Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Chembiochem. Author manuscript; available in PMC 2018 January 3.
Published in final edited form as:
PMCID: PMC5213739

A Thioacetal Photocage Designed for Dual Release: Application in the Quantitation of Therapeutic Release via Synchronous Reporter Decaging


Despite the immense potential, application of existing photocaging technology is limited by the paucity of advanced caging tools. Here we report on the design of a novel thioacetal ortho-nitrobenzaldehyde (TNB) dual arm photocage that enables control of the simultaneous release of two payloads linked to a single TNB unit. Using this cage which was prepared in a single step from commercial 6-nitroverataldehyde, three drug-fluorophore conjugates were synthesized, Taxol-TNB-Fluorescein, Taxol-TNB-Coumarin and Doxorubicin-TNB-Coumarin, and long wavelength UVA light-triggered release experiments demonstrated that dual payload release occurred with rapid decay kinetics for each conjugate. In cell based assays in vitro, dual release could also be controlled by UV exposure, resulting in increased cellular fluorescence and cytotoxicity as potent as unmodified drug towards the KB carcinoma cell line. The extent of such dual release was quantifiable by reporter fluorescence measured in situ, and found to correlate with the extent of cytotoxicity. Thus, this novel dual-arm cage strategy provides a valuable tool which enables both active control and real-time monitoring of drug activation at the delivery site.

Keywords: Caged Compounds, Dual Arm Cage, Photolysis, Controlled Release, Antitumor Agents, Fluorescence Reporter

Graphical Abstract

An external file that holds a picture, illustration, etc.
Object name is nihms835829u1.jpg


Photocaging[1] has played an instrumental role in the spatiotemporal control of biological processes.[2] It is based on the temporary inactivation of a biologically active molecule through conjugation with a photocleavable protecting group (photocage) which can be conditionally removed by light exposure to reactivate the molecule (Figure 1). The concept of photocaging has been extensively demonstrated in the control of cellular activities in vitro such as channel gating,[1] protein activation[3] and gene regulation[2, 4] as well as in the development of optogenetic animal models in vivo.[4b, 4c]

Figure 1
(a) The photocaging strategy as illustrated with a conventional single-arm photocage, ortho-nitrobenzyl (ONB); (b) A dual-arm strategy based on a thioacetal ortho-nitrobenzaldehyde (TNB) photocage that enables the synchronous release and activation of ...

Protecting groups for photocaging are primarily composed of UV or visible light-responsive aromatic ring structures that include ortho-nitrobenzene (ONB),[1]nitro-benzofuran,[5] 6-bromo-7-hydroxycoumarin (BHC),[6] quinolone,[7] and cyanine.[8] Recently, the use of these cage molecules has been expanded to areas of controlled drug release such as photopharmacology,[9] nanomedicine,[10] and image-guided drug transport.[11] In particular, the unique application of these molecules for the fluorescence-based imaging of intracellular prodrug activation[11e, 11f, 12] is enabled by the multifunctional design capacity of the photocleavable cages which allow for both a reporter fluorophore and a drug molecule to be attached to the same cleavable molecule. The spatially and temporally-controlled release of both drug and imaging molecules within target cells allows for monitoring of the extent of drug release with a reporter readout in real-time.[11a–d] Of these cages, the ONB cage, which is the most well characterized for its cleavage by one-photon[1] and two-photon[4b, 5a, 6a] absorption, has played a significant role in the advancement of controlled drug release.[2, 13]

Despite its limited penetration through the skin (depth ≤0.2 mm),[14] long wavelength UVA (315–400 nm) has ability to reach as far as the subcutaneous layer. UVA is less toxic to most mammalian cells than shorter UVB and UVC, and thus has served as a tool for active control in chemical biology such as protein activation,[3a, 5a, 15] optogenetics,[4a, 4c, 16] and photochemical internalization of biomacromolecules.[17] More recently, applications of light-controlled mechanisms for drug release[13] have been developed by numerous laboratories including ours in the design of prodrugs and drug delivery systems for various therapeutic agents including olaparib (AZD-2281),[11e] melphalan,[11f] methotrexate,[18] doxorubicin (Dox),[10b, 10c, 19] paclitaxel (taxol),[20] 5-fluorouracil,[21] tamoxifen,[4a, 8, 16] and ciprofloxacin.[22] However, efficient implementation of photocage molecules for such applications[13] is often challenging due in part to the synthetic complexity involved in drug and/or reporter conjugation which requires a multistep process. Some of the applications involving nanoparticulate carrier systems required its modification for presenting two attachable ends--one to the drug (or reporter probe), and the other to a particulate carrier.[10c, 11a, 19, 21, 23] The design of photocage molecules as linkers for non-particulate, small molecule applications is particularly difficult as compared to the carrier-based ones, as reflected by the scarcity of validated functional linkers.[11f, 24] In addition, fewer photocage linkers have been implemented for the control of drug release[2223] as compared to other types of linkers with release mechanisms triggered by cellular factors or conditions such as low pH,[13] reactive thiols[11b, 25] and metabolic enzymes.[11f, 26]

