Enzyme kinetics in acoustically levitated droplets of supercooled water: A novel approach to cryoenzymology

Article abstract of DOI:10.1021/ac048486f

The rate of the alkaline phosphatase-catalyzed hydrolysis of 4-methylumbelliferone phosphate was measured in acoustically levitated droplets of aqueous tris (50 mM) at pH 8.5 at 22 ± 2 °C and in supercooled solution at -6 ± 2 °C. At 22 °C, the rate of product formation was in excellent agreement with the rate observed in bulk solution in a cuvette, indicating that the acoustic levitation process does not alter the enzyme activity. The rate of the reaction decreased 6-fold in supercooled solution at -6 ± 2 °C. The acoustic levitator apparatus is described in detail.

Full text of DOI:10.1021/ac048486f

Anal. Chem. 2005, 77, 2558-2563
Enzyme Kinetics in Acoustically Levitated Droplets
of Supercooled Water: A Novel Approach to
Cryoenzymology
David D. Weis*,† and Jonathan D. Nardozzi‡
Department of Chemistry, Skidmore College, 815 North Broadway, Saratoga Springs, New York 12866-1632
The rate of the alkaline phosphatase-catalyzed hydrolysis
of 4-methylumbelliferone phosphate was measured in
acoustically levitated droplets of aqueous tris (50 mM)
at pH 8.5 at 22 ( 2 °C and in supercooled solution at -6
impurities. Supercooled water has been prepared in narrow
capillaries,2 in water-in-oil emulsions, and in levitated droplets.
,3
2,3
5
Indeed, the use of supercooled water has already been
exploited to a limited extent in cryoenzymology. Douzu and co-
workers prepared aqueous solutions of various enzymes in
supercooled water by forming water-in-oil emulsions with sorbitan
tristerate (Span 65) surfactant. This technique was used to
(
2 °C. At 22 °C, the rate of product formation was in
excellent agreement with the rate observed in bulk solu-
tion in a cuvette, indicating that the acoustic levitation
process does not alter the enzyme activity. The rate of the
reaction decreased 6-fold in supercooled solution at -6
3
+
measure the low-spin to high-spin ratio in camphor-bound Fe
cytochrome P450, to stabilize the Fe2 ‚O
+
intermediate in cyto-
2
chrome P450, and to follow horseradish peroxidase-CO recom-
(
2 °C. The acoustic levitator apparatus is described in
6
bination following flash photolysis. More recently, Szyperski and
detail.
co-workers have applied NMR methods for structural determina-
tion of proteins in supercooled water down to -16.5 °C in narrow
capillaries.7 Both of these methods of preparing supercooled
water have limitations. In the preparation of oil-in-water emulsions,
there is the potential for either the oil or surfactant to disrupt the
enzyme structure, perhaps even denaturing it. In the case of the
capillary method, the large volume of solution limits the extent
of supercooling.2 Levitated droplets are free of these limitations.
In acoustic levitation, small liquid or solid particles are
suspended in the nodal points of a resonant acoustic field (i.e., a
standing wave) established between a high-intensity source and
a reflector.9 Recently, there has been considerable interest in
applying acoustic levitation to problems in areas of analytical
The goal in cryoenzymology is to use low temperatures to
,8
facilitate the investigation of enzymes by slowing the rates of
reactions or trapping intermediate states such that the systems
become experimentally accessible. Since the natural solvent for
enzymes is water (or dilute aqueous solutions), the normal
freezing point of water at 0 °C imposes a limitation on cryoenzy-
mology. Typically, this limitation is circumvented by adding a
carefully selected antifreeze that is believed to preserve the
chemical and physical properties of liquid water or at least does
not perturb the function of the enzyme.1
,3
-11
A more attractive choice of cryosolvent is supercooled liquid
water. In its supercooled state, water retains its liquid properties,
1
2-15
chemistry including several papers in this journal.
(See
2
-4
though its viscosity increases significantly.
