Subscriber access provided by GAZI UNIV
Letter
Hydrolytic Enzymes as (Bio)-Logic for Wireless and Chipless Biosensors Nigel Forest Reuel, Joseph McAuliffe, Gregory A. Becht, Mehrdad Mehdizadeh, Jeffrey Munos, RuPing Wang, and William J Delaney ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.5b00259 • Publication Date (Web): 11 Feb 2016 Downloaded from http://pubs.acs.org on February 14, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Sensors is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 10
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Sensors
Hydrolytic Enzymes as (Bio)-Logic for Wireless and Chipless Biosensors Nigel F. Reuel1*, Joseph McAuliffe2, Gregory A. Becht1, Mehrdad Mehdizadeh,3 Jeffrey Munos2, RuPing Wang1, and William J. Delaney1 1. Central Research and Development – Materials Science Division – E. I. du Pont de Nemours and Company 2. Industrial Biosciences - E. I. du Pont de Nemours and Company 3. Engineering Technologies - E. I. du Pont de Nemours and Company *To whom correspondence should be directed –
[email protected] Abstract The switchable activity of allosteric, hydrolytic enzymes was used as a single-input, ‘buffer’ logic gate (performing YES and NOT) in a screen-printable biosensor. The enzyme substrate functioned as an ‘AND’ logic gate with the enzyme and co-factor as inputs. These (bio)-logic materials transduced a signal by the cofactor activating the enzyme which then degraded the substrate that formed the dielectric of a tuning capacitor in an inductor-capacitor (LC) circuit. The degradation of the substrate was engineered to shift the capacitance and thus the resonant frequency of the device. The resonant frequency was monitored wirelessly with a low-power vector network analyzer observing the S21 parameter. Proof of concept was shown with subtilisin as the enzyme, activated by calcium (100 μg/ml and 5 mM respectively) degrading a collagen substrate with a demonstrated wireless read range of up to 4 cm. Selectivity over other divalent cations (magnesium, copper II, and manganese II) and the effect of receiver motion were also shown on the wireless measurement. Keywords Allostery, enzyme, resonant, inductor, wireless, chipless, biosensor Enzymes are not only efficient catalysts1, but also selective switches for controlling transport and metabolic function within organisms. Allosteric enzymes, metalloenzymes, zymogens, and custom enzymes with de novo regulation sites2 can be viewed as logic gates, turning on or off under the influence of a variety of inputs. The activities of most enzymes can be controlled by temperature and pH3. Allosteric enzymes are activated or inhibited due to conformational changes wrought by the binding of an effector molecule4. Metalloenzymes require a strongly-bound metal cofactor to operate, such as urease binding to zinc or amylase binding to calcium5. Zymogens, or ‘pro-enzymes’ reside in a dormant form until activated by a secondary protein - a classic example is chymotrypsinogen that remains inactive until cleaved by an active trypsin6. Elementary Boolean logic gates (AND, OR, NOR, XOR, NAND, identify, inverter) composed of enzymes and their controlling inputs were first demonstrated over a decade ago7 and have since been used for a diverse range of electronic sensing applications7-8. Other (bio)-logic9 gates move back up the translation/transcription pathway and use DNA to form genetic logic gates, with AND, OR, NOR, NOT, NAND, and XOR all presented in the literature10. In this work we utilized substilisin, a hydrolytic metalloenzyme, to again demonstrate enzyme Boolean logic gates (buffer YES gate, AND gate) but this time the enzyme output (i.e. degradation rate) was directly transduced in a wireless, screen-printable inductor-capacitor circuit (LC) that was wirelessly interrogated. The enzyme degraded the capacitor dielectric which then shifted the resonant frequency of the device. The advantages, limitations, and applications of this method are 1 ACS Paragon Plus Environment
ACS Sensors
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 10
