Reagents and assay strategies for quantifying active enzyme analytes using a personal glucose meter

Article abstract of DOI:10.1039/c3cc43702g

This communication describes small molecule reagents and a rapid single-step assay for quantifying nanomolar levels of active enzyme analytes using a personal glucose meter. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3cc43702g

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Reagents and assay strategies for quantifying active
enzyme analytes using a personal glucose meter†
Cite this: Chem. Commun., 2013,
49, 6134
Hemakesh Mohapatra and Scott T. Phillips*
Received 16th May 2013,
Accepted 28th May 2013
DOI: 10.1039/c3cc43702g
www.rsc.org/chemcomm
This communication describes small molecule reagents and a rapid
single-step assay for quantifying nanomolar levels of active
enzyme analytes using a personal glucose meter.
A wide range of qualitative colorimetric activity-based assays
have been developed over the years for detecting enzymes and
other analytes via specific reactions between a substrate and
the analyte.1–3 Quantifying the colorimetric output, however,
has proven challenging, particularly in point-of-care (POC)
settings where few resources are available for conducting the
assay. Specialized hand-held readers that enable quantification
based on reflectance,4 transmittance,5 and absorbance6 have
been developed for this purpose, but they are produced at low
volume, are expensive for many POC applications, and suffer
from interference from background color and particulates that
may be in a sample. Quantification using a camera-equipped
cellular phone offers a lower cost alternative to specialized
Fig. 1 A one-step assay strategy for quantifying enzyme analytes via their
enzymatic activity in combination with a personal glucose meter.
readers, but with added concerns associated with lighting wait a fixed period of time, and then measure the quantity of
conditions, focus, and differences between cameras that may glucose in the sample.19 Thus, this strategy offers a combi-
affect the quantitative interpretation of an assay.7–9 Con- nation of ease of operation and quantification that may be
sequently, the need remains for a simple and inexpensive particularly useful in environments that lack refrigeration,
approach for quantifying activity-based assays without relying electricity, and other resources, such as remote villages in the
on measurements of color, while still building on years of developing world.20–23
development in activity-based enzyme assays.3 Herein, we
describe such a strategy using a personal glucose meter.10–18
The selectivity in the assay originates from the functionality
on the small molecule reagent, which can be tuned easily for
To enable the use of a personal glucose meter for activity- specific activity-based detection events.3,24 Fig. 2 shows four
based assays, we developed reagents that release glucose when general controlled release reagents that we prepared to illus-
a target enzyme analyte is detected. Our single-step assay is trate four methods for releasing glucose in response to a target
illustrated in Fig. 1. When the target enzyme is present in a enzyme. These four reagents are not optimized for a specific
sample, it reacts with one of the reagents to release glucose, target enzyme, but rather provide platforms on which four
which is then quantified using a personal glucose meter. general classes of enzymes can be detected in this type of
This assay requires only that the user deposit a test fluid into glucose meter-mediated assay. These classes include glyco-
a pre-loaded tube containing a small molecule assay reagent, sidases (reagent 1a), esterases (1b), phosphatases (1c), and
proteases (1d). As demonstrated by the structures in Fig. 2,
different classes of enzyme targets require different strategies
for releasing glucose.
The Pennsylvania State University, 104 Chemistry Bldg, University Park, PA 16802,
USA. E-mail: sphillips@psu.edu; Fax: +1 814 865 5235; Tel: +1 814 867 2502
To target a specific enzyme within one of these classes, the
† Electronic supplementary information (ESI) available: Synthetic procedures,
substrate (blue portion in Fig. 2) that is covalently linked to glucose
or the controlled release reagent (depending on the release strategy)
compound characterization data, experimental procedures, supporting figures,
and tables of data. See DOI: 10.1039/c3cc43702g
c
6134 Chem. Commun., 2013, 49, 6134--6136
This journal is The Royal Society of Chemistry 2013

View Article Online
Communication
ChemComm
Exposure of 1d to various quantities of penicillin-G-amidase
provided a reproducible dose-dependent response (Fig. S2,
ESI†) that provides a useful calibration curve for quantifying
penicillin-G-amidase in a sample. Similar calibration curves
were obtained for b-D-galactosidase (Fig. S4, ESI†), esterase
(Fig. S5, ESI†), and alkaline phosphatase (Fig. S6, ESI†) using
their respective small molecule assay reagents (Fig. 2). The
sensitivity provided by these initial assays is comparable to
colorimetric assays,26 but with the benefit that the color of the
sample does not interfere with the readout provided by the
personal glucose meter (as it would for a colorimetric assay).