Small volumes of
Santesson and Nilsson16 for a recent and thorough review of these
applications.) Much of the motivation for this research has been
water (<10 µL) can remain liquid down to about -42 °C (i.e., the
4
homogeneous nucleation limit) provided the water does not
(5) Kr a¨ mer, B.; Schwell, M.; H u¨ bner, O.; Vortisch, H.; Leisner, T.; R u¨ hl, E.;
contain any ice-forming impurities and is not in contact with an
ice-forming surface (i.e., heterogeneous nucleation). Since it is
the solvent that most closely resembles ordinary liquid water,
supercooled water is an ideal medium to use to slow reactions
involving proteins, as this solvent is the least disruptive to the
native structure of the protein. To avoid nucleation, the volume
of water must be small and the water must be free of ice-forming
Baumg a¨ rtel, H.; W o¨ ste, L. Ber. Bunsnges. Phys. Chem. 1996, 100, 1911-
1914.
(6) Douzou, P. In Advances in Enzymology; Meister, A., Ed.; John Wiley and
Sons: New York, 1980; Vol. 51, pp 1-74.
(7) Skalicky, J. J.; Sukumaran, D. K.; Mills, J. L.; Szyperski, T. J. Am. Chem.
Soc. 2000, 122, 3230- -3231.
(8) Skalicky, J. J.; Mills, J. L.; Sharma, S.; Szyperski, T. J. Am. Chem. Soc. 2001,
23, 388-397.
9) Brandt, E. H. Science 1989, 243, 349-355.
10) Lierke, E. G. Acustica 1996, 82, 220-237.
1
(
(
*
Corresponding author.
Present address: Department of Chemistry, MSC03 2060, 1 University of
(11) Yarin, A. L.; Pfaffenlehner, M.; Tropea, C. J. Fluid Mech. 1998, 356, 65-
91.
†
New Mexico, Albuquerque, NM 87131-0001. E-mail: dweis@unm.edu. Fax: (505)
(12) Santesson, S.; Andersson, M.; Degerman, E.; Johansson, T.; Nilsson, J.;
2
77-2902.
Nilsson, S. Anal. Chem. 2000, 72, 3412-3418.
(13) Santesson, S.; Johansson, J.; Taylor, L. S.; Levander, I.; Fox, S.; Sepaniak,
M.; Nilsson, S. Anal. Chem. 2003, 75, 2177-2180.
‡
Present address: Department of Biochemistry and Molecular Biology, SUNY
Upstate Medical University, Syracuse, NY 13210.
(
(
1) Travers, F.; Barman, T. Biochemie 1995, 77, 937-948.
2) Angell, C. A. In Water, a Comprehensive Treatise; Franks, F., Ed.; Plenum
Press: New York, 1982; Vol. 7, pp 1-81.
3) Angell, C. A. Annu. Rev. Phys. Chem. 1983, 34, 593-630.
4) Debenedetti, P. G. J. Phys.: Condens. Matter 2003, 15, R1669-R1726.
(14) Santesson, S.; Cedergren-Zeppezauer, E. S.; Johansson, T.; Laurell, T.;
Nilsson, J.; Nilsson, S. Anal. Chem. 2003, 75, 1733-1740.
(15) Santesson, S.; Ramirez, I. B.-R.; Viberg, P.; Jergil, B.; Nilsson, S. Anal. Chem.
2004, 76, 303-308.
(
(
(16) Santesson, S.; Nilsson, S. Anal. Bioanal. Chem. 2004, 378, 1704-1709.
2558 Analytical Chemistry, Vol. 77, No. 8, April 15, 2005
10.1021/ac048486f CCC: $30.25 © 2005 American Chemical Society
Published on Web 03/08/2005

that acoustic levitation provides experimental access to moderately
small volumes (∼0.1-10 µL) in a containerless (or noncontact)
environment. The containerless aspect of the system is driven by
the need to eliminate wall effects (adsorption and reaction) when
working with small volumes of dilute solutions that are routinely
1
6
encountered in bioanalytical applications. Here, the noncontact
environment of the levitator permits the supercooling of water.
As a model system, we have chosen to work with the enzyme
alkaline phosphatase (AP, EC 3.1.3.1) a nonspecific phospho-
1
7
monoesterase. In our reaction, AP cleaves the phosphate group
from 4-methylumbelliferone phosphate (MUP), converting it into
the fluorescent product coumarin 4.
