Artificial Modification of an Enzyme for ... - ACS Publications

Subscriber access provided by United Arab Emirates University | Libraries Deanship

Article

Artificial modification of an enzyme for construction of cross-reactive polyion complexes to fingerprint signatures of proteins and mammalian cells Shunsuke Tomita, Osamu Niwa, and Ryoji Kurita Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b02010 • Publication Date (Web): 22 Aug 2016 Downloaded from http://pubs.acs.org on August 23, 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.

Analytical Chemistry 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 22

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

Analytical Chemistry

Artificial modification of an enzyme for construction of cross-reactive polyion complexes to fingerprint signatures of proteins and mammalian cells Shunsuke Tomita*a, Osamu Niwaa,b, and Ryoji Kurita*a a

Biomedical Research Institute, National Institute of Advanced Industrial Science and Technology, and DAILAB, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8566, Japan

b

Advanced Science Research Laboratory, Saitama Institute of Technology, 1690 Fusaiji, Fukaya,

Saitama 369-0293, Japan

* To whom correspondence should be addressed. Tel.: +81-29-861-2634.

Fax: +81-29-861-6177.

E-mail: [email protected]; [email protected]

1

ACS Paragon Plus Environment


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 22

Abstract A novel strategy for construction of cross-reactive enzyme-based sensors have been developed based on chemical modification of lysine ε-NH3+ groups in β-Galactosidase (β-Gal) from Aspergillus oryzae with various acid anhydrides. Modification of lysine side chains markedly altered the kinetic parameters of β-Gal (KM and kcat), whereas catalytic activity remained and the tertiary structure was maintained for all modified β-Gals. The addition of cationic PEGylated polymers to reactive solutions caused the formation of polyion complexes (PICs) with marked inhibition of the modified β-Gal activity. Therefore, the obtained PICs can be used to construct cross-reactive enzyme-based sensors without any purification. With addition of each analyte protein or mammalian cell to the PIC library, the modified β-Gals were partially released from PICs, and therefore the catalytic activities showed characteristic recovery fingerprints. Linear discriminant analysis (LDA) of fingerprints generated by the combination of only three PICs enabled successful discrimination of 0.5 µg/mL human plasma proteins (albumin, immunoglobulin G, transferrin, fibrinogen, α1-antitrypsin, C-reactive protein), and semi-quantification of inflammatory biomarker proteins in buffer (0.3 – 8.1 µg/mL) and human serum (2 – 100 µg/mL) with comparable sensitivity for diagnosis in human blood samples. Moreover, we identified five mammalian (human, bovine, fetal bovine, horse, and rabbit) sera containing 2.5 µg/mL serum proteins, and three human cancer cell lines [A549 (lung), MG63 (bone), HuH7 (liver)] and a human mesenchymal stem cell line (UE7T-13) (1500 cells/mL) with 100% accuracy.

2


Page 3 of 22

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


Introduction Enzyme-based biosensors are of great importance for most analytical tasks, ranging from medical diagnostics, food safety, security, to environmental monitoring. Two main classes of enzyme-based biosensors are distinguished according to the type of interaction providing the signal generation: metabolism sensors and affinity sensors.1 Metabolism sensors, such as glucose biosensors, use enzymes as both the element for specific recognition and chemical conversion of the biological analyte to the corresponding products that can be detected using physicochemical detectors. However, many biological analytes of interest are not amenable to detection due to a lack of availability of sufficiently selective enzymes for the analyte. Therefore, affinity sensing has been considered as an alternative method, which utilizes the selective and strong binding of antibodies to a target analyte. The binding event is transduced through catalytic enzymes, such as horseradish peroxidase, conjugated to secondary antibodies. Despite its high selectivity and sensitivity, however, this approach still suffers from high production costs for the development of antibodies toward target analytes. Recently, we have developed a new type of enzyme-based biosensor,2-5 using a “cross-reactive sensor arrays.”6,7 Cross-reactive sensor arrays are based on pattern recognition of unique optical fingerprints obtained through differential interactions between analytes and a library of cross-reactive receptors. The most common approach is the use of fluorescent probes or their conjugates with quenchers, allowing the discrimination of proteins,8-16 bacteria,10,15,17-20 and mammalian cells.10,16,21-24 Instead of fluorescent probes, enzyme activity has been adopted as a signal amplifier for sensor arrays in which cross-reactivity is generated by conjugating an enzyme with gold nanoparticles possessing different hydrophobic groups25 or using artificial enzyme-linked receptors.26 In addition to the signal amplifier, our previous enzyme-based biosensing method exploited naturally occurring structural diversity of enzymes to generate cross-reactivities for fingerprinting analysis of proteins.2 Specifically, several anionic enzymes were first mixed with a cationic polymer to form polyion complexes (PICs). PIC formation turned off enzyme activity, but the further addition of analyte proteins to PICs released the enzymes and partially recovered catalytic activity the level of which reflected protein characteristics. As each enzyme has different cross-reactivity toward proteins, the use of PICs containing different enzymes offered unique fingerprints of changes in enzyme activity. The obtained fingerprints were then analyzed statistically to identify proteins using a chemometric method. However, commercially available pairs of enzymes and substrates are limited. Furthermore, not all enzymes are switched off through PIC formation (see below, Table S-1). These issues hamper the construction of effective and practical systems suitable for sensing of complex real protein solutions 3