A case in point (which is highlighted in Fig. S6, ESI†) is alkaline
phosphatase.26 At normal levels, alkaline phosphatase is present
in blood between 30–120 U LÀ1, whereas the concentration is
greater than 120 U LÀ1 in patients with impaired liver function.1
The sensitivity of the assay shown in Fig. S6 (ESI†) is sufficient
for differentiating individuals with normal and elevated levels of
alkaline phosphatase. Likewise, reagent 1c (which is used for
detecting alkaline phosphatase) is similar in structure to the
standard colorimetric reagent (p-nitrophenol phosphate27) that
Fig. 2 Structures of general activity-based controlled release reagents designed
for assays that detect four classes of enzymes by releasing glucose. The classes
are: glycosidases (1a), esterases (1b), phosphatases (1c), and proteases (1d). The
green color-coding denotes glucose that will be released upon reaction of
the reagent with the target enzyme, and blue represents the substrate for the
enzyme. Selective assays should be possible by modifying the substrate (the blue
portion of the general reagent) for the target enzyme.3
could be changed to match the substrate preference for the is used for clinical analyses, and thus can be expected to display
desired enzyme analyte.3 For example, in the case of 1d (which a similar level of selectivity for alkaline phosphatase.
is used for detecting proteases; penicillin-G-amidase is the
Given these promising results, we reasoned that shorter
model protease in this study), the substrate for penicillin- assay times would further improve the convenience of the
G-amidase is connected to glucose through a self-immolative assays for point-of-care applications. Lu,10–13 Xiang,15–17 and
linker, which is necessary to ensure that cleavage of the Yang18 employed a signal amplification step in their antibody,
C-terminus of the amide in 1d can be translated into release aptamer, and nucleic acid-based assays to increase signal and
of glucose without competing with non-specific background ultimately reduce assay times when using a personal glucose
release in the time frame of the assay. Once penicillin- meter, but here we simply mix 62 mg of glucose (for a 200 mL
G-amidase cleaves the C-terminus of the peptide bond, the assay) with the desired detection reagent prior to adding the
resulting aniline proceeds through azaquinone methide to sample that contains the desired analyte. In other words, we
release carbon dioxide and glucose. In the absence of the use pre-loaded assay tubes that contain enough glucose to
enzyme analyte, glucose oxidase (in the commercial strips that provide a low-level reading on the glucose meter. The presence
accompany personal glucose meters) is not able to process the of the target enzyme then simply elevates this reading within
glucose bound to the small molecule detection reagent, thus the detection window of the glucose meter. The consequence of
avoiding a possible source of background signal.
this standard addition approach is 2–12Â shorter assays times
Detection reagent 1d was accessed through a convergent (Fig. S7–S9, ESI†) compared with our first-generation assays
route that involved preparing the peptide substrate and the (Fig. S2, S4–S6, ESI†), since time is not wasted waiting for the
glucose components separately, and then bringing them enzymatic detection event to accumulate levels of glucose
together through a carbonate linkage (Scheme S1, ESI†). This (i.e., millimolar levels of glucose) to enable a reading using
convergent approach offers the opportunity to modify the the glucose meter.
peptide substrate as needed to target a desired protease.
Fig. S7–S9 (ESI†) show three calibration curves generated
Treatment of 1d with penicillin-G-amidase (0.5 U mLÀ1) in using this second generation approach. The calibration
0.1 M phosphate buffer (pH 7.5, 0.5% (v/v) Tween 20, 20 1C) led curves are for penicillin-G-amidase (using 1d) (Fig. S7, ESI†),
to nearly instantaneous consumption of 1d (see the stacked b-D-galactosidase (using 1a) (Fig. S8, ESI†), and alkaline phos-
LCMS traces in Fig. S1, ESI†) and formation of glucose, which phatase (using 1c) (Fig. S9, ESI†). For a 30 min assay, the limit-
was measured using a personal glucose meter. In the absence of-detection for penicillin-G-amidase is 0.01 U mLÀ1, and for a
of penicillin-G-amidase, 1d decomposed only 2% after exposure 15 min assay, the limit-of-detection for b-D-galactosidase is
to the assay solution for 1 h, and 7.5% after 5 h of exposure25 4 U mLÀ1 (38 nM). This latter level of sensitivity compares favorably
(Fig. S1, ESI†), the latter duration being 5Â longer than is with previous personal glucose meter-based assays that employ
needed for conducting the first-generation assay (Fig. S2, ESI†). signal amplification reactions as part of the assay.10,11 In our
Since reagent 1d would be stored and supplied dry before enzymatic assays, however, a secondary signal amplification
conducting an assay, we tested its stability in this context as step is unnecessary, and the assay times remain short (minutes).
well. Heating the dry reagent at 40 1C open to the air caused no The dynamic range for these standard addition assays also is
detectable decomposition after 24 h, and only 6% decomposi- noteworthy. In the case of a 5 min alkaline phosphatase assay
tion after heating for 7 days (Fig. S3, ESI†), thus illustrating its (Fig. S9, ESI†), the dynamic range is 1.5 U LÀ1 to 160 U LÀ1
stability for use in certain types of point-of-care environments. (1.3 nM to 140 nM), with a limit of detection of 1.5 U LÀ1 (1.3 nM).