Our objectives in this investigation were (1) to demonstrate that
kinetic studies are feasible in levitated droplets, (2) to demonstrate
that enzymatic activity is not disrupted in an acoustically levitated
droplet, and (3) to measure the rate of an enzyme-catalyzed
reaction in supercooled aqueous solution. We describe our
acoustic levitator and spectrofluorometer in detail, report results
demonstrating that the first two objectives have been achieved in
the AP-catalyzed MUP hydrolysis reaction, and present data on
the rate of the reaction in supercooled water at -6 °C.
Figure 1. (A) Acoustic levitator apparatus. (B) Photograph of a 10-
µL levitated droplet.
Figure 1. A resonant acoustic field (standing wave) was established
in the vertical direction (z axis) between a high-intensity sound
source and a reflector. Small (1-5 mm) droplets could be levitated
in the nodal points of the standing wave. A ∼29.5-kHz sinusoidal
signal was generated by digital function generator (Agilent model
3
3120A), amplified by a 75-W wide-band amplifier (Krohn-Hite
EXPERIMENTAL SECTION
model 7500), and impedance-matched to the ultrasonic transducer
with an impedance matching transformer (Krohn-Hite model MT-
Materials. Tris buffer (50 mM) was prepared from 2 M tris
stock (Fisher BP1759) and adjusted to pH 8.50 with ∼1 M
hydrochloric acid. Alkaline phosphatase from bovine intestinal
mucosa was obtained as a lyophilized powder (Sigma P6772).
5
6R). The ultrasonic transducer18 consisted of two lead zirconium
titanate (PZT) rings compressed in a sandwich configuration
between two aluminum cylinders. The acoustic signal was ampli-
fied through a stepped horn and emitted by a circular plate.
Electrical contact was established through a thin copper electrode
sandwiched between the two PZT rings. The two halves of the
transducer body were brought into electrical contact and grounded
via a bolt that passed through the center of the rings. A polymer
sleeve surrounding the central bolt prevented short circuiting
between the PZT rings and ground.
-
1
Aliquots of an 80.4 DEA units mL stock AP solution in 50 mM
tris/0.01% (m/v) bovine serum albumin (BSA, Fisher BP 671-1)
were kept at -20 °C until needed. To minimize variability, all
assays were prepared from aliquots from a single AP stock
solution. There was no evidence of loss of AP activity over a period
of two months. Aliquots of a stock 4-methylumbelliferone phos-
phate (MUP, Acros 41504) solution in anhydrous dimethyl
sulfoxide (DMSO) at a concentration of 0.190 M were kept at -20
Two microphones (Radio Shack model 33-3013) were used to
monitor the performance of the levitator. The amplitude of the
microphone signals was observed on an oscilloscope. The trans-
ducer microphone, positioned beneath the radiating plate, was
used to monitor the transducer output independent of the reflector
position. The resonant field microphone, positioned in the vicinity
of an antinode of the resonant field, was used to monitor the
quality of the resonant field. The transducer has a sharp resonance
°
C until needed. Standards were prepared in tris buffer from stock
solutions of coumarin 4 (Acros 40549) in methanol (HPLC grade,
Fisher) and aqueous resorufin sodium salt (Aldrich 230154). All
water used was high-purity, deionized grade (18 MΩ, 0.2 µm
filtered).
Acoustic Levitator. A diagram of the acoustic levitator
apparatus and a photograph of a levitated droplet are shown in
(
17) McComb, R. B.; G. N. Bowers, J.; Posen, S. Alkaline Phosphatase; Plenum:
New York, 1979.
(18) Trinh, E. H. Rev. Sci. Instrum. 1985, 56, 2059-2066.
Analytical Chemistry, Vol. 77, No. 8, April 15, 2005 2559

near 29.5 kHz; however, the resonant frequency varied slightly
as a function of temperature, so it was necessary to tune the
transducer to resonance just prior to levitation by adjusting the
frequency of the input signal until the transducer microphone
signal reached a maximum. A 1.9 cm diameter stainless steel
cylinder served as the reflector. The reflector was mounted on a
tilt/swivel mount (Edmund Industrial Optics K37-922) that was
used to ensure that the transducer and reflector faces were
parallel. The tilt/swivel mount was attached to a vertical (z axis)
linear stage (Newport model 423) driven by a linear actuator
(Zaber Technologies model T-LA28) controlled by in-house
software. Adjustment of the linear actuator changed the vertical
gap distance between the transducer and reflector. The linear
actuator was mounted onto a pair of orthogonal linear stages
(Newport model TSX-1D) aligned in the x and y directions. These
linear stages were used to center the reflector over the transducer.