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 22

with similar characteristics. Synthetic introduction of structural diversity into cationic polymers is one possible way to create effective cross-reactive PICs,3 which has been used successfully to discriminate complex real protein solutions, such as cell culture supernatants4 and mammalian sera.5 However, this method requires laborious synthetic efforts. The artificial acetylation of lysine ε-NH3+ groups by acid anhydrides represents a simple and inexpensive method for alternation of enzyme characteristics. Acetylation by acetic anhydride neutralizes the charge of lysine side chains and increases the net negative charge of proteins, and is therefore useful for investigating the roles of electrostatics in protein stability and function.27,28 For some proteins, such modifications increased their resistance to irreversible inactivation and aggregation without perturbing the tertiary structure.29,30 As maleic anhydride derivatives can convert lysine side chains from cationic to anionic, the modification of succinic anhydride with proteins provides a molecular label for semiconductor-based potentiometric biosensors31 and yields insight into the effects of protein charge on complex coacervation with polyelectrolytes.32 PIC micelles formed between anionic block-copolymers and charge-converted cationic proteins by modification of pH-sensitive maleic anhydrides possessed high salt stability and pH-responsive degradability in the endosome, which is useful for efficient intracellular delivery of proteins.33-35 In addition to charge neutralization or conversion, various hydrophobic groups can be introduced simultaneously into lysine side chains.36,37 In this study, we used the modification of lysine ε-NH3+ groups in an enzyme by acid anhydrides to develop a simple and inexpensive strategy for construction of highly sensitive and cross-reactive PICs (Figure 1). An anionic enzyme was artificially modified with various acid anhydrides possessing different characteristics, followed by the addition of a cationic polymer into reactive solutions to turn off the catalytic activity of modified enzymes through the formation of cross-reactive PIC libraries. Newly developed libraries enabled not only qualitative human plasma protein discrimination but also semi-quantification of inflammatory biomarker proteins, and discrimination of mammalian sera and cells.

Experimental section Materials β-Galactosidase (β-Gal) from Aspergillus oryzae, 4-methylumbelliferyl-β-D-galactopyranoside (MUG), 3-(N-morpholino)propanesulfonic acid (MOPS), hexanoic anhydride, benzoic anhydride, succinic anhydride, phthalic anhydride, 1,4-dioxiane, immunoglobulin G from human serum (IMM), α1-antitrypsin from human plasma (ANT), fibrinogen from human plasma (FIB), apotransferrin from 4


Page 5 of 22

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


human (TRA), and albumin from human serum (ALB) were obtained from Sigma Chemical Co. Acetic anhydride, guanidine hydrochloride (GdnHCl), dimethyl sulfoxide (DMSO), and acetone were obtained from Wako Pure Chemical Ind. Fluorescamine was obtained from Tokyo Chemical Industry Co., Ltd. 3-[4-(2-Hydroxyethyl)-1-piperazinyl]propanesulfonic acid (EPPS) was obtained from Dojindo Lab. Native human C-reactive protein (CRP) was obtained from AbD Serotec. Horse, rabbit, and bovine sera were obtained from Cosmo Bio Co., Ltd. Human and fetal bovine sera were obtained from Thermo Scientific. All chemicals used were of high-quality analytical grade and were used as received.

Preparation of artificially modified enzymes Lysine ε-NH3+ groups in β-Gal were modified with various acid anhydrides according to the reported procedures with slight modifications.30,33,36 Briefly, 20-µL aliquots of solutions containing various concentrations of acid anhydrides in 1,4-dioxiane were added to 230 µL of a solution containing 5.0 µM β-Gal in 100 mM EPPS (pH 9.0). After incubation for 60 minutes at room temperature, the degree of chemical modification was determined by the fluorescamine method33-35 as follows: artificially modified β-Gals diluted 5-fold with 100 mM EPPS (pH 9.0) or 20-fold with 8.4 M GdnHCl, 100 mM EPPS (pH 9.0) (150 µL, 0.92 µM or 0.23 µM, respectively) were incubated with 50 µL of 3 mg/mL fluorescamine in acetone for 5 minutes at room temperature, and the fluorescent signal at 460 nm was measured using a microplate reader (Fluoroskan Ascent; Thermo Labsystems) with excitation at 355 nm. β-Gals modified with 10 equiv. of acetic anhydride (E2), 10 equiv. of hexanoic anhydride (E3), 6 equiv. of benzoic anhydride (E4), 20 equiv. of succinic anhydride (E5), and 10 equiv. of phthalic anhydride (E6) were used in the following experiments (the degrees of the chemical modification were 91%, 97%, 89%, 94%, and 94%, respectively). Prior to enzyme assay, shaking treatment, and cross-reactive sensing, all of the solutions containing artificially modified β-Gals were diluted with 10 mM MOPS buffer (pH 7.0). The concentration of β-Gal was determined from the absorbance at 280 nm using a spectrophotometer (V-630; Jasco Corp.) with an extinction coefficient of 192075 M–1 cm–1.38

Enzyme assay Enzyme kinetics: Aliquots of 192 µL of 1.0 nM artificially modified β-Gals in 10 mM MOPS (pH 7.0) were mixed with 8 µL of 0 – 18.75 mM MUG in DMSO in 96-well plates (96 Well Black Flat-Bottom Polystyrene NBS™ Microplates; Corning Inc.), and the time course of the increase in fluorescence at 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

Page 6 of 22

460 nm was recorded using a microplate reader with excitation at 355 nm. The reaction rate was obtained from the initial slope of the fluorescence increases, which were derived from the fluorogenic products produced by hydrolysis of MUG. The kinetic parameters, the Michaelis constant (KM) and turnover number (kcat), were determined based on the Lineweaver–Burk equation,

where [E] and [S] are β-Gal and substrate concentration, respectively. In this study, the product concentration was not considered quantitatively and so the values of kcat were used for relative evaluation. To evaluate the effects of residual byproducts on the kinetic parameters, hydrolyzed acid anhydrides were prepared as described above but without β-Gal. Unmodified β-Gal was mixed with prepared solutions containing hydrolyzed acid anhydrides with the same concentrations as in the case of artificially modified β-Gals, and the time course of the fluorescence increase was measured. Inhibition assay: Quaternized polyethylene glycol-block-poly(N,N-dimethylaminoethyl methacrylate)s (PEG-b-QPAMAs) with iodoethane (P1, Mw: 11500) and benzyl bromide (P2, Mw: 13900) were synthesized as reported previously.3 The degrees of polymerization of PEG-b-QPAMA were 102 for PEG and 35 for QPAMAs. Various concentrations of PEG-b-QPAMAs were mixed with artificially modified β-Gals (0.5 nM in 10 mM MOPS, pH 7.0, and 1.0 nM in the same buffer supplemented with 1% serum). These solutions (192 µL) were loaded into each well of 96-well plates. After incubation for 30 minutes at 30°C, 8 µL of 25 mM MUG in DMSO was added to each well, and the time course of the increase in fluorescence was recorded.

