Effects of molecular confinement and crowding on horseradish peroxidase kinetics using a nanofluidic gradient mixer

Article abstract of DOI:10.1039/c5lc01413a

The effects of molecular confinement and crowding on enzyme kinetics were studied at length scales and under conditions similar to those found in biological cells. These experiments were carried out using a nanofluidic network of channels constituting a nanofluidic gradient mixer, providing the basis for measuring multiple experimental conditions simultaneously. The 100 nm × 40 μm nanochannels were wet etched directly into borosilicate glass, then annealed and characterized with fluorescein emission prior to kinetic measurements. The nanofluidic gradient mixer was then used to measure the kinetics of the conversion of the horseradish peroxidase (HRP)-catalyzed conversion of non-fluorescent Amplex Red (AR) to the fluorescent product resorufin in the presence of hydrogen peroxide (H₂O₂). The design of the gradient mixer allows reaction kinetics to be studied under multiple (five) unique solution compositions in a single experiment. To characterize the efficiency of the device the effects of confinement on HRP-catalyzed AR conversion kinetics were studied by varying the starting ratio of AR:H₂O₂. Equimolar concentrations of Amplex Red and H₂O₂ yielded the highest reaction rates followed by 2:1, 1:2, 5:1, and finally 1:5 [AR]:[H₂O₂]. Under all conditions, initial reaction velocities were decreased by excess H₂O₂. Crowding effects on kinetics were studied by increasing solution viscosity in the nanochannels in the range 1.0-1.6 cP with sucrose. Increasing the solution viscosities in these confined geometries decreases the initial reaction velocity at the highest concentration from 3.79 μM min^-1 at 1.00 cP to 0.192 μM min^-1 at 1.59 cP. Variations in reaction velocity are interpreted in the context of models for HRP catalysis and for molecular crowding.

Full text of DOI:10.1039/c5lc01413a

Lab on a Chip
View Article Online
View Journal
PAPER
Effects of molecular confinement and crowding
on horseradish peroxidase kinetics using a
nanofluidic gradient mixer†
Cite this: DOI: 10.1039/c5lc01413a
a
b
ab
William R. A. Wichert, Donghoon Han and Paul W. Bohn*
The effects of molecular confinement and crowding on enzyme kinetics were studied at length scales and
under conditions similar to those found in biological cells. These experiments were carried out using a
nanofluidic network of channels constituting a nanofluidic gradient mixer, providing the basis for measuring
multiple experimental conditions simultaneously. The 100 nm × 40 μm nanochannels were wet etched di-
rectly into borosilicate glass, then annealed and characterized with fluorescein emission prior to kinetic
measurements. The nanofluidic gradient mixer was then used to measure the kinetics of the conversion of
the horseradish peroxidase (HRP)-catalyzed conversion of non-fluorescent Amplex Red (AR) to the fluores-
2 2
cent product resorufin in the presence of hydrogen peroxide (H O ). The design of the gradient mixer al-
lows reaction kinetics to be studied under multiple (five) unique solution compositions in a single experi-
ment. To characterize the efficiency of the device the effects of confinement on HRP-catalyzed AR
conversion kinetics were studied by varying the starting ratio of AR: H
Amplex Red and H yielded the highest reaction rates followed by 2 : 1, 1 : 2, 5: 1, and finally 1 : 5
AR]: [H ]. Under all conditions, initial reaction velocities were decreased by excess H . Crowding ef-
2 2
O . Equimolar concentrations of
2 2
O
[
O
2 2
2 2
O
Received 17th November 2015,
Accepted 8th January 2016
fects on kinetics were studied by increasing solution viscosity in the nanochannels in the range 1.0–1.6 cP
with sucrose. Increasing the solution viscosities in these confined geometries decreases the initial reaction
−1
−1
velocity at the highest concentration from 3.79 μM min at 1.00 cP to 0.192 μM min at 1.59 cP. Varia-
tions in reaction velocity are interpreted in the context of models for HRP catalysis and for molecular
crowding.
DOI: 10.1039/c5lc01413a
www.rsc.org/loc
1
9–23
beam lithography.
Nanopores are usually fabricated by
Introduction
nanosphere lithography or FIB, while nanochannels are typi-
cally made by FIB milling which permits control over channel
dimensions simply by altering the ion-beam current.
