Highly Efficient Flavin–Adenine Dinucleotide Glucose Dehydrogenase

Subscriber access provided by UNIVERSITY OF THE SUNSHINE COAST

Communication

Highly efficient flavin-adenine dinucleotide glucose dehydrogenase fused to a minimal cytochrome c domain Itay Algov, Jennifer Grushka, Raz Zarivach, and Lital Alfonta J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b07011 • Publication Date (Web): 15 Sep 2017 Downloaded from http://pubs.acs.org on September 15, 2017

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.

Journal of the American Chemical Society 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 5

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

Journal of the American Chemical Society

Highly efficient flavin-adenine dinucleotide glucose dehydrogenase fused to a minimal cytochrome c domain Itay Algov†, Jennifer Grushka†, Raz Zarivach†,§, Lital Alfonta†,* †

Department of Life Sciences and Ilse Katz Institute for Nanoscale Science and Technology, BenGurion University of the Negev, Beer-Sheva 84105, Israel. §National Institute for Biotechnology in the Negev, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel. * Corresponding author: Lital Alfonta; [email protected] Supporting Information Placeholder ABSTRACT: Flavin-adenine dinucleotide (FAD) dependent glucose dehydrogenase (GDH) is a thermostable, oxygen insensitive redox enzyme that has been widely used in bioelectrochemical applications. The FAD co-factor of the enzyme is buried within the proteinaceous matrix of the enzyme which makes it almost unreachable for a direct communication with an electrode. In the present study, FAD dependent glucose dehydrogenase was fused to a natural minimal cytochrome domain in its c-terminus to achieve direct electron transfer. Herein, we introduce a fusion enzyme that can communicate with an electrode directly, without the use of a mediator molecule. The new fusion enzyme, with its direct electron transfer abilities displays superior activity to that of the native enzyme, with a kcat that is ca. three times higher than that of the native enzyme, a kcat/KM that is more than three times higher than that of GDH and five to seven times higher catalytic currents with an onset potential of (-) 0.1 V vs. Ag/AgCl, affording much higher glucose sensing selectivity than previously reported. Taking these parameters into consideration, the fusion enzyme presented herein can serve as a good candidate for blood glucose monitoring and for other glucose based bioelectrochemical systems. Redox enzymes are proteins that participate in biocatalytic processes which involve electron transfer (ET)1. Depending on their redox potential, enzyme mediated redox reactions may be used in anodes and cathodes of biofuel cells2 as well as in biosensing applications3. For the utilization of most redox enzymes in such devices, a mediator molecule should be used to mediate the ET between the enzyme and the electrode. While some improvements in mediator synthesis were recently achieved by lowering their redox potentials and allowing mediated electron transfer with minor over potentials, the use of an external redox mediator can still lead to diffusional limitations4. Redox active polymers can be used to avoid the mediator diffusion by covering electrode surface with mediator containing polymer, however, these are expensive and further complicate the system5. Another approach that enables the achievement of efficient enzyme-electrode communication is by wiring the redox enzyme to the electrode. Wiring can be achieved as non-specific, by using a chemical modification, conductive polymers6 and semi-conductive or conductive matrices such as reduced graphene oxide7 and single or multi-walled carbon nanotubes8. Another approach is site-specific wiring by inserting non-canonical amino acids in a desired site that covalently link to a linker that is bound to the electrode9. Wiring an enzyme to an electrode, help to avoid diffusion limitations, since the enzyme is attached in a close proximity and with the proper orienta-

