Multiplex analysis of enzyme kinetics and inhibition by droplet microfluidics using picoinjectors

Article abstract of DOI:10.1039/c3lc41398e

Enzyme kinetics and inhibition is important for a wide range of disciplines including pharmacology, medicine and industrial bioprocess technology. We present a novel microdroplet-based device for extensive characterization of the reaction kinetics of enzyme substrate inhibitor systems in a single experiment utilizing an integrated droplet picoinjector for bioanalysis. This device enables the scanning of multiple fluorescently-barcoded inhibitor concentrations and substrate conditions in a single, highly time-resolved experiment yielding the Michaelis constant (K_m), the turnover number (k_cat) and the enzyme inhibitor dissociation constants (k_i, k _i′). Using this device we determine K_m and k _cat for β-galactosidase and the fluorogenic substrate Resorufin β-d-galactopyranoside (RBG) to be 442 μM and 1070 s^-1, respectively. Furthermore, we examine the inhibitory effects of isopropyl-β-d-thiogalactopyranoside (IPTG) on β-galactosidase. This system has a number of potential applications, for example it could be used to screen inhibitors to pharmaceutically relevant enzymes and to characterize engineered enzyme variants for biofuels production, in both cases acquiring detailed information about the enzyme catalysis and enzyme inhibitor interaction at high throughput and low cost.

Full text of DOI:10.1039/c3lc41398e

Lab on a Chip
View Article Online
View Journal | View Issue
PAPER
Multiplex analysis of enzyme kinetics and inhibition by
droplet microfluidics using picoinjectors3
Cite this: Lab Chip, 2013, 13, 1754
Staffan L. Sjostrom,* Haakan N. Joensson and Helene Andersson Svahn
Enzyme kinetics and inhibition is important for a wide range of disciplines including pharmacology,
medicine and industrial bioprocess technology. We present a novel microdroplet-based device for
extensive characterization of the reaction kinetics of enzyme substrate inhibitor systems in a single
experiment utilizing an integrated droplet picoinjector for bioanalysis. This device enables the scanning of
multiple fluorescently-barcoded inhibitor concentrations and substrate conditions in a single, highly time-
resolved experiment yielding the Michaelis constant (Km), the turnover number (kcat) and the enzyme
inhibitor dissociation constants (ki, ki9). Using this device we determine Km and kcat for b-galactosidase and
the fluorogenic substrate Resorufin b-D-galactopyranoside (RBG) to be 442 mM and 1070 s21, respectively.
Furthermore, we examine the inhibitory effects of isopropyl-b-D-thiogalactopyranoside (IPTG) on
b-galactosidase. This system has a number of potential applications, for example it could be used to
screen inhibitors to pharmaceutically relevant enzymes and to characterize engineered enzyme variants
for biofuels production, in both cases acquiring detailed information about the enzyme catalysis and
enzyme inhibitor interaction at high throughput and low cost.
Received 20th December 2012,
Accepted 12th February 2013
DOI: 10.1039/c3lc41398e
www.rsc.org/loc
inhibit the enzymes employed.3,4 Understanding these could
improve the cost effectiveness of biofuels production.
Enzyme kinetics is generally governed by the Michaelis–
Menten equation.2,5,6
Introduction
The understanding of enzyme kinetics and inhibition is
important to a wide range of disciplines including pharmacol-
ogy, medicine and industrial biofuels production. Enzymes
catalyse the biochemical reactions in our bodies and modulate
many of the processes required for life. Many diseases can be
related to malfunctioning of these systems and enzymes thus
provide an important target for pharmaceuticals in the form of
enzyme inhibitors.1 Interactions between pharmaceuticals and
enzymes other than the target enzymes is also a cause of
unwanted side effects of many drugs.2 Thus, there is a great
need to be able to characterize enzyme kinetics and inhibition
in a high throughput manner to be able to develop better
enzymes as selective pharmaceuticals and to further the
understanding of the fundamental biological functions of
living organisms.