Utilization of nanopores in track-etched membranes has
proven to be an especially useful addition to fluidic
Lab-on-a-chip devices have been fabricated with increasing
complexity as the characteristic dimensions are reduced from
the millimeter to the micrometer scale. Devices, typically
2
4,25
1
2
3
made from glass, polystyrene, PDMS, and now 3D print-
4
26–28
ing, have been used for a wide variety of biological applica-
devices,
aiding in gas separations, protein analysis and
5
,6
29,30
tions including in vitro cellular analysis, biological commu-
nanofiltration through application of polymer brushes.
Bruening et al., for example, immobilized polyIJstyrene
sulfonate) and trypsin inside of membrane pores to increase
7
8–10
nication, and enzyme kinetics.
With the increasing
availability of nanofabrication capabilities, lab-on-a-chip de-
vices have begun to exploit nanoscale properties
1
1–13
31
in the
the speed and efficiency of protein digestion. One key issue
1
4–16
17
physical form of nanopores,
glass fiber bundles, and
affecting reaction kinetics of immobilized enzymes and other
proteins in nanoscale architectures is the protein configura-
tion and the potential for steric hindrance that needs to be
considered when utilizing surface-immobilized bio-
macromolecules. Alternatively, if reagent use issues are not
limiting, then using non-immobilized enzymes can yield
higher efficiency reactors and produce results that are more
1
8
nanochannels. Nanoscale devices have been fabricated in a
number of ways including wet-chemical etching, nanosphere
lithography, focused ion beam (FIB) milling, and electron
a
Department of Chemistry and Biochemistry, University of Notre Dame, Notre
Dame, IN 46556, USA. E-mail: pbohn@nd.edu
32,33
readily connected to in vivo kinetics.
b
Department of Chemical and Biomolecular Engineering, University of Notre
Nanoscale devices are of particular interest for studying
how molecular confinement affects reaction efficiency at cel-
Dame, Notre Dame, IN 46556, USA
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
1
7,27,34,35
c5lc01413a
lular and molecular length scales.
In previous work
This journal is © The Royal Society of Chemistry 2016
Lab Chip

View Article Online
Paper
Lab on a Chip
from our laboratory, the kinetics of horseradish peroxidase
in order to obtain multiple sets of experimental conditions
within a single experimental run. Integrating a fluidic gradi-
ent mixer on-chip has been shown to be an efficient way to
generate spatial and temporal gradients of analyte concentra-
tion for use in downstream experiments. Thus, as a first
step towards developing a robust and broadly applicable ex-
perimental system capable of probing biomacromolecular re-
actions over wide ranges of conditions simulating the cytosol,
we have developed a nanofluidic gradient mixer capable of si-
multaneously producing multiple homogeneous solution re-
action conditions at different locations on the chip.
(
HRP) immobilized in a track-etched membrane were studied
by measuring the production of a fluorescent product
resorufin) from non-fluorescent reagents (Amplex Red and
H O ) downstream of the membrane in a polydimethylsilox-
(
4
4
2
2
2
7
ane (PDMS) microchannel.
Unfortunately, because the
nanopores were oriented perpendicular to the membrane sur-
face, it was not possible to follow the progress of the reaction
spatially, a difficulty when trying to measure kinetics in real-
time. Utilizing nanofluidic channels that are aligned parallel
to the substrate surface is
a
useful alternative to
perpendicularly-oriented nanopores, because the entire
length of the nanochannel is accessible to optical interroga-
tion, and the reaction extent can be measured directly as a
function of position, and reaction velocities can be obtained
directly by imaging.
Expanding on a wet-etching method used for micro-
1
9
channels, we developed a technique for controllably etching
channels in borosilicate glass to produce channel depths
ranging from 40–100 nm, thus realizing nanofluidic geome-
tries with aspect (width-to-depth) ratios exceeding 500. This
architecture thus supports an electrokinetically-driven nano-
fluidic gradient mixer that takes two arbitrary initial concen-
trations of reactants and mixes them through a series of bi-
furcated and trifurcated channels, yielding a lateral (relative
to the direction of flow) spatial concentration gradient rang-
ing between the two starting concentrations. This gradient is
then introduced, via electrokinetic flow, into a series of five
reaction channels, arranged parallel to the glass surface (i.e.
horizontal nanochannels). Once the reactants have been in-
troduced into the reaction channels, the electrokinetic flow is
stopped and the progress of the reaction is followed by fluo-
rescence microscopic imaging. This device architecture al-
lows multiple concentrations and conditions to be tested si-
multaneously. The nanofluidic mixer was characterized by
measuring the effects of confinement and crowding on the
kinetics of the HRP-catalyzed conversion of Amplex Red to
resorufin.