tion to the electrode. On the other hand, the wiring procedure may result in loss of enzymatic activity6 in addition to wiring being complicated and expensive due to challenging chemical synthesis of linkers. Hence, the ability to engineer an enzyme that allows direct electron transfer (DET) using protein domains that afford DET, will result in a “built in” redox mediator, eliminating the need for external molecules, which is very appealing in bioelectrochemical applications10. In order to achieve an efficient DET, a few conditions should be met. First, the distance between the enzyme’s active site and the electrode should be as short as a few Ångströms11. Second, if attached to the enzyme, the ET domain should be minimal so as not to introduce additional insulation to the system by a complex proteinaceous matrix, and linked by a flexible linker to avoid interrupting the catalytic redox activity12. Nature, has already designed such an enzyme with an intrinsic minimal cytochrome, allowing an efficient DET. The enzyme cellobiose dehydrogenase (CDH) is produced by various wood-degrading fungi, this enzyme can communicate with an electrode through DET13,14. CDH consists of two domains – a large catalytic flavin adenine dinucleotide (FAD) containing domain15, and a small cytochrome b containing domain that are connected by a polypeptide linker16. That unique property of CDH inspired us to hypothesize whether it is possible to engineer a redox enzyme with an added minimal ET unit to result in DET to an electrode. Redox enzyme fusion proteins were previously introduced as a way of creating new multi-domain proteins, that can be exploited as biological tools, as well as to improve enzyme-electrode communication17. Examples of such fusion enzymes are the PQQGDH fused to c-type cytochrome domain from ethanol dehydrogenase to achieve DET18 and a heme domain of cytochrome P450 BM3 fused to a flavodoxin19. FAD-GDH from Burkholderia cepacia is a thermostable enzyme and oxygen independent, making it a useful enzyme for blood glucose monitoring20,21. The native enzyme complex consists of 3 subunits – catalytic FAD binding α subunit, a small chaperon-like γ subunit and a c-type cytochrome β subunit22. The γ subunit, is needed as a chaperon for correct folding and maturation as well as for the FAD co-factor binding23. The γ subunit also contains a Twin arginine translocase (Tat) pathway motif sequence that signals the localization of the mature protein to the periplasm of the bacteria23. The β subunit is a 43 kDa c-type multi-heme cytochrome that is able to transfer electrons from glucose oxidation to an electron acceptor. The entire FAD-GDH complex have shown DET abilities with relatively low bioelectrocatalytic currents with a redox potential of +400 mV vs. Ag/AgCl 24. The

ACS Paragon Plus Environment

Journal of the American Chemical Society

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 5

staining30 was performed to verify the presence of the heme compared to the GDH α subunit only (Figure 2, left panel). Anti Histag Western blot analysis was performed in order to verify the full-length enzyme’s expression (Figure 2, right panel). Both FGM and GDH enzymes were expressed and their respective bands appeared as expected according to their calculated size – ca. 64.494 kDa and 62.080 kDa for FGM and GDH, respectively. Ingel heme staining revealed a band for FGM only, indicating the presence of a porphyrin containing iron bound to FGM.

size of the β subunit which enables ET in the native system, reduced the efficiency of ET to electrodes, due to its size25. Hence, we reasoned that replacing the large cytochrome c by a minimal cytochrome domain (MCD) will shorten the enzyme-electrode distance and improve the DET capabilities of the enzyme26. In addition, reducing the redox potential of the enzyme from +400 mV to 0.0 mV will improve enzyme selectivity in blood glucose monitoring as long as the sensing surface is adjusted to work at pH 5.0 and not 7.0 (physiological pH) since it's the optimal operational pH of the enzyme. In the present study, a fusion enzyme was designed as a combination of a biocatalytic function from a redox enzyme domain that was fused to a natural minimal ET domain via a short polypeptide linker as shown in figure 1. The catalytic domain was the α subunit of an FAD-GDH from Burkholderia cepacia20. As the minimal ET unit we chose the c-type cytochrome domain MCR-2 from a MamP protein which originates from a magnetotactic bacteria magneto-ovoid bacterium MO-127,28. MamP is part of the magnetosome, a unique organelle that is found in magnetotactic bacteria that allows magneto-taxis to occur in these bacteria. MCR-2 is one of the shortest natural c-type cytochromes known today (23 amino-acids long)27, thus can be used to achieve DET. We chose a short and flexible peptide linker (GSGYGSG) that contains glycine for flexibility, serine for hydrophilic interactions, with a length of up to 25Å in its extended form.29 A shorter linker could have prevented a proper maturation of the porphyrin binding MCD (due to its close proximity to GDH), on the other hand a longer linker could prevent efficient ET between the two domains. In order for the designed FAD-GDH-MCD (FGM) fusion enzyme to mature properly in the host cell, FAD-GDH α subunit should correctly fold. The Enzyme’s γ subunit aids in the maturation of the α subunit and translocates the entire construct to the periplasm. Within the bacterial periplasmatic environment, the maturation of c-type cytochromes occurs with the help of a specific gene cluster called ccmA-H30. In that manner, the holo-enzyme is being transferred to the periplasm and there the MCD’s maturation process occurs.