Vmax½Sꢀ
V0 ~
(1)
½SꢀzKm
where V0, Vmax, [S] and Km are the initial reaction velocity, the
maximum reaction velocity, the substrate concentration and
the Michaelis–Menten constant respectively. The Michaelis–
Menten constant is considered one of the most important
enzyme kinetic parameters and represents the enzyme affinity
for the substrate. The maximum reaction velocity is related to
the turnover number, kcat, which represents the number of
molecules processed by an enzyme molecule per second. kcat is
derived as follows.5
Another area where understanding enzyme inhibitor inter-
actions could have a large impact is in the biofuels production
process. Production of biofuel from cellulosic biomass could
provide an alternative to fossil fuels. To achieve cost efficient
conversion of cellulosic biomass to fuel, enzymatic sacchar-
ification and fermentation is being investigated. A challenge in
these processes is that cellulosic biomass contains complex
mixtures of polysaccharides and glycosides, many of which
Vmax
kcat
~
(2)
½Eꢀ
where [E] is the enzyme concentration. The Michaelis–Menten
equation can be modified to incorporate the effect of
inhibitors as follows.2,5
Vmax½Sꢀ
ꢀ
ꢁ
ꢀ
ꢁ
V0~
(3)
½Iꢀ
ki’
½Iꢀ
½Sꢀ 1z
zKm 1z
ki
Division of Proteomics and Nanobiotechnology, Science for Life Laboratory, KTH-
Royal Institute of Technology, Stockholm, Sweden. E-mail: ssjos@kth.se
Electronic supplementary information (ESI) available. See DOI: 10.1039/
3
c3lc41398e
1754 | Lab Chip, 2013, 13, 1754–1761
This journal is ß The Royal Society of Chemistry 2013

View Article Online
Lab on a Chip
Paper
Here [I], ki and ki9 are the inhibitor concentration, enzyme
inhibitor disassociation constant and the enzyme–substrate
inhibitor disassociation constant, respectively. The ki and ki9
parameters characterize the mode of inhibition and the
enzyme affinity for the inhibitor.
amounts of reagent making it suitable for large scale screening
experiments. To demonstrate the system, we determine the
Km, kcat, ki and ki9 of b-galactosidase (b-gal) and its substrate
resorufin-b-D-galactopyranoside (RBG) inhibited by isopropyl-
b-D-thiogalactopyranoside (IPTG).
To be able to determine the kcat, Km, ki and ki9 of an enzyme
substrate inhibitor system the initial reaction velocity V0 must
be determined for combinations of multiple substrate and
inhibitor concentrations. The conventional method to perform
enzyme kinetics measurements employs the microtiter plate
platform. However, this platform has several limitations.
These experiments require several pipetting steps for each
condition leading to pipetting errors. Micro titer plate based
studies also consume prohibitive amounts of reagent to allow
for large-scale screening. Furthermore, with the conventional
platforms the time resolution that can be achieved is on the
minute scale, which makes it difficult to study fast enzymatic
reactions.
Stopped-flow is a method that can be used to study such
fast reactions. In this method, reagent and enzyme are
propelled from two syringes into a mixing chamber for
readout in a spectrophotometer cell.6 This method gives
millisecond time resolution but these platforms only perform
one experiment at a time making them unsuitable for large
scale screening experimentation.
Recently, a number of microfluidic systems have been
proposed for enzyme kinetics analysis. These systems provide
lower reagent consumption and higher precision compared to
conventional systems.7–10 However, conventional microfluidic
systems have two main drawbacks, slow mixing and Taylor
dispersion. Droplet-based microfluidic systems is an emerging
technology that can provide millisecond mixing and prevents
dispersion in addition to the advantages offered by conven-
tional microfluidic systems.11
In droplet microfluidic systems, surfactant stabilized
mono-disperse picoliter sized aqueous droplets are generated
in fluorocarbon oil. Each droplet has a unique chemical
identity after generation and can be thought of as a test tube or
a micro plate well. Droplets can be manipulated at rates of
above 1000 Hz in a multitude of ways including addition of
reagent by droplet fusion12 or picoinjection13, sorting of
droplets based on size14 and fluorescence-activated sorting of
droplets.15
Materials and methods
Materials
FC-40 FluorinertTM Electronic Liquid was obtained from 3M.