Horseradish peroxidase (HRP) is an excellent model en-
zyme, as it can accept a range of substrates, but it possesses
only a single well-characterized catalytic site, a heme pros-
3
6
thetic group. It has been immobilized to solid substrates in
a variety of configurations, retaining its activity even in the
3
7
presence of possible complications due to steric hindrance.
3
8
HRP is known to oxidize both Scopoletin as well as Amplex
3
9
Red (AR), making it possible to measure its enzymatic activ-
ity by monitoring product fluorescence, since the overall reac-
tion is fluorogenic. The AR reaction is favored over Scopoletin
due to its favorable extinction coefficient and the quantum
efficiency of the product resorufin, as well as the fact that it
exhibits an increase in fluorescence upon oxidation, rather
than a decrease, as with Scopoletin. In addition, the mecha-
nism of oxidation of AR is now well-understood, involving an
2 2
initial one electron transfer from H O to form an FeIJIV) oxo
intermediate, which is followed by a pair of 1-electron/1-
proton transfers and a final step in which two Amplex Red
radicals undergo a dismutation reaction to form one mole- Experimental
1
7
cule each of resorufin and Amplex Red. Finally, HRP is a 44
kDa protein with an approximate molecular size of 6.0 nm ×
Materials
Schott Glass D substrates were purchased from Applied
Microarrays; S1813 photoresist from Fisher Scientific; horse-
radish peroxidase, hydrogen peroxide, and sucrose from
Sigma-Aldrich; and Amplex Red from Life Technologies. All
solutions were prepared using 50 mM phosphate buffer (pH
3
.5 nm × 3.0 nm, so it occupies a significant fraction of the
height (≤100 nm) of the nanochannels developed for this
work, making it a good candidate to study the effects of mo-
lecular confinement.
In order to accurately mimic the behavior of the enzyme
in vivo, measurements of enzyme kinetics should ideally re-
produce the conditions found in the cell. The difficulty in do-
ing so is that the composition of the cellular cytosol is highly
4
0
7
.4) and deionized (DI) water (ρ ∼ 18.2 MΩ cm, EMD Milli-
pore). A 10 mM stock solution of AR was prepared by dissolv-
ing in DMSO (Fisher Scientific).
4
1,42
variable with varying effects on both molecular transport
4
3
Nanochannel fabrication
and reactivity, although current understanding suggests
that molecular crowding increases rates for reactions, like
slow association reactions, that are limited by intrinsic chem-
ical characteristics of the activation barrier, while it decreases
rates for diffusion-limited reactions. This is consistent with
the observation that diffusion is slowed in the cytosol by a
few-fold in relation to diffusion in aqueous solutions of the
Nanofluidic devices were fabricated using conventional
photolithography and wet chemical etching. Widths were
controlled precisely using a chrome photomask while depths
were controlled by varying the etch periods. Following a 60 s
vapor deposition of hexamethyldisilazane (HMDS), the gradi-
ent mixer design was transferred to a photoresist covered
4
3
−2
same viscosity but without macromolecular crowding.
glass substrate via a 5 s UV exposure at 14 mW cm (Karl
In order to efficiently probe multiple reaction conditions
in the shortest possible time, fluidic mixing may be utilized
Suss MJB 3 mask aligner). After the development of photore-
sist in AZ917-MIF developer, the nanochannels were etched
Lab Chip
This journal is © The Royal Society of Chemistry 2016

View Article Online
Lab on a Chip
Paper
with a 2 : 1 : 1 v : v : v mixture of 15 mM HF, 7.5 mM NH F, 7.5
55m, Chroma). Initial rates were then calculated by first
converting resorufin fluorescence intensity into concentration
using a calibration curve (see ESI† Fig. S2) then calculating
the initial slope of the kinetic curves (ESI† Fig. S3). Initial
starting ratios of 1 : 1, 1 : 2, 1 : 5, 2 : 1, and 5 : 1 100 μM AR :
100 μM H O with 0.025 units per mL (2.3 nM) HRP were ex-
2 2
amined. In these experiments, n : 1 represents a solution in
which n = 100 μM and the other reactant concentration is
4
mM HNO3 etch solution to a typical depth of 100 nm and
width of 40 μm. Channel dimensions were characterized
using a stylus-based profilometer (Tencor P-6, KLA-Tencor In-
struments) (ESI† Fig. S1). Fluidic access holes were drilled at
the ends of the channels as a means to introduce solutions
and permit electrode placement to drive electrokinetic flow.
