Procyanidin C1 | Acetyl CoA Carboxyla Signal

Supramolecular Sensor for Astringent Procyanidin C1: Fluorescent Artificial Tongue for Wine Components

Authors: Yui Sasaki, Satoshi Ito, Zhoujie Zhang, Xiaojun Lyu, Shin-ya Takizawa, Riku Kubota, and Tsuyoshi Minami

This manuscript has been accepted after peer review and appears as an Accepted Article online prior to editing, proofing, and formal publication of the final Version of Record (VoR). This work is currently citable by using the Digital Object Identifier (DOI) given below. The VoR will be published online in Early View as soon as possible and may be different to this Accepted Article as a result of editing. Readers should obtain the VoR from the journal website shown below when it is published to ensure accuracy of information. The authors are responsible for the content of this Accepted

Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202002262
Link to VoR: https://doi.org/10.1002/chem.202002262
Chemistry – A European Journal 10.1002/chem.202002262

COMMUNICATION

Supramolecular Sensor for Astringent Procyanidin C1: Fluorescent Artificial Tongue for Wine Components
Yui Sasaki,[a] Satoshi Ito,[a, b] Zhoujie Zhang,[a] Xiaojun Lyu,[a] Shin-ya Takizawa,[c] Riku Kubota[a] and Tsuyoshi Minami*[a]

Abstract: An artificial tongue which detects astringent components for comprehensive evaluation of taste has not been established to date. Herein, we first propose fluorescent polythiophene (PT) derivatives (S1‒S3) modified with 3-pyridinum boronic acid as supramolecular chemosensors for wine components including astringent procyanidin C1. After numerous attempts for the synthetic conditions, more than 95 mol% of the PT unit was modified with the pyridinium boronic acid moiety. To evaluate the PT derivatives as chemosensors of the artificial tongue, qualitative and quantitative analyses were performed with four types of wine components (i.e. sweet, sour, bitter, and astringent tastes) in combination with pattern recognition models. Notably, procyanidin C1 in actual wine sample was successfully detected in a quantitative manner. In other words, we have established the authentic artificial tongue by the PT based supramolecular chemosensors.

In Nature, the taste sensory system allows to recognize sweetness, saltiness, sourness, and bitterness in foods and drinks. The taste components are detected by the corresponding receptors and thus we can discriminate the taste based on pattern recognition of our tongue.[1] An artificial tongue is inspired by such mammalian system, which has promoted the progress of food analyses.[2] Indeed, various types of artificial tongues based on electronic devices[3] or chemosensor array systems[4] have been developed as taste sensors for drinks (e.g. coffee,[3c] juice,[3d],[4j] milk,[3e] whisky,[4c], beer,[4d] mineral water,[4e] tea,[4g] olive oil,[4i] and wine[3f],[4f],[4h]). Interestingly, we can also recognize the astringency of drinks even though the receptors for astringent components do not exist in the mammalian system.[5] For example, tastes of wine are classified in sweetness, sourness, bitterness, and astringent.[6] Inspired by the mammalian system for taste sensing, artificial tongues for wine analysis has been reported. Gutiérrez et al. developed an ion-sensitive field-effect transistor (ISFET)-type wine sensor.[3f] However, the parameters in the ISFET sensor for discrimination of wines were pH and mineral nutrients, meaning that the actual components of wine tastes were not included because of the limited ability of the ion-sensitive membrane. To increase the number of detectable wine components, Anslyn et al.reported a colorimetric chemosensor array consisting of a peptide-metal-indicator for flavonoids.[4f] Bunz et al. developed a fluorescence chemosensor array utilizing poly(p- phenyleneethynylene) for discrimination of alcohols, sweetness (i.e. saccharides), sourness (i.e. acids and vitamin), and mineral nutrients.[4h] These molecular sensor array systems achieved actual wine discrimination according to host-guest interactions, while the comprehensive sensing of primary wine components is still a challenging issue. Especially, the detectability of artificial tongues for the astringents such as procyanidin C1 has not been reported. Procyanidins are classified in polyphenols and contained in a variety of wines.[6] Herein, we designed and synthesized the first artificial receptor for detection of astringent procyanidin C1[7] in wine.

