l-2-Oxothiazolidine-4-carboxylic acid (OTZ), a 5-oxoproline analog, is metabolized by 5-oxoprolinase and converted to cysteine, the rate-limiting amino acid for GSH synthesis, with the release of CO2. [13C]OTZ (1.5 mg/kg) was used in 12 healthy men and women (ages 23–73 yr) to indirectly assess precursor mobilization for GSH synthesis when stores were reduced by 2 g acetaminophen. Expired breath samples were analyzed for13CO2, and results were analyzed using noncompartmental and two-compartment open minimal models. Results show an increase in13C excretion (higher OTZ hydrolysis) when GSH stores were reduced and 5-oxoprolinase substrate utilization patterns, consequently, were altered (P < 0.01). A metabolic rate index (MRI) of the OTZ probe was found to be significantly higher after reduction of GSH content by acetaminophen (P < 0.05). The difference in adaptive capacity (difference between control and postacetaminophen metabolic rate indexes) was two times as large in the young than the old subjects (P < 0.01). These data support the use of [13C]OTZ as a probe to identify individuals who may be at risk for low GSH stores or who have an impaired capacity to synthesize GSH.
- cysteine prodrug
- breath test
l-2-oxothiazolidine-4-carboxylic acid(OTZ) is a prodrug of cysteine, the amino acid believed to be rate limiting for GSH synthesis (15, 16). OTZ is a 5-oxoproline analog and therefore a substrate for the ubiquitous intracellular enzyme called 5-oxoprolinase (15, 16, 32). 5-Oxoprolinase catalyzes the reaction converting 5-oxoproline, also known as pyroglutamate, to glutamate, which is required for the first step for GSH synthesis (Fig.1). 5-Oxoprolinase links the synthesis pathway and the catabolism pathway of GSH in the γ-glutamyl cycle (15, 16). OTZ, when acted upon by 5-oxoprolinase, is converted to l-cysteine with the release of CO2 (Fig.2). A number of studies have demonstrated that OTZ administration can stimulate hepatic GSH formation both in vitro and in vivo (2, 6, 26, 31). OTZ has also been used to increase intracellular GSH stores in normal and human immunodeficiency virus-infected human volunteers (8, 19) and, more recently, to sensitize tumor cells to melphalan, a chemotherapeutic agent (3).
GSH, a ubiquitous cellular constituent, accounts for 90% of nonprotein thiols in mammalian tissues and plays an important role in the detoxification of endogenous and exogenous compounds [e.g., free radicals, peroxides, chemotherapeutic agents such as cisplatin or doxorubicin, or drugs such as acetaminophen (APAP); see Ref. 15]. GSH deficiency produced in animals leads to death within a few days (14, 15), and GSH depletion has recently been implicated in hyperglycemia-induced embryopathy (27). GSH also serves as the principal antioxidant for mitochondria, and low GSH levels are associated with increased oxidative damage to mitochondria (14). Low serum GSH levels are also reported in aging (10, 21) and are implicated in the free radical theory of aging.
Hereditary defects have been described in four of the six enzymes in the γ-glutamyl cycle (22). A deficiency of GSH synthetase, leading to a marked reduction in cellular levels of GSH, results in increased formation of γ-glutamylcysteine, which is efficiently converted to 5-oxoproline when not utilized for GSH synthesis or for transpeptidation reactions. Because 5-oxoprolinase is widely distributed in human tissues, indirect measurement of its activity using a 5-[13C]oxoproline analog such as OTZ ([13C]OTZ) and quantitating 13C accumulation in expired CO2 is expected to be an index of activity within the GSH synthesis cycle. Because 5-oxoprolinase has a low Michaelis constant (8 ± 2 μM; see Ref.17), its activity may be viewed as an indicator of the flux through the GSH cycle (4). Under conditions of adequate GSH, precursors from the transpeptidase reaction provide sufficient natural substrate (5-oxoproline) for 5-oxoprolinase, resulting in a lower rate of [13C]OTZ conversion and lower cumulative appearance of13CO2in breath. When GSH demand is high (or GSH stores low), flow is shunted toward the synthesis of γ-glutamylcysteine (15), leaving the 5-oxoprolinase pathway open for OTZ metabolism (31) and resulting in the appearance of higher amounts of13CO2. This paradigm is the basis for the use of [13C]OTZ as a probe for precursor mobilization for GSH synthesis. In this study, we used [13C]OTZ to indirectly assess precursor (cysteine) mobilization for GSH synthesis when GSH stores are reduced by an acute ingestion of APAP.
