Transposing molecular fluorescent switches into the near-IR: development of luminogenic reporter substrates for redox metabolism.

Optical reporters enable the direct and minimally invasive probing of molecular events in complex environments such as living cells and tissues. While significant achievements have been made in the area of fluorescent sensors for metal ions and small organic compounds,1 probes that report directly on enzyme activity are less developed by comparison, with the exception of reporters for hydrolases and kinases.2 Lanthanide complexes are attractive for the design of responsive probes owing to their sharp emission bands at long wavelengths and long luminescent lifetimes. While studies in this area have been dominated by the visible light emitting lanthanides Tb3+ and Eu3+,3 the development of reporters based on the near-infrared (NIR) emitting metals is still in its infancy. The ions Nd3+, Yb3+, Er3+, Ho3+, and Pr3+ have emission transitions spanning the region from 800 to 1600 nm and comprise the NIR lanthanide series.4 NIR imaging agents have particular appeal for in vivo applications since the spectral region in which they operate allows greater tissue penetration.5,6 Herein, we communicate the development of a NIR luminescent redox switch, based on a Nd(III) complex, that serves as a reporter substrate for the human aldo-ketoreductase 1C2 (AKR1C2). As part of a program aimed at the development of fluorescent and luminescent enzyme activity reporters,7 we have previously described a class of fluorogenic substrates for AKR1C2,7a,e an enzyme involved in steroid hormone metabolism and stress response pathways.8 These reporters, based on an aminocoumarin fluorophore, emit a strong green fluorescence (λmax ≈ 510 nm) upon enzymatic reduction of the corresponding nonfluorescent ketone substrate. The importance of AKR1C2 as an emerging cellular stress marker motivated us to explore the possibility of adapting the fluorogenic substrate to a NIR emitting format. Since many organic chromophores are known to sensitize lanthanide luminescence via energy transfer (ET), we envisioned a strategy whereby our coumarin substrates could be directly converted to NIR reporters by tethering them to an appropriate lanthanide complex, thus obviating the need to evolve a new substrate based on an organic NIR fluorescent scaffold (e.g., Cy5.5)6 (Figure 1). Following this approach, we were faced with two key questions. The first one is whether the ketone/alcohol transformation would effect the significant change in ET efficiency necessary to afford a good luminescent switch.9 Since the sensitization of lanthanide metals is assumed to occur via a triplet excited state of the organic chromophore,3,4 it may be expected that groups promoting intersystem crossing (from singlet to triplet state), such as the ketone, would influence the energy transfer efficiency. However, the direction of this effect was not readily predictable on the basis of previous studies.10,11 The second question is how the size and polarity of the metal complex would affect the probe’s ability to function as an enzyme substrate. We supposed that both ET efficiency and enzyme substrate capacity would be dependent on the length of the tether between the coumarin and the lanthanide complex, and we expected these parameters to be mutually antagonistic a short tether would increase ET efficiency, while a longer tether would relieve the steric bulk from the site of enzymatic activity. We also anticipated the relative orientation of the coumarin and the complex to be a factor in enzyme recognition. Guided by these design considerations, we synthesized a series of probe candidates comprised of macrocyclic Nd:DOTA chelates tethered via two-, three-, or six-carbon alkyl linkers to two regioisomeric aminocoumarin scaffolds (Figure 2, 1Nd-6Nd). The corresponding alcohol products were also prepared (7Nd-12Nd). Importantly, alcohols 7Nd-12Nd all displayed characteristic Nd3+ luminescence with the emission maximum at ∼1060 nm (excitation at 395 nm). In contrast, ketones 1Nd-6Nd had negligible emission in this region, making the ketone/alcohol pair a robust luminogenic switch in the reduction direction (Figure 3A). The relative luminescence intensity was calculated by integrating the Figure 1. The concept of adapting a fluorogenic redox reporter (left) to a NIR emitting format by the incorporation of lanthanide ions (right).