Article

pubs.acs.org/accounts

Integrated Electrochemical Microsystems for Genetic Detection of

Pathogens at the Point of Care
Kuangwen Hsieh,†,# B. Scott Ferguson,†,∇ Michael Eisenstein,‡ Kevin W. Plaxco,§,∥
and H. Tom Soh*,†,‡,§,⊥
†
Department of Mechanical Engineering, ‡Institute for Collaborative Biotechnologies, §Interdepartmental Program in Biomolecular
Science and Engineering, ∥Department of Chemistry and Biochemistry, and ⊥Materials Department, University of California, Santa
Barbara, Santa Barbara, California 93106, United States

CONSPECTUS: The capacity to achieve rapid, sensitive,

specific, quantitative, and multiplexed genetic detection of
pathogens via a robust, portable, point-of-care platform could
transform many diagnostic applications. And while contem-
porary technologies have yet to effectively achieve this goal,
the advent of microfluidics provides a potentially viable
approach to this end by enabling the integration of
sophisticated multistep biochemical assays (e.g., sample
preparation, genetic amplification, and quantitative detection)
in a monolithic, portable device from relatively small biological
samples.
Integrated electrochemical sensors offer a particularly promising solution to genetic detection because they do not require optical
instrumentation and are readily compatible with both integrated circuit and microfluidic technologies. Nevertheless, the
development of generalizable microfluidic electrochemical platforms that integrate sample preparation and amplification as well
as quantitative and multiplexed detection remains a challenging and unsolved technical problem. Recognizing this unmet need,
we have developed a series of microfluidic electrochemical DNA sensors that have progressively evolved to encompass each of
these critical functionalities.
For DNA detection, our platforms employ label-free, single-step, and sequence-specific electrochemical DNA (E-DNA) sensors,
in which an electrode-bound, redox-reporter-modified DNA “probe” generates a current change after undergoing a hybridization-
induced conformational change. After successfully integrating E-DNA sensors into a microfluidic chip format, we subsequently
incorporated on-chip genetic amplification techniques including polymerase chain reaction (PCR) and loop-mediated isothermal
amplification (LAMP) to enable genetic detection at clinically relevant target concentrations. To maximize the potential point-of-
care utility of our platforms, we have further integrated sample preparation via immunomagnetic separation, which allowed the
detection of influenza virus directly from throat swabs and developed strategies for the multiplexed detection of related bacterial
strains from the blood of septic mice. Finally, we developed an alternative electrochemical detection platform based on real-time
LAMP, which not is only capable of detecting across a broad dynamic range of target concentrations, but also greatly simplifies
quantitative measurement of nucleic acids.
These efforts represent considerable progress toward the development of a true sample-in−answer-out platform for genetic
detection of pathogens at the point of care. Given the many advantages of these systems, and the growing interest and innovative
contributions from researchers in this field, we are optimistic that iterations of these systems will arrive in clinical settings in the
foreseeable future.

■ INTRODUCTION
Genetic detection of viruses and bacteria at the “point of care”
sensitive and specific genetic detection directly from
unprocessed samples at the point of care, as evidenced by the
are critically needed in food safety testing,1 in environmental lack of Food and Drug Administration (FDA)-approved
monitoring,2 and, most importantly, in clinical diagnostics.3,4 systems.
Ideally, such detection should be sensitive, specific, quantitative, The advent of microfluidics has lowered the technology
and capable of multiplexing, which necessitates multistep assays barrier to effective, point-of-care genetic detection because it
that incorporate procedures for sample preparation, genetic enables the development of miniaturized and disposable devices
amplification, and quantitative detection. Point-of-care use adds that can perform multistep biochemical assays from small
still further technical challenges, as this requires that the sample volumes with minimal sample loss and rapid assay time
complex, multistep procedures be performed rapidly and in a
portable, robust, and user-friendly platform.5,6 As a result, Received: December 22, 2014
contemporary approaches have yet to achieve acceptably

© XXXX American Chemical Society A DOI: 10.1021/ar500456w

Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

Figure 1. Evolution of integrated electrochemical microsystems for point-of-care genetic detection. Building on a foundation of microscale sensors
for sequence-specific electrochemical DNA (E-DNA) target detection, we have subsequently integrated other critical on-chip functionalities,
including sample preparation, DNA/RNA amplification, quantitative, real-time detection, and multiplexed detection. Integrated electrochemical
microsystems that encompass these functionalities present a promising solution for genetic detection of pathogens at the point of care.

(see reviews by Chen et al.7 and Park et al.8). Unfortunately,

most existing examples rely on delicate and bulky optical
instrumentation for detection, rendering them relatively poorly
suited for point-of-care applications. Electrochemical sensors, in
contrast, offer a promising alternative for simplified genetic
detection as they eliminate the need for optical equipment, are
highly amenable to miniaturization, and can be easily interfaced
with integrated circuits and electronic instruments.9,10 Never-
theless, the development of generalizable microfluidic electro-
chemical platforms that integrate sample preparation and
amplification as well as quantitative and multiplexed detection
remains a challenging and unsolved problem.
Motivated by the above observations, we have developed a
series of microfluidic electrochemical DNA sensors that
progressively incorporate integrated sample preparation, on-
chip amplification, real-time measurement, and the capacity for
multiplexed detection and differentiation of pathogen strains
(Figure 1). In this Account, we share our perspectives on
developing such microsystems toward point-of-care use, and
illustrate how our developments fit in the broader context of Figure 2. Operating principle of E-DNA sensors. The E-DNA sensor
the field by highlighting relevant contemporary systems from comprises a redox-reporter-modified DNA probe attached to an
interrogating electrode. In the absence of target (left), the redox
the literature. Finally, we briefly discuss our current research reporter is held in proximity to the electrode, ensuring efficient
efforts and insights for future research directions.

