Leveraging an Electron-Acceptor Engineering Strategy to Regulate Excitation Dynamics of Dyes for Devising Ideal Phototherapeutic Agents in Synergistic Photodynamic/Mild-Photothermal Tumor Therapy

Introduction

Cancer, a challenging disease that has seriously harmed humanity for many years, and its diagnosis and treatment have always been a hot research area. Photon-driven therapy utilizing organic dyes, in contrast to conventional treatments such as chemotherapy, radiotherapy, and surgery, offers a promising alternative with minimal toxicity to normal tissues and precise spatiotemporal control. This innovative approach has opened new avenues for cancer treatment.^1–3 For example, indocyanine green (ICG), an Food and Drug Administration (FDA)-approved dye, has shown potential in both diagnosing and treating tumors in clinical settings or trials, highlighting its versatility in biomedical applications. However, limitations such as low photostability and inefficient generation of reactive oxygen species (ROS) may hinder the efficacy of ICG-based photon-driven therapy for tumor treatment. Consequently, researchers are committed to developing robust phototherapy systems to overcome these challenges and advance the diagnosis and treatment of tumors.

Fluorescence imaging (FLI) is a common optically based diagnostic method that has advantages of sensitive response and precise spatiotemporal control. However, in practical scenarios, the limited penetration depth of FLI significantly hinders imaging resolution and accuracy due to variations in light absorption and scattering by tissues.^4–6 Fortunately, photoacoustic imaging (PAI) partially overcomes this limitation by providing higher spatial resolution. PAI, a crucial clinical imaging technology, involves the absorption of light by molecules, leading to thermal elastic expansion and nonradiative conversion of electromagnetic energy into mechanical energy to generate mechanical waves. Subsequently, an ultrasonic transducer detects the signal to create an observable image, enabling deeper tissue imaging.^7,8 Consequently, PAI has broader prospects in clinical applications.

Phototheranotics, which include photodynamic therapy (PDT) and photothermal therapy (PTT), are recognized as effective methods for treating tumors due to their precise and controllable spatiotemporal selectivity, lack of resistance, and noninvasiveness.^9,10 In PDT, upon exposure to light of a specific wavelength, the excited photosensitizer (*PS) can generate ROS through energy or electron transfer pathways, leading to direct ablation of tumor cells by damaging biological macromolecules.^11,12 Despite the advantages of high spatial-temporal accuracy and minimal invasiveness of PDT in trials, the therapeutic efficacy of the PS depends on the O₂ content in the tumor microenvironments to a large extent. Unfortunately, the hypoxic microenvironments in solid tumor tissues are not conducive to the therapeutic outcomes of PDT.^13,14 In PTT, a photothermal agent absorbs photon energy at a specific wavelength, transitions to an excited state, and then releases energy through nonradiative pathways, thereby increasing the temperature of the tumor environment to denature intracellular proteins or destroy the plasma membrane for tumor ablation. Unlike PDT, PTT is not limited by the O₂ concentration in the tumor environment.^15,16 Furthermore, the heat generated during PTT can facilitate photothermal imaging (PTI) using a thermal imaging system, enabling real-time monitoring of the treatment process.¹⁷ However, traditional PTT may cause nonspecific heat damage to normal tissues surrounding the illuminated area due to high light power density or excessive heat production. Additionally, the upregulation of heat shock proteins in tumor cells in response to heat stress can limit the efficacy of PTT to a certain extent.^18,19 Therefore, the limitations of the single-modal strategy of treatment in clinical settings, as described above, often result in unsatisfactory therapeutic outcomes. Combining the multimodal therapeutic functions and imaging advantages of various modalities in a smart dye offers a promising strategy to address these shortcomings. By integrating the benefits of PDT and PTT, this approach can enhance the treatment efficacy. On the one hand, PTT can increase the tissue temperature within the treatment area, thus expanding blood vessels, increasing blood flow, and enhancing tissue oxygenation to boost ROS production by the PS during PDT. Simultaneously, PDT can eliminate the incomplete ablation of tumor cells due to the heat shock effect during PTT.^20–23 The synergistic effects of PDT and PTT eliminate the need for high-power laser irradiation to ensure PTT effectiveness or the maintenance of high temperatures (≥50 °C) during treatment, which can damage normal tissues. This approach also prevents the incomplete killing of cancer cells and tumor recurrence in deep tissues caused by low treatment temperatures (≤45 °C) in mild-photothermal therapy (MPTT).^24–27