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 6134--6136 6135

View Article Online
ChemComm
Communication
This range is within the clinically-relevant window for the modifying the general controlled release reagents so that they
enzyme analyte.1
are selective for a desired enzyme analyte.
While this approach of providing a base-line quantity of
This work was supported by NSF (CHE-1150969), the Penn
glucose to the assay improves the sensitivity and dynamic range State MRSEC (DMR-0820404), the Arnold and Mabel Beckman
of the assay, it also could increase the prevalence of undesired Foundation, the Camille and Henry Dreyfus Foundation, and
background signal arising from either decomposition of an Louis Martarano. S.T.P acknowledges support from the Alfred
assay reagent (e.g., 1d) or from glucose that may be present in a P. Sloan Research Fellows program.
sample. As mentioned previously, 1d decomposes B6% when
Notes and references
heated at 40 1C open to the air for 7 days. In the context of an
assay for 0.06 U mLÀ1 penicillin-G-amidase, however, reagent
1 C. A. Burtis, E. R. Ashwood and N. W. Tietz, Tietz textbook of
1d that was heated at 40 1C for 7 days provided nearly
equal assay results (using the personal glucose meter and the
calibration curve in Fig. S7, ESI†) as a comparison assay using
freshly-prepared reagent 1d: e.g., 0.05 Æ 0.00 U mLÀ1 penicillin-
G-amidase for the sample heated to 40 1C for 7 days vs. 0.07 Æ
0.01 U mLÀ1 for freshly-prepared 1d. Assuming that the 6% of
decomposed 1d releases glucose, we would expect a meter
reading that is B5% higher than the reading when 1d is not
decomposed. The accuracy of the personal glucose meter,
however, is within 20% of standard laboratory assays, therefore
this small level of decomposition of 1d is not detectable within
the sensitivity limits of the meter.
clinical chemistry, W.B. Saunders, Philadelphia, 1999.
2 D. G. Cho and J. L. Sessler, Chem. Soc. Rev., 2009, 38, 1647–1662.
3 J.-P. Goddard and J.-L. Reymond, Trends Biotechnol., 2004, 22, 363–370.
4 For an example of a commercial reflectance reader see: ESEQuant Lateral
Flow System, QIAGEN N. V. Hilden, Germany http://www.qiagen.com/
About-Us/Contact/Corporate-Contacts/OEM-Services/ESE-Instruments/
ESEQuant-Lateral-Flow-System/ (accessed March, 2013).
5 A. K. Ellerbee, S. T. Phillips, A. C. Siegel, K. A. Mirica, A. W. Martinez,
P. Striehl, N. Jain, M. Prentiss and G. M. Whitesides, Anal. Chem.,
2009, 81, 8447–8452.
6 For an example of a commercial UV/Vis spectrophotometer see:
PhotoLab UV/Vis portable spectrophotometers, Xylem Inc.: White
Plains, NY USA http://www.globalw.com/products/photolab.html
(accessed March, 2013).
7 A. W. Martinez, S. T. Phillips, E. Carrilho, S. W. Thomas, H. Sindi
and G. M. Whitesides, Anal. Chem., 2008, 80, 3699–3707.
8 L. Shen, J. A. Hagen and I. Papautsky, Lab Chip, 2012, 12, 4240–4243.
9 T. Schwaebel, O. Trapp and U. H. F. Bunz, Chem. Sci., 2013, 4,
273–281.