(
1/n) × 100 μM.
Device annealing
Blank glass substrates and glass substrates containing the
etched nanochannels were first subjected to high pressure DI
water and then sonicated to remove residual glass particles
from drilling the fluidic access holes. These substrates were
then placed in a 3 : 1 piranha (H SO : H O ) solution for 30
2 4 2 2
min, rinsed with DI water, and immersed in 3 : 1 : 1 RCA
Crowding measurements
Molecular crowding experiments were conducted using su-
crose as the viscogen over a range of concentrations/viscosi-
ties. AR (100 μM), HRP (2.3 nM), and hydrogen peroxide (100
μM) solutions were prepared using 0, 5, 10, and 15% sucrose
solutions, yielding viscosities ranging from 1.0–1.6 cP.
Analytes were introduced and studied in the same manner as
with the concentration and confinement measurements. Ini-
tial rates were again used to compare kinetic effects under
crowding conditions.
solution (NH OH : H O : H O) at 60 °C for 30 min. Finally the
4
2
2
2
glass substrates were rinsed with DI water and annealed at
60 °C for 8 h. Prior to use, devices were filled with buffer
5
and left overnight to ensure a uniform surface charge
density.
Results and discussion
Enzyme kinetics measurements
Device design and fabrication
The positions of reactants and enzyme (AR, H
controlled electrokinetically with a high voltage sequencer
HVS448, LabSmith, Inc.) using the voltage scheme in Fig. 1.
2 2
O , HRP) were
Microfluidic gradient mixers have been built and demon-
strated to generate spatial and temporal gradients of concen-
tration with a great deal of flexibility in controlling design pa-
(
Electrokinetic flow was used to introduce the reactants
through the mixer portion of the device and into the reaction
channels for a period of 5 min, after which, the external volt-
age was shut off, and fluorescence measurements were ac-
4
4
rameters with fabrication choices. The design explored here
is an electrokinetically-driven nanofluidic gradient mixer that
takes two initial reactant concentrations and mixes them
through a series of intersecting channels, yielding a lateral
(relative to the direction of flow) spatial concentration gradi-
ent ranging between the two starting concentrations. This
gradient is then introduced into a series of reaction channels,
allowing multiple concentrations to be tested simultaneously.
In the design tested here, five separate reactions channels
−
1
quired 1 min for the remainder of the experiment using an
inverted optical microscope (IX-71, Olympus, Inc.). Care was
taken to shutter the excitation source between data acquisi-
tions to avoid photooxidation of AR. A mercury arc lamp was
used as the excitation source with appropriate excitation, di-
chroic and emission filters (AT540/25x, AT565DC, AT605/
Fig. 1 Schematic diagram of the electrokinetically driven nanofluidic gradient mixer, permitting multiple concentration experiments to be carried
out simultaneously. (a) Buffer is introduced along with reactant and mixed through a series of intersecting channels to form a concentration
gradient in the lateral direction. (b) Fluorescence image of the device filled with 100 μM fluorescein and buffer acquired from the gradient column
just above the lateral diffusion region. (c) Fluorescence image of the device acquired with 100 μM fluorescein and buffer obtained in the reaction
nanochannels, illustrating that each nanochannel contains a unique, laterally-varying concentration of reactant.
This journal is © The Royal Society of Chemistry 2016
Lab Chip

View Article Online
Paper
Lab on a Chip
were built, although the number can be increased as desired
to characterize the reaction of interest. The nanofluidic mixer
was initially characterized by imaging the mixing of 100 μM
fluorescein with 50 mM pH 7.4 phosphate buffer, as shown
in Fig. 1. As a model enzymatic reaction the effects of con-
finement and crowding on the kinetics of the HRP-catalyzed
fluorogenic conversion of AR to resorufin in the presence of
an oxidant (H O here), Scheme 1, were studied. The forma-
∇·(−D∇c) + u·∇c = R
where D is the diffusion coefficient of the reactant, c is the
concentration of the reactant, and R is the consumption rate
of the reactant. Fig. 2 illustrates the simulated concentration
profile as a function of position throughout the mixer. As
shown, the simulations of gradient mixer produce a linear
concentration range across the five reaction channels, consis-
tent with fluorescence measurements shown in Fig. 1.
