Turning tryptophanase into odor-generating biosensors

Article abstract of DOI:10.1002/anie.201309684

An odor-based sensor system that exploits the metabolic enzyme tryptophanase (TPase) as the key component is reported. This enzyme is able to convert an odorless substrate like S-methyl-L-cysteine or L-tryptophan into the odorous products methyl mercaptan or indole. To make a biosensor, TPase was biotinylated so that it could be coupled with a molecular recognition element, such as an antibody, to develop an ELISA-like assay. This method was used for the detection of an antibody present in nM concentrations by the human nose. TPase can also be combined with the enzyme pyridoxal kinase (PKase) for use in a coupled assay to detect adenosine 5-triphosphate (ATP). When ATP is present in the low μM concentration range, the coupled enzymatic system generates an odor that is easily detectable by the human nose. Biotinylated TPase can be combined with various biotin-labeled molecular recognition elements, thereby enabling a broad range of applications for this odor-based reporting system. The nose knows: Tryptophanase (TPase), which converts S-methyl-L-cysteine into methyl mercaptan (smelly), was coupled to a molecular recognition element (such as an antibody) to create an odor-based biosensor. Biotinylated TPase could be combined with various biotin-labeled molecular recognition elements, thereby enabling a broad range of applications for this odor-based reporting system.

Full text of DOI:10.1002/anie.201309684

.
Angewandte
Communications
DOI: 10.1002/anie.201309684
Biosensors
Turning Tryptophanase into Odor-Generating Biosensors**
Yaqin Xu, Zhuyuan Zhang, M. Monsur Ali, Joanna Sauder, Xudong Deng, Karen Giang,
Sergio D. Aguirre, Robert Pelton, Yingfu Li,* and Carlos D. M. Filipe*
Abstract: An odor-based sensor system that exploits the
metabolic enzyme tryptophanase (TPase) as the key compo-
nent is reported. This enzyme is able to convert an odorless
substrate like S-methyl-l-cysteine or l-tryptophan into the
odorous products methyl mercaptan or indole. To make
a biosensor, TPase was biotinylated so that it could be coupled
with a molecular recognition element, such as an antibody, to
develop an ELISA-like assay. This method was used for the
detection of an antibody present in nm concentrations by the
human nose. TPase can also be combined with the enzyme
pyridoxal kinase (PKase) for use in a coupled assay to detect
adenosine 5’-triphosphate (ATP). When ATP is present in the
low mm concentration range, the coupled enzymatic system
generates an odor that is easily detectable by the human nose.
Biotinylated TPase can be combined with various biotin-
labeled molecular recognition elements, thereby enabling
a broad range of applications for this odor-based reporting
system.
with the second small molecule to generate volatile ethane-
thiol. This approach is particularly useful for detecting
enzymes through the use of specialized substrates. However,
it is not clear how it could be adapted to achieve the detection
of other types of analytes, which is the focus of the current
study.
We have been interested in exploring tryptophanase as
the key component in the construction of odor-based
biosensors (Figure 1a).[2] Tryptophanase, denoted TPase in
W
e use our sense of smell every day to detect threats such as
Figure 1. Use of tryptophanase (TPase) to construct an odor-based
sensor. a) Concept. Linking TPase to a molecular recognition compo-
nent allows the reporting of a target through odor generation. b) Odor-
generating reactions catalyzed by TPase.
fire or spoiled food. The doping of natural gas with
mercaptans for leak detection is an elegant example of the
engineering of safety systems that exploit our sense of smell.