discussed. Although subtilisin was used for this prototype, hydrolases are a broad class of enzymes that digest ester, sugar, ether, peptide, carbon, and sulfur bonds11 and thus many other enzyme and substrate candidates could be found for this technique. These enzymes could serve as the logic in a ‘chipless’ biosensor or one that does not rely on traditional integrated circuit logic. Results and Discussion Herein we demonstrated signal transduction from enzyme-based logic via the degradation of allosterically activated hydrolases to a wirelessly measurable signal with a resonant circuit where the enzyme substrate (material to be degraded by the enzyme) forms the tuning capacitor. For a single input, YES gate, the inactive enzyme resided on or in the dielectric material (Fig 1a) and for the AND gate both the enzyme and cofactor acted as inputs to the substrate (Fig 1b – accompanying truth tables shown). Once activated by the specific cofactor, the enzyme digested the material causing an increase in capacitance, either by the influx of buffer (raising the permittivity, ε) or decreasing the displacement between the plates due to structure degradation (Fig 1a-b) as dictated by the equation for capacitance (Eq 1). A resonant circuit was formed by coupling the capacitor to an inductive coil. The coil and capacitor plates were screen printed on a flexible, Kapton® substrate (Fig 1c) and folded over the substrate material and enzyme (Fig 1d). Shorting of the capacitor plates and inductive coil was prevented by a thin, insulating layer (low-density polyethylene (LDPE) film was used). This device was wirelessly monitored by observing the location of the resonant peak using a vector network analyzer in a one or two port mode (S11 and S21 parameter respectively). =
Eq (1)
A few candidate enzymes were screened for their 1) ability to be modulated by an external cofactor and 2) ability to digest a substrate producing a dielectric change. We found that amylase could be moderately controlled by calcium concentration and pH (Supplement 1) but producing a uniform, dielectric film from starch was challenging. Mutanase12 (α(1, 3) glucanase; EC 3.2.1.59) was found to readily digest α(1,3) polysaccharide films13 and produced a measurable dielectric change but was not easily modulated by external factors, such as pH (Supplement 2). The protease subtilisin (from Bacillus licheniformis, Sigma P5380) was found to fit both desired parameters. The activity of the protease was readily controlled by calcium concentration (Fig 2a) as measured by absorbance assay of free tyrosine with Folin’s reagent given off from digestion of collagen14. It is difficult to compare results from this activity assay between independent experiments due to the exact timing, mixing, and formation of precipitates and thus the relative amounts of Tyrosine produced at different enzyme concentrations is of less importance than the overall trend, indicating an increase in enzyme activity due to calcium concentration. The one deviation from this trend (5 mM Ca and 100 μg/ml) was likely due to dilution error of the stock sample used for all three replicates. The digestion of collagen film was also found to produce an appreciable change in capacitance of the film (up to 1.5 pF) as measured by a four point impedance analyzer with dielectric test fixture (Agilent 16451B – Fig 2b). The capacitance shift is better observed by plotting each of the response curves at a specified frequency (Fig 2c- 40.68 MHz is shown as an open ISM band and 25 MHz is shown as the frequency that most closely matches the wireless measurements we performed below). These curves are normalized at the second time point, as we observed a settling period immediately after the collagen and enzyme material were clamped between the parallel plates of the test fixture. The control conditions (0 mM Ca only, 5 mM Ca only, 0 mM Ca with 10 μg/ml subtilisin) all showed a small increase in capacitance, likely due to the universal effect of 2 ACS Paragon Plus Environment
Page 3 of 10
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Sensors
buffer influx (increasing the permittivity, ε, in equation 1). However, there was a clear increase in capacitance when both enzyme and calcium were present (5 mM Ca and 10 μg/ml subtilisin) and thus we continued to making a wireless device from the subtilisin, calcium, and collagen materials. Using the screen printed inductive coil circuit (Fig 1a-b) we then confirmed that this enzymatically mediated change in capacitance could be detected wirelessly as a shift in resonant frequency. First we used ceramic capacitors of known values with the screen printed circuit to determine limits in read range and detectable capacitance values. The resonant frequency was detected with a low power, portable vector network analyzer (Copper Mountain Tech S5048) by measuring the S21 parameter; this was done with transmitting and receiving antenna loops that are positioned for null mutual induction (the magnetic fluxes between the overlap area and non-overlap areas were equalized) at a specified frequency, in our case tuned to the 40.68 MHz ISM band (Supplement 3). A 6 pF capacitor was found to tune our LC circuit to this ISM band and we observed a detectable signal peak in the range of 0 to 15 cm using our