Radical‐Mediated Thiol‐Ene Strategy: Photoactivation of Thiol‐Containing Drugs in Cancer Cells

Abstract Photoactivated drugs provide an opportunity to improve efficacy alongside reducing side‐effects in the treatment of severe diseases such as cancer. Described herein is a photoactivation decaging method of isobutylene‐caged thiols through a UV‐initiated thiol‐ene reaction. The method was demonstrated with an isobutylene‐caged cysteine, cyclic disulfide‐peptide, and thiol‐containing drug, all of which were rapidly and efficiently released under mild UV irradiation in the presence of thiol sources and a photoinitiator. Importantly, it is shown that the activity of histone deacetylase inhibitor largazole can be switched off when stapled, but selectively switched on within cancer cells when irradiated with non‐phototoxic light.

Inrecent decades,p recision medicine has drawn al ot of attention for the effective treatment of various diseases, especially cancer. Currently,asaresult of alack of selectivity in the pathological sites,t he development of new,e ffective, and safe therapies remains challenging.A mong various new methods,t he recently developed light-mediated treatment is recognized as ap romising approach to achieve controlled activation of medicine at pathological sites, [1] and could significantly reduce the side effects of chemotherapy.S of ar, in the battle against cancer, several types of light-activated anticancer reagents,w hich can be switched on conditionally with irradiation, have been investigated. [2] Thes tructures of these photocaged drugs include various ultraviolet, nearinfrared, or visible responsive moieties, [3] such as o-nitrobenzyl, [2a, 4] coumarinyl ester, [5] and metal complexes. [2e,6] These photoresponsive structures offer an extensive toolbox for use in cancer therapies and other biological applications. However,i ssues and challenges remain in this field, such as using nontoxic wavelengths,a chieving rapid and efficient conversion, and improving the bioavailability of the prodrug. Therefore,t here is still demand for new designs and new developments for photomediated therapy.
Thiol-ene reactions (Scheme 1), ac onjugation between at hiol and an alkene,h ave been known since the early 1900s. [7] Thec oupling reaction proceeds through two mechanisms,n amely photoinitiated free-radical addition and catalyzed Michael addition reactions.T here are several desirable features of aclick reaction, including rapid reaction rates,e ase of implementation, high yields and rates of conversion, [8] so thiol-ene "click" reactions have been increasingly used for various applications,s uch as biofunctionalization, [9] surface and polymer modification, polymerization, [10] and so on. [8,11] Recently,i sobutylene-bridged polymer networks have been extensively studied to synthesize polymer networks through radical-initiated thiol-ene chemistry. [12] This covalently cross-linked network is able to undergo photomediated, reversible cleavage of its isobutylene backbone to allow chain rearrangement and relieve structural strain. This method has also been used to provide ar eactive handle for reversible addition and exchange of biochemical moieties under cytocompatible conditions. [12d] Key to this reaction is the isobutylene structure capable of additionfragmentation chain transfer (AFCT), in which the structure is attacked by the photoinitiated thiol radical in the presence of aphotoinitiator (PI) to release the caged thiol part.