MATERIALS AND METHODS
Materials. [13C]OTZ (patent pending) was obtained from MassTrace (Woburn, MA) and was administered under investigator-sponsored Investigational New Drug (IND) application no. 46,237. [13C]bicarbonate was also obtained from MassTrace. Ammonium 7-fluoro-benzo-2-oxa-1,3-diazole-4-sulforate (SBD-F) was obtained from Wako (Kyoto, Japan). GSH (reduced form) was purchased from Sigma Chemical and was administered under investigator-sponsored IND application no. 42,899. APAP (McNeil Laboratories, Fort Washington, PA) is an effective analgesic and antipyretic agent available over the counter and widely used by all age groups. Toxic does of APAP can cause hepatocellular necrosis (20), and the mechanism for hepatic injury is believed to be mediated by a reactive metabolite of APAP,N-acetyl-p-benzoquinoneimine, which depletes hepatocellular GSH (28).
Subjects. Twelve nonsmoking, nonobese men (n = 8) and women (n = 4), age range 23–73 yr, in general good health volunteered to participate in these studies. They were recruited from the community and selected on the basis of a medical history, thorough physical examination, and normal blood and urine tests. None was on medications known to alter GSH status. None of the postmenopausal women were on hormone replacement therapy. The premenopausal woman was studied in the follicular phase of her menstrual cycle to minimize possible variation in metabolic rate and CO2 production between studies.
Study design. Each volunteer was studied on two separate occasions. The volunteers were admitted to the General Clinical Research Center 2 days before the study and were provided with a weight-maintaining diet providing 1 g protein ⋅ kg−1 ⋅ day−1and 45–50% of energy from carbohydrates. On the morning of the study after a 10- to 12-h overnight fast, an intravenous line was placed for blood sampling. Baseline blood and expired breath samples were collected and handled as previously described (7). In one study, the volunteers received [13C]OTZ (1.5 mg/kg or 0.101 mmol/kg) orally, and expired breath and blood samples were obtained at timed intervals for 5 h after ingestion. In the second study, volunteers received 2 g APAP 60 min before the [13C]OTZ dose. Baseline blood samples in study 2 were obtained immediately before and 60 min after the APAP dose. Indirect calorimetry was obtained four times within the 5-h study to determine rates of CO2 production (Deltatrac; Datex, Helsinki, Finland).
Five of the 12 volunteers underwent a third study during which time they received a 5-h intravenous infusion of GSH (reduced form; Sigma Chemical, St. Louis, MO) at 10 mg/min (∼32 ± 2 μmol ⋅ kg−1 ⋅ h−1) immediately after drinking the [13C]OTZ. Three of the 12 volunteers were also studied with [13C]bicarbonate (0.101 mmol/kg) instead of OTZ with and without APAP.
Analytic methods. Expired breath samples were analyzed using isotope ratio-mass spectrometry for13C enrichment of expired CO2 [atoms percent excess (APE)]. Plasma amino acid concentrations were measured after precolumn derivatization with phenylisothiocyanate and separation by reversed-phase column HPLC with detection at 254 nm (Pico-Tag Analysis System; Waters, Milford, MA). Plasma GSH was measured after reduction and decoupling from plasma proteins with tri-n-butylphosphine, derivatization with SBD-F, and separation by reversed-phase HPLC with fluorometric detection (Waters Chromatography), as previously described (1).