Therefore, the abovementioned requirements inspired us to design smart phototherapeutic dyes with multimodal diagnostic and therapeutic capabilities, which posed significant challenges. According to the Jablonski diagram, first, after absorbing photon energy from the ground state (S₀) to the excited state (*S₁), the dye molecule can dissipate energy through a nonradiative pathway, enabling PTT, PTI, and PAI. Second, in the *T₁ state, the dye can convert O₂ into ROS via energy or electron transfer pathways, endowing the molecule with the PDT capabilities. However, these two processes often compete, leading to dyes exhibiting predominantly only one property. Hence, it is crucial to optimize dye structures to strike a balance that allows for both effective PTT, PTI, and PAI and PDT functionalities.^21,22 In addition, most of the developed organic PSs/photothermal agents tend to aggregate in aqueous solutions due to poor solubility, significantly reducing their imaging and therapeutic efficacy because of reducing the fluorescence quantum yield and ROS yield of dyes.^28–30 Conversely, dyes exhibiting aggregation-induced emission (AIE) characteristics demonstrate enhanced performance in the aggregated state. For example, the AIE effect significantly influences the ROS generation of PSs.^31–34 Despite numerous molecules with AIE properties being identified for tumor ablation primarily through single-modal PDT or PTT, dyes with optimal synergistic functions and multimodal imaging capabilities remain scarce (Scheme 1).

• Download figure
• Download PowerPoint

To address the above issues, in this study, we utilized an acceptor engineering approach to develop a series of near-infrared (NIR) activated dyes exhibiting AIE properties. The introduction of diphenylamine into xanthene structures resulted in dyes with enhanced AIE characteristics, PTT, PDT, and PAI potential, due to increased freely rotatable single bonds and asymmetric structures. Interestingly, the variation in the number of cyano groups in the dyes allowed for modulation of their excitation wavelengths, PDT efficacy, and PTT capabilities. Experimental results revealed that Hcy-ON demonstrated high ROS yield and heat production capability upon 760 nm laser irradiation. Molecular theory calculations indicated that Hcy-ON exhibited a significant spin–orbit coupling matrix element (SOCME) value $⟨{\text{S}}_1|{\widehat{H}}_{\text{SOC}}|{\text{T}}_3⟩$and the lowest gap between S₁ and T₁ energy levels of 0.678 eV, which correlated with its strongest ROS generation ability. Furthermore, analysis of the singlet–triplet (S–T) energy gap, electronic transition mechanisms, root mean squared displacement (RMSD) values, and Huang–Rhys factor supported the exceptional photothermal performance of Hcy-ON. In vitro investigations demonstrated that Hcy-ON possessed enhanced dual-modality imaging capabilities for PAI and PTI. Remarkably, in vivo application of the Hcy-ON dye in the combined PTT/PDT therapy, followed by a single treatment session and 760 nm laser irradiation, resulted in highly efficient MPTT and favorable prognostic outcomes. Therefore, the multifunctional dye developed in this study offers a promising solution to overcome the limitations associated with single-modality phototherapeutic agents.

Experimental Methods

In this study, all reagents were obtained from commercial sources. Nuclear magnetic resonance (NMR) spectra of all compounds were detected by Bruker Avance II 400 spectrometers (Karlsruhe, Baden-Württemberg, Switzerland). Mass spectrometric [electrospray ionization mass spectrometry (ESI-MS)] data were obtained with Synapt G2-Si HDMS (Framingham, Massachusetts, United States). Atmospheric pressure chemical ionization mass spectrometry (APCI-MS) data were obtained with Agilent 6130 (Santa Clara, California, United States). The confocal laser scanning microscope images were obtained on an Olympus FV3000-IX81 confocal laser scanning microscope (Tokyo, Japan). Photothermal testing used an FLIR E6-XT infrared thermal camera (Wilsonville, Oregon, United States), and photoacoustic (PA) testing used an InVision 128 MSOT system (Munich, Bavaria, Germany). Apoptotic cells stained with Annexin V-FITC/PI were measured by flow cytometry BD Accuri C6 plus (Franklin Lakes, New Jersey, United States). In addition, the photothermal conversion efficiency (PCE) value (η) was calculated based on previous literature reports.^35,36 For more dye characterization and experimental details, please refer to the Supporting Information. All the animal experiments involved in this study were conducted in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (8th edition, 2011), and approved by the Ethics Committee of Dalian University of Technology (Dalian, Liaoning Province, China).