■
electron transfer (eT) and a readily detectable current. The
hybridization of the target to the probe (middle) positions the
INTEGRATION OF E-DNA PROBES IN redox reporter away from the electrode, thereby reducing the current
MICROFLUIDIC CHIPS signal (right). The probe can subsequently be regenerated via rinsing
to restore the baseline current. Adapted from ref 13. Copyright 2006
The detection architecture we have employed in most of our PNAS.
platforms is the electrochemical DNA (E-DNA) sensor,11
which employs electrode-bound DNA probes modified with a
redox reporter (e.g., methylene blue, MB). Single-stranded Importantly, for our integrated microsystems, we have generally
target DNA hybridization changes the probe conformation and employed “signal-off” E-DNA sensors (i.e., target is detected by
thereby alters the electron transfer rate of the reporter, which a decrease in redox peak current) because they involve simple
can be detected via a change in redox peak current (Figure 2). and robust probe structures, can detect diverse DNA sequences
B DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

(such as unpurified polymerase chain reaction (PCR) microfluidic electrochemical DNA sensors (see reviews by
amplicons12) in a single step without labeling or exogenous Pumera et al.15 and Mir et al.16).

■
reagents,13 and can be efficiently regenerated via simple buffer
rinsing, which serves to validate that signal change is due to the INTEGRATION OF PCR WITH E-DNA
target.
A major focus of our group has been to integrate E-DNA The femtomolar or even attomolar concentrations of pathogen
sensors with a host of powerful microfluidic functionalities. DNA or RNA typically found in clinical samples falls below the
Doing so required that we develop processes for microfluidic detection limits of E-DNA and, indeed, of almost all proposed
chip and microelectrode fabrication, electrode cleaning and methods of (unamplified) DNA detection. Many research
preparation, electrochemical patterning and immobilization of groups have addressed this challenge by using PCR to amplify
E-DNA probes at designated sensor electrodes, sequence- target DNA prior to detection (see reviews by Luo and Hsing17
specific detection from multiple electrodes, and regeneration of and Duwensee et al.18), but most have kept PCR and detection
the sensing electrodes (Figure 3). All of these can now be as separate modules, rendering the process unsuitable for point-
of-care use. In response, we developed a monolithic micro-
fluidic system that combines PCR amplification, enzymatic
conversion of PCR amplicons into single-stranded DNA, and
E-DNA detection.
The Integrated Microfluidic Electrochemical DNA (IMED)
sensor19 comprises two modules: the reaction chamber and the
detection chamber. Genomic DNA and PCR reagents are
loaded into the reaction chamber with a syringe pump (Figure
4A), and on-chip PCR is performed using a temperature-
controlled thin-film heater (Figure 4B). The reverse primers
yield phosphorylated strands that are subsequently selectively
digested by lambda exonuclease20 within the reaction chamber,
efficiently producing single-stranded DNA (Figure 4C and 4D).
The sample and reagents are thoroughly mixed by syringe
pumping in and out of the PCR chamber through a dedicated
port. The single-stranded DNA is then mixed with a high-salt
buffer (Figure 4E) for optimal hybridization to E-DNA probes,
and pumped to the detection chamber for detection (Figure
4F).
To achieve successful operation of the IMED system
required that we overcome several hurdles. For example, in
order to obtain PCR efficiencies rivaling benchtop systems, we
minimized the internal surface area of PDMS in our reaction
chamber (known to cause enzyme adsorption), used the well-
known PCR-additive bovine serum albumin (BSA) to passivate
the chip surface and further reduce PCR inhibition due to
adsorption, and used precise sample loading with a syringe
pump to minimize air bubble formation (known to cause failure
of chip-based PCR). Similarly, to ensure accurate E-DNA
detection of PCR products and mitigate changes in peak
current arising from the adsorption of proteins present in the
PCR mix, we incubated the electrode surface with PCR mix
Figure 3. Specific detection of H5N1 and H1N1 in microfluidic E-
DNA sensor. H5N1 target causes a current decrease for the H5N1 containing BSA.
sensor (a) but not the H1N1 sensor (b). Conversely, H1N1 target The successful integration of PCR amplification, exonu-
yields negligible current change for the H5N1 sensor (c), but reduces clease-driven generation of single-stranded products, and,
the current at the H1N1 sensor (d). Reproduced with permission from ultimately, the detection via E-DNA into a single device
ref 14. Copyright 2008 American Chemical Society. allowed IMED to achieve unprecedented sensitivity. Using our
chips, we have detected genomic DNA from Salmonella enterica
serovar Typhimurium LT2 with a limit of detection (LOD) of
below 10 aM (∼300 copies in our 50 μL reaction chamber)
performed seamlessly within a single-chambered microfluidic (Figure 5). This is ∼2 orders of magnitude lower than that of
device.14 Using such a device, we demonstrated sequence- previously reported chip-based electrochemical methods,21−24
specific detection of DNA target sequences derived from as these platforms used less efficient, asymmetric PCR to
human (H1N1) and avian (H5N1) influenza virus (Figure 3). generate single-stranded products for detection.