We next tested whether the assays could be conducted in a
complicated fluid such as serum. Fig. S10 (ESI†) reveals that
a 5 min assay for alkaline phosphatase (using 1c) enables a
10 Y. Xiang and Y. Lu, Nat. Chem., 2011, 3, 697–703.
dynamic range of 2 U LÀ1 to 160 U LÀ1 (2 nM to 140 nM), with a 11 Y. Xiang and Y. Lu, Anal. Chem., 2012, 84, 4174–4178.
limit of detection of 2 U LÀ1 (2 nM), which is a range and level
12 Y. Xiang and Y. Lu, Anal. Chem., 2012, 84, 1975–1980.
13 Y. Xiang and Y. Lu, Chem. Commun., 2013, 49, 585–587.
14 Z. Nie, F. Deiss, X. Liu, O. Akbulut and G. M. Whitesides, Lab Chip,
of sensitivity that is nearly identical to the calibration curve
generated in buffer (Fig. S9, ESI†). We also purposely added
glucose to the serum to simulate samples that contain different
initial quantities of glucose. In each of these tests, the final
assay solution contained 20 U LÀ1 alkaline phosphatase and
either 0 mM, 6.6 mM, or 3.3 mM of added glucose (in addition
to the glucose that was present already in serum). By conduct-
ing two assays on each sample (one with and one without
reagent 1c), we were able to account for glucose in a sample and
still enable quantitative assays using the calibration curve in
Fig. S10 (see Table S10, ESI†).
In conclusion, this communication describes a convenient
assay for measuring trace levels of enzyme analytes. The assays
are reproducible and easily implemented, rapid, operate in
complex fluids such as serum, and use small molecule
reagents, thus making them amenable for use in point-of-care
2010, 10, 3163–3169.
15 J. Xu, B. Jiang, J. Xie, Y. Xiang, R. Yuan and Y. Chai, Chem. Commun.,
2012, 48, 10733–10735.
16 J. Su, J. Xu, Y. Chen, Y. Xiang, R. Yuan and Y. Chai, Chem. Commun.,
2012, 48, 6909–6911.
17 J. Su, J. Xu, Y. Chen, Y. Xiang, R. Yuan and Y. Chai, Biosens.
Bioelectron., 2013, 45, 219–222.
18 L. Yan, Z. Zhu, Y. Zou, Y. Huang, D. Liu, S. Jia, D. Xu, M. Wu,
Y. Zhou, S. Zhou and C. J. Yang, J. Am. Chem. Soc., 2013, 135,
3748–3751.
19 Background glucose levels (if expected in
a sample) can be
accounted for by testing the sample for glucose prior to conducting
an assay (ref. 10).
20 P. Yager, T. Edwards, E. Fu, K. Helton, K. Nelson, M. R. Tam and
B. H. Weigl, Nature, 2006, 442, 412–418.
21 H. Kettler, K. White and S. Hawkes, Mapping the landscape of
diagnostics for sexually transmitted infections, World Health Organi-
zation, Geneva, SW, 2004.
22 D. Mabey, R. W. Peeling, A. Ustianowski and M. D. Perkins, Nat. Rev.
Microbiol., 2004, 2, 231–240.
settings that lack refrigeration. The assays also enable measure- 23 A. W. Martinez, S. T. Phillips, G. M. Whitesides and E. Carrilho,
Anal. Chem., 2010, 82, 3–10.
24 A case in point is the substrate used in Zstatflut tests for influenza
ments of active enzymes rather than simply the total level of a
target enzyme antigen, regardless of whether it is active or not.
neuraminidase (K. E. Achyuthan, L. M. Pence, J. R. Appleman and
This type of differentiation should prove useful in the context of
liver function tests, quantifying enzymes associated with the
presence of an infectious disease, and for a variety of other
applications where a rapid, quantitative point-of-care test will
be valuable. With the advent of cell-phone-based personal
glucose meters,28 this new assay strategy may offer a powerful
alternative to colorimetric activity-based assays that are quan-
tified using cell phone cameras and telemedicine. Future
efforts will build on these proof-of-concept studies to develop
thorough and specific analytical assays for target analytes by
C. D. Shimasaki, Luminescence, 2003, 18, 131–139). This activity-
based substrate is selective for influenza-derived neuraminidases
over neuraminidases from other sources, and thus enables assays
that provide equal or better levels of selectivity for detecting the
presence of the influenza virus than complimentary antibody–
antigen assays.
25 Five hours was the duration of the stability assay.
26 S. J. Vella, P. Beattie, R. Cademartiri, A. Laromaine, A. W. Martinez,
S. T. Phillips, K. A. Mirica and G. M. Whitesides, Anal. Chem., 2012,
84, 2883–2891.
27 G. N. Bowers and R. B. McComb, Clin. Chem., 1966, 12, 70–89.
28 iBGStar blood glucosemeter, Sanofi-Aventis U.S. LLC: Bridgewater,
NJ USA http://www.ibgstar.us/ (accessed March, 2013).
c
6136 Chem. Commun., 2013, 49, 6134--6136
This journal is The Royal Society of Chemistry 2013