Circular dichroism Circular dichroism (CD) experiments were performed in a 1-mm path length quartz cuvette using a spectropolarimeter (J-720; Japan Spectroscopic Co., Ltd.). Artificially modified β-Gals were diluted 3-fold with 33 mM EPPS (pH 9.0), and the spectra were measured at 25°C. The CD spectra were corrected by subtracting the corresponding spectra of the solutions in the absence of artificially modified β-Gals. To evaluate the effects of residual byproducts on the CD spectra, unmodified β-Gals mixed with hydrolyzed acid anhydrides prepared as described above were also measured.

Shaking treatment Artificially modified β-Gals (25.0 nM in 10 mM MOPS, pH 7.0) were added to 0.5-mL microcentrifuge tubes. The samples were then vortexed at 2000 rpm for 30 minutes at room temperature using a micro tube mixer (MT-400; Tomy Digital Biology Co., Ltd.). The enzyme 6


Page 7 of 22

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


activities were measured by the assay described above.

Cross-reactive sensing Samples were prepared as follows (Figure S-1). Proteins: Concentrations of proteins were determined from the absorbance at 280 nm using a spectrophotometer with extinction coefficients of 1.427 (mg/mL)–1 cm–1 (IMM), 0.450 (mg/mL)–1 cm–1 (ANT), 1.689 (mg/mL)–1 cm–1 (FIB), 1.132 (mg/mL)–1 cm–1 (TRA), and 0.518 (mg/mL)–1 cm–1 (ALB).38 The supplied solution of 1.0 mg/mL CRP was diluted with 10 mM MOPS (pH 7.0) without any purification. Mammalian sera: Total protein in the sera was quantified using the Bradford Reagent (Sigma Chemical Co.) according to the manufacturer’s instructions, and then diluted to a total protein concentration of 25 µg/mL with 10 mM MOPS (pH 7.0). Mammalian cells: The human lung adenocarcinoma epithelial cell line (A549), human osteosarcoma cell line (MG63), human hepatoma cell line (HuH7), and human bone marrow-derived mesenchymal stem cell line (UE7T-13) were obtained from the Japanese Collection of Research Bioresources. All of the cells were grown in DMEM (Wako Pure Chemical Ind.) supplemented with 10% FBS (GE Healthcare Life Science) and 1% penicillin-streptomycin-neomycin antibiotic mixture (Life Technologies). Cells were washed with DPBS buffer (Wako Pure Chemical Ind.), trypsinized with 1× trypsin (Life Technologies), and collected in CDCHO media (Invitrogen) supplemented with 8 mM L-glutamine

(Wako Pure Chemical Ind.).

Aliquots of 172 µL of solutions containing artificially modified β-Gals and PEG-b-QPAMA were loaded into each well of 96-well plates (0.58 nM β-Gals and 7.0 nM P1 for 10 mM MOPS (pH 7.0), 1.16 nM β-Gals and 116.3 nM P2 for the same buffer supplemented with 1% human serum), followed by the addition of 20 µL of analytes. The solutions were incubated for 30 minutes at 30°C. After 8 µL of 25 mM MUG in DMSO was mixed with the solutions, the time course of the fluorescence increase was recorded using a microplate reader. The final concentrations of analytes were 0.5 µg/mL (proteins, qualitative), 0.3 – 8.1 µg/mL (proteins, semi-quantitative), 2.5 µg/mL (mammalian sera), and 1500 cells/mL (mammalian cells).

Results and discussion Preparation and characterization of artificially modified enzymes We chose β-Galactosidase from Aspergillus oryzae (β-Gal, E1) (Mw = 110 KDa, pI = 5.2) as a model enzyme, which is one of the major enzymes used in the food industry.39 β-Gal is an anionic 7



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 8 of 22

enzyme with -4.8 kDa per charge at pH 7.0 due to the presence of 58 aspartic acid, 50 glutamic acid, 50 lysine, and 33 arginine residues. Lysine side chains in E1 were simply acetylated for E2, while hydrophobic groups were simultaneously introduced for E3 (hexanoyl), E4 (benzoyl), and E6 (2-carboxybenzoyl) (Log P > 1.7). In addition, cationic lysine side chains in E1 were converted to anionic carboxyl-containing groups for E5 and E6. The byproducts of the acetylation reaction are the corresponding carboxylic acids (E2, E3, and E4) or H2O (E5 and E6). Far-UV CD spectra of unmodified E1 mixed with these byproducts were almost identical to those without the byproducts (Figure S-2). The kinetic parameters, the Michaelis constant (KM) and turnover number (kcat), determined from Lineweaver–Burk plot were not significantly different between E1 with/without the byproducts (Figure S-3). These results suggested that the influences of the byproducts on both tertiary structure and catalytic activity of β-Gal were negligible, and therefore artificially modified β-Gals were used for the following procedures without any purification; PIC formation with a block copolymer with PEG and cationic quaternized poly(N,N-dimethylaminoethyl methacrylate) (PEG-b-QPAMA)3 (Figure 1). Due to the functional groups introduced at lysine side chains, PICs containing artificially modified β-Gals may possess different cross-reactivities toward proteins or mammalian cells (Figure 1). Therefore, mixing of PICs with biological analytes was expected to generate fingerprints of changes in catalytic activity. The optimal conditions for chemical modification by acid anhydrides were first determined using an amine-reactive fluorescamine.33-35 The ratios between the fluorescence before and after the modifications were markedly decreased by the addition of all acid anhydrides. (Figure S-4A). In more detail, acid anhydrides showed 90% reduction in fluorescence at 8 equiv. (E2, E3 and E6), 6 equiv. (E4) and 14 equiv. (E5). The buried lysine residues protected from solvent or an intramolecular interaction involving lysine side chains may hinder the reactivity of the lysine ε-NH3+ groups with acid anhydrides. The assay using fluorescamine in the presence of a high concentration of denaturant (GdnHCl)40 indicated that ~95% lysine side chains were reactive. Taken together, these observations indicated that sufficient amounts of acid anhydrides against ε-NH3+ groups were used for modification of ~90% of lysine ε-NH3+ groups in β-Gal. Therefore, we used β-Gals modified by optimized concentrations of acid anhydrides (see Experimental section) in the following sensing experiments. We characterized unmodified and artificially modified β-Gals by CD, enzyme kinetics, and stability tests based on shaking treatment. To evaluate whether chemical modifications perturbed the tertiary structure of β-Gal, CD spectra were measured at the same pH as that for chemical modifications (pH 8


