Pillararene-based rare-earth luminescent probe for two-channel simultaneous detection of ROS and ATP in living cells

Abstract

Exploring the fluctuation of reactive oxygen species (ROS) and adenosine triphosphate (ATP) in living systems is significant for the early diagnosis of many diseases. However, the simultaneous detection of ROS and ATP is still challenging. Herein, a supramolecular fluorescent probe (G₃⊂CP5) was designed for two-channel simultaneous detection of H₂O₂ and ATP. G₃⊂CP5 consists of an Eu(III)-complex and water-soluble pillar[5]arene (CP5), achieving supramolecular assembly via synergistic coordinated, host–guest and amphiphilic interactions. Phenylboric acid groups in the ligand can interact with H₂O₂, and CP5 exhibits host–guest recognition toward ATP. As a consequence, the phenylboric acid groups and CP5 are the responsive moieties for H₂O₂ and ATP, respectively. Optical experiments demonstrated that G₃⊂CP5 exhibited low limits of detection and could respond to H₂O₂ and ATP in two channels to reduce mutual interference. Significantly, the probe shows good mitochondrial targeting. The fluorescent sensing of H₂O₂ and ATP in both HeLa cells and RAW264.7 macrophages was also confirmed. As a consequence, taking advantage of the superior luminescence of rare-earth complexes and the precise recognition of pillar[n]arenes, this work prepared a fluorescent probe for simultaneous detection of H₂O₂ and ATP, providing opportunities for further accurate biosensing.

---

Introduction

Reactive oxygen species (ROS) and adenosine triphosphate (ATP) are important markers of cellular oxidative stress and energy metabolism.^1–3 ROS can trigger abnormal hepatocyte death, abnormal oxidative stress and localized inflammation,^4,5 while ATP is the molecular currency of energy transfer that sustains many intracellular physiological events.⁶ The development of technologies for non-invasive, reliable, and real-time monitoring of ROS and ATP in biosystems is particularly significant for the early diagnosis of liver-related diseases, such as drug-induced liver injury, hepatic ischemia-reperfusion injury, liver fibrosis and liver cancer.^7–9 It is also important to develop convenient, accurate and reliable methods to monitor physiological and pathological processes under different conditions. In recent years, diverse fluorescent probes aiming at different factors have been reported as the most effective strategies for the early assessment of liver-related disease owing to their high sensitivity, great selectivity and rapid real-time response.^10–13 However, the precise and non-interferential detection of ATP and ROS is still a challenge.^14–16

Among the diverse fluorescent probes, probes based on rare-earth complexes have received significant attention because of their long fluorescence lifetime, large Stokes shift, narrow emission range and good resistance to photobleaching.^17–19 The long lifespan of rare-earth luminescent materials can effectively avoid the interference of intracellular spontaneous fluorescence, and it is expected to achieve accurate imaging of specific components in cells.^20–22 Thus, rare-earth probes can be promising tools for precise sensing of ATP and ROS in living systems. However, on the one hand, rare-earth complexes suffer from several problems, including the ease of bursting in water, the lack of biocompatibility and poor stability, which limit their practical applications. On the other hand, traditional rare-earth materials face problems such as cumbersome design and syntheses, poor responsive ability and the difficult regulation of functions.²³

Fortunately, supramolecular assemblies based on non-covalent interactions have dynamic controllability and intelligent responsibility, which provide the opportunity to design and synthesize complexes with dynamic optical regulation.²⁴

Based on the advantages of supramolecular assemblies, the development of rare-earth intelligent luminescent probes has also made great progress, expanding their applications to the fields of bioimaging,^25–28 detection of hypochlorous acid^29,30 and measurement of temperature.^31–33 Supramolecular assemblies can simplify the complicated design of ligands and realize the controllable syntheses of rare-earth complexes through non-covalent interactions for dynamic and intelligent characteristics.

In addition, the combination of rare-earth complexes with supramolecular macrocycles would also endow them with more advantages such as good responsiveness and precise regulation, further avoiding the water quenching, improving the luminescent performance and recognizing specific small molecules due to the macrocycle cavities.^34–36 Therefore, the collaboration of host–guest and coordinated interactions would be promising to construct fluorescent probes with accurate and adjustable biosensing, which could monitor the dynamic information including the intracellular microenvironment and substrate migration. They also have great potential for the simultaneous detection of H₂O₂ and ATP.

Due to the above inspiration, a supramolecular fluorescence probe consisting of rare-earth complex and water-soluble pillar[5]arene (CP5), i.e., Eu(TPBA)₂THB⊂CP5 (G₃⊂CP5), has been designed and prepared (Scheme 1). There are double responsive groups for the detection of H₂O₂ and ATP, respectively. Firstly, the luminescent rare-earth complexes were constructed with the tripyridine ligands, which include the phenylboric acid groups that respond to H₂O₂ and quaternary ammonium salt moieties that interact with CP5 through host–guest interactions. The supramolecular assemblies of G₃⊂CP5 could be constructed via the synergistic coordinated, amphiphilic and host–guest interactions. The phenylboric acid groups could respond to H₂O₂ and CP5 and recognize ATP through competitive binding. Optical experiments have demonstrated that G₃⊂CP5 can achieve the two-channel (blue and red emission for ATP channel and H₂O₂ channel, respectively) and simultaneous fluorescence sensing of H₂O₂ and ATP. Cellular imaging has also verified the good mitochondrial targeting and fluorescence response to H₂O₂ and ATP in both HeLa cells and RAW264.7 macrophages since they have higher levels of ROS and ATP production compared with other types of cells. This work provides a strategy to develop novel fluorescent probes with interference-free detection of different substrates in biosystems, which would further provide the opportunity to study the relationship between energy metabolism and redox homeostasis during apoptosis.