2
2
tion of the highly fluorescent product, resorufin, was moni-
tored by imaging the fluorescence from multiple nano-
channels, each exhibiting
a different composition, the
Nanofluidic channel measurements
resulting integrated fluorescence intensities then being used
to determine reaction kinetics at multiple reaction conditions
simultaneously in the nanofluidic gradient mixer.
It is by now well-appreciated that nanofluidic flow differs
substantially from flow in larger, even microfluidic, chan-
4
5
Rectangular nanochannels were fabricated in glass using
standard photolithography and wet chemical etching to pro-
duce the gradient mixer depicted in Fig. 1. The resulting de-
vice consists of a series of intersecting channels that permit
mixing of analytes and flow into the gradient column which
combines the contents of the mixing channels, yielding a
concentration gradient ranging between each of the starting
reactant concentrations. The gradient column provides each
reaction channel with a unique concentration of analytes,
allowing reactions to be carried out under multiple condi-
tions and studied simultaneously. In a typical device the
etched nanochannels were measured by profilometry to be 93
nels, so the initial studies examined the use of the nano-
fluidic gradient mixer to characterize the HRP-catalyzed con-
version of AR to resorufin in the presence of H O , utilizing
2
2
the nanofluidic gradient mixer and fluorescence microscopy.
To introduce AR, H O , and HRP, 100 V is applied across the
2
2
device (top to bottom). The resulting fluid flow mixes reac-
tants and buffer, forming a concentration gradient at the gra-
dient column that introduces unique concentrations into
each of the reaction channels. Fig. S3† shows the fluores-
cence response with time in a typical experiment. Raw fluo-
rescence data were acquired at a rate 1 min over a period of
1 h and were converted to resorufin concentration using a
standard working curve (Fig. S2†). The initial slope of the
fluorescence response was used to determine the initial reac-
tion rates, the determination of which provides a means of
characterizing the effect of starting reactant concentration ra-
tio and insight into how the nanofluidic geometry affects re-
action velocity. The results are depicted in Fig. 3, where the
−
1
±
12 nm in depth and 40 μm in width, as shown graphically
in ESI† Fig. S1.
To assess the flow dynamics and mixing behavior, finite
element simulations were performed over a two-dimensional
domain representing the geometry and dimensions of the
mixer employed in these experiments. A laminar flow,
electrostatics, stationary study using transport of dilute spe-
cies was carried out in order to simulate how samples mix
with the sheath solution. Within the mixer, the flow velocity,
u, is dictated by the incompressible Navier–Stokes equation,
0
initial reaction velocities, r , corresponding to the different
AR: H O ratios are shown in relation to the reaction nano-
2
2
channel in which they were obtained.
T
ρ(u·∇)u = ∇·[−pI + μ(∇u + (∇u) )] + F
(1)
where ρ is the fluid density, p is the pressure, I is the identity
metric, μ is the dynamic viscosity, and F is a volume force,
such as gravity. The local concentration of the reactant is the
main result followed in the simulations and is governed by
Fick's law, which is described by the diffusion and convec-
tion equation:
Fig.
2 Finite element simulations demonstrate concentration
distribution within the gradient mixer. The enlarged view shows the
species concentration for the each of the nanofluidic reaction
channels.
Scheme 1
Lab Chip
This journal is © The Royal Society of Chemistry 2016

View Article Online
Lab on a Chip
Paper
Fig. 3 Initial reaction rate as a function of initial reactant concentration ratio measured in the gradient mixer nanochannels. Initial reaction rate
represents the initial slope of product formation with respect to time in Fig. S3.† Percent concentration represents the dilution factor from the
initial 100 μM concentration of AR and H O . AR : H O
2 2 2 2
ratios of 1 : 1, 2 : 1, 1 : 2, 5 : 1, and 1 : 5 were examined. Error bars are ±1 standard deviation of
the initial rate.
Starting AR: H
and 5 : 1 AR: H
centration ratio on reaction kinetics. Under all concentra-
tions 1 : 1 AR : H yielded the highest initial reaction veloc-
2
O
2
reactant ratios of 1 : 1, 1 : 2, 1 : 5, 2 : 1,
ratio of ∼1.3 for 2 : 1 AR: H
2 2 2 2
O compared to 1 : 2 AR : H O
2
O
2
were chosen to determine the effect of con-
and an even greater difference, ∼2.4, for 5 : 1 AR: H
2 2
O com-
pared to 1 : 5 AR : H O .
