Strategies toward Organic Carbon Monoxide ... - ACS Publications

Jan 11, 2018 - Department of Chemistry and Center for Diagnostics and Therapeutics, Georgia State University, Atlanta, Georgia 30303 United. States...
8 downloads 0 Views 2MB Size
Article Cite This: Acc. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/accounts

Strategies toward Organic Carbon Monoxide Prodrugs Xingyue Ji and Binghe Wang* Department of Chemistry and Center for Diagnostics and Therapeutics, Georgia State University, Atlanta, Georgia 30303 United States CONSPECTUS: Carbon monoxide is widely acknowledged as an important gasotransmitter in the mammalian system with importance on par with that of nitric oxide. It has also been firmly established as a potential therapeutic agent with a wide range of indications including organ transplantation, cancer, bacterial infection, and inflammation-related conditions such as colitis and sepsis. One major issue in developing CO based therapeutics is its delivery in a pharmaceutically acceptable form. Currently, there are generally five forms of deliveries: inhaled CO, photosensitive CO-releasing molecules, encapsulated CO, CO dissolved in drinks, and molecules that would release CO under physiological conditions without the need for light. For over a decade, the last category only included metal-based CO releasing molecules. What had been missing were organic CO prodrugs, which release CO under physiological conditions with tunable rates and in response to various exogenous and endogenous triggers such as water, chemical reagents, esterase, ROS, and changes in pH. This Account describes our work in this area as well as the demonstration for these organic prodrugs to recapitulate CO’s pharmacological effects both in vitro and in vivo. Generally, two categories of CO prodrugs have been developed in our lab. Both can be considered as precursors of norbornadien-7-ones, which readily undergo cheletropic reaction under very mild conditions to extrude CO. The first category of CO prodrugs capitalizes on the inter- and intramolecular inverse electron demand Diels−Alder (DAinv) reaction to trigger CO release under physiological conditions. As for the bimolecular CO prodrugs, we proposed a new concept of “enrichment triggered CO release” by conjugating both components with a mitochondria-targeting moiety to achieve targeted CO delivery with improved biological outcomes in vitro and in vivo. As for the unimolecular CO prodrugs, the release half-lives can be readily tuned from minutes to days by varying the substituents on the dienone ring, the tethering linker, and the alkyne. Some significant structure−release rates relationships (SRRs) have been unveiled. An esterase-activated CO prodrug and a cascade prodrug system for co-delivery of CO and another payload have also been devised using such an intramolecular click and release strategy. The second category of CO prodrugs leverage on an elimination reaction to generate norbornadien-7-ones for CO release from norborn-2-en-7-ones. In the case of pH-sensitive ones, the CO release is triggered by β-elimination, and the release rate can be quantitatively predicted using the Hammett constant of the substituents on the leaving group. The ROS-activated ones take advantage of ROS-induced selenoxide elimination to achieve targeted CO delivery to disease sites with elevated ROS level. We strongly believe that these CO prodrugs could serve as powerful tools for CO-associated biological studies and are promising candidates for ultimate clinical applications.



INTRODUCTION CO is generated endogenously via the catabolism of heme by heme oxygenases (HO-1 and HO-2)1 and is considered a signaling molecule in mammals, providing pleiotropic pharmacological effects including cytoprotective,2 antibacterial,3 antiinflammatory,4 and anticancer effects5,6 at low and safe doses. CO largely exists in the hemoglobin-bound form, COHb, in vivo. The US Food and Drug Administration (FDA) sets 14% of serum COHb as the upper limit for clinical trials, and several clinical trials have demonstrated the safety of inhaled CO at low doses (250 ppm, 3 h/day).7,8 However, widespread clinical applications of inhaled CO face several challenges, including difficulty in controlling dosage, lack of portability, and dependence on patients’ respiratory state to deliver precise amounts.8 In addition, safety issues due to possible malfunction of a CO delivery device are nontrivial. To mitigate all these limitations, elegantly designed CO releasing molecules (CO© XXXX American Chemical Society

RMs) as donors have been reported as safer and more controllable ways to deliver CO for preclinical studies.2,9−13 These CO-RMs are largely transition metal carbonyl complexes and have greatly facilitated research about CO’s physiological, pathological, and pharmacological effects. Recently, exciting advances have been made in developing encapsulated CO-RMs, metal−organic framework (MOF) based CO-RMs, and CO dissolved in a liquid.14−20 Additionally, several photosensitive metal-free CO-RMs have also been reported.21−24 These photo-CO-RMs offers temporal controls in CO release and serve as very good research tools for in vitro studies. What had been missing among all the earlier work was a class of CO donors that are metal-free and do not require light for CO release under physiological conditions. In this Account, we Received: January 11, 2018