In this study, we designed a dual-arm photocage molecule that provides both convenience in synthesis and payload conjugation and incorporates an externally controlled active release mechanism (Figure 1b). This photocage is based on a novel thioacetal ortho-nitrobenzaldehyde (TNB) which is cleaved in response to light exposure. This photocage strategy[2425] allows for the active control of release unlike other cleavable linkers that primarily rely on chemical and metabolic mechanisms that occur passively in response to cellular factors.[13] Here we evaluated the practical application of this TNB cage for fluorescence-based monitoring of drug release by cross-tethering both an anticancer agent and a fluorescent probe to the cage. This dual conjugation strategy allows for light-controlled drug release and concomitant activation of the fluorescent reporter which can be quantified in situ. As cellular cytotoxicity is directly correlated with the intracellular concentration of drug released, information regarding the efficiency of this release system can be obtained from the fluorescence readout. Thus the intracellular kinetics and extent of drug release can be followed and measured on a real-time basis. Thus this article reports on the design of three anticancer TNB reporter conjugates with the proof of concept application validated for light-controlled drug release and real-time monitoring in cellular systems.

Results and Discussion

Rationale of TNB Design

Despite extensive use, the ONB cage presents only a single cleavage site, and is not optimal for dual-arm design which requires two attachment sites per cage. In order to address this design barrier and expand the applications of the ONB cage, we designed a TNB system in which the benzylic position is flanked by two identical arms, each connected with a C–S bond instead of a C–O bond (Figure 2). Accordingly, each of the two arms has the potential for conjugation to either a targeting ligand, drug or reporter molecule, and has the ability to release the two payloads simultaneously via light-controlled cleavage of each C–S bond. This light control adds the benefit of selective payload release, as the thioacetal which is often used as the protecting group for carbonyl compounds is stable under a wide range of physiological conditions and is only susceptible to degradation under harsh synthetic conditions with oxidative or extremely acidic reagents.[27]

Figure 2
Structures of thioacetal o-nitrobenzaldehyde (TNB)-based photocages 1, 2 and three TNB-derived compounds 35 in which each TNB molecule serves as a branching spacer flanked by paclitaxel (taxol) or doxorubicin (Dox) as a therapeutic molecule and ...

Synthesis of TNB Cages

Two types of TNB photocages, 1 and 2 (Figure 2), each terminated with carboxylic acid or alcohol, respectively, were synthesized with 76–89% yield from commercial 6-nitroverataldehyde and a requisite thiol (2.4 mol equiv) in a condensation reaction catalyzed by BF3·Et2O (1.2 mol equiv) and acetic acid (2.4 mol equiv) at ~0 °C (Scheme S2 and Scheme S3, Supporting Information): MS (ESI; m/z): 1 = 428.0 [M + Na]+; 2 = 372.0 [M + Na]+. This one-step synthetic methodology confers rapid access to various TNB photocages on a multigram scale. It is similarly applicable to other aromatic aldehyde precursors used in photocaging such as 7-(diethylamino)coumarin-4-carboxaldehyde which was readily converted to its thioacetal form with 3-mercaptopropionic acid (not shown): MS (ESI; m/z): 462.0 [M + Na]+, 440.0 [M + H]+. In addition, each TNB photocage has a symmetric thioacetal which provides equal reactivity in each of its two arms during dual payload conjugation.

Synthesis of TNB Conjugates

We illustrate the synthetic convenience of this TNB cage strategy for dual payload conjugation by synthesizing three TNB conjugates 35. First, 1 TNB in which each arm is terminated with a carboxylic acid was coupled with taxol at its C2′-OH to prepare an ester-based Taxol-TNB conjugate 6 (Scheme S2, Supporting Information). The other carboxylic acid remaining in the intermediate was subsequently conjugated with a reporter probe including fluorescein methyl ester (fluorescein) or 4-methyl-7-coumarinol (coumarin) through an ester linkage. This sequential reaction led to 3 Taxol-TNB-Fluorescein and 4 Taxol-TNB-Coumarin, respectively, and each conjugate was fully characterized by mass spectrometry, 1H NMR spectroscopy and fluorescence spectroscopy (Supporting Information): purity by analytical ultraperformance liquid chromatography (UPLC): ≥95% (3, 4); ESI HRMS calcd for 3 (C83H80N2O25S2) 1569.4564 [M + H]+, found 1569.4542; calcd for 4 (C72H74N2O23S2) 1399.4203 [M + H]+, found 1399.4145; fluorescence spectroscopy 3: λex = 480 (±5) nm, λem = 520 (±5) nm; 4: λex = 365 nm, λem = 445 nm.

Second, 2 TNB which is functionalized with an alcohol in its arm was derivatized by coupling with 4-methyl-7-coumarinol through carbonate formation, and then with Dox through a carbamate linkage, resulting in 5 Dox-TNB-Coumarin (Scheme S3, Supporting Information). This TNB conjugate was characterized as above: ESI HRMS calcd for 5 (C52H52N2O22S2) 1138.2791 [M+NH4]+, found 1138.2777; analytical UPLC purity (≥95%), and fluorescence spectroscopy 5: λex = 365 nm, λem = 445 nm.

