Introduction

Abstract Introduction Results Discussion References

Phenolphthalein, a triphenylmethane derivative (fig. 1), is widely used as a laxative and cathartic agent and is the active ingredient in many commercial products that are available for use (and possible abuse by some bulimics) as over-the-counter, nonprescription formulations (1-4). A recent report of the National Toxicology Program found evidence that phenolphthalein was a carcinogen based on 2-yr feeding studies (5). Earlier subchronic toxicity studies of phenolphthalein given orally to rats and mice in food found little evidence of toxicity in rats but did find toxic effects in mice. Changes found in the mouse reproductive system included depressed testis and right epididymal weights and sperm density, elevated production of abnormal sperm, and morphological changes in seminiferous tubules. Hematopoietic changes included bone marrow hypoplasia, increased splenetic hematopoiesis in males, and an elevated incidence of micronucleated erythrocytes (6, 7).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Molecular structure of phenolphthalein: 3,3-bis(4-hydroxyphenyl)-1-(3H)-isobenzo-furanone.

The pharmacologydosage, absorption, distribution, biotransformation, and excretionof phenolphthalein and related diphenylmethane laxatives has been reviewed (5, 8). Phenolphthalein is believed to function as a laxative by stimulating the synthesis and release of prostaglandins, as suggested by the ability of prostaglandin inhibitors such as aspirin and indomethacin to inhibit the laxative action of phenolphthalein in mice (9). At ordinary dose levels, phenolphthalein appears to be metabolized to its glucuronide conjugate in the intestinal lumen and the liver and excreted largely as the glucuronide. Animal studies have shown that phenolphthalein is excreted almost entirely in the bile as phenolphthalein glucuronide (>98%) (10-12). Phenolphthalein is glucuronidated in intestinal mucosa at an activity about 20% of that in the liver on a per gram of organ basis (13). When either phenolphthalein or phenolphthalein glucuronide was administered iv to male rats, they were both rapidly excreted in bile as the glucuronide conjugate. At dose levels of 30 mg phenolphthalein, 99.6% was excreted as phenolphthalein glucuronide and 0.4% as phenolphthalein itself (14).

The extent and duration of phenolphthalein laxation seems to be enhanced by extensive enterohepatic recirculation after hydrolysis of the glucuronide by bacterial glucuronidase in the intestine. In experiments with [³H]phenolphthalein administered to female Wistar albino rats, 90% was eliminated predominantly as the glucuronide within 3 hr after ip injection; return of labeled phenolphthalein glucuronide to duodena of bile-duct-cannulated animals demonstrated 85% enterohepatic recirculation in 24 hr. Pretreatment of animals with antibiotics to suppress intestinal microflora decreased the enterohepatic recirculation rate to 22% (15).

The phenolic substituents of phenolphthalein (fig. 1) make it a likely candidate for peroxidase metabolism(scheme 1), which provided the impetus for the present study. Similar reactions with phenolic compounds are well known, with free radical intermediates having been detected with EPR during peroxidase-catalyzed metabolism (16-22). Lactoperoxidase is a typical mammalian peroxidase found in milk, saliva, and tears as well as the mammary glands, and has a biochemistry resembling that of HRP² and many other peroxidases. In general, mammalian peroxidases are activated to compound I enzyme intermediates by hydrogen peroxide, which can subsequently oxidize many phenolic substrates to their corresponding free radicals (23).

View larger version (10K): [in this window] [in a new window]   Scheme 1.   Metabolism by peroxidase.

View larger version (12K): [in this window] [in a new window]   Scheme 2.   Formation of the phenolphthalein phenoxyl metabolite and the subsequent oxidation of GSH, NADH, and ascorbate.

Phenoxyl radical metabolites can react with suitable biochemical reductants such as GSH and NAD(P)H to regenerate the parent phenolic compound and a reductant-derived radical such as GS^· or NAD^· which reacts to form superoxide (24). Therefore, one molecule of phenol can undergo multiple cycles of oxidation and reduction in a process termed "futile metabolism," with the concomitant production of superoxide in quantities that greatly exceed the concentration of the phenolic compound.