2
2
2 2
O
ity in each of the reaction channels, and the reaction velocity
Crowding measurements
increases, as expected, with concentration from ∼6 μM to
∼
97 μM for all reactant ratios. At the lowest overall concen-
tration (highest dilution), 1 : 1 AR: H O exhibits the largest
2 2
Altering solution viscosity has been shown to affect macromo-
lecular reaction kinetics. Increasing the solution viscosity is
the most straightforward way to increase molecular crowding,
and it has been shown to monotonically decrease the rates of
diffusion-limited reactions and to increase them for reaction-
r0, while the other ratios yield statistically indistinguishable
rates. Under these conditions we estimate a second order rate
−
1
−1
constant, k = 6.8 ± 0.7 M s . For the remainder of the con-
4
3,46,47
ditions r decreases in the order, 2 : 1, 1 : 2, 5 : 1, and 1 : 5, i.e.
limited situations.
Sucrose solutions with compositions
0
the larger the concentration mismatch the slower the initial
velocity. In addition, in comparing 1 : n to n : 1 (n = 2, 5) AR :
H O ratios, the condition with excess H O yields slower ve-
0, 5, 10, and 15 weight% sucrose in buffer (yielding viscosities
of 1.0, 1.14, 1.33, and 1.59 cP, respectively) were used to in-
crease the viscosity and therefore test its effect on r in the
2
2
2
2
0
locities across all conditions. This is consistent with the fact
that under all conditions there is sufficient H to carry out
the initial HRP oxidation to the FeIJIV) oxo species. After this,
HRP-catalyzed conversion of AR to resorufin. The voltage
scheme and reactant introduction/dilution/mixing schemes
used for these experiments were as depicted in Fig. 1. Fluores-
cence data were collected at a rate of 1 min for 1 h, and the
fluorescence data were again converted to resorufin concen-
tration using the calibration in Fig. S2† and fit to give initial
reaction velocities to compare against viscosity.
2 2
O
−
1
the active site catalyzes a one-electron transfer from Amplex
1
7
Red yielding a radical intermediate. Within this context, de-
creasing the amount of AR available for the second step
would result in a slower reaction velocity.
2
7
Previous work from our laboratory demonstrated that
HRP-catalyzed conversion of AR to resorufin using
immobilized HRP in a vertically-oriented (i.e. perpendicular
to the substrate plane) nanopore yields higher initial rates,
r0, than free solution or planar bulk surface measurements.
These experiments also compared non-stoichiometric AR :
Fig. 4 shows the initial nanochannel reaction rates plotted
against reactant concentration, expressed as a percentage of
the initial 100 μM concentration, using 1 : 1 AR: H O ratios
2
2
for all viscosities. As expected, at all viscosities increasing the
concentration of 1 : 1 AR: H O increases r . Importantly, in-
2
2
0
creasing solution viscosity lowers the initial reaction veloci-
−
1
−1
H
2
O
2
ratios and yielded initial reaction rates ∼1.6 times
higher when AR is in excess compared to excess H O , similar
ties, ranging from 0.01 μM min to 0.19 μM min at the
−
1 −1
highest value, η = 1.59 cP, to 0.21 μM min
min for η = 1.00 cP. These observations are consistent with
l
to 3.79 μM
2
2
−
1
to results obtained here with the nanofluidic mixer, i.e. a
This journal is © The Royal Society of Chemistry 2016
Lab Chip

View Article Online
Paper
Lab on a Chip
Fig. 4 Viscosity measurements. Initial reaction velocity as a function of initial reactant concentration measured in the reaction nanochannels.
Initial reaction rate represents the slope of the product formation per unit time. Sucrose compositions range from 0–15% and had calculated
viscosities from 1.0–1.6 cP. Error bars are ±1 standard deviation.
the HRP-catalyzed reaction being diffusion controlled under
the conditions explored here, however the ∼20-fold differ-
0
ence in r between η = 1.00 cP and η = 1.59 cP greatly exceeds
simple expectations based on differences in diffusion alone,
as predicted, for example, by the Stokes–Einstein equation (D
consistent with recent developments in understanding the
mechanism of HRP-catalyzed oxidation of AR to resorufin. In-
creasing the viscosity inside the nanochannels results in a de-
crease in reaction velocity, consistent with diffusion-control,
but the magnitude of the decrease suggests a crowding-
induced effect on the mechanism of enzyme catalysis.
∼
1/η). This suggests that in addition to its simple effect on
reactant transport, crowding causes a change in the detailed
dynamics governing the mechanism of HRP-catalyzed AR
oxidation.
