Amido/alkoxy–aryl–aryl–picolinate push–pull antennas for two-photon sensitization of Eu³⁺ luminescence

Abstract

Luminescent two-photon (2P) absorbing lanthanide(III) complexes hold great promise for microsocpy imaging of biological samples. Conjugating such a complex to well-chosen cell penetrating peptides (CPP) allows its controlled delivery to the cytosol of live cells. However, alkoxy–phenyl–ethynyl–picolinate, one of the best antennae for 2P sensitization of Eu³⁺, undergoes side reactions at its ethynyl group during peptide synthesis or in biological media and thus cannot be used to create such a conjugate. In this article, we evaluate the effect of substituting the ethynyl group by a phenyl one. We describe the synthesis of conjugates of the TAT CPP with Eu³⁺ complexes featuring amido–phenyl–phenyl–picolinamide, alkoxy–phenyl–phenyl–picolinamide and amido–phenyl–phenyl–picolinate ter-aryl antennae and compare their spectroscopic properties to those of analogues with bi-aryl antennae, including the amido–phenyl–picolinamide already used for 2P live cell imaging. The absorption spectrum of the ter-aryl antennae is red-shifted and better covers the active spectral range for 2P excitation by a Ti-sapphire laser. Among compounds with ter-aryl antennae, those with an amido electron donating group are the most interesting, showing brightness ca. 4 times higher than their bi-aryl counterparts, and similar to the ethynyl-containing antenna. 2P microscopy imaging of live cells incubated with the TAT-Eu³⁺ conjugate and dFFLIPTAT, a non-luminescent CPP that promotes cytosolic delivery, showed diffuse cytosolic staining of the Eu³⁺ probe. The ter-aryl-based probes showed superior performances compared to bi-aryl, with ca. 80% of the cells showing Eu³⁺ staining of the cytosol.

---

Introduction

Lanthanide(III) (Ln³⁺) complexes have luminescent properties due to f–f transitions that are very interesting for biological applications, including fine emission bands (line widths of a few nanometers) at fixed wavelengths characteristic of each lanthanide, long luminescence lifetimes and high resistance to photobleaching.^1–6 A major drawback to their use is the low efficiency of light absorption by Ln³⁺ with ε values in the order of 1–5 M⁻¹ cm⁻¹ for f–f transitions. This can be overcome by introducing a chromophore close to the lanthanide, capable of transferring to the lanthanide ion the energy it has acquired through light excitation, to bring it into its emissive excited state. This chromophore is called an antenna and the process of sensitizing the luminescence of the lanthanide is referred to as the antenna effect.^7–10 Sensitization of Ln³⁺ that emit in the visible (Tb³⁺, Dy³⁺, Eu³⁺ and Sm³⁺) requires an antenna whose donor excited levels are above the Ln³⁺ excited emissive level (typically 2000–5000 cm⁻¹ above). Consequently, for these lanthanides, it is mandatory to use antennae that are excited with UV light, which has the disadvantage of causing damage to biological samples. One solution to this problem is to use antennae that can be excited by two-photon (2P) absorption, a non-linear optical phenomenon that involves the simultaneous absorption of two photons instead of one, with half the energy required for excitation by a single photon.^11–14 As a result, the excitation is shifted from the UV to the near infrared, causing less damage to cells, for instance.

Over the last fifteen years, we have demonstrated that it is possible to perform 2P microscopy of fixed cells (i.e. cells treated with PFA or MeOH resulting in membrane permeabilization, but also causing cell death) with Ln³⁺ complexes based on a macrocyclic ligand TACN(pic)₃ (Fig. 1) and equipped with 2P-absorbing push–pull antennae.^13,15–17 More recently, we have used Ln³⁺ complexes based on a DO3Apic ligand equipped with a push–pull antenna and conjugated to a cell-penetrating peptide that enables the Ln complex to be delivered into the cytosol of living cells.^18–20 2P microscopy imaging of live cells was performed using p-amido- or p-methoxy–phenyl–picolinamide antennae (A1 and A2 in Fig. 1, respectively) coordinated to the Ln³⁺. These antennae show an absorption band in the UV (Fig. 1) with a maximum at 315 nm (λ_max) and that extends up ca. 360 nm (λ_cut-off). The Ti:sapphire lasers used for 2P excitation have an excitation wavelength between about 700 nm and 1050 nm, corresponding to the 350–525 nm range for 1P absorption. This means that only the tail of the absorption band of the Ln³⁺ complex is likely to be active for 2P excitation by the Ti:sapphire laser (Fig. 1). Additionally, the 2P cross-section of these antennae is small (35 GM at 700 nm for the TACN-based complex with three antennae, i.e. ca. 12 GM per antenna).

Fig. 1 Eu complexes with TACN-trispicolinate and DO3A-picolinate/picolimamide ligands, sensitizing push–pull antennas and their absorption spectra.

A better antenna is the alkoxy–phenyl–ethynyl–picolinate antenna (A3 in Fig. 1), with λ_max and λ_cut-off values around 325 nm and 370 nm, respectively, and a 2P cross-section of ca. 50 GM at 700 nm.^21,22 However, its alkyne group is not completely stable and can react with nucleophiles (water, alcohols, thiols, phosphines, etc.) under certain conditions. For instance, under acidic conditions used during the preparation of Ln³⁺ complex-peptide conjugates (TFA treatment for protecting group removal), complete hydration of the triple bond or addition of other nucleophiles (e.g. the thioanisol scavenger) is observed.¹⁸ A solution to red-shift the absorption of the aryl-picolinate antenna was recently proposed, that consists in replacing the p-methoxy–phenyl–ethynyl (antenna A3 in Fig. 1) substituent of the picolinate with a p-thioanisolyl substituent (antenna A4 in Fig. 1).²³ Such an antenna shows λ_max and λ_cut-off values of ca. 330 nm and 375 nm, respectively, similar to the p-methoxy–aryl–ethynyl–picolinate antenna. A 2P cross-section of 35 GM at 700 nm was measured for the TACN-based complex that features three antennae. This accounts for ca. 12 GM per antenna. In this article, we explore an alternative solution: replacing the ethynyl linker with an aryl group. We describe the preparation and the luminescence properties of several Eu³⁺ complexes featuring p-amido- or p-methoxy–aryl–aryl–picolinamide antennae (A5 and A6 in Fig. 1, respectively), denoted here as ter-aryl antennae, in comparison with the corresponding bi-aryl p-amido- or p-methoxy–aryl–picolinamide antennae.