Scheme 1 The schematic of the preparation of G₃⊂CP5 and the two-channel detection of H₂O₂ and ATP in living cells.

Results and discussion

Syntheses and characterization

Design and syntheses of rare-earth complexes. Firstly, two ligands, i.e., 4-([2,2′: 6′,2″-bipyridin]-4′-yl)phenylboronic acid (TPBA, compound 1) and hydrophobic N,N,N-trimethyl-2-(3-(4,4,4-trifluoro-3-oxobutanoyl)-9H-carbazol-9-yl)ethan-1-amine (THB, compound 5), were successfully synthesized and fully characterized by ¹H NMR, ¹³C NMR and mass spectroscopy (Schemes S1, S2, and Fig. S1–S12, ESI†). Then, TPBA, THB and the well-known ligand for rare-earth complexes, i.e., 2-thiophenyl trifluoroacetone (TTA), were used to successfully prepare three complexes, including G₁, G₂ and G₃, by coordinating with Eu(III). They were determined by high-resolution mass spectrometry (HRMS), elemental analysis (Table S1, ESI†), ultraviolet-visible absorption (UV-Vis) spectroscopy and Fourier-transform infrared spectroscopy (FT-IR). The peak at 497.102 could be observed in the HRMS spectrum of G₁ (Fig. S13, ESI†), which was consistent with a theoretical mass-to-charge ratio of 496.7 (theoretical total molecular weight is 1490.1), indicating that G₁ was a nine-coordination composed of a europium ion, a TPBA ligand, two TTA ligands and a THB ligand. Similarly, the complex G₂ consisted of one europium ion, one TPBA ligand, two o-phenanthroline ligands and one THB ligand losing an iodide ion, while G₃ consisted of a europium ion, two TPBA ligands and one THB ligand losing an iodide ion, according to their corresponding HRMS spectra (Fig. S14 and S15, ESI†). Besides, the experimental results in Table S1 (ESI†) prove that G₁–G₄ all have good purity.

As shown in Fig. S17(A), (B) and (C) (ESI†), the assembled complexes G₁, G₂ and G₃ exhibited different morphologies such as irregular, rod-like and spherical shapes, respectively. Furthermore, the successful synthesis of the complexes was also verified by the UV-Vis spectra and FT-IR spectra (Fig. S17, ESI†). The complex G₁ had a maximum absorbance at 278 nm and a broader peak at 342 nm, which was slightly blue shifted compared with the absorption of the ligand at 344 nm (n → π* transition) and 283 nm (π → π* transition). This shift could be ascribed to the formation of coordinated bonding and energy transfer between the ligand and Eu(III). The complex G₂ possessed a sharp absorption at 278 nm and a broad peak from 327 to 363 nm, containing the absorption of TPBA and THB ligands at 327 nm and 342 nm, respectively. The broader and red-shifted peak at 363 nm compared with individual THB could be ascribed to the coordination and energy transfer. Besides, the complex G₃ showed a very similar absorption spectrum compared with G₂ because of their similar ligands and weak UV absorption of o-phenanthroline ligands. Moreover, as shown in the FT-IR spectra (Fig. S17(G)–(I), ESI†), the peak at 1585 cm⁻¹ for all the three complexes belonged to the enol form, which shifted to the higher wavenumber after coordinating with the central Eu³⁺ ion. All the above results demonstrated that the three Eu(III)-complexes (G₁, G₂ and G₃) were successfully synthesized.