■
When challenged with H5N1 target, only the H5N1 sensor
responded with a 38% signal change, while conversely, only the
H1N1 sensor responded to H1N1 target with a 46% signal INTEGRATION OF SAMPLE PREPARATION
change. We detected these targets in a high-salt buffer at a Although IMED greatly improved the sensitivity of on-chip
concentration of 400 nM, which is representative of single- pathogen detection, it nevertheless requires preassay isolation
stranded DNA concentrations achievable via asymmetric of pathogen DNA from samples. This increases the time
PCR.12 Our results were comparable to other contemporary required for diagnosis and creates additional work for end-users
C DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

Figure 4. IMED assay overview. (A) Template DNA is added to a PCR reagent mixture containing phosphorylated reverse primers and (B) PCR
amplified. (C) Lambda exonuclease is mixed with the product and (D) digests the phosphorylated strands. (E) MgCl2 is added to optimize
hybridization conditions. (F) Before introducing the sample, baseline sensor redox current is measured. Next, the single-stranded DNA product
hybridizes with the E-DNA probe, modulating the redox current signal. Finally, the E-DNA probe is regenerated to verify target hybridization.
Reprinted with permission from ref 19. Copyright 2009 American Chemical Society.

as well as opportunities for contamination or sample damage. tion, we used a similar workflow as in the IMED system to
Other existing electrochemical microsystems likewise exhibited perform on-chip RT-PCR, lambda exonuclease-mediated
a limited capacity for working directly with biological samples. single-strand generation, and sequence-specific E-DNA detec-
For example, Yamanaka combined reverse-transcriptase PCR tion.
(RT-PCR) and electrochemical sensors to detect influenza A, MIMED achieved highly sensitive H1N1 virus detection
but, as with IMED, the viral RNA was purified on the benchtop from throat swab samples within 3.5 h. Whereas virus-free
prior to on-chip amplification and detection.25 Likewise, the negative control samples produced <1% signal change (Figure
microsystem presented by Safavieh et al. could rapidly detect 7A), we measured sensor signals of 28%, 21%, and 4.2% from
Escherichia coli, but only from benchtop-purified urine samples respectively spiked with H1N1 virus at 1000, 100, or
samples.26 10 TCID50 (median tissue culture infective dose) (Figure 7B−
We therefore developed the Magnetic Integrated Micro- D). This confirmed that MIMED can achieve unambiguous
fluidic Electrochemical Detector (MIMED), which incorporates detection at concentrations as low as 10 TCID50, or 4 orders of
robust sample preparation functionality to overcome this magnitude lower than the clinical titers seen in typical throat
bottleneck (Figure 6A).27 The MIMED device employs a swab samples (∼105 TCID50).29
high-gradient magnetic field to enable immunomagnetic target
capture, concentration, and purification (Figure 6B and C),
followed by efficient on-chip RT-PCR (Figure 6D−F), single-
■ MULTIPLEXED DIFFERENTIATION OF TWO
BACTERIA STRAINS
stranded DNA generation (Figure 6G), and sequence-specific MIMED offers an integrated solution for directly detecting
E-DNA detection (Figure 6H). single pathogens in complex samples, but we saw even greater
Sample preparation in the MIMED system incorporates viral value in simultaneously detecting and discriminating multiple
RNA isolation and stabilization, magnetic-based concentration, different pathogens from patient samples within a single chip.
and continuous washing, all of which are critical for the success To the best of our knowledge, no other electrochemical
of the assay. To isolate and stabilize intact viral RNA, we microsystem to date has demonstrated such detection
incubated the throat swab sample into a cocktail containing: capabilities. For example, Yeung et al. achieved duplexed
(1) a nonionic detergent for dissolving the viral envelope and detection of Escherichia coli and Bacillus subtilis at concen-
releasing intact ribonucleoprotein (RNP)-containing target trations equivalent to 1 × 105 cells/mL, but only from culture
RNA,28 (2) an RNA stabilizer, and (3) antibody-coated broth.23
magnetic beads for capturing the released RNP. We To achieve this, we developed an integrated device that
subsequently injected the sample into the device, where high performs loop-mediated isothermal amplification (LAMP)30 to
magnetic field gradients captured and concentrated the bead- “universally” amplify common gene regions of closely related
bound viral particles within the device, enabling us to Salmonella strains directly from blood samples and sub-
continuously wash away cellular debris and other interferents sequently achieves detection and strain discrimination with
that may inhibit RT-PCR. As such, MIMED sample preparation two sequence-specific E-DNA probes functionalized on two
essentially matched the ideal, lossless positive control (viral distinct electrodes (Figure 8).31 As a demonstration, we
particles doped directly into PCR mix), as measured by designed our assay to detect S. enterica subsp. enterica serovars
benchtop, real-time RT-PCR. Following this sample prepara- Typhimurium and Choleraesuis (causative agents of enter-
D DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

Figure 5. IMED detection of Salmonella genomic DNA. (A) The no-

template negative control yielded <1% change in current (red)
compared to the baseline (blue). Probe regeneration reset the sensor
to within 98% of its initial state (green). The (B) 100 aM and (C) 10
aM samples produced a 52% and 12% signal change, respectively,
Figure 6. MIMED achieves sample-to-answer genetic detection of
relative to the baseline. Each measurement was validated via sensor
H1N1 virus from throat swabs. (A) The device features a sample
regeneration (green), as well as benchtop-prepared zero-template
preparation/reaction chamber, an E-DNA detection chamber, and
negative controls, which resulted in drops of 1% and 0%, respectively
three fluidic ports: sample/buffer/reagent input (left), waste output
(purple). Reprinted with permission from ref 19. Copyright 2009
(center), and E-DNA product output (right). (B) A throat swab is
American Chemical Society.
collected and combined with influenza virus and antibody-coated
magnetic beads in a tube containing RNA stabilizer. (C) The sample is
ocolitis and sepsis in humans32) directly from unprocessed, pumped into the sample preparation chamber, where external magnets
whole blood of infected mice. capture, concentrate and purify labeled viral RNPs. (D) RT-PCR mix
We employed LAMP rather than PCR as the amplification is injected, and (E) The chip is heated to denature the RNP and
technique in this work due to its sensitivity, robustness, release the RNA. (F, G) RT-PCR is performed on-chip, followed by
lambda exonuclease-mediated single-strand generation. (H) The
isothermal reaction condition, and most importantly, the rapid product is then pumped into the detection chamber for detection.
assay turnaround time, which is crucial for point-of-care Reprinted with permission from ref 27. Copyright 2011 American
diagnostics. Specifically, our LAMP was sufficiently robust to Chemical Society.
work in up to 10% blood by volume with essentially no sample
preparation and produced a high concentration of single-
stranded amplicons that can be detected with E-DNA without
the need for an additional step to generate single-stranded LAMP (∼1 h), this considerably shortened the assay
products. Coupled with the inherently fast reaction speed of turnaround time from 3.5 h (in the case of MIMED) to <2 h.
E DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

and finally immobilized the exposed second electrode with S.