Polydimethylsiloxane base (PDMS) and curing agent was
obtained from Dow Coroning. EA surfactant was obtained
from RainDance Technologies. Resorufin, resorufin-b-D-galactopyr-
anoside (RBG), b-galactosidase from E. coli (768 units mg21) (See ESI
3
for definition of unit), 3-(N-morpholino)propanesulfonic acid buffer
(MOPS), rhodamine B and fluorescein was obtained from Sigma-
Aldrich. Low melting solder was obtained from Indium Corp.
Isopropyl-b-D-thiogalactopyranoside (IPTG) was obtained from Alfa
Aesar. Microscope glass slides was obtained from Vwr.
Manufacturing of chips
Microfluidics chips were manufactured in PDMS and glass
according to standard two-layer soft lithography techniques.25
Briefly, masters were fabricated on 4’’ silicon wafers using SU8
photoresist. To cast the replicates, liquid PDMS was mixed
with curing agent in a 9 : 1 ratio and poured onto the master
template. After curing for 4 h at 65 uC, the PDMS was peeled
off and holes for the inlets were punched with biopsy
punchers (Harris Uni-Core) and cleaned using scotch tape.
The PDMS slabs were subsequently sonicated in an ultrasonic
bath (15 min) to remove debris. The PDMS slabs were surface
activated in an oxygen plasma (Femto Scientific) and bonded
to glass microscope slides. The glass/PDMS devices were
placed on a hot plate held at 80 uC and melted low melting
point solder was injected into purpose-designed channels to
form on-chip electrodes. The electrodes were connected with
copper wire before being removed from the hot plate to allow
the metal electrodes to solidify.
Experimental setup
The experimental setup consists of four syringe pumps
(NEMESYS, Cetoni GmbH) used to actuate syringes that are
connected to the microfluidic chip with polyether ether ketone
(PEEK) tubing. The microfluidic chip is placed in the center of
an xy-adjustable table in a custom-built microscope setup. A
488 nm laser can be focused through the microscope objective
and targeted to a specific point in the chip, enabling laser-
induced fluorescence measurements by two photo multiplier
tube modules (PMTs) (Hamamatsu) detecting emitted light
filtered through band pass filters at 525 ¡ 20 nm and 593 ¡
20 nm respectively. A high voltage amplifier (TREK Inc) unit is
connected to on-chip electrodes by copper wires and amplifies
a signal supplied by a computer to generate an electric field on
chip.
Such droplets have been shown to be biocompatible16 and
have enabled the miniaturization of many biochemical
analysis methods in the areas of single cell analysis,17
diagnostics18 and directed evolution.19 The droplets can also
be barcoded to allow for highly multiplex assays.17,20 Droplet
microfluidics has been successfully demonstrated for enzyme
kinetics and dose response assays in a number of recent
publications.11,21–24
In this paper, we present a droplet-based microfluidic
platform to extract the key enzyme kinetic and inhibition
constants in a single experiment utilizing an integrated
droplet picoinjector (Fig. 1). Our system allows for multiplex
experiments, has good time resolution and consumes low
This journal is ß The Royal Society of Chemistry 2013
Lab Chip, 2013, 13, 1754–1761 | 1755

View Article Online
Paper
Lab on a Chip
Fig. 1 A) Assay workflow. From the top left: enzyme solution is mixed with enzyme + inhibitor + barcode solution to produce a library of chemically distinct droplet
populations that each contain the same enzyme concentration but various concentrations of inhibitor barcoded by a fluorescent green dye. The library is collected in a
syringe and reinjected into the second circuit. The droplets are spaced with fluorocarbon oil and each droplet is injected with a defined amount of fluorogenic
enzyme substrate using a picoinjector module. The droplets are incubated on-chip in a serpentine channel where the enzyme processes the substrate into orange
fluorescent product. Droplet fluorescence is measured in two channels (525 nm and 593 nm), which allows for detection of both the barcode and the assay product,
at the six designated measurement points (MP1, MP2 … MP6) to gather one data set. The concentration of the fluorogenic substrate supplied by the picoinjector
module is subsequently changed and data sets are gathered for multiple defined substrate concentrations. The data can be used to extract reaction velocities for
multiple combinations of substrate and inhibitor concentrations, which in turn are used to determine the kinetic parameters of the enzyme substrate inhibitor system.