Design and syntheses of host–guest complexes. In order to develop an efficient probe for the simultaneous detection of H₂O₂ and ATP, a water-soluble macrocycle, i.e., CP5 with good recognition toward ATP, was introduced to co-assemble with the Eu(III)-complexes. CP5 has strong binding affinity toward the quaternary ammonium salt moiety of THB. As shown in Fig. 1(A), after the addition of CP5, the complex G₁ co-assembled with CP5 and turned into uniform and hemispherical shapes via the synergetic host–guest and hydrophilic interactions. The complex G₂ developed a uniform and spherical morphology with a rough surface and a size of about 500 nm after interacting with CP5 instead of its inherently elongated rod-like structures. Then, spherical morphologies with size of 200–300 nm for the complex G₃ after forming host–guest assemblies with CP5 were also constructed. Dynamic light scattering (DLS) data showed that the average sizes of G₁⊂CP5, G₂⊂CP5 and G₃⊂CP5 were 325.72 nm, 425.01 nm and 287.04 nm, respectively. Typically, the size of G₃⊂CP5 showed a significant reduction compared with the individual assembly of G₃ of about 400–500 nm. In addition, all the three rare-earth complexes (G₁, G₂ and G₃) suffered the problems of aggregation and sedimentation in aqueous solution. After co-assembling with CP5 via host–guest interactions, G₁⊂CP5, G₂⊂CP5 and G₃⊂CP5 exhibited obviously better dispersion, indicating that the introduction of CP5 can also improve their dispersibility. Moreover, according to the analyses of the zeta potentials in Fig. 1(B), the three self-assembled complexes had positive charges of about 44.19 mV, 5.83 mV, and 55.94 mV, while G₁⊂CP5, G₂⊂CP5 and G₃⊂CP5 exhibited a negatively charged surface with the values of −28.73 mV, −30.35 mV, and −34.32 mV, respectively. It suggested that CP5 could effectively diminish the surface potential due to its inherent electron-rich cavities and rims, which would benefit their biological circulation. It was proved that G₃⊂CP5 possessed the most regular morphology, the smallest size with good dispersibility and a negatively charged surface, which could further satisfy the requirement of biological applications. Thus, G₃⊂CP5 was the most promising bio-probe to be utilized in biosystems and was chosen for subsequent experiments.

Fig. 1 (A) SEM images of G₁⊂CP5, G₂⊂CP5 and G₃⊂CP5. The top right inset shows the dynamic light scattering (DLS) images of the three host–guest complexes. (B) Zeta potentials of the three complexes (G₁, G₂, G₃) and the supramolecular assemblies with CP5 (G₁⊂CP5, G₂⊂CP5 and G₃⊂CP5) in water. (C) TEM images and elemental mapping of G₃⊂CP5. (D) Fluorescence lifetime spectra of G₃ and G₃⊂CP5 (λ_ex = 331 nm).

The transmission electron microscopy (TEM) images and elemental mapping further confirmed the uniform distribution of G₃⊂CP5 with the Eu(III) strategically positioned at the core (Fig. 1(C)). Finally, the fluorescence lifetimes of the Eu³⁺ excited state of G₃ and G₃⊂CP5 were also measured to be 0.52 ms and 0.48 ms at an excitation wavelength of 331 nm, respectively (Fig. 1(D)). The results showed that G₃⊂CP5 exhibited a very slight decrease in lifetime compared with G₃, indicating that CP5 had a negligible effect on the fluorescence lifetime of the supramolecular assemblies.

Exploration of the detection mechanism

Study of the host–guest interaction between CP5 and THB. In order to understand the relationship of each component and explore the mechanism of detection, the host–guest interactions between CP5 and THB were investigated by ¹H NMR titration in D₂O. As shown in Fig. 2 and Fig. S18(A), ESI,† the signals of phenyl protons (H₁) in CP5 shifted downfield slightly upon the gradual addition of CP5, resulting from the interaction between CP5 and THB. In contrast, the signals derived from the methyl protons (H_a) of THB shifted upfield due to the shielding effect of the electron-rich cavities in CP5. This clearly demonstrated the inclusion of THB into the cavity of CP5. The point of intersection in the Job plot was determined to be 0.77 (Fig. S18(C), ESI†), which suggested that the stoichiometric ratio of THB⊂CP5 was 1 : 1. The interaction of CP5-THB was also examined by UV-Vis titration, which also indicated the 1 : 1 binding stoichiometry (Fig. S18D and E, ESI†). ITC data showed that the binding constant of THB⊂CP5 was 5.92 × 10⁴ M⁻¹ (Fig. S19A, ESI†). All the above results indicated that there was a good host–guest interaction between CP5 and THB with a binding ratio of 1 : 1. Benefiting from the efficient binding affinity of CP5 and THB, the supramolecular assembly of G₃⊂CP5 could be successfully constructed via the synergistic host–guest interaction and amphiphilic effect.

Fig. 2 (A) Partial ¹H NMR spectra for the titration of THB⊂CP5 (400 MHz, D₂O, THB (G, 1 mM) and CP5 (H, 0–2 mM)). (B) Partial ¹H NMR spectra for the titration of ATP⊂CP5 (400 MHz, D₂O, ATP (G, 1 mM) and CP5 (H, 0–2 mM)).

Study of the host–guest interaction between CP5 and ATP. The possibility of ATP detection was largely dependent on the efficient binding affinity between pillar[n]arenes and ATP, which has been reported in the literatures.^37,38 Accordingly, the ATP recognition by CP5 was also studied here by ¹H NMR titration in D₂O. On increasing the concentration of CP5, the signals of phenyl protons (H₁) in CP5 shifted slightly downfield. Meanwhile, the signals derived from the pyridine protons (H_a) of ATP shifted downfield. The stoichiometric ratio of ATP⊂CP5 was determined to be 1 : 1 (Fig. S18(F), ESI†). This demonstrated that ATP was included in the cavity of CP5. ITC data showed that the binding constant of ATP⊂CP5 was 2.33 × 10⁸ M⁻¹ (Fig. S19B, ESI†), which was larger than that of THB, indicating that the binding ability between ATP and CP5 was stronger than that of THB.