Photolysis of TNB Cages

UV–vis absorption spectra of the two TNB cages 1 and 2 were measured in an aqueous medium (Figure 3). Each showed a λmax at 346 nm (ε = 5,950 (1); ε = 4,292 (2) M−1 cm−1) which is slightly longer than the λmax value of 340 nm for a conventional ONB cage terminated with benzylic alcohol.[18b] This bathochromic shift might be attributable to the effect of the sulfur atom placed on the TNB cage which is less electron withdrawing than the benzylic oxygen in the ONB system.[28] The photolysis of TNB cages 1 and 2 was performed by UVA exposure (365 nm), and its progress was monitored by 1H NMR spectroscopy, UV–vis spectrometry and UPLC, each providing evidence supportive of the cleavage mechanism proposed in Figure 3a.

Figure 3
(a) A proposed mechanism for the photolysis of 1 TNB(CO2H) by long wavelength UVA light (365 nm); (b) 1H NMR spectral traces (500 MHz) and proton assignment for the photolysis products of 1 (2.5 mM in 20% CD3OD/D2O); (c) UV–vis spectral traces ...

First, 1H NMR analysis of photolysed 1 showed time-dependent disappearance of its benzylic proton (δ = 6.02 ppm) at the thioacetal group and concomitant growth of new aromatic signals in the upper field (Figure 3b). This is indicative of the oxidation of the thioacetal to 2-nitrosobenzoic thioester which is believed to undergo subsequent hydrolysis to the 2-nitrosobenzoic acid and 3-mercaptopropionic acid (detected as an oxidized disulfide form; Figure S1, Supporting Information). In addition to such hydrolytic cleavage, we anticipate that other types of chemical cleavage or ligation[29] may occur when it is cleaved in a cellular environment through reacting with amines and thiols. Second, UV–vis spectral traces acquired for each cage showed a rapid decrease in absorbance at the λmax of each compound for light exposure times of up to 2 min (Figure 3c), suggesting rapid photolytic cleavage of the TNB cage. The quantum yield (Φuncaging) of uncaging was determined by ferrioxalate actinometry[30] for 1 and 2 which showed 0.20 and 0.19, respectively, as summarized in Table S1. Such spectral and photophysical features were similarly observed in the photolysis of the comparable ONB cage (Φuncaging = 0.01–0.7).[6b] Third, UPLC analysis for photolysed 1 showed the appearance of three new peaks, each at a shorter retention time (Figure S1, Supporting Information), which is consistent with the formation of smaller molecular species including the nitrosobenzoic acid and its (thio)esters. Collectively, these data point to the photocleavage of the TNB cage, yielding a 2-nitrosobenzoic acid product with the concomitant release of two thiol-terminated spacer molecules.

Light-controlled Release Kinetics

First as a representative example, 3 Taxol-TNB-Fluorescein in an aqueous solution (1:1 water/methanol) was exposed to UVA light (365 nm), and fluorescence spectra were measured (Figure 4). The emission intensities were compared at 520 nm (λem for fluorescein), showing an increase in emission as a function of exposure time (4-fold increase over a 10 min period). We attribute this fluorescence change to the release of fluorescein in its free phenolic form which is characterized by stronger fluorescence emission than the unreleased conjugated form. The increase in fluorescence is consistent with the growth of an absorption peak at 485 nm (Figure S2b, Supporting Information) which is indicative of fluorescein release. UPLC analysis was also performed with the photolysed solutions. Area under the curve (AUC) analyses indicated that the photolysis of 3 occurs rapidly in a time-dependent manner with a decay half-life (t1/2) of <2 min and Φuncaging of 0.05 (Figure 4b). The UPLC analysis in combination with LC–MS analysis enabled to detect the release of two payloads, fluorescein and taxol (Figure S2, Supporting Information).

Figure 4
Photolysis of 3 Taxol-TNB-Fluorescein, 4 Taxol-TNB-Coumarin and 5 Dox-TNB-Coumarin in solution. (a) A proposed mechanism for the light-mediated dual payload release in 3 and the structures of precursor compounds (A–D). (b–d) Release kinetics ...

We believe that such dual release occurs through the proposed mechanisms described in Figure 4 in which a thiol-terminated precursor (C, D) is formed initially by light-triggered C–S bond cleavage followed by intramolecular self-immolation[11f, 31] which leads to the release of the free payload species. This release mechanism can also produce two other intermediates (A, B), and these thiol ester-containing intermediates can be each then converted to the thiol-terminated precursor (C, D) through hydrolysis, transthioesterification[29, 32] or amidation[29] by water, thiols, amines or other nucleophilic biomolecules in the cell. This mechanism is supported by the results of the LC–MS analysis performed for Taxol-TNB (a reference compound without a linked reporter molecule) after 2 min of irradiation which showed the release of a precursor A along with taxol (Figure S3, Supporting Information). Our proposed mechanism for taxol release from the precursor D is also in agreement with the drug release from a thiol-terminated alkanoyl taxoid[33] (a homologue to the precursor D) via self-immolative cyclization which has been pioneered by Ojima, et al.[31a] and others[11c, 34] using thiol-terminated prodrugs. However, despite the same mechanism, the rate of payload release from precursor C and D might be not identical due to the difference in their leaving group capability. Thus, the release of fluorescein that occurs via a phenoxide anion (pKa = 6.3–6.8)[35] could be faster than that of paclitaxel that occurs via an alkoxide anion (pKa >10). In summary, we believe that the present data in combination with existing knowledge are supportive of the dual payload release via the self-immolative mechanism proposed here.

