Synthesis and Evaluation of Phosphorodithioate-Based Hydrogen Sulfide Donors

Abstract

A series of O-aryl- and alkyl-substituted phosphorodithioates were designed and synthesized as hydrogen sulfide (H₂S) donors. H₂S released capability of these compounds was evaluated by fluorescence methods. O-aryl substituted donors showed slow and sustained H₂S release while O-alkylated compounds showed very weak H₂S release capability. We also evaluated donors’ protective effects against hydrogen peroxide (H₂O₂)-induced oxidative damage in myocytes and donors’ toxicity toward B16BL6 mouse melanoma cells.

Hydrogen sulfide (H₂S) was known as a toxic pollutant for years. However, since 2000 this gaseous molecule has been classified as an important cell signaling molecule, much like nitric oxide (NO). Literature published in the past few years increasingly suggests that H₂S is a mediator of many physiological and/or pathological processes.^1–5 The production of H₂S in mammalian systems has been attributed to at least three endogenous enzymes:^6–9 cystathionine β-synthase, cystathionine γ-lyase, and 3-mercaptopyruvate sulfur-transferase. These enzymes use cysteine or cysteine derivatives as substrates and convert them into H₂S within different organs and tissues. In addition to these enzymatic pathways, there are also a range of comparably simple chemical events which may liberate H₂S from the intracellular pool of ‘labile’ sulfur, for instance from the ‘sulfane sulfur’ pool (compounds containing sulfur atoms bound only to other sulfur atoms).¹⁰ While the exact mechanisms of action of H₂S are still under investigation, some chemical and biochemical catabolic reactions of H₂S have been disclosed that may be responsible for its biological functions. For example, H₂S reacts readily with methemoglobin to form sulfhemoglobin, which might act as a metabolic sink for H₂S. H₂S is a powerful reducing agent and is likely to be consumed by endogenous oxidant species, such as peroxynitrite, superoxide, and hydrogen peroxide.^11–13 H₂S can also promote protein S-sulfhydration providing a possible mechanism whereby H₂S alters the functions of a wide range of proteins and modulates signaling.^14–16 It is likely that many more important reactions of H₂S are to be discovered. Nevertheless, the production of endogenous H₂S and the exogenous administration of H₂S have been demonstrated to exert protective effects in many pathologies.^1–5 For example, H₂S has been shown to relax vascular smooth muscle, induce vasodilation of isolated blood vessels, and reduce blood pressure. H₂S can also inhibit leukocyte adherence in mesenteric microcirculation during vascular inflammation in rats, suggesting H₂S is a potent anti-inflammatory molecule. Additionally, it has become evident that H₂S is a potent antioxidant and, under chronic conditions, can up-regulate antioxidant defense. These results strongly suggest that modulation of H₂S levels could have potential therapeutic values.

In H₂S research, researchers typically use sulfide salts (NaHS or Na₂S) as H₂S source. The pK_a values for the first and second dissociation steps of H₂S are 7.0 and >12.0, respectively.^17–18 In aqueous state under the physiological pH of 7.4, the ratio of HS⁻/H₂S is ~3:1 as long as sulfide salt solution is prepared. Sulfide salts are therefore considered as short-lasting H₂S donors as they release H₂S quickly. The rapid release of H₂S may cause acute changes in blood pressure and may exert toxic actions. In addition, sulfide concentrations in aqueous solution can rapidly decrease due to volatilization,¹⁹ thus significantly limiting the utility of these chemical precursors.

Due to these considerations, researchers have started to use synthetic H₂S-releasing agents (i.e., H₂S donors) to explore the biological functions of H₂S.^20–22 Currently there are six types of H₂S donors known in literature (Scheme 1): 1) a Lawesson’s reagent derivative named GYY4137;²³ 2) garlic-derived organic polysulfides such as diallyl trisulfide (DATS),²⁴ 3) the dithiolthione moiety²⁵ 4) a series of N-(benzoyl)-thiobenzamide derivatives as thiol-activated H₂S donors,²⁶ 5) S-acylated perthiol based donors,²⁷ 6) amino acid-based thioacids in the presence of bicarbonate buffers.²⁸ Among these donors, GYY4137 is probably the most well-known donor. GYY4137 has a phosphorodithioate core structure and H₂S release from this compound is due to hydrolysis. It is considered as a slow-release donor. GYY4137 has shown some H₂S-relevant biological activities. For example, it relaxes arotas, vasodilates the preconstricted kidney, and exhibits antihypertensive activity in rats. It can also stimulate heart contraction by interaction with endogenous NO generation.²⁹ Although well-applied in biological studies, only one donor (i.e. GYY4137) with fixed H₂S release capability may not fulfill the requirements of different biological applications. We envisioned that the phosphorodithioate template have rich chemistry related to H₂S to be explored. Structure modifications on phosphorodithioate may lead to H₂S release capability change and in turn lead to their biological activity changes. Herein we report the synthesis and evaluation of a series of O-substituted phosphorodithioate-based H₂S donors.