Since both phenolphthalein and its glucuronide have phenolic groups that are likely to be substrates for peroxidases and since the exposure of phenolphthalein and its glucuronide to endogeneous peroxidases is likely to be extended by extensive enterohepatic recirculation, we have undertaken in vitro studies to investigate the possible peroxidase metabolism of phenolphthalein and its glucuronide to phenoxyl radicals with the associated production of biochemical reductant-derived free radicals and superoxide.

Materials and Methods

Chemicals and Biochemicals. Ascorbic acid, ACS Reagent Grade, DTPA, GSH, HRP [EC 1.11.1.7], NADH, and phenolphthalein glucuronic acid sodium salt were purchased from Sigma Chemical Co. (St. Louis, MO) and were used as received. Phenolphthalein was obtained from Fisher (Raleigh, NC). Catalase (from beef liver, 65,000 U/mg crystalline suspension in water) and superoxide dismutase (SOD; from bovine erythrocytes, 3000 U/mg lyophilizate according to the method of McCord and Fridovich (25)) were purchased from Boehringer-Mannheim (Indianapolis, IN) and were used as received. The spin trap DMPO, also purchased from Sigma, was vacuum-distilled twice at ambient temperature and stored at 70°C until use. Lactoperoxidase [EC 1.11.1.7] (LPO; from bovine milk, 87 U/mg) was purchased from Worthington Biochemical Corporation (Freehold, NJ) and was used as received. The buffer for all reactions was 100 mM enzyme grade tris(hydroxymethyl)-aminomethane (Bethesda Research Laboratories, Bethesda, MD) adjusted to pH 7.4 with hydrochloric acid (Tris-HCl). The Tris-HCl buffer was treated with Chelex 100 resin (Bio-Rad, Hercules, CA) to remove adventitious transition metal ions and contained 50 µM DTPA as a further precaution against trace metal catalysis.

Electron Spin Resonance (EPR) Experiments. Room-temperature EPR spectroscopy was used to detect and identify free radical intermediates formed during reactions. Although some of the free radicals in these experiments were too short lived to be detected directly in static incubations, certain intermediates were detected using the spin-trapping technique. The DMPO spin trap was present in all experiments (unless otherwise specified) to react with the short-lived free radical intermediate to form a relatively long-lived DMPO radical adduct. The original free radical intermediate was then identified by the resulting EPR hyperfine coupling constants of its DMPO adduct.

Oxygen Consumption Experiments. Oxygen consumption measurements were made using a Clarke-type oxygen electrode fitted to a 1.8 ml Gilson sample cell and monitored by a Yellow Springs Instrument Company Model 53 oxygen monitor. Oxygen concentration in the samples as a function of time was recorded by a personal computer interfaced to the oxygen monitor with a Data Translation DT2801 data acquisition board. The oxygen consumption data were plotted by importing the ASCII data files into the Origin program (MicroCal Software, Inc., Northampton, MA).

	Results

Abstract Introduction Results Discussion References

Direct EPR Observation of Radicals from Phenolphthalein Metabolism by HRP. A fast-flow reaction mixture containing phenolphthalein (fig. 1), HRP, and hydrogen peroxide produced EPR signals that seemed to be a superposition of spectra from two species (fig. 2A). Control studies demonstrated that no signal was observed in the absence of phenolphthalein (fig. 2D) or HRP (fig. 2E), and a much weaker signal was observed in the absence of added hydrogen peroxide (fig. 2C). Computer analysis of the composite spectrum showed it to be a relatively broad singlet (with peak-to-peak line width of 3.6 gauss) comprising 86% of the observed spectrum, while the remaining 14% resulted from a species with two sets of two equivalent hydrogen atoms with a^H = 6.56 gauss and a^H = 1.79 gauss, respectively. These parameters permitted a successful simulation of the composite spectrum (fig. 2B).

