Skip to main content
NCBI home page
MedChemComm logoLink to MedChemComm
. 2017 Jan 30;8(3):606–615. doi: 10.1039/c7md00014f

Comparison of free-radical inhibiting antioxidant properties of carvedilol and its phenolic metabolites† ‡

Thomas C Malig a, Mitchell R Ashkin a, Austin L Burman a, Manuel Barday a, Belinda J M Heyne a, Thomas G Back a,✉
PMCID: PMC6071783  PMID: 30108776

graphic file with name c7md00014f-ga.jpgCarvedilol is a widely prescribed drug for the treatment of heart failure and the prevention of associated ventricular arrhythmias.

Abstract

Carvedilol is a widely prescribed drug for the treatment of heart failure and the prevention of associated ventricular arrhythmias. It has also been reported to function as a biological antioxidant via hydrogen atom transfer from its carbazole N–H moiety to chain-propagating radicals. Metabolites of the drug include phenolic derivatives, such as 3-hydroxy-, 4′-hydroxy- and 5′-hydroxycarvedilol, which are also potential antioxidants. A comparison of the radical-inhibiting activities of the parent drug and the three metabolites was carried out in two separate assays. In the first, hydrogen atom transfer from these four compounds to the stable radical DPPH was measured by the decrease in the UV-visible absorption at 515 nm of the latter. The known radical inhibitors BHT, 4-hydroxycarbazole and α-tocopherol were employed as benchmarks in parallel experiments. In the second assay, inhibition of the photoinduced free-radical 1,2-addition of Se-phenyl p-tolueneselenosulfonate to cyclopropylacetylene, along with competing ring-opening of the cyclopropane ring, was monitored by 1H NMR spectroscopy in the presence of the carvedilol-based and benchmark antioxidants. In both assays, carvedilol displayed negligible antioxidant activity, while the three metabolites all proved superior radical inhibitors to BHT, with radical-quenching abilities in the order 3-hydroxy- > 5′-hydroxy > 4′-hydroxycarvedilol. Among the metabolites, 3-hydroxycarvedilol displayed even stronger activity in both assays than α-tocopherol, the best of the benchmark antioxidants. These results suggest that the radical-inhibiting antioxidant properties that have been attributed to carvedilol are largely or exclusively due to its metabolites and not to the parent drug itself.

Introduction

Heart failure is a debilitating chronic disease and a leading cause of death, particularly in developed countries. For example, this disease affects ca. six million people in the United States, with 550 000 new cases diagnosed each year1 and associated health expenditures estimated at over $30 billion in 2012.2 Furthermore, heart failure patients are vulnerable to ventricular arrhythmias (VAs) that can result in sudden death. While several therapies exist, treatment with beta-blockers is among the most common and carvedilol (1), a potent nonselective beta-blocker, is often the drug of choice.3 Furthermore, it was recently discovered that carvedilol is unique among the beta-blocker class of drugs in its ability to suppress store overload induced calcium release (SOICR) in cardiac myocytes through its ability to bind to the ryanodine receptor (RyR2).4 A defective RyR2 ion channel can result in irregular gating and release of abnormal Ca2+ waves from the sarcoplasmic reticulum into the cytosol of cardiomyocytes, ultimately triggering VAs. This finding led in turn to the design of several novel carvedilol analogues that retain the SOICR-suppressing activity of the parent compound while reducing beta-blockade by ca. 1000-fold.4,5 While carvedilol is generally administered as its racemate, it was also found that the (R)-enantiomer of carvedilol retains the SOICR-inhibiting properties, along with considerably diminished binding to the beta-adrenoreceptor.6 Since carvedilol is ca. nanomolar in its beta-blocking ability, but micromolar in its effect on SOICR, side effects stemming from excessive adrenoreceptor inhibition (e.g. bradycardia, hypotension), associated with the relatively high doses required for SOICR suppression by racemic carvedilol, can be significantly decreased by administration of the (R)-enantiomer, while the desired SOICR attenuation remains intact. These observations raise the possibility of new therapies for heart failure and associated VAs, based on the use of more selective carvedilol analogues.

Moreover, the heart and cardiovascular system in general are particularly prone to oxidative stress. A third beneficial function of carvedilol, in addition to beta-blockade and SOICR suppression, is its potential antioxidant activity. For example, one investigation of ferrous ion-mediated lipid peroxidation indicated that carvedilol serves as an antioxidant and free-radical trap,7 while a subsequent study concluded that the antioxidant activity of carvedilol was the result of its ability to sequester ferric ions and not of its radical scavenging behaviour.8 It has also been reported that carvedilol protects cardiomyocytes from free-radical damage and prevents leakage of calcium ions from the ryanodine receptor.9 In another investigation,10 production of the prostaglandin 8-isoPGF2α was monitored as a measure of oxidative stress in an endothelial cell culture assay and carvedilol was observed to lower 8-isoPGF2α levels. In contrast, however, no evidence was found that carvedilol inhibits oxidative stress in healthy human volunteers.10 Thus, the precise role of carvedilol as a biological antioxidant remains unclear.