Results and Discussion

Synthesis and spectral characterization

The changes in electron concentration of the lowest unoccupied molecular orbital (LUMO) have a more significant impact on the electron bandgap than the wide electron delocalization of the highest occupied molecular orbital (HOMO), leading to a greater effect of the electron-acceptor portion in dye molecules.³⁷ Furthermore, the intense stretching vibration of cyano groups is less affected by the external environment compared to intramolecular rotation, thereby favoring intramolecular heat generation in aggregated states.³⁸ Thus, by employing a tuning electron-acceptor engineering strategy for dyes, as shown in Figure 1, electron-acceptor segments containing different numbers of cyano groups were incorporated into molecules. This was achieved through a Knoevenagel condensation reaction to obtain three donor-π-acceptor (D-π-A) structured analogs (namely Hcy-OO, Hcy-ON, and Hcy-NN) with AIE characteristics. All compounds underwent comprehensive characterization using ¹H-NMR, ¹³C-NMR, high-resolution mass spectrometry, and APCI-MS techniques ( Supporting Information Figures S1–S11).

• Download figure
• Download PowerPoint

In general, using a π-bridge to connect an electron-donor (D) and an electron-acceptor (A) to form a D-π-A structure can reduce the electronic energy level gap and promote the intersystem crossing (ISC) process. This leads to a red shift in wavelength and enhances the generation yield of ROS.^39,40 Building on these insights, we initially investigated the photophysical properties of these compounds in different solvents. Figure 2a–c and Supporting Information Figure S12 illustrate a noticeable red shift in both absorption and fluorescence emission peaks from Hcy-OO to Hcy-NN. With an increase in the number of cyano groups in the electron-acceptor segments, the electron-withdrawing capacity is enhanced. It is noteworthy that Hcy-ON and Hcy-NN exhibit exceptional absorption properties within the 700–800 nm range, which not only satisfies the requirements for in vivo PAI systems but also enhances tissue penetration of light, thereby improving treatment efficacy.^21,22 Moreover, owing to the presence of freely rotatable diphenylamine groups in the dihydroxanthene-based scaffold, we further explored their AIE characteristics in tetrahydrofuran (THF) solutions with phosphate-buffered saline (PBS) fractions (f_PBS) varying from 0 to 100%. As anticipated, with the increase in the PBS content, the fluorescence intensities of the three molecules gradually increased (Figure 2d). This phenomenon can be attributed to the fact that the energy dissipation of dye excited states is primarily dominated by active intramolecular motion in the solution phase, while such motion is restricted during molecular aggregation. This observation distinctly showcases the AIE properties of the aforementioned dyes.

• Download figure
• Download PowerPoint

To further evaluate the ROS generation capacity of the three dyes in solution, 2′,7′-dichlorodihydrofluorescein (DCFH) was used as an ROS fluorescence indicator. As shown in Figure 2e,f and Supporting Information Figure S13, the dyes exhibit different ROS generation capabilities, with Hcy-ON notably enhancing the fluorescence intensity of DCFH the most after exposure to the 760 nm laser. Additionally, the emission intensity of DCFH exceeded 200 times, indicating superior ROS generation efficiency of Hcy-ON compared to the other dyes. Moreover, commercial PSs Chlorin e6 (Ce6) and Rose Bengal (RB) showed minimal fluorescence enhancement under identical lighting conditions owing to their short excitation wavelengths, further demonstrating the advantages of the designed dyes in longer wavelength phototherapy. Furthermore, 1,3-diphenylisobenzofuran (DPBF), hydroxyphenyl fluorescein (HPF), and dihydrorhodamine 123 (DHR 123) were used as indicators for ¹O₂, •OH, and O₂^•−, respectively, to further evaluate the types of ROS generated by the designed compounds (Figure 2g–i and Supporting Information Figure S14). These results clearly indicate that Hcy-ON not only effectively produced ¹O₂ but also exhibited superior •OH generation capacity, indicating dual characteristics of Hcy-ON as a type I and II PS. As a result, Hcy-ON shows promise in addressing hypoxia in PDT, rendering it more suitable for treating hypoxic solid tumors.⁴¹

