Xueqi
Tian
a,
Shengke
Li
b,
Krishnasamy
Velmurugan
a,
Zhihang
Bai
a,
Qian
Liu
a,
Kaiya
Wang
a,
Minzan
Zuo
*a and
Xiao-Yu
Hu
*a
aCollege of Materials Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, 211106, China. E-mail: minzanzuo@nuaa.edu.cn; huxy@nuaa.edu.cn
bState Key Laboratory of Quality Research in Chinese Medicine, Institute of Chinese Medical Sciences, University of Macau, Macau, 999078, China
First published on 24th March 2023
Photosensitizers (PSs), featuring switchable properties to achieve tunable reactive oxygen species (ROS) generation, are highly desirable in photodynamic treatment and antibacterial applications. A supramolecular strategy combined with aggregation-induced emission (AIE) characteristics for synergistic design of an effective PS has emerged as a promising approach to reduce its side effects. Herein, we have constructed a ternary supramolecular PS based on a two-step Förster resonance energy transfer (FRET) process using an AIE-active pillar[5]arene host, a spiropyran derivative (SP-G) guest, and Nile blue (NiB) dye in aqueous media. The reversible isomerization process of SP-G endows this PS with the “on–off” switchability. Under UV light irradiation, the guest SP-G (non-emissive closed-form) can be converted to merocyanine MC-G (emissive open-form). And the in situ formed MC-G can function as the first energy acceptor and an efficient PS for ROS generation. Meanwhile, the introduction of NiB promotes a synergistic ROS-generation activity, leading to a two-step dynamic FRET process in response to UV and visible light irradiation. Moreover, this photoswitchable system with excellent controllable fluorescence performance and ROS-generation ability can be applied to inactivation of cancer cells and bacteria. The present work on the basis of a supramolecular strategy provides a proof-of-concept and new insight for the fabrication of smart PS materials.
In general, traditional PSs always suffer from aggregation-caused fluorescence quenching in biological environments, and are not suitable for precise PDT treatment. Benefiting from the discovery of aggregation-induced emission (AIE), AIE-based materials have non-emissive or weak fluorescence in solution but exhibit strong emission in the aggregated state due to restricted intramolecular rotations and vibrations, resulting in strong luminous properties and high photostability. Thus, PS-related nanomaterials with attractive AIE properties could serve as an ideal alternative for conventional PSs.17–21 Additionally, Förster resonance energy transfer (FRET) is perceived as an efficient strategy for PSs to achieve various wave-bands through switchable energy transfer effects.22–25 Therefore, the development of efficient supramolecular PSs that combine excellent AIE properties with photoswitchability based on host–guest interactions and the FRET process is highly appealing.
Inspired by our previous work, the water-soluble meso-tetraphenylethene (TPE)-functionalized water-soluble pillar[5]arene (m-TPEWP5) can act as an AIE host molecule to fabricate smart biomedical materials.26 The structure of m-TPEWP5 possesses two key factors: (1) the TPE motif embedded in the pillararene backbone endows it with excellent AIE properties and (2) the presence of a pillar[5]arene cavity can encapsulate various guest molecules via non-covalent interactions, thus inhibiting their aggregation and improving the water-solubility of guest molecules. By virtue of these desirable features, m-TPEWP5 was selected as an ideal candidate for fabricating an AIE-based PS. It is well known that spiropyran derivatives have been widely used as efficient photoswitches owing to their excellent photochromic abilities, i.e. the non-emissive closed form of spiropyran (SP) and the emissive open form of merocyanine (MC) can be transformed reversibly by UV-Vis light irradiation.27–29 Notably, upon UV light irradiation, SP converts to MC, accompanied by the appearance of a new absorption band at 520 nm, which perfectly overlaps with the emission band of TPE.30–32 Accordingly, in this case, a possible reversible “on–off” FRET process will be activated or inactivated between the TPE-based energy donor and these two isomers of the spiropyran-based acceptor. Therefore, we presumed that the host molecule m-TPEWP5 could be used as a FRET donor and the open form of guest MC-G could be used as an acceptor to fabricate an effective FRET system. Moreover, Nile blue (NiB) was introduced as a second acceptor, thus realizing a two-step energy transfer process with switchability.33 In addition to the properties mentioned above, another attractive feature of spiropyran derivatives as well as NiB is that they can both function as photosensitizers to generate ROS during the photoswitching process.34,35 Thus, the photoswitchable two-step FRET process makes the system a suitable candidate for controllable PDT treatment in response to changes in light wavelength.