Figure 2. Left: In-gel heme staining showing the presence of a heme molecule in FGM compared to its absence in GDH. Right: Anti 6xHis-tag Western blot analysis verifying full length protein expression. FGM catalytic redox activity and heme peroxidase activity were measured biochemically in solution and compared to GDH as shown in figure S3. FGM has oxidized D-glucose as was measured by FAD-GDH activity assay in 50 mM Tris-base (pH 7.0), 0.6 mM 2,6-dichloroindophenol (DCIP) and 0.6 mM phenazine methyl sulfate (PMS) in 37˚C31. Absorbance (λ=610 nm) was monitored using a plate reader while the oxidized DCIP (blue color) was reduced by FGM to its reduced form (colorless). FGM has also showed peroxidase activity, measured by heme activity assay in 1 mM 3,3′-dimethoxybenzidine (DMB) and 1 mM of hydrogen peroxide30. Absorbance in 455 nm was monitored for DMB oxidation by MCD, resulting in an oxidized DMB (red color). Heme activity assay results indicate that FGM indeed binds a heme group while no heme molecules are bound by GDH. Absorbance measurements of protein sample spectrum have revealed a peak in absorbance at 408 nm for FGM and no peak for GDH, indicating the presence of heme c in FGM (figure S4). A408nm/A280nm ratio was calculated to be 0.4 for FGM expressed in the presence of pEC86 plasmid, compared to 0.2 for FGM expressed in the absence of this plasmid, indicating more efficient heme attachment and maturation in the presence of the helper plasmid. The apparent kinetic and thermodynamic parameters of FGM were calculated using GDH biochemical activity assay in 37˚C (Table 1). FGM and GDH concentrations were first determined spectrophotometrically using a standard Bradford assay. One enzyme activity unit was defined as the amount of enzyme oxidizing 1 µM of substrate per minute. We have calculated the molar absorption coefficient of DCIP according to a calibration curve to be 4.7 cm-1mM-1. By using the biochemical activity assay and DCIP molar absorption coefficient, FGM and GDH specific activity were calculated to be 16 mU in the reduction assay mixture. Lineweaver-Burk plots (figure S7) were used in order to calculate the kinetic and thermodynamic parameters of the enzyme. KMapp values were determined as 157±5 µM and 174±9 µM for FGM and GDH, respectively, showing slightly different affinity of the enzymes towards the substrate. GDH’s KMapp value is lower than previously reported values but yet in the same order of magnitude of some21,32,33. kcatapp value was ca. three times higher for FGM compared to GDH indicating faster oxidation of D-glucose by FGM. FGM also showed more than three times higher catalytic efficiency (kcatapp/KMapp) compared to GDH (Table 1). Next, we measured the electrochemical activity of the enzymes to deter-

Figure 1. A 3D model of FAD-GDH-MCD (FGM) based on the structure of GDH, predicted by homology to formate oxidase using Swiss-model (ID:3Q9T) and on the structure of MCD from MamP crystal structure (ID:4JJ0). FAD-GDH from burkholderia cepacia (green) is presented with its FAD binding motif (orange) and the FAD co-factor (light orange). MCD model (cyan) and the linker (grey) were cut from MamP known 3D structure and include the Heme binding motif (red) and a heme molecule (pink). The fusion shown in this figure was manually generated using PyMOL software. Fusion enzyme’s engineered DNA sequence was cloned into pTrcHis6A2 expression vector and was co-transformed into E-coli BL21 with pEC86 plasmid for constitutive ccmA-H operon expression31. FGM was overexpressed in the bacterial expression system and then purified by utilizing immobilized metal affinity chromatography purification system (figure S2). In-gel heme

2

ACS Paragon Plus Environment

Page 3 of 5

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

Journal of the American Chemical Society As shown in table 2, the electrochemical KMapp is 2.8±0.6 mM for GDH and 1.4±0.3 mM for FGM, indicating twice as much higher affinity of FGM towards glucose compared to GDH affinity towards glucose, from reasons that we cannot explain. However, the imax value is five times higher for FGM compared to GDH – 2.0±0.5 µA·cm-2 and 0.4±0.2 µA·cm-2, respectively. The difference in the imax value is indicative that the DET efficiency is different between the two enzymes where FGM shows five to seven times higher current then GDH for the same glucose concentrations (figure S5).

mine whether the addition of the minimal cytochrome domain will improve enzyme-electrode ET.