Results and discussion

Design and synthesis of the probes

In a previous article, we have described the conjugate mTAT[Eu·L-Ar-NHAc] (Fig. 2A). This compound comprises an Eu³⁺ ion bound to a cyclen-based macrocyclic ligand, DO3Apic, featuring a π-extended picolinamide antenna. The picolinamide moiety is substituted by a p-acetamido–phenyl group, in para position relative to the pyridine, the whole forming a bi-aryl push–pull antenna for 2P absorption. The Eu³⁺ complex is attached to a TAT peptide, which provides water solubility and cell penetration properties. Our first target probe was therefore the ter-aryl analogue mTAT[Eu·L-Ar-Ar-NHAc] (Fig. 2A). Conjugates mTAT[Eu·L-Ar-OMe] and mTAT[Eu·L-Ar-Ar-OMe] with a methoxy electron donating group instead of the acetamide group were also investigated (Fig. 2A).

Fig. 2 Chemical structures of (A) mTAT[Ln·L-R] conjugates (R = Ar-OMe, Ar-NHAc, Ar-Ar-OMe and Ar-Ar-NHAc; Ln = Eu³⁺ or Gd³⁺) and (B) the mTAT[SucNHArArL·Ln] conjugate.

The three new probes were synthesized as already reported for mTAT[Eu·L-Ar-NHAc].¹⁹ The iodinated compound 1 (Fig. 3A) was engaged in a Miyaura–Suzuki coupling with borylated aromatics 3b–d to yield pro-ligands L-R(tBu)₃ after hydrolysis of the picolinic methyl ester. All pro-ligands were obtained in similar yields (34–48%) after HPLC purification. Preparation of the conjugates was performed as already described.¹⁹ L-R(tBu)₃ pro-ligands were then coupled to the selectively deprotected lysine side chain of the peptide, on resin using PyBOP/DIEA activation (Fig. S6 of ESI†). The peptide was cleaved from the resin and all protecting groups, including the macrocyclic ligand tBu ones, were removed by TFA/scavenger treatment for 4 h. The conjugates were purified by HPLC and metalation with Eu³⁺ was performed in water at pH 8 in the presence of excess EuCl₃ to give mTAT[Eu·L-Ar-OMe], mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[Eu·L-Ar-Ar-OMe] as pure compounds after HPLC purification as attested by LCMS analysis (Fig. S8 of ESI†).

Fig. 3 Synthetic pathways for pro-ligands (A) L-R(tBu)₃ (R = Ar-NHAc,¹⁹ Ar-OMe, Ar-Ar-NHAc and Ar-Ar-OMe) and (B) L-Ar-Ar-NHSuc(Me)₄. The carboxylic acid group used for peptide conjugation is highlighted in red.

Due to the disappointing luminescence properties of mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[Eu·L-Ar-Ar-OMe] (vide infra), we have also investigated the influence of the peptide anchoring position onto the complex. For this purpose, mTAT[SucNH-Ar-Ar-L·Eu] was prepared, in which the ter-aryl antenna is attached to the peptide through the electron donating group rather than through the picolinate moiety. This results in a negatively charged Eu³⁺ complex with a picolinate coordinating group rather than a neutral complex with a picolinamide group (Fig. 2). The synthesis of the pro-ligand L-Ar-Ar-NHSuc(Me)₄ and its TAT conjugate were performed starting from the iodinated compound 2. It was coupled to boronic ester 3e to form L-Ar-Ar-NH₂(Me)₄, followed by treatment with succinic anhydride to give pro-ligand L-Ar-Ar-NHSuc(Me)₄ (Fig. 3B), which was then coupled to the mTAT peptide on resin (Fig. S7 of ESI†). After resin cleavage in TFA and HPLC purification, the methyl protecting groups of the carboxylates were removed using a NaOH 2 M treatment for 15 minutes prior to metalation with EuCl₃ in water, which gave mTAT[SucNH-Ar-Ar-L·Eu].

Photophysical properties

The photophysical properties of these probes were investigated in PBS buffer (pH 7.4). Spectroscopic data are summarized in Table 1.

Table 1 Spectroscopic characterizations in PBS pH 7.4^a

Compound	λ max; λcut-off/nm	E(S1);^bE(T1)^c/cm^−1	ε at λmax/M^−1 cm^−1	Φ Eu	Φ ^EuEu	η sens	σ 2P at 720 nm/GM	B 2P at 720 nm/GM	τ Eu (H2O, aerated); τEu (H2O, de-oxygenated); τEu (D2O, aerated)/ms	τ R/ms
a Error is estimated ±5% on ε values and ±10% on ΦLn and σ2P. Error on τLn is estimated ± 0.03 ms. b Energy of the excited singlet state is determined from the λcut-off. c Energy of the excited triplet state is estimated from the wavelength at half-maximum on the onset of the time-gated phosphorescence spectrum of the Gd^3+ analogue in PBS/glycerol 9 : 1 v/v recorded at 77 K.
mTAT[Eu·L-Ar-NHAc]	315; 366	S1: 27 300; T1: 22 000	20 000	0.15	0.22	0.67	5.3	0.8	1.05; 1.05; 1.63	4.68
mTAT[Eu·L-Ar-OMe]	318; 364	S1: 27 500; T1: 22 000	21 000	0.15	0.22	0.67	3.5	0.5	1.05; 1.05; 1.65	4.56
mTAT[Eu·L-Ar-Ar-NHAc]	329; 384	S1: 26 000; T1: 20 400	29 000	0.075	0.16	0.48	33	2.5	0.71; 0.76; 0.99	4.56
mTAT[Eu·L-Ar-Ar-OMe]	333; 390	S1: 25 600; T1: 20 400	24 000	0.035	0.086	0.41	41	1.4	0.39 (93%), 0.77 (7%); 0.40 (92%), 0.78 (8%); 0.46 (89%), 0.89 (11%)	4.56
mTAT[SucNH-Ar-Ar-L·Eu]	326; 381	S1: 26 300; T1: 20 700	34 000	0.15	0.22	0.67	24	3.6	1.02 1.03 1.51	4.54

---

Absorption. All compounds show a strong absorption in the UV assigned to an intra-ligand charge transfer (ILCT) transition within the antenna. The absorption spectra of the two probes with bi-aryl antennas, mTAT[Eu·L-Ar-NHAc] and mTAT[Eu·L-Ar-OMe], are almost identical with maximum and cut-off wavelengths at ca. 316 nm and 365 nm, respectively (Fig. 4A and B). In comparison with the bi-aryl compounds, their two ter-aryl analogues, mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[Eu·L-Ar-Ar-OMe] have red-shifted absorption, ca. 15 and 20 nm, respectively (Fig. 4A). The former has a higher molar absorption. The third ter-aryl-based compound, mTAT[SucNH-Ar-Ar-L·Eu], anchored through the electron-donating group, is also red-shifted compared to the bi-aryl compounds but less than the two other ter-aryl compounds (Fig. 4A), but it presents the highest molar absorption among ter-aryl compounds (Fig. 4A and Table 1). Therefore, all ter-aryl compounds have a λ_max around 330 nm (±5 nm), close to the ones displayed by p-alkoxyphenyl–ethynyl–picolinate and thioanisolyl–picolinate antennas. Interestingly, at 350 nm and above, the absorption of the three ter-aryl compounds is very similar and significantly higher than the bi-aryl ones.