Calculations. The appearance of13CO2in breath after an oral bolus (Fig. 3) is characterized by a rapid, almost linear rise after a short lag time to accommodate variable gastric emptying, followed by an exponential decay, which when plotted logarithmically appears to be at least biphasic. The goal of this study was to describe this mass flow of tracer from [13C]OTZ to CO2 enrichment vs. time in a polyexponential functional form. This would enable us to1) determine the overall area under the curve (AUC) and net recovery of tracer as a percentage of dose,2) describe the postabsorptive enrichment changes independent of the lag and absorption phases, and3) derive an index of OTZ conversion outside of the central (plasma) compartment as the correlate for the effect of the APAP “insult.”
Pharmacokinetic analysis by nonlinear, weighted least squares (25) of multiple exponential models showed that the metabolic process of OTZ hydrolysis in vivo is best described by the expression for zero-order absorption and biexponential decay (9) Equation 1during tracer absorption when time (t, min) is less than or equal to the zero-order absorption time (T); and Equation 2during tracer disappearance. Caand Cd are13C enrichments in breath expressed as APE times 1,000, lag is the time before the start of absorption, andC 1,C 2,k 1,k 2,T, and lag are the unknown parameters to be solved.
This numeric model (29) was chosen over all others, including those with first-order absorption of tracer, because it afforded the lowest root mean squared scaled by degrees of freedom for each model (SE), lowest Akaike Information Criterion (AIC) for adequacy of the exponential term selection, highestF-test value for goodness of fit, and normally distributed residuals, especially during the absorption phase (30). The weighting scheme selected was based on the squared inverse of the dependent variable. Representative fitting values on a cohort of six young male subjects, including the data depicted in Fig. 3[average fitted parameter, average coefficient of variation (CV) of individual weighted fits, and CV of average fits, respectively] were 1) for the basal or pre-APAP:C 1, 66.5 (12 and 35%); k 1, 0.0088 (12 and 18%);C 2, 57.3 (15 and 16%); k 2, 0.0466 (13 and 47%); T, 43.2 (5 and 36%); lag, 11.6 (3 and 100%); and 2) for the post-APAP period:C 1, 59.2 (11 and 30%); k 1, 0.0080 (15 and 20%);C 2, 31.2 (14 and 54%); k 2, 0.0222 (10 and 43%); T, 48.2 (7 and 30%); lag, 9.2 (5 and 52%). For this numeric model, the weighted SE, AIC, and F-test (data points, parameters) were as follows: pre-APAP, 0.06, 0.87, and 43,000; post-APAP, 0.13, 0.95, and 39,000, respectively. For the corresponding monoexponential, zero-order absorption model, the commensurate goodness-of-fit criteria were as follows: pre-APAP, 1.3, 2.6, and 1,600 and post-APAP, 2.1, 2.9, and 1,200, respectively. Similar mono-, bi-, and triexponential models with first-order absorption on average showed unacceptable fitting characteristics as follows: pre-APAP, >4, >3, and <800; post-APAP, >3, >3, and <1,000.
In terms of a physiological, compartmental description, the functional form described above is consistent with the kinetics of xenobiotic disposal shown by the minimal two-compartment open model with central compartment elimination, insensible losses by alternate routes of disposal, and zero-order input, depicted in Fig.4.
According to Wagner's derivation (29, 30), the postabsorptive tracer kinetics of OTZ metabolism are readily described by stripping away the effects of lag and absorption. Thus rearranging and combining Eqs. 1 and 2 yields Eq.3 Equation 3 in which Cm represents the rate of tracer disappearance from compartments 1 and 2 that would result from an instantaneous dose input intocompartment 1.