Table 1. Apparent biochemical kinetic parameters of FGM and GDH kcatapp (s-1)

KMapp (µM)

kcatapp/KMapp (s-1·mM-1)

GDH

1.7±0.1

174±9

9.6±0

FGM

5.2±0.1

157±5

33±0

For the electrochemical measurements, we have used a standard 3 electrode electrochemical cell with 0.9 mm graphite rod as the auxiliary electrode, 3 M KCl saturated Ag/AgCl reference electrode and 3 mm diameter glassy carbon electrode (GCE) as the working electrode. About 0.16 mU of FGM or GDH enzyme solution were added to the GCE surface and dried in 4˚C overnight. The electrodes surface was then covered with a cellulose dialysis membrane tightened to the surface with an O-ring to keep the enzyme close to the electrode surface during measurements and to avoid enzyme diffusion to the surrounding buffer. Cyclic voltammetry (CV) measurements were preformed to compare the enzyme-electrode communication of FGM to that of GDH. Optimal pH value for electrochemical measurements was tested by measuring the catalytic current in pH values of 3.6-7.0 and found to be 5.0. Measurements were performed in phosphate-citrate buffer pH=5.0 at room temperature and a scan rate of 100 mV/sec for both enzymes in the presence or absence of 5 mM glucose solution. It can be seen in figure 3 that the CVs of both enzymes before the addition of glucose are almost identical. No clear anodic or cathodic peaks were identified for both enzymes in any scan rate. After the addition of glucose to a final concentration of 5 mM, FGM has demonstrated a much higher electrocatalytic current compared to that of GDH with an onset potential of ca. (-) 150 mV. The fact that no catalytic current was observed at this high scan rate using GDH, but a significant catalytic current evolved using FGM, is an indication of fast ET rates of FGM. That observed ET efficiency is probably due to the addition of the minimal cytochrome domain that mediates ET from the buried FAD co-factor to GCE. To identify a peak originating from the MCD domain, square-wave voltammetry (SWV) was performed. As can be seen in figure 4, the voltammogram of an electrode with FGM, revealed a peak around -230 mV, while the GDH had no observable peak at this potential. According to a potentiometric titration conducted for MamP MCD by Chang and coworkers34, the protein redox potential was determined to be -89 mV ± 11 mV vs. NHE, this value is close to the value of the peak observed for FGM at -230 mV vs. Ag/AgCl. Given a 70% amino acids sequence identity between the two peptides. In addition, a small peak for FAD emerges in the SWV measurements for both enzymes at ca. -550 mV vs. Ag/AgCl that could be assigned for a partially exposed FAD, due to partial misfolding of the enzyme at the electrode surface. Chronoamperometric measurements were performed to determine the apparent kinetic parameters of FGM compared to GDH. The current was measured vs. successive glucose additions. A potential of 0.0 mV was applied during the measurements. The current for each glucose concentration was determined and is presented in figure 5. Using the linear part of the transient curves and a Lineweaver-Burk transformation – the electrochemical KMapp and imax values were calculated (figure S8).

Figure 3. CVs of GCE/GDH and GCE/FGM in the presence (+) and absence (-) of 5 mM glucose. The measurements were performed in 150 mM phosphate-citrate buffer, pH=5.0 at room temperature vs. Ag/AgCl as a reference electrode at a scan rate of 100 mV s-1.