Light-controlled release studies were also performed with 4 Taxol-TNB-Coumarin and 5 Dox-TNB-Coumarin, and photolysed solutions were characterized as summarized in Figure 4 and by UV–vis spectrometry and LC–MS analysis (Figure S4, Figure S5; Supporting Information). Photolysis of these led to an increase in fluorescence intensity of 2.4 and 4.7-fold, respectively, relative to the level before light exposure (Figure 4). This enhanced fluorescence occurs with the rapid disappearance of each TNB construct with decay half-lives (t1/2) of <2 min in UPLC analysis and Φuncaging of 0.07 (4) and 0.08 (5). Their UV–vis spectra indicated exposure time-dependent rapid changes in absorption features. The LC–MS analysis of selected photolysed solutions provided evidence supportive of the release of the free drug molecule and coumarin. We believe that this release also occurs via an intramolecular cyclization reaction involving an ester[31a, 33] (4) or carbonate[11c, 34] (5) bond by the nucleophilic terminal thiol on the drug or reporter precursor. It is notable that the quantum efficiency (Φuncaging) of uncaging determined for each conjugate 3 to 5 shows ~3–4-fold decrease relative to its parent TNB cage as presented above. This decrease might be attributable to photon absorption (such as a protective “antenna effect”[36]) by a conjugated payload molecule such as fluorescein, coumarin and dox given their significant molar absorptivity (ε) at the irradiation (UVA) wavelength (Figure S2–S5).

Intracellular Uptake and Light Triggered Release

To determine whether the TNB conjugates could be taken up intracellularly, and subsequently cleaved through light activation after uptake, cellular fluorescence was analyzed using flow cytometry. KB cells, an epithelial carcinoma line overexpressing the folic acid (FA) receptor (FAR(+)) were used as a model system. Cells were incubated with 1.5 μM of 3 Taxol-TNB-Fluorescein in FA free cell culture media for 0, 2, 6, and 24 hr at 37 °C (Figure 5). At the end of the incubation period, unincorporated conjugate was removed by washing, and the cells were then exposed to long wavelength UVA (365 nm) light for 2 min. For the incubation time of 0 h, the conjugate was added to the cells and immediately washed off prior to light exposure. Cellular fluorescein fluorescence (mean fluorescence intensity (MFI)) was then measured by flow cytometry. The percentage of cells with fluorescence greater than the intrinsic fluorescence of untreated cells was also determined (Figure 5c). 3 Taxol-TNB-Fluorescein was taken up intracellularly within 2 hr of incubation, and continued to be taken up over 24 hr as evidenced by the slight increase in the cellular MFI and the percentage of cells with high fluorescence in the absence of light exposure, attributable to the fluorescence of the uncleaved conjugate (Figure 5c). Upon light exposure, a large increase in the cellular MFI and percentage of cells containing increased fluorescence was observed for cells incubated with 3 (e.g., 6.3 fold increase in MFI, and 6.0 fold increase in the % cells with increased fluorescence for cells treated with 1.5 μM of 3 with 6 hr of incubation prior to light exposure). The increase in cellular fluorescence induced by light exposure increased in magnitude with longer preincubation times between 0–6 hr, reflecting the accumulation of the conjugate intracellularly (Figure 5b, Figure 5c). These results confirm that a significant amount of the intact Taxol-TNB-Fluorescein conjugate is taken up intracellularly and is retained within the cell in a relatively intact state that maintains the fluorescein reporter in a fluorescently quenched state. Furthermore, these results demonstrate that UV mediated photocleavage of the TNB construct can occur intracellularly. The decrease in MFI and % of fluorescent cells between 6 and 24 hr, however, is attributable to cell death due to the cytotoxic effects of the drug conjugate.

Figure 5
Determination of intracellular uptake by flow cytometric analysis. KB cells were incubated with 1.5 μM of 3 Taxol-TNB-Fluorescein for the indicated times (t = 0, 2, 6, and 24 hr), washed to remove extracellular conjugate, and then exposed to UVA ...

Correlation Between Fluorescence and Cytotoxicity In Vitro

To confirm the functional activity of the TNB conjugates and the released drug, their cytotoxic properties were evaluated in KB cells with or without exposure to UV. A microplate assay which allows measurement of fluorescence on a real-time basis on live cells, and measurement of the viability of the same cells by an XTT assay was employed instead of the fluorimeter which was used for photolysis in solution (Figure 4). FAR(+) KB cells were treated with 3 Taxol-TNB-Fluorescein, 4 Taxol-TNB-Coumarin and 5 Dox-TNB-Coumarin at various concentrations and exposed briefly to UVA light (365 nm) for 2 min (Figure 6). The fluorescence of the reporter molecule was measured immediately. A large increase in either fluorescein or coumarin reporter fluorescence (several orders of magnitude) was seen for all three conjugates upon UV irradiation relative to treated cells that were not exposed to UV. Fluorescein and coumarin conjugates had similar low fluorescence emission intensities prior to light activation at their respective excitation and emission wavelengths. Interestingly, for the Taxol-TNB conjugates, the fluorescein containing conjugate, 3, displayed an increase in fluorescence of much greater magnitude compared to the coumarin conjugate 4 (Figure 6a, Figure 6b). This lower fluorescence exhibited by 4 might be explained, in part, by its lower increase in fluorescence in solution after UV exposure (Figure 4b, Figure 4c).