View larger version (34K): [in this window] [in a new window]   Fig. 2.   The ESR fast-flow spectra of phenolphthalein-derived radicals produced in a system of phenolphthalein, hydrogen peroxide, and horseradish peroxidase. The concentrations of H2O2, HRP, and phenolphthalein in the flat cell were 12.5 mM, 15 U/ml, and ca. 0.2 mM. Equal volumes of phenolphthalein/H2O2 and HRP in deoxygenated, pH 7.4 Tris buffer solutions were mixed milliseconds before entering the flat cell at a total flow rate of 60 ml/min. The solutions also contained about 1% ethanol. (A) Complete system with phenolphthalein, H2O2, and HRP. (B) Optimized computer simulation composed of 86% of a species with a single broad line (peak-to-peak line width, 3.6 G) plus 14% of a species with characteristic phenoxyl radical hyperfine structure (aoH = 6.56 G (2H) and amH = 1.79 G (2H)). (C) As in A, but no H2O2. (D) As in (A), but no H2O2 and with addition of 50 µg/ml catalase. (E) As in A, but no phenolphthalein. (F). As in (A), but no HRP.

Phenolphthalein Metabolism in the Presence of Glutathione. Glutathione thiyl free radical formation was mediated by lactoperoxidase-catalyzed metabolism of phenolphthalein as demonstrated by spin-trapping with DMPO. A four-line EPR spectrum was observed in an incubation that contained phenolphthalein, GSH, lactoperoxidase, and hydrogen peroxide at pH 7.4 (fig. 3A). Computer analysis and simulation of the characteristic four-line EPR spectrum, corresponding to hyperfine coupling constants a^N = 15.1 and a^H = 16.2 G, was consistent with formation of the DMPO/SG (28, 29). If phenolphthalein was not present in the incubation, DMPO/^·SG was detected at a much lower concentration owing to direct oxidation of GSH by lactoperoxidase (28, 29) (fig. 3B). In the absence of any of the incubation constituents (GSH, DMPO, or lactoperoxidase), no free radicals were detected (figs. 3C, 3D, and 3E, respectively). The omission of hydrogen peroxide resulted in a much lower concentration of DMPO/SG (fig. 3F). These EPR results are consistent with the formation of a phenolphthalein phenoxyl metabolite and the subsequent oxidation of GSH to its thiyl radical metabolite (scheme 2).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   ESR spectrum of the DMPO/glutathione thiyl radical adduct detected in a solution containing GSH, phenolphthalein, hydrogen peroxide and lactoperoxidase. (A) Complete system containing 10 µM phenolphthalein, 32 µM H2O2, 1 mM GSH, 10.6 µg/ml (0.85 U/ml) lactoperoxidase and 100 mM DMPO in 100 mM Tris-HCl buffer, pH 7.4. The sample also contained about 1% ethanol from the phenolphthalein stock solution. (B). As in (A), but without addition of phenolphthalein. (C). As in (A), but without glutathione. (D) As in (A), but without DMPO. (E) As in (A), but without lactoperoxidase. (F) As in (A), but without hydrogen peroxide. The instrumental conditions were as follows: 20 mW microwave power, 1 G modulation amplitude, 41 ms time constant and a scan rate of 0.833 G/s.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Effect of catalase and SOD on the ESR spectrum of the DMPO/glutathione thiyl radical adduct in a solution containing GSH, phenolphthalein, hydrogen peroxide, and lactoperoxidase. (A) Complete system as in fig. 3A. (B) As in (A), but with 150 µg/ml (6,500 U/ml) catalase. (C) As in (A), but without hydrogen peroxide. (D) As in (A), but without hydrogen peroxide and with the addition of 150 µg/ml (6500 U/ml) catalase added. (E) As in (A), but with 150 µg/ml (750 U/ml) SOD added. The instrumental conditions were as follows: 20 mW microwave power, 1 G modulation amplitude, 41 ms time constant and a scan rate of 0.833 G/s.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Effects of ascorbate and DMPO on the rate of oxygen consumption during the oxidation of phenolphthalein by lactoperoxidase in the presence of GSH. The two sets of vertical bars mark the time of reaction initiation by the addition of lactoperoxidase and subsequent additions (if any) approximately 1 min later. (A) Complete system containing 10 mM reduced glutathione, 10 µM phenolphthalein, 200 µM H2O2, and 60 µg/ml lactoperoxidase in 100 mM Tris-HCl buffer, pH 7.4, in a final volume of 1.8 ml. The reaction was initiated by the addition of lactoperoxidase. (B) As in (A), but with 100 mM DMPO added 1 min before initiation. (C) As in (A), but with 1 mM ascorbate added 1 min after initiation. (D) As in (A), but with 100 mM DMPO added 1 min after initiation. (E) As in (A), but with 1 mM ascorbate added 1 min before initiation. (F) As in (A), but in the absence of phenolphthalein.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Effect of catalase and SOD on the rate of oxygen consumption during the oxidation of phenolphthalein by lactoperoxidase in the presence of GSH. The two vertical bars mark the time of reaction initiation by the addition of lactoperoxidase and subsequent additions (if any) approximately 1 min later. (A) Complete system as in fig. 5A. The reaction was initiated by the addition of lactoperoxidase. (B) As in (A), but without H2O2. (C) As in (A), but with 90 µg/ml (450 U/ml) SOD added 1 min after initiation. (D) As in (A), but with 90 µg/ml (450 U/ml) SOD added 1 min before initiation. (E) As in (A), but with 75 µg/ml (3250 U/ml) catalase added 1 min after initiation.