rudimentary VNA antennas (Fig 3a and Supplement 4). Note that this maximum read range (ability to observe a device peak in the S21 curve) only occurs if the device resonant frequency matches the null mutual induction frequency of the VNA loops. Further optimization of VNA antenna geometry (size, shape, and frequency matching) would boost read range. We used a panel of capacitors (2-1000 pF) to observe the extent of resonant frequency shift in the S21 measurement (Fig 3b). The resonant peak was very clear in the 2-100 pF capacitance range and the data collected adheres to the theoretical relationship of resonant frequency (Eq 2) with a single fit value of 2.86 μH for the inductance of this coil. Having already determined that the single film collagen capacitors range in the 10-25 pF range (Fig 2b), it was expected that the device peak could be observed in the 0-100 MHz sweep range and that a modest change in capacitance could be detected (as seen with the ceramic capacitors Fig 3b). The enzymatic response on the collagen film was tested in the wireless LC circuit setup by adding 20 μl of the logic inputs (calcium and/or enzyme) to the collagen substrate, closing the device, and observing the S21 response. The controls (0 mM Ca only, 5 mM Ca only, 0 mM Ca with 10 μg/ml subtilisin) did not produce a shift, whereas adding enzyme in the presence of calcium (5 mM and 100 μg/ml subtilisin) produced a shift of 888 kHz in 20 min which translates to a 1.2 pF change (Eq 2), commensurate with our findings from the impedance analyzer (Fig 2b). The concentration of subtilisin in these experiments increased from 10 μg/ml used in the Cp-D measurements (Fig 2b) to 100 μg/ml in order to speed up the response time (from 100 min to 20 min). This could be done as the VNA setup used automated data acquisition whereas the Cp-D data was gathered manually (single traces saved to disk). =
1
Eq (2)
2 √
Next we evaluated the selectivity of the subtilisin logic gate by comparing the calcium response to other divalent cations, such as magnesium, copper II, and manganese II (Fig 3d - Supplement 5 for full curves). Magnesium produced a smaller shift (299 kHz) and copper II smaller still (65 kHz) whereas manganese produced a 132 kHz shift in the opposite direction, indicating a small decrease in capacitance. We confirmed this unexpected result with a second trial (Supplement 5) and are not sure what the mechanism might be. However this general downward trend in enzyme activity is supported by work in the literature15 that shows decreasing activity in subtilisin as its associated cation moves from calcium to magnesium, copper, and manganese. Finally we addressed the limitations of our wireless measurement system by observing the effect of movement and rotation on the sensor response. The resonant frequency (measured as the inflection 3 ACS Paragon Plus Environment
ACS Sensors
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 10
point of the S21 curve – see Fig 3c) varies from experiment to experiment as each piece of collagen used is not the exact same size, nor is the positioning of the device in front of the VNA coils the same between experiments. However, because the response is a frequency shift, it is tolerant of small displacements, angles between receiver and transmitter, as well as variation in substrate size. However when one moves or rotates the device substantially, the response curves do shift in the frequency space (Supplement 6). However, we found that fitting the response to a four parameter logistics curve (Eq 3) gives parameters that track the motion of the receiver. The difference in the minimum and maximum asymptote (A and D respectively) yields the span of the response, the inflection point (C) corresponds to the resonant frequency, and the slope of the curve at the inflection point is reported as well (B). Displacement can be determined by observing the change in the response span and then determining the expected shift in frequency (both linearly dependent – see supplement 6). If rotation is anticipated, the same process is used, however the relation between span, inflection point, and rotation angle are non-linear (supplement 6). Even though these corrections could be applied to account for motion the best method for wireless measurement in practice would be a setup where the orientation and displacement through the read cycle are fixed (for dynamic data) or returned to similar positions (for intermittent collection on objects in motion). = Methods
−
1 +
+
Eq (3)
Enzyme and Buffer Preparation A 50mM potassium phosphate buffer, pH 7.5 was made with 11.4 mg/ml potassium phosphate dibasic (Sigma p5504), trihydrate in low calcium water (