Fig. 4 1P Spectroscopic characterization of mTAT[Eu·L-R] (R = Ar-AcNH, Ar-OMe, Ar-Ar-NHAc and Ar-Ar-OMe) and mTAT[SucNH-Ar-Ar-L·Eu] probes in PBS: (A) Absorption spectra, (B) normalized emission spectra (λ_ex = 315 nm for mTAT[Eu·L-Ar-NHAc] and mTAT[Eu·L-Ar-OMe] and 330 nm for the others) recorded in PBS pH 7.4.

Eu³⁺ emission. All five compounds show a characteristic Eu³⁺ emission with bands at 580, 590, 615, 650 and 700 nm corresponding to the ⁵D₀ → ⁷F_J transitions (J = 0–4) when excited into the ILCT band (Fig. 4B and Fig. S9†). Emission spectra of Eu³⁺ complexes are known to be very sensitive to the coordination sphere.²⁴ The five Eu³⁺ emission spectra are very similar. This is in agreement with an identical N₅O₄ coordination environment forced by the DO3Apicolinamide/picolinate ligand. Indeed, the para substituent of the picolinate group has only minimal effect on the coordination and on the Eu³⁺ emission spectrum. In all cases, the excitation spectrum (Fig. S9†) matches well the absorption spectrum indicating Eu³⁺ sensitization through the ILCT band of the antenna. While the conjugates with the bi-aryl antennae, mTAT[Eu·L-Ar-NHAc] and mTAT[Eu·L-Ar-OMe], have an Eu³⁺ emission quantum yield, Φ_Eu, of 0.15 and an Eu³⁺ emission lifetime, τ_Eu, of 1.05 ms, their ter-aryl analogues, mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[Eu·L-Ar-Ar-OMe], show significantly lower quantum yields (0.075 and 0.035, respectively, Table 1 and Fig. S10†) and shorter emission lifetimes (0.7 and 0.4 ms, respectively, Table 1). The shorter lifetime indicates an additional non-radiative de-excitation pathway for the Eu³⁺ excited state of the latter two, while the lower quantum yield is the result of either a lower sensitization efficiency of the antenna or an additional de-activation pathway or both. Strikingly, mTAT[SucNH-Ar-Ar-L·Eu] shows Eu³⁺ emission quantum (0.15) and lifetimes (1.02 ms) almost identical to those of the bi-aryl compounds.

Mechanistic insights. Common pathways for Ln³⁺ excited states de-activation are energy transfer to the overtones of O–H vibrations, especially those of coordinated water molecules, and back energy transfer to the antenna excited triplet states.^1,8 In order to gain more insight into the deactivation mechanism, the number of coordinated water molecules, q, and the excited triplet state energy were determined. Regarding q, the Eu³⁺ emission lifetimes were measured in PBS buffer prepared with H₂O and D₂O (Table 1) and q was determined from Parker's equation q = 1.2 × (1/τ_Eu(H₂O) − 1/τ_Eu(D₂O) – c) with τ_Eu in ms and with c = 0.325 to account for the picolinamide ligands, which includes a coordinated CONHR group (Table S1†).²⁵ In all cases, the q value was zero (Table S1†), confirming that the DO3Apicolinamide ligand saturates Eu³⁺, whatever the antenna, in agreement with the similarity of the Eu³⁺ spectra. The excited triplet state energy was determined from the emission spectra at 77 K of Gd³⁺ analogue in frozen PBS/glycerol 9/1 mixture (Fig. S11†). The triplet state of mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[Eu·L-Ar-Ar-OMe] are ca. 1500 cm⁻¹ lower than those of their bi-aryl counterparts but ca. 3000 cm⁻¹ higher than the Eu^{3+ 5}D₀ excited state. Such an energy gap precludes any back energy transfer from the Eu^{3+ 5}D₀ state to the antenna triplet state, which is also confirmed by the absence of O₂ sensitivity of the Eu³⁺ emission lifetime (Table 1). The additional de-activation processes at play in mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[Eu·L-Ar-Ar-OMe] are not due to additional coordinated water molecules or a low-lying triplet state and remain unexplained.

In the case of Eu³⁺, the metal-centred emission quantum yield (Φ^Eu_Eu, also named intrinsic quantum yield) can be determined from the emission spectrum using eqn (1) and (2), where τ_R is the radiative lifetime of Eu³⁺, n is the refractive index of the medium, I_tot and I_MD are the total area of the corrected Eu³⁺ emission spectrum and the area of the ⁵D₀ → ⁷F₁ transition band, respectively, and A_MD,0 is the spontaneous emission probability for the ⁵D₀ → ⁷F₁ transition.²⁶A_MD,0 was initially assumed to be constant and equal to 14.65 s⁻¹,²⁶ but significant deviations from this value were recently reported.²⁷ Therefore, these equations have to be used with caution when comparing complexes. Nevertheless, (i) as all the complexes described here are based on the DO3Apic chelator, (ii) as they have nearly identical emission spectra and (iii) as they have the same hydration state (q = 0), the derived τ_R and Φ^Eu_Eu values can be compared. Their values are given in Table 1.

Φ^Eu_Eu = τ_Eu/τ_R (1)

1/τ_R = A_MD,0 × n³ × (I_tot/I_MD) (2)

In agreement with their nearly identical emission spectra, all compounds have the same τ_R value, ca. 4.6 ms, and the differences in their intrinsic quantum yield values follow the differences in the τ_Eu values.

The sensitizing efficiency, η_sens, which quantifies the efficiency of the electronic energy transfer from the antenna to the Eu³⁺, can be calculated from eqn (3).