Figure 6
Fluorescence intensity (a–c) measured by a microplate reader and cell viability (d–f) of FAR(+) KB cells treated with 3, 4 or 5 with or without exposure to long wavelength UV (365 nm) for 2 min (UV = 0, 2 min). Fluorescence was measured ...

The increase in reporter fluorescence upon light exposure correlated with the concomitant increase in cytotoxicity of the drug conjugates. Cytotoxicity displayed a dose-dependency for both non UV and UV treated cells. Prior to UV exposure, 3 Taxol-TNB-Fluorescein, 4 Taxol-TNB-Coumarin and 5 Dox-TNB-Coumarin exhibited minimal cytotoxicity (IC50 values of 0.24 μM, 0.23 μM, and 9.22 μM, respectively) as compared to unconjugated doxorubicin and taxol (0.04 μM, 0.02 μM, respectively; Figure S6), suggesting that conjugation can reduce systemic toxicity. Upon UV exposure, the cytotoxicity of the conjugates increased dramatically, decreasing the IC50 values by an order of magnitude (IC50 values for 3, 4, 5: 0.04 μM, 0.05 μM, and 0.84 μM, respectively). As controls, no effect of light exposure on cell viability or fluorescence intensity (inset) was observed for either free taxol or doxorubicin, both of which were highly cytotoxic at even low nM concentration (Figure S6). These results demonstrate the ability to actively control the cytotoxicity of these constructs temporally and support the utility of fluorescence readouts as a means for monitoring the amount of drug transported to the target cell and accordingly, for predicting tumor cell viability in situ.

Real-time Release Kinetics

The positive correlation observed between fluorescence and cytotoxicity above is supportive of the ability to exert spatiotemporal control on drug release and perform quantitative monitoring of the extent of release indirectly via fluorescence measurements in real-time. In order to characterize the release kinetics in cellular systems, we determined the extent of additional reporter release (reflective of drug release), over time after the initial UV stimulus by following the increase in the fluorescence of live cells over 24 hr of incubation at 37 °C (Figure 7). Incubation of cells treated with 3 or 4 and exposed to UV (exposure time = 0.5 or 2 min) led to a rapid increase in fluorescence within the first 3 hr of incubation, followed by a small gradual increase over the remaining 24 hr. The endpoint fluorescence after 24 hr was more than two-fold the fluorescence measured immediately after UV exposure. The continued release of reporter and drug for an extended period after the initial UV stimulus was not due to cleavage of the cage by cellular factors but this slow gradual increase is likely due to the self-immolative conversion of transient dye intermediates to a fully fluorescent dye molecule, as similar release kinetics were observed in the control which lacked cells (Figure 7d).

Figure 7
Real-time monitoring of reporter fluorescence intensities under three different exposure conditions. (a–c) The fluorescence of FAR(+) KB cells treated with 3 Taxol-TNB-Fluorescein, 4 Taxol-TNB-Coumarin or 5 Dox-TNB-Coumarin, each at 1.5 μM, ...

Interestingly, in contrast to conjugates 3 and 4, conjugate 5 showed only a minimal additional increase in fluorescence over the course of 24 hr from the initial jump in fluorescence immediately after 2 min of UV exposure. This suggests that 2 min of UV exposure was sufficient to rapidly and fully release all of the coumarin in 5. Thus dye release from light exposed 5 is likely to occur faster than from 3 or 4. We believe that such differences among the three TNB conjugates might be attributable to the mechanism and rate of the self-immolation reaction[11c, 34] by which the fluorescent dye molecule is converted from its respective thiol precursor. The thiol precursor generated from 5 has a coumarin moiety attached to the spacer through a more chemically reactive carbonate functionality which is thus more labile to nucleophilic displacement reactions than the ester functionality present in the thiol precursors from 3 and 4.

Control cells which were treated with the conjugates but were unexposed to UV showed only minimal changes in bulk reporter fluorescence over 3 hr of incubation, indicating stability of 35 during the fluorescence measurement time period in the fluorescence assay in Figure 6 and confirming the role of UV in reporter activation. However, prolonged incubation for 24 hr led to some increase in reporter fluorescence for conjugates 3–5 even without light exposure, although the fluorescence at 24 hr was still <30–50% that of the light exposed cells after 24 hr. This “dark” release of reporter and putatively drug, accounts for the degree of cytotoxicity observed for the non-UV treated cells in Figure 6. Such partial increase suggests the possibility of dye release via other non-photochemical pathways including ones not triggered by cleavage of the C–S bond in the TNB cage such as intermolecular nucleophilic attack at the carbonate or ester bond by water, free amines, and thiols. Involvement of these passive mechanisms might similarly occur in the release of the drug payload. In particular, TNB conjugates 3 and 4 each carry taxol attached through an ester bond which is reportedly labile and cleaved by hydrolysis and lysosomal enzymes.[31a, 37] However, the stability of these constructs over several hours makes them functional over the time frame required for them to reach the target site for photoactivation applications.

