English

• The interaction between dyes and surrounding molecules and the process of photoexcitation lead to changes in the conjugated system or molecular configuration of dye molecules, resulting in fluorescence changes, which create the signal transduction mechanism of fluorescent probes [1-3]. With the coordinated development of labeling techniques and dye chemistry, fluorescent probes have been continuously updated, from general environment-sensitive fluorescence changes to selective fluorescence responses after molecular recognition for specific target molecules [4,5]. The most typical fluorophore with molecular recognition fluorescence response performance is the fluorescent protein [6]. The molecular recognition between its endogenous chromophore and the fluorescent protein cavity effectively inhibits the rotation of the chromophore in the protein cavity and produces a unique phenolic hydroxyl dehydrogenation reaction. The hydroxyl dehydrogenation reaction ensures that the chromophore can only have fluorescence emission ability in the GFP cavity [7]. Such chromophores that activate fluorescence through molecular recognition are called fluorogens. Drawing on the pattern of recognition and binding of such endogenous chromophores and biomolecules to emit light, in recent years, relying on the screening of substrate molecules by antibodies or aptamers, non-covalently bound fluorescent antibodies and aptamers activated against exogenous chromophores has been developed [8-11]. Depending on genetic coding and enzymatic reactions, covalently linked protein tags, such as the widely used SNAP-tag [11,12] and Halo-tag [13], have been developed. This strategy of combining and activating fluorogen emission has stimulated the study of fluorescent probes that recognize biological macromolecules (including proteins, nucleic acids and carbohydrates) with low background fluorescence, but how to realize molecular recognition and activate fluorescence is a challenge in the research of such fluorescent probes [14-18].

  Inhibition of twisted intramolecular charge transfer (TICT) is one of the main mechanisms by which molecular recognition activates fluorescence, with dialiphatic substituted amino groups being the most commonly used rotors (Fig. 1a) [19-21]. The structure of such molecules is characterized by the direct conjugation of tertiary amine groups in planar rigid chromophores, common fluorophores such as coumarin, naphthalimide and rhodamine. The lone pair of electrons of the tertiary amine is delocalized to the aromatic system of the chromophore due to the effect of electron donating and acceptance, forming a partial C=N double bond. After being excited by light, an electron is completely intramolecularly transferred to the acceptor, and the tertiary amine group rotates to form a vertical configuration with the chromophore to stabilize the zwitterion with intramolecular charge separation. However, such molecules typically have high background fluorescence and lack strong binding to biomolecules due to their large size and structural rigidity.

  Figure 1

  

  Figure 1.  (a) The fluorophores with dialkylamine as a TICT rotator. (b) Aniline as a TICT rotor in methine dyes. (c) HSA probe with aniline as a TICT rotator in this work.

  DownLoad: Full-Size Img PowerPoint

  Reducing the size of the chromophore, and increasing the flexibility of the molecule is a conventional improvement idea (Fig. 1b). The fluorophore of the fluorescent protein is connected a methine group to a single benzene ring conjugated a system as the electron donor and acceptor, respectively. The existence of the methine group increases the flexibility when it is combined with biomolecules, and the size of the conjugated structure of the single benzene ring is moderate. This type of methionine chromophore has the structural characteristics of strong binding force with biological macromolecules. Inspired by fluorescent protein chromophores, aniline-substituted methine chromophores, with typical structures such as DMABI and Malachite Green, have been shown to have strong binding to both proteins and nucleic acids [22,23]. More importantly, aniline acts as an electron-donating group to form a quinoid structure, and the tertiary amine group and the aniline as a whole can co-rotate during the intramolecular charge transfer process in the excited state, which can more effectively quench the fluorescence. The background fluorescence of these chromophores is very low, and the fluorescence intensity increases by 2–3 orders of magnitude after binding to proteins or nucleic acids. In one of our recent works, the importance of the methine structure in binding to biomacromolecules and activating fluorescence was also demonstrated through the modification of the methine structure in the fluorophore by linking electron-donating and electron-withdrawing groups.