Diphenylamines (2, Fig. 1) are well-known industrial antioxidants because of their ability to inhibit undesired radical chain reactions in polymers, fuels, lubricants and other materials, generally by means of hydrogen donation from the N–H group to a chain-propagating radical species.11 The structural similarity of diphenylamines to the carbazole moiety of carvedilol supports the expectation that carvedilol could similarly serve as a radical-inhibiting antioxidant. However, carbazole and 2-hydroxycarbazole were inactive in inhibiting oxidation of methyl linoleate, while several 3-hydroxycarbazoles proved more effective in this assay.12 Several carbazole alkaloids such as carazostatin (3, Fig. 1),13 have also been shown to exhibit antioxidant activity in various assays, typically augmented by the presence of phenolic hydroxyl substituents. Furthermore, carvedilol is rapidly metabolized in mammals to a variety of products, including hydroxylated species, cleavage products and glucuronidates.14 In humans, radiolabelling experiments indicated that only 9% of the unchanged drug remained in the plasma after 1.5 hours, while only 2% of the excreted labelled compounds consisted of the intact drug.14a Among the metabolic pathways that have been reported are the CYP-mediated oxidations of the aromatic groups of carvedilol to their phenolic counterparts, including the 4′- and 5′-hydroxy derivatives 4 and 5, and 3-hydroxycarvedilol (6) (Fig. 1).15 Since various phenols are widely recognized as radical inhibitors and biological antioxidants (e.g. alpha-tocopherol), it is reasonable that phenolic carvedilol metabolites could contribute to, and perhaps exceed, any in vivo antioxidant properties displayed by the parent drug. Several earlier studies of the effects of 6 (also known as BM-910228 or SB211475) on pathologies where oxidative stress is implicated reported that this metabolite suppresses lipid peroxidation,16 mitigates post-ischemic injury17 and provides neuroprotective benefits.18 Moreover, 6 proved less effective than 1 in protecting against ischemic reperfusion injury,19 while both compounds prevented hydroxyl radical-mediated cardiac contractile dysfunction.20 Neither drug was effective in preventing nephropathy in kidney transplants.21 It is possible that some of the above results are unrelated to the antioxidant activities of 1 and 6, while the antioxidant behaviour of metabolites 4 and 5, and of other hydroxylated metabolites of carvedilol, has been less studied. This uncertainty underscores the need for an improved mechanistic understanding of the relative abilities of carvedilol and its phenolic metabolites to function as radical-inhibiting antioxidants.

Fig. 1. Structures of carvedilol, three phenolic metabolites and some structurally related compounds.

Fig. 1

Open in a new tab

We recently reported the synthesis of phenols 4–6, in order to determine their SOICR-inhibiting properties, and found that they are comparable to those of 1.22 Thus, the metabolism of carvedilol to these products neither significantly activates nor deactivates its ability to suppress SOICR. With these compounds in hand, we now report the relative free radical inhibiting abilities of 1, 4–6 and several benchmark compounds used for comparison in two separate assays.

Results and discussion

The ability of phenolic compounds to act as radical-inhibiting antioxidants stems from their ability to transfer a hydrogen atom to a propagating radical species in a process such as lipid peroxidation. This disrupts the radical chain process and generates a less reactive phenoxy radical that is stabilized by resonance and, in some cases, by steric protection that increases its kinetic persistence.

One method for evaluating the activity of phenolic or other radical inhibitors is to measure their ability to transfer hydrogen to diphenylpicrylhydrazyl (DPPH, 7, Scheme 1). The latter is a stable free radical with a strong UV/visible absorption at 515 nm, while the corresponding hydrazine (DPPH-H, 8) is essentially transparent at that wavelength. Thus, by monitoring the intensity of the DPPH absorption with time, it is possible to measure the decrease in its concentration and therefore to estimate the radical-inhibiting ability of putative antioxidants (Scheme 1).23

Scheme 1.

Scheme 1

Open in a new tab

In order to obtain a standard for comparison, we first tested the hydrogen-transfer abilities of well-known antioxidants that include the widely used industrial antioxidant butylated hydroxytoluene (BHT, 9), the phenolic carbazole derivative 10, as well as the biological antioxidant α-tocopherol (vitamin E, 11) (Fig. 2). The results are displayed in Fig. 3–6, where in each case the concentration of DPPH is plotted vs. time. A standard calibration plot of absorbance vs. DPPH concentration (see Fig. S1 of the ESI‡) was used to measure DPPH concentrations during the assays and a control experiment with DPPH (50 μM) alone under the same conditions showed no decrease in its absorbance after 60 min.

Fig. 2. Structures of benchmark antioxidants 9–11.

Fig. 2

Open in a new tab

Fig. 3. DPPH concentration vs. time in the presence of varying amounts of BHT (9) in methanol.

Fig. 3

Open in a new tab

Fig. 4. DPPH concentration vs. time in the presence of varying amounts of 4-hydroxycarbazole (10) in methanol.

Fig. 4

Open in a new tab

Fig. 5. DPPH concentration vs. time in the presence of varying amounts of α-tocopherol (11) in methanol.

Fig. 5

Open in a new tab

Fig. 6. Summary of results using 0.5 equiv. of each benchmark antioxidant in methanol.

Fig. 6

Open in a new tab

In Fig. 3–6, the results are shown at three different initial antioxidant concentrations (5 μM, 25 μM and 50 μM), representing 0.1, 0.5 and 1.0 equiv. relative to DPPH (50 μM). The time (t1/2) required by each antioxidant to decrease the concentration of DPPH to 25 μM, (half of its initial concentration), was also measured for each antioxidant for the plots where 0.5 equiv. were added.24 Moreover, some phenolic radical-inhibiting antioxidants can quench more than 1 equiv. of DPPH, as the phenoxy radicals derived from such species after hydrogen transfer are able to continue reacting with additional DPPH.25,26 These results are summarized in Table 1.

Table 1. DPPH assay of carvedilol (1), its metabolites 4–6 and benchmark compounds 9–11.