B) Micrograph showing one of the six measurement points with lower depth (30 mm) than the incubation channel (110 mm) which allows for laser-induced
fluorescence measurements with single droplet resolution. C) Micrograph showing one of the many constrictions in the serpentine channel that enhances mixing of
droplet and thereby improves consistency of incubation times.
times.26 Six measurement points in the incubation line
Microfluidic circuits
allowed for laser-induced fluorescence measurement with
single droplet resolution with detection at two wavelengths
The microfluidic system consists of two circuits (Fig. S1 and
S2, ESI for design schematics). In the first circuit, mono-
3
by PMTs. Each measurement point corresponds to a defined
incubation time post injection of enzyme substrate. The
incubation time between two consecutive measurement points
was calculated to 15.4 s for the total flow rate used and the
channel volume. The second circuit was operated with the
following flow rates: 100 mL h21 fluorocarbon oil, 50 mL h21
emulsion and 40 mL h21 aqueous phase through the
picoinjector channel.
disperse 33 pL aqueous droplets in fluorocarbon oil are
generated at a rate of 840 Hz. The droplets are stabilized by
2.5% (w/w) EA surfactant in the fluorocarbon oil phase. The
droplet generation circuit is operated at flow rates of 100 mL
h21 total aqueous phase and 200 mL h21 FC-40 fluorocarbon
oil.
The emulsion that was generated on the first circuit is
collected off-chip in a glass syringe (Hamilton) and transferred
to the second circuit. The droplets are flowed onto the second
circuit, spaced with fluorocarbon oil and then injected with
one droplet volume of enzyme substrate using a picoinjector
module. The droplets are subsequently incubated on-chip in
an incubation line with increased depth (110 mm) and width
(200 mm). The incubation line contains regular constrictions to
improve droplet mixing and ensure reliable droplet incubation
Results
Characterization of picoinjector module
The function of the picoinjector module is to inject a defined
volume of liquid into each droplet that passes in the main
1756 | Lab Chip, 2013, 13, 1754–1761
This journal is ß The Royal Society of Chemistry 2013

View Article Online
Lab on a Chip
Paper
channel (Fig. 2A). The picoinjector module was characterized
by injecting fluorescein solution into droplets and subse-
quently measuring their fluorescence downstream of the
injection point. Flow rates, mode of actuation and strength
of electric field were varied to determine the parameters which
ensured consistent injection. The key parameter to determine
consistent and predictable injection for each respective
condition is the coefficient of variance (CV) of the mean
fluorescence of the droplets post injection. For the initial
experiments, droplets were generated on-chip just before the
picoinjector module. Under these conditions the picoinjector
worked consistently over a range of conditions with a low CV
2–6%. Actuation of the picoinjector by syringe pump was
compared to actuation by pressure controller and no discern-
able difference in performance was found. The strength of
electric field required to yield injection of liquid into droplets
was evaluated and similarly to earlier reports a voltage
threshold value for injection was found.13 Increasing the
voltage beyond the threshold had no observable effect. The
threshold voltage was found to be 200 V.
In addition we evaluated the performance of the picoinjector
module when injecting fluorescein (lmax,em = 521 nm) into a
pre-generated, off-chip collected and reinjected emulsion
containing rhodamine B (lmax,em = 625 nm). By measuring
the droplet fluorescence in two optical channels (centered 525
nm and 593 nm) after the picoinjector we could determine
that all droplets were injected with fluorescein from the
picoinjector and that only very infrequently were droplets
generated from the picoinjector channel (,1% of total events).
These picoinjector droplets are easily discriminated by their
size. By varying the relative flow rate of the picoinjector
channel to the emulsion various volume fractions could be
injected into the droplets. Injected fractions ranging from 20
to 60% of the resulting droplet volume were evaluated. Typical
CV of the 525 nm fluorescence of the droplets post-injection
was 6% for the flow rates tested (Fig. 2B). Furthermore, the
picoinjector was tested for an extended period of time and was
shown to consistently inject droplets with a CV of 4.4% over
the course of 30 min.