Phenolphthalein Metabolism in the Presence of NADH. A reaction mixture containing NADH, phenolphthalein, lactoperoxidase, and DMPO exhibited an EPR spectrum characteristic of the DMPO/superoxide radical adduct (fig. 7A). This spectrum could be simulated with hyperfine coupling constant values in close agreement with those previously published (30, 31): a^N = 14.1, a^H = 11.3, and a^H = 1.2 G. When phenolphthalein was omitted from the reaction mixture, the DMPO/superoxide radical adduct decreased over threefold (fig. 7B). No signal was detected in the absence of either NADH (fig. 7C) or lactoperoxidase (fig. 7D). The signal was inhibited in a concentration-dependent manner by catalase, which showed that trace amounts of hydrogen peroxide, probably as a consequence of NADH autoxidation, were present and required for the reaction (figs. 7E and 7F). The generation of hydrogen peroxide in solutions of NADH has been noted in other peroxidase investigations (32). DMPO/superoxide radical adduct was inhibited in a concentration-dependent manner by SOD, with 5 U/ml sufficient for ca. 60% inhibition (fig. 7G) and 98 U/ml sufficient for complete inhibition (fig. 7H).

View larger version (40K): [in this window] [in a new window]   Fig. 7.   ESR spectrum of the DMPO-superoxide radical adduct detected in a solution containing NADH, phenolphthalein, and lactoperoxidase. (A) Complete system containing 10 µM phenolphthalein, 10 mM NADH, 21.2 µg/ml (1.8 U/ml) lactoperoxidase and 100 mM DMPO in 100 mM Tris-HCl buffer, pH 7.4. The sample also contained about 1% ethanol from the phenolphthalein stock solution. (B) As in (A), but without addition of phenolphthalein. (C) As in (A), but without NADH. (D) As in (A), but without lactoperoxidase. (E) As in (A), but with the addition of 50 µg/ml (2170 U/ml) catalase. (F) As in (A), but with the addition of 100 µg/ml (4340 U/ml) catalase. (G) As in (A), but with the addition of 0.08 µg/ml (5.2 U/ml) SOD. (H) As in (A), but with the addition of 1.5 µg/ml (98 U/ml) SOD. The instrument conditions were as follows: 20 mW microwave power, 1 G modulation amplitude, 21 ms time constant, and a scan rate of 1.11 G/s.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Effect of ascorbate and DMPO on the rate of oxygen consumption during the oxidation of phenolphthalein by lactoperoxidase in the presence of NADH. The two sets of larger vertical bars mark the time of reaction initiation by addition of lactoperoxidase and subsequent additions (if any) 2 min later. The short vertical bar marks the time of additions (if any) made prior to reaction initiation. (A) Complete system containing 3 mM NADH, 10 µM phenolphthalein, and 20 µg/ml (2.8 U/ml) lactoperoxidase in 100 mM Tris-HCl buffer, pH 7.4, in a final volume of 1.8 ml. The reaction was initiated by the addition of lactoperoxidase. (B) As in (A), but with 100 mM DMPO added 2 min after initiation. (C) As in (A), but with 1 mM ascorbate added 2 min after initiation. (D) As in (A), but with 100 mM DMPO added 1 min before initiation. (E) As in (A), but in the absence of phenolphthalein. (F) As in (A), but with 1 mM ascorbate added 1 min before initiation.

