Communication pubs.acs.org/JACS
Sense and Release: A Thiol-Responsive Flavonol-Based Photonically Driven Carbon Monoxide-Releasing Molecule That Operates via a Multiple-Input AND Logic Gate Tatiana Soboleva,† Hector J. Esquer,‡ Abby D. Benninghoff,‡ and Lisa M. Berreau*,† †
Department of Chemistry and Biochemistry, Utah State University, Logan, Utah 84322-0300, United States Department of Animal, Dairy and Veterinary Sciences, Utah State University, Logan, Utah 84322-4815, United States
‡
S Supporting Information *
signal correlation patterns which follow a molecular-scale AND logic gate type approach.19−22 Herein we demonstrate that a novel sense-of-logic photonically driven carbon monoxidereleasing molecule (SL-photoCORM) based on an extended flavonol framework enables fluorescence sensing of biothiol levels followed by controlled CO release within a cellular environment. This approach represents the merging of three fields of significant current interest in terms of designing molecular tools for applications in biological systems. Specifically, it is a combination of logically enabled molecular design, analyte sensing, and a controlled gasotransmitter release. Inspired by the work of Liu et al.,23 we hypothesized that functionalizing the extended flavonol 124 via the addition of an acryloyl appendage at the 3-OH position (Scheme 1, reaction 1)
ABSTRACT: Molecular structures capable of intracellular information processing that couple responses from biomarker signals to the release of drug molecules represent intelligent delivery systems. Herein we report a chemophotonically driven, sense-of-logic carbon monoxide-releasing molecule (SL-photoCORM). This extended flavonol motif operates via an AND logic gate by first sensing the cellular environment via detection of thiols and then releasing CO when triggered with visible light and O2. Overall, this approach couples the detection of a cellular redox biomarker with the ability to release a smallmolecule gasotransmitter known to trigger pathways involved in pro- and anti-inflammatory effects in a dosedependent manner. Significantly, the fluorescence properties of the flavonol-based SL-photoCORM produce a series of chemophotonic input responses via two photochromatic switches (blue-to-green and green-to-colorless), leading to trackability and spatiotemporal control of CO release. Examination of the O2 requirements of the CO release step revealed that the SL-photoCORM is suitable for use under conditions of variable cellular levels of O2. These combined properties within a single-molecular framework advance the field of CO-releasing molecules by providing feedback on the diversity and complexity of the cellular environment prior to CO release.
Scheme 1. Preparation and Reactivity of SL-photoCORM 3a
a
SL-photoCORM 3 is prepared via (1) treatment of 1 with acryloyl chloride, (2) thiol sensing reaction of 3 to regenerate 1, and (3) visible light-induced CO release reactivity of 1 to generate 2, which is not a fluorescent molecule.24,26 The quantum yield for release of CO from 1 is 0.010(3) in 1:1 DMSO:PBS at pH = 7.4.26
T
he redox balance within cells is a summary of outputs from multiple pathways involving the generation and consumption of molecular entities, such as reactive oxygen species (ROS), and is an indicator of cellular metabolic activity and disease.1−5 Insight into the redox state of cells can be gained by probing factors such as cellular ROS levels or other molecular markers produced as a response to oxidative stress, such as biothiol levels.5−13 Gasotransmitters, such as carbon monoxide (CO), are known to trigger pathways involved in pro-/antiinflammatory effects in a dose-dependent manner.14−16 Incorporation of a biothiol level assessment prior to the triggered release of CO will enable a high degree of control in studies of the roles of this gasotransmitter in biology. To date, no unimolecular structure has been reported that combines these functionalities and is suitable for use in a biological environment.17,18 We hypothesized that such a structure could be achieved through the development of a structural motif with specific input−output © 2017 American Chemical Society
would enable fluorescent thiol sensing via Michael addition, followed by an intramolecular cyclization reaction (Scheme 1, reaction 2).25 This would result in unmasking of the COreleasing unit in 1 for subsequent visible light-induced reactivity (Scheme 1, reaction 3). In order to test this hypothesis, we synthesized and characterized 3 (Figures S1−S7). The presence of the acryloyl appendage produces a blue shift in the absorption and emission features of 3 relative to those exhibited by 1 (Figure S8). Specifically, compound 3 has a lowest energy absorption Received: April 22, 2017 Published: July 5, 2017 9435
DOI: 10.1021/jacs.7b04077 J. Am. Chem. Soc. 2017, 139, 9435−9438
Communication