Herein, we reported the successful fabrication of a photoswitchable AIE-active supramolecular photosensitizer with red emission using a two-step hierarchical assembly process (Scheme 1). Firstly, incorporating the guest MC-G into the m-TPEWP5 host led to the formation of a host–guest self-assembled system (m-TPEWP5⊃MC-G). And then, NiB was loaded into the above assembly to construct a ternary m-TPEWP5⊃MC-G-NiB nanoparticle system (NP2). In this system, the energy could transfer from the m-TPEWP5 donor (λem = 465 nm) to the MC-G 1st acceptor (λem = 625 nm) and then to the NiB 2nd acceptor (λem = 675 nm) with high energy transfer efficiency through a two-step sequential FRET process. Importantly, high ROS generation efficiency could be achieved due to the synergistic enhancement of ROS-generation ability of MC-G and NiB, which provided a critical guarantee for the NP2 system as an ideal PS. To this end, UV-induced optical conversion from SP-G (non-emissive closed-form) to MC-G (emissive open-form) activated the FRET and ROS generation processes in the NP2 system to achieve the photodynamic inactivation of cancer cells and bacteria. In the present work, we proposed an efficient supramolecular strategy to develop a photoswitchable PS, which could synergistically incorporate the dynamic FRET process and controllable ROS-generation ability for potential applications in anticancer and antibacterial treatments.
![]() | ||
Scheme 1 A schematic illustration of the construction of an AIE-based ROS-generation system in aqueous solution. |
By introducing a water-soluble sulfonate terminal into SP-G, the water solubility of SP-G is improved and a binding site is also afforded for the efficient binding with m-TPEWP5. Based on our previous work, we evidenced that the alkyl chain with the sulfonate end of the guest could thread into the cavity of m-TPEWP5, which confirmed the formation of a host–guest complex between m-TPEWP5 and SP-G.26 The binding stoichiometry ratio between m-TPEWP5 and SP-G was determined to be 1:
1 based on the Job's plot analysis (Fig. S7, ESI†). And the binding constant (Ka) was calculated to be (2.30 ± 0.34) × 104 M−1 by using a non-linear curve-fitting method (Fig. S8, ESI†).36 The driving forces for the formation of a m-TPEWP5⊃SP-G supramolecular complex were attributed to the synergistic hydrophobic and electrostatic interactions. Therefore, we deduced that m-TPEWP5 and MC-G could also form a 1
:
1 inclusion complex with the sulfonate terminal as the binding site. Since the formed m-TPEWP5⊃MC-G complex has an amphiphilic structure, the best molar ratio between m-TPEWP5 and MC-G for constructing supramolecular nanoaggregates was further determined to be 1
:
4 based on fluorescent titration experiments (Fig. S9, ESI†). Subsequently, the morphology and size of the nanoaggregates formed by m-TPEWP5⊃MC-G (1
:
4) in aqueous solution were investigated by transmission electron microscopy (TEM) and dynamic light scattering (DLS) studies (Fig. 1a). From the TEM image, m-TPEWP5⊃MC-G (with a 1
:
4 ratio) showed nano-aggregated spherical particles with an average diameter (d) of around 250 nm, which was in good agreement with the result obtained from the DLS measurements (d = 276 nm).
Given that SP-G can undergo reversible isomerization from non-emissive SP-G to emissive MC-G, the photophysical properties of m-TPEWP5 and MC-G were characterized by UV-Vis and fluorescence spectroscopies. To confirm the first energy transfer process from the m-TPEWP5 donor to the MC-G acceptor, the normalized absorption and emission spectra of MC-G and m-TPEWP5 were plotted. As shown in Fig. 1b, the normalized spectra showed a good overlapping between MC-G and m-TPEWP5, which is an essential prerequisite for the FRET process. Next, we investigated the 1st energy transfer process at the best molar ratio of the m-TPEWP5 donor (1 equiv.) and MC-G acceptor (4 equiv.) under UV light irradiation. Upon extending the UV irradiation time (0 to 5 min), the fluorescence intensity of the m-TPEWP5 donor at 465 nm decreased gradually, while the fluorescence emission of the MC-G acceptor at 625 nm increased accordingly, together with the observation of an isosbestic point at 610 nm, indicating the first-step FRET process took place from the donor to the acceptor (Fig. 1c). Furthermore, the fluorescence color changed from blue-green of m-TPEWP5 to pink-red of the m-TPEWP5⊃MC-G assembly (Fig. 1c inset). Considering that the FRET efficiency is relevant to the ratios of the donor and acceptor, we further investigated the energy transfer efficiencies (ϕET) with different m-TPEWP5/MC-G ratios, and the maximum energy transfer efficiency was calculated to be 70.2% at a molar ratio of donor/acceptor = 1:
4, which was consistent with the best molar ratio (Fig. S10, ESI†).