Surprisingly, there were no reports of taking advantage of our
olfactory sense for the design of a biosensor until Mohapatra
and Phillips recently reported an odor-based biosensor that
uses two small molecules and glucose oxidase to detect
enzymes as analytes.[1] The first small molecule is a glucose
derivative specifically designed to be a substrate for the
enzyme to be detected. The presence of this enzyme leads to
the generation of glucose; glucose then reacts with glucose
oxidase to produce hydrogen peroxide, which in turn reacts
this report, is a metabolic enzyme that is capable of
converting l-tryptophan and S-methyl-l-cysteine into indole
and methyl mercaptan, respectively (Figure 1b).[3] Indole and
methyl mercaptan can be easily detected by the human nose,
with threshold concentrations of 0.02 ppb (1.0 ꢀ 10À12 m) and
0.2 ppb (1.0 ꢀ 10À11m), respectively.[4] Therefore, the combi-
nation of TPase/l-tryptophan or TPase/S-methyl-l-cysteine
offers an attractive strategy for producing odors to which the
human nose is extremely sensitive. When linked to a molec-
ular recognition element (such as an antibody or aptamer),
TPase should allow for the odor-based detection of a wide
variety of molecular targets. To demonstrate such an appli-
cation, we herein describe the coupling of an antibody to
TPase and the use of the resulting biosensor to detect
a protein target.
TPase is also an enzyme that requires pyridoxal phosphate
(PLP) as a coenzyme; therefore, an analyte that can be linked
to the production of PLP would also be compatible with
a TPase-derived odor-based detection system. To this end, we
also developed an odor-based biosensor for adenosine
triphosphates (ATP), an important biological cofactor.
We first investigated the biotinylation of TPase in order to
establish a simple and general method for coupling a chosen
antibody to TPase. The biotinilyated TPase can be linked to
a biotinylated antibody through avidin, a tetrameric biotin-
binding protein. The biotinylation of TPase was accomplished
[*] Y. Xu, Z. Zhang, J. Sauder, X. Deng, K. Giang, Prof. Dr. R. Pelton,
Prof. Dr. C. D. M. Filipe
Department of Chemical Engineering, McMaster University
Hamilton, Ontario, L8S 4L7 (Canada)
E-mail: filipec@mcmaster.ca
Dr. M. M. Ali, S. D. Aguirre, Prof. Dr. Y. Li
Department of Biochemistry and Biomedical Sciences
McMaster University, Hamilton, Ontario, L8S 4L7 (Canada)
E-mail: liying@mcmaster.ca
[**] This work was supported by research grants from the Natural
Sciences and Engineering Research Council of Canada (NSERC) and
SENTINEL Bioactive Paper Network. We thank Professor Martin
Safo at Virginia Commonwealth University for the plasmid (pET22-
hPLK) to express pyridoxal kinase. We thank Dr. Weian Zhao for
critical reading of the manuscript and Dr. Rita Sassi for contributing
artwork.
Supporting information for this article, including experimental
details, is available on the WWW under http://dx.doi.org/10.1002/
anie.201309684.
2620
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 2620 –2622

Angewandte
Chemie
by using N-hydroxysulfosuccinimide biotin as a biotinylated
reagent.
l-tryptophan as the substrate (to give indole as the product),
and comparable results were obtained (see the Supporting
Information). These findings show that TLISA can report the
presence of a target analyte of interest with high sensitivity.
The ideal biosensor should have a fast response time. To
determine how long it would take to generate a measurable
signal, we examined the effect of different amounts of TPase
on the production of methyl mercaptan from S-methyl-l-
cysteine (Figure 2c). As expected, increasing the amount of
TPase led to faster generation of methyl mercaptan. Even
5 mg TPase can generate 100 ppb methyl mercaptan (500 ꢀ the
threshold level for the human nose) in less than 5 min
(Figure 2c), thus indicating that our system has a quick
response time. A similar test was also done with l-tryptophan
(see the Supporting Information).
To demonstrate protein detection through the use of an
antibody, we selected anti-rabbit IgG (ARIgG) as the anti-
body, which specifically recognizes rabbit IgG (RIgG) as the
target. To achieve target detection, we devised a detection
procedure that we termed TPase-linked immunosorbent
assay (TLISA; Figure 2a). Briefly, RIgG was allowed to
Next, we performed odor-based experiments to determine
the minimum amount of TPase that can actually be detected
by the human nose as a way to measure the sensory threshold
of the TPase-based sensing system. Following established
protocols,[5] we prepared a pre-reaction mixture that con-
tained S-methyl-l-cysteine (1 mm) and PLP (0.1 mm). This
mixture was then used to set up a series of testing sample sets,
each of which consisted two identical blanks (i.e. without
TPase) and a positive sample with varying amounts of TPase
(note: having two blanks served to decrease the probability of
identifying the positive sample by guessing). Each panelist
was asked to evaluate the entire collection of sample sets in
order of sequentially increasing concentrations of TPase and
identify the positive samples.