Chemical structures of polythiophene-based chemosensors and target wine components.

Polythiophene (PT) is a representative π-conjugated polymer backbone as a chemosensor,[8] while PT-based chemosensor arrays are still rare.[9] PT derivatives modified with artificial receptors offer not only multi-binding sites in the single chemosensor but efficient amplification of the molecular recognition information based on molecular wire effects.[10] In this regard, the PT derivatives were utilized only for the qualitative detection and one analyte (i.e. saccharides[11] or catechin[12]) per one chemosensor. However, for the comprehensive taste evaluation, a variety of the contributing components such as astringent (procyanidin C1 and catechin), bitterness (caftaric acid), sweetness (glucose and fructose), and sourness (lactic acid, tartaric acid and malic acid) should be detected simultaneously and quantitatively. All of these compounds contain diol units as common functional groups (Figure 1), we thus decided to synthesize PT derivatives (S1, S2, and S3) functionalized with pyridinium boronic acid which possesses a binding affinity for diols.[13] The pyridinium unit endows not only amphiphilicity but high solubility in aqueous media. Upon binding to the diol-

Chemistry – A European Journal

COMMUNICATION

10.1002/chem.202002262
containing compounds, the chemosensors S1‒S3 become zwitterionic, resulting in self-aggregation. The self-aggregation is reflected to the fluorescence quenching profile of S1‒S3. Thus, the quenching profile should depend on types of the wine components. In addition, the different linker lengths among S1, S2, and S3 offer multi-fluorescence responses originated from a combination of the chemosensors and wine components. The obtained fingerprint like fluorescence patterns deserve comprehensive analysis of 8 types of wine components utilizing pattern recognition techniques (i.e. linear discriminant analysis (LDA)[4a],[14]). Finally, we demonstrated a quantitative analysis for astringent procyanidin C1 in the real red wine using machine learning algorithm (i.e. support vector machine (SVM)[15],[16]).

