New Paper from Dr. McFarlandDr. McFarland's latest paper explores the limitations of NIR light activated Ruthenium photosensitizers.
Ruthenium Photosensitizers for NIR PDT Require Lowest-Lying Triplet Intraligand (3IL) Excited States ABSTRACT
A family of complexes of the type [Ru(tpbn)(IP-R)(4-pic)]Cl2 (tbpn=2,2′-(4-(tert-butyl)pyridine-2,6-diyl)bis(1,8-napthyridine); 4-pic=4-picoline; IP-R=imidazo[4,5-f][1,10]phenanthroline attached to an aromatic group R for 2–8 and H for 1) were prepared as near-infrared (NIR) absorbing coordination complexes to test whether triplet intraligand excited states (3IL) of higher energy than the lowest-lying triplet metal-to-ligand charge transfer excited states (3MLCT) could effectively generate cytotoxic singlet oxygen (1O2) and elicit in vitro photodynamic therapy (PDT) effects. Aromatic groups ranged from benzene to anthracene, with corresponding triplet state energies that were all significantly higher (approximately 3.7–1.8 eV) than the 3MLCT state estimated at 1.5 eV. Complexes 1–8 absorbed NIR light, with their longest-wavelength peak maxima occurring near 725 nm that extended out to 800 nm. The 1O2 quantum yields for the aromatic-containing compounds were extremely small (ΦΔ=0.07), with correspondingly modest in vitro photocytotoxicities. All compounds were nontoxic without a light trigger, with dark EC50 values >60 µM and most values closer to 100 or greater. EC50 values with visible light were 5–6 (PI=15–20), 7–10 (PI=8–11), and 10–15 µM (PI=6–8) in SKMEL28, A375, and B16F10 cancer cell lines, respectively. With NIR light, these values were even less: 11–16 (PI=5–9), 16–50 (PI=2–6), and 15–19 µM (PI=4–6) in SKMEL28, A375, and B16F10 cancer cell lines, respectively. While measurable, the modest activities and absence of any trend between the 3IL energies and values for ΦΔ or PI demonstrate that 3IL states with energies above the lowest-lying 3MLCT states do not contribute to the overall excited state dynamics responsible for potent PDT effects in previous studies. Lowest-lying 3MLCT states in this family of NIR-absorbing photosensitizers do not produce the requisite 1O2 for effective in vitro photocytotoxic effects, underscoring the need to install 3IL states that are lower in energy than the lowest-lying 3MLCT states in order the create potent NIR-activatable Ru(II) complexes for PDT.
INTRODUCTION
Photodynamic therapy (PDT) is a promising technique that offers some advantages over conventional cancer therapy (i.e., surgery, chemotherapy, and radiotherapy) due to its capacity for spatial and temporal control through the delivery of light.(1), (2), (3), (4), (5) The reactive oxygen species (ROS) generated in PDT result from the interaction of a photosensitizer (PS) with light and oxygen, and these ROS are responsible for the specific cytotoxicity effects on cancerous tumor cells. In the PDT process, the PS is photoexcited from the ground state (S0) to an excited singlet state (S1), which then undergoes intersystem crossing (ISC) to a longer-lived excited triplet state (T1) that can interact with oxygen to generate ROS in one of two mechanisms.6 The Type I mechanism is an electron transfer reaction that ultimately produces ROS such as hydrogen peroxide (H2O2), superoxide (O2−) or hydroxyl radicals (•OH). The Type II mechanism is an energy transfer process between 3PS and ground state triplet oxygen (3O2) to produce singlet oxygen (1O2), and this mechanism is considered the major contributor to the PDT effect. The antitumor action of PDT arises from (i) direct damage to the tumor, (ii) disruption of tumor vasculature, and (iii) initiation of an immune system response.(7), (8), (9), (10), (11), (12)
Photofrin was the first PS approved for PDT13 and remains the clinical standard in PDT. It is a mixture of oligomeric tetrapyrroles that is activated with 630-nm red light. The success of Photofrin has inspired a variety of organic PSs based on the tetrapyrrolic structure14 (e.g., porphyrins, chlorins, bacteriochlorins, phthalocyanins, etc.) as single-compound PSs, but none has surpassed Photofrin in terms of clinical utility. More recently metal-containing tetrapyrroles have also received attention.15
One goal of next-generation PSs is to extend the light absorption window to tissue-penetrating near infrared (NIR) wavelengths12,(16), (17), (18) (∼700900 nm) and to impart activity in hypoxic conditions.(19), (20), (21), (22) Ruthenium (Ru) polypyridyl complexes constitute an attractive scaffold in in this regard since their photophysics are well established(23), (24), (25), (26), (27) and easily tunable through the design of special coordinated ligands and their combinations around the Ru(II) center. It is within this context that our group has developed numerous PSs,12,(18), (19), (20),(28), (29), (30), (31), (32), (33), (34), (35), (36), (37), (38), (39) including TLD1433,40 that exploit low-lying intraligand (3IL) and 3ILCT (intraligand charge transfer) excited states with long lifetimes and high 1O2 quantum yields. The strategy is proving effective: TLD1433 is the first Ru(II) PS to progress to phase II human clinical trials (NCT03945162) for treating cancer with PDT. Clinically, TLD1433 is activated with green light (532 nm) and shows some photocytotoxicity with red light (630 nm), but none at wavelengths near 700 nm without special formulations.41 While some NIR-absorbing Ru scaffolds are known,17,42,43 their photocytotoxicity has not been widely studied. The present work stems from the idea that effective NIR PDT in Ru(II) complexes can be achieved by combining π-expansive ligands that produce the desirable low-lying 3IL or 3ILCT states with ligands that promote NIR absorption via low-lying 3MLCT states.12,18 The key is to design the complexes such that the longer-lived 3IL or 3ILCT state can be efficiently populated by positioning its energy lower than the shorter-lived 3MLCT state, thereby facilitating cytotoxic pathways such as the production of 1O2.