View larger version (19K): [in this window] [in a new window]   Fig. 9.   Effect of catalase and SOD on the rate of oxygen consumption during the oxidation of phenolphthalein by lactoperoxidase in the presence of NADH. The two sets of larger vertical bars mark the time of reaction initiation by addition of lactoperoxidase and subsequent additions (if any) 2 min later. The short vertical bar marks the time of additions (if any) made prior to reaction initiation. (A) Complete system as in Figure 8A. (B) As in (A), but with 90 µg/ml (450 U/ml) SOD added 2 min after initiation. (C) As in (A), but with 50 µg/ml (3250 U/ml) catalase added 2 min after initiation. (D) As in (A), but in the absence of phenolphthalein. (E) As in (A), but with 90 µg/ml (450 U/ml) SOD added 1 min before initiation. (F) As in (A), but with 50 µg/ml (3250 U/ml) catalase added 1 min before initiation.

Phenolphthalein Metabolism in the Presence of Ascorbate. Ascorbate is facilely oxidized by a one-electron transfer to produce the relatively stable ascorbate anion radical, which can be detected directly by its EPR spectrum. A reaction mixture of phenolphthalein, ascorbate, lactoperoxidase, and hydrogen peroxide gave a strong EPR spectrum that was characteristic of the ascorbate anion radical (fig. 10A). The observed two-line EPR spectrum had the characteristic hyperfine coupling constant a^H = 1.79 G in agreement with the 1.76 G reported in the literature (34). When the reaction mixture included GSH, no change was observed in the EPR signal (fig. 10B), which indicated that the phenolphthalein phenoxyl radical reacted with ascorbate instead of GSH. When phenolphthalein was omitted from the incubation, the ascorbate anion radical decreased twofold (fig. 10C). Omission of lactoperoxidase from the reaction mixture (fig. 10D) produced only the residual ascorbate radical anion signal characteristic of the ascorbic acid solution in buffer (fig. 10G). Omission of ascorbate from the reaction mixture produced no EPR signal at all (fig. 10E), and omission of hydrogen peroxide from the incubation produced only a weak, transient EPR signal, presumably from trace amounts of hydrogen peroxide (fig. 10F). When the reaction occurred in the absence of added hydrogen peroxide but with added catalase (to remove traces of hydrogen peroxide), only the residual ascorbate radical anion signal was observed (fig. 11C). The presence of SOD in the reaction mixture had no effect on the EPR signal intensity (fig. 11D), which suggested that superoxide was not involved in the reactions that produce the ascorbate radical anion.

View larger version (19K): [in this window] [in a new window]   Fig. 10.   ESR spectrum of ascorbate anion radical generated during the oxidation of phenolphthalein by lactoperoxidase. (A) Sample containing 5 µM phenolphthalein, 1 mM ascorbate, 16 µM hydrogen peroxide and 2.65 µg/ml (0.23 U/ml) lactoperoxidase. (B) As in (A), but with 1 mM glutathione added. (C) As in (A), but without addition of phenolphthalein. (D) As in (A), but without lactoperoxidase. (E) As in (A), but without ascorbate. (F) As in (A), but without hydrogen peroxide. (G) 1 mM ascorbate in buffer. The instrumental conditions were as follows: 20 mW microwave power, 1 G modulation amplitude, 160 ms time constant and a scan rate of 0.333 G/s. The field center was offset to higher field so the ascorbate ESR spectrum would be scanned early in the spectral sweep.