Scheme 1.

Structures of H₂S donors.

In the structure of GYY4137, there is a phenyl-phosphorus linkage via a C-P bond. We decided to replace the C-P bond with O-substitution and to explore H₂S production from the resulted analogs. To this end, a four-step synthesis was developed (Scheme 2). Freshly distilled trichlorophosphine 1 was treated with 1, 2-dithioethane in 3:1 ratio in the absence of solvent and base. The desired product 2-chloro-1,3,2-dithiaphospholane 2 was obtained in almost quantitative yield by simply removing excess 1 upon distillation. Without further purification, compound 2 was treated with aniline followed by sulfurization with elemental sulfur to provide the key intermediate: 2-amine-1,3,2-dithiaphospholan-2-sulfide 4. Finally compound 4 was reacted with different phenols and alcohols in the presence of 1,8-diazobicyclo[5.4.0]undec-7-ene (DBU) to afford a series of phosphorodithioate analogs 5 as DBU salts. In all, five O-substituted compounds 5a–5e were prepared and their structures are shown in Scheme 2.

Scheme 2.

Synthesis of phosphorodithioate-based H₂S donors.

With these compounds in hand, we turned to test their H₂S release capabilities. In comparison GYY4137 was also tested. Previously the measurement of H₂S release from GYY4137 was mainly done by using the standard methylene blue method.^23,30 However strong acidic conditions were involved in the methylene blue method. It is known that the hydrolysis of phosphorodithioates is affected by pH changes and hydrolysis under highly acidic conditions is much faster than under neutral pH.³¹ Therefore methylene blue method is not appropriate for evaluating phosphorodithioate-based donors. Recently a number of fluorescent probes have been developed for H₂S detection^32–34 and these probes are used under neutral pH. We envisioned that such probes could be useful for evaluating H₂S donors. Dansyl azide (DNS-Az), a fluorescent probe developed by the Wang group,³⁵ was selected in this study due to its fast reaction rate and easy of synthesis. In our experiments, 100 μM of each donor was dissolved in a mixed MeCN/phosphate buffer solution (1:1 v/v, pH 7.4) containing DNS-Az (200 μM). The changes of fluorescence emission spectra at 535 nm (λ_ex = 340 nm) were then monitored for three hours at room temperature. Finally the fluorescence signals were converted to H₂S concentrations based on a reference curve obtained with a series of Na₂S standard solutions. Our results are summarized in Figure 1. GYY4137 was first tested as a reference compound. In consistent with previous reports, this donor showed slow and low H₂S release. After 3 h GYY4137 produced approximately 600 nM H₂S, which counted for ~0.6% of initial material. Using the same protocol the newly synthesized donors 5a–5e were also tested. Three compounds 5a–5c showed similar activities as GYY4137. They released H₂S in a slow and sustain fashion and the amounts of H₂S production were also similar to that of GYY4137. Compounds 5d and 5e only released trace amount of H₂S (data not shown). Presumably the O-alkyl substitutions led to increased stability of phosphorodithioates and therefore decrease the tendency of the hydrolysis to generate H₂S.

Figure 1.

Time dependent H₂S releasing profiles of donors 5a–5c and GYY4137. Donors 5d and 5e released trace amounts of H₂S (data not shown).

Having demonstrated H₂S release from phosphorodithioate-based donors in buffers, we wondered whether they could produce H₂S in cells. To address this question, we conducted cell image experiments using H9c2 myocytes. One example using donor 5a is shown in Figure 2. H9c2 cells were incubated with donor 5a under different concentrations (0, 100, and 200 μM) for 24 h. Then a H₂S-specific fluorescent probe, WSP-1,³⁶ was applied into the cells to monitor the production of H₂S. As expected, donor-treated cells (Figure 2B and 2C) shown much enhanced fluorescent signals compared to vehicle treated cells (Figure 2A). These results suggest that phosphorodithioate-based donors indeed can sustainably release H₂S in cells.