The best estimate threshold (BET) for each panelist was
calculated as the geometric mean of the highest concentration
that was not recognized and the next concentration identified
correctly. The BET for the entire panel (five people in our
case) was determined as the geometric mean of individual
BETs. The BET for TPase assay was found to be 3.6 nm
(Table S3 in the Supporting Information). We applied the
same method, but with varying amounts of RIgG, for
determining the BET for the TLISA assay, which was found
to be 85 nm (Table S4).
Coupled enzymatic assays are widely used in bioanalysis
and this strategy is particularly useful for analyzing enzyme-
catalyzed reactions for which the substrate and product are
difficult or impossible to measure.[6] The idea is to use
a secondary enzyme to transform the product of the first
reaction into a product that can be readily measured.
Motivated by the wide use of coupled enzymatic assays and
the general importance of ATP in the context of biological
systems, we decided to develop a coupled enzymatic assay
that uses odor as a readout for the presence of ATP. ATP is
the universal energy source in living organisms. It disappears
in less than two hours after cell death, and the amount of ATP
per cell is generally constant.[7] ATP detection is often used as
a general assay for monitoring drinking-water cleanliness[8]
and detecting food spoilage,[9] as well as for hygiene monitor-
ing in hospitals.[10]
Figure 2. TPase-linked immunosorbent assay (TLISA) for protein detec-
tion using an antibody. a) Schematic of TLISA. The target is RIgG and
the select selected antibody is ARIgG. The biosensor was assembled
by combining biotinylated TPase, avidin, and biotinylated ARIgG. The
addition of S-methyl-l-cysteine (or l-tryptophan) and PLP leads to the
formation of methyl mercaptan (or indole). b) Methyl mercaptan
(MM) concentration in the headspace generated for a test sample
(containing 440 mg RIgG) and a control (no RIgG) after one hour. Each
experiment was done in triplicate. c) Concentration of methyl mercap-
tan (MM) as a function of time for samples containing different
amounts of TPase, 15 mm S-methyl-l-cysteine and 0.1 mm PLP in KDP
buffer. Each data point is the average of three measurements. The
straight line represents the detection threshold for the human nose
(0.2 ppb).
bind to magnetic beads (MB) containing ARIgG. After
washing and blocking with super block T20 buffer, the MB
were incubated with pre-assembled TPase–Avidin–ARIgG
reporter and then washed. This was followed by the addition
of S-methyl-l-cysteine (or l-tryptophan) and PLP, which led
to the generation of methyl mercaptan (or indole) that can be
detected by a human nose. For quantification purposes,
a photoionization detector (ToxiRAE Plus PID) was also
used to analyze methyl mercaptan in the headspace. It is
noteworthy that the photoionization detector has a limit of
detection of 0.1 ppm, which means that the human nose is
about 500 times more sensitive than this instrument.
The concentration of methyl mercaptan reached 1.3 ppm
(Figure 2b) in the test sample containing RIgG. This concen-
tration is 6500 times higher than the threshold level for the
human nose. By contrast, no methyl mercaptan was detected
when RIgG was absent (Control in Figure 2b). We also tested
TPase is a tetrameric enzyme, the activity of which activity
is dependent on PLP.[11] Moreover, we found that the activity
of TPase can be modulated by the concentration of PLP
(Figure S3 in the Supporting Information). Because PLP can
Angew. Chem. Int. Ed. 2014, 53, 2620 –2622
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2621

.
Angewandte
Communications
be synthesized from pyridoxal by pyridoxal kinase (PKase)
using ATP as a cofactor, an odor-based sensor for ATP can be
conveniently constructed by coupling PKase and TPase and
their respective substrates pyridoxal and S-methyl-l-cysteine
(Figure 3a).