View larger version (17K): [in this window] [in a new window]   Fig. 11.   Effect of catalase and SOD on the ascorbate anion radical generated during the oxidation of phenolphthalein by lactoperoxidase. (A) Complete system as in (10A). (B) As in (A), but without hydrogen peroxide. (C) As in (A), but without hydrogen peroxide and with 50 µg/ml (3250 U/ml) catalase added. (D) As in (A), but without addition of lactoperoxidase. (E) As in (A), but with 1.5 µg/ml (7.5 U/ml) SOD added. The instrumental conditions were the same as in fig. 10.

Metabolism of Phenolphthalein Glucuronic Acid. Glutathione thiyl free radical was mediated by lactoperoxidase-catalyzed metabolism of phenolphthalein glucuronide as demonstrated by spin trapping with DMPO. As with phenolphthalein, a four-line EPR spectrum of DMPO/^·SG was observed in an incubation that contained phenolphthalein glucuronide, GSH, lactoperoxidase, and hydrogen peroxide (fig. 12A). If phenolphthalein glucuronide was not present in the incubation, the DMPO/SG was detected at a lower concentration owing to the direct oxidation of GSH by lactoperoxidase (28, 29) (fig. 12B). When any of the incubation constituents (DMPO, GSH, or lactoperoxidase) was absent, no free radicals were detected (figs. 12C, 12D, and 12E, respectively). The omission of hydrogen peroxide resulted in a much lower concentration of DMPO/SG (fig. 12F). These EPR results are consistent with the formation of a phenolphthalein glucuronide phenoxyl metabolite and the subsequent oxidation of GSH to its thiyl radical metabolite.

View larger version (17K): [in this window] [in a new window]   Fig. 12.   ESR spectrum of the DMPO/glutathione thiyl radical adduct detected in a solution containing GSH, phenolphthalein glucuronide, hydrogen peroxide and lactoperoxidase. (A) Complete system containing 50 µM phenolphthalein glucuronide, 16 µM hydrogen peroxide, 1 mM GSH, 6.5 µg/ml lactoperoxidase and 100 mM DMPO in 100 mM Tris-HCl buffer, pH 7.4. (B) As in (A), but without addition of phenolphthalein glucuronide. (C) As in (A), but without DMPO. (D) As in (A), but without GSH. (E) As in (A), but without lactoperoxidase. (F) As in (A), but without hydrogen peroxide. The instrument conditions were as follows: 20 mW microwave power, 1 G modulation amplitude, 41 ms time constant and a scan rate of 0.833 G/s.

View larger version (17K): [in this window] [in a new window]   Fig. 13.   Effect of catalase and SOD on the rate of oxygen consumption during the oxidation of phenolphthalein glucuronide by lactoperoxidase in the presence of NADH. The two sets of larger vertical bars mark the time of reaction initiation by addition of lactoperoxidase and subsequent additions (if any) 2 min later. The short vertical bar marks the time of additions (if any) made prior to reaction initiation. (A) Complete system containing 3 mM NADH, 100 µM phenolphthalein glucuronide, and 32.2 µg/ml (2.8 U/ml) lactoperoxidase in 100 mM Tris-HCl buffer, pH 7.4, in a final volume of 1.8 ml. The reaction was initiated by the addition of lactoperoxidase. (B) As in (A), but with 90 µg/ml (450 U/ml) SOD added 2 min after initiation. (C) As in (A), but with 50 µg/ml (3250 U/ml) catalase added 2 min after initiation. (D) As in (A), but with 90 µg/ml (450 U/ml) SOD added 1 min before initiation. (E) As in (A), but with 50 µg/ml (3250 U/ml) catalase added 1 min before initiation.