A

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research summarize exciting work in the last four years in the development of organic CO prodrugs that (1) release CO under physiological conditions without the need for light, (2) have tunable and predictable release rates, (3) offer controlled release with various chemical or biological triggers (such as esterase, pH, or reactive oxygen species), (4) offer the possibility for organelle (mitochondria) or biomarker-based targeting, (5) generate a fluorescent side product after CO release for easy monitoring of the release process, (6) recapitulate CO-associated biological effects both in vitro and in vivo, and (7) are amenable to structural optimization and medicinal chemistry work for further development. A focus of this Account is to describe the design strategy to stimulate future activity in this area. Because of the vast experience of the pharmaceutical industry in developing small organic molecules, we hope that these organic CO prodrugs would help advance CO-based therapeutics. Developing bioreversible prodrug derivatization methods for CO represents very unique challenges because of the lack of an easily bioreversible “handle” as in the cases of amines, alcohols, and carboxylic acids. Earlier work on metal-containing CO-RMs took advantage of the reversible complexation of CO to transition metals. However, for the design of organic CO prodrugs, likely the release would involve the cleavage of a carbon−carbon bond, which is much harder under physiological conditions. The photosensitive metal-free CO-RMs rely on the ability of photons to cleave C−C bonds in certain structures. For prodrugs that would release CO under physiological conditions, we thought about using pericyclic reactions, which are well-known for the ability to easily break and form C−C bond under mild conditions.25 Specifically, our design strategy relies on the ability of norbornadien-7-ones to undergo facile cheletropic reaction to extrude CO under mild conditions.26 Therefore, precursors of norbornadien-7-ones could serve as stable CO prodrugs. Two major strategies were employed in our design to make precursors of norbornadien-7ones, namely, an inter- and intramolecular “click and release” strategy and a β-elimination-based strategy. In the following sections, these two strategies are discussed in detail.

Figure 1. A bimolecular CO prodrug strategy.

have been several fluorescent probes developed for tracking CO in vitro and even in vivo.30,31 We also used one of those (COP1) for studying intracellular CO release from our other CO prodrugs (see below). The second order rate constant can be tuned from 0.11 to 12.2 M−1 s−1 by using different strained alkynes or dienones, indicating the tunability, albeit indirect, of the CO release rate.32 Tethering a hydrophilic mannose moiety also allowed for the minimization of membrane permeability, resulting in compounds 1b and 2b with no cytotoxicity toward Raw 264.7 cells at 1 mM. These prodrugs also decreased the lipopolysaccharide (LPS)-induced tumor necrosis factor (TNF) secretion in Raw 264.7 cells, a strong indication of their antiinflammatory effects.33 The demonstration of the initial chemical feasibility initiated two subsequent directions in terms of bringing the organic CO prodrugs closer to biological applications: (1) enrichmenttriggered release through the use of a bimolecular system and (2) bringing the diene and dienophile into a single molecule to create unimolecular CO prodrug systems, which are discussed in the following sections. Generally, bimolecular prodrugs have limited utility in therapeutic applications due to the difficulties in synchronizing the pharmacokinetic profiles of the two reagents for in vivo studies. However, one could actually take advantage of bimolecular CO release (Figures 1 and 2) to address a very difficult problem in targeted prodrug delivery and activation by using enrichment-triggered release.34 In doing so, one could dose the compound at such a low concentration that would not allow the reaction to happen at a meaningful rate. Upon enrichment at a desired site, the bimolecular reaction would significantly accelerate, leading to CO release. Such an approach would require the tethering of a targeting moiety to allow for enrichment. It has been reported that CO’s site of action is largely in the mitochondria.35 Thus, we tethered the two reaction components in Figure 1 with triphenylphosphonium (TPP), which targets mitochondria. We reasoned that at low concentrations in solution (e.g., 10 μM), the click reaction with a rate constant of ∼0.2 M−1 s−1 would happen at an extremely slow rate (t1/2 > 130 h), which ensures minimal CO release. It is well-known that TPP conjugation could lead to mitochondria enrichment by up to 1000-fold.36 If we consider a modest enrichment to about 500 μM, then the release half-life becomes ∼1.9 h, which is well within the release rates that allow pharmacological effects when dosed in the micromolar ranges. Thus, compounds 1d and 2c each with a TPP moiety (Figure 2) were synthesized.34 Additionally, the product after CO release is fluorescent, making it easy to monitor CO release. Subsequent imaging experiments showed that cells cotreated with compounds 1c and 2a without TPP conjugation



CO PRODRUGS BASED ON A “CLICK AND RELEASE” STRATEGY Norbornadien-7-ones can be generated by an inverse electrondemand Diels−Alder reaction (DAinv) between an alkyne and a cyclopentadienone. However, such reactions normally require high temperatures.27 Because the reaction rate constant of such reactions is dependent on the energy gap between the HOMO of the alkyne and the LUMO of the dienone,28 we reasoned that such reactions could be bioorthogonally performed under physiological conditions for CO release by increasing the HOMO energy level of the alkyne or decreasing the LUMO energy level of the dienone. With this in mind, we tested the reaction between teteraphenylcyclopentadienone 1a and bicyclo[6.1.0]non-4-yn-9-ylmethanol (BCN) 2a (Figure 1), a strained alkyne with high HOMO energy level.28 As expected, the reaction went smoothly in MeOH at room temperature with a second order reaction rate constant of 0.6 M−1 s−1. The CO release was confirmed both by a CO−myoglobin assay and a commercially available carbon monoxide detector. Sodium dithionite used in the CO−myoglobin assay is known to trigger CO release from some metal-based CO-RMs.29 However, it did not induce CO release from compound 1a in the absence of 2 because the chemistry is completely different. Recently, there B

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Figure 2. Enrichment triggered CO release within mitochondria.