  In this paper, we introduced aniline into the difluoroboron β-diketonate chromophore via a methine bridge, resulting in N-BF₂ that is almost non-fluorescent in aqueous solution (Fig. 1c). N-BF₂ has fast and strong binding to HSA or BSA, accompanied by the activation of fluorescence (fluorescence enhancement 90 (HSA)/112 (BSA)-fold). This intermolecular binding was found to be rapid and reversible in solution and intracellularly. Since N-BF2 also has lipid droplet-targeted ability, the complex of N-BF₂HSA realizes the regulation of reversible lipid droplets staining in cells.

  Firstly, we inspected the quenching effect of aniline-rotor in N-BF₂, through the analysis of fluorescence spectra depending on different solvents and viscosity. Similar with the most traditional d-π-A dyes, N-BF₂ shown solvochromic effect with obvious red-shift in fluorescence emission from 541 nm in dioxane to 591 nm in DMSO (Fig. S1 in Supporting information). And it should be noted that N-BF₂ exhibited negligible fluorescence in strong polar solvent (ϕ < 0.01) which indicated the facile formation of TICT and its low background (Fig. 2a and Table S1 in Supporting information). This was further illustrated through the electron density changing in LUMO and HOMO (Fig. 2b). Once excited, the electron density of aniline decreased and flowed toward difluoroboron β-diketonate thus enhancing the charge separation which favored to form TICT. While with the increase of viscosity, the corresponding emission intensity of N-BF₂ significantly enhanced (Fig. 2a). Such viscosity-dependent fluorescence intensification is also a typical characteristic of TICT emission. These results indicated potential applications of N-BF₂ in developing fluorogenic probes based on restricted rotation.

  Figure 2

  

  Figure 2.  (a) Fluorescence spectra of N-BF₂ in different proportions of glycerol/EtOH mixture and H₂O. (b) Molecular orbital amplitude plots of HOMO and LUMO energy levels of N-BF₂. (c) The absorption and (d) fluorescence spectra of N-BF₂ upon addition of HSA/BSA and pretreated with ibuprofen (200 µmol/L). (e) Time course of fluorescence intensity of 5 µmol/L N-BF₂ in the presence of 5 µmol/L HSA and BSA. (f) The Job's plot of N-BF₂ and HSA at difffferent ratios, total concentration of N-BF₂ and HSA was 10 µmol/L. λ_ex = 460 nm.

  DownLoad: Full-Size Img PowerPoint

  Surprisingly, the rotation of aniline-rotor in N-BF₂ can be inhibited upon binding HSA and BSA thus inducing great fluorescence enhancement. After the addition of HSA or BSA, the maximum absorption of N-BF₂ at 513 nm slightly blue-shifted to 506 nm and 507 nm, respectively (Fig. 2c). Similarly, the emission at 583 nm blue-shifted to 568 nm and 559 nm, respectively, which indicated the polar microenvironment of cavity binding the chromophore (Fig. 2d). In the meanwhile, the fluorescence intensity increased more than 90-fold (HSA) and 112-fold (BSA), respectively (Table S2 in Supporting information). We also found N-BF₂ could sense HSA and BSA in ~1.2 s much faster than those reported HSA/BSA probes (Fig. 2e and Table S3 in Supporting information). Furthermore, the Job's plot was conducted through monitoring fluorescence intensity at 568/559 nm, and indicated the N-BF₂/HSA and N-BF₂/BSA complexes all had 1:1 stoichiometry (Fig. 2f and Fig. S2 in Supporting information).