Choleraesuis specific probe.
The resultant device readily differentiated the two strains in
unprocessed blood from septic mice at clinically relevant levels
of <1000 colony-forming units (CFU)/mL (Figure 9). Animals

Figure 7. Sensitive, sample-to-answer detection of H1N1 from throat

swabs with MIMED. Relative to the 0.5% peak current change
observed for (A) our virus-free negative control, swab samples Figure 9. LAMP chip-based detection and discrimination of
containing (B) 1000, (C) 100, and (D) 10 TCID50 return clearly Salmonella serovars. Our LAMP chip could detect and differentiate
detectable peak current changes of 28, 21, and 4.2%, respectively. All S. Typhimurium from S. Choleraesuis from blood samples derived
sensors could be regenerated to baseline levels, verifying the presence from septic mice infected with these bacteria. In contrast, samples
of the target in the sample. Reprinted with permission from ref 27. from uninfected or Y. pseudotuberculosis infected mice yielded minimal
Copyright 2011 American Chemical Society. signal change. CFU/mL for each sample was measured by direct
colony counting. Reproduced from ref 31. Copyright 2013 American
Society for Microbiology.

After the LAMP reaction, detection of strain-specific internal

amplicon sequences was achieved with two sequence-specific E- were infected intraperitoneally with 1000 CFU of S.
DNA probes functionalized on two distinct electrodes using an Typhimurium, S. Choleraesuis or Yersinia pseudotuberculosis,
established “differential probe labeling” technique.14 In this, we with uninfected mice as negative controls. We added blood
first immobilized S. Typhimurium specific probe onto both collected from the tail vein at day 5 postinfection to the LAMP
electrodes, then selectively desorbed the probe molecules from reaction mixture and loaded it into the LAMP chip, with an
the second electrode by applying a positive potential sweep, additional aliquot of blood reserved to quantify CFUs by direct

Figure 8. Overview of the LAMP chip assay. (A) Unprocessed, whole blood from infected animals is introduced into the chip’s amplification
chamber along with LAMP reagents and heated at 65 °C. The reaction mixture containing single-stranded amplicons is then pushed into (B) the
electrochemical detection chamber. This chamber contains a duplexed electrode array that supports simultaneous, sequence-specific electrochemical
detection by selectively hybridizing with amplicons from S. Typhimurium or S. Choleraesuis, (C) generating a detectable decrease in current.
Reprinted from ref 31. Copyright 2013 American Society for Microbiology.

F DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

plating. Our LAMP chip could discriminate S. Typhimurium As the reaction proceeds, we initiate a series of time-course
from S. Choleraesuis in blood containing between 8 × 102 and voltammetry scans, measuring the current every minute
6.9 × 104 CFU/mL. In contrast, samples from uninfected or Y. throughout the reaction. We generally observed an initial
pseudotuberculosis-infected mice (at 1.1 × 103 CFU/mL decrease in the redox current traces in the first 10 min for all of
bacterial load) yielded minimal signal change, validating our our reactions, regardless of the presence of DNA target, but
assay.

■
only LAMP reactions containing the target of interest triggered
REAL-TIME ELECTROCHEMICAL DETECTION a further decrease, yielding a sigmoid pattern resembling the
The detection systems described above could yield a valuable reaction kinetics typically observed in real-time PCR and
yes/no answer or semiquantitative results, but do not possess LAMP (Figure 11A). This sigmoidal curve enabled us to define
the fully quantitative capabilities that would be desirable in a
clinical diagnostics setting. This is because the final
concentration of amplification products generally does not
correlate well with the initial copy number present in a sample.
In response, we have developed the microfluidic electro-
chemical quantitative (MEQ)-LAMP platform,33 which allows
for continuous electrochemical monitoring of LAMP reaction
progress in real-time, facilitating quantitative measurements.
MEQ-LAMP leverages the fact that the MB redox reporter can
intercalate into double-stranded DNA to directly detect
production of double-stranded LAMP amplicons. The MEQ-
LAMP chip features a single chamber for both amplification
and detection (Figure 10A). Initially, MB molecules doped into

Figure 10. Overview of MEQ-LAMP. (A) The MEQ-LAMP chip

features a single chamber for both amplification and electrochemical
detection. (B) MB enables real-time monitoring of the LAMP
reaction. Initially, MB molecules doped into the reaction mix freely
encounter the gold working electrode and transfer electrons,
producing a measurable current. As the reaction progresses,
intercalation of MB molecules into the LAMP amplicons segregates
them from the electrode, proportionally decreasing the current.
Reprinted with permission from ref 33. Copyright 2012 Wiley.