In continuation, we further examined the 2nd energy transfer process. After careful evaluation, NiB was chosen as a decent second acceptor for the m-TPEWP5⊃MC-G (1:
4) system, since the absorption peak of NiB overlaps very well with the emission peak of m-TPEWP5⊃MC-G assembly (Fig. 1e). Upon gradual addition of NiB (0 to 0.1 equiv.) into the m-TPEWP5⊃MC-G system, the emission band at 675 nm ascribed to NiB increased gradually, while the emission peak of m-TPEWP5⊃MC-G assembly at 465 nm and 625 nm decreased under UV light irradiation, accompanied by the visual fluorescence color changes from pink-red to deep red, representing the characteristic FRET process (Fig. 1f inset). And the maximum energy transfer efficiency reached 62.5% at a m-TPEWP5
:
MC-G
:
NiB ratio of 1
:
4
:
0.1 in aqueous solution. Thus, m-TPEWP5⊃MC-G (1
:
4) with a NiB loading ratio of 0.1 equiv. was used for the following experiments. Meanwhile, the TEM data of m-TPEWP5⊃MC-G-NiB (1
:
4
:
0.1) also showed spherical structures with a similar diameter to the DLS results (Fig. 1d). In comparison, the m-TPEWP5⊃MC-G-NiB nanoparticles, in which NiB could be encapsulated in the hydrophobic region showed a slightly larger average size (d = 286 nm) than the m-TPEWP5⊃MC-G nanoparticles. Additionally, the stability of the m-TPEWP5⊃SP-G and m-TPEWP5⊃SP-G-NiB nanoparticles was also evidenced by TEM analysis, which showed that the size and shape of the assemblies were not significantly affected by three cycles of irradiation (Fig. S11, ESI†). In addition, the zeta potential of the m-TPEWP5⊃MC-G-NiB assembly was examined to be 49.10 mV (Fig. S12, ESI†), indicating that the surface of the nanoparticles was positively charged and such an electrostatic repulsive force can prevent their agglomeration and further improve the stability of the nanoparticles in aqueous solution. And the obtained m-TPEWP5⊃MC-G-NiB nanoparticles demonstrated good stability that they had no significant change in the particle diameter for 4 weeks (Fig. S13 and S14, ESI†).
Subsequently, control experiments were conducted to check that the “off” state corresponds to visible light irradiation. For the m-TPEWP5⊃SP-G (1:
4) system, no emission peak could be observed at 625 nm (MC-G, 1st acceptor) under continuous visible light irradiation (>500 nm, 20 mW cm−2, Fig. S15a, ESI†). Similarly, the fluorescence spectra of the m-TPEWP5⊃SP-G-NiB (1
:
4
:
0.1) system showed only slight changes at 465 nm and 625 nm (m-TPEWP5⊃SP-G donor), but no isosbestic changes or peak enhancement was observed at 675 nm (NiB, 2nd acceptor, Fig. S15b, ESI†). Overall, the above results demonstrated that the FRET process can be turned off by visible light irradiation.
To further prove the FRET process, fluorescence lifetime decay experiments were carried out. First, the average fluorescence lifetime of m-TPEWP5 was determined to be τ = 13.10 ns by fitting the decay curve as a double exponential decay curve. While assembling with MC-G, the lifetimes obviously decreased to τ = 11.37 ns due to the energy transfer from m-TPEWP5 to MC-G. After loading NiB into the m-TPEWP5⊃MC-G assembly, the lifetime of the m-TPEWP5⊃MC-G-NiB (1:
4
:
0.1) ternary system reduced to τ = 7.97 ns, thus further confirming the two-step sequential energy transfer (Fig. 2a). Furthermore, it was worth noting that the fluorescence quantum yield of m-TPEWP5 was only 3.24%, which significantly increased to 14.43% and 22.06% after co-assembling with MC-G and then loading with NiB, respectively (Table S1, ESI†). The above results corroborated that the m-TPEWP5⊃MC-G and m-TPEWP5⊃MC-G-NiB systems feature typical FRET processes with moderate quantum yields.