Journal of the American Chemical Society band centered at 366 nm, versus 410 nm in 1 (Figure S8, top). Excitation at the absorbance maximum of 3 produces a single emission feature at 490 nm (Figure S8, bottom). The most intense emission for 1 (λex = 410 nm) is significantly red-shifted (λem ≈ 575 nm) and is associated with an excited-state intramolecular proton transfer (ESIPT) tautomer (T*, Figure S8, bottom).26 The distinct emission features of 3 and 1 indicate that a photochromatic switch is possible within this system. Based on the prior results reported by Liu et al.23 indicating the unmasking of the flavonol 3-OH moiety upon reaction with thiols, we anticipated a unidirectional thiol-responsive conversion of 3 to 1, thus building up 1 as a molecular memory for thiol levels in the cellular environment. To evaluate the thiol reactivity of 3, we first performed absorption and emission spectral analyses of the reaction of 3 with Cys, Hcy, and GSH (Figures S9−S12). Time-dependent studies over the course of 60 min (Figure 1) showed that
Importantly, the second set of inputs, specifically photons from visible light and O2, which will convert 1 to 2 with CO release, must not lead to the independent photodegradation of 3 in order to prevent a false positive response in the AND logic gate sequence. To address this issue, we performed a control experiment wherein the photostability of 3 upon illumination with visible light in the presence of O2 was evaluated. As shown in Figure 2, while the unmasked flavonol 1 undergoes complete
Figure 2. Changes in absorption of 3 and 1 upon illumination with visible light (λill = 419 nm, intensity = 2450 lx) under air for 8 min at 37 °C in CH3CN:PBS (1:1 v/v, w/3% DMSO, 10 mM, pH = 7.4). Time interval between spectral acquisitions is 1 min.
reaction within ∼8 min of illumination at λ = 419 nm (intensity = 2450 lx), minimal reactivity was observed for 3, indicating its suitability for application in cellular studies and corresponding to a negative response (value = 0) in the binary code of a truth table (Figure 5, vide infra). The conversion of 1 to 2 (Scheme 1, reaction 3) results in the loss of all absorbance features in the visible region.24,26 This means that the product remaining following quantitative CO release (determined by GC headspace measurement)24 is colorless and will have no emission when excited at wavelengths >400 nm. Thus, the conversion of 1 to 2 is a second photochromatic switch that results in CO release with the loss of visible light emission features. As such, this is the erasing of the memory buildup of 1 and indicates the completion of an ordered sequence of inputs resulting in CO release. We note that the CO release reaction of 1 likely proceeds via formation of a singlet excited state which subsequently undergoes intersystem crossing to a triplet excited-state species that reacts with 3O2.31 Cytotoxicity studies of 3 were performed using the A549 cell line. An MTT colorimetric assay was used to determine an IC50 value for 3 (62 μM, Figure S18). The unmasked flavonol 1 has an IC50 that is slightly lower (41.5 μM),24 indicating that inclusion of the acryloyl tail makes 3 less toxic relative to 1. We note that the CO-release product 2 (Scheme 1, reaction 2) was previously shown to be nontoxic.24 Overall, these studies demonstrated that the compounds were suitable for further studies in A549 cells at concentrations below the IC50 value. We next examined the properties of 3 in A549 cells using confocal microscopy. After 1 h of incubation of 3 at a concentration of 25 μM, the A549 cells showed good uptake, as indicated by the intense emission in the blue channel (Figure 3). Notably, emission is also evident in the green channel, which indicates the formation of the deprotected 1 via reaction of 3 with intracellular thiols. Importantly, a control reaction involving cells incubated in N-ethylmaleimide (NEM)-treated media (Figure 3) produced similar results, indicating fast uptake of 3 and that thiol
Figure 1. Plot of the time-dependent reactivity of 3 with excess cysteine (Cys), homocysteine (Hcy), or glutathione (GSH) (3:thiol = 1:10). Conditions: CH3CN:PBS = 1:1 v/v, with 3% DMSO, 10 mM, pH = 7.4, 37 °C).
treatment of 3 with a 10-fold excess of cysteine results in an increase in absorbance at 410 nm that is complete within ∼3 min at 37 °C. The pseudo-first-order rate constant for this reaction is 0.013 s−1 (Figure S13). The reaction involving Hcy with 3 requires >1 h to produce a similar concentration of 1. The reaction involving GSH is very slow and proceeds only to 90% conversion to 2 within ∼600 s. Reducing the amount of O2 present to 1% of the headspace slows this reaction, which reaches ∼45% completion within 1400 s. However, as the maximum concentration of 1 generated within the cellular environment will not surpass 25 μM (corresponding to the concentration of 3 co-incubated with the cells), we performed an additional experiment with 1% O2 content and [1] = 0.025 mM. Monitoring the fluorescence emission of 1 at 575 nm showed that the conversion of 1 to 2 was complete within