In summary, we have developed an odor-based sensor
system that exploits TPase as the key component. This
enzyme is able to convert the substrates S-methyl-l-cysteine
and l-tryptophan into methyl mercaptan and indole, respec-
tively, both of which can be easily detected by the human
nose. The catalytic power of the enzyme ensures rapid signal
generation. In addition, the use of biotinylated TPase allows
easy coupling of TPase to a molecular recognition element,
such as an antibody, for the development of an ELISA-like
assay. Moreover, we have shown that TPase can be coupled to
PKase for the creation of a system that is capable of detecting
ATP, which may find use in monitoring water and food
quality. The system presented herein is modular in nature and
the ability to combine the biotinylated TPase reporter with
virtually any molecular recognition element should allow for
a broad range of applications for this odor-based reporting
system.
Figure 3. ATP detection by using a coupled enzymatic odor-based
sensor. a) Coupled enzyme reactions. Pyridoxal kinase converts pyr-
idoxal into PLP, which serves as a cofactor to activate TPase to
produce methyl mercaptan (MM). b) The specificity of the ATP sensor.
Samples (700 mL) contained 0.2 mm of the respective nucleoside
triphosphate, 40 mg TPase, 50 mg PKase, 0.1 mgmLÀ1 pyridoxal, and
15 mm S-methyl-l-cysteine (or l-tryptophan). Each data point is the
average of three measurements.
Received: November 7, 2013
Published online: February 4, 2014
Keywords: biosensors · methyl mercaptan ·
.
molecular recognition · olfactory detection · tryptophanase
To validate the coupled assay and to demonstrate its
specificity towards ATP, we performed the following experi-
ment. First we prepared parallel samples containing PKase,
TPase, pyridoxal, and S-methyl-l-cysteine. Each solution was
then mixed with ATP, guanosine triphosphate (GTP), cytidine
triphosphate (CTP), or uridine triphosphate (UTP). For the
control sample, no nucleoside triphosphate was added.
Finally, the methyl mercaptan concentration in the headspace
was measured with the ToxiRAE Plus PID and the results are
shown in Figure 3b. As expected, only ATP was able to
induce the production of methyl mercaptan. The concentra-
tion of methyl mercaptan in the headspace was 49.1 ppm,
a value much larger than the 0.2 ppb detection limit for the
human nose. A similar experiment was performed with l-
tryptohpan as the substrate (see the Supporting Information).
We also determined the BET for ATP with a panel of 5
people and with S-methyl-l-cysteine as the substrate. The
BET was found to be 0.32 mm (Table S5).
[1] H. Mohapatra, S. T. Phillips, Angew. Chem. 2012, 124, 11307;
Angew. Chem. Int. Ed. 2012, 51, 11145.
[2] Y. Xu, Ph.D. Thesis, McMaster University (Hamilton, Canada),
2011.
[3] W. Newton, Y. Morino, E. Snell, J. Biol. Chem. 1965, 240, 1211.
[4] J. Greenman, M. EL-Maaytah, J. Duffield, P. Spencer, M.
Rosenberg, D. Corry, S. Saad, P. Lenton, G. Majerus, S.
Nachnani, J. Am. Dent. Assoc. 2005, 136, 749.
[5] a) ASTM Standard E 679 – 04, Annual Book of Standards,
Vol. 15.08, ASTM International, Conshocken, 2008, pp. 36;
b) H. T. Lawless in Food Science Text Series, 2nd ed. (Ed.:
D. R. Heldman), Springer, Heidelberg, 2010.
[6] H. U. Bergmeyer, Methods of Enzymatic Analyis, 3rd ed.,
Academic Press, London, 1983.
[7] L. Selan, F. Berlutti, C. Passariello, M. C. Thaller, G. Renzini, J.
Clin. Microbiol. 1992, 30, 1739.
[8] R. A. Deininger, J. Lee, Field Anal. Chem. Technol. 2001, 5, 185.
[9] M. W. Griffiths, J. Dairy Sci. 1993, 76, 3118.
[10] K. Seeger, M. W. Griffiths, J. Food Prot. 1994, 57, 509.
[11] a) C. Suelter, E. Snell, J. Biol. Chem. 1977, 252, 1852; b) F.
Happold, A. Struyvenberg, Biochem. J. 1954, 58, 379.
2622 www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 2620 –2622