Antioxidant t 1/2 a (s) DPPH consumed b , c
BHT (9) 840 2.6
4-Hydroxycarbazole (10) 180 3.0
α-Tocopherol (11) 26 2.0
Carvedilol (1) No reaction 0.0
4′-Hydroxycarbazole (4) 23 2.6
5′-Hydroxycarbazole (5) <10 5.6
3-Hydroxycarbazole (6) Instant d 2.5
Open in a new tab

aTime required for the decrease in DPPH concentration from 50 μM to 25 μM by 25 μM (0.5 equiv.) of antioxidant.

bNumber of moles of DPPH consumed per mol of antioxidant at equilibrium.

cInitial concentration of antioxidant was 5 μM (0.1 equiv.).

dThe reaction was nearly complete by the time the first measurement could be made.

The effects of BHT are shown at the three initial concentrations in Fig. 3. When 0.1 and 0.5 equiv. of BHT were used, equilibration was slow and still incomplete after 60 min, but with 1.0 equiv., the DPPH concentration reached a steady state at 7.3 μM after 60 min. The t1/2 with 0.5 equiv. of BHT was 14 min (840 s). Fig. 3 also reveals that BHT consumed more than its equivalent amount of DPPH when added at the two lower concentrations, with each mol of BHT quenching ca. 2.6 mol of DPPH after 60 min when introduced at an initial concentration of 5 μM (0.1 equiv.). This is consistent with the literature, where a similar value of 2.8 equiv. of DPPH consumed by each equiv. of BHT was reported.25 4-Hydroxycarbazole (10) behaved similarly (Fig. 4), but with a faster initial reaction rate with t1/2 = 180 s with 0.5 equiv. and equilibrium again attained within 60 min, resulting in a final concentration of DPPH of 8.1 μM with 1.0 equiv. of 10. When an initial concentration of 5 μM (0.1 equiv.) of 10 was employed, 3.0 mol of DPPH were consumed for each mol of 10, thus surpassing the ratio of 2.6 observed with BHT (9). α-Tocopherol (11), proved to be the most potent benchmark antioxidant with t1/2 = 26 s with 0.5 equiv., and a final concentration of DPPH of only 3.7 μM when using 1.0 equiv. (Fig. 5). Moreover, equilibrium was established in less than 5 min, even with the lowest concentration of 11. At the lowest concentration, α-tocopherol consumed only 2.0 mol of DPPH for each mol of the antioxidant, again consistent with the literature.26Fig. 6 shows a superimposition of the effects of 0.5 equiv. of the three benchmark antioxidants to provide a more illustrative comparison.

The results of the DPPH assay of carvedilol (1) and its three phenolic metabolites 4–6 under the same conditions as employed for the above benchmarks are shown in Fig. 7–10, respectively, and the results are summarized in Table 1. To our surprise, carvedilol itself displayed negligible antioxidant activity, where even with an equimolar concentration of the drug relative to DPPH, there was insignificant reduction in DPPH concentration, compared to a control reaction with no antioxidant present (Fig. 7). In striking contrast, 4′-hydroxycarvedilol (4) proved to be a potent radical-inhibiting antioxidant. Fig. 8 shows that 0.5 equiv. of 4 afforded a t1/2 of 23 s, comparable to the results obtained with α-tocopherol. At the lowest initial concentration, each mol of compound 4 consumed 2.6 mol of DPPH. 5′-Hydroxycarvedilol (5) was even more effective than 4, with the two higher concentrations employing 0.5 and 1.0 equiv. reaching equilibrium within ca. 1 min (Fig. 9). Furthermore, when 0.1 equiv. of 5 was used, the antioxidant consumed ca. 5.8 mol of DPPH per mol of 5. Finally, the reaction of 3-hydroxycarvedilol (6) with DPPH was essentially instantaneous, as all three concentrations attained equilibrium within the time required to record the first data point (Fig. 10).27 Metabolite 6, however, consumed only 2.5 equivalents of DPPH at its lowest concentration. An overlay of the plots of the three metabolites 4–6 at the medium 0.5 equiv. level is provided in Fig. 11 to facilitate comparison.

Fig. 7. DPPH concentration vs. time in the presence and absence of 1 equiv. of carvedilol in methanol.

Fig. 7

Open in a new tab

Fig. 8. DPPH concentration vs. time in the presence of varying amounts of 4′-hydroxycarvedilol (4) in methanol.

Fig. 8

Open in a new tab

Fig. 9. DPPH concentration vs. time in the presence of varying amounts of 5′-hydroxycarvedilol (5) in methanol.

Fig. 9

Open in a new tab

Fig. 10. DPPH concentration vs. time in the presence of varying amounts of 3-hydroxycarvedilol (6) in methanol.

Fig. 10

Open in a new tab

Fig. 11. Summary of results using 0.5 equiv. of carvedilol metabolites 4–6 in methanol.

Fig. 11

Open in a new tab

These observations clearly indicate that in the DPPH assay, carvedilol (1) has negligible radical-inhibiting antioxidant activity, while its three phenolic metabolites (4–6) are powerful antioxidants, even when compared to α-tocopherol, the best of the three benchmark compounds employed for comparison. Nevertheless, several limitations to this assay have been reported,11a such as interference from single-electron transfers from the corresponding phenoxide anion to the radical DPPH, which independently affords DPPH-H (8) upon protonation. The relatively high acidity of phenols ensures the presence of significant concentrations of phenoxide ions in alkaline or even neutral media. Thus, in order to verify the above results, a second, different assay was devised, based on the ability of putative radical inhibitors to interrupt propagation steps in a radical chain reaction. Since compounds 4–6 appeared to be highly competent in the DPPH assay, a particularly fast and efficient radical chain process was desired for this purpose.