Page 9 of 22

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


9.0) just after the modification of β-Gal. Chemical modification of E2 – E5 only slightly changed the CD spectra around 220 nm, while the secondary structure of E6 was almost the same as that of E1 (Figure S4-B). There were small positive (E4) and negative (E6) Cotton effects around 240 – 260 nm, possibly due to the modified aromatic groups. CD analyses suggested that large changes in charge and hydrophobicity on the β-Gal surface induced only slight perturbation of secondary structure of β-Gal. Hydrolytic activity assays for artificially modified β-Gals were performed to investigate the influence of chemical modifications on catalytic activity. The values of KM for E2, E5, and E6 were higher than those of unmodified E1 in the following order: E6 > E5 > E2 (Figures 2C and S-5A), suggesting that both the acetylation of lysine side chains and the introduction of further anionic groups decreased the substrate affinity toward β-Gal. On the other hand, increased hydrophobicity of E6 against E5 likely increased KM, but the values of KM for E3 and E4 remained unchanged compared to that of E1 despite the significant modification of hydrophobic hexyl (E3) and aromatic (E4) groups at lysine side chains. It was assumed that increases in hydrophobicity for simply neutralized E3 and E4 were favorable for the binding of relatively hydrophobic substrate (MUG), which possibly cancelled the unfavorable effect of lysine acetylation. The values of kcat showed a different tendency compared with KM (Figures 2D and S-5B). E5 and E6 had nearly 2-fold higher values of kcat than E1, but that of E2 was lower, indicating that the acetylation of lysine side chains reduced the catalytic rate constant as well as substrate affinity. In contrast to substrate affinity, however, increases in anionic charges on the enzyme surface may facilitate enzyme catalysis, and therefore the favorable effect of increases in anionic charges on kcat probably exceeded the unfavorable effects of lysine side chain acetylation. It should be noted that the modification of hydrophobic groups (E3 and E4) significantly decreased kcat. All of these effects were due to slight perturbation of overall or local tertiary structures (Figure 2B) and significant alteration of electrostatic or hydrophobic nature of the enzyme surface. Further investigations are required to gain a sufficient understanding of the mechanisms governing the influences of chemical modifications on enzyme activities, which will be useful information for the development of simple and inexpensive means to improve enzyme performance or possibly tuning substrate specificity of enzymes. More importantly, the catalytic activity of β-Gal remained regardless of the marked alteration of the surface characteristics of β-Gal by acetylation and modification of various anionic and hydrophobic groups. We finally investigated the stability of artificially modified β-Gals based on shaking treatment. It is well known that mechanical stress, such as shaking, is a major cause of aggregation and inactivation of proteins during manufacturing and transport.41 Interestingly, the catalytic activities of all modified 9



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 10 of 22

β-Gals were almost fully maintained after vigorous shaking treatment (Figure S-6), suggesting that sufficient stability against mechanical stress was retained in β-Gals after significant chemical modifications. This stability is advantageous to obtain reproducible fingerprints derived from PICs containing the modified β-Gals after construction of a sensing system, as discussed below.

Construction of cross-reactive enzyme-based sensors β-Gal has a net negative charge under these buffer conditions (pH 7.0). Chemical modification by acid anhydrides causes neutralization (E2, E3, and E4) and charge conversion (E5 and E6) of cationic lysine ε-NH3+ groups. P1 possessed 35 PAMA units quaternized by iodoethanol, which were positively charged regardless of pH. The number of cationic groups were therefore nearly twofolds higher than the net charge of E1 and comparable to those of E2 – E6 at pH 7.0, resulting in the inhibition of both unmodified E1 and artificially modified β-Gals (E2 – E6) through multiple electrostatic interactions (Figure 3). P1 showed 90% inhibition at 20 equiv. (E1), 8 equiv. (E2 and E3), 3 equiv. (E4), 12 equiv. (E5), and 6 equiv. (E6). P1 had a greater effect on the decrease in activity of artificially modified β-Gals compared with E1, suggesting that the increase in net negative charge of β-Gal by acetylation of lysine side chains provided higher affinity toward cationic P1. The influences of modified groups on the inhibitory effect were likely less than those of acetylation. However, P1 seemed to act as the strongest inhibitor toward E4 in which aromatic groups were introduced. Therefore, hydrophobic interactions also contributed to increased affinity between P1 and artificially modified β-Gals, in agreement with a previous report.4 These differences in chemical modifications would provide diverse interactions between PICs and biological analytes.

Sensing of biological analytes After determination of the optimal β-Gals and P1 ratio, we prepared cross-reactive enzyme-based sensors by mixing appropriate stoichiometry of artificially modified β-Gals and P1, and examined the ability of this sensor to identify five of the most abundant human plasma proteins—albumin (ALB), immunoglobulin G (IMM), α1-antitrypsin (ANT), fibrinogen (FIB), and transferrin (TRA)—and C-reactive protein (CRP), an important marker for the diagnosis of infection and inflammation,42 as initial targets (Figure S-1A). The data matrix obtained through addition of 0.5 µg/mL proteins to PICs (6 PICs × 6 proteins × 6 replicates) was analyzed by linear discriminant analysis (LDA), which is a well-established chemometric method for multigroup classification.6,7 Stepwise analysis with different PIC set(s) (i.e., Jackknife classification procedure) was initially used to select an effective subset of 10