  To further understand the binding between HSA and N-BF₂, drug competition and molecular docking experiments were next performed. Phenylbutazone and ibuprofen, which have been known to bind HSA/BSA at site Ⅰ and site Ⅱ, respectively, are used to inspect the binding model. Both N-BF₂/HSA and N-BF₂/BSA complexes shown slight decrease in presence of 50–200 µmol/L phenylbutazone. However, the addition of ibuprofen could induce the fluorescence intensity of N-BF₂ sharply decreased to 30% (Fig. 3a). It indicated that N-BF₂ most likely to bind HSA/BSA at narrow site Ⅱ which considered as the common drug sites. The following docking study between N-BF₂ and HSA/BSA also demonstrated this result (Fig. 3b and Fig. S3 in Supporting information). It revealed that oxygen in difluoroboron β-diketonate could formed two stable hydrogen bonds of 2.1 Å and 2.5 Å with the basic amino acid Arg-410 and Lys-414, which made N-BF₂ specifically bind site Ⅱ. On the other side, hydrophobic interactions of aniline unit with surrounding neutral amino acids, Phe-403, Val-433, Leu-453, Ala-449, Cys-438, Cys-392 and Ile-388, were found thus inhibiting the rotation of aniline and enhancing fluorescence intensity of N-BF₂. Inspired by the stable binding model, titration experiment and double logarithmic equation was further conducted to examine the limit of detection (LOD) and binding constant (K_b). With the addition of HSA, the fluorescence intensity gradually enhanced and the emission wavelength shifted from 583 nm to 568 nm (Fig. 3c). A linear relationship (R² = 0.9845) was found between fluorescence intensity of N-BF₂ and HSA concentration in the range of 0–10 µmol/L (Fig. 3d). Based on 3σ/k, the corresponding limit of detection (LOD) for HSA was calculated to be 1.21 mg/mL (Fig. 3d). Besides, with the addition of N-BF₂, the fluorescence intensity at 340 nm gradually decreased and the concentration-dependent fluorescence changes were utilized to find the binding constant, the K_b was 4.61 × 10⁵ L/mol (Figs. 3e and f). LOD and K_b of N-BF₂ toward BSA was also inspected to be 0.93 mg/mL and 4.67 × 10⁵ L/mol which indicated the strong chelation between HSA/BSAand N-BF₂ (Figs. S4 and S5 in Supporting information). As an endogenous substance in vivo, HSA not only has high stability but also acts as a drug carrier which can increase the water solubility of drugs and regulate drug release. Therefore, we used N-BF₂/HSA complexes as a trigger to control the entry and exit of N-BF₂ in living cells. As shown in Figs. 4a–c, N-BF₂ could specifically light up lipid droplets in HeLa cells. With the addition of 30 µmol/L HSA, the fluorescence of N-BF₂ in lipid droplets disappeared, indicating that N-BF₂ gradually dissociated from lipid droplets and flowed out of cells to bind HSA (Figs. 4d–f). Ibuprofen could bind HSA with higher affinity. Then the addition of ibuprofen replaced N-BF₂ from HSA and the released N-BF₂ could re-cross the cell membrane and go back into lipid droplets which had been confirmed by the recovery of fluorescence in lipid droplets (Figs. 4g–i). So, it could be concluded that the complex of N-BF₂/HSA indeed realized the regulation of reversible lipid droplet staining in cells.

  Figure 3

  

  Figure 3.  (a) Normalized fluorescence intensity of N-BF₂ (5 µmol/L) in presence of HSA (5 µmol/L) or BSA (5 µmol/L) which pretreated with different concentrations of site-specific marker (0–200 µmol/L ibuprofen or phenylbutazone). (b) Binding modes of N-BF₂ with contacting residues in binding site Ⅱ of HSA. (c) Fluorescence spectra of N-BF₂ (5 µmol/L) in presence of different concentration of HSA (0–10 µmol/L). (d) The linear relationship of fluorescence intensity at 568 nm with concentration of HSA (0–10 µmol/L). λ_ex = 460 nm. (e) Fluorescence emission spectra of HSA (5 µmol/L) in presence of increasing concentration of N-BF₂ (0–10 µmol/L) and (f) double logarithmic relationship of fluorescence intensity at 340 nm with the concentration of N-BF₂ (0–10 µmol/L). λ_ex = 280 nm.

