English

• Owing to their fascinating bright emissions, aggregation-induced emission (AIE) fluorophores have extensive applications in optical chemistry, materials science, and biomedical imaging field [1-5]. Structurally, AIE moieties featuring multiple rotatable phenyl rings (e.g., tetraphenylethene (TPE) and triphenylamine (TPA)) have enabled AIEgens to advance biomedical imaging as rotor-based agents [6-8]. Recently, the rotor probes have enabled the visualization of dynamic processes involving fluid-phase diffusion in cells and membrane [9-16], protein aggregation and misfolding [17-19], ferroptosis [20] and organelle interactions [21]. They also facilitate in vivo detection of viscous-related diseases including liver injury, inflammation, neurodegenerative diseases, and cancer [22-24].

  In principle, AIE-rotor probes enhanced fluorescence through the restriction of intramolecular motions (RIM), which arises from two main factors [25-28] (1) aggregated form-induced fluorescence, and (2) restricted form-induced fluorescence in the monomeric form. In the aggregated state, non-radiative decay pathways are suppressed through self-aggregation. Conversely, in the monomeric state, restriction can result from interactions with the environment or other molecules, that provide responsiveness to the surrounding environment. The interplay between aggregation and molecular rotor effects significantly influences the photophysical properties of AIEgens [29-31]. However, the quantum-mechanism underling AIEgens-enhanced fluorescence through aggregation or rotor effects is not well understood. Therefore, it is crucial to understand the mechanisms of aggregation-activation and rotor restriction in AIE probes. The development of new systems to overcome this challenge is expected to pave the way for new applications and expand the use of AIE rotors.

  To investigate the functional behaviors of AIEgens, the well-known AIEgen TPE was integrated into cyanine systems, valued for their low toxicity and versatile structures [32-34], making them crucial for clinical and preclinical applications (e.g., cyanine 5 (Cy5), indocyanine green (ICG), pafolacianine (OTL-38)) [35-38]. The TPE-cyanine system provides a beneficial case to study TPE's behavior; however, a method for exclusively controlling TPE-cyanine's fluorescence, which is activated through rotor restriction or aggregation, remains unclear. In this work, by tuning the methine chain length from Cy3 to Cy5 to Cy7, we demonstrate that the frontier energies of cyanines allow effective control over the TPE's behavior in TPE-cyanine systems. We propose a new mechanism to dominate the function of TPE by adjusting the energy difference (ΔE (DA)) between TPE and cyanines (Fig. 1). Consequently, TPE-cyanine can exhibit fluorescence through the AIE effect in Cy3, the rotor effect in Cy5, which is uniquely devoid of aggregation activation, and no effect in Cy7. Interestingly, TPE-Cy5 (Cy5PT) showed a 90-fold enhancement in the rotor-restricted state, which responded to tumor rigidity both in vitro and in vivo. Thus, this study not only provides a method to control the function of AIEgnen, but also expands the use of AIE rotors in tumor biophysics.

  Figure 1

  

  Figure 1.  Mechanism for AIEgen, rotor-specific AIEgen and ACQgen of TPE-cyanines under theoretical calculations. The figure illustrates tuning cyanine chromophore via the methine-chain length to control TPE's behavior in TPE-cyanine systems.

  DownLoad: Full-Size Img PowerPoint

  Structurally, cyanines can function as vinyl (CH═CH) rotor probes because of the ability of the heterocyclic rings to rotate around the methine chain [20,22]. However, they exhibit aggregation-caused quenching (ACQ) effect. To mitigate this drawback, TPE can act as electron donors in cyanine scaffolds [39-42], or be introduced as a rotor-decorator into D-π-A cyanine systems, as demonstrated in previous studies including our own [43,44]. Initially, we introduced TPE groups on the methine chain of the substrates trimethine cyanine (Cy3P), pentamethine cyanine (Cy5P) and heptamethine canine (Cy7P) to generate TPE-cyanines (denoted as Cy3PT, Cy5PT, and Cy7PT, respectively) (Fig. 2a). The success of these reactions was confirmed by high resolution mass spectrometry (HRMS), ¹H nuclear magnetic resonance (NMR), and ¹³C NMR analyses of the resulting conjugates (Figs. S1–S18 and Table S1 in Supporting information). However, AIE fluorophores typically exhibit fluorescence enhancement in response to poorly soluble solvents (e.g., water for organic dyes), which has thus far limited the in vivo applications. To address this issue, short hydrophilic polyethylene glycol (PEG) moieties were incorporated into the cyanine substrates to improve their water solubility and promote good biocompatibility. Subsequently, we investigated their aggregation fluorescence in aqueous environments. Due to their good solubility in water, polyacrylic acid (PAA) was used to induce cyanine aggregation. Compared to their monomeric forms, Cy5PT and Cy7PT showed no obvious fluorescence enhancement in the aggregates (Fig. 2b). Interestingly, Cy3PT uniquely exhibited a 76-fold fluorescence enhancement in aggregates compared with its monomer. These results suggested that TPE predominately displayed an aggregation-enhancement effect in the Cy3PT scaffold, which was in accordance with the AIE rule.