The percentage of the OTZ dose recovered in breath was computed from the cumulative integral of Eqs. 1 and 2 taken as a continuous function from 0 to 300 min multiplied by the CO2production rate of each volunteer measured by indirect calorimetry. This recovery factor (Rf) was used to adjust the amount of the input tracer dose in subsequent calculations for the decrementing effect of insensible losses (urinary excretion of OTZ, biological fixation of13CO2) upon the mass of13CO2metabolite in the system. We defined the metabolic rate index (MRI, nmol ⋅ kg−1 ⋅ min−1) for OTZ as the dose per kilogram body weight (D) of OTZ, adjusted for insensible losses (Rf), multiplied by the fractional catabolic rate outside of the central compartment (FCRmc, min−1), a parameter equivalent to the inverse of the mean residence time (MRTmc, min) of tracer outside the central compartment, and determined it by noncompartmental manipulation of Eq. 3 , in accordance with Eq. 4 Equation 4
For noncompartmental analysis, the area under the curve (AUC) was computed for the13CO2enrichment versus time graph by numeric integration, and the apparent replacement rate (RR, mmol/min; or clearance equivalent of CO2 derived from OTZ) was computed as the tracer dose divided by AUC. The apparent metabolic distribution pool (MD, mmol) was calculated as the tracer dose times AUMC divided by AUC2, and system mean residence time (MRT, min) was calculated as AUMC divided by AUC where AUMC is the area under the moment curve of time multiplied by enrichment vs. time.
Statistics. Results are presented as means ± SD. Comparisons between tests were made using the pairedt-test. Group differences based on age or gender were compared using a two-samplet-test.
The volunteers (8 men and 4 women) ranged in age from 23 to 73 yr. The women were older (58 ± 14 vs. 36 ± 16 yr) and shorter (163 ± 3 vs. 174 ± 5 cm) than the men but of similar weight (72 ± 7 vs. 78 ± 6 kg) and body mass index (27 ± 3 vs. 26 ± 3 kg/m2). When the subjects were grouped according to age (young 23–41 yr; old ≥53 yr), they differed only in height but not weight or body mass index. Consequently, all volunteers received approximately the same dose of [13C]OTZ (0.75 mmol). Indirect calorimetry results were similar in both studies, as shown in Table 1. Expected differences in the rate of CO2 production and energy expenditure between young and old or men and women were found.
With regard to the noncompartmental kinetics over the entire time course (absorption plus disposition phase), there was a significant increase in the AUC of13CO2excretion and a slight decrease in the apparent RR (clearance) of CO2 derived from the OTZ after APAP compared with the study in which [13C]OTZ was ingested alone (control; Fig. 5). The percentage of the dose excreted 5 h after ingestion did not significantly differ among subjects or studies, averaging ∼75%. The mass of exchangeable bicarbonate (apparent whole body distribution pool, MD) did not differ between the two studies (1,155 ± 181 vs. 1,116 ± 199 mmol) but was significantly lower in women compared with men (962 vs. 1,230 mmol, respectively, P < 0.05) and old compared with young subjects (1,004 vs. 1,230 mmol, respectively,P < 0.01), reflecting the age- and gender-related differences in body metabolic “space” and bicarbonate volume of distribution. Plasma levels of APAP 4 h after ingestion averaged 12 ± 6 μg/ml, with one individual contributing to the variability with a level of 28 μg/ml. Time-to-peak13CO2excretion averaged 45 min in both the OTZ alone (45 ± 10) and OTZ plus APAP (45 ± 14) studies, and MRT did not differ (99 ± 10 vs. 102 ± 10 min, respectively).