Photothermal and photoacoustic evaluation in solvents

The temperature changes of Hcy-OO, Hcy-ON, and Hcy-NN solutions under the same power density were measured to evaluate their photothermal efficiency. As shown in Figure 3a,b, Hcy-ON (ΔT = 29.1 °C) and Hcy-NN (ΔT = 18.5 °C) significantly increased the solution temperature under continuous irradiation of the 760 nm laser (800 mW/cm²), while Hcy-OO exhibited weak photothermal performance under irradiation of a 671 nm laser (800 mW/cm²) owing to its short absorption wavelength. This difference may be attributed to the introduction of cyano groups, where the intramolecular vibrations and electron-withdrawing ability of –C≡N bonds play a crucial role in regulating the ISC that competes with nonradiative relaxation. Additionally, we observed that the photostability of Hcy-NN was poor during previous ROS testing. Even under low light power conditions (100 mW/cm²), its characteristic absorption peak was significantly reduced, and noticeable color changes in the solution before and after illumination could be observed with the naked eye ( Supporting Information Figure S15a). This observation may explain why Hcy-ON exhibits superior photothermal performance compared to Hcy-NN, despite the latter having more cyano groups.

• Download figure
• Download PowerPoint

Therefore, in the following analysis, we focused on assessing the photothermal efficiency of Hcy-ON. The results shown in Figure 3c,d indicate a positive correlation between the rate of temperature increase in the Hcy-ON aqueous solution and both its concentration and laser power intensity. This implies that, during PTT, precise regulation of the heat produced by Hcy-ON for MPTT is achievable. Notably, based on the heating–cooling profiles ( Supporting Information Figure S15b,c), the PCE of Hcy-ON can reach 49.91%. While the FDA-approved ICG dye exhibited a higher ΔT value under identical conditions (η = 72.7%), the photothermal stability of Hcy-ON surpassed that of ICG, maintaining the photothermal efficacy even after undergoing five heating–cooling cycles. In contrast, ICG underwent complete photodegradation and lost its heat-generating ability after the third exposure to the laser ( Supporting Information Figure S15d).

We incorporated diphenylamine into three dyes as an electron donor, along with multiple freely rotatable single bonds, which facilitated the thermal conversion of the molecule. The irregular conformations of these dyes increased the intermolecular distance. As a result, intramolecular rotations can be retained even in the aggregated state with loose intermolecular packing, thereby promoting heat generation in the aggregated state.²⁶ Moreover, we gradually substituted the carbonyl group of the molecules with electron-deficient malononitrile as the electron acceptor, resulting in dyes ranging from Hcy-OO to Hcy-ON and Hcy-NN. Compared with the intramolecular rotation, the vigorous stretching vibration of –C≡N was less affected by the external environment, thus favoring the intramolecular heat generation in the aggregated state.³⁸ Additionally, the changes in electron-concentrated LUMO had a more significant effect on electronic bandgaps than changes in the broadly electron-delocalized HOMO.³⁷ This emphasizes the influence of the electron-acceptor components on molecular properties. This highlights the value of introducing electron-withdrawing groups rich in –C≡N to enhance the thermogenic capacity of the dyes. To assess the heat-generating ability of these molecules with AIE properties through intramolecular motions in solution, various tests were conducted on the three molecules.

The nonradiative relaxation behavior of molecules upon reaching the excited state provides molecules with both photothermal and PA properties. Building on the excellent photothermal effect of Hcy-ON, we further investigated the PA properties of Hcy-OO, Hcy-ON, and Hcy-NN. PAI was performed in a biological tissue phantom. As shown in Figure 3e, Hcy-ON and Hcy-OO exhibited the highest and lowest PA intensity, respectively. Additionally, a linear relationship was observed between the PA signal intensity and the molecule concentration (Figure 3f). In conclusion, the excellent PA imaging capability of Hcy-ON results in dual-modal PTI/PAI.