ROS production ability is a critical evaluation factor that needs to be considered in the PDT process. Thus, the ROS-generating capabilities of m-TPEWP5⊃MC-G (NP1) and m-TPEWP5⊃MC-G-NiB (NP2) were evaluated by using 9,10-anthracenediyl-bis(methylene)dimalonic acid (ABDA) as a ROS indicator.37,38 Commercially available Rose Bengal (RB) was used as a standard photosensitizer. With the generation of singlet oxygen, the absorbance of ABDA at 400 nm significantly decreased in the presence of MC-active NP1, MC-active NP2, and RB under visible light irradiation (Fig. S16, ESI†). As shown in Fig. 2b, the effective generation of ROS from MC-active NP1 and NP2 could be clearly observed, and NP1 exhibited nearly equal ROS productivity to RB. Impressively, MC-active NP2 exhibited excellent ROS generation ability, and the consumption of ABDA was analyzed to be 0.17 mM in the presence of MC-active NP2 after irradiating for 180 s, but it was only 0.11 mM in the case of RB, which was 1.54-fold higher than that of RB. This may be attributed to the loading of NiB into NP2, which can be evidenced from two factors: (1) NiB, one of the widely studied fluorescent dyes, has some weak phototoxicity and can be used as a cationic PS;39–43 (2) NiB is protected by the formation of nanoassemblies, which can stabilize its triplet states and improve the ROS-generation capacity during irradiation. Furthermore, in order to test the synergistic ROS generation efficiency of the NP2 system, free MC-G and NiB were also estimated in aqueous solution. As shown in Fig. S16 (ESI†), both of them possessed ROS-generation ability but with low efficiency. Singlet oxygen quantum yield (ϕΔ) is an important indicator to evaluate the singlet-oxygen generating capability. The ϕΔ values of NP2, NP1, MC-G, and NiB were calculated by using 1,3-diphenylisobenzofuran (DPBF) as a probe with RB (ϕΔ = 75.0%) as the reference.34 And the singlet oxygen quantum yields for NP2, NP1, MC-G, and NiB were calculated to be 85.2%, 71.2%, 37.5%, and 18.9%, respectively (Fig. S17, ESI†). The high ϕΔ value of NP2 may be caused by the appropriate encapsulation of MC-G and NiB, which can inhibit their aggregate-induced quenching effect thus to synergistically enhance the singlet oxygen quantum yields. The above results demonstrated the excellent ROS generation performance of NP2, indicating that it is a potential contender for PDT applications.
Considering the UV light activated ROS generation and excellent tunable red emission (up to 700 nm) of this system, we further investigate the in vitro fluorescence imaging of NP1 and NP2 at the cellular level after different UV-Vis light alternative irradiation cycles (each cycle contains UV (365 nm, 32 mW cm−2) and visible light (>500 nm, 20 mW cm−2) alternative irradiation). In detail, UV light irradiation for a short period of time is used to activate the photosensitizer process, while the ROS generation process is actually regulated by visible light. After incubating NP1 and NP2 with HeLa cells for 4 h, obvious red emissions could be observed with UV-visible irradiation for different cycles, which indicated that the nanoparticles could be taken up very well by cancer cells. Moreover, the red fluorescence was completely quenched with visible light irradiation, indicating the efficient photoswitching of spiropyran from the MC to SP form in the NP1 and NP2 systems (Fig. 3 and Fig. S18, ESI†). The red fluorescence could be turned on and off again in the following 2nd and 3rd cycles of alternating irradiation. The intense red fluorescence confirmed that MC-active NPs could exist in cells during photoconversion and participate in the cellular internalization process. As a comparison, free MC-G exhibited very weak emission under UV light irradiation (Fig. S19, ESI†), due to its low fluorescence quantum yield and low cell internalization efficiency compared to NP1 and NP2. Thus, a reversible two-step FRET process enables real-time monitoring of the reversible ROS “on–off” process.
In order to further confirm the photoswitchable intracellular ROS generation capability, MC-active NP1 and NP2 treated HeLa cells were incubated with a readily available ROS probe, singlet oxygen sensor green (SOSG). Interestingly, strong green fluorescence (MC-state) could be observed in the cells after visible light irradiation. It is worth mentioning that there was no noticeable fluorescence (SP-state) in the cells after irradiation with visible light, resulting in the “off” state of ROS switch (Fig. 4 and Fig. S20, ESI†). To further confirm this observation, another ROS probe 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) was also employed to evaluate the controllable production of ROS (Fig. S21 and S22, ESI†).44 The above results evidenced that high ROS generation in living cells could be well controlled by desired light irradiation, and the resulting smart nanoparticles could be potentially used for PDT applications.