Study of the mechanism for H₂O₂ detection. H₂O₂ is important for cell growth, proliferation and differentiation. Abnormal changes in H₂O₂ may lead to various diseases. The real-time monitoring of H₂O₂ in living cells can help detect some possible cellular lesions in time.

Thus, the development of a precise probe for the detection of H₂O₂ is also significant. Herein, we designed and synthesized a novel probe based on the Eu-complex with phenylboronic acid as a responsive moiety for H₂O₂ detection, in which the luminescence of Eu(III) could be sensitized by terpyridine and β-diketone ligands for real-time fluorescence sensing.

In comparison, the relationship between the fluorescence intensity of G₂ (at a constant concentration of 10 µM) and H₂O₂ (over the range from 0 to 10 mM) was carefully studied, as depicted in Fig. S20 (ESI†). On increasing the concentration of H₂O₂, the fluorescence intensity of G₂ diminished, signifying that it was “turned off” in response to H₂O₂. In addition, the fluorescence excitation and emission spectra of G₂ were also measured with a fixed slit width of 3 nm. The maximum excitation was at the 370 nm wavelength and the maximum emission wavelength was 618 nm. The ethanol solution of G₂ (10 µM) exhibited the typical emission of Eu(III) under 370 nm excitation. The narrow and strong linear emission was at 618 nm corresponding to the ⁵D₀ → ⁷F₂ transition. Several side peaks were distributed at 581 nm, 592 nm, 598 nm and 699 nm corresponding to the ⁵D₀ → ⁷F₁, ⁵D₀ → ⁷F₂, ⁵D₀ → ⁷F₃, and ⁵D₀ → ⁷F₄ transitions of Eu(III), respectively.

Another rare-earth complex, i.e., G₄ with phenol instead of the phenylboronic acid group, was also synthesized as a control to determine the typical fluorescence sensing of complexes with the TPBA ligand toward H₂O₂. As shown in Fig. S20(D) and (E) (ESI†), G₄ exhibited emission at 491 nm with a very weak shoulder peak at 615 nm, which was the characteristic emission of Eu(III). The fluorescence intensity of G₄ remained relatively stable upon increasing the concentration of H₂O₂, suggesting that G₄ had no response toward H₂O₂. This further confirmed that the TPBA ligand with the phenylboronic acid group played a crucial role in H₂O₂ detection. The phenylboronic acid group underwent oxidation to form a phenol group after reacting with H₂O₂, causing the breaking of energy transfer between TPBA and Eu(III), further turning off the fluorescence.

In summary, it could be envisaged that the supramolecular assembly of G₃⊂CP5 could act as a fluorescent probe for the simultaneous detection of H₂O₂ and ATP because CP5 exhibited good recognition toward ATP and the ligand containing phenylboronic acid groups in complex G₃ was responsive toward H₂O₂.

Optical properties and response to H₂O₂ and ATP

Determining the limits of detection. Compared to the assembly of G₂⊂CP5 with a size of ∼425 nm, the assemblies of G₁⊂CP5 and G₃⊂CP5 possessed smaller sizes, exhibiting more applicability as bioprobes. Thus, we focused on studying the optical properties of these two assemblies and their limits of detection for ATP and H₂O₂.

As shown in Fig. S21(A) and (B) (ESI†), on increasing the concentration of H₂O₂ from 0 to 2 mM, the fluorescence intensity of G₁⊂CP5 at 614 nm (the characteristic peak of Eu(III), ⁵D₀ → ⁷F₂) decreased and showed a “turn-off” behavior. The plots and fitted curve of the fluorescence intensity at 614 nm were then found to match the regression equation of y = −226467.90992x + 869435.7917 (Fig. S21(C), ESI†). The limit of detection (LOD) for G₁⊂CP5 toward H₂O₂ was calculated to be 0.36 mM (LOD = 3σ/K, σ is the standard deviation of the blank sample and K is the slope of the calibration curve).

Under the same condition, the fluorescence intensity of G₁⊂CP5 at 479 nm gradually increased upon the addition of ATP with the concentration ranging from 0 to 2 mM (Fig. S21(D) and (E), ESI†), suggesting that G₁⊂CP5 showed a “turn-on” behavior for ATP detection. Concurrently, the characteristic emission of Eu(III) at 614 nm also exhibited decreased fluorescence intensity. The plots and fitted curve of G₁⊂CP5 at 479 nm were then found to match the regression equation of y = 59432.6087x + 239568.91304 (Fig. S21(F), ESI†). The LOD for G₁⊂CP5 toward ATP was calculated to be 0.18 mM.

As for the assembly of G₃⊂CP5, the LODs for both H₂O₂ and ATP were also determined using the same experimental procedures as above (Fig. S22 and S23, ESI†). G₃⊂CP5 possessed significantly lower LODs for H₂O₂ and ATP with 0.11 mM and 0.05 mM, respectively, which could be ascribed to the higher proportion of detection moieties in G₃⊂CP5 compared with G₁⊂CP5. Thus, considering the lower LODs and better performance in detection, G₃⊂CP5 was chosen as the final and typical probe for subsequent investigation.