Page 11 of 22

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


sensors. Accuracies of 61% – 89% using one PIC were observed, while 10 of 56 combinations provided 100% accuracy (Table S-2). In particular, only the sensor consisting of PICs containing E1 (P1/E1), E2 (P1/E2), and E6 (P1/E6) achieved 100% accuracy for not only human plasma proteins (Table S-2) but other biological analytes (see below, Figure S-7). Higher turnover number of β-Gal among commercially available hydrolytic enzymes allowed the discrimination of 0.5 µg/mL plasma proteins ranging from 1.3 nM for FIB (Mw: 387 kDa) to 11.3 nM for ANT (Mw: 44 kDa), which were substantially lower concentrations than recently reported fluorometric array-based protein sensors (50 nM – 100 µM).13-16 Figure 4A shows the changes in enzyme activity for P1/E1, P1/E2, and P1/E6 upon addition of 0.5 µg/mL human plasma proteins. These data were converted to discriminant scores that were used to provide a graphical representation of LDA clustering in fingerprints (Figure 4B). Each dot represents the fingerprint of a single protein analyte to the PIC sensor (P1/E1, P1/E2, and P1/E6). Each cluster was successfully separated in three-dimensional space. The first discriminant score provided the best discrimination among the classes, which accounted for 92.9% of the total variance. Although driving forces involved in the formation of PICs were mainly electrostatic interactions, the first discriminant scores were not well correlated with isoelectric points (pI) of proteins (r = –0.605) (Figure S-8A). In addition, the mixing of FIB with P1/E6 unexpectedly decreased the enzyme activity (Figure 4A), indicating a strong interaction between FIB and E6 despite the negative charges on both FIB and E6. Taking into account the low correlation coefficient between the first discriminant scores and the surface hydrophobicity of proteins (Φsurface) (r = –0.134) (Figure S-8B), various characteristics of modified β-Gal surfaces played significant roles in PIC-protein interactions. For our previous PIC library, the responses of human plasma proteins were mainly attributed to electrostatic interactions.2,3 Therefore, the design of PIC libraries based on chemical modifications of an enzyme would be a simple and effective approach to generate diverse cross-reactivity toward human plasma proteins. These advantages may arise from (i) generation of unusual supercharging and aromatic moieties on the enzyme surface by a significant amount of lysine modification and (ii) the alteration of surface characteristics of β-Gal directly related to conversion of the binding event into a readable fluorescent signal through its own catalytic activity, as indicated in Figure 2D. We further examined the applicability of the constructed cross-reactive enzyme-based sensor consisting of P1/E1, P1/E2, and P1/E6 using other biological analytes, i.e., FIB and CRP in different concentrations, mammalian sera and cells. The FIB and CRP levels in patients are expected to be useful for the diagnosis of cardiovascular disease,43 disseminated intravascular coagulation,44 and chronic 11



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 12 of 22

obstructive pulmonary disease,45 and so we tested whether the cross-reactive enzyme-based sensor could quantify FIB and CRP (0.3 to 8.1 µg/mL). The fingerprints of changes in enzyme activity from different concentrations of FIB and CRP were found to be characteristic of each protein, and the responses increased differently with increasing concentrations (Figure S-9A). These fingerprints were subjected to further LDA analysis, and provided 100% accuracy with no overlapping of clusters (Figure 5A). Interestingly, LDA yielded a relatively smooth progression from 0.3 to 8.1 µg/mL; FIB clusters moved along the y-axis with increasing concentration, while movement along the x-axis was observed for CRP. Cross-reactive enzyme-based sensors also provide a versatile platform for identification of real biological analytes, such as mammalian sera5 and cells. Five sera from different mammalian sources (human, horse, rabbit, bovine, and fetal bovine) were selected to explore the discrimination of the first real biological analytes. The responses obtained by addition of 2.5 µg/mL sera were analyzed, and again, 100% correct classification was achieved (Figure 5B and S-9B). Similarly, our sensor identified three human cancer cell lines [A549 (lung), MG63 (bone), and HuH7 (liver)] and a human mesenchymal stem cell line (UE7T-13) with 100% accuracy (Figure 5C and S-9C) at levels as low as only 300 cells, which was more sensitive than recently proposed sensors (1000 – 5000 cells).10,16,23,24 Note that UE7T-13 was well separated from cancer cell lines, suggesting that our sensor may be able to recognize differences in the surface characteristics of different cell types. Early identification of cell types or states is required for the diagnosis and treatment of cancer.46 Currently available methods for identification of cell types are generally based on the specific interactions between antibodies and biomarkers present on the cell surface.47,48 Therefore, our results indicated the potential of our cross-reactive enzyme-based sensors allowing markerless characterization of cells for diagnosis. Preparation of artificially modified enzymes does not require laborious synthetic efforts, such as separation and purification. Therefore, the present strategy using PICs containing artificially modified enzymes would be useful to create a library of cross-reactive receptors appropriate for target biological analytes. In addition, amplification of the interactions between PICs and samples through enzyme reaction enables more sensitive identification than sensing systems based on fluorescent changes of synthetic fluorescent probes or their conjugates with quenchers.13,16,23,24 Furthermore, combination of the enzyme-based strategy with electrochemical devices49 is expected to lead to the development of more practical cross-reactive sensing systems.

Hierarchical clustering analysis for β-Gals/P1 system 12


Page 13 of 22

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


As described above, P1/E1, P1/E2, and P1/E6 were the best combinations to discriminate biological analytes (plasma proteins, FIB and CRP in different concentrations, mammalian sera and cells) used in this study (Figure S-7). To understand the structural origins, hierarchical clustering analysis (HCA), unsupervised method semi-quantitatively providing similarities or differences between samples based on the Euclidean distance,6,7 were employed (Figure 6). Changes in enzyme activities for 22 biological analytes were divided by the root mean square of corresponding PIC data for normalization to be able to compare artificially modified β-Gals with different catalytic properties. Interestingly, PICs were clustered according to the charge numbers of each lysine residue in β-Gals, indicating that differences in cross-reactivity were mainly responsible for the alteration of enzyme surface characteristics caused by changes in electrostatics. With regard to neutralized lysine cluster (P1/E2, P1/E3, and P1/E4), the variations in responses for P1/E3 and P1/E4 were higher than those for P1/E2, because of the lower kcat values (Figure 2D). As a consequence, P1/E2 may be more effective in neutralized lysine cluster to discriminate biological analytes. Although E5 and E6 showed similar high kcat values (Figure 2D), additional aromatic groups of E6 probably rendered good differentiability due to changes in cross-reactivity. LDA of the data obtained using six PICs supported the above interpretations; the distances between clusters in discriminant score plots except for mammalian cells, were not markedly increased (Figure S-10) compared to those with the three selected PICs (P1/E1, P1/E2, and P1/E6) (Figures 4B and 5), indicating that PICs categorized in the same cluster shown in Figure 6 had similar cross-reactivity toward the analytes used in this study. It should be noted that pI of plasma proteins showed low correlations with the first discriminant scores, as described above (Figure S-8A), even though the charge number of lysine side chains was important for discrimination of biological analytes, indicating that the charge conversion of lysine residues simultaneously affected various surface characteristics of enzymes, including hydrophobic and hydrogen bonding abilities.