Figure 4. SWVs of GCE/GDH and GCE/FGM. The measurements were performed in 150 mM phosphate-citrate buffer, pH=5.0 vs. Ag/AgCl reference electrode with 5 mV steps, amplitude of 10mV and a frequency of 5Hz. Background GCE current was subtracted from the signal. In addition to improved kinetic and thermodynamic parameters of FGM compared to GDH, the specificity of the enzyme towards glucose was determined using chronoamperometry while biasing the electrode potential to 0.0 mV vs. Ag/AgCl (a very low potential compared to the potentials used with GDH (ranging from +200 - +400 mV) Since under these relatively high potentials, many physiologically interfering molecules can be oxidized either by the electrode itself or by the enzyme, it is very beneficial to use a much lower potential under which only glucose will be bioelectrochemically oxidized. In order to determine the GCE/FGM selectivity, 3.6 mM of glucose were added followed by physiologically relevant interfering molecules in their relevant concentrations such as the sugars: maltose, xylose, lactose and galactose in addition to ascorbic acid and acetaminophen. It can be seen in figure S6 that when the applied electrode potential was 300 mV

3

ACS Paragon Plus Environment

Journal of the American Chemical Society

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

ascorbic acid was a major interfering molecule (due to its electrode oxidation under this potential) while the electrode biasing potential was lowered to 0.0 mV, hardly any interference was detected when 0.17 mM of ascorbic acid were added to a low concentration of 3.6 mM glucose. These results are an indication of a major advantage for the use of FGM as a selective glucose sensing enzyme.

Page 4 of 5

search was supported in part by an ISF research grant number: 232/13 (L.A.).

REFERENCES (1) (2) (3) (4) (5) (6) (7) (8) (9)

(10) (11) (12)

Figure 5. Steady state currents from chronoamperometry measurements of GCE/GDH (●) and GCE/FGM (■) using different glucose concentrations.

(13) (14)

Table 2. Apparent electrochemical parameters of FGM and GDH

(15) (16)

KMapp (mM)

imax (µA·cm-2)

GDH

2.8±0.6

0.4±0.2

(17) (18)

FGM

1.4±0.3

2.0±0.5

(19) (20)

GDH from Burkholderia Cepacia is a good candidate for glucose sensing applications because of its stability in high temperatures and insensitivity to oxygen. By adding a minimal cytochrome domain to its c-terminus and by the ability to ribosomally express the enzyme in its mature form with the aid of a helper ccm maturation plasmid. We were able to show an improved DET by showing higher bioeletrocatalytic currents compared to GDH with almost the same affinity to the substrate with a potential bias as low as 0.0 mV (vs. Ag/AgCl), which makes it more accurate for glucose biosensing with improved sensitivity and selectivity than previously reported for GDH. Moreover, the same approach could be taken for other redox enzymes with a buried enzymatic active site, provided the redox potentials are compatible. A thorough investigation of FGM ET properties will be possible in the future when the enzyme will be site specifically wired to an electrode.

(21) (22) (23) (24) (25) (26) (27)

(28) (29)

ASSOCIATED CONTENT Supporting Information

(30) (31)

Including experimental details, plasmids, constructs and strains is available free of charge on the ACS Publications website.

(32)

AUTHOR INFORMATION Corresponding Author

(33)

Lital Alfonta: [email protected] (34)

Notes The authors declare no competing financial interests.