Acknowledgements
This work was supported by the Department of Energy under
grant DE FG02 07ER15851 (WRAW) and a subcontract from
Oak Ridge National Laboratory (DE-AC05-00OR22725;
Subaward # 4000132808 – DH). The authors gratefully ac-
knowledge the assistance of S. Branagan in the initial design
of the gradient mixer. Fabrication of the devices studied here
was accomplished at the Notre Dame Nanofabrication Facility
whose generous support is gratefully acknowledged.
Conclusions
We have demonstrated a simple way to conduct multiple, si-
multaneous enzyme kinetics experiments using a nanofluidic
gradient mixer. Starting with just two reactant concentra-
tions, intersecting channels meet and mix the reactants to
form a concentration gradient in the gradient column of the
device. This concentration gradient is used to introduce dif-
ferent concentrations of analyte into a series of laterally
displaced reaction nanochannels. The design of the mixer
with a lateral diffusion zone at the end of the gradient mixing
portion of the device allows for efficient an smooth concen-
tration variations across the device that are consistent with
the results of finite-element modeling calculations. This de-
sign may be altered to increase the degree of mixing by
changing the structure of the mixing region or the number of
reaction conditions, by simply increasing the number of reac-
tion nanochannels.
References
1 L. Yi, X. Wang, R. Dhumpa, A. M. Schrell, N. Mukhitov and
M. G. Roper, Lab Chip, 2015, 15, 823–832.
2 A. S. Johnson, K. B. Anderson, S. T. Halpin, D. C.
Kirkpatrick, D. M. Spence and R. S. Martin, Analyst,
2013, 138, 129–136.
3 D. C. Duffy, J. C. McDonald, O. J. A. Schueller and G. M.
Whitesides, Anal. Chem., 1998, 70, 4974–4984.
4 K. B. Anderson, S. Y. Lockwood, R. S. Martin and D. M.
Spence, Anal. Chem., 2013, 85, 5622–5626.
Altering the AR: H O ratio was observed to have a dra-
2
2
matic effect on the initial reaction velocities in this nano-
fluidic geometry, with concentration mismatches showing a
decrease in initial velocity in all cases, with a stronger de-
crease when H O is in excess. These observations are
5 K. B. Anderson, S. T. Halpin, A. S. Johnson, R. S. Martin and
D. M. Spence, Analyst, 2013, 138, 137–143.
6 R. S. Martin, P. D. Root and D. M. Spence, Analyst,
2006, 131, 1197–1206.
2
2
Lab Chip
This journal is © The Royal Society of Chemistry 2016

View Article Online
Lab on a Chip
Paper
7
8
9
M. G. Roper and C. Guillo, Anal. Bioanal. Chem., 2009, 393,
59–465.
A. G. Hadd, D. E. Raymond, J. W. Halliwell, S. C. Jacobson
and J. M. Ramsey, Anal. Chem., 1997, 69, 3407–3412.
27 Z. Wang, T. L. King, S. P. Branagan and P. W. Bohn, Analyst,
2009, 134, 851–859.
28 L. R. Gibson and P. W. Bohn, Interface Focus, 2013, 3, 20120096.
29 M. L. Bruening, D. M. Dotzauer, P. Jain, L. Ouyang and G. L.
Baker, Langmuir, 2008, 24, 7663–7673.
4
G. H. Seong, J. Heo and R. M. Crooks, Anal. Chem., 2003, 75,
3
161–3167.
30 P. Jain, G. L. Baker and M. L. Bruening, in Annual Review of
Analytical Chemistry, Annual Reviews, Palo Alto, 2009, vol. 2,
pp. 387–408.
1
1
1
0 D. C. Duffy, H. L. Gillis, J. Lin, N. F. Sheppard and G. J.
Kellogg, Anal. Chem., 1999, 71, 4669–4678.
1 T. C. Kuo, D. M. Cannon, M. A. Shannon, P. W. Bohn and
J. V. Sweedler, Sens. Actuators, A, 2003, 102, 223–233.
2 A. Piruska, S. P. Branagan, A. B. Minnis, Z. Wang, D. M.
Cropek, J. V. Sweedler and P. W. Bohn, Lab Chip, 2010, 10,
31 F. Xu, W. H. Wang, Y. J. Tan and M. L. Bruening, Anal.
Chem., 2010, 82, 10045–10051.