Some time ago, we reported that the 1,2-addition of Se-phenyl p-tolueneselenosulfonate (12) to acetylenes proceeds via a free-radical addition when initiated by its thermolysis in the presence of AIBN28 or by photolysis.29 Among such selenosulfonation reactions, we observed that cyclopropylacetylene undergoes both 1,2-addition to afford 14 as the dominant product, along with ring-opening of the cyclopropane moiety to produce the allenic minor product 16via radical 15 (Scheme 2).30 The ring-opening of cyclopropyl-substituted radicals31 and the group transfer of the phenylseleno moiety from various sources to chain-propagating radicals32,33 have been previously investigated, while authentic samples of the two products 14 and 16 were easily prepared via our previous method.30 Thus, the ability of carvedilol, its metabolites and corresponding benchmarks to inhibit the process shown in Scheme 2 was expected to provide additional insight into their relative radical-inhibiting antioxidant properties.

Scheme 2.

Scheme 2

Open in a new tab

The assay was performed by initiating the process with UV light in a photoreactor equipped with 254 nm lamps and a turntable. The reactions were monitored via1H NMR spectroscopy in CDCl3 solution, in which the disappearance of the aromatic signals from the p-toluenesulfonyl group of 12 at δ 7.2 ppm and the appearance of the well-separated corresponding signals from 14 and 16 at δ 7.7 and δ 7.8 ppm, respectively, was measured by integration (see Fig. 12). The disappearance of the acetylenic proton of cyclopropylacetylene at δ 1.8 ppm was employed to follow its consumption. Overlap of the vinylic and allenic signals from 14 and 16 at ca. δ 5.9 ppm made these less desirable for monitoring the progress of the reaction. Dimethyl sulfone was employed as an internal standard. Each assay was performed under argon after degassing the solution by several freeze–pump–thaw cycles. The kinetic plot for the control selenosulfonation of cyclopropylacetylene in the absence of an antioxidant is shown in Fig. 13, where the disappearance of the selenosulfonate is accompanied by the simultaneous formation of 14 and 16, produced in the final ratio of ca. 7.0 : 1. The consumption of cyclopropylacetylene tracked that of the selenosulfonate closely, but is not shown for the sake of clarity. The reaction was 50% complete, based on the mol fraction of remaining 12, in 4.6 min (t1/2) and the selenosulfonate was fully consumed within 30 min. The disappearance of the selenosulfonate followed first order kinetics (see Fig. S2 in the ESI‡), consistent with the homolysis of 12 as the expected rate-determining step in Scheme 2.

Fig. 12. 1H NMR spectra of the reaction of selenosulfonate 12 with cyclopropylacetylene. A) At start (t = 0 min); B) at t = 5 min; C) at end (t = 30 min).

Fig. 12

Open in a new tab

Fig. 13. Kinetic plot for the selenosulfonation of cyclopropylacetylene with no inhibitor (control).

Fig. 13

Open in a new tab

The same reaction was then repeated in the presence of 1 mol% of each of the benchmark antioxidants 9–11, carvedilol (1) and its metabolites 4–6, as shown in Fig. 14–20, respectively, along with the plot for the disappearance of selenosulfonate 12 under control conditions (dashed line) in each assay for comparison. The times required for 50% and full completion of each reaction are summarized in Table 2. The results in the presence of BHT (9) (Fig. 14) indicate only a slight retardation of the process compared to the control reaction, with t1/2 = 7.9 min and the time required for completion of ca. 45 min. 4-Hydroxycarbazole (10) proved a more effective radical inhibitor (Fig. 15), with t1/2 = 32.4 min and complete consumption of 12 reached at ca. 90 min. While α-tocopherol (11) displayed a comparable t1/2 = 27.8 min compared to that of 10, it is noteworthy that the reaction was almost completely inhibited for the first 20 min, following which the disappearance of 12 and appearance of the products 14 and 16 proceeded at a much faster pace. This lag time is consistent with the powerful radical-inhibiting properties of α-tocopherol, which persisted until it was fully consumed (only 1 mol% was present initially), following which the reaction resumed a more normal course.

Fig. 14. Kinetic plot for the selenosulfonation of cyclopropylacetylene in the presence of 1 mol% of BHT (9).

Fig. 14

Open in a new tab

Fig. 15. Kinetic plot for the selenosulfonation of cyclopropylacetylene in the presence of 1 mol% of 4-hydroxycarbazole (10).

Fig. 15

Open in a new tab

Fig. 16. Kinetic plot for the selenosulfonation of cyclopropylacetylene in the presence of 1 mol% of α-tocopherol (11).

Fig. 16

Open in a new tab

Fig. 17. Kinetic plot for the selenosulfonation of cyclopropylacetylene in the presence of 1 mol% of carvedilol (1).

Fig. 17

Open in a new tab

Fig. 18. Kinetic plot for the selenosulfonation of cyclopropylacetylene in the presence of 1 mol% of 4′-hydroxycarvedilol (4).

Fig. 18

Open in a new tab

Fig. 19. Kinetic plot for the selenosulfonation of cyclopropylacetylene in the presence of 1 mol% of 5′-hydroxycarvedilol (5).

Fig. 19

Open in a new tab

Fig. 20. Kinetic plot for the selenosulfonation of cyclopropylacetylene in the presence of 1 mol% of 3-hydroxycarvedilol (6).

Fig. 20

Open in a new tab

Table 2. Radical inhibition of the selenosulfonation of cyclopropylacetylene by carvedilol (1), its metabolites 4–6 and benchmark compounds 9–11.