Figure 3. Fluorescent images of RAW 264.7 cells cotreated with CO prodrugs. (a−d) 1c + 2 (a, 1 μM; b, 5 μM); 2c + 1d (c, 1 μM; d, 5 μM). (e−h) Compound 2c (5 μM), 1d (2.5 μM), and MT-deep red (50 nM): (e) bright field; (f) red channel; (g) DAPI channel; (h) Merged image of images e, f, and g.

did not present any blue fluorescence at low concentrations (1 and 5 μM). In contrast, strong blue fluorescence was observed dose-dependently in cells treated with 1d and 2c with TPP conjugation. With various control experiments, the significantly enhanced fluorescence intensity in cells treated with 1d and 2c is safely ascribed to TPP-induced enrichment in the mitochondria and the subsequent increase in reaction rate. Additional colocalization experiments using a fluorescent mitochondrion-tracker (MT-deep red) further confirmed that the click reaction happened primarily in mitochondria (Figure 3). In addition, cotreatment of 1d and 2c dose-dependently inhibited LPS-induced TNF secretion with an IC50 value of only 5 μM, while no such effects were observed for 1c and 2a even at a higher concentration (10 μM). It is worth noting that 5 μM is the lowest IC50 value ever observed with all the other organic CO prodrugs in our lab under the same conditions, indicating the benefits of targeted delivery to mitochondria. Further, we tested these compounds in an APAP (acetaminophen) induced liver injury model. Indeed, cotreatment with 1d and 2c (0.4 mg/kg) significantly suppressed APAP-induced liver injury as indicated by a 50% reduction in alanine transaminase (ALT) level as compared to the vehicle (DMSO) group. Meanwhile no positive results were observed under treatment with 1c and 2a at the same dosage. The coinjection of 1d and 2c (4 mg/kg) also resulted in a sustained COHb level of 4% from 20 min to 1 h postinjection (baseline: 2%). Altogether, these results firmly established the proof-ofconcept of using enrichment to trigger CO release for targeted delivery to mitochondria. It can be envisioned that by using

different targeting moieties, CO delivery can be targeted to various tissues or organs or disease sites. Although targeted CO delivery is preferable and has been achieved,37 most of the CO-associated indications are obtained without the need for targeted CO delivery, because CO is very permeant, and accessing a target by CO is not an issue, even without targeting.8 Further, CO has a comparable or higher safety margin when compared with nutrients and drugs including potassium, glucose, and insulin.8 Therefore, there is nothing special about CO in terms of the need to target. Much the same as traditional pharmaceuticals such as acetaminophen, aspirin, ibuprofen, doxorubicin, and remicade, targeted delivery is desired but not essential for application.38 The bimolecular approach fully demonstrated the chemical feasibility of using DAinv to decage CO from dienones. However, the unique challenge of controlling the pharmacokinetics of two components and thus applying such a strategy to in vivo studies means that it is only applicable in unique situations. Therefore, we have also designed a unimolecular system in Figure 4.39,40 There are several issues to consider in such a design. First, it is critical that the prodrug is stable during synthesis and storage and yet readily releases CO under physiological conditions. Fortuitously, Breslow and others had already demonstrated that water and glycoproteins significantly accelerate Diels−Alder reactions by several thousand fold, presumably through hydrophobically driving the diene and dienophile closer.41,42 Such findings form the theoretical foundation of the observed stability difference of the CO prodrugs in water and in organic C

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

dimethyl group (R5 = R6 = Me in scaffold I) on the tether linker significantly accelerates the cycloaddition rate (>1000-fold). It should be noted that replacement of the phenyl ring at the R1 position (scaffold I) with a methylthiol group does not affect the release rate, indicating tolerability of substituent’s electronic properties. Besides, other factors also allow for tuning of release rate. For example, an internal alkyne (R4 = Me in scaffold I) is less reactive than a terminal alkyne, and two separated phenyl rings at position R2 and R3 (scaffold I) substantially accelerate release rate compared to the fused naphthalene substitution. These SRRs can serve as a powerful guide for the design of new CO prodrugs with specific release rates. CO release from such prodrugs was validated both by a CO−myoglobin assay and a reported fluorescent CO probe, COP-1.30 The oral administration of several prodrugs could lead to a serum COHb level as high as 10% for 2 h in mice (unpublished data). In addition, one representative CO prodrug (BW-CO-103, R1 = Ph, R2, R3 = naphthalene, R4−R6 = H, n = 2, X = -N-iso-Pr, scaffold I) was shown to recapitulate CO’s anti-inflammatory effects both in LPS-challenged Raw 264.7 cells and in a trinitrobenzenesulfonic acid (TNBS)-induced colitis mouse model. These results firmly demonstrated the feasibility of using intramolecular DAinv to deliver CO for biological applications. One issue is the need to improve the water solubility of these CO prodrugs because of their somewhat hydrophobic nature and the current need to use solubilizing agent such as Solutol. However, we have developed conjugation method for tethering functional groups for improved water solubility. It should be noted that the issue of “drug-likeness” for a prodrug or drug delivery system is substantially different from that of an active ingredient. This is especially true for the prodrug of a gasotransmitter. CO is highly diffusive and permeant.8,38 Thus, reaching the desired target is not an issue. As a result, the desire is for the prodrug (and the product after CO release) not to be membrane permeable in order to reduce side effects. The issue of solubility, protein-binding, and even metabolism need to be considered in the context of CO release rate, but not in the traditional sense of “activity”, since CO is the only bioactive component. It should also be noted that the issue of CO toxicity has been thoroughly analyzed in a prior publication.8 CO’s safety margin is no less than other pharmaceuticals and nutrients such as insulin, glucose, potassium, and nitric oxide. Thus, there is nothing special

Figure 4. Unimolecular CO prodrugs using intramolecular DAinv for CO release.