Φ_Eu = η_sens × Φ^Eu_Eu (3)

The conjugates with the bi-aryl antennae, mTAT[Eu·L-Ar-NHAc] and mTAT[Eu·L-Ar-OMe], show the same sensitization efficiency, 0.67, but their ter-aryl analogues, mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[Eu·L-Ar-Ar-OMe], show significantly lower η_sens values, i.e. 0.48 and 0.41, respectively. With push–pull antennas, photoinduced electron transfer (PeT) from the excited antenna to the Eu³⁺ is likely to compete with the electronic energy transfer.^20,28 The Gibbs energy of the PeT process can be evaluated using eqn (4).^29–33

ΔG⁰_eT = (E⁰_D − E⁰_A) − E* + ΔG_S+C (4)

where E⁰_D is the oxidation potential of the electron donor (anisole or acetanilide moieties), E⁰_A is the reduction potential of the acceptor (Eu³⁺), E* is the excited state energy of the antenna, i.e. E(S₁) in Table 1. The last term accounts for solvation and coulombic interactions within the donor/acceptor pair after electron transfer. It can be estimated ca. −0.15 eV.^31–33 The oxidation potential of anisole and acetanilide donors were determined at E⁰_ox = 1.8 and 1.9 V per NHE by cyclic voltammetry measurements using N-(4′-iodo-[1,1′-biphenyl]-4-yl)acetamide and 4-bromo-4′-methoxy-1,1′-biphenyl as model compounds, in line with values reported in the literature.³⁴ The reduction potential of Eu³⁺ bound to the DO3Apicolinamide chelator has been determined at −1.2 V per NHE. From the E(S1) values in Table 1, we can calculate that −0.6 eV < ΔG⁰_eT < −0.3 eV for the mTAT[Eu·L-R] conjugates, indicating that PeT is possible within these compounds. A difference in the rate constants of the PeT and the energy transfer processes may explain the difference observed in η_sens within this family. A PeT process between the excited Eu³⁺ and the antenna can also happen. The Gibbs energy of such a process can be calculated taking for E* the energy of the ⁵D₀ state of Eu³⁺, i.e. 2.14 eV. In this case, ΔG⁰_eT > 0.6 eV, indicating that such a PeT process is not possible and cannot be involved in the de-activation of the Eu^{3+ 5}D₀ excited state. Therefore, a PeT process may be responsible for the lower η_sens in mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[Eu·L-Ar-Ar-OMe] but not for the lower τ_Eu, which remains to be clarified.

Influence of the peptide anchoring point. At this stage, we reasoned that the efficiency of PeT process could be lowered by shifting from a neutral Eu³⁺ complex to an anionic one. Indeed, it has been demonstrated that the reduction potential of Eu³⁺ decreases by ca. 100 mV replacing an amide by a carboxylate in the coordination sphere.²⁸ This prompted us to move the peptide anchoring point on the complex to the electron donating group of the antenna in order to have a picolinate instead of a picolinamide in the Eu³⁺ coordination sphere and thus we designed mTAT[SucNH-Ar-Ar-L·Eu]. Interestingly, this modification restores both a high sensitization efficiency (η_sens = 0.67) and a long Eu³⁺ emission lifetime (τ_Eu = 1.05 ms), comparable to those of the bi-aryl antennae. Consequently, mTAT[SucNH-Ar-Ar-L·Eu] is the most interesting ter-aryl compound: its absorption is red-shifted compared to the bi-aryl compounds but its overall quantum yield remains unchanged (Φ_Eu = 0.15).

Two-photon absorption. The 2P-absorption properties of the conjugates were studied in PBS using the two-photon excited fluorescence method with a Ti:sapphire laser source. When the conjugates in PBS solution are excited at 730 nm with the Ti:sapphire laser, the Eu³⁺ emission is detected with a quadratic dependence of its intensity on the laser power, which confirms biphotonic excitation (Fig. 5A). The 2P absorption spectra are show in Fig. 5B and compared to the 1P absorption spectrum in Fig. S12.† For all compounds, an excellent agreement is observed between the wavelength-doubled 1P absorption spectrum and the 2P absorption spectrum above 680 nm. This indicates that the ILCT transition of the antenna that is responsible for Eu³⁺ sensitization is 2P allowed as expected for non-centrosymmetric compounds. At 720 nm, the excitation wavelength used in 2P microscopy experiments (vide infra), the 2P cross-sections (σ_2P) of the three ter-aryl-based compounds are much higher than those of the bi-aryl ones (24–41 GM vs. 3.5–5.3 GM, respectively), showing a significant improvement of the 2P absorption efficiency by extending the antenna with a phenyl group. Among ter-aryl compounds, mTAT[Eu·L-Ar-Ar-OMe] has the highest σ_2P and mTAT[SucNH-Ar-Ar-L·Eu] the lowest. However, taking into account the Eu³⁺ emission quantum yields, the reversed trend is observed concerning the 2P brightness, B_2P = σ_2P × Φ_Eu (Table 1). The cross-section values at 720 nm of the three ter-aryl compounds are comparable to those determined for the alkoxy–phenyl–ethynyl–picolinate antenna (antenna A3 in Fig. 1, ca. 35 GM per antenna at 720 nm) and higher than those of the thioanisolyl–picolinate antenna (antenna A4 in Fig. 1, ca. 8 GM per antenna at 720 nm) in related Eu³⁺ complexes. This demonstrates that the replacement of the ethynyl group by a phenyl in alkoxy–phenyl–ethynyl–picolinate antennas is a valuable solution to reactivity issues encountered with the alkyne group.

Fig. 5 2P spectroscopy: (A) quadratic power dependence of the Eu³⁺ emission (λ_ex = 730 nm) for mTAT[Eu·L-R] (R = Ar-OMe, Ar-Ar-NHAc and Ar-Ar-OMe) and mTAT[SucNH-Ar-Ar-L·Eu] in PBS. Data were fitted using I = A × Pⁿ yielding n = 1.85, 1.87, 1.99 and 1.97, respectively. (B) 2P absorption spectra measured in PBS. The vertical dotted line corresponds to the 2P excitation wavelength used for cell microscopy imaging. Measurements were performed in PBS pH 7.4.

Two-photon microscopy

The mTAT conjugates with the ter-aryl antennae and an amido electron donating group showed better emission properties compared to the one with a methoxy group. Therefore, two-photon microscopy (2PM) on live HeLa cells was attempted with conjugates featuring complexes [Eu·L-Ar-NHAc], [Eu·L-Ar-Ar-NHAc] and [SucNH-Ar-Ar-L·Eu]. We were interested to know if cytosolic delivery of the Eu³⁺ probe could be achieved with these conjugates. 2P excitation was performed at 720 nm and emission was collected with an avalanche photodiode (APD) for better sensitivity or an array of photomultipliers (PMT) to achieve spectral detection. Under 720 nm 2P excitation, the main contributors to the emission are the Eu³⁺ probe and the cell autofluorescence arising from 2P-excited NAD(P)H and FAD.^19,35–37 Cell autofluorescence mostly originates from mitochondria and has a perinuclear distribution. It is characterized by a broad emission with a maximum at ca. 480 nm extending up to 620 nm (Fig. 6B). The fine Eu^{3+ 5}D₀ → ⁷F_J emission bands at 595 nm (J = 1) and 615 nm can easily be discriminated from the cell autofluorescence by spectral detection.

