Mitochondrion-Targeting, Environment-Sensitive Red Fluorescent

Oct 23, 2017 - Mitochondrial vicinal dithiol-containing proteins (VDPs) are key regulators in cellular redox homeostasis and useful markers for diagno...
0 downloads 11 Views 2MB Size
Download PDF
Letter Cite This: Anal. Chem. 2017, 89, 11203-11207

pubs.acs.org/ac

Mitochondrion-Targeting, Environment-Sensitive Red Fluorescent Probe for Highly Sensitive Detection and Imaging of Vicinal DithiolContaining Proteins Feng Liu, Hai-Juan Liu, Xian-Jun Liu, Wen Chen, Fenglin Wang,* Ru-Qin Yu, and Jian-Hui Jiang* Institute of Chemical Biology and Nanomedicine, State Key Laboratory of Chemo/Bio-Sensing and Chemometrics, College of Chemistry and Chemical Engineering, Hunan University, Changsha 410082, P. R. China S Supporting Information *

ABSTRACT: Mitochondrial vicinal dithiol-containing proteins (VDPs) are key regulators in cellular redox homeostasis and useful markers for diagnostics of redox-dependent diseases. Current probes fail to target mitochondrial VDPs and show limited sensitivity and response rate. We develop a novel fluorescent probe using an engineered benzoxadiazole fluorophore that allows selective targeting of mitochondria and exhibits highly sensitive environment responsiveness. This probe is almost nonfluorescent in aqueous media, while delivering intense fluorescence upon binding to VDPs via a cyclic dithiaarsane ligand. The fluorescence probe is shown to have rapid response within 30 s and high sensitivity for detecting reduced bovine serum albumin (rBSA) in the concentration range from 0 to 0.1 μM with a detection limit of 2 nM. To our knowledge, this is the first fluorescence probe for VDPs which exhibits deep red emission, instantaneous response, high turn-on fluorescence ratio, and specific mitochondrial localization. It may provide a new tool for in situ monitoring mitochondrial VDPs.

M

and precludes real-time monitoring of intracellular VDPs. An activatable design that delivers “turn-on” fluorescence on binding VDPs is to use the cyclic dithiaarsane as a linker for a pair of fluorescence resonant energy transfer (FRET) donor and acceptor.14 This design allows real-time and wash-free monitoring of VDPs, however, and still suffers from complexity in synthesis and relatively high background due to limited quenching efficiency of FRET pairs. Environmental sensitive dyes, which change their fluorescent properties in response to microenvironmental variations such as viscosity and polarity15 afford a promising platform for direct design of activatable probes of VDPs. Because environmental sensitive dyes are able to show very high signal-to-background ratio on binding to the proteins, this design offers the possibility of detecting and imaging VDPs with better sensitivity and improved contrast.16,17 Nevertheless, no fluorescent probes have been reported for specifically detection and imaging of VDPs in mitochondria. Herein, we report the first demonstration of a mitochondrion-targeting and activatable fluorescent probe for highly sensitive and very rapid detection and imaging of mitochondrial VDPs in living cells, as illustrated in Scheme 1. In order to develop new environmental sensitive dyes with mitochondriontargeting ability and highly sensitive environment responsiveness, we start to engineer nitrobenzoxadiazole (NBD), a well-

itochondria play essential roles in many fundamental cellular events, including fatty acid synthesis, Ca2+ homeostasis, and apoptotic signaling. As the major source of ROS, mitochondria are particularly susceptible to oxidative damage, and tight regulations in the redox-dependent processes are central to the organelles’ functions.1,2 Aberrant redox metabolism is one of the major causes for many diseases such as neurodegenerative diseases,3 diabetes,4 and cancer.5 Vicinal dithiol-containing proteins (VDPs) are key regulators in mitochondrial redox homeostasis through their interactions with reactive oxygen and nitrogen species and electrophiles.6 Reversible interconversion of the protein vicinal dithiols with their oxidized form represents a crucial mechanism to regulate the functions of VDPs and control the mitochondrial metabolic flux for redox signaling.7 The protein vicinal dithiols are also critical linking centers for stabilization of protein structures and have close association with protein functions.8,9 Therefore, detection and tracking of the VDPs, especially in mitochondria, are of great significance for diagnostics of redox-related diseases and elucidation of their pathophysiology.10 Fluorescent probes provide a useful tool for detection and tracking of VDPs in living systems. A general strategy of these probes is to use trivalent arsenical moieties that can bind with vicinal dithiols with high specificity over other monothiols.11 A straightforward design is to conjugate an “always-on” fluorophore to a cyclic dithiaarsane binding group for direct visualization of VDPs in living cells.12,13 A possible issue with this design is the background fluorescence from unreacted or nonspecifically bound probes, which requires tedious washing © 2017 American Chemical Society