solvents (or neat). Second, there is the question of what kind of alkyne to use. It turned out that the entropic advantage of the intramolecular system (Figure 4) is so significant that the reaction does not need an activated alkyne (Figure 1). As a matter of fact, even internal alkynes, which are considered much less reactive than terminal alkynes in intermolecular DAinv,27 are suitable. The third question is the tunability of the CO release rates. In this regard, we also reasoned that the cycloaddition/CO release rate could be easily tuned by varying the substituents on the dienone ring and the tethering linker by controlling entropic factors. Several candidate CO prodrugs with different scaffolds (I and II) were synthesized.41,42 These compounds are stable in the solid state at room temperature for weeks and yet start releasing CO upon dissolution in a mixed aqueous solution. In some examples with R2 and R3 being a naphthalene group, a blue fluorescent reporter compound is also formed upon CO release, allowing for real-time monitoring of CO release. In addition, the CO release half-life can be easily tuned from minutes to days by varying the nature and the substituents on the tethering linkers, and some significant insights into the structure−release rate relationships (SRRs) were obtained. Generally, entropic factors play dominant roles in determining the cycloaddition rate. For example, an amide linker (X = NR in scaffold I) is more entropically favorable compared to an ester linkage (X = O in scaffold I) due to the more rigid nature of the amide bond, 5-membered ring formation (n = 1) is more favorable than 6-membered ring formation (n = 2), and gem-

Figure 5. Stimuli activated CO prodrugs. D

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

use β-elimination as a way to generate the double bond. In doing so, an aldehyde and a leaving group were appended at the C5 and C6 position, respectively (Figure 7, 14a−e).45 Despite

about developing CO as a pharmaceutical agent in terms of adverse effects. To develop prodrugs with added control in CO release, we designed a general approach toward esterase-sensitive organic CO prodrugs.43 The general idea was to introduce conformational constraints to “hold” the alkyne moiety away from the cyclopentadienone moiety to prevent cycloaddition until the constraint is released (Figure 5). We envisioned using an esterase-sensitive cleavable linker to impose the needed conformational constraints. Thus, we synthesized two such CO prodrugs (8a,b). As shown in Figure 5, the alkyne is “held back” from the dienone moiety with a 7-membered lactone ring, which is optimal because it strikes a good balance between the need to use a small ring to impose conformational constraints and a large enough ring to make the ester itself entropically unfavorable and easily cleavable. Indeed, prodrugs 8a and 8b release CO in the presence of porcine liver esterase (PLE) with a half-life of 1 or 4 h, respectively. However, in the absence of PLE, the half-life is over 17 h. In addition, cycloaddition product 11 without lactone hydrolysis was not observed in the course of the experiment, indicating that the 7membered lactone ring indeed restricted the alkyne from reacting with the dienone moiety. In addition, CO release was confirmed by using COP-1 and was shown to suppress TNF production in LPS challenged Raw 264.7 cells. The success of these two examples also suggests that one could use other cleavable linkers (-Y−X-), such as disulfide or diazo bond, for release of CO in response to other triggers for potentially targeted delivery.

Figure 7. Correlation of CO release rate (k) to Hammett constant (σ).

being a superior leaving group, we decided to avoid halide as a leaving group due to the formation of hydrogen halide. Therefore, substituted phenols were employed as leaving groups to balance reactivity and stability, and several potential CO prodrugs 14a−e were thus synthesized. The subsequent CO release studies showed that these compounds are extremely stable in organic solvent and can survive the harsh conditions (e.g., reflux in CH2Cl2) used for their preparation. Additionally, no reaction was found for CO prodrug 14a (R = H) even after 1 week of incubation in CDCl3 at 37 °C, indicating the superb stability of these CO prodrugs for long-term storage. Moreover, these compounds readily underwent β-elimination reaction with the concomitant release of one equivalent of CO in a mixed aqueous solution (pH = 7.4). The half-lives of these CO prodrugs correlated very well with the Hammett constant of the R groups on the phenyl ring (Figure 7), allowing for the prediction of release rates for new CO prodrugs with different R groups. In addition, because β-elimination is base-catalyzed, CO release from these prodrugs is sensitive to the pH of the buffer solution used. One representative CO prodrug 14b (R = F) released CO with a half-life of only 0.65 h in PBS buffer (pH = 7.4). In contrast, the half-life of CO release in simulated gastric fluid (pH = ∼1) is over 9 h. Two representative CO prodrugs (14b and 14c, R = Me) showed intracellular CO release and TNF suppression in LPS-challenged Raw 264.7 cells. Based on a similar strategy, Larsen and co-workers devised several organic CO prodrugs with the leaving group and electron withdrawing group being bromide and amide, respectively (Figure 8).46 Compounds 15a,b release CO in



ORGANIC CO PRODRUGS BASED ON ELIMINATION REACTIONS TO GENERATE NORBORNADIENE-7-ONES All the aforementioned CO prodrugs represent the first category of organic prodrugs and firmly established the feasibility of using norbornadien-7-ones as a surrogate for CO. As is true in any drug discovery efforts, structural diversity is crucial to the eventual success. Therefore, we also worked on developing β-elimination approaches leading to pH- and reactive oxygen species (ROS)-sensitive CO prodrugs. Our design is based on the observation that unlike norbornadien-7-ones 12, norborn-2-en-7-ones 14 are highly stable.44 Therefore, we reasoned that norborn-2-en-7-ones 14 could serve as feasible CO prodrugs, provided that a double bond between the C5 and C6 positions can be easily formed under physiological conditions (Figure 6). First, we decided to

Figure 8. Water-soluble organic CO prodrugs based on β-elimination.