Theoretical calculation of ROS generation and photothermal performance

Density functional theory (DFT) and time-dependent DFT (TD-DFT) calculations were performed to investigate variations in ROS generation and photothermal performance resulting from distinct photophysical properties of three dyes. As shown in Figure 4a, the donor and acceptor planes of Hcy-ON and Hcy-NN were relatively twisted, favoring their photothermal performance. In contrast, Hcy-OO exhibited inferior photothermal performance. Moreover, in the ground state, the LUMO of all three dyes is mainly located in the electron-acceptor region, while the HOMO exhibited delocalized electron distribution, causing a red shift in the absorption/emission spectra of the dyes. It is noteworthy that Hcy-NN (2.19 eV) and Hcy-ON (2.36 eV) featured a narrower HOMO–LUMO band gap, thereby further shifting their absorption toward the NIR range (650–950 nm). Conversely, the largest energy gap observed for Hcy-OO (2.60 eV) corresponded to the shortest absorption wavelength. These computational findings align with the absorption spectra of the three molecules, ranging from longer (780 nm) to shorter (660 nm) wavelengths (Figure 2a–c).

• Download figure
• Download PowerPoint

Next, to investigate their photothermal potentials, we calculated the optimized S₀ and S₁ geometries of Hcy-OO, Hcy-ON, and Hcy-NN, and the results are displayed in Figure 4b. These molecules exhibited significant geometric changes, with relatively high RMSD values. These observations indicate pronounced intramolecular mobility and robust nonradiative transition ability of the molecules, resulting in efficient photothermal conversion. To compare the nonradiative transition rates, we utilized Gaussian16 to calculate the Huang–Rhys factors for the S₁–S₀ transitions of the three molecules (Figure 4d). Notably, the analysis revealed that Hcy-ON exhibited the highest maximum Huang–Rhys factor (>70), signifying superior nonradiative transition efficiency.

We further studied the S–T energy gaps and electronic transition processes of the three dyes using DFT calculations and a simplified Jablonski energy level diagram. Additionally, we determined the spin–orbit coupling (SOC) rates with ORCA 5.0.4 to explore the variations in photothermal properties and ROS generation ability of the dyes. Notably, significant SOC was observed between the S₁–T₃ levels of Hcy-ON, as shown in Figure 4c. The SOCME value was $⟨{\text{S}}_1|{\widehat{H}}_{\text{SOC}}|{\text{T}}_3⟩=0.725{\text{cm}}^{-1}$ , with a minimal S₁–T₂ energy gap of only 0.02 eV. According to Fermi’s golden rule, the intersystem crossing rate (K_ISC) is directly proportional to the SOC rate (${\widehat{H}}_{\text{SOC}}$) and inversely proportional to the energy gap between the S and T levels. This indicates an efficient ISC, and the electrons reaching the T₃ and T₂ level subsequently undergo internal conversion (IC), dissipating energy as heat, and eventually transitioning to T₁. These results, in conjunction with Figure 4d, indicate that Hcy-ON exhibits superior photothermal performance. For Hcy-OO, the low ${\widehat{H}}_{\text{SOC}}$ value indicates poor ISC efficiency; consequently, only a small number of electrons transition from T₂ to T₁, generating heat through IC. However, the maximum Huang–Rhys factor for Hcy-OO exceeded that of Hcy-NN, contrary to experimental findings. We speculated that this discrepancy may be attributed to the heightened fluorescence intensity of Hcy-OO, leading to the dissipation of absorbed energy through radiative transitions and minimal efficiency in IC processes. It is noteworthy that although the nonradiative transition ability of Hcy-NN from S₁ to S₀ is slightly low, the S₁–T₃ energy gap is minimal (0.01 eV). The SOCME value $⟨{\text{S}}_1|{\widehat{H}}_{\text{SOC}}|{\text{T}}_2⟩=1.057{\text{cm}}^{-1}$ indicates that electron transitions from the S₁ level to the T₃ and T₂ levels are highly probable, generating more heat through IC before finally transitioning to the T₁ state. This, together with Figure 4d, confirms that Hcy-NN has marginally inferior photothermal performance compared to Hcy-ON. Moreover, Hcy-ON exhibited the lowest $\Delta {\text{E}}_{\text{S1}}{}_{\text{-T1}}$ value (0.678 eV), while those of Hcy-NN and Hcy-OO were 0.714 and 0.837 eV, respectively. Considering the ${\widehat{H}}_{\text{SOC}}$ value and ISC process analysis, it is evident that the ROS generation ability of Hcy-ON, Hcy-NN, and Hcy-OO decreased in that order. In summary, Hcy-ON has stronger cyano electron-withdrawing groups than Hcy-OO, as well as enhanced photostability, reduced electron cloud overlaps at the HOMO and LUMO energy levels, and a high RMSD value, compared to Hcy-NN. As a result, Hcy-ON exhibits superior intramolecular charge transfer (ICT), a more efficient ISC process, and a nonradiative transition/IC process. These results collectively confirm that Hcy-ON exhibits the strongest ROS generation ability and the best photothermal/PA performance, making it a multifunctional diagnostic and therapeutic dye with significant potential. Therefore, strategic modifications involving cyano groups with strong electron-withdrawing ability, along with ingeniously modified donor–acceptor structure and dye conformation, can lead to dye molecules with improved performance.