Inspired by this AFCT reaction, we hypothesized that the the isobutylene structure could be used as abridging graft to cage thiol-containing drugs and allow further controlled activation of anticancer drugs by means of aradical-mediated thiol-ene mechanism (Scheme 1). Previously,w er eported ao ne-pot macrocyclization strategy [with tris(2-carboxyethyl)phosphine (TCEP)] for thiol-containing peptides by using an isobutylene graft, which can significantly enhance both membrane permeability and binding activity of the corresponding macrocycles. [13] Thei sobutylene graft can be rapidly and efficiently installed onto reduced thiol groups in abiocompatible manner because of the high reactivity of the bis(bromo)isobutylene.O ur research began with N-tertbutoxycarbonyl-l-cysteine methyl ester (N-Boc-Cys-OMe 1), which was protected with 3-bromo-2-bromomethyl-1propene (Table 1a nd Supporting Information). Thes tapled cysteine 2 was then screened under as eries of reaction conditions.F irstly,d ifferent PIs were tested with the same amounts of thiol source,1 -thio-b-d-glucose tetraacetate (4AcGlcSH), and 2 under irradiation at 365 nm. Thereaction with the water-soluble PI 2,2'-azobis[2-(2-imidazolin-2yl)propane]dihydrochloride (Vazo 44) did not progress after two hours.H owever,t he other PI, 2,2-dimethoxy-2-phenylacetophenone (DPAP) successfully promoted the reaction and released 1. b-Mercaptoethanol (BME) was also tested as the thiol compound under the same reaction conditions.T his reaction released the cysteine in aslightly lower yield, which suggests that different thiol sources could be used to promote the reaction. However,w hen N,N-dimethylformamide (DMF) was used as solvent, two mixed disulfides resulting from the reaction between the 1-thioglucose and N-Boccysteine methyl ester 1' were observed. To evaluate the reaction and calculate the isolated yield, TCEP was added to reduce the disulfides after the reaction. Under these reaction conditions,t he reaction was complete within 15 minutes and gave ar elatively high yield (65 %). Moreover,t he reaction was also rapid and efficient when glutathione was used as athiol source in amixture of DMF and water as the solvent, which shows that the reaction is practical for an aqueous environment. Then, the feasibility of our strategy was investigated with an isobutylene cyclized 5-mer peptide that bears two terminal cysteines and was synthesized by solidphase peptide synthesis. [13] Theshort peptide was completely converted into the disulfide derivative within 15 minutes under the same reaction conditions described previously ( Figure 1).
To demonstrate that the release of the thiol compound occurs by means of ar adical-mediated thiol-ene mechanism, as eries of control experiments were conducted (Table 2a nd Supporting Information). As shown in Table 2, UV irradiation (entry 2), athiol source (entry 3), and aPIare essential for the decaging reaction. When the radical scavenger (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO) was added to the reaction, it terminated the reaction by forming an intermediate with the DPAP fragmentation radical, which suggests that the reaction is mediated by radicals.
Thep roposed mechanism for the radical-mediated thiolene decaging reaction was studied by quantum mechanical calculations using abbreviated thiol models (see the Supporting Information) and is shown as Scheme 2. After generation of the thiol radical by the PI under UV irradiation, the isobutylene grafted structure undergoes afast thiol-ene anti-Markovnikov addition with ac alculated activation energy of DG°% 14 kcal mol À1 at the PCM(H 2 O)/M06-2X/6-31+ ++ + G(2,p) theory level, to generate as ymmetric tertiarycarbon-centered radical intermediate.T hen, the unstable radical intermediate undergoes a b-scission at av ery similar reaction rate,r egenerating the isobutylene linkage and resulting in am ixed caged compound, which can be again attacked by another thiol radical following the same mechanism to release the other unit of the caged thiol compound. Thep rocess is nearly thermoneutral and reversible until two [a] Amixture of disulfides formed by reaction of cysteine and 1thioglucose and two cysteines. GSH = glutathione;PI= photoinitiator.