TRIS-sucrose buffer at 37 °C with half-lives of 19 and 75 min, respectively. The CO release from such compounds is also pH sensitive. However, unlike compounds 14a−e, which release CO at relatively slow rate in buffer at low pH (4−6), no CO release was observed for compounds 15a,b in slightly acidic buffer (pH < 7). Importantly, compound 15a was able to

Figure 6. CO prodrugs based on β-elimination. E

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Figure 9. ROS-activated CO prodrugs.

release CO both in MDCK cells and in rats and recapitulated CO-related vasorelaxation effect on precontracted rat aortic rings, demonstrating the feasibility of using such compounds as CO prodrugs for therapeutic applications. It is anticipated that CO prodrugs as such could be employed for localized delivery to the lower gastric intestinal (GI) tract by leveraging the pH changes in the GI tract. Such a feature is highly desirable for delivery to the lower GI and can be used for treatment of inflammation-related bowel diseases, which primarily affect the lower GI tract. It is well-known that certain pathological conditions such as inflammation, cancer, and infection are accompanied by an elevated level of ROS.47,48 Therefore, we also undertook an effort to develop ROS-sensitive CO prodrugs for targeted delivery. Similar to the pH-sensitive CO prodrugs in design principle, we devised two ROS-sensitive CO prodrugs 17a,b with a phenyselenium group appended at the C5 position.49 Without the substitution of an electron withdrawing group at the C6 position, compounds 17a,b are not subjected to βelimination to release CO and, hence, are very stable during synthesis and storage. However, phenylselenium is very sensitive to ROS, leading to the formation of selenium-oxide compounds, which can readily undergo a syn-elimination to form a double bond between the C5 and C6 positions for subsequent CO release (Figure 9). Therefore, CO prodrugs as such can target disease sites with elevated ROS levels. As expected, no CO release was observed from compounds 17a,b in the absence of ROS. However, in the presence of ROS such as hypochlorite, both compounds underwent selenium oxidation and subsequent selenoxide elimination to release CO. The subsequent cell imaging studies with COP-1 showed that compound 17b released CO only in cells with elevated ROS level, such as cancer cells (HeLa cells) and inflammatory cells (LPS-challenged Raw 264.7 cells), and no significant CO release was observed in normal cells (H9C2) or naive Raw 264.7 cells (Figure 10). These results firmly established the ability for compound 17b to target CO delivery to cells with elevated levels of ROS. Moreover, compound 17b sensitized HeLa cells but not the normal cells (H9C2) to Dox treatment, demonstrating the benefits of using an ROS trigger for siteselective prodrug activation. In summary, the second category of CO prodrugs are superior in terms of site-selective activation and controlled delivery of CO. Since the substituents on the norborn-2-en-7one scaffold do not impact the CO release profiles except the ones required for β-elimination, they can be removed to increase atom efficiency and water solubility. Meanwhile, for the pH-sensitive CO prodrugs, different combinations of electron withdrawing groups and leaving groups at C5 and 6 positions should also be probed to avoid the usage of an aldehyde group, which is a nucleophilic center with potential toxicity concerns.

Figure 10. Fluorescent imaging of CO release of 17b in different type of cells (a−h) Raw 264.7 cells; (i−l) H9C2 cells; (m−p) HeLa cells: (a) COP-1 only (1 μM); (c) LPS (1 μg/mL) + COP-1 (1 μM); (e) 17b (50 μM) + COP-1 (1 μM); (g) 17b (50 μM) + COP-1 (1 μM) + LPS (1 μg/mL); (i) COP-1 only (1 μM); (k) COP-1 (1 μM) + 17b (50 μM); (m) COP-1 only (1 μM); (o) COP-1 (1 μM) + 17b (50 μM).



A CASCADE PRODRUG SYSTEM FOR CO-DELIVERY OF CO AND A DRUG PAYLOAD CO has been shown to present sensitization effects when used in combination with antibiotics, anticancer agents, or antiinflammatory agents.5,50,51 Therefore, we devised a general system for the co-delivery of CO and another payload for potential combinational therapy.52 As shown in Figure 11, the initial intramolecular cycloaddition leads to the release of CO and the formation of a fluorescent compound, 23, which is poised to undergo lactonization to release the appended payload. The formed fluorescent compound 24 can be employed to track the release of CO and the payload intracellularly. As a proof of concept, benzyl alcohol or metronidazole was appended to such a system (21a,b). As expected, CO along with benzyl alcohol or metronidazole were successfully released upon dissolution in DMSO/PBS at 37 °C. The CO release from prodrug 21b was also studied in Raw 264.7 cells by leveraging the fluorescent nature of the byproduct 24 and COP-1. Additionally, compound 21b showed much enhanced antibacterial effect against Helicobacter pylori with a MIC90 value of 0.16 μg/mL, as compared to 2.5 μg/mL for metronidazole, indicating the beneficial effects of combination treatment.



PERSPECTIVE AND OUTLOOK All the examples described demonstrate the potential and advantage of formulating CO “in a pill” or an injection form, as F

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Figure 11. Cascade prodrug system for the co-delivery of CO and another payload.



compared to inhaled form. However, currently available CO donors still have features that need improvement, and additional medicinal chemistry work is still needed to finetune the developability of these compounds. Some of the remaining issues include solubility, atom efficiency, and the presence of an aldehyde group. Fortunately, results so far suggest that the core scaffold can tolerate extensive modifications to address these issues. It will also be important to examine the toxicity issues of the side product after CO release for eventual applications as well as the pharmacokinetic issues with these prodrugs. With the rapidly increasing interest in CO as a gasotransmitter, we hope that the collective effort by many in the field will lead to a new category of CO-based therapeutics in clinical trials.