Poly{3-[4-(pyridinium-3-boronic acid)butyl]thiophene} bromide (S1) and poly{3-[5-(pyridinium-3-boronic acid)pentyl]thiophene} bromide (S2) were newly synthesized in this study (See the supporting information (SI)). Although poly{3- [6-(pyridinium-3-boronic acid)hexyl]thiophene} bromide (S3) was previously reported,[11] the synthesis was not reproducible and thus requests a revisit of the reaction conditions. The following describes our process to find the reaction conditions for S3. Initially, poly(3-bromohexylthiophene) was reacted with 3- pyridylboronic acid (3-PyBA) in N,N-dimethylformamide at 70 °C for 3 days according to the previous report.[11] However, 1H NMR analysis of the resulting PT derivative showed a complete elimination of the boronic acid moiety (Figure S4), which hindered further analytical characterization and study on the molecular recognition ability.
The elimination would arise from heating of the reaction system, which produces boroxine derivatives and water.[17] Water molecules cause the elimination of the boronic acid moiety from the aromatic ring.[18] Thus, we attempted to suppress the elimination reaction using toluene as a solvent. The elimination rate was reduced to 42 mol% by the reaction at 120 °C for 44 h (see the SI). To further suppress the elimination reaction, we widely investigated the effects of molar equivalence of 3-PyBA, solvents, and reaction time. After much trials, we found that the reaction with 10 eq. of 3-PyBA in anhydrous chloroform at 70 °C for 48 h afforded elimination rate down to 1.5 mol% (see the SI). On the other hand, 3.5 mol% of 3-hexylbromothiophene units were remained unreacted. The reaction progress and the suppression of boronic acid elimination were found to be trade-off.
Moreover, the 1H NMR spectrum of S3 in this study was completely different from that of the previous report.[11] We expect that our insight will be helpful for derivatization of a variety of boronic acid-attached polymers.
The PT derivatives S1‒S3 were dissolved in various aqueous-methanol media. Thus, optical characteristics of the obtained chemosensors were evaluated by UV-vis and fluorescence measurements in the aqueous solution. The absorption and emission maxima of S1‒S3 were observed at 420 nm and 570 nm, respectively. Moreover, the emission quantum yield (Φ) and emission lifetime (τ) were estimated to be Φ = 12‒ 15% and τ = 0.47‒0.55 ns, respectively (see the SI). Next, we investigated the sensing ability of the chemosensors by target titrations. The UV-vis spectra of S1‒S3 upon the addition of wine components showed redshifts with isosbestic points. A representative example of the UV-vis titration is shown in Figure S7. The optical change stemmed from an expansion of π- conjugate length in the PT, which further induced the self- aggregation. Accompanied with the UV-vis spectral shift, the fluorescence intensity was decreased. For example, the titration
of fructose is shown in Figure 2A. The fluorescence intensity of S1 was slightly quenched upon the addition of fructose at mmol/L levels, which was the reasonable result in comparison with previously reported fluorescence chemosensors for saccharides in an artificial wine.[4k] On the other hand, a drastic quenching was observed with an increase in the concentration of procyanidin C1, and the fluorescence quenching was saturated at 40 nM (Figure 2B). It should be noted that the astringent procyanidin C1 exhibited the highest sensitivity due to the multi-binding sites. Furthermore, unlike fructose and procyanidin C1, a sigmoidal response was observed in a fluorescence titration for tartaric acid accompanied with a redshift (Δλem = 25 nm) (Figure 2C). The observed redshift was probably caused by changes in polarity surrounding PT backbone and subsequent planarization.[19] Further addition of tartaric acid (> 0.6 mM) resulted in slight revert back to the original wavelength (Figure 2C). Interestingly, the optical responses depend on chemical structures of the wine components, meaning that the sigmoidal profile was obtained only for compounds containing carboxy groups (i.e. sour and bitter components). The observed sigmoidal feature would stem presumably from the multi-interactions between α- hydroxycarboxylate moieties of the components and boronic acid moieties of the PT derivatives.[20]
In this regard, the estimated Stern-Volmer quenching constants (KSV)[21] reflect the above-mentioned responses mainly arose from the static quenching process. Among the wine components, the higher KSV values were obtained for the astringent compounds. As expected, the KSV of the trimeric phenol compound procyanidin C1 was higher than that of catechin, suggesting that multi-interactions between the PT derivative and procyanidin C1 contributed to the high sensitivity. In addition, the KSV for procyanidin C1 exhibited 105 times higher than that of fructose. Overall, the order of the response was determined as follows: astringent (~107 M-1) > bitterness (~105 M-1) > sourness (102~103 M-1) > sweetness (101~102 M-1) (Table 1). The determined Ksv values match the concentration scale of corresponding components,[4f],[22] suggesting that four types of the taste components can be comprehensively detected by S1‒S3.

To establish an easy-to-handle and robust artificial tongue for wine analysis, a solid-state chemosensor array chip was employed in this assay. The chemosensors (S1, S2, and S3) in a hydorgel were dispensed in a microwell (0.2 μL/well) by a robotic dispenser to fabricate a uniform microarray chip. The fluorescence image patterns of the chemosensor array chip were recorded by a CCD camera (see the SI).[23] In other words, the proposed chemosensor array system would serve a simple wine analysis without any large analytical apparatuses. The datasets for pattern recognition contain information of fluorescence intensities for each excitation wavelength. Totally 3840 datasets

Table 1. The Stern-Volmer quenching constants (Ksv [M-1]) obtained from fluorescence titration.
Target wine components S1 S2 S3

Procyanidin C1 5.8 ×107 4.6 ×107 8.4 ×107
Catechin 2.9 ×107 9.3 ×106 2.6 ×107
Caftaric acid 3.4 ×105 3.1 ×105 3.1 ×105
L-Tartaric acid 1.6 ×103 1.1 ×103 4.2 ×103
L-Malic acid 1.7 ×102 5.5 ×102 3.3 ×102
L-Lactic acid 2.4 ×102 2.8 ×102 2.8 ×102
D-Glucose 3.2 ×101 3.6 ×101 5.6 ×101
D-Fructose 1.1 ×102 1.9 ×102 1.5 ×102