Finally, we tested if the picoinjector could be used to add
reagent to a barcoded droplet library while preserving the
integrity of the barcode. A droplet library with droplets of five
distinct fluorescein concentrations was prepared and the
droplets were injected with one droplet volume of RBG using
the picoinjector. The 525 nm fluorescence of the droplets post-
injection was evaluated and it was found that the barcode of
the five droplet populations was still resolvable after picoinjec-
tion (Fig. 2C–D).
Multiplex enzyme assay
The microfluidic system was demonstrated using the well
known model enzyme b-galactosidase (b-gal), which cleaves
the substrate resorufin-b-galactopyranoside (RBG) into fluor-
escent resorufin (lmax,em = 589 nm) and galactose. The enzyme
reaction was inhibited by isopropyl-b-thiogalactopyranoside
(IPTG) and the inhibitor concentration was barcoded using
fluorescein (lmax,em = 521 nm) (Fig. 3).
A droplet library with five chemically distinct populations
was created by mixing two solutions (pH 6.9), both of which
contained the same enzyme concentration (0.86 mg mL21) but
differed in that one solution also contained IPTG (300 mM) and
fluorescein (5 mM). The solutions were mixed in a T-junction
immediately prior to the droplet generation chip at the ratios
Fig. 2 Picoinjector module characterization. A) Micrograph of the picoinjector
module. The function of the picoinjector is to add a defined volume of reagent
to each droplet as it passes the vertical nozzle from left to right. Darker areas at
the bottom are electrodes. B) Graph showing mean fluorescence of droplets
after picoinjection of various volumes fractions of fluorescein into reinjected
emulsion containing rhodamine B. Error bars denote one standard deviation.
Each mean value is constructed from >40 000 droplet events. C) Histogram
showing fluorescence of a droplet library with five distinct concentrations of
fluorescein after generation. N.B. the population with no fluorescein content
could not be detected. D) Histogram showing the fluorescence of the same
droplet library as in C) after picoinjection of one droplet volume of RBG into
each droplet. Note, the population with no fluorescein can now be detected as
the RBG injected gives a signal in the 593 nm channel.
This journal is ß The Royal Society of Chemistry 2013
Lab Chip, 2013, 13, 1754–1761 | 1757

View Article Online
Paper
Lab on a Chip
of 100%, 75%, 50%, 25% and 0% of the first solution with the
remainder constituting of the second solution. The differential
fluorescein concentration of the resulting droplet populations
allowed for later identification of the inhibitor concentration
of each droplet from the fluorescence intensity in the 525 nm
channel.
The droplet library was transferred to the second circuit
where the droplets were spaced with fluorocarbon oil and each
droplet was injected with RBG as it passed the picoinjector
module (300 Vp2p, 25 kHz). The enzyme reaction progress
could be followed on-chip by measuring droplet fluorescence
in the spatially separated measurement points, each measure-
ment point representing a defined incubation time. The 593
nm fluorescence intensity of each droplet event could be
correlated to the resorufin product concentration using a
standard curve (Fig. 4A). The inhibitor concentration of each
droplet event could be decoded according to its 525 nm
Fig. 3 Model enzyme–inhibitor system assayed. Resorufin galactopyranoside
(RBG) is hydrolyzed by b-galactosidase into the fluorescent product resorufin
(lmax,em = 589 nm) and galactose. The enzyme reaction is inhibited by isopropyl
thiogalactopyranoside (IPTG).
Fig. 4 Droplet data collection and barcode decoding. A) Graph showing a representative raw data set, each black dot representing one droplet event. y-axis shows
resorufin concentration derived from the measured 593 nm fluorescence and x-axis shows the each of the six measurement points, which corresponds to the
incubation time. B) Histogram of barcode fluorescence (525 nm channel) for all of the droplet events shown in A. Color highlights the five droplet populations with
different inhibitor concentrations. C) Raw droplet event data decoded for inhibitor content by gating the raw data set for the colored areas in B. D) Data sets were
gathered for the five substrate concentrations: 17, 54, 92, 129 and 167 mM RBG in a single experiment by changing the concentration of substrate supplied through
the picoinjector. ESI Fig. S33 shows data sets for all five substrate concentrations.