Sensing of inflammatory biomarker proteins in human serum Sensing of inflammatory biomarker proteins in human serum is of critical importance for practical applications compared with that in simple buffer solutions (Figure 5A). It is estimated that >10000 proteins are commonly present in human serum with high overall content (60 – 80 mg/mL),51 generating a challenging complex matrix. To test our sensor array, physiologically relevant concentrations of inflammatory marker proteins (CRP and FIB) spiked into human serum were selected as analytes. However, P1 did not effectively inhibit β-Gals in the presence of human serum due to the presence of competing serum proteins. Therefore, we used more hydrophobic P2 instead of P1, which 13



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 14 of 22

was expected to strongly inhibit β-Gals in such complex solutions.3 Although the activity of 1.0 nM unmodified E1 was not decreased even in the presence of 300 nM P2 in 10 mM MOPS (pH 7.0) with 1% human serum, P2 inhibited the activity of E2-E6 to 70% – 85% (Figure S-11A). Using three selected PICs (100 nM P2 with 1.0 nM E3, E4, and E6), discrimination accuracy of 100% were achieved for CRP (2.0 to 8.0 µg/mL) and FIB (25 and 100 µg/mL) (Figures 7 and S-11B). It has been suggested that 14 µM (~5 mg/mL) FIB was should be used as a cutoff point for inflammatory diseases,44 and active inflammation and bacterial infection produce CPR levels of 40 to 200 µg/mL,52 which are comparable to the concentrations of stock sample solutions used in the present study (≥200 µg/mL CRP and ≥2.5 mg/mL FIB in human serum).

Conclusions In summary, we have developed a novel strategy for construction of cross-reactive enzyme-based sensors using PICs between cationic PEGylated polymers and artificially modified anionic enzymes. Chemical modifications of lysine ε-NH3+ groups of native β-Gal by acid anhydrides altered kinetic parameters of β-Gal (KM and kcat), whereas catalytic activity was retained. We employed the differential changes in enzyme activities of modified β-Gals obtained by the addition of each analyte protein or mammalian cell to the PICs between artificially modified β-Gals and PEGylated cationic polymers. The cross-reactive enzyme-based sensor was a versatile platform for highly sensitive identification of various biological analytes, i.e., human plasma proteins, different concentrations of inflammatory biomarker proteins, mammalian sera and cells in buffer and human serum. Charge conversion and introduction of hydrophobic groups into lysine side chains facilitated inhibition of catalytic activity through the formation of PICs. Therefore, enzymes that are not inhibited through PIC formation (Table S-1) are expected to be available for cross-reactive sensor elements. Taken together, this newly developed simple and inexpensive strategy for construction of cross-reactive enzyme-based sensors holds great promise for identification of dilute biofluids with closely related characteristics, such as secreted protein-containing cell culture medium.4 Further explorations of these applications are currently underway in our laboratory.

Acknowledgments We thank Prof. Dr Kentaro Shiraki (Faculty of Pure and Applied Sciences, University of Tsukuba) for the help with CD measurements. This work was supported by a Grant-in-Aid for JSPS Young Scientists (B, 26810074) 14


Page 15 of 22

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


Supporting Information Lists of anionic enzymes examined for sensing, profiles of biological analytes used in this study, statistical analysis, and additional characterization of β-Gals are available free of charge via the Internet at http://pubs.acs.org.

References ( 1)

Kirsch, J.; Siltanen, C.; Zhou, Q.; Revzin, A.; Simonian, A. Chem. Soc. Rev. 2013, 42, 8733-8768.

( 2)

Tomita, S.; Yoshimoto, K. Chem. Commun. 2013, 49, 10430-10432.

( 3)

Tomita, S.; Soejima, T.; Shiraki, K.; Yoshimoto, K. Analyst 2014, 139, 6100-6103.

( 4)

Tomita, S.; Sakao, M.; Kurita, R.; Niwa, O.; Yoshimoto, K. Chem. Sci. 2015, 6, 5831-5836.

( 5)

Tomita, S.; Yokoyama, S.; Kurita, R.; Niwa, O.; Yoshimoto, K. Anal. Sci. 2016, 32, 237-240.

( 6)

Anzenbacher, P. Jr; Lubal, P.; Bucek, P.; Palacios, M.A.; Kozelkova, M.E. Chem. Soc. Rev. 2010, 39, 3954-3979.

( 7)

Askim, J.R.; Mahmoudi, M.; Suslick, K.S. Chem. Soc. Rev. 2013, 42, 8649-8682.

( 8)

You, C.C.; Miranda, O.R.; Gider, B.; Ghosh, P.S.; Kim, I.B.; Erdogan, B.; Krovi, S.A.; Bunz, U.H.; Rotello, V.M. Nat. Nanotechnol. 2007, 2, 318-323.

( 9)

De, M.; Rana, S.; Akpinar, H.; Miranda, O.R.; Arvizo, R.R.; Bunz, U.H.F.; Rotello, V.M. Nat. Chem. 2009, 1, 461-465.

(10)

Pei, H.; Li, J.; Lv, M.; Wang, J.; Gao, J.; Lu, J.; Li, Y.; Huang, Q.; Hu, J.; Fan, C. J. Am. Chem. Soc. 2012, 134, 13843-13849.

(11)

Zamora-Olivares, D.; Kaoud, T.S.; Jose, J.; Ellington, A.; Dalby, K.N.; Anslyn, E.V. Angew. Chem. Int. Ed. 2014, 53, 14064-14068.