Compound t 1/2 (min) Time to completion (min)
No inhibitor 4.6 30
BHT (9) 7.9 45
4-Hydroxycarbazole (10) 32.4 90
α-Tocopherol (11) 27.8 60
Carvedilol (1) 4.6 30
4′-Hydroxycarvedilol (4) 8.7 45
5′-Hydroxycarvedilol (5) 13.8 90
3-Hydroxycarvedilol (6) 64.6 180
Open in a new tab

When carvedilol (1) was subjected to this assay under identical conditions, it had no discernible effect on the rate of the selenosulfonation reaction (Fig. 17). The plot for the consumption of 12 superimposed closely on that of the control reaction. This confirms the finding from the DPPH assay, where carvedilol displayed essentially no antioxidant activity. The 4′-hydroxy metabolite 4 (Fig. 18) revealed weak activity, comparable to that of BHT (Fig. 14), with t1/2 = 8.7 min and a completion time of 45 min. Inhibition of the selenosulfonation reaction was more effective with the 5′-hydroxy derivative 5 (Fig. 19), where a longer t1/2 of 13.8 min and a completion time of ca. 90 min were observed. Finally, 3-hydroxycarvedilol (6) (Fig. 20) resembled α-tocopherol (Fig. 16) in that it displayed a similar lag time, where it very effectively quenched the radical chain process for the first 20 minutes, followed by a more rapid reaction rate. However, a longer t1/2 of 64.6 min and overall completion time of >180 min indicate that 6 is even more effective than α-tocopherol in this assay.

It is also noteworthy that the ratio of the amount of 1,2-addition product 14 to that of the allene 16 was relatively constant throughout these experiments, within the range 5.94 : 1 to 7.02 : 1. Moreover, the rate of ring-opening of the cyclopropylcarbinyl radical 17 (Fig. 21) was reported31a to be 1.3 × 108 s–1 at 25 °C and this process has been used as a radical clock reaction to measure the rates of competing processes.34 While cyclopropylvinyl radicals such as 13 in Scheme 2 have been less thoroughly studied, Baines et al.35 reported that radical 18 (Fig. 21) undergoes ring-opening with an even faster rate constant of 1.6 ± 0.2 × 1010 s–1 at 20 °C. If it is assumed that the ring-opening of radical 13 is roughly comparable to that of 18,36 then it can be concluded that the phenylseleno group transfer from selenosulfonate 12 to radical 13 in the 1,2-addition to cyclopropylacetylene is an exceptionally facile process, as the corresponding product 14 consistently dominated over the ring-opened allene 16.

Fig. 21. Structures 17–20.

Fig. 21

Open in a new tab

The relatively slow initial homolysis of the selenosulfonate 12 and its unimolecular decomposition suggest that the resulting p-toluenesulfonyl radical is the principal target for hydrogen-donating antioxidants, resulting in the expected formation of the corresponding sulfinic acid 19 (Fig. 21).37 However, we also considered the possibility that the vinyl radical 13 might react to some extent with hydrogen atom donors. This possibility was excluded through the failure to observe, by either GC-MS or HPLC analysis, the expected product 20 (Fig. 21) in the reaction mixture, even when the quantity of the antioxidant was increased from 1% to 50%. An authentic sample of 20 was prepared (see ESI‡) and could be detected under these conditions even at very low concentrations. We were also unable to detect the sulfinic acid 19 (Fig. 21) in the reaction mixture by these methods, but that does not rule out its formation, as sulfinic acids are known to undergo conjugate additions to other vinyl sulfones,38 similar to those produced in the present process, as well as self-condensation to afford α-sulfinyl sulfones, that in turn decompose readily to a variety of other products.39 We therefore postulate that hydrogen atom transfer to the initially-formed sulfonyl radical is indeed the principal pathway for inhibition of the overall process shown in Scheme 2.

Conclusions

The relative antioxidant abilities of carvedilol and three of its hydroxylated metabolites, along with three common phenolic radical-inhibiting antioxidants employed for comparison, were measured in two separate assays. The first assay revealed that the three benchmark antioxidants were effective in quenching the DPPH radical in the order BHT (9) < 4-hydroxycarbazole (10) < α-tocopherol (11). The three phenolic carvedilol derivatives displayed activities in the order 4′-hydroxycarvedilol (4) < 5′-hydroxycarvedilol (5) ≪ 3-hydroxycarvedilol (6). All three compounds were comparable to (in the case of 4) or more active (in the case of 5 and 6) than even the best benchmark oxidant 11.

In the second assay, based on inhibition of the selenosulfonation of cyclopropylacetylene, BHT exhibited a relatively modest effect on this free-radical chain reaction, while both 4-hydroxycarbazole and α-tocopherol proved to be potent inhibitors of this process. The three metabolites displayed the same order of antioxidant activity as in the DPPH assay, with 3-hydroxycarvedilol affording by far the strongest inhibitory effect, with nearly complete suppression of the process during an initial lag time of ca. 20 min.

Although several reports have adumbrated the antioxidant properties of carvedilol (1), to our surprise, the parent drug showed essentially no discernible antioxidant activity in either of the two assays described here. While our in vitro studies may not reflect the antioxidant properties of these compounds in a more complex in vivo environment, they strongly suggest that it is not carvedilol itself, but rather its phenolic metabolites that provide most or all of the protective antioxidant activity associated with the drug.