View larger version (17K): [in this window] [in a new window]   Fig. 14.   Comparison of the rates of oxygen consumption during the oxidation of phenolphthalein and phenolphthalein glucuronide in the presence of GSH or NADH. The vertical bar marks the time of reaction initiation by addition of lactoperoxidase. (A) Oxygen uptake was determined in a complete system of 10 µM phenolphthalein, 10 mM GSH, 60 µg/ml lactoperoxidase, and 200 µM hydrogen peroxide. (B) Oxygen uptake was determined in a complete system of 10 µM phenolphthalein, 3 mM NADH, and 20 µg/ml lactoperoxidase. (C) Same as (A), except phenolphthalein was replaced by 100 µM phenolphthalein glucuronide. (D) Same as (B), except phenolphthalein was replaced by 100 µM phenolphthalein glucuronide.

	Discussion

Abstract Introduction Results Discussion References

Phenolphthalein (fig. 1) has two phenolic substituents that would be expected to undergo reactions characteristic of phenols. When amines and phenols are oxidized in a peroxidase-hydrogen peroxide system, varied polymeric products are formed (35). Because the phenolic moieties of phenolphthalein lack substituents ortho to the hydroxyl function, phenolphthalein should be oxidized readily to relatively unstable phenoxyl free radicals by chemical and biochemical oxidizing agents. The EPR results for fast-flow studies of phenolphthalein oxidation by HRP (fig. 2A) confirmed the observation of a primary phenolphthalein phenoxyl radical. A large fraction (86%) of the observed signal was a result of a second species with a large line width and a predominantly Gaussian line shape. This signal probably arises from polymeric free radical species that were formed from the primary phenoxyl free radical. The smaller fraction (ca. 14%) of the observed signal arises from a phenoxyl radical with hyperfine coupling constants as follows: a_o^H = 6.56 G (2H) and a_m^H = 1.79 G (2H). Typical literature values for phenoxyl hyperfine coupling constants are ca. 6 G for ortho protons and ca. 2 G for meta protons (36, 37), in close agreement with these results for the phenolphthalein system.

Evidence for the peroxidase-mediated metabolism of phenolphthalein to form phenoxyl radicals is supported by the direct EPR observation of a phenoxyl radical (fig. 2). Additional support is provided by EPR observations of the indirect consequences of phenoxyl radical production such as the production of glutathione thiyl (figs. 3 and 4), superoxide (fig. 7), and ascorbate radicals (figs. 10 and 11).

Scheme 1 represents the peroxidase-mediated metabolism of phenolphthalein to form phenolphthalein phenoxyl radicals. Each cycle of the peroxidase in scheme 1 results in the conversion of two phenolic substituents into phenoxyl radicals. If some subsequent reaction occurs to reform the original phenol species, the possibility for a cyclic chain reaction exists. In biological systems, the reducing equivalents required to reduce the phenoxyl radical can be provided by the endogenous reductants GSH, NADH, and ascorbate. Depletion of these species in vivo is evidence of oxidative stress. The extent of endogenous reductant free radical enhancement is seen in figs. 3, 7, and 10 by comparing the intensity of the EPR signal that is generated in the presence of phenolphthalein in the system with that generated in its absence. In the presence of endogenous reductants, the phenolphthalein phenoxyl radical is reduced to the parent compound, and subsequent product formation such as polymers will be prevented.

Scheme 2 outlines the sequence of possible reactions that can lead to oxidative stress (23). In the presence of GSH and in agreement with the scheme, metabolism of phenolphthalein led to enhanced formation of the glutathione thiyl radical and the stimulation of oxygen consumption in the glutathione/lactoperoxidase/hydrogen peroxide system. Some oxygen may also be consumed by the addition of the glutathione thiyl radical to oxygen to produce the glutathione peroxyl radical (GSOO^·) (38, 39). Most glutathione thiyl radical is thought to react with another GSH (GS^-) to form the glutathione disulfide radical anion, which then reacts with molecular oxygen to produce oxidized glutathione and superoxide radical anion (23).

