Gold‐Triggered Uncaging Chemistry in Living Systems

Abstract Recent advances in bioorthogonal catalysis are increasing the capacity of researchers to manipulate the fate of molecules in complex biological systems. A bioorthogonal uncaging strategy is presented, which is triggered by heterogeneous gold catalysis and facilitates the activation of a structurally diverse range of therapeutics in cancer cell culture. Furthermore, this solid‐supported catalytic system enabled locally controlled release of a fluorescent dye into the brain of a zebrafish for the first time, offering a novel way to modulate the activity of bioorthogonal reagents in the most fragile and complex organs.

Seminal works showcased the capabilities of foreign transition metal catalysts in mediated chemoselective transformations within cells. [1] More recently,t he emerging field of bioorthogonal catalysis [2] has produced aw ealth of creativity in av ariety of applications,r anging from biomolecule labeling, [3a-c] metabolite detection [3d] and intra/subcellular probe release, [3e-h] to in situ enzyme [3i,j] and prodrug activation. [3k-o] Substoichiometric activity and access to agreater diversity of chemical processes and functionalities are some of the added benefits provided by abiotic transition metals to the current bioorthogonal toolbox, thus expanding the boundaries of this central field of research. [4] One of the latest additions to this area was recently reported by Tanaka and coworkers, [5] who developed an ovel strategy for in vivo protein labeling mediated by glycoalbumin-bound gold(III) complexes (Scheme 1). Despite recent advances in the field, many challenges lie ahead as transition metal catalysts show limited biocompatibility in living systems in terms of catalytic versatility and inherent toxicity.
Gold catalysis has received enormous attention in organic synthesis over the last decades. [6] Among the chemical properties of gold that stand out are its preference to coordinate with alkynes in the presence of other functional groups,i ncluding alkenes. [6,7] Solid-supported gold nanoparticles (Au-NP) have also attracted the interest of chemists searching for greener catalysts,b ecause such catalysts are recyclable,g old is safe to handle,a nd it demonstrates ar emarkable ability to catalytically mediate oxidative reactions at or even below ambient temperatures. [7] In the chemical biology field, however, the chemistry of gold is dominated by near-covalent SÀAu bonding. [8] This spontaneous bond formation provides the basis for the bottom-up self-assembly of monolayers functionalized with am ultitude Scheme 1. a) Au III -mediated bioorthogonala midation reported by Tanaka and coworkers. [5] b) The solid-supported gold-catalyzed uncaging strategy developed in this work.
of small molecules and biomolecules at the surface of the metal-a highly reliable process that has found widespread application in nanotechnology,b iotechnology,a nd theranostics. [8] Because of the high affinity of thiols for gold and their ubiquitous presence in peptides and proteins,t he attractive alkynophile properties of Au-NP pass unnoticed in the biological milieu. We envisioned that embedding Au-NP in as olid support would serve to protect the metal nanostructures from large thiol-rich biomolecules,w hile allowing alkyne-functionalized small molecules to enter and undergo gold-mediated chemistry even in biological systems.I mportantly,b ased on the high biocompatibility of metallic gold, such ad evice would be optimal to catalyze bioorthogonal transformations in vivo.I nasuitable shape and size,t his functional device could potentially be implanted by asurgeon at the anatomical site of alocalized cancer (for example,the brain) and enable the local "manufacture" of different drugs-in ac atalytic fashion-from systemically administered innocuous starting materials.T his unique delivery method would offer the benefits of drug release systems [9] (that is,focal treatment and reduced general side effects) with fewer limitations (for example,e xtended lifetime and access to multiple therapies).
Towards this goal, herein we report the first example of bond-cleavage chemistry mediated by heterogeneous gold catalysts in living systems (Scheme 1)-a previously overlooked chemical reactivity of gold that facilitates the efficient bioorthogonal uncaging of various clinically approved anticancer drugs in cancer cell culture and the first intracranial activation of ab ioorthogonal probe in zebrafish.