1758 | Lab Chip, 2013, 13, 1754–1761
This journal is ß The Royal Society of Chemistry 2013

View Article Online
Lab on a Chip
Paper
fluorescent signal (Fig. 4B–C). A data set was gathered by
recording 20 s of two-channel droplet fluorescence data at
each of the six designated measurement points in the
incubation channel. The concentration of RBG injected into
the droplets through the picoinjector was subsequently
changed by varying the flow rate of the two syringes (33 and
333 mM RBG) that fed the picoinjector inlet through an off-
chip T-junction. Data sets were acquired for a total of five
substrate concentrations ranging from 17 to 167 mM of RBG
(Fig. 4D and S3 ). Thus, the enzyme reaction could be studied
3
at combinations of five inhibitor and five substrate concentra-
tions. Furthermore, a total of three replicate data sets were
gathered for each substrate condition and the replicate data
sets were found to be in good agreement for each of the
substrate concentrations (see ESI for further discussion).
3
After decoding the droplet events for inhibitor concentration
and stratifying for substrate concentration a line could be fit to
the data using simple linear regression, the slope of this line
representing the reaction velocity. One linear fit was produced
for each combination of inhibitor and substrate concentration
to extract 25 unique reaction velocities (Fig. 5A–B and S4, ESI ).
3
The reaction velocities had the unit mM per measurement
point but could be converted to mM s21 knowing the total flow
rate and the channel volume between each of the measure-
ment points.
Subsequently, a line was fit to the reaction velocities
extracted for each of the five inhibitor concentrations using
the Lineweaver–Burke equation, a linearized version the
Michaelis–Menten equation (eqn (3)), traditionally used to
extract the enzyme kinetic constants.
1
1
Km,app
1
~
z
(4)
V0
½Sꢀ Vmax,app
Vmax,app
where Km,app and Vmax,app are the apparent Km and Vmax values
respectively and can be extracted from the intercept and the
slope of the Lineweaver–Burk plot (Fig. 5C).
For the no inhibitor case the Km,app and Vmax,app represents
the true Km and Vmax values which were determined as: Km
=
Fig. 5 Enzyme kinetics and inhibition analysis A) Reaction velocities were
442 ¡ 99 mM and Vmax = 0.99 ¡ 0.20 mM s21. The Km value
agrees well with previously published values for b-gal and RBG,
which range between 320 and 650 mM.8,9,21,27 The turnover
number was calculated from Vmax using the enzyme concen-
tration and eqn (2) as kcat = 1070 ¡ 210 s21. The turnover
number is more difficult to compare to previously published
results because it is highly dependent on the purity of the
extracted from the data sets after gating for inhibitor concentration with simple
linear regression to yield 25 unique reaction velocities, one for each
combination of substrate concentration and inhibitor concentration. The
uncertainty in the values of the extracted velocities ranged from ¡0.4–2.2%
using a 95% confidence interval. Three replicate data sets for each substrate
concentration was acquired and agreed well with an average CV in extracted
velocity of 3% between replicates. ESI Fig. S43 shows the corresponding reaction
velocity graph for all substrate concentrations. The linear regression was fit to all
droplet events for respective inhibitor and substrate conc., even though the data
are shown as a mean value and one standard deviation in the graph. B) Table
showing the 25 reaction velocities extracted in A) for the various combinations
of inhibitor and substrate concentrations. C) A Lineweaver–Burke plot was
produced using the obtained reaction velocities for each of the five inhibitor
concentrations to extract the enzyme kinetics constants and the enzyme
inhibitor disassociation constants. The no inhibitor case was used to determine
Km and kcat and the inhibited cases were used to determine ki and ki9.
enzyme.2 One publication has reported the kcat to be 730 s21 27
For the inhibited cases the Km,app and Vmax,app is given as
.
ꢀ
ꢁ
Km,app
Km
½Iꢀ
~
1z
(5)
Vmax,app
Vmax
ki
and
ꢀ
ꢁ
1
1
½Iꢀ
ki’
~
1z
(6)
Vmax,app
Vmax
and ki9 = 62 mM respectively (r2 of 0.96 and 0.98 respectively).