Angewandte Chemie
Zuschriften decaged radical thiols collapse forming astable disulfide bond (Scheme 2). With all this knowledge in hand and to demonstrate that this strategy is practical to activate drugs in vitro,weapplied our method to the potent histone deacetylase inhibitor (HDAC)l argazole.T he cyclic depsipeptide largazole is am arine natural product, and its derivatives are recognized as promising potential anticancer therapeutics.L argazole possesses remarkable and preferential growth-inhibitory activity against cancer cell lines relative to corresponding nontransformed cells. [14] Theo ctanoyl tail in largazole has better cell permeability than the active free thiol species, largazole thiol, the latter being formed inside cells by esterase or lipase-based cleavage of the octanoyl residue. [15] Thef ree thiol group binds to the active site Zn 2+ -domain within the HDAC enzyme,and results in apotent inhibitory effect. Thus, the thiol-ene decaging strategy can be used to protect the thiol group,i mprove the cell-permeability,a nd allow controlled activation with UV.Therefore,wesynthesized three largazole derivatives:l argazole,l argazole thiol, and stapled largazole (see the Supporting Information). Ther esult of the parallel artificial membrane permeability assay indicated that stapled largazole is ah ighly passively permeable compound (logP e À5.29 and P e 5.3 10 À6 cm sec À1 ;see Table S4 in the Supporting Information). Then, the stapled largazole was reacted with 1-thio-b-d-glucose tetraacetate and DPAP under UV irradiation for 15 minutes (Figure 2a). Full conversion of the stapled largazole was observed in the HPLC trace along with the appearance of al argazole thiol signal (Figure 2b). Next, the growth-inhibitory activity was evaluated with human colon carcinoma cell lines,H CT-116 ( Figure 2c). As expected, [16] largazole (GI 50 1.433 nm)i sm ore potent than the corresponding free thiol species (GI 50 185.1 nm)owing to its octanoyl side-chain which improves cell permeability and allows facile presentation of the free thiol within the cell. [15] Thes tapled largazole (GI 50 407.7 nm)i samuch less potent compound because the thiol group is protected by the isobutylene structure,w hich prevents binding with the Zn 2+ domain in cells.B efore testing the decaging conditions in cells,weinvestigated the toxicity of DPAP and phototoxicity of the light in terms of the power and the irradiation time (see Figure S7). As et of cytocompatible conditions,1 5minute irradiation at 80 W, were chosen to conduct further investigations.
Thedecaging reaction of stapled largazole was tested with HCT-116 cells at 150 nm.M cCoys5 Ac ulturing medium contains cysteine and glutathione,sonoother thiol source was added to the medium. Thec ell viabilities of the three drug groups with/without UV irradiation were consistent with the growth-inhibitory assay (Figure 2d). TheU V-irradiated premixed group of stapled largazole and DPAP showed significantly lower cell viability than the corresponding nonirradiation group,t he stapled largazole group,a nd the largazole thiol group.T oc onfirm the results,afluorometric HDAC activity assay was conducted with HCT-116 cell lysates ( Figure 2e). As ignificant decrease of fluorescence was observed in the UV-irradiated premixed group relative to the control groups and the non-irradiation group.B oth the cell viability and the enzyme activity results indicated that the stapled largazole was successfully activated by UV light.
In summary,w eh ave developed ar apid and efficient thiol-ene-based photoactivation strategy for thiol-containing drugs caged using isobutylene.The radical-mediated reaction, that is triggered by UV light, undergoes at hiol-ene addition step to form an unstable radical intermediate which is further Figure 2. The photoactivation of isobutylene-grafted largazole thiol. a) Thiol-ene decaging reaction of stapled largazole with 1-thio-b-dglucose tetraacetate. HPLC traces of the reaction mixture. Blue: stapled largazole;r ed:reaction mixture after 15 min. b) Growth inhibitory effects of largazole, largazole thiol, and stapled largazole on HCT-116 colon carcinoma cells:for largazole GI 50 1.433 nm, for largazole thiol GI 50 185.1 nm, and for stapled largazole GI 50 407.7 nm. c) Cell survival under different conditions. SL = stapled largazole, UV condition,3 65 nm 80 W, 15 min. d) HDAC activity assay of control group and UV group. Data are representative of three independent tests and analyzed by the two-tailed unpaired Student's t-test. **, P < 0.01;* ***, P < 0.0001. Error bars reflect one standard deviation from the mean. Scheme 2. Proposed mechanism for the radical-mediated thiol-isobutylene decaging reaction. cleaved by b-scission to release the caged thiols.W ea pplied this method to various substrates,s uch as N-Boc-cysteine, ac ysteine-containing peptide,a nd the HDAC inhibitor largazole,a nd showed that the caged thiol molecules,u nlike their free counterparts,display high membrane permeability. Thes uccessful activation of largazole in HCT-116 cells demonstrates the potential as ad rug delivery and activation method for cancer therapy.Invivo,this strategy may explore radical initiators that are often characteristic of cancer cells to achieve targeted drug activation. And finally,f urther investigation of the use of this strategy on proteins could also be useful for photocontrolled protein activation and drug release from antibody-drug conjugates.