Received: September 12, 2017 Accepted: October 23, 2017 Published: October 23, 2017 11203

DOI: 10.1021/acs.analchem.7b03724 Anal. Chem. 2017, 89, 11203−11207

Letter

Analytical Chemistry

the optimal responsiveness, we synthesized three BD fluorophores, BDI, BDB, and BDP by conjugating BD with indolium, benzindolium, and pyridinium, respectively (Scheme S1 in the Supporting Information). Their structures were characterized with 1H NMR, 13C NMR, and MS, respectively (Figures S1−S9 in the Supporting Information). Their responsiveness to environmental viscosity and polarity were investigated with polarity-modulated mixtures of water and 1,4dioxane and viscosity-modulated mixtures of water and glycerol, respectively. As the orientation polarizability (Δf) of the solution decreased from 0.32 (100% water) to 0.292 (20% water),19 the fluorescence intensity of BDI at 630 nm increased by a factor of 33, which indicates that it is a highly polaritysensitive fluorescent dye (Figure 1a,b). BDB and BDP also

Scheme 1. (a) Structures of Donor−acceptor Environmental Sensitive Dyes and (b) Illustration of Fluorescence Turn-On Mechanism for BDI Probe in Detecting Mitochondrial VDPs

known environment-sensitive solvatochromic fluorophore with green emission.18 We hypothesize that extension of the conjugated system in NBD by substituting other electronwithdrawing moieties for the nitro group can render new fluorophores undergoing intramolecular charge transfer (ICT) at the excited state. Motivated by this hypothesis, we design benzoxadiazole (BD) derivatives with three different conjugated cationic electron-withdrawing moieties, indolium, benzindolium, and pyridinium. These new fluorophores have an electron donor and acceptor structure, which can create a highly dipolar and twisted excited state with sensitive fluorescent responsiveness to microenvironmental viscosity and polarity.15,18 Moreover, the conjugated cationic hydrophobic electron-withdrawing moieties allow the probes to exhibit specific mitochondrion-targeting ability and red-shifted emission with low autofluorescence background. We find that the BD derivatives all display desirably high signal-to-background ratio in detecting microenvironmental viscosity and polarity variations, and indolium conjugated BD (BDI) shows better responsiveness than benzindolium conjugated BD (BDB) and pyridinium conjugated BD (BDP). On the basis of the better responsiveness of BDI, a novel fluorescence probe for mitochondrial VDPs is designed by further derivation of BDI with a cyclic dithiaarsane ligand. In the absence of VDPs, the BDI probe merely delivers very weak fluorescence. Upon binding of VDPs in mitochondria, the BDI fluorophore is brought into a hydrophobic protein pocket, activating intense fluorescence as the tracer for mitochondrial VDPs. To our knowledge, this is the first activatable fluorescence probe that enables selective targeting and imaging of mitochondrial VDPs. Because the BDI fluorophore exhibits very high signal-tobackground ratio in detecting microenvironmental viscosity and polarity variations, this probe may provide a highly sensitive tool for in situ monitoring mitochondrial VDPs. The rational behind environmental sensitive probes for VDPs is that vicinal dithiols are present in hydrophobic pockets of the protein, which are characterized by decreased microenvironmental polarity and increased viscosity.16,17 Hence, environmental sensitive dyes with higher responsiveness to polarity and viscosity typically imply improved sensitivity for the VDP probes. To obtain an environmental sensitive dye with

Figure 1. (a) Fluorescence spectra of BDI (1.0 μM) in water/1,4dioxane solvent mixtures with different fractions of 1,4-dioxane (fd); (b) fluorescence enhancement ratio (F/F0) of BDI, BDB, and BDP (1.0 μM) in water/1,4-dioxane solvent mixtures with different fractions of 1,4-dioxane (fd); (c) fluorescence spectra of BDI in water/glycerol solvent mixtures with different fractions of glycerol ( fd); (d) F/F0 of BDI, BDB, and BDP in water/glycerol solvent mixtures with different fractions of glycerol ( fd).