(1) Ryter, S. W.; Alam, J.; Choi, A. M. Heme oxygenase-1/carbon monoxide: from basic science to therapeutic applications. Physiol. Rev. 2006, 86, 583−650. (2) Motterlini, R.; Otterbein, L. E. The therapeutic potential of carbon monoxide. Nat. Rev. Drug Discovery 2010, 9, 728−743. (3) Davidge, K. S.; Sanguinetti, G.; Yee, C. H.; Cox, A. G.; McLeod, C. W.; Monk, C. E.; Mann, B. E.; Motterlini, R.; Poole, R. K. Carbon monoxide-releasing antibacterial molecules target respiration and global transcriptional regulators. J. Biol. Chem. 2009, 284, 4516−4524. (4) Otterbein, L. E.; Bach, F. H.; Alam, J.; Soares, M.; Tao Lu, H.; Wysk, M.; Davis, R. J.; Flavell, R. A.; Choi, A. M. Carbon monoxide has anti-inflammatory effects involving the mitogen-activated protein kinase pathway. Nat. Med. 2000, 6, 422−428. (5) Wegiel, B.; Gallo, D.; Csizmadia, E.; Harris, C.; Belcher, J.; Vercellotti, G. M.; Penacho, N.; Seth, P.; Sukhatme, V.; Ahmed, A.; Pandolfi, P. P.; Helczynski, L.; Bjartell, A.; Persson, J. L.; Otterbein, L. E. Carbon monoxide expedites metabolic exhaustion to inhibit tumor growth. Cancer Res. 2013, 73, 7009−7021. (6) Kawahara, B.; Moller, T.; Hu-Moore, K.; Carrington, S.; Faull, K. F.; Sen, S.; Mascharak, P. K. Attenuation of Antioxidant Capacity in Human Breast Cancer Cells by Carbon Monoxide through Inhibition of Cystathionine β-Synthase Activity: Implications in Chemotherapeutic Drug Sensitivity. J. Med. Chem. 2017, 60, 8000−8010. (7) Mayr, F. B.; Spiel, A.; Leitner, J.; Marsik, C.; Germann, P.; Ullrich, R.; Wagner, O.; Jilma, B. Effects of carbon monoxide inhalation during experimental endotoxemia in humans. Am. J. Respir. Crit. Care Med. 2005, 171, 354−360. (8) Ji, X.; Damera, K.; Zheng, Y.; Yu, B.; Otterbein, L. E.; Wang, B. Toward carbon monoxide-based therapeutics: critical drug delivery and developability issues. J. Pharm. Sci. 2016, 105, 406−416. (9) Chakraborty, I.; Carrington, S. J.; Mascharak, P. K. Design strategies to improve the sensitivity of photoactive metal carbonyl complexes (photoCORMs) to visible light and their potential as COdonors to biological targets. Acc. Chem. Res. 2014, 47, 2603−2611. (10) Heinemann, S. H.; Hoshi, T.; Westerhausen, M.; Schiller, A. Carbon monoxide–physiology, detection and controlled release. Chem. Commun. 2014, 50, 3644−3660. (11) Romao, C. C.; Blattler, W. A.; Seixas, J. D.; Bernardes, G. J. Developing drug molecules for therapy with carbon monoxide. Chem. Soc. Rev. 2012, 41, 3571−3583. (12) Schatzschneider, U. Novel lead structures and activation mechanisms for CO-releasing molecules (CORMs). Br. J. Pharmacol. 2015, 172, 1638−1650. (13) Garcia-Gallego, S.; Bernardes, G. J. Carbon-monoxide-releasing molecules for the delivery of therapeutic CO in vivo. Angew. Chem., Int. Ed. 2014, 53, 9712−9721. (14) Nguyen, D.; Boyer, C. Macromolecular and Inorganic Nanomaterials Scaffolds for Carbon Monoxide Delivery: Recent

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Binghe Wang: 0000-0002-2200-5270 Notes

The authors declare no competing financial interest. Biographies Xingyue Ji received his Ph.D. in 2011 at Institute of Medicinal and Biotechnology, Chinese Academy of Medical Sciences and Peking Union Medical College. Thereafter, he took a position at Institute of Medicinal and Biotechnology as an assistant professor. In 2013, he joined Dr. Binghe Wang’s lab as a postdoctoral fellow at Georgia State University. Binghe Wang is Regents’ Professor of Chemistry, Georgia Research Alliance Eminent Scholar in Drug Discovery, Georgia Cancer Coalition Distinguished Cancer Scholar, and Associate Dean for Research, Innovation, and Graduate Studies in the College of Arts and Sciences at Georgia State University. His research interests are in the areas of drug discovery, drug delivery, and new diagnostics.