Cytotoxicity and ROS generation in vitro

Based on the results of testing three dye molecules in vitro, Hcy-ON has demonstrated significant potential in PTT/PDT synergistic therapy. Subsequently, we investigated the cytotoxicity of the three molecules on cancer cells under NIR light irradiation in vitro through the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay.

As shown in Figure 5a–c, Hcy-OO and Hcy-ON showed no significant toxicity under dark conditions; however, when the concentration exceeded 50 μg/mL, Hcy-NN exhibited significant toxicity, suggesting that only Hcy-OO and Hcy-ON had preferable biocompatibility. Under light exposure, only Hcy-ON exhibited excellent phototoxicity with an IC₅₀ value of 30.2 μg/mL. Real-time temperature monitoring in a 96-well plate ( Supporting Information Figure S16) revealed that at a concentration of 100 μg/mL, the maximum temperature under illumination could reach 56.6 °C, indicating satisfactory PTT efficacy in cells. Additionally, 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) imaging confirmed the production of ROS within cells. As shown in Figure 5d, MCF-7 cells incubated with Hcy-OO and Hcy-NN showed no significant fluorescence under illumination, whereas cells treated with Hcy-ON exhibited noticeable fluorescence. These imaging results align with the solution test results (Figure 2e,f), collectively verifying that Hcy-ON possessed remarkable ROS generation capability in live cells. Upon irradiation, the heat and ROS produced by Hcy-ON effectively eradicated tumor cells, highlighting its superior PTT/PDT synergistic killing potential among the three molecules in vitro.

• Download figure
• Download PowerPoint

The ablation effect of Hcy-ON on cancer cells was visually evaluated using a dual staining method. In this staining method, calcein acetate methyl ester (calcein-AM, green channel) and propidium iodide (PI, red channel) were used to distinguish live and dead cells via confocal microscopy. In Figure 5e, both the control group and the nonilluminated Hcy-ON group exhibited strong green fluorescence, indicating that the light dose used was relatively safe and Hcy-ON had good biocompatibility. Remarkably, upon co-culturing MCF-7 cells with Hcy-ON under 760 nm light, a significant increase in red fluorescence was observed in nearly all cells observed significant red fluorescence. This indicates cell death, as PI enters the damaged cell membrane and specifically binds to the DNA double helix released during nuclear fission. Based on previous test results, Hcy-ON was chosen for all subsequent tests owing to its exceptional photothermal ability, highest PA signal intensity, and strongest ROS generation ability.

Study on the mechanism of Hcy-ON-induced cell death

After confirming the potent cytotoxicity of Hcy-ON against cancer cells in vitro, the induced cell death pathway under NIR light irradiation was further elucidated. For this evaluation, Annexin V-FITC and PI was used to stain the cells. As shown in Figure 5f, no fluorescence was observed in the control, light, and Hcy-ON groups under confocal microscopy. In stark contrast, the “ Hcy-ON + Light” group exhibited significant green fluorescence on the cell membrane, indicating specific binding of Annexin V to exposed phosphatidylserine that flips to the outer membrane. Additionally, red fluorescence was observed on the nucleus, signifying membrane damage and nucleus rupture, as PI enters the cell and binds to free DNA. This dual fluorescence pattern indicates the complete disintegration of cells and an advanced stage of apoptosis, confirming that Hcy-ON induces cell death via NIR photon activation.