  DownLoad: Full-Size Img PowerPoint

  Figure 4

  

  Figure 4.  Confocal cellular imaging of 1 µmol/L N-BF₂ in HeLa cells in presence or absence of HSA and ibuprofen. [HSA] = 30 µmol/L. [Ibuprofen] = 30 µmol/L. Scale bar = 10 µm.

  DownLoad: Full-Size Img PowerPoint

  In conclusion, we developed a fluorogenic HSA/BSA probe through introducing aniline and a methine bridge into the difluoroboron β-diketonate chromophore. Due to the inhibition of aniline rotation, N-BF₂ could light up HSA/BSA in 1.2 s with 90/112-fold fluorescence enhancements. Furthermore, the stable hydrogen bonds made N-BF₂ exhibit high sensitivity with HSA and K_b towards site Ⅱ of HSA/BSA, with LOD 1.21/0.93 mg/mL, K_b 4.61 × 10⁵/4.67 × 10⁵ L/mol, which had been confirmed by molecular docking experiments. Finally, we realized the regulation of reversible lipid droplets staining through binding and dissociation between N-BF₂ and HSA in living cells, which may be applied in drug site-specific delivery.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgment

  This work is supported by the National Natural Science Foundation of China (Nos. 22078314, 21878286, 21908216, 22078201, U1908202).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.04.070.

  
  
• 1. [1]

     H. Singh, K. Tiwari, R. Tiwari, et al., Chem. Rev. 119(2019) 11718–11760. doi: 10.1021/acs.chemrev.9b00379

  2. [2]

     Z. Xu, J. Yoon, D.R. Spring, Chem. Soc. Rev. 39(2010) 1996–2006. doi: 10.1039/b916287a

  3. [3]

     W. Zhou, X. Fang, Q. Qiao, et al., Chin. Chem. Lett. 32(2021) 943–946.

  4. [4]

     J. Chen, W. Liu, X. Fang, et al., Chin. Chem. Lett. 33(2022) 5042–5046. doi: 10.1016/j.cclet.2022.03.120

  5. [5]

     J. Chen, C. Wang, W. Liu, et al., Angew. Chem. Int. Ed. 60(2021) 25104–25113. doi: 10.1002/anie.202111052

  6. [6]

     R.Y. Tsien, Angew. Chem. Int. Ed. 48(2009) 5612–5626. doi: 10.1002/anie.200901916

  7. [7]

     D.P. Barondeau, C.D. Putnam, C.J. Kassmann, et al., Proc. Natl. Acad. Sci. U. S. A. 100(2003) 12111–12116. doi: 10.1073/pnas.2133463100

  8. [8]

     C. Szent-Gyorgyi, B.F. Schmidt, Y. Creeger, et al., Nat. Biotechnol. 26(2008) 235–240. doi: 10.1038/nbt1368

  9. [9]

     E. Gallo, Bioconjug. Chem. 31(2020) 16–27. doi: 10.1021/acs.bioconjchem.9b00710

  10. [10]

      L. Jullien, A. Gautier, Methods Appl. Fluoresc. 3(2015) 042007. doi: 10.1088/2050-6120/3/4/042007

  11. [11]

      S. Xu, H.Y. Hu, Acta Pharmacol. Sin. 8(2018) 339–348. doi: 10.1016/j.apsb.2018.02.001

  12. [12]

      W. Liu, Q. Qiao, J. Zheng, et al., Biosens. Bioelectron. 176(2021) 112886. doi: 10.1016/j.bios.2020.112886

  13. [13]

      C.G. England, H. Luo, W. Cai, Bioconjug. Chem. 26(2015) 975–986. doi: 10.1021/acs.bioconjchem.5b00191

  14. [14]