For comparison, a previous publication has reported ki = 0.08
mM.28
Knowing the value of Km and Vmax, the ki and the ki9 could be
extracted with linear regression from the intercept and slope
values of the inhibited cases using eqn (5) and (6) as ki = 49 mM
This journal is ß The Royal Society of Chemistry 2013
Lab Chip, 2013, 13, 1754–1761 | 1759

View Article Online
Paper
One limitation in the determination of the kinetic para-
Lab on a Chip
3 J. M. Fox, S. E. Levine, H. W. Blanch and D. S. Clark,
Biotechnol. J., 2012, 7, 361–373.
4 B. Yang, Z. Dai, S.-Y. Ding and C. E. Wyman, Biofuels, 2011,
2, 421–450.
5 D. Voet, J. G. Voet and W. P. Pratt, Fundamentals of biochemistry,
John Wiley & Sons, Inc., 1999.
6 K. Wilson and J. Walker, ed., Principles and techniques of
biochemistry and molecular biology (6th Ed.), John Wiley &
Sons Inc., 2005.
7 D. C. Duffy, H. L. Gillis, J. Lin, N. F. Sheppard and G.
J. Kellogg, Anal. Chem., 1999, 71, 4669–4678.
8 S. Jambovane, E. C. Duin, S.-K. Kim and J. W. Hong, Anal.
Chem., 2009, 81, 3239–3245.
9 A. G. Hadd, D. E. Raymond, J. W. Halliwell, S. C. Jacobson
and J. M. Ramsey, Anal. Chem., 1997, 69, 3407–3412.
10 M. Kerby and R.-L. Chien, Electrophoresis, 2001, 22, 3916–3923.
11 H. Song and R. F. Ismagilov, J. Am. Chem. Soc., 2003, 125,
14613–14619.
12 M. Chabert, K. D. Dorfman and J.-L. Viovy, Electrophoresis,
2005, 26, 3706–3715.
13 A. R. Abate, T. Hung, P. Mary, J. J. Agresti and D. A. Weitz,
Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 19163.
meters for the b-gal and RBG system is the limited solubility of
the enzyme substrate.9 Ideally, the assay should be performed
at a range of substrate concentrations both above and below
the Km of the substrate but the concentration range was
limited by the solubility of RBG.2 Since this cannot be done
due to the solubility limit of RBG, a higher uncertainty is
expected in the intercept value of the Lineweaver–Burke plot
fits, which may introduce a bias in the determination of Vmax
and ki9.
Conclusions
We present a novel system capable of extensive enzyme
kinetics and inhibition characterization in a single experiment
using picoliter droplets. This microfluidic system enables
multiplex experiments where the data is decoded using a
fluorescent barcode. The small amount of reagent used in the
experiments provides a substantial cost advantage to conven-
tional micro plate based methods. We also demonstrate the
use of a picoinjector module to deliver reagent to microfluidic
droplets and we show that the picoinjector module is both
accurate and robust. By increasing the complexity of the
barcode, this system could be used to screen a wide variety of
inhibitors and/or inhibitor concentrations in a single experi-
ment for example to measure inhibition of pharmaceutically
relevant enzymes, which constitutes the targets of almost half
of the of orally administered pharmaceuticals today.1 The
system could also be used in enzyme engineering to
conveniently screen the performance of new enzyme variants
for desired attributes for example in biofuels production
where enzymes are designed to avoid inhibition by naturally
occurring inhibitors in the feedstock. The main limitation of
the system is that a fluorescent product is required for readout
of the kinetics, which is a typical requirement for most droplet
microfluidic assays. Given that a fluorescent product is
available, the device could be used to study multiplex reaction
kinetics of just about any biochemical or chemical system in a
cost-efficient and high-throughput manner with minor mod-
ifications to the workflow.
´
14 H. N. Joensson, M. Uhlen and H. A. Svahn, Lab Chip, 2011,
11, 1305–1310.