(12)

Xu, S.; Lu, X.; Yao, C.; Huang, F.; Jiang, H.; Hua, W.; Na, N.; Liu, H.; Ouyang, J. Anal. Chem. 2014, 86, 11634-11639.

(13)

Yuan, Z.; Du, Y.; Tseng, Y.T.; Peng, M.; Cai, N.; He, Y.; Chang, H.T.; Yeung, E.S. Anal. Chem. 2015, 87, 4253-4259.

(14)

Ran, X.; Pu, F.; Ren, J.; Qu, X. Chem. Commun. 2015, 51, 2675-2678.

(15)

Chen, S.; Wei, L.; Chen, X.W.; Wang, J.H. Anal. Chem. 2015, 87, 10902-10909.

(16)

Xu, Q.; Zhang, Y.; Tang, B.; Zhang, C.Y. Anal. Chem. 2016, 88, 2051-2058.

(17)

Phillips, R.L.; Miranda, O.R.; You, C.C.; Rotello, V.M.; Bunz, U.H. Angew. Chem. Int. Ed. 15



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 16 of 22

2008, 47, 2590-2594. (18)

Duarte, A.; Slutsky, M.; Hanrahan, G.; Mello, C.H.; Bazan, G.C. Chem. Eur. J. 2012, 18, 756-759.

(19)

Tao, Y.; Ran, X.; Ren, J.; Qu, X. Small 2014, 10, 1667-3671.

(20)

Wan, Y.; Sun, Y.; Qi, P.; Wang, P.; Zhang, D. Biosen. Bioelectron. 2014, 55, 289-293.

(21)

Bajaj, A.; Miranda, O.R.; Kim, I.B.; Phillips, R.L.; Jerry, D.J.; Bunz, U.H.; Rotello, V.M. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 10912-10916.

(22)

Bajaj, A.; Miranda, O.R.; Phillips, R.; Kim, I.B.; Jerry, D.J.; Bunz, U.H.; Rotello, V.M. J. Am. Chem. Soc. 2010, 132, 1018-1022.

(23)

Tao, Y.; Auguste, D.T. Biosens. Bioelectron. 2016, 81, 431-437.

(24)

Rana, S.; Elci, S.G.; Mout, R.; Singla, A.K.; Yazdani, M.; Bender, M.; Bajaj, A.; Saha, K.; Bunz, U.H.; Jirik, F.R.; Rotello, V.M. J. Am. Chem. Soc. 2016, 138, 4522-4529.

(25)

Miranda, O.R.; Chen, H.T.; You, C.C.; Mortenson, D.E.; Yang, X.C.; Bunz, U.H.; Rotello, V.M. J. Am. Chem. Soc. 2010, 132, 5285-5289.

(26)

Selvakumar, K.; Motiei, L.; Margulies, D. J. Am. Chem. Soc. 2015, 137, 4892-4895.

(27)

Gitlin, I.; Carbeck, J.D.; Whitesides, G.M. Angew. Chem. Int. Ed. 2006, 45, 3022-3060.

(28)

Shaw, B.F.; Moustakas, D.T.; Whitelegge, J.P.; Faull, K.F. Adv. Protein Chem. Struct. Biol. 2010, 79, 127-164.

(29)

Gitlin, I.; Gudiksen, K.L.; Whitesides, G.M. J. Phys. Chem. B 2006, 110, 2372-2377.

(30)

Shaw, B.F.; Schneider, G.F.; Bilgicer, B.; Kaufman, G.K.; Neveu, J.M.; Lane, W.S.; Whitelegge, J.P.; Whitesides, G.M. Protein Sci. 2008, 17, 1446-1455.

(31)

Goda, T.; Miyahara, Y. Anal. Chem. 2010, 82, 8946-8953.

(32)

Obermeyer, A.C.; Mills, C.E.; Dong, X.H.; Flores, R.J.; Olsen, B.D. Soft Matter 2016, 12, 3570-3581.

(33)

Lee, Y.; Ishii, T.; Cabral, H.; Kim, H.J.; Seo, J.H.; Nishiyama, N.; Oshima, H.; Osada, K.; Kataoka, K. Angew. Chem. Int. Ed. 2009, 48, 5309-5312.

(34)

Lee, Y.; Ishii, T.; Kim, H.J.; Nishiyama, N.; Hayakawa, Y.; Itaka, K.; Kataoka, K. Angew. Chem. Int. Ed. 2010, 49, 2552-2555.

(35)

Kim, A.; Miura, Y.; Ishii, T.; Mutaf, O.F.; Nishiyama, N.; Cabral, H.; Kataoka, K. Biomacromolecules 2016, 17, 446-453.

(36)

Gudiksen, K.L.; Gitlin, I.; Moustakas, D.T.; Whitesides, G.M. Biophys. J. 2006, 91, 298-310.

(37)

Lee, A.; Mirica, K.A.; Whitesides, G.M. J. Phys. Chem. B 2011, 115, 1199-1210. 16


Page 17 of 22

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


(38)

Pace, C.N.; Vajdos, F.; Fee, L.; Grimsley, G.; Gray, T. Protein Sci. 1995, 4, 2411-2423.

(39)

Husain, Q. Crit. Rev. Biotechnol. 2010, 30, 41-62.

(40)

MacMillan-Crow, L.A.; Crow, J.P.; Thompson, J.A. Biochemistry 1998, 37, 1613-1622.

(41)

Mahler, H.C.; Friess, W.; Grauschopf, U.; Kiese, S. J. Pharm. Sci. 2009, 98, 2909-2934.

(42)

Pepys, M.B.; Hirschfield, G.M. J. Clin. Invest. 2003, 111, 1805-1812.