  Figure 2

  

  Figure 2.  Spectra properties of TPE-cyanines. (a) Structures of TPE-cyanines Cy3PT, Cy5PT, and Cy7PT. (b) Normalized emission spectra of TPE-cyanines Cy3PT, Cy5PT, and Cy7PT in monomer (pure water) and aggregates (PAA 0.5 wt% in water) in aqueous environment. (c) Fluorescence emission spectra of Cy3PT, Cy5PT, and Cy7PT in different volume fractions of glycerol/H₂O (f_Gly) solutions. (d) Viscosity sensitivity of Cy3PT, Cy5PT, and Cy7PT in Gly/H₂O mixtures. The linear relationship of double logarithmic plots of fluorescence intensity (F) vs. viscosity (η).

  DownLoad: Full-Size Img PowerPoint

  Next, the spectral properties of the restricted monomeric dyes were investigated in different glycerol fractions. Glycerol (Gly) was used as a standard solvent to analyze the fluorescence enhancement sensitivity upon rotational restriction of the rotor-based fluorophores. Notably, unlike previous studies utilizing glycerol-organic systems (e.g., methanol and tetrahydrofuran) [20,45], a glycerol-water system was used in our study to achieve favorable applications in physiological environments. As shown in Figs. S19–S21 (Supporting information), the substrate cyanines Cy3P, Cy5P, and Cy7P exhibited 6-, 4-, and 3-fold fluorescence enhancements in the glycerol–water mixture (f_G 90%), respectively compared to pure water. Remarkably, Cy3PT and Cy5PT demonstrated more substantial enhancement, with 103- and 90-fold improvements, respectively, in the same glycerol–water system. Surprisingly, Cy7PT exhibited only a modest 4-fold enhancement in glycerol-water systems, similar to the substrate cyanine Cy7P with a 3-fold enhancement on its own (Fig. 2c and Fig. S21). These findings suggest that TPE does not function effectively as a rotor or AIE property in the Cy7PT system. Additionally, Cy3PT (x = 0.77) and Cy5PT (x = 0.73) exhibited excellent viscosity sensitivity compared with Cy7PT (x = 0.27) (Fig. 2d). Notably, TPE selectively exhibited its fluorescence in the restricted monomeric form but not in the aggregated state within the Cy5PT system. To understand the mechanism behind this phenomenon, we performed theoretical calculations and analyzed the frontier orbital energies of the dyes.

  Detailed density functional theory (DFT) calculations were employed to elucidate why the different phenomena were observed for Cy3P, Cy5P, and Cy7P after the incorporation of TPE. Sketches of the molecular orbitals (MOs) showed that the MOs in the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) were substantially localized on the substrate cyanine skeletons, and the MOs in HOMO-1 were mainly distributed in the TPE group for Cy3PT, Cy5PT, and Cy7PT (Fig. 3a). For Cy3PT, the time-dependent density functional theory/polarizable continuum model (TD-DFT/PCM) calculations suggested that the HOMO to LUMO dominated the electronic transition with the largest oscillator strength (f = 1.3831) in aqueous solution (Table S2 in Supporting information). However, the HOMO-1 to LUMO dominated the electronic transition, with the largest oscillator strength (f = 1.3914) for Cy5PT in aqueous solution. Obviously, Cy5PT excitation involves a twisted intramolecular charge transfer (TICT) process from the TPE group to the substrate pentamethine cyanine skeleton, where the substrate cyanine skeleton and the TPE group act as the electron acceptor and donor, respectively. As for Cy7PT, the electronic transition was generated via HOMO and LUMO with the largest oscillator strength (f = 2.1237). However, the TPE group exhibits a low electron density in the LUMO. This observation suggests a role exchange between TPE and cyanine Cy7P, where TPE acts as the acceptor, which is distinct from its role in Cy5PT. Therefore, calculations demonstrated that TICT from TPE (donor) to cyanine skeleton (acceptor) in the excited state was the main factor for TPE to function solely as a rotor without AIE property in the Cy5PT scaffold (Fig. 3a).

  Figure 3

  

  Figure 3.  Mechanism for AIE, rotor-specific AIE and ACQ phenomenon of TPE-cyanines under theoretical calculations. (a) Molecular orbitals (LUMOs and HOMOs) of Cy3PT, Cy5PT and Cy7PT calculated by PCM/TD-DFT method. The PCM model is utilized for calculation in aqueous solution. (b) Band gaps (ΔE_H–L and ΔE_H–H) between TPE and substrate cyanines calculated by TD-DFT method. Hydrogen atoms and PEG chain substituents are omitted for clarity.