exhibited similar trends and their fluorescence intensities at 624 and 594 nm increased by a factor of 10 and 12, respectively (Figure 1b and Figure S10 in the Supporting Information). Furthermore, BDI, BDB, and BDP are also responsive to solvent viscosity. Their maximal intensities increased gradually as the viscosity increases from 0.894 cP (water) to 950 cP (99% glycerol) at 25 °C,20 with BDI presented the largest enhancement fact of 35 (Figure 1c,d and Figure S10 in the Supporting Information). We chose BDI for further studies as it exhibited the greatest environment sensitivity. BDI was further derived with a carboxylic linker for functionalization. 2-(4-Aminophenyl)-1,3,2-dithiarsolane (PAO-EDT) was chosen as the recognition unit for VDPs as it can selectively differentiate vicinal dithiols from other forms of thiols by exchanging 1,2-ethanedithiol in cyclic dithiaarsanes with the vicinal dithiols in proteins.21 BDI probe was obtained by conjugating PAO-EDT with the BDI derivative via a flexible linker (three methylene groups) to minimize steric hindrance.22 Control probe without the cyclic dithiaarsane group was also synthesized. The structures of the BDI probe and the Control probe were also confirmed with 1H NMR, 13C NMR, and MS, respectively (Figures S11−S16 in the Supporting Information). Both BDI probe and Control probe are still responsive to polarity variation (Figure S17 in the Supporting Information). 11204

DOI: 10.1021/acs.analchem.7b03724 Anal. Chem. 2017, 89, 11203−11207

Letter

Analytical Chemistry

increase in fluorescence was attributed to the hydrophobic domain of rBSA, we denatured rBSA with guanidine hydrochloride (GndHCl), which can induce conformation changes and disrupt hydrophobic pocket of rBSA.23 The fluorescence intensity was significantly decreased upon addition of GndHCl (Figure S21 in the Supporting Information), which indicates that the fluorescence emission was due to the hydrophobic pocket of rBSA. The fluorescence was not completely diminished, possibly because only a proportion of rBSA was denatured.24 These results demonstrate that our probe can specifically react with VDPs and turn on its fluorescence due to binding induced microenvironment variation. The robustness and selectivity BDI probe were then investigated. The response of BDI probe was investigated in the presence of cysteine, GSH, Hcy, and ascorbic acid. Its fluorescence response did not exhibit noticeable differences (Figure S22 in the Supporting Information). These results indicate that our probe is resistant to reducing agents and have a much higher affinity toward vicinal dithiols as compared to monothiols. Furthermore, its selectivity toward different amino acids (Hcy, Tyr, Thr, Arg, Glu, Lys, Met, Ser), GSH, ascorbic acid, and proteins (lysozome, ovabumin, pepsin, trypsin, BSA, HSA) were also studied. There was minor fluorescence increase when BDI probe was added to the above substances (Figure S23 in the Supporting Information). The negligible fluorescence enhancement was ascribed to that the cyclic dithiaarsane ring was highly specific toward dithiols, and there were no vicinal dithiols on the tested substances.16,17 In contrast, the fluorescence increased dramatically upon the addition of rBSA. These results further confirm that BDI probe is highly robust and specific to VDPs. The response time of BDI probe toward rBSA was studied by measuring fluorescence intensity at 630 nm in real time. Figure 2c shows that the fluorescence signal increases dramatically and reaches a plateau within 30 s upon rBSA addition. The reaction rate was calculated to be 0.028 s−1 (Figure S24 in the Supporting Information). In contrast, the free probe had a negligible background signal with no noticeable variation under the same conditions (Figure 2c). Compared with all the previously reported probes for VDPs, our probe affords the most rapid response toward rBSA, which might be attributed to the use of a flexible methylene linker minimizing the steric hindrance and facilitating a faster reaction. The BDI probe also afforded high sensitivity toward rBSA. With increasing concentrations of rBSA, the fluorescence signals gradually increased and reached a plateau at 2.5 μM (Figure 2d). There was a 35-fold intensity increase as the concentrations of rBSA varied from 0 to 2.5 μM. The calibration curve shows that the emission intensity at 630 nm was proportional to rBSA concentration up to 0.1 μM with a detection limit of 2 nM (Figure S25 in the Supporting Information). The absorption spectrum of BDI probe slightly increased upon incremental addition of rBSA with a slight red shift (Figure S26 in the Supporting Information), which is consistent with the solvatochromic properties of environment sensitive dyes.15 These results demonstrate that BDI probe can detect rBSA quantitatively, which could correlate with the concentrations of VDPs. The ability of BDI probe for endogenous visualization of VDPs in living cells was investigated, using MCF-7 cells as a model cell line. First, the cytotoxicity of the BDI probe was evaluated with a standard MTT assay. The cells had over 85% viability for concentrations of the BDI probe up to 20 μM,