In order to enhance our understanding of this non-photochemical background release, we investigated the effect of a model thiol, 3-mercaptopropionic acid (1 mM ± UV) on the release of coumarin using TNB-Coumarin as a model system (Figure S7). When added at a ~5 molar equiv, the thiol lacked the ability to trigger the release of coumarin in the absence of UVA light exposure. In combination with light exposure, the thiol played a clear role in altering the distribution of thioester precursors and accordingly made an indirect contribution to the dye release. This result, though observed in a simple model system, is supportive of the earlier results observed in the cellular system (Figure 7). Lastly, intracellular esterase activity may contribute to the non-photochemical background release as these are well known for their role in the hydrolytic activation and intracellular trapping of cell-permeable flurogenic ester molecules.[38] A full understanding of all of the contributors to non-photochemical release requires the thorough design of various model studies, and constitutes the objective of future studies.

In summary, the extent of dye release was conveniently measured in real-time by fluorescence in situ. The resulting kinetics of dye release suggests that the release occurs via a mechanism triggered by the photolytic C–S cleavage of the TNB cage and subsequently by other non-photochemical mechanisms. Thus measurements made shortly after UV activation represents the extent of drug released as triggered by the light stimulus, while measurements made over an extended incubation time represents release of drug collectively from a combination of the photochemical and other mechanisms.


Major challenges facing the development of controlled drug release systems include the paucity of cage technologies that allow for active control and for gauging in situ how much drug is released from its cage system at a given dose.[11d, 39] We have designed a novel TNB-based, dual arm cage with the unique ability to trigger light-controlled synchronous release of two payloads. Compared to existing bifunctional linkers,[11b, 11c, 11f, 2526, 40] it offers synthetic convenience, an active mechanism for the control of payload release by light exposure, and rapid release kinetics comparable to an ONB cage system.[10b, 13] Its practicality for fluorescence-based monitoring of drug release for in vitro cellular applications was investigated here by its conjugation with taxol or doxorubicin along with a conditionally fluorescent reporter. The conjugates exhibited a strong correlation between drug release and fluorescence measured in situ on a real-time basis. Furthermore, TNB conjugation was also able to maintain the drug compounds in a reduced cytotoxic state in the conjugated form, and the conjugates were taken up intracellularly at doses that were as cytotoxic as the unmodified drug upon UV activation.

We believe that this new strategy will make a broad impact on the advancement of fluorescence-based release control in a wide range of applications ranging from those involving small molecule drug-reporter conjugates to nanoparticulate delivery systems. This strategy offers important tools to the expanding field of photocaging[2, 6b, 9] which has been applied not only for organic molecules but also for metal ions using photolabile metal chelators[41] such as acetal-based ONB[42] which employs a similar fragmentation reaction at its acetal carbon. Finally, we believe that this TNB strategy has strong potential for use as a new platform for photocaging thiols which can be used in the active control of cellular activities that are mediated by bioactive thiols including cysteine among others. These applications range from the control of redox biology,[43] enzymes,[15, 44] and ion channel permeability.[45] As discussed above, the values of Φuncaging which were determined for TNB-caged small thiols including 1, 2 (Table S1) are comparable to ONB caged molecules[6b] and are supportive of the compatibility of such application. Future efforts will focus on extending and validating the scope of this TNB strategy in photocaged small molecules for chemical biology, and in nanoscale systems designed for receptor targeted anticancer therapeutic agents.

Supplementary Material

Supporting Information


This work was supported by the Michigan Nanotechnology Institute for Medicine and Biological Sciences, and in part by the NIH National Cancer Institute 1R21CA191428, and the British Council and Department for Business Innovation & Skills through the Global Innovation Initiative.


Experimental Section

Details for compound synthesis, characterization, release kinetics and cell study are described in the Supporting Information.