the LAMP reaction mix freely encounter the gold working Figure 11. MEQ-LAMP accurately quantifies genomic DNA target
electrode and transfer electrons, producing a measurable copy number. (A) Normalized real-time current traces for a negative
current. As the reaction progresses, MB intercalates into the control sample (black) and a sample containing S. Typhimurium
double-stranded amplicons. This segregates MB from the genomic DNA (red). Only the pathogen-containing sample generates
electrode, decreasing the redox current in a manner that a sharp current decrease in Region 2 of the current trace, with the
enables tracking of the reaction in real time (Figure 10B). sigmoidal behavior similar to typical real-time PCR kinetics. (B) By
taking the derivative of the current trace and defining the signal
MEQ-LAMP is considerably simpler than our previous threshold as the local minimum in the derivative curve, we can
platforms; this detection mechanism completely obviates the determine the required reaction time to threshold (tTH, vertical dashed
need for E-DNA probes or probe-target hybridization, though it line). (C) 10-fold serial dilutions of S. Typhimurium DNA (ranging
requires a new strategy for data processing. Once genomic from 1.6 × 104 to 1.6 × 101 copies) result in traces with distinct tTH,
DNA and the MB-doped LAMP reaction mixture are loaded, separated by approximately 10 min. Reprinted with permission from
the chip is mounted onto a block heater maintained at 65 °C. ref 33. Copyright 2012 Wiley.

G DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

the signal threshold for each MEQ-LAMP reaction at the local (e.g., Gram-positive bacteria) and biological samples that
minimum in the corresponding current derivative trace (i.e., dI/ require specialized processing (e.g., sputum).
dt). We defined the time to threshold (tTH) as the time Expanded multiplexing would be far more efficient than
required for a particular sample reaction to reach the signal single-pathogen assays, and could facilitate treatment by
threshold (Figure 11B), analogous to the threshold cycle (CT) enabling discrimination of distinct pathogens that cause similar
concept in real-time PCR. disease states (e.g., sepsis). By some estimates, genetic testing
Using this strategy, MEQ-LAMP measures Salmonella for ∼8−30 different microbial strains could diagnose 80−90%
genomic DNA with a dynamic range spanning 3 orders of of all serious infections, facilitating proper antibiotic selection.41
magnitude and a LOD as low as 16 copies in <50 min. Different Performing multiple pathogen-specific reactions in parallel is
initial copy numbers of S. Typhimurium DNA (10-fold serial one solution, but this approach can be inefficient and cost-
dilutions ranging from 1.6 × 104 to 1.6 × 101 copies) generated prohibitive for large numbers of targets. Alternatively, one
temporally distinct local minima separated by approximately 10 might employ universal primers to amplify sequences that can
min for each of the current derivative traces, with a distinct tTH be subsequently differentiated via either sequence-specific
for each initial target copy number (Figure 11C). MEQ- probes (as demonstrated with our LAMP chip and else-
LAMP’s speed, sensitivity, and quantitative capabilities where42−45) or labeled with different redox reporters.46,47
surpassed other real-time electrochemical amplification plat- However, the extent to which this approach can be scaled up
forms (see review by Patterson et al.34), thus triggering growing remains to be seen.
interests in this approach. Other research groups have System integration, automation, and miniaturization will also
subsequently developed platforms employing similar strategies play a critical role.48 Current microsystems generally still
to achieve impressive performance.35−38 For example, Ahmed require manual intervention (e.g., reagent loading) and
et al.36 have reported detection of Staphylococcus aureus and relatively bulky peripheral instruments (e.g., syringe pumps,
Escherichia coli at LODs of 30 and 20 copies μL−1, respectively, heating blocks, and potentiostats). Potential solutions to this
in 30 min, and Luo et al.37 and Safavieh et al.38 have both challenge may be integrated platforms that support multistep
demonstrated real-time, multiplexed on-chip electrochemical assays and electrochemical detection in a hands-free fashion,
LAMP. such as centrifugal microfluidic devices,49 bubble-mediated

■ CONCLUSIONS AND OUTLOOK

In this Account, we have described our evolving efforts in
reagent transfer devices,50 and paper microfluidic devices.51 We
note that reagents must also be embedded in such devices to
ensure true point-of-care use. In parallel, portable instruments
developing integrated electrochemical microsystems for the such as commercially available, miniature, USB-powered
genetic detection of pathogens at the point of care. Building on potentiostats and custom-developed potentiostats52,53 have
a foundation of microscale sensors employing redox-tagged, grown increasingly popular as means to reduce system
conformation-switching E-DNA probes realized in microfluidic footprint.
chips, we have subsequently incorporated on-chip genetic In sum, the past several years have witnessed significant
amplification to enable detection at clinically relevant advances in integrated electrochemical microsystems, and the
concentrations. The next generation of devices further technological barriers that once thwarted the development of
simplified clinical application by integrating sample preparation these systems have been lowered considerably. Given their
via immunomagnetic separation or even direct amplification in many advantages, the promising recent advances described
blood via LAMP. We have further demonstrated the potential above, and the growing number of research groups making
for multiplexed pathogen differentiation by discriminating two valuable contributions to this field, we are optimistic that future
bacterial strains in blood. Most recently, we have achieved iterations of integrated electrochemical microsystems will
rapid, quantitative, real-time detection across a broad dynamic address the remaining challenges in sample preparation,
range of nucleic acid concentrations via the MEQ-LAMP multiplexed detection, and system integration, and thereby
platform. This approach has now become popular in this field, bring the power of genetic detection of pathogens to the point
and we foresee exciting potential to incorporate sample of care in the foreseeable future.
preparation and multiplexed detection capabilities into the
MEQ-LAMP platform.
Though the design of an integrated electrochemical micro-
■ AUTHOR INFORMATION
Corresponding Author
system must be dictated by the specific applicationthe
pathogen or pathogens of interest, the type of biological *E-mail: [email protected].
sample, and the type of sample preparation (see specific Present Addresses
applications provided by Park et al.5 and Niemz et al.6)we #
K.H.: Department of Mechanical Engineering, Johns Hopkins
see several general areas to further extend the utility of University, Baltimore, Maryland 21218.
integrated electrochemical microsystems and accelerate their ∇
B.S.F.: Aptitude Medical Systems, Santa Barbara, California
implementation at the point of care. Effective sample 93105.
preparation remains a critical challenge.6,39,40 Direct target
Notes
amplification in biological samples offers one simple solution
without adding to the assay turnaround time, but still-greater The authors declare no competing financial interest.
sensitivity is required to detect the limiting quantities of
Biographies
pathogen genetic material in microliter-scale sample volumes.
Immunomagnetic purification and concentration are effective, Kuangwen Hsieh received his B.S. in Electrical Engineering from
but require additional reagents and steps that may prove University of Maryland, College Park and Ph.D. from University of
problematic in point-of-care settings. Furthermore, alternative California, Santa Barbara under the guidance of H. Tom Soh. He is
approaches are needed for pathogens that are difficult to lyse currently a postdoctoral researcher at Johns Hopkins University. His

H DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

research interests focus on integrated electrochemical microsystems (9) Drummond, T. G.; Hill, M. G.; Barton, J. K. Electrochemical
and microfluidic droplet technologies. DNA sensors. Nat. Biotechnol. 2003, 21, 1192−1199.
(10) Wang, J. Electrochemical biosensors: Towards point-of-care
B. Scott Ferguson received his B.Eng. in Engineering Physics from cancer diagnostics. Biosens. Bioelectron. 2006, 21, 1887−1892.
McMaster University and Ph.D. from the University of California, (11) Fan, C.; Plaxco, K. W.; Heeger, A. J. Electrochemical
Santa Barbara under the guidance of H. Tom Soh. He is currently the interrogation of conformational changes as a reagentless method for
CEO at Aptitude Medical Systems Inc. His research interests focus on the sequence-specific detection of DNA. Proc. Natl. Acad. Sci. U. S. A.
directed evolution and real-time aptamer-based integrated electro- 2003, 100, 9134−9137.
chemical microsystems. (12) Lubin, A. A.; Plaxco, K. W. Folding-based electrochemical
biosensors: The case for responsive nucleic acid architectures. Acc.
Michael Eisenstein received his B.S. in biology from Brown University
Chem. Res. 2010, 43, 496−505.
and M.S. in cellular and molecular biology from Rockefeller University. (13) Lai, R. Y.; Lagally, E. T.; Lee, S. H.; Soh, H. T.; Plaxco, K. W.;
He currently works as a freelance science journalist, manuscript editor, Heeger, A. J. Rapid, sequence-specific detection of unpurified PCR
and research consultant, with a primary specialization in genetics and amplicons via a reusable, electrochemical sensor. Proc. Natl. Acad. Sci.
biotechnology. U. S. A. 2006, 103, 4017−4021.
Kevin W. Plaxco received his B.S. with a double major in Chemistry (14) Pavlovic, E.; Lai, R. Y.; Wu, T. T.; Ferguson, B. S.; Sun, R.;
and Biochemistry from the University of California Riverside before Plaxco, K. W.; Soh, H. T. Microfluidic device architecture for
moving on to complete a Ph.D. in Molecular Biology at Caltech. He is electrochemical patterning and detection of multiple DNA sequences.
Langmuir 2008, 24, 1102−1107.
currently a professor of Chemistry and Biochemistry at UC-Santa
(15) Pumera, M.; Merkoci, A.; Alegret, S. New materials for
Barbara where his research focuses on molecular biophysics and
electrochemical sensing VII. Microfluidic chip platforms. TrAC, Trends
bioengineering. Anal. Chem. 2006, 25, 219−235.
H. Tom Soh received his B.S. with a double major in Mechanical (16) Mir, M.; Homs, A.; Samitier, J. Integrated electrochemical DNA
Engineering and Materials Science with Distinction from Cornell biosensors for lab-on-a-chip devices. Electrophoresis 2009, 30, 3386−
University and his Ph.D. in Electrical Engineering from Stanford 3397.
University. He is currently the Ruth Garland Professor at UC-Santa (17) Luo, X.; Hsing, I.-M. Electrochemical techniques on sequence-
Barbara. His current research interests are in biosensors and directed specific PCR amplicon detection for point-of-care applications. Analyst
2009, 134, 1957−1964.
evolution of molecules.

■
(18) Duwensee, H.; Mix, M.; Flechsig, G.-U. Sequence-specific
electrochemical detection of nucleic acids in real samples. Bioanal. Rev.
ACKNOWLEDGMENTS 2010, 2, 103−114.
We gratefully acknowledge our colleagues from the Soh Lab (19) Ferguson, B. S.; Buchsbaum, S. F.; Swensen, J. S.; Hsieh, K.;
Lou, X.; Soh, H. T. Integrated microfluidic eectrochemical DNA
and the Plaxco Lab, with particular recognition going to
sensor. Anal. Chem. 2009, 81, 6503−6508.
Elizabeth Pavlovic and Adriana S. Patterson, who spearheaded (20) Little, J. W. An exonuclease induced by bacteriophage lambda.
the development of two key platforms described in this II. Nature of the enzymatic reaction. J. Biol. Chem. 1967, 242, 679−
Account. We are also grateful for the financial support of the 686.
Garland Initiative, ARO Institute for Collaborative Biotechnol- (21) Umek, R. M.; Lin, S. W.; Vielmetter, J.; Terbrueggen, R. H.;
ogies (W911F-09-D-0001, W81XWH-09-0698), the National Irvine, B.; Yu, C. J.; Kayyem, J. F.; Yowanto, H.; Blackburn, G. F.;
Institutes of Health (U54 DK093467, U01 HL099773), and the Farkas, D. H.; Chen, Y. P. Electronic detection of nucleic acids: a
W. M. Keck Foundation Medical Research Program. versatile platform for molecular diagnostics. J. Mol. Diagn. 2001, 3,