The fabricated photoswitchable NP1 and NP2 with continuous ROS generation ability may have great potential to cause cellular apoptosis. For the NP1 and NP2 systems, their SP-states exhibit good stability to natural light and are not affected by visible light irradiation, thus they can avoid both photocytotoxicity and dark cytotoxicity. The cytotoxicity of NP1 and NP2 against the HeLa cell line was tested using the Cell Count Kit-8 (CCK-8) assay.45 As shown in Fig. 5a and b, the cell viability exceeds 90% without irradiation, suggesting good cytocompatibility for NP1 and NP2. In the control group, only the MC-G guest was treated with HeLa cells and the cell viability was about 90% (Fig. S23, ESI†), indicating that the MC-G guest exhibited very weak ROS generation ability. However, once exposed to alternative irradiation cycles (each cycle contains UV (365 nm, 32 mW cm−2) and visible light (>500 nm, 20 mW cm−2) alternative irradiation), ROS would be accumulated efficiently in NP1 and NP2, leading to a high cytotoxicity against HeLa cells. With further increase in the concentration and irradiation cycles, the cell viability decreased to 60% and 45% for NP1 and NP2, respectively. The results demonstrated that NP2 exhibited higher photocytotoxicity than NP1. As shown in Fig. 5c, d and Fig. S24 (ESI†), the flow cytometric analysis also confirmed an increase in the number of apoptotic cells after different irradiation cycles.
Considering the high ROS production process of NP2 that can be activated by UV light in a short period of time, while this process is actually regulated by visible light, NP2 may serve as a more suitable PS for photodynamic antibacterial therapy (PDAT). Encouraged by the above ROS generation results and considering the positively charged property of NP2, we put forward to examine the potential performance of NP2 in bacterial inactivation. Due to the presence of quaternary ammonium groups in m-TPEWP5, NP2 should exhibit broad-spectrum antibacterial activity against Gram-positive and Gram-negative bacteria with negatively charged membranes.46–49 Bacteriostatic effects of NP2 against both Gram-negative Escherichia coli (E. coli) and Gram-positive Staphylococcus aureus (S. aureus) were further investigated. As shown in Fig. 6a, NP2 exhibited moderate activity against bacteria without irradiation, probably due to the intrinsic antibacterial activity of the trimethylamine moiety in m-TPEWP5. However, NP2 achieved greatly enhanced bacterial lethality in E. coli and S. aureus after UV-visible light irradiation. The bacterial inhibition ratio was gradually increased to 59.8% (E. coli) and 55.4% (S. aureus) after the 1st irradiation cycle, while the bacteria-killing efficiency was remarkably increased up to 90.9% (E. coli) and 93.1% (S. aureus) after the 3rd irradiation cycle, respectively, suggesting the irradiation cycle-dependent ROS production capability of NP2 (Fig. 6b). More importantly, these antibacterial studies demonstrated that NP2 exhibited excellent antibacterial activity against both Gram-positive and Gram-negative bacteria and that the antibacterial capacity is regulated by the simple UV-visible light irradiation. In order to afford more insights into the mechanism of the bacteriostatic process, zeta potential measurements were performed to investigate the interactions between NP2 and E. coli. With the addition of NP2 into E. coli, the zeta potentials gradually changed from −19.1 mV to −5.1 mV, which confirmed the successful binding of NP2 to the E. coli surface (Fig. S25, ESI†). Moreover, a scanning electron microscope (SEM) was used to further visualize the morphological changes of NP2-treated bacteria, and it was observed that partial membranes of the E. coli became rough and irregular (Fig. S26, ESI†). Therefore, based on the above observations, a possible antibacterial mechanism was proposed as shown in Fig. 6c, and the integrity of the bacterial membrane was destroyed through the electrostatic interactions between the positively charged m-TPEWP5 and the negatively charged bacterial membrane. Meanwhile, the ternary NP2 system could generate ROS under irradiation to inhibit bacterial proliferation, thus achieving a synergistic effect.
Footnote |
† Electronic supplementary information (ESI) available: Experimental data, NMR spectra etc. See DOI: https://doi.org/10.1039/d3qm00153a |
This journal is © the Partner Organisations 2023 |