Study of the detection efficiency of G₃⊂CP5 for H₂O₂ and ATP. To further investigate the effect of introducing CP5 on the detection efficiency of H₂O₂ and the two-channel detection of H₂O₂ and ATP, the fluorescence performance of G₃⊂CP5 and the individual complex G₃ for H₂O₂ were explored and compared. On increasing the concentration of H₂O₂, the fluorescence intensity of G₃⊂CP5 at 615 nm showed a gradual decrease with an LOD of 0.2 mM, which was lower than that of complex G₃ (0.26 mM) (Fig. 3(A and B)). This demonstrated that the introduction of CP5 could improve the efficiency of detection for H₂O₂. Furthermore, the fluorescence intensity of probe G₃⊂CP5 was measured to detect various ROS including H₂O₂, ¹O₂, ClO⁻, and –OH. The PBS solutions of G₃⊂CP5 were added into the solution of ROS, respectively. On comparing the decreasing fluorescence intensity at 614 nm, it was found that the intensities of G₃⊂CP5 were decreased, indicating that the probe has the ability to detect all types of ROS including H₂O₂.

Fig. 3 Detection of ATP and H₂O₂ by the probes in PBS buffer solution (c = 10 µM): (A) fluorescence emission spectra of G₃⊂CP5 in PBS buffer with the addition of different concentrations of H₂O₂ (0–2 mM). (B) Linear fit of fluorescence intensity at 614 nm versus the concentration of H₂O₂. (C) Fluorescence emission spectra of G₃⊂CP5 in PBS buffer solution with the addition of different concentrations ATP (0–2 mM). (D) Linear fit of fluorescence intensity at 465 nm versus the concentration of ATP. Spectral crosstalk assessment experiments for G₃⊂CP5 (c = 10 µM): (E) fluorescence emission spectra measured in the presence of 2 mM ATP after adding different concentrations of H₂O₂ (0–2 mM) to the probe (inset: the fluorescence spectra before and after the addition of 2 mM ATP). (F) Linear fit of fluorescence intensity at 492 nm versus the concentration of H₂O₂. (G) Fluorescence emission spectra were measured in the presence of 2 mM H₂O₂ after adding different concentrations of ATP (0–2 mM) to the probe (inset: the fluorescence spectra before and after the addition of 2 mM H₂O₂). (H) Linear fit of fluorescence intensity at 490 nm versus the concentration of added ATP (λ_ex = 331 nm, slit width: 3 nm).

Moreover, the LOD of G₃⊂CP5 for ATP detection was also determined to be 0.07 mM, largely lower than that of G₃ to be 0.35 mM (Fig. 3(C) and (D)). This also demonstrated that the co-assembly of G₃ and CP5 through host–guest interaction would intensely improve the efficiency of detection for ATP.

Study of the two-channel and simultaneous detection for H₂O₂ and ATP using G₃⊂CP5. Considering the probe was mainly applied in the biological systems, the spectral crosstalk resistance was evaluated in vitro using PBS buffer with a pH value of 7.4 to simulate the physiological environment. The fluorescence response of G₃⊂CP5 to ATP in the H₂O₂ channel was firstly investigated. As depicted in Fig. 3(E and F), the intensity of the red fluorescence at 614 nm was essentially unchanged when ATP was added, but the blue fluorescence at 475 nm was significantly enhanced, indicating that G₃⊂CP5 could effectively respond to ATP in the ATP channel. Then, the detection of H₂O₂ was also evaluated in the presence of ATP upon the addition of different amounts of H₂O₂. The fluorescence intensity at 492 nm (ATP channel) was gradually enhanced, while the fluorescence intensity at 614 nm (H₂O₂ channel) was unchanged, indicating that the probe could respond to H₂O₂ in the ATP channel in the presence of large amounts of ATP. Notably, large amounts of H₂O₂ did not cause an increase in the ATP channel (Fig. 3(G) and (H)). However, when detecting ATP in the presence of H₂O₂, only a gradual increase of blue fluorescence could be observed on adding different amounts of ATP, suggesting that the ATP channel of G₃⊂CP5 was not sensitive to the addition of H₂O₂ and could only collect the ATP-related signals. These results confirmed that the presence of ATP would interfere with the detection of H₂O₂ but still gave output signals in the ATP channel. The presence of H₂O₂ would not interfere with the detection of ATP. Thus, the two-channel simultaneous detection of H₂O₂ and ATP could be achieved.

Targeted imaging and detection in living cells

Fluorescence imaging and mitochondria targeting. Considering the superior fluorescence of rare-earth complexes, we envisage that G₃⊂CP5 could act as a good fluorescence probe for intracellular imaging and detection.