The responsiveness of BDI probe toward VDPs was then investigated using reduced BSA (rBSA) as the model protein. rBSA has eight vicinal cysteine pairs, while BSA and oxidized BSA (oBSA) do not have vicinal thiols.14,17 The fluorescence signal of BDI probe was quite low in the absence of rBSA or in the presence of oxidized BSA (oBSA), and it was remarkably increased with a maximum emission wavelength of 630 nm in the presence of rBSA (Figure 2a). The fluorescence increase

Figure 2. (a) Fluorescence spectra of BDI probe (1.0 μM) toward rBSA (2.5 μM), oBSA (2.5 μM), and alone. (b) Specificity of BDI probe to VDPs confirmed by SDS-PAGE. (c) Time course of the fluorescence intensity of BDI probe (1.0 μM) with and without rBSA. (d) Fluorescence response of BDI probe (1.0 μM) to different concentrations of rBSA.

could be reversed with the addition of 1,2-ethanedithiol (EDT) (Figure S18 in the Supporting Information). Meanwhile, there was negligible fluorescence signal when the Control probe was treated with rBSA (Figure S19 in the Supporting Information). These results indicate that the fluorescence enhancement was due to the selective binding of vicinal dithiols on rBSA to the trivalent arsenical of BDI probe, which brought the probe from polar aqueous media into the hydrophobic pocket of protein. The absence of dithiols on oBSA could not react with BDI probe; hence, no fluorescence was enhanced. These results were further confirmed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). An intense red fluorescent band was observed in the lane loaded with rBSA and BDI probe, whereas no visible emission was observed in the lanes loaded with BDI probe and BSA or oBSA because of the absence of vicinal dithiols. There was also no fluorescence in the lane loaded with rBSA and the Control probe without the cyclic dithiaarsane ring (Figure 2b). Coomassie brilliant blue (CBB) staining confirms that the fluorescence was due to the formation of rBSA-BDI probe complex. These results further implied that the cyclic dithiaarsane ring in BDI probe could specifically react with the vicinal dithiols and bring the fluorophore into the hydrophobic pocket of BSA. Furthermore, the response of BDI probe toward other reduced forms of proteins such as human serum albumin (HSA), ovalbumin, lysozyme, pepsin, and trypsin were also studied. The reduced HSA (rHSA) displayed a remarkable increase in fluorescence as it is similar to that of rBSA, while the other reduced forms of proteins exhibited negligible fluorescence increase, due to the lack of vicinal dithiols (Figure S20 in the Supporting Information). In addition, to further demonstrate that the 11205

DOI: 10.1021/acs.analchem.7b03724 Anal. Chem. 2017, 89, 11203−11207

Letter

Analytical Chemistry which indicates that our probe is not toxic to MCF-7 cells (Figure S27 in the Supporting Information). Then, 1 μM of BDI probe was incubated with MCF-7 cells for 20 min in PBS, containing 0.2% DMSO as a cosolvent. Strong red fluorescence could be seen from the cells incubated with BDI probe as shown more clearly in the merged image (Figure 3a1−a2). By

Figure 3. Confocal fluorescence imaging of cellular VDPs in MCF-7 cells: (a1−a2) cells treated with BDI probe (1.0 μM); (b1−b2) cells treated with Control probe (1.0 μM); (c1−c2) cells pretreated with DTT (10 mM) followed by incubation with BDI probe; (d1−d2) cells pretreated with PAO (30 μM) followed by incubation with BDI probe. Scale bars: 20 μm.