It is noteworthy that the decrease in mitochondrial membrane potential (MMP) is a common feature of cell apoptosis.⁴² Therefore, we used the MMP-dependent probe JC-1 to further investigate the mode of cell death. In healthy cells, the MMP is high, and JC-1 forms aggregates (JC-1 aggregates) in the mitochondrial matrix, emitting red fluorescence under confocal microscopy. When cells are damaged, the MMP depolarizes, resulting in a decrease in MMP. Consequently, JC-1 is unable to aggregate and instead exists in its monomer form (JC-1 monomer), which is associated with green fluorescence observed during confocal imaging. Using carbonyl cyanide 3-chlorophenylhydrazine (CCCP) as a control, which serves as a mitochondrial oxidative phosphorylation uncoupling agent, can alter the permeability of mitochondrial inner membrane to H⁺, resulting in the loss of MMP. As shown in Figure 5g, confocal imaging of MCF-7 cells co-cultured with JC-1 revealed red fluorescence in the control and Hcy-ON groups without light exposure, indicating the presence of JC-1 in an aggregated state and healthy cells. Upon NIR light irradiation, cells co-incubated with Hcy-ON exhibited green fluorescence, indicating that JC-1 existed in the monomeric form and damaged MMP, similar to the results observed in the CCCP group with apoptosis inducers. Subsequently, Annexin V-FITC/PI staining through flow cytometry demonstrated that MCF-7 cells co-cultured with Hcy-ON experienced significant apoptosis under 760 nm irradiation compared to the control group, while cells co-cultured solely with Hcy-ON remained undamaged (Figure 5h–j). These findings were aligned with the results of JC-1 staining and AV/PI staining. Overall, it can be concluded that Hcy-ON exhibited excellent biocompatibility, effectively targeting cancer cells and inducing apoptosis upon NIR laser activation.

In vivo multimodal imaging and synergistic PTT/PDT

Compared to FLI, PAI offers superior penetration depth, addressing the limitations of FLI in this aspect. Therefore, we used PAI to initially visualize tumor cells in vitro. MCF-7 cells were incubated with varying concentrations of Hcy-ON, followed by rinsing with PBS and replacing with fresh culture medium. Subsequently, the cell fluid and culture medium (control) were packaged into small tubes for PAI. In Figure 6a, clear PA signals were observed in cells treated with Hcy-ON under short pulse laser irradiation, along with distinct cell contours. These findings were consistent with the results obtained from phantom tests (Figure 3e,f), demonstrating a positive correlation between the intensity of the PA signal and the concentrations of Hcy-ON (Figure 6b). This confirms the excellent PAI ability of Hcy-ON in vitro.

• Download figure
• Download PowerPoint

After achieving promising initial results in vitro, driven by the efficient nonradiative transition/IC of Hcy-ON, we further evaluated the PA signal generation ability of Hcy-ON in vivo. Following intratumoral injection of Hcy-ON, PA signal production was monitored by capturing PA signal images at different time points. As shown in Figure 6c,d, no PA signal was detected at the tumor site before injection. Subsequently, PA signals became visible in the tumor region immediately postinjection; however, due to incomplete penetration, the PA signal intensity and coverage were not significant. Within 30 min, complete infiltration of Hcy-ON into the tumor resulted in an expanded PA imaging range and peak PA signal intensity. Strong PA signals persisted even 150 min after postinjection, indicating that Hcy-ON has the potential for prolonged in vivo PAI. Additionally, as shown in Figure 6e–g, under 600 mW/cm² illumination, the tumor area temperature marginally increased in the control group, whereas a significant temperature rise was evident in the illuminated group, which was clearly visible through a photothermal camera. It is noteworthy that at an optical power density of 300 mW/cm², the tumor area temperature reached 44.8 °C, escalating to 58.1 °C at 600 mW/cm², underscoring its exceptional PCE. Consequently, we opted for 300 mW/cm² as a relatively mild setting for subsequent MPTT tumor ablation. This combined PDT with lower temperature PTT approach not only circumvents incompletely tumor eradication but also mitigates nonspecific damage to neighboring tissues. In summary, Hcy-ON exhibits capability for multimodal imaging, including PTI and PAI, in MCF-7 tumor-bearing mice, along with synergistic PTT/PDT.

After confirming the efficacy of Hcy-ON dual-modal imaging and the MPTT treatment plan, we aimed to achieve synergistic PTT/PDT of human breast cancer (MCF-7) in mice using Hcy-ON. To ensure the enrichment of Hcy-ON in tumors, nude mice were treated with a single intratumoral injection, and tumor volume, mice weight, and tumor appearance data were recorded every 2 days. The treatment protocol, depicted in Figure 7a, involved a 20-min 760 nm laser (300 mW/cm², 360 J/cm²) irradiation following a 1 h intratumoral injection. By the second day post-treatment, mice exhibited scabbing and darkening of the tumor due to ablation. Subsequently, while tumors in the control and Hcy-ON groups grew rapidly, the “ Hcy-ON + Light” group demonstrated a significant tumor-inhibitory effect; that is, the tumors gradually faded away from black scabs and healed to a great extent (Figure 7b–d).