Thus, the performance of fluorescence imaging was studied using confocal laser scanning microscopy (CLSM) and super-resolution fluorescence microscopy. Firstly, the cytotoxicity of G₃⊂CP5 was evaluated by the CCK-8 assay kit in RAW264.7 and HeLa cells, respectively. As shown in Fig. S25 (ESI†), G₃⊂CP5 exhibited no obvious toxicity toward both RAW264.7 macrophages and HeLa cells. The survival rates of RAW264.7 macrophages and HeLa cells remained above 92% and 86% after 24 h incubation, demonstrating the good biocompatibility of G₃⊂CP5 as a bioprobe. In order to evaluate the cell damage caused by UV light, the cytotoxicity of G₃⊂CP5 was evaluated by the CCK-8 assay kit in cells under 365 nm and 400 nm irradiation (Fig. S26, ESI†). Upon the irradiation of pulsed UV light, no cytotoxicity was observed in both RAW264.7 macrophages and HeLa cells, suggesting that UV light irradiation had negligible damaging effect on the cells.

Subsequently, cell uptake experiments were also conducted to verify the imaging of G₃⊂CP5 in living cells. As shown in Fig. 4(A), HeLa cells were shaped with clear cell contours, while the RAW264.7 macrophages in the same state were smaller in size and appeared in various shapes. This is because the RAW264.7 macrophages are semi-suspended cells. The suspended cells are circular and the attached cells are spindle-shaped. The green fluorescence of G₃⊂CP5 under 488 nm excitation could be observed. The green fluorescence did not overlap with the blue fluorescence of DAPI nuclear dye after staining, indicating that the probe was taken up in the cytoplasm and did not enter the nucleus. In addition, the probe exhibited stronger green fluorescence in HeLa cells compared with that in RAW264.7 macrophages, resulting from the higher level of ATP in the HeLa cells. In addition, the detailed cellular uptake of G₃⊂CP5 was also determined by the CLSM images (Fig. S27, ESI†), indicating that the maximum cellular uptake of G₃⊂CP5 was reached in about 6 h for both HeLa cells and RAW264.7 cells.

Fig. 4 Confocal laser scanning microscopy (CLSM) images of G₃⊂CP5 in HeLa cells and RAW264.7 cells. (A) Cellular uptake assay. (B) CLSM images of co-localization and targeted imaging in mitochondria. (C) The Pearson's correlation of mitochondrial targeting. (D) Fluorescence images of mitochondria observed through super-resolution fluorescence microscopy.

In order to verify the targeting ability of the probe, the co-localization and targeting experiments were also conducted in RAW264.7 macrophages and HeLa cells, respectively. Since the mitochondria are composed of double-layer membranes and the potential is positive outside the membrane, the negatively charged G₃⊂CP5 can target the mitochondria by electrostatic interaction with the outside membrane. Moreover, mitochondrion is one of the most important subcellular organelles to make energy and the primary place for the aerobic respiration of cells. On the one hand, mitochondria are in charge of the common pathway of final oxidation, including the tricarboxylic acid cycle and oxidative phosphorylation. Oxidative phosphorylation is a step that reduces oxygen and synthesizes ATP. Moreover, the inner membrane protein of mitochondria is the vector for ADP and ATP. On the other hand, more than 90% of the oxygen inhaled into the body is consumed by the mitochondria. Oxygen molecules will produce radical oxygen species. Under certain conditions, mitochondria can produce too many radical oxygen species, which will lead to the formation of tumor metastases. Thus, the mitochondrial targeting of the probe is crucial to detect ROS and ATP in living cells as well as monitor the energy metabolic process. As shown in Fig. 4(B) of the CLSM images, the green fluorescence of G₃⊂CP5 could be detected in the green channel upon 488 nm excitation.

As indicated by the co-staining results with a commercial mitochondrial agent (Mito-Tracker Red), G₃⊂CP5 specifically lightened up the mitochondria region. Furthermore, Pearson's correlation coefficient was calculated to be 0.75 for HeLa cells and 0.91 for RAW264.7 macrophages, respectively (Fig. 4(C)). The above results indicated that G₃⊂CP5 had an excellent ability for mitochondria targeting. Besides, the Pearson's correlation coefficient of G₃⊂CP5 in the HeLa cells was higher than that for the RAW264.7 macrophages, which is ascribed to the higher potential of the mitochondrial membrane in the HeLa cells. At the same time, better mitochondria targeting also resulted from the active metabolism and a large amount of ATP production. Furthermore, the lysosome targeting of G₃⊂CP5 was also measured. As shown in Fig. S28 (ESI†), G₃⊂CP5 exhibited weak light in the lysosome region. Pearson's correlation coefficients were calculated to be 0.39 for HeLa cells and 0.52 for RAW264.7 macrophages, indicating that the probe G₃⊂CP5 does not target the lysosome. The results confirmed the ATP and H₂O₂ detection mainly occurred in the mitochondria.