32 S. J. Li, C. Wang, Z. Q. Wu, J. J. Xu, X. H. Xia and H. Y.
Chen, Chem. – Eur. J., 2010, 16, 10186–10194.
33 C. Wang, D. K. Ye, Y. Y. Wang, T. Lu and X. H. Xia, Lab
Chip, 2013, 13, 1546–1553.
34 J. Zhao, S. P. Branagan and P. W. Bohn, Appl. Spectrosc.,
2012, 66, 163–169.
35 J. C. Yu, Y. J. Zhang and S. Q. Liu, Biosens. Bioelectron.,
2014, 55, 307–312.
36 G. I. Berglund, G. H. Carlsson, A. T. Smith, H. Szoke, A.
Henriksen and J. Hajdu, Nature, 2002, 417, 463–468.
37 C. Wang, S. J. Li, Z. Q. Wu, J. J. Xu, H. Y. Chen and X. H.
Xia, Lab Chip, 2010, 10, 639–646.
38 A. Boveris, E. Martino and A. O. M. Stoppani, Anal. Biochem.,
1977, 80, 145–158.
39 M. J. Zhou, Z. J. Diwu, N. PanchukVoloshina and R. P.
Haugland, Anal. Biochem., 1997, 253, 162–168.
40 W. C. Yeon, B. Kannan, T. Wohland and V. Ng, Langmuir,
2008, 24, 12142–12149.
41 A. S. Verkman, Trends Biochem. Sci., 2002, 27, 27–33.
42 J. A. Dix and A. S. Verkman, Annu. Rev. Biophys., 2008, 37,
247–263.
1
237–1244.
1
1
1
1
3 D. G. Haywood, A. Saha-Shah, L. A. Baker and S. C.
Jacobson, Anal. Chem., 2015, 87, 172–187.
4 L. R. Gibson, S. P. Branagan and P. W. Bohn, Small, 2013, 9,
0–97.
5 C. X. Ma, L. P. Zaino and P. W. Bohn, Chem. Sci., 2015, 6,
173–3179.
9
3
6 Y. H. Lanyon, G. De Marzi, Y. E. Watson, A. J. Quinn, J. P.
Gleeson, G. Redmond and D. W. M. Arrigan, Anal. Chem.,
2
007, 79, 3048–3055.
7 H. H. Gorris and D. R. Walt, J. Am. Chem. Soc., 2009, 131,
277–6282.
8 L. D. Menard and J. M. Ramsey, Nano Lett., 2011, 11,
12–517.
9 C. A. Baker and M. G. Roper, J. Chromatogr. A, 2010, 1217,
743–4748.
0 D. G. Haywood, Z. D. Harms and S. C. Jacobson, Anal.
Chem., 2014, 86, 11174–11180.
1 C. X. Ma, N. M. Contento, L. R. Gibson and P. W. Bohn, ACS
Nano, 2013, 7, 5483–5490.
1
1
1
2
2
2
6
5
4
43 H. X. Zhou, G. N. Rivas and A. P. Minton, Annu. Rev.
Biophys., 2008, 37, 375–397.
44 D. Amarie, J. A. Glazier and S. C. Jacobson, Anal. Chem.,
2007, 79, 9471–9477.
45 A. Piruska, M. Gong, J. V. Sweedler and P. W. Bohn, Chem.
Soc. Rev., 2009, 38, 1–13.
46 Y. Li, Y. Z. Xu, X. J. Feng and B. F. Liu, Anal. Chem.,
2012, 84, 9025–9032.
2 J. M. Perry, K. M. Zhou, Z. D. Harms and S. C. Jacobson, ACS
Nano, 2010, 4, 3897–3902.
2
2
3 P. Mao and J. Y. Han, Lab Chip, 2005, 5, 837–844.
4 Z. D. Harms, D. G. Haywood, A. R. Kneller, L. Selzer, A.
Zlotnick and S. C. Jacobson, Anal. Chem., 2015, 87, 699–705.
5 S. P. Branagan, N. M. Contento and P. W. Bohn, J. Am.
Chem. Soc., 2012, 134, 8617–8624.
2
2
6 J. de Jong, R. G. H. Lammertink and M. Wessling, Lab Chip,
47 C. Liu, Y. Li, Y. Li, P. Chen, X. Feng, W. Du and B.-F. Liu,
Talanta, 2016, 149, 237–243.
2
006, 6, 1125–1139.
This journal is © The Royal Society of Chemistry 2016
Lab Chip