This article is protected by copyright. All rights reserved.
Chemistry – A European Journal

COMMUNICATION
10.1002/chem.202002262

Fluorescence spectra of (A) S1, (B) S2, and (C) S3 upon the addition of fructose, procyanidin C1, and tartaric acid, respectively. Each titration was performed in a mixture of phosphate buffer (10 mM) and MeOH (1:9, v/v) at 7.0 at 25 ○C. [S1] = [S2] = [S3] = 10 μM, [D-fructose] = 0‒10 mM, [procyanidin C1]
= 0‒0.3 μM, [L-tartaric acid] = 0‒1.5 mM. λex = 460 nm. The titration isotherms were obtained by gathering the maximum emission (λem = 573 nm) intensities at various target concentrations.(3 chemosensors × 8 analytes × 20 repetitions × 8 recording conditions) were collected after data preprocessing (i.e. the Student’s t-test[24] and analysis of variance (ANOVA)[25]) for the qualitative assay. The ANOVA result implied a significance of alkyl chains in the PT chemosensors, that is, the longer side- chains offered larger optical changes including fluorescence intensities and wavelength changes toward a multi-recognition for wine components (see the SI). Furthermore, the LDA offers a statistic classification of wine components and/or concentrations by decreasing a dimension of the data matrix. The LDA canonical score plots for the simultaneous discrimination of 8 types of wine components were shown in Figure 3. Based on the general contents of wine components, the concentrations for the LDA was set as follows: 10 μM for bitterness and astringent, and 30 mM for sourness and sweetness.[4f] The 9 clusters (8 analytes and control) were classified with 99% accuracy in complete aqueous media. In this study, we have achieved the simultaneous qualitative multi-recognition including astringent procyanidin C1 with high accuracy utilizing fluorescence PT derivatives S1, S2, and S3.
Semi-quantitative assay was subsequently conducted with three types of components which are often employed for wine analysis. In this assay, fructose, tartaric acid, procyanidin C1 and those concentrations were discriminated by LDA. As shown in Figure 4, 3 types of species were discriminated with 99% accuracy. Importantly, the cluster distribution of each target exhibited the concentration-dependent manner. Unlike fructose and procyanidin C1, the direction for the cluster distribution of tartaric acid was lateral, which would potentially arise from the sigmoidal response.

LDA canonical plots with 99% confidence ellipsoids for the qualitative analysis of 8 types of wine components and control. 20 repetitions were conducted for each analyte. LDA results of the semi-quantitative assay for D-fructose (square), L- tartaric acid (diamond), and procyanidin C1 (triangle) with 99% confidence ellipsoids. 20 repetitions were conducted for each target.

This article is protected by copyright. All rights reserved.
Chemistry – A European Journal

COMMUNICATION
Finally, the quantitative analyses in complicated environments such as mixtures of wine components and a commercially available wine were demonstrated. The SVM is one of the machine learning algorithms which serve as a powerful pattern recognition model allowing to predict concentrations of unknown samples by calibration lines obtained from the training data.[15],[16] The mixtures consist of various concentrations of fructose, tartaric acid, and procyanidin C1 (see the SI). The SVM results including 3 species and each 2 concentrations (totally 6 types of predictions) indicated a highly accurate analysis even in the mixtures (see the SI). To achieve further authentic wine- analysis utilizing the supramolecular artificial tongue based on the PT derivatives, we conducted a regression analysis of the procyanidin C1 concentration in the red wine (Cabernet Sauvignon). Totally 9 datapoints were selected as calibration dataset (black squares). The predicted clusters of procyanidin C1 at 4 μM and 14 μM were distributed on the obtained calibration line, which indicates the successful recognition and prediction of procyanidin C1 in the wine samples (Figure 5). Overall, we established the authentic artificial tongue by the integration strategy with polymer-based supramolecular chemosensors and pattern recognition models.
Results of SVM regression of procyanidin C1 in the red wine. The values of the root-mean-square errors of calibration (RMSEC) and prediction (RMSEP) confirm the high accuracy of the built prediction model.