  DownLoad: Full-Size Img PowerPoint

  Calculation offered deeper insight that the HOMO_D energy of TPE (donor) was higher than the LUMO_A energy of substrate Cy5P (acceptor), giving rise to minimal gap (ΔE_H–L (DA)) between HOMO_D and LUMO_A for Cy5PT (Fig. 3b, Table S3 in Supporting information). This process agrees with the intramolecular charge transfer (ICT) [46,47]. In contrast, the higher LUMO_A of the substrate Cy3P inhibited TICT from TPE to the cyanine skeleton (Fig. 3b). Thus, the emission process of Cy3PT involved a direct transition from LUMO to HOMO with the largest oscillator strength (f = 1.3831, Table S2), allowing aggregates-induced fluorescence recovery. Upon changing the cyanine substrate to Cy7P, the energy gap between the HOMO and LUMO of cyanine decreased further, allowing electrons in the HOMO of TPE to be transferred to the HOMO of Cy7P (Fig. 3b). This process is consistent with the acceptor-excited photoinduced electron transfer (a-PET) mechanism [48-50]. As a result, TPE functioned neither rotor nor AIE-effect in Cy7PT. Our data revealed that tuning the frontier orbital energy of the cyanine chromophore via the methine-chain length of cyanines allowed effectively control of the TPE's behavior in TPE-cyanine systems. In summary, only the rotor function without the AIE effect was endowed in Cy5PT by TPE, which offered an effective way to eliminate the interference of background signal enhancement from AIEgen aggregation in an aqueous rotor-imaging system.

  Cholesterol, an essential lipid in the plasma membrane, directly influences membrane fluidity and elasticity [51,52]. Reduced membrane cholesterol levels contribute to cell rigidity, resulting in a stiffening effect. Therefore, we speculated that the rotor fluorophore, Cy5PT, could be employed to detect the changes in cell and tissue elasticity. As proof-of-concept, we treated 4T1 cells with methyl-β-cyclodextrin (MeβCD), a compound that augments cell stiffness by depleting cholesterol in the membrane lipid bilayer. Initially, we explored the impact of cholesterol modulation on the dynamics of F-actin, a critical component of the actin cytoskeleton that shapes cell structure and governs mechanical properties [12]. Our results showed that F-actin content was approximately two-fold higher in the presence of MeβCD compared to native cells, as observed via confocal laser scanning microscopy (CLSM) (Figs. 4a and b). Moreover, following membrane cholesterol depletion by MeβCD, F-actin underwent reorganization into prominent and long filaments, forming a robust network throughout the cells. This organization of F-action contributes to increased cellular rigidity and reduced intracellular fluidity, potentially facilitating the activation of rotor fluorophores.

  Figure 4

  

  Figure 4.  Cholesterol level controls rotor-imaging in 4T1 cells. (a) CLSM images of F-actin in MeβCD-treated or native 4T1 cells. 4T1 cells were native or pre-treated with MeβCD (cholesterol capturer, 5 mmol/L, 1.5 h). (b) Qualitative analysis of fluorescence in a. Native vs. MeβCD: P = 0.0008. (c) CLSM imaging of 4T1 cell incubated with different probes (5 µmol/L) for 1.5 h Cell nuclei were stained with Hoechst. CLSM images of cells were recorded at λ_ex = 405 nm for Hoechst (blue), λ_ex = 640 nm for NIR (red) dyes. (d) Signal intensity ratios of MeβCD/Native for Cy3PT, Cy5PT and Cy7PT in c. Cy3PT vs. Cy5PT, P = 0.0084; Cy5PT vs. Cy7PT, P = 0.0105. Data are presented as mean values ± SD from three separate measurements and analyzed by two-sided student's t-test. n = 3 independent samples per group. P < 0.05, **P < 0.01, ***P < 0.001. Fluorescence images were captured at λ_ex/λ_em: 561/570 nm for Cy3PT, λ_ex/λ_em: 640/665 nm for Cy5PT and λ_ex/λ_em: 640/767 nm for Cy7PT.

