenkirchen, Germany, Oct. 3-6.

157.

Pu, Q., Oyesanya, O., Thompson, B., Liu, S. and Alvarez, J. (2007) On-chip
micropatterning of plastic (cylic olefin copolymer, COC) microfluidic channels for the
fabrication of biomolecule microarrays using photografting methods. Langmuir, 23,
1577-1583.

158.

Jain, N. and Roy, I. (2009) Effect of trehalose on protein structure. Protein Science, 18,
24-36.

159.

Garcia, E., Kirkham, J., Hatch, A., Hawkins, K. and Yager, P. (2004) Controlled
microfluidic reconstitution of functional protein from an anhydrous storage depot. Lab on
a Chip, 4, 78-82.

160.

Furuberg, L., Mielnik, M., Gulliksen, A., Solli, L., Johansen, I., Voitel, J., Baier, T.,
Riegger, L. and Karlsen, F. (2008) RNA amplification chip with parallel microchannels
and droplet positioning using capillary valves. Microsystem Technologies-Micro-and
Nanosystems-Information Storage and Processing Systems, 673-681.

161.

Fang, X.E., Liu, Y.Y., Kong, J.L. and Jiang, X.Y. (2010) Loop-Mediated Isothermal
Amplification Integrated on Microfluidic Chips for Point-of-Care Quantitative Detection
of Pathogens. Analytical Chemistry, 82, 3002-3006.

162.

Ramalingam, N., Rui, Z., Liu, H.B., Dai, C.C., Kaushik, R., Ratnaharika, B. and Gong,
H.Q. (2010) Real-time PCR-based microfluidic array chip for simultaneous detection of
multiple waterborne pathogens. Sensors and Actuators B-Chemical, 145, 543-552.
163

163.

Cady, N.C., Stelick, S., Kunnavakkam, M.V. and Batt, C.A. (2005) Real-time PCR
detection of Listeria monocytogenes using an integrated microfluidics platform. Sensors
and Actuators B-Chemical, 107, 332-341.

164

Kwok, P.-Y. and Chen, X. (2003) Detection of single nucleotide polymorphisms. Current
issues in molecular biology, 5, 43-60.

165.

Kim, S. and Misra, A. (2007) SNP genotyping: Technologies and biomedical
applications. Annual Review of Biomedical Engineering, 9, 289-320.

166.

Sandgren, A., Strong, M., Muthukrishnan, P., Weiner, B., Church, G. and Murray, M.
(2009) Tuberculosis Drug Resistance Mutation Database. PLoS Medicine, 6, 132-136.

167.

Frackman, S., Kobs, G., Simpson, D. and Storts, D. (1998) Betaine and DMSO:
Enhancing Agentsfor PCR. Promega Notes, 65, 27-30.

168.

Jensen, M., Fukushima, M. and Davis, R. (2010) DMSO and Betaine Greatly Improve
Amplification of GC-Rich Constructs in De Novo Synthesis. PLoS One, 5.

169.

You, Y., Moreira, B.G., Behlke, M.A. and Owczarzy, R. (2006) Design of LNA probes
that improve mismatch discrimination. Nucleic Acids Research, 34, e60.

170

Ramirez M, Cowart K, Campbell P, Morlock G, Sikes D, Winchell J, Posey J. (2010)
Rapid detection of multidrug-resistant Mycobacterium tuberculosis by use of real-time
PCR and high-resolution melt analysis. Journal of Clinical Microbiology. 48, 4003-9.

171

Brugè F, Littarru GP, Silvestrini L, Mancuso T, Tiano L. (2009) A novel Real Time PCR
strategy to detect SOD3 SNP using LNA probes. Mutation Research, 669,80-4.

172

http://www.neb.com/nebecomm/tech_reference/TmCalc/Default.asp#.T3S94GH20SA

164