In summary, we synthesized the PT derivatives S1‒S3 modified with 3-PyBA as fluorescent chemosensors for wine components. Synthetic procedures were discovered to introduce more than 95 mol% of 3-PyBA in the PT units. Our numerous trials for PT derivatives would provide important insights for a variety of boronic acid-attached polymers. The PT derivatives exhibited various optical responses depending on the chemical structures of analytes, resulting in the different magnitude of the Stern- Volmer quenching constants. The detectability of the PT derivatives was suitable for wine components in practical concentration range. Interestingly, S1‒S3 exhibited strongest response to procyanidin C1 among the components. The difference in the length of side chains offered optical responses including fluorescence intensities and wavelength changes upon multi-recognition for wine components, which allowed us to achieve simultaneous qualitative and quantitative analysis with LDA. Most importantly, the concentration of procyanidin C1 was
410.1002/chem.202002262 successfully predicted in the real red wine (Cabernet Sauvignon). To the best of our knowledge, this is the first report on the comprehensive detection of wine components including the astringents based on supramolecular chemosensors. We believe that our strategy will lead to the authentic artificial tongue for accurate prediction of taste.

Acknowledgements

TM thanks the Japan Society for the Promotion of Science (JSPS), Grant-in-Aid for Young Scientists (A), No. 17H04882. YS thanks the JSPS, Research Fellow for Young Scientists (DC1), No.18J21190.

Conflict of Interest

The authors declare no conflict of interest.

Keywords: sensors • molecular recognition • polythiophene • wine • pattern recognition