• Download figure
• Download PowerPoint

During the treatment period, the mice maintained stable weight, and no significant mutations were observed (Figure 7e). Subsequently, we further evaluated the therapeutic effect of Hcy-ON through histological and immunohistochemical studies (Figure 7f). Hematoxylin and eosin (H&E) staining results revealed that the treatment effect of Hcy-ON, mediated by NIR laser, and the generation of ROS, caused extensive damage to tumor tissue, resulting in significant abnormalities in tumor cells. The light-treated group exhibited numerous vacuoles and obvious nuclear condensation, while the other two groups showed dense tumor cell populations. The terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay demonstrated abundant red fluorescence exclusively in the light-treated group, confirming severe apoptosis of tumor tissue cells following the combined treatment of Hcy-ON with PDT/PTT. Ki67 staining analyses of the excised tumor tissues revealed that only the “light +  Hcy-ON” treated group exhibited a low labeling rate. This finding indicates that the PDT/PTT synergistic treatment with Hcy-ON has a favorable prognosis and a reduced likelihood of tumor recurrence. Moreover, all histological analyses showed no pathological changes or morphological abnormalities in other organs and tissues ( Supporting Information Figure S17), reaffirming the excellent biocompatibility of Hcy-ON. In conclusion, these results clearly demonstrate that Hcy-ON is a versatile dye with promising applications for dual-modal PAI/PTI-guided PDT/PTT synergistic cancer treatment.

Conclusions

Through the optimization of an electron-withdrawing engineering strategy, this study introduced three new multifunctional dyes featuring AIE characteristics, namely Hcy-OO, Hcy-ON, and Hcy-NN. We accurately regulated molecular properties at the level of molecular conformation and donor–acceptor electronic structure by varying the numbers of cyano groups. Following comprehensive molecular theory calculations and a series of in vitro and in vivo studies, we found that Hcy-ON exhibited the best performances among the dyes used in this study. Notably, owing to its distinctive D-π-A architecture and cyano group modulation, Hcy-ON demonstrated NIR excitation at 760 nm and a high ROS yield (${}^1\text{O}{}_2$, •OH), showcasing excellent type I & II PDT capability. The presence of multiple freely rotating single bonds and electron-withdrawing groups, endows Hcy-ON with excellent PTT capability as well as dual-modal imaging capabilities in PAI and PTI. Compared with other traditional theranostics dyes, Hcy-ON exhibited an extended excitation wavelength and superior PAI performance, effectively addressing the problem of insufficient imaging penetration depth. Additionally, Hcy-ON demonstrated the capacity for PAI/PTI and combined PTT/PDT, enabling integrated diagnosis and treatment guided by dual-modal imaging. While the investigation of Hcy-ON is presently confined to tumor models in cellular and murine systems, the synergistic effects of dual-modal imaging and combined therapies hold promise for revolutionizing single-modality imaging or treatment approaches in clinical settings.

Supporting Information

Supporting Information is available and includes additional details on general methods, the experimental section, supporting figures ( Figures S1–S17).

Conflict of Interest

The authors declare that they have no competing financial interests.

Funding Information

This work was supported by the National Key Research and Development Program of China (grant no. 2023YFB3810300), the National Natural Science Foundation of China (grant nos. 22078050, 22278061, 22378050, 22378051, and 22090011), the Science and Technology Plan Project of Liaoning Province (grant no. 2023JH2/101700296), the Fundamental Research Funds for the Central Universities (grant nos. DUT24ZD117 and DUT24LAB105), and the Open Research Fund from the State Key Laboratory of Chemo/Biosensing and Chemometrics of Hunan University (grant no. 20240604). J.Y. thanks to the Nano & Material Technology Development Program through the National Research Foundation of Korea (NRF) funded by Ministry of Science and ICT (grant no. RS-2024-00407093) and the NRF for the grant funded by the Korean government (MSIT) (grant no. 2022R1A2C3005420).