Once the transducer was tuned to resonance, the reflector-
transducer vertical distance was adjusted until the signal from the
resonant field microphone signal reached a maximum. Initially,
the reflector position was adjusted using all five degrees of
freedom (i.e., x, y, z, and the tilt and swivel angles) until stable
levitation was obtained. Once the reflector was aligned, only the
vertical distance required adjustment.
The entire levitator assembly was mounted to a custom-built
aluminum scaffold. The scaffold and optics (see below) were
mounted onto an aluminum breadboard and enclosed in a light-
tight insulated housing. The temperature inside the housing could
Figure 2. Schematic diagram of the optical components.
source was a deuterium tungsten/halogen dual light source
(Ocean Optics DH2000). The excitation light passed through a
variable long-pass filter (Ocean Optics model LVF-H) adjusted to
minimize MUP fluorescence (50% cutoff at 321 nm). A pair of fiber-
optic collimating lenses (Ocean Optics model 74-UV) served as
the excitation and emission optics. The lenses were positioned 3
cm from the droplet and mounted on xyz stages (Newport MT-
XYZ) so that they could be aligned with the droplet. Fluorescence
emission spectra of the droplet were collected using a fiber-optic
spectrometer with a charge-coupled device array detector (Ocean
Optics USB2000). All optical connections were made with 600-
µm-diameter optical fibers (Ocean Optics model P-600-2-UV-VIS);
a solarization-resistant fiber was used between the light source
and the long-pass filter (Ocean Optics model P600-2SR). Cuvette
enzyme assays were performed in a 0.3 cm × 1 cm × 4.5 cm
quartz cuvette using the same optical setup except that a cuvette
holder (Ocean Optics model CUV-FL-DA) took the place of the
levitated droplet. In kinetic assays, an emission spectrum was
collected every 8.2 s. A software-controlled shutter in the light
source was kept closed between measurements to prevent
photodegradation of MUP and coumarin 4.
Typical fluorescence emission spectra of MUP and coumarin
4 are shown in Figure 3. The emission wavelengths used to
monitor MUP, coumarin 4, and resorufin (an internal standard,
see Results and Discussion) were 388, 475, and 581 nm, respec-
tively. Although only weakly fluorescent, the MUP substrate was
present at a high concentration in the reaction mixture. The
shoulder of the MUP fluorescence emission overlapped with the
maxima of the coumarin 4 product emission (447 and 475 nm).
Since MUP was consumed over the course of the reaction, its
fluorescence signal at the coumarin 4 emission wavelength (475
nm) was not constant; thus, it was necessary to subtract it. A two-
wavelength mixture analysis method using the emission maxima
of both MUP and coumarin 4 was used to subtract the MUP
contribution to the fluorescence signal at the coumarin 4 emission
wavelength. This is described in the Supporting Information.
be varied between -18 and 25 °C by circulating a refrigerated
solution of 50% ethylene glycol through three 6-m coils of 1/4-in.-
diameter copper tubing placed in three of the corners of the
housing. A circulating chiller (Neslab RTE-140) was used to
refrigerate the circulating fluid. To slow droplet evaporation in
2
2 °C experiments, the humidity inside the housing was main-
tained at 60-70% RH by flowing humidified nitrogen through the
housing. The nitrogen was humidified by bubbling the gas
through water. Four small fans inside the housing provided mixing
to eliminate temperature and humidity gradients. The air flows
were not directly incident upon the levitated droplet. The humidity
and air temperature inside the chamber were measured with a
digital hygrometer (Fisher model 11-661-7B).
Aqueous droplets (10-15 µL, 2.7-3.1-mm diameter) were
injected into an acoustic node using a micropipet (Gilson model
P20) with a gel loading tip (Rainin GT-060). To reduce adhesion
between the droplet and the tip, the tip was rendered hydrophobic
by briefly immersing it in a 0.5% (m/v) solution of paraffin in
1
9
petroleum ether and then allowing it to dry. The pipet tip was
positioned just outside of an acoustic node so that as the liquid
emerged from the tip it formed a droplet in the node. The droplet
was dislodged from the pipet tip by increasing the amplitude of
the input signal. Once the droplet was dislodged, the signal
amplitude was reduced. A door on the side of the housing and an
observation window on the front permitted the injection of droplets
into the levitator. To maintain temperature and humidity condi-
tions inside the chamber during droplet injection, the pipet was
inserted into the chamber through a pair of polyethylene curtains.