[1] J. Chandrashekar, M. A. Hoon, N. J. P. Ryba, C. S. Zuker, Nature 2006,
444, 288–294.
[2] J. J. Lavigne, S. Savoy, M. B. Clevenger, J. E. Ritchie, B. McDoniel, S.-
J. Yoo, E. V. Anslyn, J. T. McDevitt, J. B. Shear, D. Neikirk, J. Am. Chem. Soc. 1998, 120, 6429–6430.
[3] For reviews, see: a) P. Ciosek, W. Wroblewski, Analyst 2007, 132, 963– 978; b) A. R. D. Rosa, F. Leone, F. Cheli, V. Chiofalo, J. Food Eng. 2017, 210, 62–75. For examples, see: c) Y. G. Vlasov, A. V. Legin, A. M. Rudnitskaya, A. D’Amico, C. Di Natale, Sens. Actuators B Chem. 2000, 65, 235–236; d) L. Lvova, S. S. Kim, A. Legin, Y. Vlasov, J. S. Yang, G.
S. Cha, H. Nam, Anal. Chim. Acta 2002, 468, 303–314; e) M. Y. M. Sim,
T. J. Shya, M. N. Ahmad, A. Y. M. Shakaff, A. R. Othman, M. S. Hitam, Sensors 2003, 3, 340–349; f) M. Gutiérrez, A. Llobera, J. Vila-Planas, F. Capdevila, S. Demming, S. Büttgenbach, S. Mínguez, C. Jiménez- Jorquera, Analyst 2010, 135, 1718–1725.
[4] For reviews, see: a) P. Anzenbacher, Jr., P. Lubal, P. Buček, M. A. Palacios, M. E. Kozelkova, Chem. Soc. Rev. 2010, 39, 3954–3979; b) J. Wu, B. Kwon, W. Liu, E. V. Anslyn, P. Wang, J. S. Kim, Chem. Rev. 2015, 115, 7893–7943. For examples, see: c) S. L. Wiskur, E. V. Anslyn, J. Am. Chem. Soc. 2001, 123, 10109‒10110; d) C. Zhang, D. P. Bailey, K. S. Suslick, J. Agric. Food Chem. 2006, 54, 4925‒4931; e) M. A. Palacios,
Z. Wang, V. A. Montes, G. V. Zyryanov, P. Anzenbacher, Jr., J. Am. Chem. Soc. 2008, 130, 10307‒10314; f) A. P. Umali, S. E. LeBoeuf, R.
W. Newberry, S. Kim, L. Tran, W. A. Rome, T. Tian, D. Taing, J. Hong,
M. Kwan, H. Heymann, E. V. Anslyn, Chem. Sci. 2011, 2, 439‒445; g)
W. Niu, Luminescence 2013, 28, 239‒243; h) J. Han, M. Bender, K. Seehafer, U. H. Bunz, Angew. Chem. Int. Ed. 2016, 55, 7689‒7692; i) D. Calabria, M. Mirasoli, M. Guardigli, P. Simoni, M. Zangheri, P. Severi, C. Caliceti, A. Roda, Sens. Actuator B 2020, 305, 127522.
[5] S. Soares, E. Brandao, N. Mateus, V. de Freitas, Crit. Rev. Food Sci. Nutr. 2017, 57, 937–948.
[6] a) R. A. Arnold, A. C. Noble, V. L. Singleton, J. Agric. Food Chem. 1980, 28, 675–678; b) N. Martin, Food Quality and Preference 2002, 13, 295– 305.
[7] J. Ricardo-da-Silva, J. P. Rosec, M. Bourzeix, J. Mourgues, M.Moutounet,
Vitis -Geilweilerhof- 1992, 31, 55–63.
[8] a) E. J. Lemieux, M. Leclerc in Conjugated Polyelectrolytes: Fundamentals and Applications (Eds: B. Liu, G. C. Bazan), Wiley-VCH, Weinheim, 2013, pp. 231. For examples, see: b) H. A. Ho, M. Leclerc, J. Am. Chem. Soc. 2003, 125, 4412–4413; c) C. Li, M. Numata, M. Takeuchi, S. Shinkai, Angew. Chem. Int. Ed. 2005, 44, 6371–6374; d) T. Minami, Y. Kubo, Chem. Asian J. 2010, 5, 605–611; e) M.-P. Plante, E. Bérubé, L. Bissonette, M. Bergeron, M. Leclerc, ACS Appl. Mater. Inter.