Solid-supported gold catalysts (Figure 1a,b) were prepared by in situ generation of Au-NP within ap olyethylene glycol (PEG)-grafted low-cross-linked polystyrene matrix. In short, amino-functionalized TentaGel HL resins of 75 microns in diameter (Rapp Polymere GmbH) were treated with tetrachloroauric(III) acid and sodium hydroxide in water:tetrahydrofuran (THF) followed by reduction with tetrakis(hydroxymethyl)phosphonium chloride (THPC;s ee full synthesis in the Supporting Information). THPC was used to control particle growth and distribution because of its relatively mild reducing properties. [10] High-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) images of ultramicrotome cross-sections of the resins showed uniformly dispersed polyhedral nanocrystals of 30 nm average diameter across the polymer framework ( Figure 1b;Supporting Information, Figure S1). X-ray photoelectron spectroscopy (XPS,S upporting Information) analysis detected an incremental gradient concentration of gold from the surface to the core of the resins,w ith an Au 0 /Au d+ ratio ranging from 7( periphery) to 19 (interior).
To investigate the properties of the [Au]-resins in physiological conditions,t he devices were incubated with nonfluorescent compound 1,w hich releases strongly fluorescent Rhodamine 110 (2)  devices and their capacity to activate multiple doses of masked reagent-although ag radual decay in activity was observed over time.
To better understand the enhanced catalytic activity of the [Au]-resins in the presence of serum, we investigated the influence of OH, SH, and NH 2 groups (nucleophilic groups found in serum proteins) in the conversion of 1 into 2 by adding excess ethylene glycol (3a), 2-mercaptoethanol (3b), and ethanolamine (3c), respectively.T hese low molecular weight chemicals were used to facilitate diffusion throughout the resins,thereby maximizing their interaction with internal Au-NP.A ll the reactions were carried out in PBS at 37 8 8C.
Although 3a had ar elatively minor effect on the catalytic capacity of the devices,substantial variations in activity were observed in the presence of 3b and 3c (Figure 1c). Thiol 3b suppressed the reactivity of the resins almost completely, whereas addition of amine 3c boosted catalytic activity significantly.T he combined presence of an excess of both 3b and 3c resulted in low levels of fluorescence intensity.This prompted us to investigate the influence of glutathione in the reaction, an atural reductant that contains aS Ha nd an NH 2 group in its structure.Since glutathione is found ubiquitously in the human plasma and the interstitial fluid at ac oncentration of 2-20 mm, [11] we studied the reaction of 1 and [Au]resins at concentrations ranging from 10 to 400 mm.Asshown in the Supporting Information, Figure S5, increasing glutathione levels up to 50 mm promoted the reaction, whereas greater concentrations led to as ubstantial reduction in fluorescence intensity.N otably,al arge regain in catalytic activity was achieved upon addition of an extra milligram of [Au]-resins to the inhibited reactions.I nc ontrast, if the concentration of probe 1 was augmented, no significant increase in fluorescence generation was observed. These results indicate that S À Au bonding of glutathione molecules on the surface of the Au-NP promote the dealkylation reaction until asaturation threshold is reached (see rationale in Scheme 2a). Over the saturation limit, gold-bound biomolecules will coat most active sites on the Au-NP surface, thus hindering gold-alkyne coordination. As eries of tests carried out to monitor and analyze the reaction (Supporting Information, Figures S6 and S7) corroborated that Rhodamine 2 was the main product of the reaction, along with intermediates that could correspond to organogold species. However,noreaction byproducts were isolated or identified, which points to the production of short-lived compounds. Based on these experimental observations,w et entatively propose adealkylation pathway whereby gold acts as a p-acid to activate the nucleophilic addition of biomolecules onto the terminal alkyne group,leading to release of the leaving group (for example,adye or drug) and the generation of reactive allenyl byproducts (Scheme 2b)t hat isomerize or hydrolyze under the reaction conditions.
Prior to testing the catalytic properties of the devices in cell culture,v iability assays (PrestoBlue reagent) were performed to determine the tolerance of cells to the presence of solid-supported gold. As anticipated, [Au]-resins were found to be fully biocompatible at the concentrations tested (Supporting Information, Figure S8).