When phenolphthalein is metabolized by lactoperoxidase in the presence of NADH, as described in scheme 2, more superoxide radical anion is trapped by DMPO than in a system lacking phenolphthalein (figs. 7A and 7B). Oxygen consumption in this system is suppressed at SOD concentrations as low as 0.075 U/ml, which indicates that superoxide is produced in a chain reaction mechanism. Phenolphthalein phenoxyl radical initiates the chain reaction by reacting with NADH to produce NAD^·. On protonation, superoxide can oxidize additional NADH with a second-order rate constant of k = 1.8 × 10⁵ M^-1 s^-1, thus producing NAD^· and continuing the chain reaction (40). Other phenolic compounds such as scopoletin (31) and acetaminophen (41) are known to stimulate NADH oxidation by horseradish peroxidase.

When ascorbate is added to either the phenolphthalein/GSH/lactoperoxidase system or the phenolphthalein/NADH/lactoperoxidase system, oxygen consumption is prevented almost entirely if the addition is before initiation of the reaction or suppressed almost immediately if the addition is made after initiation of the reaction. Ascorbate radical anion production is found to increase in a phenolphthalein-dependent manner during the metabolism of phenolphthalein by lactoperoxidase. These observations are consistent with the formation of a phenolphthalein phenoxyl radical intermediate during the metabolism as shown in scheme 2. The ascorbate radical anion does not react with oxygen and, therefore, cannot contribute to increased oxidative stress.

The principal metabolite of phenolphthalein, phenolphthalein glucuronide, has a glucuronic acid moiety conjugated to one of the phenolic hydroxyl groups of phenolphthalein, while the second phenolic hydroxyl group remains unsubstituted and potentially available to form a phenoxyl radical. In these studies phenolphthalein glucuronide underwent reactions qualitatively similar to those observed for phenolphthalein itself. It stimulated the formation of glutathione thiyl radical, superoxide, and ascorbate radical anion and showed qualitatively the same oxygen consumption behavior as phenolphthalein (fig. 13). Although phenolphthalein glucuronide was significantly less active than phenolphthalein itself in these systems modeling oxidative stress (fig. 14), its in vivo concentration is over 100 times greater and either compound could be a source of in vivo free radical formation.

The reason for reduced activity of the glucuronide is not known, but Shiga and Imaizumi (17) reported that formation of phenoxyl radical by peroxidase oxidation depended not on the molecular sizes of the phenols but on their redox potentials. The presence of two oxidizable phenol groups in phenolphthalein, but only one in its glucuronide, is probably also a factor.

The results presented in these studies demonstrate that futile metabolism of micromolar quantities of phenolphthalein and phenolphthalein glucuronide can catalyze the oxidation of much greater concentrations of biochemical reducing cofactors such as ascorbate, glutathione, and NAD(P)H, with concomitant production of superoxide by subsequent reaction with oxygen of glutathione- and NAD(P)H-derived free radicals.

In the presence of catalytic SOD, superoxide is converted rapidly to hydrogen peroxide and molecular oxygen. The hydrogen peroxide so formed can react with peroxidases reinitiating the whole process (schemes 1 and 2). In the presence of reduced iron, the hydrogen peroxide thus produced can be a source of hydroxyl radical as a consequence of Fenton chemistry. Highly reactive hydroxyl radical has been found to react with DNA to produce many different products (42). In addition, oxygen-derived radicals have a pivotal role in tumor promotion (43-45). Although the facile free radical metabolism of phenolphthalein is suggestive, no evidence exists linking this free radical formation to the carcinogenicity of phenolphthalein.

The futile nature of this free radical metabolism, which does not lead to stable products, calls into question the very existence of this metabolism in vivo. In fact, no metabolites of phenolphthalein other than its glucuronide have been reported. Nevertheless, there is indirect evidence that phenolphthalein is an in vivo peroxidase substrate.

A number of investigations have found that phenolphthalein stimulates intestinal prostaglandin formation. In fact, the laxative effect of phenolphthalein is proposed to be the result of this increased prostaglandin formation (46-49). The ability of compounds to stimulate prostaglandin formation correlates directly with their ability to act as peroxidase substrates (50). The observation that phenolphthalein stimulates prostaglandin formation in vivo, thereby causing the laxative effect of phenolphthalein, is indirect evidence that phenolphthalein is metabolized to a phenoxyl free radical in vivo.