(43)

Kaptoge, S.; Di Angelantonio, E.; Pennells, L.; Wood, A.M.; White, I.R.; Gao, P.; Walker, M.; Thompson, A.; Sarwar, N.; Caslake, M.; Butterworth, A.S.; Amouyel, P.; Assmann, G.; Bakker, S.J.; Barr, E.L.; Barrett-Connor, E.; Benjamin, E.J.; Bjorkelund, C.; Brenner, H.; Brunner, E.; Clarke, R.; Cooper, J.A.; Cremer, P.; Cushman, M.; Dagenais, G.R.; D'Agostino, R.B. Sr; Dankner, R.; Davey-Smith, G.; Deeg, D.; Dekker, J.M.; Engstrom, G.; Folsom, A.R.; Fowkes, F.G.; Gallacher, J.; Gaziano, J.M.; Giampaoli, S.; Gillum, R.F.; Hofman, A.; Howard, B.V.; Ingelsson, E.; Iso, H.; Jorgensen, T.; Kiechl, S.; Kitamura, A.; Kiyohara, Y.; Koenig, W.; Kromhout, D.; Kuller, L.H.; Lawlor, D.A.; Meade, T.W.; Nissinen, A.; Nordestgaard, B.G.; Onat, A.; Panagiotakos, D.B.; Psaty, B.M.; Rodriguez, B.; Rosengren, A.; Salomaa, V.; Kauhanen, J.; Salonen, J.T.; Shaffer, J.A.; Shea, S.; Ford, I.; Stehouwer, C.D.; Strandberg, T.E.; Tipping, R.W.; Tosetto, A.; Wassertheil-Smoller, S.; Wennberg, P.; Westendorp, R.G.; Whincup, P.H.; Wilhelmsen, L.; Woodward, M.; Lowe, G.D.; Wareham, N.J.; Khaw, K.T.; Sattar, N.; Packard, C.J.; Gudnason, V.; Ridker, P.M.; Pepys, M.B.; Thompson, S.G.; Danesh, N. Engl. J. Med. 2012, 367, 1310-1320.

(44)

Windsperger, K.; Lehner, R. Am. J. Obstet. Gynecol. 2013, 208, 118.e1-118.e7.

(45)

Thomsen, M.; Ingebrigtsen, T.S.; Marott, J.L.; Dahl, M.; Lange, P.; Vestbo, J.; Nordestgaard, B.G. J. Am. Med. Assoc. 2013, 309, 2353-2361.

(46)

Pantel, K.; Brakenhoff, R.H.; Brandt, B. Nat. Rev. Cancer 2008, 8, 329-340.

(47)

Tol, J.; Punt, C.J. Clin. Thir. 2010, 32, 437-453.

(48)

Alix-Panabières, C.; Pantel, K. Lab Chip 2014, 14, 57-62.

(49)

Dykstra, P.H.; Roy, V.; Byrd, C.; Bentley, W.E.; Ghodssi, R. Anal. Chem. 2012, 83, 5920-5927.

(50)

Adkins, J.N.; Varnum, S.M.; Auberry, K.J.; Moore, R.J.; Angell, N.H.; Smith, R.D.; Springer, D.L.; Pounds, J.G. Mol. Cell Proteomics 2002, 1, 947-955.

(51)

Clyne, B.; Olshaker, J.S. J. Emerg. Med. 1999, 17, 1019-1025.

(52)

Tetko, I.V.; Gasteiger, J.; Todeschini, R.; Mauri, A.; Livingstone, D.; Ertl, P.; Palyulin, V.A.; Radchenko, E.V.; Zefirov, N.S.; Makarenko, A.S.; Tanchuk, V.Y.; Prokopenko, V.V. J. Comput. Aid. Mol. Des. 2005, 19, 453-463. 17



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 18 of 22

Figure 1. A strategy for construction of cross-reactive enzyme-based sensors for discrimination of biological analytes. Lysine ε-NH3+ groups in β-Gal were artificially modified with acid anhydrides in aqueous solutions, followed by PIC formation with PEG-b-QPAMAs, and then PICs were mixed with biological analytes to obtain unique fingerprints of changes in enzyme activity for each analyte. Log P values of lysine side chains shown in parentheses were calculated using the program ALOGPs.52

Figure 2. Kinetic parameters of β-Gals modified with various acid anhydrides. (A) The Michaelis constant (KM) and (B) turnover number (kcat) of artificially modified β-Gals. Values are shown as means ± S.E. (n = 3, *p < 0.05, **p < 0.01, ***p < 0.001 against E1).

18


Page 19 of 22

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


Figure 3. Normalized activities of artificially modified β-Gals in the presence of P1. Various concentrations of P1 were added to solutions containing 0.5 nM β-Gals in 10 mM MOPS (pH 7.0). Values are shown as means ± S.E. (n = 3).

Figure 4. Sensing of 0.5 µg/mL human plasma proteins using three PICs (P1/E1, P1/E2, and P1/E6). (A) Fingerprints of changes in enzyme activity for six human plasma proteins. Values are shown as means ± S.D. (n = 6). pI and Φsurface values of plasma proteins are shown. The values were divided by the root mean square of corresponding PIC data to facilitate visual comparison between each of the PIC results. (B) Discriminant score plot. The centers of points classified in identical proteins were connected to facilitate a visual comparison between clusters. 19



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 20 of 22

Figure 5. Discriminant score plot for (A) the quantitative response of FIB and CRP (0.3 – 8.1 µg/mL), (B) mammalian sera containing 2.5 µg/mL serum proteins, and (C) 1500 cells/mL mammalian cells using three PICs (P1/E1, P1/E2, and P1/E6). The ellipses represent confidence intervals (± 1 S.D.) for the individual analytes.

Figure 6. Clustering analysis of responses obtained with cross-reactive enzyme-based sensors. Hierarchical clustering dendrogram based on the Euclidean distance along with Ward method was created using the data set (6 PICs × 22 biological analytes × 6 replicates) divided by the root mean square of corresponding PIC data.

20


Page 21 of 22

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


Figure 7. Discriminant score plot for CRP (2.0 – 8.0 µg/mL) and FIB (25 and 100 µg/mL) in 10 mM MOPS (pH 7.0) with 1% human serum using three PICs (P2/E3, P2/E4, and P2/E6). The ellipses represent confidence intervals (± 1 S.D.) for the individual analytes.

21



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 22 of 22

For TOC only

22


Artificial Modification of an Enzyme for ... - ACS Publications

Recommend Documents