  DownLoad: Full-Size Img PowerPoint

  Next, the 4T1 cells were incubated with the TPE-cyanines (Cy3PT, Cy5PT, and Cy7PT) to evaluate whether their fluorescence responded to changes in cell mechanics. Live-cell imaging was performed using CLSM. As shown in Fig. S22 (Supporting information), negligible changes in fluorescence signals were observed after MeβCD treatment for the three substrate cyanines Cy3P, Cy5P and Cy7P, respectively. However, when incubating Cy5PT in cells, MeβCD-treated cells (5 mmol/L for 1.5 h) displayed a notable increase in fluorescence by CLSM, compared to the native cells (Fig. 4c and Fig. S24 in Supporting information). Qualitative analysis of fluorescence in Fig. 4c demonstrated that Cy5PT-treated cells showed a ca. 2-fold increase in fluorescence intensity in the presence of MeβCD than native cells (Fig. 4d). In contrast, for Cy3PT and Cy7PT, did not show statistically significant difference following MeβCD treatment (Fig. 4d, Figs. S23 and S25 in Supporting information). Although Cy3PT exhibited strong fluorescence signals in cells, the AIE-effect interfered with its rotor-imaging capability. In contrast, Cy5PT can overcome this limitation as the high AIE background can be effectively minimized. Consequently, Cy5PT has emerged as a promising rotor fluorophore for studying cell and tissue elasticity.

  Encouraged by these in vitro results, we then validated the use of Cy5PT by in vivo imaging, along with Cy3PT and Cy7PT, in 4T1 tumor-bearing mouse models. All animal experiments were performed under the guidelines and approved by with the Animal Experimental Welfare & Ethical Inspection Committee (Laboratory Animal Center of Fourth Military Medical University, China). The mice were administered either MeβCD to alter their tumor cholesterol levels or untreated (native group), followed by probes intratumorally (Fig. 5a). Initially, shear rheology experiments in 4T1 tumors were performed to investigate whether cholesterol levels influenced tumor elasticity, where the elastic storage modulus is a physical index for evaluating tissue rigidity [53]. Fig. 5b illustrated that the values of the storage modulus in MeβCD-treated tumors (ca. 1.9 kPa) were significantly higher than those in native tumors (ca. 0.9 kPa). Furthermore, histological staining with the cholesterol-specific fluorescent dye, Filipin Ⅲ, revealed higher cholesterol content in the native tumor tissues than in MeβCD-treated tumor tissues (Fig. 5c). Meanwhile, no statistically significant difference in cholesterol levels was observed in the paired muscle tissues (Figs. S26 and S27 in Supporting information). A positive correlation (R = 0.74) was found between the storage modulus and Filipin Ⅲ in tumors (Fig. 5d). The results emphasize the critical role of cholesterol in tumor elasticity, indicating that cholesterol depletion by MeβCD can enhance tumor rigidity.

  Figure 5

  

  Figure 5.  NIR imaging for 4T1 tumor-bearing mice after intratumor injection of probes Cy3PT, Cy5PT and Cy7PT. (a) Timeline of MeβCD, probe injection and NIR imaging. Mice were pre-injected with MeβCD twice (0.2 mg/16–18 g, dissolved in 25 µL of PBS) or none, followed by injection of probes (Cy5PT and Cy7PT: 0.2 nmol/16–18 g, Cy3PT: 1 nmol/16–18 g, dissolved in 25 µL of PBS) later. (b) Storage modulus in 4T1 tumors measured by shear rheology experiments. Native vs. MeβCD, P < 0.001. (c) Fluorescence analysis of cholesterol level in tumor tissues stained with Filipin Ⅲ (shown in blue color). (d) Correlation between fluorescence intensity of Filipin Ⅲ and storage modulus values in tumors. Data are presented as mean values ± SD and analysed by two-sided student's t-test. R² and P were derived using a simple linear regression model. The gray error bands show the 95% confidence intervals of the fitted line by two-tailed Student's t-test. (e) NIR imaging was performed at various timepoints (0, 0.08, 1, 2, 4, 8, 12, 24, 48 h) intratumor-injection of probes. (f) Fluorescence signal intensity in tumor tissue over time of Cy3PT, Cy5PT or Cy7PT group, n = 3. Cy3PT: native vs. MeβCD, P = 0.2329; Cy5PT: native vs. MeβCD, P = 0.0123; Cy7PT: native vs. MeβCD, P = 0.9784. (g) Correlation between fluorescence intensity of Cy5PT and storage modulus values in tumors. (h) Fluorescence intensity correlation between Cy5PT and cholesterol levels. P < 0.05, **P < 0.01. ns, not significant. Mice NIR images were recorded at λ_ex/λ_em: 560/620 nm for Cy3PT, λ_ex/λ_em: 660/710 nm for Cy5PT and λ_ex/λ_em: 680/790 nm for Cy7PT.