      W. Liu, R. Li, F. Deng, et al., ACS Appl. Bio Mater. 4(2021) 2104–2112. doi: 10.1021/acsabm.0c01269

  15. [15]

      W. Chi, Q. Qiao, C. Wang, et al., Angew. Chem. Int. Ed. 59(2020) 20215–20223. doi: 10.1002/anie.202010169

  16. [16]

      F. Deng, Q. Qiao, J. Li, et al., J. Phys. Chem. B 124(2020) 7467–7474. doi: 10.1021/acs.jpcb.0c05642

  17. [17]

      F. Deng, L. Liu, W. Huang, et al., Spectrochim. Acta A Mol. Biomol. Spectrosc. 240(2020) 118466–118474. doi: 10.1016/j.saa.2020.118466

  18. [18]

      Q. Qiao, W. Liu, J. Chen, et al., Angew. Chem. Int. Ed. 61(2022) e202202961.

  19. [19]

      C. Wang, W. Chi, Q. Qiao, et al., Chem. Soc. Rev. 50(2021) 12656–12678. doi: 10.1039/D1CS00239B

  20. [20]

      W. Liu, J. Chen, Q. Qiao, et al., Chin. Chem. Lett. 33(2022) 4943–4947. doi: 10.1016/j.cclet.2022.03.121

  21. [21]

      X. Li, J. Zheng, W. Liu, et al., Chin. Chem. Lett. 31(2020) 2937–2940. doi: 10.1016/j.cclet.2020.05.043

  22. [22]

      K. Ferreira, H.Y. Hu, V. Fetz, et al., Angew. Chem. Int. Ed. 56(2017) 8272–8276. doi: 10.1002/anie.201701358

  23. [23]

      X. Tan, T.P. Constantin, K.L. Sloane, et al., J. Am. Chem. Soc. 139(2017) 9001–9009. doi: 10.1021/jacs.7b04211

• 

  Figure 1  (a) The fluorophores with dialkylamine as a TICT rotator. (b) Aniline as a TICT rotor in methine dyes. (c) HSA probe with aniline as a TICT rotator in this work.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Fluorescence spectra of N-BF₂ in different proportions of glycerol/EtOH mixture and H₂O. (b) Molecular orbital amplitude plots of HOMO and LUMO energy levels of N-BF₂. (c) The absorption and (d) fluorescence spectra of N-BF₂ upon addition of HSA/BSA and pretreated with ibuprofen (200 µmol/L). (e) Time course of fluorescence intensity of 5 µmol/L N-BF₂ in the presence of 5 µmol/L HSA and BSA. (f) The Job's plot of N-BF₂ and HSA at difffferent ratios, total concentration of N-BF₂ and HSA was 10 µmol/L. λ_ex = 460 nm.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Normalized fluorescence intensity of N-BF₂ (5 µmol/L) in presence of HSA (5 µmol/L) or BSA (5 µmol/L) which pretreated with different concentrations of site-specific marker (0–200 µmol/L ibuprofen or phenylbutazone). (b) Binding modes of N-BF₂ with contacting residues in binding site Ⅱ of HSA. (c) Fluorescence spectra of N-BF₂ (5 µmol/L) in presence of different concentration of HSA (0–10 µmol/L). (d) The linear relationship of fluorescence intensity at 568 nm with concentration of HSA (0–10 µmol/L). λ_ex = 460 nm. (e) Fluorescence emission spectra of HSA (5 µmol/L) in presence of increasing concentration of N-BF₂ (0–10 µmol/L) and (f) double logarithmic relationship of fluorescence intensity at 340 nm with the concentration of N-BF₂ (0–10 µmol/L). λ_ex = 280 nm.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Confocal cellular imaging of 1 µmol/L N-BF₂ in HeLa cells in presence or absence of HSA and ibuprofen. [HSA] = 30 µmol/L. [Ibuprofen] = 30 µmol/L. Scale bar = 10 µm.

  下载: 全尺寸图片 幻灯片
•