Fig. 6 2PM imaging (λ_ex = 720 nm) of living HeLa cells incubated 1 h with mTAT[SucNH-Ar-ArL·Eu] (5 μM) in RPMI medium. (A) Left panel: differential interference contrast (DIC) image; Middle panel: luminescence image recorded with 420–680 nm bp APD detection (arbitrarily coloured in yellow); Right panel: merge. Scale bars correspond to 10 μm. (B) Left panel: 2P-excited emission spectra (detected with a PMT array and averaged over the whole cell surface) of the cells outlined in red, blue and green in panel A; Middle and right panels: autofluorescence and Eu³⁺ emission maps obtained by linear unmixing of a 2P-excited spectral image recorded with the PMT.

mTAT probes. HeLa cells were incubated 1 h with the mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[SucNH-Ar-Ar-L·Eu] at 5 μM. Both compounds behaved very similarly. Fig. 6 shows the 2PM images obtained for the latter. As shown in Fig. 6A, the majority of cells show dominant perinuclear emission but some cells also show a faint diffuse emission in the entire cell, including the nucleus. Spectral detection revealed that both autofluorescence and Eu³⁺ signals are present. Linear unmixing of both spectral components shows that the perinuclear emission is autofluorescence while the diffuse emission corresponds to Eu³⁺, which is overall very weak (Fig. 6B).

Dimeric TAT (dTAT) probes. Inspired by the work of Pellois,³⁸ we have shown by 2PM that dTAT[Ln·L] conjugates (Ln = Tb and Eu), comprising two Ln³⁺ complexes bound to a TAT dimer, efficiently reach the cytosol of living cells and give 2PM images with intense and diffuse Ln³⁺ emission within the entire cell.¹⁹ Hence, we prepared compounds dTAT[Eu·L-Ar-Ar-NHAc] and dTAT[SucNH-Ar-Ar-L·Eu] in order to evaluate their ability to stain the cytosol of live cells (Fig. 7A and ESI† for synthesis details).

Fig. 7 Chemical structures of (A) luminescent Eu³⁺ probes based on a dimeric TAT peptide and (B) non luminescent dimeric TAT derivative dFFLIPTAT.

First, we examined their cytotoxicity on HeLa cells by the MTT proliferation assay in order to evaluate the concentration that could be used for 2PM experiments. While the previously described dTAT[Eu·L-Ar-NHAc] conjugate was not cytotoxic up to 20 μM (IC50 > 50 μM), both dTAT[Eu·L-Ar-Ar-NHAc] and dTAT[SucNH-Ar-Ar-L·Eu] were found to be more toxic (Fig. 8A), with an IC50 of 6 ± 1 and 12 ± 2 μM, respectively. This was not unexpected since we have observed that a higher antenna hydrophobicity in these compounds causes higher cytotoxicity.^19,20 From the MTT assays, we found that the maximal nontoxic concentration that could be used for dTAT[Eu·L-Ar-Ar-NHAc] and dTAT[SucNH-Ar-Ar-L·Eu] is 2.5 μM. 2PM was performed on HeLa cells incubated 1 h with the probes at 2 μM. With dTAT[SucNH-Ar-Ar-L·Eu], an intense punctate emission was observed within the cell (Fig. 8B and Fig. S13 of ESI†), which was assigned to Eu³⁺ emission by spectral detection (Fig. 8C). dTAT-based probes were shown to accumulate in endosomes when incubated at a concentration ≤2.5 μM.^19,38 Although co-localization with an endo/lysosome tracker was not attempted here, it is likely to be the case here again. A similar staining was obtained with dTAT[Eu·L-Ar-Ar-NHAc]. Therefore, the dTAT strategy is not pertinent for ter-aryl antennae in order to achieve cytosolic delivery.

Fig. 8 (A) MTT proliferation assays performed on HeLa cells with dTAT[Eu·L-Ar-NHAc] (green),¹⁹ dTAT[Eu·L-Ar-Ar-NHAc] (blue) and dTAT[SucNH-Ar-Ar-L·Eu] (red). (B) 2PM imaging (λ_ex = 720 nm) of HeLa cells incubated 1 h with dTAT[SucNH-Ar-Ar-L·Eu] (2 μM) in RPMI medium, showing (top) the luminescence image recorded with 420–680 nm bp APD detection and (bottom) the superimposition of DIC and luminescence images. (C) Typical emission spectrum emanating from puncta. Scale bars correspond to 10 μm.

mTAT probes co-incubated with dFFLIPTAT. More recently, we have shown that the internalization of a related probe, mTAT[Eu·L-Ar-Cz],²⁰ with a N-carbazolyl–phenyl–picolinamide antenna, could be boosted by co-incubation with the dFFLIPTAT peptide, a non-luminescent TAT dimer introduced by Pellois (Fig. 7B).³⁹ This peptide was reported as an endosomolytic agent, that makes co-incubated species reach the cytosol. This strategy was tested with mTAT[Eu·L-Ar-NHAc], mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[SucNH-Ar-Ar-L·Eu]. Prior to 2PM, a MTT proliferation assay was performed in order to assess the toxicity of mixtures of dFFLIPTAT (1.5 μM) and the conjugates (5 μM). Cells incubated for 1 h with dFFLIPTAT alone at 1.5 μM showed good viability (ca. 90%).^20,39 In co-incubation experiments, the cell viability was similar (Fig. S14 of ESI†), indicating that the mTAT conjugates do not increase toxicity. Therefore, incubation of HeLa cells was performed with a mixture of dFFLIPTAT (1.5 μM) and mTAT[Eu·L-Ar-NHAc], mTAT[Eu·L-Ar-Ar-NHAc] or mTAT[SucNH-Ar-Ar-L·Eu] (5 μM) for 1 h and the 2P imaging properties of the three probes were compared.

Fig. 9A shows typical images obtained with mTAT[SucNH-Ar-Ar-L·Eu] in co-incubation with dFFLIPTAT. Contrarily to images obtained with the mTAT[SucNH-Ar-Ar-L·Eu] probe alone (Fig. 6), those obtained with dFFLIPTAT, show a majority of cells with dominant diffuse emission. In many cells, the perinuclear distribution characteristic of autofluorescence is hardly identified, hidden by the intense diffuse emission. Spectral detection (Fig. 9B) confirmed that the latter corresponds to Eu³⁺. Comparison of Fig. 6 and 9 clearly indicates that dFFLIPTAT boosts the internalization of the Eu³⁺ probe.

Fig. 9 2PM imaging (λ_ex = 720 nm) of living HeLa cells incubated 1 h with mTAT[SucNH-Ar-ArL·Eu] (5 μM) and dFFLIPTAT (1.5 μM) in RPMI medium. (A) Left panel: DIC image; Middle panel: luminescence image recorded with 420–680 nm bp APD detection (arbitrarily coloured in yellow); Right panel: merge. Scale bars correspond to 10 μm. (B) Left panel: 2P-excited emission spectra (detected with a PMT array and averaged over the whole cell surface) of the cells outlined in red, blue and green in panel A; Middle and right panels: autofluorescence and Eu³⁺ emission maps (coloured in cyan and red, respectively) obtained by linear unmixing of a 2P-excited spectral image recorded with the PMT.