Fluorescence Spectroscopy. The configuration of the optical
components is shown in Figure 2. The fluorescence excitation
(
19) Welter, E.; Neidhart, B. Fresenius J. Anal. Chem. 1997, 357, 345-350.
560 Analytical Chemistry, Vol. 77, No. 8, April 15, 2005
2

Figure 3. Fluorescence emission spectra of levitated 15-µL droplets
of 8 µM coumarin 4 and 200 µM MUP standards. Both droplets also
contained resorufin internal standard at a concentration of 200 nM.
The vertical lines show the emission wavelengths used for quantitative
measurements.
Enzyme Assay. The rate of the AP-catalyzed hydrolysis
reaction was measured at pH 8.50 in 50 mM tris buffer. The
-
1
concentration of AP was 0.017 DEA units mL (∼33 pM for the
homodimer). The MUP substrate concentration was 200 µM
(
∼20K
m
, unpublished results). A small amount of dimethyl
-
1
sulfoxide (0.1 vol %) and BSA (21 µg L ) was incorporated into
the reaction mixture because they were present in the stock
solutions of MUP and AP, respectively. Resorufin was added as
an internal standard (see Results and Discussion) at a concentra-
tion of 200 nM. The reaction was started by substrate initiation.
In levitation experiments, substrate initiation was followed im-
mediately by injection of a droplet of the reaction mixture into
the levitator. There was no evidence of MUP hydrolysis in the
absence of AP, and no emission above background at 475 nm
Figure 4. Fluorescence emission intensity versus time for a 15-µL
levitated droplet containing 8 µM coumarin 4 and 200 nM resorufin.
Each trace was normalized so that the initial value was 100%. (A)
Coumarin 4 emission at 475 nm and resorufin emission at 581 nm.
(B) Internally standardized by taking the coumarin 4-to-resorufin
emission ratio from (A).
What is of particular importance is that, despite the instabilities,
the coumarin 4 and resorufin emission track each other well. Thus,
resorufin, or another fluorophore that is easily resolved from the
analyte, can serve as an internal standard. By holding the
concentration of resorufin constant, the ratio of coumarin 4-to-
resorufin emission can serve as an analytical signal that it is
independent of both droplet size and position. Figure 4B, which
shows the ratio of the two signals shown in Figure 4A, demon-
strates the improved stability of the analytical signal over a 5-min
period. A typical internally standardized linear calibration at 22
(coumarin 4 wavelength) was observed in the absence of MUP.
Safety Considerations. The acoustic levitator described here
has an exposed high-voltage electrode. Touching the electrode
could result in serious injury. It is recommended that all exposed
metal parts be grounded to reduce the chance of electric shock.
The authors have not observed any ill effects from exposure to
high-intensity ultrasound.
RESULTS AND DISCUSSION
2
2
°
C gave R ) 0.998 and calibration at -6 °C gave R ) 0.94.
Kinetics of AP at Room Temperature. Under the conditions
Calibration. Although droplets could be held in the levitator
routinely for 10 min, instabilities in the levitator resulted in both
short-term fluctuations in the fluorescence signal (i.e., wiggle) and
long-term drift. These effects are evident in the plots of normalized
coumarin 4 and resorufin emission versus time shown in Figure
described here, the AP-catalyzed reaction is consistent with the
Michaelis-Menten kinetic mechanism20 with a K
≈ 10 µM
unpublished results). At a substate concentration of 200 µM, the
M
(
reaction takes place under saturated conditions at a maximal rate
determined solely by the AP concentration. Over the 300-s time
period of these experiments, there is no significant decrease in
substrate concentration so the rate of the reaction will be constant.
4
A. The instabilities cause changes in the position of the droplet.
Since the excitation and emission optical paths are fixed in place,
these instabilities result in instability in the observed emission
intensity. The droplet wiggle is most likely caused by the action
of the fans and by the flow of humidified gas in the enclosure.