REFERENCES

ACKNOWLEDGMENTS

We express our gratitude to the Georgia Research Alliance, Georgia State University, through various graduate fellowships and to the NIH for their general support of the work in our lab. G

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Developments and Future Trends. ACS Biomater. Sci. Eng. 2015, 1, 895−913. (15) Diring, S.; Carne-Sanchez, A.; Zhang, J.; Ikemura, S.; Kim, C.; Inaba, H.; Kitagawa, S.; Furukawa, S. Light responsive metal-organic frameworks as controllable CO-releasing cell culture substrates. Chem. Sci. 2017, 8, 2381−2386. (16) Carmona, F. J.; Rojas, S.; Romao, C. C.; Navarro, J. A. R.; Barea, E.; Maldonado, C. R. One-pot preparation of a novel CO-releasing material based on a CO-releasing molecule@metal-organic framework system. Chem. Commun. 2017, 53, 6581−6584. (17) Inaba, H.; Fujita, K.; Ueno, T. Design of biomaterials for intracellular delivery of carbon monoxide. Biomater. Sci. 2015, 3, 1423−1438. (18) Gomperts, E.; Belcher, J. D.; Otterbein, L. E.; Coates, T. D.; Wood, J.; Skolnick, B. E.; Levy, H.; Vercellotti, G. M. The role of carbon monoxide and heme oxygenase in the prevention of sickle cell disease vaso-occlusive crises. Am. J. Hematol. 2017, 92, 569−582. (19) Steiger, C.; Uchiyama, K.; Takagi, T.; Mizushima, K.; Higashimura, Y.; Gutmann, M.; Hermann, C.; Botov, S.; Schmalz, H. G.; Naito, Y.; Meinel, L. Prevention of colitis by controlled oral drug delivery of carbon monoxide. J. Controlled Release 2016, 239, 128−136. (20) America’s Seed Fund Home Page. https://www.sbir.gov/ sbirsearch/detail/1199003 (accessed Feb. 14, 2018). (21) Palao, E.; Slanina, T.; Muchova, L.; Solomek, T.; Vitek, L.; Klan, P. Transition-metal-free CO-releasing BODIPY derivatives activatable by visible to NIR light as promising bioactive molecules. J. Am. Chem. Soc. 2016, 138, 126−133. (22) Soboleva, T.; Esquer, H. J.; Benninghoff, A. D.; Berreau, L. M. Sense and Release: A Thiol-Responsive Flavonol-Based Photonically Driven Carbon Monoxide-Releasing Molecule That Operates via a Multiple-Input AND Logic Gate. J. Am. Chem. Soc. 2017, 139, 9435− 9438. (23) Antony, L. A.; Slanina, T.; Sebej, P.; Solomek, T.; Klan, P. Fluorescein analogue xanthene-9-carboxylic acid: a transition-metalfree CO releasing molecule activated by green light. Org. Lett. 2013, 15, 4552−4555. (24) Peng, P.; Wang, C.; Shi, Z.; Johns, V. K.; Ma, L.; Oyer, J.; Copik, A.; Igarashi, R.; Liao, Y. Visible-light activatable organic CO-releasing molecules (PhotoCORMs) that simultaneously generate fluorophores. Org. Biomol. Chem. 2013, 11, 6671−6674. (25) Tantillo, D. J. Reaction mechanisms Part (ii) Pericyclic reactions. Annu. Rep. Prog. Chem., Sect. B: Org. Chem. 2006, 102, 269−289. (26) Zhang, J.; Ho, D. M.; Pascal, R. A. Synthesis of an Extremely Crowded Naphthalene via a Stable Norbornadienone. J. Am. Chem. Soc. 2001, 123, 10919−10926. (27) Fujihara, T.; Tomike, Y.; Ohtake, T.; Terao, J.; Tsuji, Y. Ruthenium-catalyzed ring-closing metathesis accelerated by long-range steric effect. Chem. Commun. 2011, 47, 9699−9701. (28) Chen, W.; Wang, D.; Dai, C.; Hamelberg, D.; Wang, B. Clicking 1,2,4,5-tetrazine and cyclooctynes with tunable reaction rates. Chem. Commun. 2012, 48, 1736−1738. (29) McLean, S.; Mann, B. E.; Poole, R. K. Sulfite species enhance carbon monoxide release from CO-releasing molecules: implications for the deoxymyoglobin assay of activity. Anal. Biochem. 2012, 427, 36−40. (30) Michel, B. W.; Lippert, A. R.; Chang, C. J. A Reaction-Based Fluorescent Probe for Selective Imaging of Carbon Monoxide in Living Cells Using a Palladium-Mediated Carbonylation. J. Am. Chem. Soc. 2012, 134, 15668−15671. (31) Liu, K.; Kong, X.; Ma, Y.; Lin, W. Rational Design of a Robust Fluorescent Probe for the Detection of Endogenous Carbon Monoxide in Living Zebrafish Embryos and Mouse Tissue. Angew. Chem., Int. Ed. 2017, 56, 13489−13492. (32) Ji, X.; Ji, K.; Chittavong, V.; Aghoghovbia, R. E.; Zhu, M.; Wang, B. Click and Fluoresce: A Bioorthogonally Activated Smart Probe for Wash-Free Fluorescent Labeling of Biomolecules. J. Org. Chem. 2017, 82, 1471−1476.