The three mTAT[Eu·L-Ar-NHAc], mTAT[Eu·L-Ar-Ar-NHAc] and mTAT[SucNH-Ar-Ar-L·Eu] probes in co-incubation with dFFLIPTAT gave similar results but the Eu³⁺ emission appeared lower by visual inspection in the case of mTAT[Eu·L-Ar-NHAc], with the perinuclear autofluorescence more easily discernible. Fig. 10A shows typical 2PM images obtained with each probe using APD detection and band-pass (bp) filtering and recorded in the same conditions (laser power, pixel dwell time …), the same day. Two detection channels were used: (i) 470–540 nm range (blue channel) that is specific for autofluorescence and (ii) 580–690 nm (red channel) that comprises the two Eu^{3+ 5}D₀ → ⁷F_J (J = 1 and 2) emission bands but also the red tail of the autofluorescence. A majority of cells show a diffuse emission within the cell in the red channel, with autofluorescence distribution hardly detectable, indicating that Eu³⁺ emission level is above the one of the red-edge of the autofluorescence emission. Fig. 10D quantifies the percentage of cells with diffuse staining as determined by visual inspection. For all conjugates it is above 60% and it increases in the order mTAT[Eu·L-Ar-NHAc] < mTAT[Eu·L-Ar-Ar-NHAc] < mTAT[SucNH-Ar-Ar-L·Eu]. With the latter ca. 90% of the cells show an unambiguous diffuse Eu³⁺ staining. The intensity of Eu³⁺ emission vary strongly from cell-to-cell. This is typical of CPP-based luminescent probes.³⁸ In order to compare the three compounds, a quantitative assessment of Eu³⁺ staining was performed. For this purpose, the mean intensity (red channel) per pixel was measured using ImageJ in ca. 80 cells for cells treated with each compound and for cells not incubated with any of the compounds, as a control. It has been reported that CPP-based luminescent probes can adhere to the coverglass of the cell culture chamber providing a background signal outside cells, especially when 2P imaging is done close to the surface.^19,40 Therefore, the background mean intensity per pixel was determined in several (>50) cell-free areas of the approximate size of a cell. Fig. 10B shows the mean intensity per pixel measured in cells and in background areas (Bg) for the three conjugates and in control experiments (Ctrl). First, for the three conjugates, this quantification shows a great dispersion in the staining level among cells, with the mean intensity spanning an order of magnitude. This confirms the visual inspection. For all three conjugates, the staining is significantly higher than the background (P value <0.0001), the staining of the two conjugates with a ter-aryl antenna is similar (P = 0.24) but significantly higher than that of mTAT[Eu·L-Ar-NHAc] with the bi-aryl antenna (P < 0.0001). Note also that the background is also higher in the case of the two ter-aryl antenna-based compounds (P < 0.0001). This suggests an identical behaviour for all conjugates but with an improved emission in the case of the ter-aryl compounds in agreement with their higher brightness. However, a higher cell penetration efficiency cannot be ruled out to explain this higher staining. Finally, the percentage of cells unambiguously stained with the Eu³⁺ probe (Fig. 10C) was determined by counting the cells with a mean intensity per pixel above the limit of detection (LoD) defined by eqn (5),

LoD = <I_Bg> + I_AF + 3 × SD(I_Bg) (5)

where <I_Bg> and SD(I_Bg) are the mean value and the standard deviation of the mean intensity per pixel in background areas for experiments performed with the conjugate. I_AF is the difference between the mean values of intensity in cells and background areas in the control experiment, corresponding to the intensity of the red tail of the autofluorescence (I_AF was measured to be 1.55). For the three conjugates, this percentage increases in the order mTAT[Eu·L-Ar-NHAc] < mTAT[Eu·L-Ar-Ar-NHAc] < mTAT[SucNH-Ar-Ar-L·Eu] and for each conjugate, it matches well the percentage of cells with diffuse emission as determined by visual inspection (Fig. 10D). This indicates that the improvement in the Eu³⁺ emission intensity in the case of the ter-ayl antennae-based conjugates is strong enough to be estimated qualitatively by visual inspection. Overall, these 2P microscopy experiments demonstrate the benefits of switching from a bi-aryl to a ter-aryl antenna for the Eu³⁺ complexes.

Fig. 10 (A) 2PM imaging (λ_ex = 720 nm) of living HeLa cells incubated 1 h with dFFLIPTAT (1.5 μM) and (a) mTAT[Eu·L-Ar-NHAc], (b) mTAT[Eu·L-Ar-Ar-NHAc] or (c) mTAT[SucNH-Ar-Ar-L·Eu] (5 μM) in RPMI medium. From left to right: DIC image, autofluorescence emission (APD, 470–540 nm bp), Eu³⁺ emission (APD, 580–690 nm bp), and merged emission images. (B) Mean intensity per pixel in cells and background (Bg) for the three conjugates and for the control experiment (Ctrl); the black line shows the median value. (C) Percentage of cells unambiguously labelled with Eu³⁺, determined from the limit of detection. (D) Percentage of cells with diffuse emission, as determined by visual inspection. It was determined by three of the co-authors, the mean is plotted and the range is indicated by error bars.

The quantitative data presented here allow comparison of two cell internalization strategies relying on (1) TAT dimers, with, on the one hand, dTAT[Eu·L-Ar-NHAc], and (2) TAT monomer + dFFLIPTAT, with, on the other hand, mTAT[Eu·L-Ar-NHAc], both conjugates having the same antenna. The former strategy, with the probe at 5 μM, yielded ca. 35% of cells with unambiguous diffuse cytosolic staining of Eu³⁺ (ref. 19) while the latter strategy gave ca. 60%, at the same probe concentration, i.e. with half the amount of Eu³⁺ complex. This means that superior cytosolic delivery is achieved by co-incubation of the mTAT probe with dFFLIPTAT compared to the dTAT probe. Combining the dFFLIPTAT strategy with a better ter-aryl antenna, enhances even more the number of cells with detectable diffuse staining.