Figure 2.

H₂S production of donor 5a in H9c2 cells. Cells were incubated with vehicle (A), 100 μM 5a (B), and 200 μM 5a (C) for 24 h. After removal of excess donor, 100 μM of a H₂S fluorescent probe (WSP-1) was added. Images were taken afterwards.

In our imaging experiments, we did not observe shape changes of the cells after the treatment with donors, which indicates the donors may not be toxic under the doses we used. To further understand phosphorodithioate-based donors’ cytotoxicity, we tested H9c2 cell viability upon treating with donors 5a–5c. NaHS and GYY4137 were also tested as controls. The cells were treated with 50, 100, and 200 μM donors for 24 h and then cell viability was determined using cell counter kit (CCK)-8 assay. As shown in Figure 3, 5a, 5b and NaHS did not show significant toxicity to H9c2 cells under these doses. 5c exhibited some toxicity when concentration increased. Interestingly GYY4137 was safe at 50 and 100 μM, but showed strongest toxicity at 200 μM.

Figure 3.

Effects of H₂S donors on cell viability. H9c2 cells were treated with different concentrations of 5a–5c, GYY4137, and NaHS for 24 h. CCK-8 assay was performed to measure cell viability. Data were shown as mean ± SD (n = 4). ^#P < 0.01 versus control group.

Figure 5.

Mouse melanoma B16BL6 cells were treated with indicated donors in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin, respectively. Cells were cultured for 4 days and analyzed by a cell viability analyzer.

It is known that H₂S exhibits dose-dependent improvement of cell viability following oxidative injury.³⁷ We hypothesized that phosphorodithioate-based H₂S donors could exert similar protective effects. 5a and 5b were then selected to explore their protective effects against H₂O₂-induced oxidative damage in H9c2 cells. Again GYY4137 and NaHS were tested as controls. Three concentrations (50, 100, and 200 μM) for each donors were used. For GYY4137 only two concentrations (50 and 100 μM) were tested because of its strong toxicity at 200 μM. In these experiments, cells were incubated with each donor for 24 h and then challenged with 150 μM H₂O₂ and cultured for 5 h. After that cell viability was determined by CCK-8 assay. As shown in Figure 4, cell viability dropped to about 65% if H₂S donors were absent. In the presence of donors, however, cell viability increased significantly. These results suggest that H₂S donors indeed have some protective effects against oxidative injury. It should be noted that we did not observe significant protective effects when donors were used at lower concentrations (Figure S4 in supporting information).

Figure 4.

H9c2 cells viability after oxidative injury by H₂O₂ in the presence of donors 5a, 5b, GYY4137, and NaHS. Data were shown as mean ± SD (n = 4). ^*P < 0.01 versus control.

In 2011 Moore and co-workers showed that the slow-releasing donor GYY4137 exhibited some interesting anti-cancer activities in cells and mice xenograft.³⁰ We wondered if our donors would show similar activity. Therefore we tested the effects of 5a, 5b, and GYY4137 on cell growth and viability of mouse melanoma B16BL6 cell line. Two different concentrations (100 μM and 200 μM) were applied. Cell growth inhibitory effects were monitored for a 4-day culture period. As shown in Figure 5, 5a showed moderate toxicity at both concentrations while 5b showed relatively strong toxicity at 200 μM. Interestingly, GYY4137 showed the strongest activity to this cell line, which may suggest the C-P linkage is crucial for potency. However, at this moment it is still uncertain if these donors’ inhibitory effects on B16BL6 cells are associated with their H₂S production. Since H₂S generation is a very slow process, the inhibition may come from donor compounds themselves. Nevertheless more studies are needed to elucidate the detail mechanism.

In conclusion, here we report the design and synthesis of a series of O-arylated and alkylated phosphorodithioate-based H₂S donors. Their H₂S releasing capabilities were evaluated by fluorescence methods. N,O-Diarylated donors showed slow and sustainable H₂S generation which is similar to the well-known donor GYY4137, whereas O-alkylated donors exhibited very weak H₂S production. In addition, we explored some H₂S relevant activities of the donors. Compounds 5a and 5b exhibited notable protective effects against H₂O₂-induced oxidative damage in H9c2 myocytes. These donors also showed inhibitory effects against B16BL6 cancer cell proliferation but detailed mechanisms are to be explored. We are now actively studying other phosphorodithioates-based donors and their applications in H₂S research.