Detection of H₂O₂ and ATP in living cells. The ability of G₃⊂CP5 to detect ATP and H₂O₂ in living cells was also evaluated based on their great mitochondrial targeting. As depicted in Fig. 5, compared with individual G₃⊂CP5 without either the addition of H₂O₂ or ATP, G₃⊂CP5 + H₂O₂ exhibited relatively increased red fluorescence in the red channel and weaker fluorescence in the blue channel upon the 407 nm laser excitation. G₃⊂CP5 + ATP showed stronger blue fluorescence and nearly unchanged red fluorescence. Moreover, both G₃⊂CP5 + H₂O₂ and G₃⊂CP5 + ATP possessed stronger red and blue fluorescence in the two channels, indicating that the HeLa cells had an inherently strong metabolism of H₂O₂ and ATP. As for the control groups, no fluorescence could be detected in the groups of PBS, PBS + H₂O₂ and PBS + ATP both in the HeLa cells and RAW264.7 macrophages because of the absence of the fluorescent probe (Fig. S29 and S30, ESI†). As a consequence, the two-channel simultaneous fluorescence detection of H₂O₂ and ATP could be achieved in living cells using G₃⊂CP5 as the bio-probe. G₃⊂CP5 had excellent biocompatibility, targeted imaging in the mitochondria and simultaneous detection of H₂O₂ and ATP. Thus, it paves a new strategy for the design of two-channel bio-probes, providing opportunities to study the interactions between substrates in living cells.

Fig. 5 Fluorescence detection of additional H₂O₂ and ATP in HeLa cells and RAW264.7 cells (red channel: 570–620 nm; blue channel: 425–475 nm).

Conclusions

In this work, we designed and synthesized a novel rare-earth supramolecular fluorescent probe (G₃⊂CP5). The real-time fluorescence detection of H₂O₂ and ATP in living cells was achieved with low LODs. On the one hand, the electron-rich cavities of CP5 formed a stable inclusion with the Eu(III)-complexes by binding with the quaternary ammonium salts moieties. The host–guest supramolecular assembly realized a reduction in energy dissipation, improvement of the fluorescence efficiency and regulation of the luminescence properties. On the other hand, the introduction of water-soluble CP5 improved the water solubility and biocompatibility. G₃⊂CP5 could respond to H₂O₂ and ATP in the two-channel mode due to the interacting moieties and host–guest recognition. Significantly, G₃⊂CP5 exhibited excellent mitochondrial targeting. The intracellular detection of H₂O₂ and ATP was also achieved in both HeLa cells and RAW264.7 macrophages. This work paves the way for the development of fluorescent probes for multiple substrates detection based on supramolecular assemblies. The design provides opportunities to monitor the dynamic information such as intracellular microenvironment and substrate migration, further facilitating the early diagnosis of diseases.

Author contributions

Y. Tang and N. Song conceived the project; S. Bao conducted experiments in consultation with Y. Tang; S. Bao, C. Wang, Y. Kou and D. Lv were responsible for the syntheses and characterization of materials; Y. Guo and D. Lv completed the cell experiments. S. Bao, Y. Yang and N. Song prepared the draft of manuscript; Y. Tang and N. Song contributed to the editing of the manuscript.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We acknowledge the financial support from the National Natural Science Foundation of China (22221001, 22131007), the Science and Technology Major Plan of Gansu Province (23ZDGA012), and the 111 project (B20027).