contrast, cells incubated with the Control probe did not display any fluorescence signals (Figure 3b1−b2). The above results indicate that BDI probe could selectively react with endogenous VDPs in living cells via the cyclic dithiaarsane ligand, subsequently turn on its fluorescence emission. Furthermore, the cells were pretreated with dithiothreitol (DTT) or p-aminophenylarsenoxide (PAO), then incubated with BDI probe. Fluorescence images were obtained under identical conditions. There was an obvious increase in fluorescence for the DTT pretreated cells, whereas the fluorescence for PAO pretreated cells was decreased (Figure 3c1−c2, d1−d2 and Figures S28 and S29 in the Supporting Information). This is because DTT is a reductant which can increase the amount of endogenous VDPs, while PAO can specifically react with the vicinal dithiols and prevent their reaction with BDI probe. These results demonstrate that BDI probe can specifically detect endogenous VDPs in living cells. To test that our probe is capable of targeting and detecting mitochondrial VDPs, we further investigated the subcellular localization of BDI probe labeled VDPs in MCF-7 cells by fluorescence colocalization assay. Mito-Tracker green, Hoechst 33342 blue and Lyso-Tracker green were introduced to stain the mitochondria, nucleus, and lysosomes, respectively. MCF-7 cells were pretreated with BDI probe and then incubated with Mito-Tracker green, Hoechst 33342 blue, and Lyso-Tracker green, respectively. Fluorescence images were obtained with the confocal microscope using different excitation wavelengths and collection channels as specified in the Supporting Information. Fluorescence signals from BDI probe and Mito-Tracker green overlap well with a correlation coefficient 0.92, which is more clearly revealed in the line profile analysis (Figure 4a1−d1 and Figure S30 in the Supporting Information). However, the fluorescence signal of BDI probe does not synchronize with those of Lyso-Tracker green or Hoechst 33342 blue (Figure 4a2−d2, a3−d3) with correlation coefficients of 0.26 and 0.21, respectively. These results demonstrate that BDI probe is localized in mitochondria and specifically labels the mitochon-

Figure 4. Intracellular localization of BDI probe in MCF-7 cells. (a1, b1, c1) Cells were pretreated with BDI probe (1.0 μM) followed by Mito-Tracker Green (1.0 μM). (a2, b2, c2) Cells were pretreated with BDI probe (1.0 μM) followed by Lyso-Tracker Green (0.05 μM). (a3, b3, c3) Cells were pretreated with BDI probe (1.0 μM) followed by Hoechst 33342 (0.5 μg mL−1). Scale bars: 10 μm. (d1, d2, d3) Fluorescence intensity variation of the white line across the cells in the merged images.

drial VDPs. Therefore, BDI probe has the potential to detect and image mitochondrial VDPs. In conclusion, the BDI probe has been developed as a turnon fluorescent probe for specific detection of VDPs, using BD as the fluorescent reporter, 1,3,2-dithiarsolane group as the recognition ligand, and indolium as the mitochondrial targeting unit. The probe has negligible fluorescence in its free form but exhibits strong fluorescence intensity upon reaction with VDPs. The probe exhibits high sensitivity, instantaneous response, high selectivity, and high specificity toward VDPs. It is demonstrated that BDI probe is mitochondria specific and can be specifically applied for detecting mitochondrial VDPs. This is the first VDPs probe that possesses excellent analytical merits and mitochondrial specificity.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.7b03724. Experimental methods including synthetic procedures, compound characterizations, fluorescence measurements, cell culture, and fluorescence imaging as well as 1H NMR, 13C NMR, and HRMS (ESI) spectra (PDF) 11206

DOI: 10.1021/acs.analchem.7b03724 Anal. Chem. 2017, 89, 11203−11207

Letter

Analytical Chemistry

■

AUTHOR INFORMATION

Corresponding Authors

*Fax: +86-731-88821916. E-mail: [email protected]. *E-mail: [email protected]. ORCID

Ru-Qin Yu: 0000-0002-7412-8360 Jian-Hui Jiang: 0000-0003-1594-4023 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by NSFC (Grants 21527810, 21205034, 21521063) and the Scientific Research Funding from Hunan University (Grant 531107050917).