Conclusions

Eu³⁺ based luminescent bioprobes with p-amido- or p-methoxy–aryl–picolinate antennae (i.e. bi-aryl antennae A1 and A2) that we have described previously suffer from a weak absorption above 350 nm. In this article, we have evaluated the benefit of inserting an additional aryl ring within these antennae to red shift the absorption and achieve superior 1P and 2P photophysical properties. Three compounds with ter-aryl antennae, mTAT[Eu·L-Ar-Ar-NHAc], mTAT[Eu·L-Ar-Ar-OMe] and mTAT[SucNH-Ar-Ar-L·Eu], were synthesized and studied, differing in the nature of the electron donating group and/or in the peptide grafting point. From our photophysical study, the best compromise between the red shift of the excitation wavelength and the 2P brightness is compound mTAT[SucNH-Ar-Ar-L·Eu]. The [SucNH-Ar-Ar-L·Eu] complex shows photophysical properties similar to those that can be obtained using the related p-alkoxy–aryl–ethynyl–picolinate (antenna A3), which cannot be incorporated into peptide conjugates due to side reactions arising with the alkyne group during peptide synthesis (especially, acidic deprotection treatments). Comparatively, the ter-aryl antenna is fully stable in peptide synthesis conditions. The increased 2P brightness at 720 nm gained by passing from a bi-aryl antenna to a ter-aryl antenna is likely to be responsible for the better staining and detection achieved with mTAT[SucNH-Ar-Ar-L·Eu] in 2PM experiments. This article demonstrates that the ter-aryl p-amido–aryl–aryl–picolinate antenna is an interesting alternative to p-alkoxy–aryl–ethynyl–picolinate for the 2P sensitization of Eu³⁺ luminescence.

Author contributions

Conceptualization: O. S., O. M.; investigation: B. C., A. G., L. B., A. N., G. M., S. E., V. M.-F., O. S.; validation: O. S., A. G., A. B., O. M., V. M.-F., D. B., J. K. M.; writing (original draft): O. S.; writing (review & editing): all authors. Visualization: O. S.

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

Authors acknowledge the Agence Nationale de la Recherche (ANR-18-CE06-0022 and ANR-21-CE29-0018), the Labex ARCANE, CBH-EUR-GS (ANR-17-EURE-0003) and the CEA FOCUS Biomarqueurs program for financial support. Imaging experiments were done on Microcell core facility of the Institute for Advanced Biosciences (UGA – Inserm U1209 – CNRS 5309). This facility belongs to the IBISA-ISdV platform, member of the national infrastructure France-BioImaging supported by the French National Research Agency (ANR-10-INBS-04).