COMMUNICATION
10.1002/chem.202002262

2013, 5, 4544–4548; f) D. Cheng, Y. Li, J. Wang, Y. Sun, L. Jin, C. Lib,
Y. Lu, Chem. Commun. 2015, 51, 8544–8546; g) S Sabury, G. S. Collier,
M. N. Ericson, S. M. Kilbey, Polym. Chem. 2020, 11, 820–829.
[9] a) M. S. Maynor, T. L. Nelson, C. O’Sulliva, J. J. Lavigne, Org. Lett. 2007, 9, 3217–3220; b) C. Li, G. Shi, ACS Appl. Mater. Interfaces 2013, 5, 4503–4510.
[10] For reviews, see: a) T. M. Swager, Acc. Chem. Res. 1998, 31, 201–207;
b) S. W. Thomas, G. D. Joly, T. M. Swager, Chem. Rev. 2007, 107, 1339–1386. For examples, see: c) L. Chen, D.W. McBranch, H.-L. Wang,
R. Helgeson, F. Wudl, D.G. Whitten, Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12287–12292; d) H. Yuan, B. Wang, F. Lv, L. Liu, S. Wang, Adv. Mater. 2014, 26, 6978‒6982; e) C.-H. Wang, E. E. Nesterov, Chem. Commun. 2019, 55, 8955‒8958.
[11] C. Xue, F. Cai, H. Liu, Chem. – Eur. J. 2008, 14, 1648‒1653.
[12] Y. Liu, X. Meng, M. Pei, G. Zhang, H. Li, Anal. Methods 2014, 6, 5812‒ 5817.
[13] For reviews, see: a) T. D. James, K. R. A. S. Sandanayake, S. Shinkai, Angew. Chem. Int. Ed. 1996, 35, 1911‒1922; b) S. D. Bull, M. G. Davidson, J. M. H. van den Elsen, J. S. Fossey, A. T. A. Jenkins, Y.-B. Jiang, Y. Kubo, F. Marken, K. Sakurai, J. Zhao, T. D. James, Acc. Chem. Res. 2013, 46, 312‒326; c) X. Liang and M. Bonizzoni, J. Mater. Chem. B 2016, 4, 3094‒3103. For examples, see: (d) Y. Sasaki, É. Leclerc, V. Hamedpour, R. Kubota, S. Takizawa, Y. Sakai, T. Minami, Anal. Chem. 2019, 91, 15570‒15576. e) X. Liang, M. Trentle, V. Kozlovskaya, E. Kharlampieva, M. Bonizzoni, ACS Appl. Polym. Mater. 2019, 1, 1341‒ 1349.
[14] a) R. G. Brereton, Applied Chemometrics for Scientists, Wiley-VCH, 2007, Chichester. 13; b) Z. Li, J. R. Askim, K. S. Suslick, Chem. Rev. 2019, 119, 231‒292; c) J. Tropp, M. H. Ihde, A. K. Williams, N. J. White,
N. Eedugurala, N. C. Bell, J. D. Azoulay, M. Bonizzoni, Chem. Sci. 2019,
10, 10247‒10255.
[15] L. H. Hamel, Knowledge Discovery with Support Vector Machines, Wiley,
2009, Hoboken.
[16] T. Minami, N. A. Esipenko, B. Zhang, M. E. Kozelkova, L. Isaacs, R. Nishiyabu, Y. Kubo, P. Anzenbacher, Jr., Am. Chem. Soc. 2012, 134, 20021‒20024.
[17] A. L. Korich, P. M. Iovine, Dalton Trans. 2010, 39, 1423‒1431.
[18] a) P. A. Cox, A. G. Leach, A. D. Campbell, G. C. Lloyd-Jones, J. Am. Chem. Soc. 2016, 138, 9145‒9157; b) P. A. Cox, M. Reid, A. G. Leach,
A. D. Campbell, E. J. King, G. C. Lloyd-Jones, J. Am. Chem. Soc. 2017,
139, 13156–13165.
[19] J. Zessin, F. Fischer, A. Heerwig, A. Kick, S. Boye, M. Stamm, A. Kiriy, M. Mertig, Nano Lett. 2017, 17, 5163–5170.
[20] C. W. Gray, Jr., T. A. Houston, J. Org. Chem. 2002, 67, 5426–5428. [21] J. Keizer, J. Am. Chem. Soc. 1983, 105, 1494–1498.
[22] M.-H. Salagoïty-Auguste, A. Bertrand, J. Sci. Food Agric. 1984, 35, 1241–1247.
[23] a) Y. Liu, T. Minami, R. Nishiyabu, Z. Wang, P. Anzenbacher Jr., J. Am. Chem. Soc. 2013, 135, 7705–7712; b) T. Minami, Y. Liu, A. Akdeniz, P. Koutnik, N. A. Esipenko, R. Nishiyabu, Y. Kubo, P. Anzenbacher Jr., J. Am. Chem. Soc. 2014, 136, 11396–11401.
[24] J.-M. Mermet, R. Kellner, M. Otto, M. Valcarcel, M. Widmer, Analytical Chemistry: A Modern Approach to Procyanidin C1 Analytical Science, Wiley, 2004, New York.
[25] L. Ståhle, S. Wold, Chemom. Intell. Lab. Syst. 1989, 6, 259–272.

COMMUNICATION

Entry for the Table of Contents
We propose the fluorescent polythiophene derivatives for comprehensive taste evaluation of wine. Especially, the first detection of astringent procyanidin C1 by chemosensors has been achieved at the practical concentration range in the real wine. Furthermore, the regression analysis for the astringent was successfully carried out in real red wine. Our strategy leads to “the chemosommelier” for the accurate prediction of taste.