1. Walker JW, McCray JA, Hess GP. Biochemistry. 1986;25:1799–1805. [PubMed]
2. Mayer G, Heckel A. Angew Chem, Int Ed. 2006;45:4900–4921. [PubMed]
3. a) Lemke EA, Summerer D, Geierstanger BH, Brittain SM, Schultz PG. Nat Chem Biol. 2007;3:769–772. [PubMed]b) Goguen BN, Aemissegger A, Imperiali B. J Am Chem Soc. 2011;133:11038–11041. [PubMed]
4. a) Link KH, Shi Y, Koh JT. J Am Chem Soc. 2005;127:13088–13089. [PubMed]b) Neveu P, Aujard I, Benbrahim C, Le Saux T, Allemand JF, Vriz S, Bensimon D, Jullien L. Angew Chem, Int Ed. 2008;47:3744–3746. [PubMed]c) Lu X, Agasti SS, Vinegoni C, Waterman P, DePinho RA, Weissleder R. Bioconjugate Chem. 2012;23:1945–1951. [PMC free article] [PubMed]
5. a) Momotake A, Lindegger N, Niggli E, Barsotti RJ, Ellis-Davies GCR. Nat Methods. 2006;3:35–40. [PubMed]b) Mahmoodi MM, Abate-Pella D, Pundsack TJ, Palsuledesai CC, Goff PC, Blank DA, Distefano MD. J Am Chem Soc. 2016;138:5848–5859. [PubMed]
6. a) Furuta T, Wang SSH, Dantzker JL, Dore TM, Bybee WJ, Callaway EM, Denk W, Tsien RY. Proc Natl Acad Sci U S A. 1999;96:1193–1200. [PubMed]b) Klán P, Šolomek T, Bochet CG, Blanc A, Givens R, Rubina M, Popik V, Kostikov A, Wirz J. Chem Rev (Washington, DC, U S) 2012;113:119–191.
7. Li YM, Shi J, Cai R, Chen X, Luo ZF, Guo QX. J Photochem Photobiol, A. 2010;211:129–134.
8. Gorka AP, Nani RR, Zhu J, Mackem S, Schnermann MJ. J Am Chem Soc. 2014;136:14153–14159. [PMC free article] [PubMed]
9. Lerch MM, Hansen MJ, van Dam GM, Szymanski W, Feringa BL. Angew Chem, Int Ed. 2016;55:2–24.
10. a) Chien YH, Chou YL, Wang SW, Hung ST, Liau MC, Chao YJ, Su CH, Yeh CS. ACS Nano. 2013;7:8516–8528. [PubMed]b) Choi SK, Thomas T, Li M, Kotlyar A, Desai A, Baker JR., Jr Chem Commun (Cambridge, U K) 2010;46:2632–2634. [PMC free article] [PubMed]c) Wong PT, Chen D, Tang S, Yanik S, Payne M, Mukherjee J, Coulter A, Tang K, Tao K, Sun K, Baker JR, Jr, Choi SK. Small. 2015;11:6078–6090. [PubMed]
11. a) Impellizzeri S, McCaughan B, Callan JF, Raymo FM. J Am Chem Soc. 2012;134:2276–2283. [PubMed]b) Maiti S, Park N, Han JH, Jeon HM, Lee JH, Bhuniya S, Kang C, Kim JS. J Am Chem Soc. 2013;135:4567–4572. [PubMed]c) Yang Z, Lee JH, Jeon HM, Han JH, Park N, He Y, Lee H, Hong KS, Kang C, Kim JS. J Am Chem Soc. 2013;135:11657–11662. [PubMed]d) Lanza GM, Moonen C, Baker JR, Chang E, Cheng Z, Grodzinski P, Ferrara K, Hynynen K, Kelloff G, Lee YEK, Patri AK, Sept D, Schnitzer JE, Wood BJ, Zhang M, Zheng G, Farahani K. Wiley Interdiscip Rev: Nanomed Nanobiotechnol. 2014;6:1–14. [PubMed]e) Agasti SS, Laughney AM, Kohler RH, Weissleder R. Chem Commun (Cambridge, U K) 2013;49:11050–11052. [PMC free article] [PubMed]f) Weinstain R, Segal E, Satchi-Fainaro R, Shabat D. Chem Commun (Cambridge, U K) 2010;46:553–555. [PubMed]
12. Redy-Keisar O, Ferber S, Satchi-Fainaro R, Shabat D. ChemMedChem. 2015;10:999–1007. [PubMed]
13. Wong PT, Choi SK. Chem Rev (Washington, DC, U S) 2015;115:3388–3432. [PubMed]
14. Anderson RR, Parrish JA. J Invest Dermatol. 1981;77:13–19. [PubMed]
15. a) Abate-Pella D, Zeliadt NA, Ochocki JD, Warmka JK, Dore TM, Blank DA, Wattenberg EV, Distefano MD. ChemBioChem. 2012;13:1009–1016. [PubMed]b) Nguyen DP, Mahesh M, Elsässer SJ, Hancock SM, Uttamapinant C, Chin JW. J Am Chem Soc. 2014;136:2240–2243. [PubMed]
16. a) Inlay MA, Choe V, Bharathi S, Fernhoff NB, Baker JR, Weissman IL, Choi SK. Chem Commun (Cambridge, U K) 2013;49:4971–4973. [PMC free article] [PubMed]b) Faal T, Wong P, Tang S, Coulter A, Chen Y, Tu CH, Baker JR, Choi SK, Inlay MA. Mol BioSyst. 2015;11:783–790. [PubMed]c) Sinha DK, Neveu P, Gagey N, Aujard I, Saux TL, Rampon C, Gauron C, Kawakami K, Leucht C, Bally-Cuif L, Volovitch M, Bensimon D, Jullien L, Vriz S. Zebrafish. 2010;7:199–204. [PubMed]
17. Selbo PK, Weyergang A, Høgset A, Norum O-J, Berstad MB, Vikdal M, Berg K. J Controlled Release. 2010;148:2–12. [PubMed]