  DownLoad: Full-Size Img PowerPoint

  Subsequently, the region of interest (ROI) was continuously monitored for 24 h to assess the tumor fluorescence (Fig. 5e). After 1 h intratumoral injection of Cy5PT, the tumors in the MeβCD-treated group exhibited a 2.4-fold increase in fluorescence intensity compared to the native group at 1 h post-injection (HPI). This elevated fluorescence was sustained and remained approximately 2-fold higher up to 48 HPI (Fig. 5f). This observation suggests that the rotor effect of Cy5PT is reactivated by the enhanced tumor rigidity. Consistent with the in vivo data, Cy5PT fluorescence signals were positively correlated with the storage modulus (Fig. 5g, R = 0.89) and relative cholesterol levels (Fig. 5h, R = 0.89). Conversely, the MeβCD-treated groups exhibited negligible changes in Cy7PT signals in contrast to the native tumors (Fig. 5f).

  Notably, when administering the same low dosage of Cy3PT (0.2 nmol) as Cy5PT, we observed a negligible fluorescence signal at the tumor sites (Fig. S28 in Supporting information). However, upon increasing the dosage to 1 nmol (Fig. 5e), Cy3PT exhibited approximately 1.5-fold higher signal in the MeβCD group compared to the native tumor at 2 HPI. These findings suggested a significant contribution of the AIE-effect of Cy3PT to the observed signal, even at low dosages. Collectively, these findings indicated that Cy5PT can serve as a sensitive TPE-rotor cyanine probe for in vivo near-infrared (NIR) imaging, effectively minimizing the aggregation background. More broadly, our rotor imaging strategy suggests that alterations driven by cholesterol levels in the tumor microenvironment can significantly influence the rigidity of solid tumors, both at the cellular and in vivo levels.

  In summary, we developed a novel TICT/PET strategy to modulate the functions of TPE within cyanine scaffolds, allowing TPE to function as a rotor or exhibit AIE properties. By tuning the methine-chain length from Cy3 to Cy5 to Cy7, we demonstrated that the frontier energies of cyanines could regulate the function of TPE. In the Cy3PT scaffold, the high LUMO energy of cyanine inhibited the TICT from TPE to the cyanine skeleton, resulting in exhibiting AIEgen properties. Extending methine-chain length minimized the gap ΔE_H–L (DA) for Cy5PT, enabling ICT mechanism. This adjustment allowed the TPE to function solely as a rotor without the aggregation activation of Cy5PT. Further reduction of the energy gap led to the a-PET mechanism in Cy7PT, causing TPE to exhibit neither rotor nor AIE-effect.

  In vitro results demonstrated that Cy5PT was activated in the rotor-restricted state, showing a 90-fold enhancement in glycerol–water mixtures. This excellent rotor effect endowed Cy5PT with the ability to respond to changes in the cellular cytoskeleton and elasticity altered by cholesterol levels. Consequently, Cy5PT can serve as a sensitive TPE-rotor cyanine probe for in vivo NIR imaging, effectively minimizing the aggregation-induced background interference. The in vivo results further revealed that the fluorescence signal of Cy5PT positively correlated with the storage modulus values in tumors. Our rotor imaging strategy indicated that cholesterol-driven alterations in the tumor microenvironment could significantly influenced the rigidity of solid tumors. We anticipate that further advancements in sensitive TPE rotor visualization will provide a novel avenue for understanding the biophysical behavior of tumors.

  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.

  CRediT authorship contribution statement

  Xianghan Zhang: Writing – original draft, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation. Yuan Qin: Methodology, Formal analysis, Data curation. Huaicong Zhang: Methodology, Formal analysis, Data curation. Yutian Cao: Validation, Methodology. Haixing Zhu: Validation, Methodology. Yingdi Tang: Visualization, Validation. Zimeng Ma: Validation. Zehua Li: Validation. Jialin Zhou: Visualization. Qunyan Dong: Visualization. Peng Yang: Formal analysis. Yuqiong Xia: Writing – review & editing. Zhongliang Wang: Writing – review & editing, Supervision, Funding acquisition, Conceptualization.

  Acknowledgments

  This work was supported by National Natural Science Foundation of China (Nos. 32371433 and W2411083), the National Key Research and Development Program of China (No. 2022YFB3203800), Guang Dong Basic and Applied Basic Research Foundation (No. 2023A1515030207), Key Research and Development Program of Shaanxi (No. 2024SF2-GJHX-30), Innovation Capability Support Program of Shaanxi (No. 2022TD-52), Dual-chain Integration Special Program of Qin Chuang Yuan Construction (No. 23LLRH0070), Xidian University Specially Funded Project for Interdisciplinary Exploration (Nos. TZJH2024035, TZJH2024031).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110715.