Experimental

BHT, 4-hydroxycarbazole, α-tocopherol, DPPH, cyclopropylacetylene and carvedilol were obtained from commercial sources. Selenosulfonate 12,40 dimethyl sulfone41 and metabolites 4–622 were prepared as described previously. The preparation and characterization of authentic samples of the 1,2-addition product 14, the allenic sulfone 16 and the cyclopropylvinyl sulfone 20 are provided in the ESI.‡

DPPH assay

A calibration curve of absorbance at λ = 515 nm vs. DPPH concentration was employed for the measurement of DPPH concentrations during the assays (see ESI‡). The molar extinction coefficient of DPPH in methanol under these conditions was calculated to be 1.06 × 104 M–1 cm–1 (lit.42 1.09 × 104 M–1 cm–1). A control experiment in the absence of an antioxidant showed a decrease in concentration of DPPH from 50.2 μM to 49.0 μM after 1 h, considered to be negligible. In a typical experiment 2.50 mL of a 70 μM DPPH solution in methanol was added to a 3.5 mL cuvette at 30 °C. To the cuvette containing the DPPH solution, 1.0 mL of the antioxidant solution in methanol at 30 °C was then added in order to make up a final solution containing 50 μM DPPH and an initial molar ratio of either 10 : 1, 2 : 1, or 1 : 1 of DPPH : antioxidant. The temperature was maintained at 30 °C throughout the experiment and absorbance readings began immediately at a rate of 4 scans per minute. Each assay was performed in triplicate.

Selenosulfonation inhibition assay

A stock solution containing selenosulfonate 12 (150 mM), cyclopropylacetylene (150 mM) and dimethyl sulfone (50 mM) as the internal standard was prepared in CDCl3. The solution was degassed by several freeze–pump–thaw cycles with the introduction of argon. Aliquots of 50 μL were measured into a series of NMR tubes and diluted with 450 μL of CDCl3. The NMR tubes were placed in a Rayonet UV reactor equipped with eight 254 nm lamps, a rotating carousel, and a cooling fan that maintained the temperature at 28 °C. An NMR tube was withdrawn at each of the required times and analyzed by 1H NMR integration. Each assay was performed in triplicate. Error bars in Fig. 14–20 indicate ±1 standard deviation.

Supplementary Material

Click here for additional data file. (405.1KB, pdf)

Acknowledgments

We thank the Natural Sciences and Engineering Research Council of Canada for financial support. T. C. M. thanks the University of Calgary and the Province of Alberta for a Department of Chemistry Entrance Scholarship and a Queen Elizabeth II scholarship. M. R. A. and A. L. B. acknowledge support from the University of Calgary via the Program for Undergraduate Research Experience (PURE). M. B. was an exchange student from France with support from the Université Pierre et Marie Curie. We thank Prof. T. C. Sutherland for the use of his UV-visible spectrometer.

Footnotes

†The authors declare no competing interests.

‡Electronic supplementary information (ESI) available. See DOI: 10.1039/c7md00014f