18. a) Choi SK, Thomas TP, Li MH, Desai A, Kotlyar A, Baker JR. Photochem Photobiol Sci. 2012;11:653–660. [PubMed]b) Choi SK, Verma M, Silpe J, Moody RE, Tang K, Hanson JJ, Baker JR., Jr Bioorg Med Chem. 2012;20:1281–1290. [PubMed]
19. Dcona MM, Yu Q, Capobianco JA, Hartman MCT. Chem Commun (Cambridge, U K) 2015;51:8477–8479. [PMC free article] [PubMed]
20. a) Noguchi M, Skwarczynski M, Prakash H, Hirota S, Kimura T, Hayashi Y, Kiso Y. Bioorg Med Chem. 2008;16:5389–5397. [PubMed]b) Skwarczynski M, Noguchi M, Hirota S, Sohma Y, Kimura T, Hayashi Y, Kiso Y. Bioorg Med Chem Lett. 2006;16:4492–4496. [PubMed]
21. Agasti SS, Chompoosor A, You C-C, Ghosh P, Kim CK, Rotello VM. J Am Chem Soc. 2009;131:5728–5729. [PMC free article] [PubMed]
22. Wong P, Tang S, Mukherjee J, Tang K, Gam K, Isham D, Murat C, Sun R, Baker JR, Choi SK. Chem Commun (Cambridge, U K) 2016;52:10357–10360. [PMC free article] [PubMed]
23. a) Lin Q, Bao C, Yang Y, Liang Q, Zhang D, Cheng S, Zhu L. Adv Mater (Weinheim, Ger) 2013;25:1981–1986.b) Yang Y, Shao Q, Deng R, Wang C, Teng X, Cheng K, Cheng Z, Huang L, Liu Z, Liu X, Xing B. Angew Chem, Int Ed. 2012;51:3125–3129. [PubMed]
24. a) Karthik S, Prashanth Kumar BN, Gangopadhyay M, Mandal M, Singh NDP. J Mater Chem B. 2015;3:728–732.b) Yamazoe S, Liu Q, McQuade LE, Deiters A, Chen JK. Angew Chem, Int Ed. 2014;53:10114–10118. [PMC free article] [PubMed]
25. Santra S, Kaittanis C, Santiesteban OJ, Perez JM. J Am Chem Soc. 2011;133:16680–16688. [PMC free article] [PubMed]
26. Yuan Y, Zhang C-J, Gao M, Zhang R, Tang BZ, Liu B. Angew Chem, Int Ed. 2015;54:1780–1786. [PubMed]
27. Nicolaou KC, Mathison CJN, Montagnon T. Angew Chem, Int Ed. 2003;42:4077–4082. [PubMed]
28. Katritzky AR, Xu Y-J, Vakulenko AV, Wilcox AL, Bley KR. J Org Chem. 2003;68:9100–9104. [PubMed]
29. Dawson PE, Churchill MJ, Ghadiri MR, Kent SBH. J Am Chem Soc. 1997;119:4325–4329.
30. Hatchard CG, Parker CA. Proc R Soc London, Ser A. 1956;235:518–536.
31. a) Ojima I. Acc Chem Res. 2008;41:108–119. [PubMed]b) Labruère R, Alouane A, Le Saux T, Aujard I, Pelupessy P, Gautier A, Dubruille S, Schmidt F, Jullien L. Angew Chem, Int Ed. 2012;51:9344–9347. [PubMed]
32. Hori Y, Norinobu T, Sato M, Arita K, Shirakawa M, Kikuchi K. J Am Chem Soc. 2013;135:12360–12365. [PubMed]
33. Ojima I, Geng X, Wu X, Qu C, Borella CP, Xie H, Wilhelm SD, Leece BA, Bartle LM, Goldmacher VS, Chari RVJ. J Med Chem. 2002;45:5620–5623. [PubMed]
34. a) Vlahov IR, Vite GD, Kleindl PJ, Wang Y, Santhapuram HKR, You F, Howard SJ, Kim SH, Lee FFY, Leamon CP. Bioorg Med Chem Lett. 2010;20:4578–4581. [PubMed]b) Kock A, Zuwala K, Smith AAA, Ruiz-Sanchis P, Wohl BM, Tolstrup M, Zelikin AN. Chem Commun (Cambridge, U K) 2014;50:14498–14500. [PubMed]
35. Lavis LD, Rutkoski TJ, Raines RT. Anal Chem (Washington, DC, U S) 2007;79:6775–6782. [PMC free article] [PubMed]
36. Staleva H, Komenda J, Shukla MK, Šlouf V, Kaňa R, Polívka T, Sobotka R. Nat Chem Biol. 2015;11:287–291. [PubMed]
37. Ndungu JM, Lu YJ, Zhu S, Yang C, Wang X, Chen G, Shin DM, Snyder JP, Shoji M, Sun A. J Med Chem. 2010;53:3127–3132. [PMC free article] [PubMed]
38. Laurent A, Debart F, Lamb N, Rayner B. Bioconjugate Chem. 1997;8:856–861. [PubMed]
39. Lammers T, Aime S, Hennink WE, Storm G, Kiessling F. Acc Chem Res. 2011;44:1029–1038. [PubMed]
40. Bhuniya S, Maiti S, Kim E-J, Lee H, Sessler JL, Hong KS, Kim JS. Angew Chem, Int Ed. 2014;53:4469–4474. [PubMed]
41. a) Ellis-Davies GCR. Chem Rev (Washington, DC, U S) 2008;108:1603–1613. [PubMed]b) a) Adams SR, Tsien RY. Annu Rev Physiol. 1993;55:755–784. [PubMed]
42. Ellis-Davies GCR, Kaplan JH. J Org Chem. 1988;53:1966–1969.
43. Poole LB. Free Radical Biol Med. 2015;80:148–157. [PMC free article] [PubMed]
44. Qi JH, Zhang L, Wang J, Lu M, Wang XM, Jin ZJ. Cell Res. 1996;6:47–53.
45. Alansary D, Schmidt B, Dörr K, Bogeski I, Rieger H, Kless A, Niemeyer BA. Sci Rep. 2016;6:33347. [PMC free article] [PubMed]