Droplet evaporation is the source of the observed long-term drift.
(20) Connors, K. A. Chemical Kinetics: The Study of Reaction Rates in Solution;
VCH Publishers: New York, 1990.
Analytical Chemistry, Vol. 77, No. 8, April 15, 2005 2561

Refining the Prototype Instrument. There are some limita-
tions to this preliminary study on AP using our prototype
instrument that need to be corrected. The most significant is the
limited degree of supercooling. While we have attempted to
measure AP activity in supercooled water at lower temperatures,
the stability of the droplet in the levitator became poor. This may
have been caused by temperature instability within the transducer
itself since the resonant frequency is temperature-dependent. This
process has a complex feedback since the transducer will draw
Table 1. Rate of AP-catalyzed MUP Hydrolysis
temperature
°C)
rate ( s
no. of
meas
-
1
2
(
volume
(nM s
)
mean R ( s
2
2 ( 2
2 ( 2
6 ( 2
∼1 mL
15 µL
12 µL
11.8 ( 0.4
12.0 ( 0.7
2.1 ( 0.2
0.99946 ( 0.00005
0.959 ( 0.075
0.617 ( 0.049
5
5
3
2
-
Table 1 summarizes data from a number of measurements of
the reaction rate at 22 ( 2 °C in a 15-µL levitated droplet, in ∼1
mL of solution in a quartz cuvette, and in a 12-µL levitated droplet
at -6 ( 2 °C. The rate of the reaction was determined by linear
1
8
less current off-resonance causing it to cool slightly. Trinh
described an acoustic levitator with automatic feedback control
of the reflector position, transducer frequency, and signal ampli-
tude that functioned over the -40 to 150 °C temperature range.
Incorporation of such feedback control will greatly improve the
stability of the levitated droplet and enable levitation at lower
temperatures. A second refinement that is necessary is a direct
determination of the droplet temperature. This has already been
regression; mean values of the square of the correlation coefficient
2
(
R ) and its standard deviation (s) are also listed in Table 1. The
rate of coumarin 4 formation in the levitated droplet (12.0 ( 0.7
-
1
nM s ) was in excellent agreement with the rate observed the
-1
cuvette (11.8 ( 0.4 nM s ). Under the saturating conditions used
3+
21
reported in levitated droplets using Eu (EDTA) fluorescence
here, the rate of the reaction is diffusion-limited. Streaming flow
emission25 and La S‚Eu phosphorescence lifetime26 measure-
2
2
2 2
O
around the levitated droplet can cause circulation or agitation
of the fluid within the droplet. This circulation might result in an
elevated rate for this diffusion-limited reaction if it significantly
increased the rate of mass transfer within the droplet. However,
the consistency of the reaction rate in the cuvette and the levitated
droplet at 22 ( 2 °C indicates that the circulation is negligible
and that the levitation process does not perturb the kinetics of
the reaction.
ments. Another necessary improvement is that the pH of the
supercooled droplet must be determined accurately. While the
temperature-dependence of the K values of many buffer systems
a
are known in the temperature range of liquid water, it is not clear
if these data could be extrapolated for use with deeply supercooled
water. Spectrometric determination of pH using a fluorescent
a
indicator presents a conundrum since both the K of the indicator
and the weak acid form of the buffer will be unknown in the
supercooled solution as both are temperature-dependent. How-
ever, Douzou27 described a stepwise method of establishing a pH
scale in supercooled liquids starting with 0.01 M HCl. Since 0.01
M HCl is fully dissociated, the proton activity can be estimated
using extended Debye-H u¨ ckel theory. This HCl solution then
serves as an accurate standard on which to establish a supercooled
pH scale. An alternating series of indicators and buffers with
Kinetics of AP in Supercooled Solution. Measurements on
the rate of the reaction in supercooled water have also been made.
At -6 ( 2 °C, the rate of coumarin 4 formation was found to be
-
1
2
.1 ( 0.2 nM s (Table 1). At -6 °C, the levitated droplets were
not as stable in the levitator as droplets at 22 °C. This is evident
2
in the lower R value, which indicates more scatter in the data.