■
74−84.
REFERENCES (22) Liu, R. H.; Yang, J.; Lenigk, R.; Bonanno, J.; Grodzinski, P. Self-
contained, fully integrated biochip for sample preparation, polymerase
(1) Olsen, J. E. DNA-based methods for detection of food-borne
chain reaction amplification, and DNA microarray detection. Anal.
bacterial pathogens. Food Res. Int. 2000, 33, 257−266.
Chem. 2004, 76, 1824−1831.
(2) Wang, J.; Rivas, G.; Cai, X.; Palecek, E.; Nielsen, P.; Shiraishi, H.;
(23) Yeung, S.-W.; Lee, T. M.-H.; Cai, H.; Hsing, I.-M. A DNA
Dontha, N.; Luo, D.; Parrado, C.; Chicharro, M.; Farias, P. A. M.;
biochip for on-the-spot multiplexed pathogen identification. Nucleic
Valera, F. S.; Grant, D. H.; Ozsoz, M.; Flair, M. N. DNA
electrochemical biosensors for environmental monitoring. A review. Acids Res. 2006, 34, e118.
Anal. Chim. Acta 1997, 347, 1−8. (24) Lee, T. M.-H.; Carles, M. C.; Hsing, I.-M. Microfabricated PCR-
(3) Yang, S.; Rothman, R. E. PCR-based diagnostics for infectious electrochemical device for simultaneous DNA amplification and
diseases: uses, limitations, and future applications in acute-care detection. Lab Chip 2003, 3, 100−105.
settings. Lancet Infect. Dis. 2004, 4, 337−348. (25) Yamanaka, K.; Saito, M.; Kondoh, K.; Hossain, M. M.; Koketsu,
(4) Kelley, S. O.; Mirkin, C. A.; Walt, D. R.; Ismagilov, R. F.; Toner, R.; Sasaki, T.; Nagatani, N.; Ikuta, K.; Tamiya, E. Rapid detection for
M.; Sargent, E. H. Advancing the speed, sensitivity and accuracy of primary screening of influenza A virus: microfluidic RT-PCR chip and
biomolecular detection using multi-length-scale engineering. Nat. electrochemical DNA sensor. Analyst 2011, 136, 2064−2068.
Nanotechnol. 2014, 9, 969−980. (26) Safavieh, M.; Ahmed, M. U.; Tolba, M.; Zourob, M. Microfluidic
(5) Park, S.; Zhang, Y.; Lin, S.; Wang, T. H.; Yang, S. Advances in electrochemical assay for rapid detection and quantification of
microfluidic PCR for point-of-care infectious disease diagnostics. Escherichia coli. Biosens. Bioelectron. 2012, 31, 523−528.
Biotechnol. Adv. 2011, 29, 830−839. (27) Ferguson, B. S.; Buchsbaum, S. F.; Wu, T. T.; Hsieh, K.; Xiao,
(6) Niemz, A.; Ferguson, T. M.; Boyle, D. S. Point-of-care nucleic Y.; Sun, R.; Soh, H. T. Genetic analysis of H1N1 influenza virus from
acid testing for infectious diseases. Trends Biotechnol. 2011, 29, 240− throat swab samples in a microfluidic system for point-of-care
250. diagnostics. J. Am. Chem. Soc. 2011, 133, 9129−9135.
(7) Chen, L.; Manz, A.; Day, P. J. R. Total nucleic acid analysis (28) Kawakami, K.; Ishihama, A. RNA polymerase of influenza virus.
integrated on microfluidic devices. Lab Chip 2007, 7, 1413−1423. III. Isolation of RNA polymerase-RNA complexes from influenza virus
(8) Park, B. H.; Kim, Y. T.; Jung, J. H.; Seo, T. S. Integration of PR8. J. Biochem. 1983, 93, 989−996.
sample pretreatment, μPCR, and detection for a total genetic analysis (29) Fouchier, R. A. M.; Bestebroer, T. M.; Herfst, S.; Van der Kemp,
microsystem. Microchim. Acta 2014, 181, 1655−1668. L.; Rimmelzwaan, G. F.; Osterhaus, A. D. M. E. Detection of influenza

I DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research Article

A viruses from different species by PCR amplification of conserved (49) Kim, T. H.; Abi-Samra, K.; Sunkara, V.; Park, D. K.; Amasia, M.;
sequences in the matrix gene. J. Clin. Microbiol. 2000, 38, 4096−4101. Kim, N.; Kim, J.; Kim, H.; Madou, M.; Cho, Y. K. Flow-enhanced
(30) Notomi, T.; Okayama, H.; Masubuchi, H.; Yonekawa, T.; electrochemical immunosensors on centrifugal microfluidic platforms.
Watanabe, K.; Amino, N.; Hase, T. Loop-mediated isothermal Lab Chip 2013, 13, 3747−3754.
amplification of DNA. Nucleic Acids Res. 2000, 28, e63. (50) Yang, F.; Zuo, X. L.; Li, Z. H.; Deng, W. P.; Shi, J. Y.; Zhang, G.
(31) Patterson, A. S.; Heithoff, D. M.; Ferguson, B. S.; Soh, H. T.; J.; Huang, Q.; Song, S. P.; Fan, C. H. A bubble-mediated intelligent
Mahan, M. J.; Plaxco, K. W. Microfluidic chip-based detection and microscale electrochemical device for single-step quantitative bio-
intraspecies strain discrimination of Salmonella serovars derived from assays. Adv. Mater. 2014, 26, 4671−4676.
whole blood of septic mice. Appl. Environ. Microbiol. 2013, 79, 2302− (51) Cunningham, J. C.; Brenes, N. J.; Crooks, R. M. Paper
2311. electrochemical device for detection of DNA and thrombin by target-
(32) Coburn, B.; Grassl, G. A.; Finlay, B. B. Salmonella, the host and induced conformational switching. Anal. Chem. 2014, 86, 6166−6170.
disease: a brief review. Immunol. Cell Biol. 2007, 85, 112−118. (52) Kwakye, S.; Goral, V. N.; Baeumner, A. J. Electrochemical
(33) Hsieh, K.; Patterson, A. S.; Ferguson, B. S.; Plaxco, K. W.; Soh, microfluidic biosensor for nucleic acid detection with integrated
H. T. Rapid, sensitive, and quantitative detection of pathogenic DNA minipotentiostat. Biosens. Bioelectron. 2006, 21, 2217−2223.
at the point of care through microfluidic electrochemical quantitative (53) Rowe, A. A.; Bonham, A. J.; White, R. J.; Zimmer, M. P.; Yadgar,
loop-mediated isothermal amplification. Angew. Chem., Int. Ed. 2012, R. J.; Hobza, T. M.; Honea, J. W.; Ben-Yaacov, I.; Plaxco, K. W.
51, 4896−4900. CheapStat: An open-source, “do-it-yourself” potentiostat for analytical
(34) Patterson, A. S.; Hsieh, K.; Soh, H. T.; Plaxco, K. W. and educational applications. PLoS One 2011, 6, e23783.
Electrochemical real-time nucleic acid amplification: towards point-of-
care quantification of pathogens. Trends Biotechnol. 2013, 31, 704−
712.
(35) Jiang, D.; Xiang, G.; Liu, C.; Yu, J.; Liu, L.; Pu, X. Development
of a cyclic voltammetry method for DNA electrochemical detection on
microfluidic gene chip. Int. J. Electrochem. Sci. 2012, 7, 10607−10619.
(36) Ahmed, M. U.; Nahar, S.; Safavieh, M.; Zourob, M. Real-time
electrochemical detection of pathogen DNA using electrostatic
interaction of a redox probe. Analyst 2013, 138, 907−915.
(37) Luo, J.; Fang, X.; Ye, D.; Li, H.; Chen, H.; Zhang, S.; Kong, J. A
real-time microfluidic multiplex electrochemical loop-mediated iso-
thermal amplification chip for differentiating bacteria. Biosens.
Bioelectron. 2014, 60, 84−91.
(38) Safavieh, M.; Ahmed, M. U.; Ng, A.; Zourob, M. High-
throughput real-time electrochemical monitoring of LAMP for
pathogenic bacteria detection. Biosens. Bioelectron. 2014, 58, 101−106.
(39) Ritzi-Lehnert, M. Development of chip-compatible sample
preparation for diagnosis of infectious diseases. Expert Rev. Mol. Diagn.
2012, 12, 189−206.
(40) Reinholt, S. J.; Baeumner, A. J. Microfluidic isolation of nucleic
acids. Angew. Chem., Int. Ed. 2014, 53, 13988−14001.
(41) Gubala, V.; Harris, L. F.; Ricco, A. J.; Tan, M. X.; Williams, D. E.
Point of care diagnostics: status and future. Anal. Chem. 2012, 84,
487−515.
(42) Slinker, J. D.; Muren, N. B.; Gorodetsky, A. A.; Barton, J. K.
Multiplexed DNA-modified electrodes. J. Am. Chem. Soc. 2010, 132,
2769−2774.
(43) Henry, O. Y.; Fragoso, A.; Beni, V.; Laboria, N.; Sanchez, J. L.
A.; Latta, D.; Von Germar, F.; Drese, K.; Katakis, I.; O’Sullivan, C. K.
Design and testing of a packaged microfluidic cell for the multiplexed
electrochemical detection of cancer markers. Electrophoresis 2009, 30,
3398−3405.
(44) White, R. J.; Kallewaard, H. M.; Hsieh, W.; Patterson, A. S.;
Kasehagen, J. B.; Cash, K. J.; Uzawa, T.; Soh, H. T.; Plaxco, K. W.
Wash-free, electrochemical platform for the quantitative, multiplexed
detection of specific antibodies. Anal. Chem. 2012, 84, 1098−1103.
(45) Yang, A. H. J.; Hsieh, K.; Patterson, A. S.; Ferguson, B. S.;
Eisenstein, M.; Plaxco, K. W.; Soh, H. T. Accurate zygote-specific
discrimination of single-nucleotide polymorphisms using microfluidic
electrochemical DNA melting curves. Angew. Chem., Int. Ed. 2014, 53,
3163−3167.
(46) Wang, J.; Liu, G.; Merkoci, A. Electrochemical coding
technology for simultaneous detection of multiple DNA targets. J.
Am. Chem. Soc. 2003, 125, 3214−3215.
(47) Kang, D.; White, R. J.; Xia, F.; Zuo, X.; Vallee-Belisle, A.; Plaxco,
K. W. DNA biomolecular-electronic encoder and decoder devices
constructed by multiplex biosensors. NPG Asia Mater. 2012, 4, e1.
(48) Sin, M. L. Y.; Gao, J.; Liao, J. C.; Wong, P. K. System integration
- A major step toward lab on a chip. J. Biol. Eng. 2011, 5
DOI: 10.1186/1754-1611-5-6.

J DOI: 10.1021/ar500456w
Acc. Chem. Res. XXXX, XXX, XXX−XXX