  
  
• 1. [1]

     G. Obaid, J.P. Celli, M. Broekgaarden, et al., Nat. Rev. Bioeng. 2 (2024) 752– 769. doi: 10.1038/s44222-024-00196-z

  2. [2]

     C. Chen, X. Zhang, Z. Gao, G. Feng, D. Ding, Nat. Protoc. 19 (2024) 2408–2434. doi: 10.1038/s41596-024-00990-4

  3. [3]

     H. Li, B. Jin, Y. Wang, et al., Adv. Mater. 35 (2023) 2210085. doi: 10.1002/adma.202210085

  4. [4]

     X.T. Gong, J. Zhuang, K.C. Chong, et al., Adv. Mater. 36 (2024) 2402853. doi: 10.1002/adma.202402853

  5. [5]

     H. Zhao, N. Li, C. Ma, et al., Chin. Chem. Lett. 34 (2023) 107699. doi: 10.1016/j.cclet.2022.07.042

  6. [6]

     Z. Zhao, W. He, B.Z. Tang, Acc. Mater. Res. 2 (2021) 1251–1260. doi: 10.1021/accountsmr.1c00202

  7. [7]

     C. Chen, X. Ni, H.W. Tian, et al., Angew. Chem. Int. Ed. 59 (2020) 10008–10012. doi: 10.1002/anie.201916430

  8. [8]

     N. Wang, H. Yao, Q. Tao, et al., Chin. Chem. Lett. 33 (2022) 252–256. doi: 10.1016/j.cclet.2021.06.092

  9. [9]

     X. Cheng, Y. Pu, S. Ye, et al., Adv. Mater. 36 (2024) 2305779. doi: 10.1002/adma.202305779

  10. [10]

      R. Miao, J. Li, C. Wang, et al., Adv. Sci. 9 (2022) 2104609. doi: 10.1002/advs.202104609

  11. [11]

      M.K. Kuimova, S.W. Botchway, A.W. Parker, et al., Nat. Chem. 1 (2009) 69–73. doi: 10.1038/nchem.120

  12. [12]

      I.A. Ioannou, N.J. Brooks, M.K. Kuimova, Y. Elani, JACS Au 4 (2024) 2041–2049. doi: 10.1021/jacsau.4c00237

  13. [13]

      X. Peng, Z. Yang, J. Wang, et al., J. Am. Chem. Soc. 133 (2011) 6626–6635. doi: 10.1021/ja1104014

  14. [14]

      L. Hao, Z.W. Li, D.Y. Zhang, et al., Chem. Sci. 10 (2019) 1285–1293. doi: 10.1039/C8SC04242J

  15. [15]

      M. Paez-Perez, M.K. Kuimova, Angew. Chem. Int. Ed. 63 (2024) e202311233. doi: 10.1002/anie.202311233

  16. [16]

      M. Paez-Perez, I. Lopez-Duarte, A. Vysniauskas, N.J. Brooks, M.K. Kuimova, Chem. Sci. 12 (2021) 2604–2613. doi: 10.1039/D0SC05874B

  17. [17]

      C.H. Wolstenholme, H. Hu, S. Ye, et al., J. Am. Chem. Soc. 142 (2020) 17515–17523. doi: 10.1021/jacs.0c07245

  18. [18]

      S. Ye, H. Zhang, J. Fei, C.H. Wolstenholme, X. Zhang, Angew. Chem. Int. Ed. 60 (2021) 1339–1346. doi: 10.1002/anie.202011108

  19. [19]

      J.E. Chambers, N. Zubkov, M. Kubánková, et al., Sci. Adv. 8 (2022) eabm2094. doi: 10.1126/sciadv.abm2094

  20. [20]

      H. Li, W. Shi, X. Li, et al., J. Am. Chem. Soc. 141 (2019) 18301–18307. doi: 10.1021/jacs.9b09722

  21. [21]

      Q. Chen, H. Fang, X. Shao, et al., Nat. Commun. 11 (2020) 6290. doi: 10.1038/s41467-020-20067-6

  22. [22]

      J. Liu, W. Zhang, C. Zhou, et al., J. Am. Chem. Soc. 144 (2022) 13586–13599. doi: 10.1021/jacs.2c03832

  23. [23]

      S. Wang, W.X. Ren, J.T. Hou, et al., Chem. Soc. Rev. 50 (2021) 8887–8902. doi: 10.1039/D1CS00083G

  24. [24]

      J. Yin, M. Peng, W. Lin, Anal. Chem. 91 (2019) 8415–8421. doi: 10.1021/acs.analchem.9b01293

  25. [25]

      F. Würthner, Angew. Chem. Int. Ed. 59 (2020) 14192–14196. doi: 10.1002/anie.202007525

  26. [26]

      L.L. Yang, H. Wang, J. Zhang, et al., Nat. Commun. 15 (2024) 999. doi: 10.1038/s41467-024-45271-6

  27. [27]

      J. Li, J. Wang, H. Li, et al., Chem. Soc. Rev. 49 (2020) 1144–1172. doi: 10.1039/C9CS00495E

  28. [28]