References

1. J.-C. G. Bünzli and S. V. Eliseeva, in Lanthanide Luminescence, ed. P. Hänninen and H. Härmä, Springer Berlin Heidelberg, 2011, pp. 1–45.
2. J.-C. G. Bünzli, Lanthanide light for biology and medical diagnosis, J. Lumin., 2016, 170, 866–878.
3. M. Sy, A. Nonat, N. Hildebrandt and L. J. Charbonnière, Lanthanide-based luminescence biolabelling, Chem. Commun., 2016, 52, 5080–5095.
4. E. Mathieu, A. Sipos, E. Demeyere, D. Phipps, D. Sakaveli and K. E. Borbas, Lanthanide-based tools for the investigation of cellular environments, Chem. Commun., 2018, 54, 10021–10035.
5. G.-Q. Jin, Y. Ning, J.-X. Geng, Z.-F. Jiang, Y. Wang and J.-L. Zhang, Joining the journey to near infrared (NIR) imaging: the emerging role of lanthanides in the designing of molecular probes, Inorg. Chem. Front., 2020, 7, 289–299.
6. I. Martinić, S. V. Eliseeva and S. Petoud, Near-infrared emitting probes for biological imaging: Organic fluorophores, quantum dots, fluorescent proteins, lanthanide(III) complexes and nanomaterials, J. Lumin., 2017, 189, 19–43.
7. S. I. Weissman, Intramolecular Energy Transfer: The Fluorescence of Complexes of Europium, J. Chem. Phys., 1942, 10, 214–217.
8. J.-C. G. Bünzli, On the design of highly luminescent lanthanide complexes, Coord. Chem. Rev., 2015, 293, 19–47.
9. A. D'Aléo, F. Pointillart, L. Ouahab, C. Andraud and O. Maury, Charge transfer excited states sensitization of lanthanide emitting from the visible to the near-infra-red, Coord. Chem. Rev., 2012, 256, 1604–1620.
10. W. Thor, H.-Y. Kai, Y.-H. Yeung, Y. Wu, T.-L. Cheung, L. K. B. Tam, Y. Zhang, L. J. Charbonnière, P. A. Tanner and K.-L. Wong, Unearthing the Real-Time Excited State Dynamics from Antenna to Rare Earth Ions Using Ultrafast Transient Absorption, JACS Au, 2024, 4, 3813–3822.
11. G. Piszczek, B. P. Maliwal, I. Gryczynski, J. Dattelbaum and J. R. Lakowicz, Multiphoton ligand-enhanced excitation of lanthanides, J. Fluoresc., 2001, 11, 101–107.
12. M. H. V. Werts, N. Nerambourg, D. Pélégry, Y. L. Grand and M. Blanchard-Desce, Action cross sections of two-photon excited luminescence of some Eu(III) and Tb(III) complexes, Photochem. Photobiol. Sci., 2005, 4, 531–538.
13. A. Picot, A. D'Aleo, P. L. Baldeck, A. Grichine, A. Duperray, C. Andraud and O. Maury, Long-lived two-photon excited luminescence of water-soluble europium complex: Applications in biological imaging using two-photon scanning microscopy, J. Am. Chem. Soc., 2008, 130, 1532–1533.
14. G.-L. Law, K.-L. Wong, C. W.-Y. Man, S.-W. Tsao and W.-T. Wong, A two-photon europium complex as specific endoplasmic reticulum probe, J. Biophotonics, 2009, 2, 718–724.
15. A. D'Aléo, A. Bourdolle, S. Brustlein, T. Fauquier, A. Grichine, A. Duperray, P. L. Baldeck, C. Andraud, S. Brasselet and O. Maury, Ytterbium-Based Bioprobes for Near-Infrared Two-Photon Scanning Laser Microscopy Imaging, Angew. Chem., Int. Ed., 2012, 51, 6622–6625.
16. A. T. Bui, A. Roux, A. Grichine, A. Duperray, C. Andraud and O. Maury, Twisted Charge-Transfer Antennae for Ultra-Bright Terbium(III) and Dysprosium(III) Bioprobes, Chem. – Eur. J., 2018, 24, 3408–3412.
17. A. T. Bui, A. Grichine, S. Brasselet, A. Duperray, C. Andraud and O. Maury, Unexpected Efficiency of a Luminescent Samarium(III) Complex for Combined Visible and Near-Infrared Biphotonic Microscopy, Chem. – Eur. J., 2015, 21, 17757–17761.
18. J.-H. Choi, G. Fremy, T. Charnay, N. Fayad, J. Pécaut, S. Erbek, N. Hildebrandt, V. Martel-Frachet, A. Grichine and O. Sénèque, Luminescent Peptide/Lanthanide(III) Complex Conjugates with Push–Pull Antennas: Application to One- and Two-Photon Microscopy Imaging, Inorg. Chem., 2022, 61, 20674–20689.
19. K. P. Malikidogo, T. Charnay, D. Ndiaye, J.-H. Choi, L. Bridou, B. Chartier, S. Erbek, G. Micouin, A. Banyasz, O. Maury, V. Martel-Frachet, A. Grichine and O. Sénèque, Efficient cytosolic delivery of luminescent lanthanide bioprobes in live cells for two-photon microscopy, Chem. Sci., 2024, 15, 9694–9702.
20. J.-H. Choi, A. Nhari, T. Charnay, B. Chartier, L. Bridou, G. Micouin, O. Maury, A. Banyasz, S. Erbek, A. Grichine, V. Martel-Frachet, F. Thomas, J. K. Molloy and O. Sénèque, Carbazole-Based Eu³⁺ Complexes for Two-Photon Microscopy Imaging of Live Cells, Inorg. Chem., 2025, 64, 2006–2019.
21. A. T. Bui, M. Beyler, Y.-Y. Liao, A. Grichine, A. Duperray, J.-C. Mulatier, B. L. Guennic, C. Andraud, O. Maury and R. Tripier, Cationic Two-Photon Lanthanide Bioprobes Able to Accumulate in Live Cells, Inorg. Chem., 2016, 55, 7020–7025.
22. S. Mizzoni, S. Ruggieri, A. Sickinger, F. Riobé, L. Guy, M. Roux, G. Micouin, A. Banyasz, O. Maury, B. Baguenard, A. Bensalah-Ledoux, S. Guy, A. Grichine, X.-N. Nguyen, A. Cimarelli, M. Sanadar, A. Melchior and F. Piccinelli, Circularly polarized activity from two photon excitable europium and samarium chiral bioprobes, J. Mater. Chem. C, 2023, 11, 4188–4202.
23. D. Akl, L. Bridou, M. Hojorat, G. Micouin, S. R. Kiraev, F. Riobé, S. Denis-Quanquin, A. Banyasz and O. Maury, Comprehensive Photophysical and Nonlinear Spectroscopic Study of Thioanisolyl-Picolinate Triazacyclononane Lanthanide Complexes, Eur. J. Inorg. Chem., 2024, 27, e202300785.
24. K. Binnemans, Interpretation of europium(III) spectra, Coord. Chem. Rev., 2015, 295, 1–45.
25. A. Beeby, I. M. Clarkson, R. S. Dickins, S. Faulkner, D. Parker, L. Royle, A. S. de Sousa, J. A. G. Williams and M. Woods, Non-radiative deactivation of the excited states of europium, terbium and ytterbium complexes by proximate energy-matched OH, NH and CH oscillators: an improved luminescence method for establishing solution hydration states, J. Chem. Soc., Perkin Trans. 2, 1999, 493–504.
26. M. H. V. Werts, R. T. F. Jukes and J. W. Verhoeven, The emission spectrum and the radiative lifetime of Eu³⁺ in luminescent lanthanide complexes, Phys. Chem. Chem. Phys., 2002, 4, 1542–1548.
27. N. Kofod, L. G. Nielsen and T. J. Sørensen, Temperature Dependence of Fundamental Photophysical Properties of [Eu(MeOH-d4)₉]³⁺ Solvates and [Eu·DOTA(MeOH-d4)]⁻ Complexes, J. Phys. Chem. A, 2021, 125, 8347–8357.
28. D. Kovacs, E. Mathieu, S. R. Kiraev, J. A. L. Wells, E. Demeyere, A. Sipos and K. E. Borbas, Coordination Environment-Controlled Photoinduced Electron Transfer Quenching in Luminescent Europium Complexes, J. Am. Chem. Soc., 2020, 142, 13190–13200.
29. A. Weller, Electron-transfer and complex formation in the excited state, Pure Appl. Chem., 1968, 16, 115–124.
30. W. D. Horrocks, J. P. Bolender, W. D. Smith and R. M. Supkowski, Photosensitized near infrared luminescence of ytterbium(III) in proteins and complexes occurs via an internal redox process, J. Am. Chem. Soc., 1997, 119, 5972–5973.
31. A. Beeby, S. Faulkner and J. A. G. Williams, pH Dependence of the energy transfer mechanism in a phenanthridine-appended ytterbium complex, J. Chem. Soc., Dalton Trans., 2002, 1918–1922.
32. D. Kocsi, D. Kovacs, J. A. L. Wells and K. E. Borbas, Reduced quenching effect of pyridine ligands in highly luminescent Ln(III) complexes: the role of tertiary amide linkers, Dalton Trans., 2021, 50, 16670–16677.
33. D. Parker, J. D. Fradgley, M. Delbianco, M. Starck, J. W. Walton and J. M. Zwier, Comparative analysis of lanthanide excited state quenching by electronic energy and electron transfer processes, Faraday Discuss., 2022, 234, 159–174.
34. P. Luo, E. C. Feinberg, G. Guirado, S. Farid and J. P. Dinnocenzo, Accurate Oxidation Potentials of 40 Benzene and Biphenyl Derivatives with Heteroatom Substituents, J. Org. Chem., 2014, 79, 9297–9304.
35. Y. Qin and Y. Xia, Simultaneous Two-Photon Fluorescence Microscopy of NADH and FAD Using Pixel-to-Pixel Wavelength-Switching, Front. Phys., 2021, 9, 642302.
36. R. Cao, H. K. Wallrabe and A. Periasamy, Multiphoton FLIM imaging of NAD(P)H and FAD with one excitation wavelength, J. Biomed. Opt., 2020, 25, 014510.
37. I. A. Gorbunova, M. K. Danilova, M. E. Sasin, V. P. Belik, D. P. Golyshev and O. S. Vasyutinskii, Determination of fluorescence quantum yields and decay times of NADH and FAD in water–alcohol mixtures: The analysis of radiative and nonradiative relaxation pathways, J. Photochem. Photobiol., A, 2023, 436, 114388.
38. A. Erazo-Oliveras, K. Najjar, L. Dayani, T.-Y. Wang, G. A. Johnson and J.-P. Pellois, Protein delivery into live cells by incubation with an endosomolytic agent, Nat. Methods, 2014, 11, 861–867.
39. J. Allen and J.-P. Pellois, Hydrophobicity is a key determinant in the activity of arginine-rich cell penetrating peptides, Sci. Rep., 2022, 12, 15981.
40. M. Serulla, P. Anees, A. Hallaj, E. Trofimenko, T. Kalia, Y. Krishnan and C. Widmann, Plasma membrane depolarization reveals endosomal escape incapacity of cell-penetrating peptides, Eur. J. Pharm. Biopharm., 2023, 184, 116–124.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5qi00333d

---