There is a 6-fold slowdown in the rate of the reaction going from
2
a
2 to -6 °C. At -6 °C, however, the pH is not 8.5, since the K of
a
increasingly basic pK values is then used to calibrate the pH scale.
tris is temperature-dependent. While we cannot directly determine
the temperature of the levitated droplet, three lines of evidence
indicate that the droplet is indeed supercooled. First, some
levitated droplets were observed to freeze rapidly (<1 min) in
the levitator, presumably caused by heterogeneous nucleation.
This indicates that cooling of the droplet in the levitator is rapid.
Second, when freezing occurred, the droplet became opaque and
the fluorescence emission spectrum was overwhelmed by scat-
tering of the excitation light. Visual inspection of the droplets used
to measure the rate at -6 ( 2 °C did not indicate freezing, nor
was light scattering observed in the emission spectrum. Finally,
there was no evidence of a slowdown in the rate of coumarin 4
formation over the course of the reaction as would be expected if
the droplet cooled slowly. Bovine intestinal AP has already been
shown to be active down to much lower temperatures, but only
in the presence of significant amounts of antifreeze: Champion
Finally, droplet evaporation limits the time scale of the experi-
ments and causes the concentration of the reactants to increase
over time. This can be overcome by raising the humidity inside
the chamber or by adding microdroplets of water to the levitated
droplet over the course of the experiment.12
CONCLUSIONS
For a number of years, the use of supercooled water and
aqueous solutions has been of interest in the field of enzymology.
Acoustic levitation permits the preparation of supercooled aqueous
solutions without the need to resort to antifreeze, water-in-oil
emulsions, or narrow capillaries. We have described here an
acoustic levitator apparatus and reported its first use in a kinetic
study on an enzyme. The rate of the AP-catalyzed hydrolysis of
MUP under saturating conditions (i.e., high substrate concentra-
tion) at 22 ( 2 °C in a levitated droplet was found to be same as
in bulk solution, indicating that the levitation process does not
disrupt the enzyme kinetics. A 6-fold decrease in the rate was
23
et al. observed AP activity in concentrated sucrose solutions (10-
8 wt %) down -24 °C using the substrate p-nitrophenyl phos-
5
2
4
phate, and Bragger et al. observed AP activity down to -100 °C
in 60% DMSO/20% ethylene glycol/20% water with MUP as the
substrate.
(24) Bragger, J. M.; Dunn, R. V.; Daniel, R. M. Biochim. Biophys. Acta 2000,
1
480, 278-282.
(25) Seaver, M.; Peele, J. R. Appl. Opt. 1990, 29, 4956-4961.
(
(
(
21) Simopoulos, T. T.; Jencks, W. P. Biochemistry 1994, 33, 10375-10380.
22) Trinh, E. H.; Robey, J. L. Phys. Fluids 1994, 6, 3567-3579.
23) Champion, D.; Blond, G.; Meste, M. L.; Simatos, D. J. Agric. Food Chem.
(26) Omrane, A.; Santesson, S.; Ald e´ n, M.; Nilsson, S. Lab Chip 2004, 4, 287-
291.
(27) Douzou, P. Cryobiochemistry: An Introduction; Academic Press: New York,
2000, 48, 4942-4947.
1977.
2562 Analytical Chemistry, Vol. 77, No. 8, April 15, 2005

observed when the reaction was carried out in supercooled
aqueous solution at -6 °C.
and also assisted in their design. Financial support for this
research was provided by Skidmore College and by Ocean Optics
Inc.
ACKNOWLEDGMENT
SUPPORTING INFORMATION AVAILABLE
We thank Prof. R. Glynn Holt of the Department of Aerospace
and Mechanical Engineering at Boston University for the loan of
an ultrasonic transducer and for technical assistance. Sanna Malick
characterized the Michaelis-Menten kinetics of the AP-catalyzed
hydrolysis of MUP in bulk solution at room temperature. John
Myers of the Physics Machine Shop at Skidmore College
constructed the custom components used in the levitator apparatus
Derivation of a two-wavelength mixture analysis method for
overlapping emission spectra. This material is available free of
charge via the Internet at http://pubs.acs.org.
Received for review October 13, 2004. Accepted January
31, 2005.
AC048486F
Analytical Chemistry, Vol. 77, No. 8, April 15, 2005 2563