(33) Wang, D.; Viennois, E.; Ji, K.; Damera, K.; Draganov, A.; Zheng, Y.; Dai, C.; Merlin, D.; Wang, B. A Click-and-Release Approach to CO Prodrugs. Chem. Commun. 2014, 50, 15890−15893. (34) Zheng, Y.; Ji, X.; Yu, B.; Ji, K.; Gallo, D.; Csizmadia, E.; Zhu, M.; Roy Choudhury, M.; De La Cruz, L.; Chittavong, V.; Pan, Z.; Yuan, Z.; Otterbein, L. E.; Wang, B. Enrichment-triggered Prodrug Activation Demonstrated through Mitochondria-targeted Delivery of Doxorubicin and Carbon Monoxide. Nat. Chem. 2018, DOI: 10.1038/s41557018-0055-2. (35) Motterlini, R.; Foresti, R. Biological Signaling by Carbon Monoxide and Carbon Monoxide-Releasing Molecules (CO-RMs). Am. J. Physiol. Cell Physiol. 2017, 312, C302−C313. (36) Starenki, D.; Park, J.-I. Mitochondria-Targeted Nitroxide, MitoCP, Suppresses Medullary Thyroid Carcinoma Cell Survival In Vitro and In Vivo. J. Clin. Endocrinol. Metab. 2013, 98, 1529−1540. (37) Chaves-Ferreira, M.; Albuquerque, I. S.; Matak-Vinkovic, D.; Coelho, A. C.; Carvalho, S. M.; Saraiva, L. M.; Romão, C. C.; Bernardes, G. J. L. Spontaneous CO Release from RuII(CO)2− Protein Complexes in Aqueous Solution, Cells, and Mice. Angew. Chem., Int. Ed. 2015, 54, 1172−1175. (38) Wang, B.; Hu, L.; Siahaan, T.: Drug Delivery: Principles and Applications, 2nd ed.; John Wiley and Sons: Hoboken, NJ, 2016. (39) Ji, X.; Zhou, C.; Ji, K.; Aghoghovbia, R. E.; Pan, Z.; Chittavong, V.; Ke, B.; Wang, B. Click and Release: A Chemical Strategy toward Developing Gasotransmitter Prodrugs by Using an Intramolecular Diels−Alder Reaction. Angew. Chem., Int. Ed. 2016, 55, 15846−15851. (40) Pan, Z.; Chittavong, V.; Li, W.; Zhang, J.; Ji, K.; Zhu, M.; Ji, X.; Wang, B. Organic CO Prodrugs: Structure-CO-Release Rate Relationship Studies. Chem. - Eur. J. 2017, 23, 9838−9845. (41) Breslow, R. Hydrophobic Effects on Simple Organic Reactions in Water. Acc. Chem. Res. 1991, 24, 159−164. (42) Palomo, J. M. Diels−Alder Cycloaddition in Protein Chemistry. Eur. J. Org. Chem. 2010, 2010, 6303−6314. (43) Ji, X.; Ji, K.; Chittavong, V.; Yu, B.; Pan, Z.; Wang, B. An Esterase-Activated Click and Release Approach to Metal-Free COProdrugs. Chem. Commun. 2017, 53, 8296−8299. (44) Eguchi, S.; Ishiura, K.; Noda, T.; Sasaki, T. Synthesis and C y c l o a d d i t i o n R e a c t i o ns o f H o m o a d a m a n t a no [ 4 , 5 - c ] cyclopentadienones. A Facile Route to [4,5]-Fused homoadamantanobenzene Derivatives. J. Org. Chem. 1987, 52, 496−500. (45) Ji, X.; De La Cruz, L. K. C.; Pan, Z.; Chittavong, V.; Wang, B. pH-Sensitive Metal-Free Carbon Monoxide Prodrugs with Tunable and Predictable Release Rates. Chem. Commun. 2017, 53, 9628−9631. (46) Kueh, J. T. B.; Stanley, N. J.; Hewitt, R. J.; Woods, L. M.; Larsen, L.; Harrison, J. C.; Rennison, D.; Brimble, M. A.; Sammut, I. A.; Larsen, D. S. Norborn-2-en-7-ones as Physiologically-Triggered Carbon Monoxide-Releasing Prodrugs. Chem. Sci. 2017, 8, 5454− 5459. (47) Bhattacharyya, A.; Chattopadhyay, R.; Mitra, S.; Crowe, S. E. Oxidative Stress: An Essential Factor in the Pathogenesis of Gastrointestinal Mucosal Diseases. Physiol. Rev. 2014, 94, 329−354. (48) Trachootham, D.; Alexandre, J.; Huang, P. Targeting Cancer Cells by ROS-Mediated Mechanisms: A Radical Therapeutic Approach? Nat. Rev. Drug Discovery 2009, 8, 579−591. (49) Pan, Z.; Zhang, J.; Ji, K.; Chittavong, V.; Ji, X.; Wang, B. Organic CO Prodrugs Activated by Endogenous ROS. Org. Lett. 2018, 20, 8− 11. (50) Tavares, A. F.; Parente, M. R.; Justino, M. C.; Oleastro, M.; Nobre, L. S.; Saraiva, L. M. The Bactericidal Activity of Carbon Monoxide-Releasing Molecules against Helicobacter pylori. PLoS One 2013, 8, e83157. (51) Nikam, A.; Ollivier, A.; Rivard, M.; Wilson, J. L.; Mebarki, K.; Martens, T.; Dubois-Randé, J.-L.; Motterlini, R.; Foresti, R. Diverse Nrf2 Activators Coordinated to Cobalt Carbonyls Induce Heme Oxygenase-1 and Release Carbon Monoxide in Vitro and in Vivo. J. Med. Chem. 2016, 59, 756−762. (52) De La Cruz, L. K. C.; Benoit, S.; Pan, Z.; Yu, B.; Maier, R. J.; Ji, X.; Wang, B. Click, Release, and Fluoresce: A Cascade Prodrug for CoH

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research delivery of Carbon Monoxide, a Drug Payload, and a Fluorescent Reporter. Org. Lett. 2018, 20, 897−900.

I

DOI: 10.1021/acs.accounts.8b00019 Acc. Chem. Res. XXXX, XXX, XXX−XXX