      J. Mei, Y. Hong, J.W.Y. Lam, et al., Adv. Mater. 26 (2014) 5429–5479. doi: 10.1002/adma.201401356

  29. [29]

      H. Qian, M.E. Cousins, E.H. Horak, et al., Nat. Chem. 9 (2017) 83–87. doi: 10.1038/nchem.2612

  30. [30]

      Y. Yang, X. Su, C.N. Carroll, I. Aprahamian, Chem. Sci. 3 (2012) 610–613. doi: 10.1039/C1SC00658D

  31. [31]

      D. Tu, J. Zhang, Y. Zhang, et al., J. Am. Chem. Soc. 143 (2021) 11820–11827. doi: 10.1021/jacs.1c05647

  32. [32]

      X. Zhang, J. Gao, Y. Tang, et al., Nat. Commun. 13 (2022) 3513. doi: 10.1038/s41467-022-31136-3

  33. [33]

      Q. Jia, R. Zhang, Y. Wang, et al., Sci. Bull. 67 (2022) 288–298. doi: 10.1016/j.scib.2021.11.003

  34. [34]

      Y. Zhang, C. Yan, Q. Zheng, et al., Angew. Chem. Int. Ed. 60 (2021) 9553–9561. doi: 10.1002/anie.202017349

  35. [35]

      J.L. Tanyi, L.M. Randall, S.K. Chambers, et al., Clin. Oncol. 41 (2023) 276–284.

  36. [36]

      R. Voelker, JAMA 327 (2022) 27.

  37. [37]

      G.T. Kennedy, F.S. Azari, E. Bernstein, et al., JAMA Surg. 156 (2021) 1043–1050. doi: 10.1001/jamasurg.2021.3757

  38. [38]

      S. Blair, M. Garcia, T. Davis, et al., Sci. Transl. Med. 13 (2021) eaaw7067. doi: 10.1126/scitranslmed.aaw7067

  39. [39]

      A.N. Ramya, M.M. Joseph, J.B. Nair, et al., ACS Appl. Mater. Interfaces 8 (2016) 10220–10225. doi: 10.1021/acsami.6b01908

  40. [40]

      D.T. Jayaram, S. Ramos-Romero, B.H. Shankar, et al., ACS Chem. Biol. 11 (2015) 104–112.

  41. [41]

      S. Chen, Y. Hong, Y. Zeng, et al., Chem. Eur. J. 21 (2015) 4315–4320. doi: 10.1002/chem.201405658

  42. [42]

      S. Chen, Y. Hong, Y. Liu, et al., J. Am. Chem. Soc. 135 (2013) 4926–4929. doi: 10.1021/ja400337p

  43. [43]

      X. Zhang, B. Wang, Y. Xia, et al., ACS Appl. Mater. Interfaces 10 (2018) 25146–25153. doi: 10.1021/acsami.8b07727

  44. [44]

      M. Fang, S. Xia, J. Bi, et al., Chem. Commun. 54 (2018) 1133–1136. doi: 10.1039/C7CC08986D

  45. [45]

      D. Ding, K. Li, B. Liu, B.Z. Tang, Acc. Chem. Res. 46 (2013) 2441–2453. doi: 10.1021/ar3003464

  46. [46]

      H. Niu, J. Liu, H.M. O’Connor, et al., Chem. Soc. Rev. 52 (2023) 2322–2357. doi: 10.1039/D1CS01097B

  47. [47]

      H. Miyasaka, Acc. Chem. Res. 46 (2013) 248–257. doi: 10.1021/ar300102t

  48. [48]

      W. Chi, J. Chen, W. Liu, et al., J. Am. Chem. Soc. 142 (2020) 6777–6785. doi: 10.1021/jacs.0c01473

  49. [49]

      N. Meher, P.K. Iyer, Angew. Chem. Int. Ed. 57 (2018) 8488–8492. doi: 10.1002/anie.201802842

  50. [50]

      X. Hu, Z. Fang, F. Sun, et al., Angew. Chem. Int. Ed. 63 (2024) e202401036. doi: 10.1002/anie.202401036

  51. [51]

      K. Lei, A. Kurum, M. Kaynak, et al., Nat. Biomed. Eng. 5 (2021) 1411–1425. doi: 10.1038/s41551-021-00826-6

  52. [52]

      S. Chakraborty, M. Doktorova, T.R. Molugu, et al., Proc. Natl. Acad. Sci. U. S. A. 117 (2020) 21896–21905. doi: 10.1073/pnas.2004807117

  53. [53]