Notes and references

1. L. Lu, H.-J. Sun, Y.-T. Zeng, Y. Shao, M. V. Bermeshev, Y. Zhao, B. Sun, Z.-J. Chen, X.-K. Ren and M. Zhu, J. Mater. Chem. C, 2020, 8, 10422–10430.
2. K. Sakakibara, Y. Takahashi, R. Nishiyabu and Y. Kubo, J. Mater. Chem. C, 2017, 5, 3684–3691.
3. Q. Zan, K. Zhao, R. Li, Y. Yang, X. Yang, W. Li, G. Zhang, C. Dong, S. Shuang and L. Fan, Anal. Chem., 2024, 96, 10488–10495.
4. D. Andina, J. C. Leroux and P. Luciani, Chem. – Eur. J., 2017, 23, 13549–13573.
5. Y. Tian, S. Liu, W. Cao, P. Wu, Z. Chen and H. Xiong, Anal. Chem., 2022, 94, 11321–11328.
6. P. Lin, S. Jiang, T. Liu, X. Yuan, K. Luo, C. Xie, X. Zhao and L. Zhou, Analyst, 2024, 149, 638–664.
7. L.-Y. Geng, Y. Zhao, E. Kamya, J.-T. Guo, B. Sun, Y.-K. Feng, M.-F. Zhu and X.-K. Ren, J. Mater. Chem. C, 2019, 7, 2640–2645.
8. J. Liu, W. Zhang, X. Wang, Q. Ding, C. Wu, W. Zhang, L. Wu, T. D. James, P. Li and B. Tang, J. Am. Chem. Soc., 2023, 145, 19662–19675.
9. H. Wang, R. Zhang, K. R. Bridle, A. Jayachandran, J. A. Thomas, W. Zhang, J. Yuan, Z. P. Xu, D. H. G. Crawford, X. Liang, X. Liu and M. S. Roberts, Sci. Rep., 2017, 7, 45374.
10. C. Y. Ang, S. Y. Tan, S. Wu, Q. Qu, M. F. E. Wong, Z. Luo, P.-Z. Li, S. Tamil Selvan and Y. Zhao, J. Mater. Chem. C, 2016, 4, 2761–2774.
11. H. Chen, D. Zou, W. Cao, Y. Cui, G. Qian and Z. Liao, J. Mater. Chem. C, 2024, 12, 11012–11019.
12. G. George, C. S. Edwards, J. I. Hayes, L. Yu, S. R. Ede, J. Wen and Z. Luo, J. Mater. Chem. C, 2019, 7, 14949–14961.
13. K. Wang, R. Guo, X.-Y. Chen, Y.-S. Yang, L.-Q. Qiao and M.-L. Wang, Chem. Eng. J., 2023, 455, 140491.
14. L. Jiang, H.-Y. Chen, C.-H. He, H.-B. Xu, Z.-R. Zhou, M.-S. Wu, E. K. Fodjo, Y. He, M. E. Hafez, R.-C. Qian and D.-W. Li, Anal. Chem., 2023, 95, 3507–3515.
15. Y. Shi, M. Tang, C. Sun, Y. Pan, L. Liu, Y. Long and H. Zheng, CCS Chem., 2019, 1, 373–383.
16. V. P. Sruthi and S. Senthilkumar, J. Mater. Chem. C, 2024, 12, 8924–8934.
17. H. Chen, J. Cao, P. Zhou, X. Li, Y. Xie, W. Liu and Y. Tang, Biosens. Bioelectron., 2018, 122, 1–7.
18. Z. Zhao, J. Ru, P. Zhou, Y. Wang, C. Shan, X. Yang, J. Cao, W. Liu, H. Guo and Y. Tang, Dalton Trans., 2019, 48, 16952–16960.
19. X. Zheng, R. Fan, K. Xing, A. Wang, X. Du, P. Wang and Y. Yang, J. Mater. Chem. C, 2018, 6, 6229–6239.
20. C. Liu, J. Liu, H. Ma, W. Zhang, B. Song, L. Guo, R. Zhang and J. Yuan, Methods, 2019, 168, 102–108.
21. R. Sivakumar and N. Y. Lee, Coord. Chem. Rev., 2024, 501, 215563.
22. P. Su, L. Liang, T. Wang, P. Zhou, J. Cao, W.-S. Liu and Y. Tang, Chem. Eng. J., 2021, 413, 127408.
23. L. Yu, C. He, Q. Zheng, L. Feng, L. Xiong and Y. Xiao, J. Mater. Chem. C, 2020, 8, 3562–3570.
24. Y. Zhou, H.-Y. Zhang, Z.-Y. Zhang and Y. Liu, J. Am. Chem. Soc., 2017, 139, 7168–7171.
25. H. Chen, Y. Xie, A. M. Kirillov, L. Liu, M. Yu, W. Liu and Y. Tang, Chem. Commun., 2015, 51, 5036–5039.
26. K. Luan, R. Meng, C. Shan, J. Cao, J. Jia, W. Liu and Y. Tang, Anal. Chem., 2018, 90, 3600–3607.
27. P. Su, X. Wang, T. Wang, X. Feng, M. Zhang, L. Liang, J. Cao, W. Liu and Y. Tang, Talanta, 2021, 225, 122063.
28. X. Zhai, Y. Kou, L. Liang, P. Liang, P. Su and Y. Tang, Inorg. Chem., 2023, 62, 18533–18542.
29. P.-R. Su, T. Wang, P.-P. Zhou, X.-X. Yang, X.-X. Feng, M.-N. Zhang, L.-J. Liang, Y. Tang and C.-H. Yan, Natl. Sci. Rev., 2022, 9, nwab016.
30. L. Tian, H. Ma, B. Song, Z. Dai, X. Zheng, R. Zhang, K. Chen and J. Yuan, Talanta, 2020, 212, 120760.
31. H. Yan, H. Ni, Y. Yang, C. Shan, X. Yang, X. Li, J. Cao, W. Wu, W. Liu and Y. Tang, Chin. Chem. Lett., 2020, 31, 1792–1796.
32. X. Zhai, P. Feng, N. Song, G. Zhao, Q. Liu, L. Liu, M. Tang and Y. Tang, Inorg. Chem. Front., 2022, 9, 1406–1415.
33. W. Zhang, Y. G. Abou El-Reash, L. Ding, Z. Lin, Y. Lian, B. Song, J. Yuan and X.-D. Wang, Microchim. Acta, 2018, 185, 533.
34. X.-S. Li, Y.-F. Li, J.-R. Wu, X.-Y. Lou, J. Han, J. Qin and Y.-W. Yang, J. Mater. Chem. A, 2020, 8, 3651–3657.
35. X. Wang, J. Yang, X. Sun, H. Yu, F. Yan, K. Meguellati, Z. Cheng, H. Zhang and Y.-W. Yang, Chem. Commun., 2018, 54, 12990–12993.
36. W.-L. Zhou, Y. Chen, W. Lin and Y. Liu, Chem. Commun., 2021, 57, 11443–11456.
37. J. Zhou, M. Chen and G. Diao, Chem. Commun., 2014, 50, 11954–11956.
38. G. Yu, J. Zhou, J. Shen, G. Tang and F. Huang, Chem. Sci., 2016, 7, 4073–4078.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tc03476g

‡ The authors contributed equally to this work.

---