References

  1. Bui A. L., Horwich T. B., Fonarow G. C. Nat. Rev. Cardiol. 2011;8:30–41. doi: 10.1038/nrcardio.2010.165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Cook C., Cole G., Asaria P., Jabbour R., Francis D. P. Int. J. Cardiol. 2014;171:368–376. doi: 10.1016/j.ijcard.2013.12.028. [DOI] [PubMed] [Google Scholar]
  3. (a) Poole-Wilson P. A., Swedberg K., Cleland J. G., DiLenarda A., Hanrath P., Komajda M., Lubsen J., Lutiger B., Metra M., Remme W. J., Torp-Pedersen C., Scherhag A., Skene A. Lancet. 2003;362:7–13. doi: 10.1016/S0140-6736(03)13800-7. [DOI] [PubMed] [Google Scholar]; (b) Foody J. M., Farrell M. H., Krumholz H. M. JAMA, J. Am. Med. Assoc. 2002;287:883–889. doi: 10.1001/jama.287.7.883. [DOI] [PubMed] [Google Scholar]; (c) Remme W. J. Cardiovasc. Drugs Ther. 2010;24:351–358. doi: 10.1007/s10557-010-6247-7. [DOI] [PubMed] [Google Scholar]
  4. Zhou Q., Xiao J., Jiang D., Wang R., Vembaiyan K., Wang A., Smith C. D., Xie C., Chen W., Zhang J., Tian X., Jones P. P., Zhong X., Guo A., Chen H., Zhang L., Zhu W., Yang D., Li X., Chen J., Gillis A. M., Duff H. J., Cheng H., Feldman A. M., Song L.-S., Fill M., Back T. G., Chen S. R. W. Nat. Med. 2011;17:1003–1010. doi: 10.1038/nm.2406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Smith C. D., Wang A., Vembaiyan K., Zhang J., Xie C., Zhou Q., Wu G., Chen S. R. W., Back T. G. J. Med. Chem. 2013;56:8626–8655. doi: 10.1021/jm401090a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Zhang J., Zhou Q., Smith C. D., Chen H., Tan Z., Chen B., Nani A., Wu G., Song L.-S., Fill M., Back T. G., Chen S. R. W. Biochem. J. 2015;470:233–242. doi: 10.1042/BJ20150548. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Yue T. L., Cheng H. Y., Lysko P. G., McKenna P. J., Feuerstein R., Gu J. L., Lysko K. A., Davis L. L., Feuerstein G. J. Pharmacol. Exp. Ther. 1992;263:92–98. [PubMed] [Google Scholar]
  8. Noguchi N., Nishino K., Niki E. Biochem. Pharmacol. 2000;59:1069–1076. doi: 10.1016/s0006-2952(99)00417-7. [DOI] [PubMed] [Google Scholar]
  9. Mochizuki M., Yano M., Oda T., Tateishi H., Kobayashi S., Yamamoto T., Ikeda Y., Ohkusa T., Ikemoto N., Matsuzaki M. J. Am. Coll. Cardiol. 2007;49:1722–1732. doi: 10.1016/j.jacc.2007.01.064. [DOI] [PubMed] [Google Scholar]
  10. Fahlbusch S. A., Tsikas D., Mehls C., Gutzki F.-M., Böger R. H., Frölich J. C., Stichtenoth D. O. Eur. J. Clin. Pharmacol. 2004;60:83–88. doi: 10.1007/s00228-004-0729-0. [DOI] [PubMed] [Google Scholar]
  11. (a) For a recent review of the mechanistic aspects of the radical chain-inhibiting and antioxidant properties of diphenylamines, phenols and related species, see: Ingold K. U., Pratt D. A., Chem. Rev., 2014, 114 , 9022 –9046 . [DOI] [PubMed] [Google Scholar]; (b) For a recent structure-activity study of substituted diphenylamines as antioxidants, see: Hanthorn J. J., Valgimigli L., Pratt D. A., J. Am. Chem. Soc., 2012, 134 , 8306 –8309 . [DOI] [PubMed] [Google Scholar]
  12. Iwatsuki M., Komuro E., Niki E. Bull. Chem. Soc. Jpn. 1995;68:620–624. [Google Scholar]
  13. For examples of carbazole alkaloids with antioxidant properties, see: Hieda Y., Anraku M., Choshi T., Tomida H., Fujioka H., Hatae N., Hori O., Hirose J., Hibino S., Bioorg. Med. Chem. Lett., 2014, 24 , 3530 –3533 . [DOI] [PubMed] [Google Scholar]
  14. (a) Neugebauer G., Neubert P. Eur. J. Drug Metab. Pharmacokinet. 1991;16:257–260. doi: 10.1007/BF03189969. [DOI] [PubMed] [Google Scholar]; (b) Schaefer W. H., Politowski J., Hwang B., Dixon F., Goalwin A., Gutzait L., Anderson K., DeBrosse C., Bean M., Rhodes G. R. Drug Metab. Dispos. 1998;26:958–969. [PubMed] [Google Scholar]; (c) Oldham H. G., Clarke S. E. Drug Metab. Dispos. 1997;25:970–977. [PubMed] [Google Scholar]; (d) Lim H.-K., Chen J., Sensenhauser C., Cook K., Subrahmanyam V. Rapid Commun. Mass Spectrom. 2007;21:1821–1832. doi: 10.1002/rcm.3024. [DOI] [PubMed] [Google Scholar]
  15. (a) Li J., Wang L., Wang S., Chen M., Gu E., Hu G., Ge R. J. Chromatogr., B. 2015;974:138–146. doi: 10.1016/j.jchromb.2014.10.037. [DOI] [PubMed] [Google Scholar]; (b) Furlong M. T., He B., Mylott W., Zhao S., Mariannino T., Shen J., Stouffer B. J. Pharm. Biomed. Anal. 2012;70:574–579. doi: 10.1016/j.jpba.2012.05.026. [DOI] [PubMed] [Google Scholar]; (c) Chandiran I. S., Jayaveera K. N., Reddy R. S. J. Chem. Pharm. Res. 2011;3:341–353. [Google Scholar]; (d) Myung S.-W., Jo C.-H. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2005;822:70–77. doi: 10.1016/j.jchromb.2005.05.023. [DOI] [PubMed] [Google Scholar]; (e) Zhou H.-H., Wood A. J. J. Clin. Pharmacol. Ther. 1995;57:518–524. doi: 10.1016/0009-9236(95)90036-5. [DOI] [PubMed] [Google Scholar]
  16. (a) Santos D. J. S. L., Moreno A. J. M. Biochem. Pharmacol. 2001;61:155–164. doi: 10.1016/s0006-2952(00)00522-0. [DOI] [PubMed] [Google Scholar]; (b) Abreu R. M. V., Santos D. J. S. L., Moreno A. J. M. J. Pharmacol. Exp. Ther. 2000;295:1022–1030. [PubMed] [Google Scholar]
  17. (a) Christopher T. A., Lopez B. L., Ma X.-L., Feuerstein G. Z., Ruffolo Jr. R. R., Yue T.-L. Br. J. Pharmacol. 1998;123:292–298. doi: 10.1038/sj.bjp.0701598. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Kramer J. H., Weglicki W. B. Free Radical Biol. Med. 1996;21:813–825. doi: 10.1016/0891-5849(96)00184-0. [DOI] [PubMed] [Google Scholar]