      W. Fan, K. Adebowale, L. Váncza, et al., Nature 626 (2024) 635–642. doi: 10.1038/s41586-023-06991-9

• 

  Figure 1  Mechanism for AIEgen, rotor-specific AIEgen and ACQgen of TPE-cyanines under theoretical calculations. The figure illustrates tuning cyanine chromophore via the methine-chain length to control TPE's behavior in TPE-cyanine systems.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Spectra properties of TPE-cyanines. (a) Structures of TPE-cyanines Cy3PT, Cy5PT, and Cy7PT. (b) Normalized emission spectra of TPE-cyanines Cy3PT, Cy5PT, and Cy7PT in monomer (pure water) and aggregates (PAA 0.5 wt% in water) in aqueous environment. (c) Fluorescence emission spectra of Cy3PT, Cy5PT, and Cy7PT in different volume fractions of glycerol/H₂O (f_Gly) solutions. (d) Viscosity sensitivity of Cy3PT, Cy5PT, and Cy7PT in Gly/H₂O mixtures. The linear relationship of double logarithmic plots of fluorescence intensity (F) vs. viscosity (η).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Mechanism for AIE, rotor-specific AIE and ACQ phenomenon of TPE-cyanines under theoretical calculations. (a) Molecular orbitals (LUMOs and HOMOs) of Cy3PT, Cy5PT and Cy7PT calculated by PCM/TD-DFT method. The PCM model is utilized for calculation in aqueous solution. (b) Band gaps (ΔE_H–L and ΔE_H–H) between TPE and substrate cyanines calculated by TD-DFT method. Hydrogen atoms and PEG chain substituents are omitted for clarity.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Cholesterol level controls rotor-imaging in 4T1 cells. (a) CLSM images of F-actin in MeβCD-treated or native 4T1 cells. 4T1 cells were native or pre-treated with MeβCD (cholesterol capturer, 5 mmol/L, 1.5 h). (b) Qualitative analysis of fluorescence in a. Native vs. MeβCD: P = 0.0008. (c) CLSM imaging of 4T1 cell incubated with different probes (5 µmol/L) for 1.5 h Cell nuclei were stained with Hoechst. CLSM images of cells were recorded at λ_ex = 405 nm for Hoechst (blue), λ_ex = 640 nm for NIR (red) dyes. (d) Signal intensity ratios of MeβCD/Native for Cy3PT, Cy5PT and Cy7PT in c. Cy3PT vs. Cy5PT, P = 0.0084; Cy5PT vs. Cy7PT, P = 0.0105. Data are presented as mean values ± SD from three separate measurements and analyzed by two-sided student's t-test. n = 3 independent samples per group. P < 0.05, **P < 0.01, ***P < 0.001. Fluorescence images were captured at λ_ex/λ_em: 561/570 nm for Cy3PT, λ_ex/λ_em: 640/665 nm for Cy5PT and λ_ex/λ_em: 640/767 nm for Cy7PT.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  NIR imaging for 4T1 tumor-bearing mice after intratumor injection of probes Cy3PT, Cy5PT and Cy7PT. (a) Timeline of MeβCD, probe injection and NIR imaging. Mice were pre-injected with MeβCD twice (0.2 mg/16–18 g, dissolved in 25 µL of PBS) or none, followed by injection of probes (Cy5PT and Cy7PT: 0.2 nmol/16–18 g, Cy3PT: 1 nmol/16–18 g, dissolved in 25 µL of PBS) later. (b) Storage modulus in 4T1 tumors measured by shear rheology experiments. Native vs. MeβCD, P < 0.001. (c) Fluorescence analysis of cholesterol level in tumor tissues stained with Filipin Ⅲ (shown in blue color). (d) Correlation between fluorescence intensity of Filipin Ⅲ and storage modulus values in tumors. Data are presented as mean values ± SD and analysed by two-sided student's t-test. R² and P were derived using a simple linear regression model. The gray error bands show the 95% confidence intervals of the fitted line by two-tailed Student's t-test. (e) NIR imaging was performed at various timepoints (0, 0.08, 1, 2, 4, 8, 12, 24, 48 h) intratumor-injection of probes. (f) Fluorescence signal intensity in tumor tissue over time of Cy3PT, Cy5PT or Cy7PT group, n = 3. Cy3PT: native vs. MeβCD, P = 0.2329; Cy5PT: native vs. MeβCD, P = 0.0123; Cy7PT: native vs. MeβCD, P = 0.9784. (g) Correlation between fluorescence intensity of Cy5PT and storage modulus values in tumors. (h) Fluorescence intensity correlation between Cy5PT and cholesterol levels. P < 0.05, **P < 0.01. ns, not significant. Mice NIR images were recorded at λ_ex/λ_em: 560/620 nm for Cy3PT, λ_ex/λ_em: 660/710 nm for Cy5PT and λ_ex/λ_em: 680/790 nm for Cy7PT.

  下载: 全尺寸图片 幻灯片
•