ACKNOWLEDGMENT

Léger, C.; Bertrand, P. Chem. Rev. 2008, 108, 2379. Minteer, S. D.; Liaw, B. Y.; Cooney, M. J. Curr. Opin. Biotechnol. 2007, 18, 228. Zebda, A.; Gondran, C.; Le Goff, A.; Holzinger, M.;Cinquin,P.; Cosnier, S. Nat. Commun. 2011, 2, 370. Elmgren, M.; Lindquist, S. E.; Henriksson, G. J. Electroanal. Chem. 1992, 341, 257. Tasca, F.; Gorton, L.; Harreither, W.; Haltrich, D.; Ludwig, R.; Noll, G. J. Phys. Chem.. 2008, 112, 13668. Heller, A. Acc. Chem. Res. 1990, 23, 128. Ravenna, Y.; Xia, L.; Gun, J.; Mikhaylov, A. A.; Medvedev, A. G.; Lev, O.; Alfonta, L. Anal. Chem. 2015, 87, 9567. Halámková, L.; Halámek, J.; Bocharova, V.; Szczupak, A.; Alfonta, L.; Katz, E. J. Am. Chem. Soc. 2012, 134, 5040. Amir, L.; Carnally, S. A.; Rayo, J.; Rosenne, S.; Yerushalmi, S. M.; Schlesinger, O.; Meijler, M. M.; Alfonta, L. J. Am. Chem. Soc. 2012, 135, 70. Wang, J. Chem. Rev. 2008, 108, 814. Marcus, R. A.; Sutin, N. BBA Rev. Bioenerg. 1985, 811, 265. Gorton, L.; Lindgren, A.; Larsson, T.; Munteanu, F. D.; Ruzgas, T.; Gazaryan, I. Anal. Chim. Acta 1999, 400, 91. Harreither, W.; Felice, A. K. G.; Paukner, R.; Gorton, L.; Ludwig, R.; Sygmund, C. Biotechnol. J. 2012, 7, 1359. Gal, I.; Schlesinger, O.; Amir, L.; Alfonta, L. Bioelectrochemistry 2016, 112, 53. Hallberg, B. M.; Henriksson, G.; Pettersson, G.; Divne, C. J. Mol. Biol. 2002, 315, 421. Hallberg, B. M.; Bergfors, T.; Bäckbro, K.; Pettersson, G.; Henriksson, G.; Divne, C. Structure 2000, 8, 79. Gilardi, G.; Fantuzzi, A. Trends Biotechnol. 2001, 19, 468. Okuda, J.; Sode, K. Biochem. Biophys. Res. Commun. 2004, 314, 793. Gilardi, G.; Meharenna, Y. T.; Tsotsou, G. E.; Sadeghi, S. J.; Fairhead, M.; Giannini, S. Biosens. Bioelectron. 2002, 17, 133. Okuda-Shimazaki, J.; Kakehi, N.; Yamazaki, T.; Tomiyama, M.; Sode, K. Biotechnol. Lett. 2008, 30, 1753. Inose, K.; Fujikawa, M.; Yamazaki, T.; Kojima, K.; Sode, K. Biochim. Biophys. Acta 2003, 1645, 133. Tsuya, T.; Ferri, S.; Fujikawa, M.; Yamaoka, H.; Sode, K. J. Biotechnol. 2006, 123, 127. Yamaoka, H.; Ferri, S.; Fujikawa, M.; Sode, K. Biotechnol Lett 2004, 26, 1757. Yamazaki, T.; Okuda-Shimazaki, J.; Sakata, C.; Tsuya, T.; Sode, K. Anal. Lett. 2008, 41, 2363. Willner, I.; Katz, E. Angew. Chem. Int. Ed. Engl. 2000, 39, 1180. Wong, T. S.; Schwaneberg, U. Curr. Opin. Biotechnol. 2003, 14, 590. Siponen, M. I.; Legrand, P.; Widdrat, M.; Jones, S. R.; Zhang, W.-J.; Chang, M. C. Y.; Faivre, D.; Arnoux, P.; Pignol, D. Nature 2013, 502, 681. Blakemore, R. Science 1975, 190, 377. Chen, X.; Zaro J. L.; Shen, W. C. Adv. Drug. Deliv. Rev. 2013 65, 1357. Sanders, C.; Lill, H. Biochim. Biophys. Acta 2000, 1459, 131. Braun, M.; Thöny-Meyer, L. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 12830. Yamashita, Y.; Ferri, S.; Huynh, M. L.; Shimizu, H.; Yamaoka, H.; Sode, K. Enzyme Microb. Technol. 2013, 52, 123. Fapyane, D.; Lee, S.J.; Kang, S.H.; Lim, D.H.; Cho, K.K.; Nam, T.; Ahn, J.P.; Ahn, J. H.; Kim, S. W.; Chang, I. S. Phys. Chem. Chem. Phys. 2013, 15, 9508. Jones, S. R.; Wilson, T. D.; Brown, M. E.; Rahn-Lee, L.; Yu, Y.; Fredriksen, L. L.; Ozyamak, E.; Komeili, A.; Chang, M. C. Y. Proc. Natl. Acad. Sci. U.S.A. 2015, 112, 3904.

We would like to greatly acknowledge advice and pEC86 ccm plasmid from Prof. D. Pignol, and Dr. Chen Gutmann. This re-

4

ACS Paragon Plus Environment

Page 5 of 5

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

Journal of the American Chemical Society

Insert Table of Contents artwork here

5

ACS Paragon Plus Environment