■

REFERENCES

(1) Handy, D.; Loscalzo, J. Antioxid. Redox Signaling 2012, 16, 1323− 1367. (2) Kang, J.; Pervaiz, S. Biochem. Res. Int. 2012, 2012, 1. (3) Nakamura, T.; Lipton, S. Cell Death Differ. 2011, 18, 1478. (4) Houstis, N.; Rosen, E. D.; Lander, E. S. Nature 2006, 440, 944− 948. (5) Valko, M.; Rhodes, C.; Moncol, J.; Izakovic, M. M.; Mazur, M. Chem.-Biol. Interact. 2006, 160, 1−40. (6) Willems, P. H.; Rossignol, R.; Dieteren, C. E.; Murphy, M. P.; Koopman, W. J. Cell Metab. 2015, 22, 207−218. (7) Adler, V.; Yin, Z.; Tew, K. D.; Ronai, Z. E. Oncogene 1999, 18, 6104−6111. (8) Cook, K. M.; Hogg, P. J. Antioxid. Redox Signaling 2013, 18, 1987−2015. (9) Borges, C. R.; Lake, D. F. Antioxid. Redox Signaling 2014, 21, 392−395. (10) Bachi, A.; Dalle-Donne, I.; Scaloni, A. Chem. Rev. 2013, 113, 596−698. (11) Requejo, R.; Chouchani, E. T.; James, A. M.; Prime, T. A.; Lilley, K. S.; Fearnley, I. M.; Murphy, M. P. Arch. Biochem. Biophys. 2010, 504, 228−235. (12) Huang, C.; Yin, Q.; Zhu, W. P.; Yang, Y.; Wang, X.; Qian, X. H.; Xu, Y. Angew. Chem., Int. Ed. 2011, 50, 7551−7556. (13) Huang, C.; Yin, Q.; Meng, J.; Zhu, W. P.; Yang, Y.; Qian, X. H.; Xu, Y. F. Chem. - Eur. J. 2013, 19, 7739−7747. (14) Huang, C.; Jia, T.; Tang, M. F.; Yin, Q.; Zhu, W. P.; Zhang, C.; Yang, Y.; Jia, N. Q.; Xu, Y. F.; Qian, X. H. J. Am. Chem. Soc. 2014, 136, 14237−14244. (15) Klymchenko, A. Acc. Chem. Res. 2017, 50, 366−375. (16) Wang, Y. Y.; Zhong, Y. G.; Wang, Q.; Yang, X. F.; Li, Z.; Li, H. Anal. Chem. 2016, 88, 10237−10244. (17) Wang, Y. Y.; Yang, X. F.; Zhong, Y. G.; Gong, X. Y.; Li, Z.; Li, H. Chem. Sci. 2016, 7, 518−524. (18) Loving, G.; Sainlos, M.; Imperiali, B. Trends Biotechnol. 2010, 28, 73−83. (19) Jiang, N.; Fan, J. L.; Xu, F.; Peng, X. J.; Mu, H. Y.; Wang, J. Y.; Xiong, X. Q. Angew. Chem., Int. Ed. 2015, 54, 2510−2514. (20) Zhang, G.; Sun, Y.; He, X. Q.; Zhang, W. J.; Tian, M. G.; Feng, R. Q.; Zhang, R. Y.; Li, X. C.; Guo, L. F.; Yu, X. Q.; Zhang, S. L. Anal. Chem. 2015, 87, 12088−12095. (21) Griffin, B.; Adams, S.; Tsien, R. Science 1998, 281, 269−272. (22) Zhang, J.; Campbell, R.; Ting, A. Y.; Tsien, R. Nat. Rev. Mol. Cell Biol. 2002, 3, 906. (23) Zhdanova, N. G.; Maksimov, E. G.; Arutyunyan, A. M.; Fadeev, V. V.; Shirshin, E. A. Spectrochim. Acta, Part A 2017, 174, 223−229. (24) Hong, Y.; Feng, C.; Yu, Y.; Liu, J.; Lam, J. W. Y.; Luo, K. Q.; Tang, B. Z. Anal. Chem. 2010, 82, 7035−7043.

11207

DOI: 10.1021/acs.analchem.7b03724 Anal. Chem. 2017, 89, 11203−11207