Herein, a series of Ru(II) complexes of the form [Ru(NNN)(NN)(N)]Cl2 (Chart 1) were prepared, whereby the chromophoric tridentate NNN and monodentate N ligands were held constant. The combination of the tbpn and 4-pic (tbpn=2,2′-(4-(tert-butyl)pyridine-2,6-diyl)bis(1,8-napthyridine); 4-pic=4-picoline) ligands was selected based on our previous studies and serves to shift the energy of the 1MLCT absorption into the NIR, while maintaining a relatively stable coordination complex. The bidentate ligand NN was based on a common imidazo[4,5-f][1,10]phenanthroline (IP) ligand attached to variable aromatic groups R (IP-R).29,30,44 The R groups were chosen as π-expansive triplet reservoirs with known energies to test whether their 3IL states were of sufficiently low energy to initiate photocytotoxic effects toward cancer cells. The complexes were purposely designed to have 3IL energies higher than the lowest-energy 3MLCT states. Our hypothesis was that the 3IL energy should be lower than the 3MLCT energy to contribute to the excited state dynamics and elicit photocytotoxic effects. The present family is designed to disprove this hypothesis if any of the members are significantly photocytotoxic.
.
.
.
.
CONCLUSIONS
The goal of this study was to explore a series of [Ru(tpbn)(NN)(4-pic)]Cl2 complexes derived from NN=IP-R, using the pendant R group to systematically vary the energy of the 3IL state owing to the well-established triplet energies of aromatic chromophores. In addition to R=H as a reference lacking an aromatic R group, the groups ranged from benzene as the least π-expansive unit to pyrene and anthracene as the most π-extended, providing 3IL energies that ranged from about 1.8 to 3.7 eV.45,58 These energies were higher than the 3MLCT state estimated at 1.5 eV in previous studies. As such, the series afforded the opportunity to test whether 3IL states that are higher in energy than the lowest-lying 3MLCT state in this NIR-absorbing motif lead to effective 1O2 generation and in vitro PDT effects. The quantum yields for 1O2 were very low at 6–13% and did not correlate with the 3IL energies as we have observed previously for families of Ru(II) and Os(II) complexes that include members with lowest-lying 3IL states. The in vitro PDT effects were also poor in all three cell lines tested with all light conditions. Even for the most potent visible light condition using the most sensitive cell line, the PI values for compounds with aromatic R groups were only in the 15–20 range. Otherwise, the PI values with NIR light were in the range of 5–9 (SKMEL28), 2–6 (A375), and 4–6 (B16F10).
This study demonstrates that NIR-absorbing complexes of the type [Ru(tpbn)(NN)(4-pic)]Cl2 based on the IP-R ligand as NN, whereby R is an aromatic chromophore with triplet energy above the lowest-lying 3MLCT, have low 1O2 quantum yields and poor in vitro PDT effects. We have previously established that lowest-lying 3IL and mixed 3IL/3ILCT states (with Ru(II) and Os(II) compounds that absorb visible light) are superior to 3MLCT states for 1O2 generation and in vitro PDT effects. In such systems, it is relatively straightforward to choose π-expansive chromophores that contribute lowest-lying 3IL or 3IL/3ILCT states because the Ru(II) and Os(II) polypyridyl complexes have their 3MLCT energies near 2.1 and 1.7 eV, respectively. It becomes more challenging to maintain the ligand-localized excited states as the lowest-energy states when designing analogous NIR-absorbing complexes because the NIR absorption is achieved by shifting the energy of the 1MLCT state into the NIR, which necessarily shifts the 3MLCT state to even lower energies. This is problematic for two reasons. First, lowest-lying 3MLCT states dominate the excited state dynamics and nonradiative decay back to the ground state through thermal relaxation competes effectively with 1O2 sensitization due to the energy gap law and high SOC associated with the metal character of these states. Second, the degree of π-expansion needed to establish even lower-lying 3IL states creates the need for custom design and complicated syntheses and the resulting extremely hydrophobic compounds are prone to agglomeration, which is exacerbated in aqueous solutions. One of the few exceptions involved the use of the dppn ligand. [Ru(tpbn)(dppn)(4-pic)]Cl2 has an estimated 3IL energy of 1.3 eV and thus excited state dynamics controlled by the 3IL state, whereby 1O2 sensitization was effective and in vitro PDT effects very potent with visible and NIR light.
The present series demonstrates that aromatic groups with higher triplet energies compromise the photocytotoxic activity and suggests that the 3IL energy should be lower than the 3MLCT energy to contribute to the excited state dynamics and elicit potent photocytotoxic effects. Importantly, this study in combination with our prior studies, underscores a fundamental limit for shifting the absorption into the NIR using 1MLCT states. We previously proposed that there is a minimum 3MLCT–3IL energy gap required for effective 3IL population that would dictate how far into the NIR the absorption could be shifted while maintaining high 1O2 quantum yields and good photocytotoxicity. With the tbpn/4-pic Ru(II) scaffold, we suggest that the longest-wavelength NIR absorption maximum should not be longer than about 715–720 nm and the 3IL energy should be well below 1.5 eV.