15 J. C. Baret, O. J. Miller, V. Taly, M. Ryckelynck, A. El-Harrak,
L. Frenz, C. Rick, M. L. Samuels, J. B. Hutchison, J. J. Agresti, D.
R. Link, D. A. Weitz and A. D. Griffiths, Lab Chip, 2009, 9,
1850–1858.
16 C. Holtze, A. C. Rowat, J. J. Agresti, J. B. Hutchison, F.
E. Angile, C. H. Schmitz, S. Koster, H. Duan, K. J. Humphry,
R. A. Scanga, J. S. Johnson, D. Pisignano and D. A. Weitz, Lab
Chip, 2008, 8, 1632–1639.
17 E. Brouzes, M. Medkova, N. Savenelli, D. Marran, M. Twardowski,
J. B. Hutchison, J. M. Rothberg, D. R. Link, N. Perrimon and M.
L. Samuels, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 14195–14200.
18 D. Pekin, Y. Skhiri, J. C. Baret, D. Le Corre, L. Mazutis, C. B. Salem,
F. Millot, A. El Harrak, J. B. Hutchison, J. W. Larson, D. R. Link,
P. Laurent-Puig, A. D. Griffiths and V. Taly, Lab Chip, 2011, 11,
2156–2166.
19 J. J. Agresti, E. Antipov, A. R. Abate, K. Ahn, A. C. Rowat, J.-
C. Baret, M. Marquez, A. M. Klibanov, A. D. Griffiths and D.
A. Weitz, Proc. Natl. Acad. Sci. U. S. A., 2010, 107,
4004–4009.
20 H. N. Joensson, M. L. Samuels, E. R. Brouzes, M. Medkova,
´
M. Uhlen, D. R. Link and H. Andersson-Svahn, Angew.
Chem., Int. Ed., 2009, 48, 2518–2521.
Acknowledgements
21 M.-P. N. Bui, C. A. Li, K. N. Han, J. Choo, E. K. Lee and G.
H. Seong, Anal. Chem., 2011, 83, 1603–1608.
22 J. Clausell-Tormos, A. D. Griffiths and C. A. Merten, Lab
Chip, 2010, 10, 1302–1307.
23 O. J. Miller, A. E. Harrak, T. Mangeat, J.-C. Baret, L. Frenz,
B. E. Debs, E. Mayot, M. L. Samuels, E. K. Rooney, P. Dieu,
M. Galvan, D. R. Link and A. D. Griffiths, Proc. Natl. Acad.
Sci. U. S. A., 2011, 109, 378–383.
This research was supported by The Novo Nordisk Foundation
Center for Biosustainability (CFB) and the ProNova VINN
Excellence Centre for Protein Technology. We would like to
acknowledge RainDance Technologies for generously provid-
ing the surfactant used in the experiments.
24 C.-H. Chen, M. A. Miller, A. Sarkar, M. T. Beste, K. B. Isaacson, D.
A. Lauffenburger, L. G. Griffith and J. Han, J. Am. Chem. Soc., 2013,
135, 1645–1648.
25 J. C. McDonald, J. R. Anderson, D. T. Chiu, H. Wu, O.
J. Schueller and G. M. Whitesides, Electrophoresis, 2000, 21,
27–40.
References
1 R. A. Copeland, M. R. Harpel and P. J. Tummino, Expert
Opin. Ther. Targets, 2007, 11, 967–978.
2 R. A. Copeland, Enzymes: a practical introduction to
structure, mechanism and data analysis, Wiley-VCH, 2000.
1760 | Lab Chip, 2013, 13, 1754–1761
This journal is ß The Royal Society of Chemistry 2013

View Article Online
Lab on a Chip
Paper
26 L. Frenz, K. Blank, E. Brouzes and A. D. Griffiths, Lab Chip,
2009, 9, 1344–1348.
27 J. Hofmann and M. Sernetz, Anal. Chim. Acta, 1984, 163,
28 D. H. Juers, B. Rob, M. L. Dugdale, N. Rahimzadeh,
C. Giang, M. Lee, B. W. Matthews and R. E. Huber, Protein
Sci., 2009, 18, 1281–1292.
67–72.
This journal is ß The Royal Society of Chemistry 2013
Lab Chip, 2013, 13, 1754–1761 | 1761