Subsequently,the bioorthogonal [Au]-triggered release of as tructurally diverse selection of clinically used anticancer drugs was investigated in culture with human lung cancer A549 cells.T hree different drug precursors were tested (see Figure 2a): Pro-FUdR [12a] (4a), POB-SAHA [12b] (4b), and N-Poc-DOX( 4c;anovel drug precursor inspired by prior designs [3m, 13] ). Cells were treated with 4a-c and [Au]-resins separately (negative controls) or in combination (activation assay), and unmodified drugs 5a-c used as the positive controls.R emarkably,w hile prodrugs 4a-c did not elicit any effect on their own, potent anticancer activity was displayed in combination with [Au]-resins ( Figure 2a); unequivocal evidence that the active drugs are released in situ by heterogeneous gold chemistry.R euse of the [Au]-resins in three consecutive prodrug activation cycles confirmed the capacity of the devices to activate multiple drug doses (Supporting Information, Figure S9). Thesynthesis of drugs 5a-c was also verified in vitro (Supporting Information, Figure S10), confirming the capability of gold to cleave both O-a nd N-propargyl groups from ar ange of molecules based on structurally different scaffolds.These studies support apotential application scenario where gold-functionalized implants could be used in situ to modulate the spatiotemporal generation of chemotherapeutics from inactive precursors in the treatment of localized malignancies such as brain or prostate cancer.
Encouraged by the biocompatibility and catalytic properties of the [Au]-resins,weembarked on an innovative study to evaluate the capacity of the devices to convert nonfluorescent 1 into Rhodamine 2 inside the cranium of zebrafish embryos. Firstly,asingle [Au]-resin was carefully transplanted into the optic tectum, as mall anatomical cavity [14] of the brain of zebrafish larvae.After the surgery,either reagent 1 (activation assay) or just dimethyl sulfoxide (DMSO;n egative control) were added to the medium and embryos imaged at 24 h(n = 5). Because of its lipophilicity,p rodye 1 can enter the zebrafish through the skin and/or by ingestion and distribute systemically,b ut will only be converted into fluorescent compound 2 upon reaction with the [Au]-bead. As shown in Figure 2b,astrong green fluorescent signal originating from the [Au]-resin was observed only when incubated with 1, confirming the local generation of Rhodamine 2.P rolongation of the study by three additional days corroborated previous observations regarding the sustained functionality of the devices (Supporting Information, Figure S11). This study, which represents the first bioorthogonal organometallic reaction to be locally performed in the brain of al iving animal, indicates that heterogeneous gold catalysts have the capacity to mediate in vivo bioorthogonal release of functional reagents in as patially controlled manner.
In conclusion, we have developed ah eterogeneous catalytic system that enables access to chemical properties of Au-NP that were previously out of our reach in biological environments.S uch devices triggered the bioorthogonal uncaging of as tructurally diverse selection of cytotoxic precursors through an unexplored chemical reactivity of gold, providing an ovel and safe method to activate therapeutics by nonbiological chemical stimuli. [3k-o,13, 15] Furthermore,t his solid-supported catalyst enabled-for the first time-the locally controlled release of afluorescent dye in the brain of az ebrafish. This notable breakthrough expands our capacity to chemically modulate the activity of bioorthogonal reagents in the most fragile and complex organs. (Principalss cholarship), respectively,f or financial support. L.H. was supported by aCRUKPhD Fellowship.D .S.thanks CRUK for aCareer Establishment Award. V. S. ,S .I.,and J.S. thank CIBER-BBN for financial support. CIBER-BBN is an initiative funded by the VI National R&D&i Plan 2008-2011 financed by the Instituto de Salud Carlos III with assistance Figure 2. a) Gold-triggered activation of prodrugs 4a-c in A549 cancer cell culture. Negative controls:[ Au]-resins(1mgmL À1 ); 4a-c (10, 100, and 1 mm,r espectively). Positive control: 5a-c (10, 100, and 1 mm,r espectively). Prodrug activation assay:[Au]-resins + 4a-c (10, 100, and 1 mm, respectively). Cell viability was measured at day 4u sing PrestoBlue reagent. Error bars: AE SD from n = 3. b) Bioorthogonal gold-mediated release of green fluorescentRhodamine1 10 from precursor 1 in the brain of azebrafish. The presence of the [Au]-resin is indicated with aw hite arrow. Study of fluorescence intensity shows high statistical significance compared to the negative control (DMSO).