  18. Lysko P. G., Lysko K. A., Webb C. L., Feuerstein G., Mason P. E., Walter M. F., Mason R. P. Biochem. Pharmacol. 1998;56:1645–1656. doi: 10.1016/s0006-2952(98)00275-5. [DOI] [PubMed] [Google Scholar]
  19. Brunvand H., Liu G., Ma X.-L., Yue T.-L., Ruffolo Jr. R. R., Feuerstein G. Z. Eur. J. Pharmacol. 1998;356:193–198. doi: 10.1016/s0014-2999(98)00494-4. [DOI] [PubMed] [Google Scholar]
  20. Flesch M., Maack C., Cremers B., Bäumer A. T., Südkamp M., Böhm M. Circulation. 1999;100:346–353. doi: 10.1161/01.cir.100.4.346. [DOI] [PubMed] [Google Scholar]
  21. Götteman U., Oltersdorf J., Schaub M., Knoll T., Back W. E., van der Woude F. J., Braun C. J. Cardiovasc. Pharmacol. 2003;42:442–450. doi: 10.1097/00005344-200309000-00017. [DOI] [PubMed] [Google Scholar]
  22. Malig T., Xiao Z., Chen S. R. W., Back T. G. Bioorg. Med. Chem. Lett. 2016;26:149–153. doi: 10.1016/j.bmcl.2015.11.008. [DOI] [PubMed] [Google Scholar]
  23. (a) Blois M. S. Nature. 1958;181:1199–1200. [Google Scholar]; (b) For a review, see: Kedare S. B., Singh R. P., J. Food Sci. Technol., 2011, 48 , 412 –422 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Attempts to obtain t1/2 when an equimolar amount of the antioxidant was added resulted in reactions that were too fast to measure accurately in some cases
  25. Bondet V., Brand-Williams W., Berset C. LWT–Food Sci. Technol. 1997;30:609–615. [Google Scholar]
  26. Prevc T., Levart A., Cigić I. K., Salobir J., Ulrih N. P., Cigić B. Molecules. 2015;20:14777–14790. doi: 10.3390/molecules200814777. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Minor irregularities in the early stages of the kinetic plots of 4, 5 and 6 are attributed to limitations in reactant mixing efficiency and rapid reaction times
  28. (a) Back T. G., Collins S., Kerr R. G. J. Org. Chem. 1983;48:3077–3084. [Google Scholar]; (b) Back T. G., Collins S., Gokhale U., Law K.-W. J. Org. Chem. 1983;48:4776–4779. [Google Scholar]
  29. (a) Back T. G., Nakajima K. J. Org. Chem. 1998;63:6566–6571. [Google Scholar]; (b) Back T. G., Lai E. K. Y., Muralidharan K. R. J. Org. Chem. 1990;55:4595–4602. [Google Scholar]
  30. Back T. G., Muralidharan K. R. J. Org. Chem. 1989;54:121–125. [Google Scholar]
  31. (a) Maillard B., Forrest D., Ingold K. U. J. Am. Chem. Soc. 1976;98:7024–7026. [Google Scholar]; (b) Mathew L., Warkentin J. J. Am. Chem. Soc. 1986;108:7981–7984. [Google Scholar]
  32. For reviews of radical reactions of selenium and tellurium compounds, see: ; (a) Back T. G. and Press D. J., in The Chemistry of Organic Selenium and Tellurium Compounds, ed. Z. Rappoport, Wiley, Chichester, U.K., 2012, ch. 8, vol. 3 (Part 1), pp. 357–522. [Google Scholar]; (b) Renaud P., in Organoselenium Chemistry – Modern Developments in Organic Synthesis, ed. T. Wirth, Springer, Heidelberg, Top. Curr. Chem., 2000, vol. 208, pp. 81–112. [Google Scholar]; (c) Back T. G., in Organoselenium Chemistry – A Practical Approach, ed. T. G. Back, Oxford University Press, Oxford, U.K., 1999, ch. 9, pp. 173–191. [Google Scholar]
  33. For example, the rate of phenylseleno group transfer to the 5-hexenyl radical from selenosulfonate 12 has been reported to be 3.0 × 106 L mol–1 s–1: Russell G. A., Tashtoush H., J. Am. Chem. Soc., 1983, 105 , 1398 –1399 . [Google Scholar]
  34. (a) Griller D., Ingold K. U. Acc. Chem. Res. 1980;13:317–323. [Google Scholar]; (b) Newcomb M., Encyclopedia of Radicals in Chemistry, Biology and Materials, Wiley, Chichester, U.K., 2012, ch. 1, pp. 107–124. [Google Scholar]
  35. Milnes K. K., Gottschling S. E., Baines K. M. Org. Biomol. Chem. 2004;2:3530–3534. doi: 10.1039/b412676a. [DOI] [PubMed] [Google Scholar]
  36. Resonance stabilization by the phenyl substituent of the radical formed by ring-opening of 18 is expected to facilitate this process. The precise effect of the sulfonyl group on the rate of the similar opening of 13 is uncertain
  37. Hydrogen donation from the antioxidant to the phenylseleno radical is unlikely to be competitive with donation to the sulfonyl radical due to the lower reactivity of the former. Moreover, selenols are in turn excellent hydrogen atom donors (see ref. 32)
  38. Truce W. E., Wellisch E. J. Am. Chem. Soc. 1952;74:5177–5179. [Google Scholar]
  39. (a) Oae S. and Kunieda N., in Organic Chemistry of Sulfur, ed. S. Oae, Plenum Press, New York, 1977, ch. 11, pp. 603–647. [Google Scholar]; (b) Kice J. L., Pawlowski N. E. J. Am. Chem. Soc. 1964;86:4898–4904. [Google Scholar]; (c) Guaraldi G., Venier C. G. J. Org. Chem. 1966;31:3561–3567. [Google Scholar]
  40. Back T. G., Collins S., Krishna M. V. Can. J. Chem. 1987;65:38–42. [Google Scholar]
  41. Choban A. F., Yurchuk I. R., Lyavinets A. S. Russ. J. Gen. Chem. 2008;78:2071–2074. [Google Scholar]
  42. Lebeau J., Furman C., Bernier J.-L., Duriez P., Teissier E., Cotelle N. Free Radical Biol. Med. 2000;29:900–912. doi: 10.1016/s0891-5849(00)00390-7. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Click here for additional data file. (405.1KB, pdf)

Articles from MedChemComm are provided here courtesy of Royal Society of Chemistry

RESOURCES