In the subset of five individuals who received the exogenous GSH infusion, none of the calculated parameters differed significantly. The AUC in the five individuals studied was 6,484 ± 664 APE ⋅ 10−3 ⋅ min, RR was 12 ± 1 mmol/min, MRT was 104 ± 7 min, MD was 1,308 ± 168 mmol, peak time was 47 ± 14 min, and dose recovery was 76 ± 2%. In the three volunteers who received [13C]bicarbonate instead of [13C]OTZ in separate studies, there was no significant difference in the AUC of13CO2excretion (6,248 ± 733 vs. 6,355 ± 448 APE ⋅ 10−3 ⋅ min), MRT (79 ± 4 vs. 76 ± 6 min), or dose recovery (64 ± 5 vs. 61 ± 3%) without or with APAP, respectively.
Plasma concentrations of glutamic acid, cysteine, glycine, and GSH [54 ± 25, 34 ± 3, 163 ± 16, and 3.5 ± 0.9 μM, respectively] did not vary significantly between studies or over time in the studies with or without APAP. As previously reported, plasma cysteine and glycine both increased during exogenous GSH administration from baseline values of 31 ± 2 and 170 ± 36 to 41 ± 5 and 206 ± 21 μM, respectively, at 150 min after the initiation of the GSH infusion. Plasma GSH concentrations nearly doubled after 180 min of exogenous GSH administration.
From the modeling analysis, the results expressed in terms of nanomoles of OTZ metabolized per kilogram body weight per minute for the basal (control) and post-APAP (+APAP) conditions and the stimulatory difference between +APAP and control are shown in Table2. In all subjects, the MRI of the OTZ probe was significantly higher (P < 0.05) after the GSH-depleting insult (+APAP) than before (control). The young set showed a lower basal MRI and a higher level of OTZ utilization after APAP, consistent with the view that GSH homeostasis was altered by the aging process, possibly because of a more prolonged exposure to oxidant stress with advancing age. The difference in adaptive capacity, as measured by the stimulatory parameter, was two times as large in the young than in the old subjects and was highly significant (P < 0.01).
The amount or rate of appearance of isotopically labeled CO2 generated from a substrate containing the labeled carbon may be indicative of different physiological functions related to the labeled substrate. A classic application is the use of [13C]aminopyrine to distinguish among different degrees of liver disease affecting the hepatic P-450 demethylation of [13C]aminopyrine and subsequent production of13CO2(18). Investigators have recently used labeled α-keto analogs of the branched-chain amino acids to probe hepatic mitochondrial function by measuring the exhalation of labeled CO2 (11). Taken as a whole in the context of these various breath test diagnostic paradigms, the present data support the hypothesis that the [13C]OTZ effectively tracks the course of GSH depletion, even after a mild insult, and can also be used to estimate the relative capacity of the GSH production machinery in cohorts with different oxidant stress histories. Under basal (normal) conditions, intracellular GSH titers are sufficiently high to inhibit γ-glutamylcysteine synthesis and GSH synthesis so that the substrate flux of glutamyl precursors can be expected to be shunted through the transpeptidase and cyclotransferase enzymes, producing oxoproline and presumably saturating the 5-oxoprolinase gateway. If a xenobiotic analog of oxoproline, namely [13C]OTZ, is introduced under these basal conditions, a lower proportion of it will be hydrolyzed in the presence of a surplus of endogenous oxoproline; therefore, the appearance of label derived from OTZ will be subject to greater dilution and correspondingly lower enrichment. However, in response to an acute GSH-depleting stimulus, renewed synthesis of γ-glutamylcysteine and of GSH draws the glutamyl precursors away from the substrate cycling shunt, presumably to replenish both intracellular and extracellular GSH losses. Under these circumstances, intracellular titers of oxoproline diminish; therefore, a greater proportion of OTZ is hydrolyzed to13CO2by 5-oxoprolinase. Because the expression and concentration of 5-oxoprolinase is unaffected by GSH status, a higher rate of13CO2appearance and concomitant increase in the area under the breath test curve were found.
This demonstrates for the first time the possibility of monitoring the metabolism of an oral tracer as an index for precursor mobilization in a ubiquitous metabolic pathway, i.e., the γ-glutamyl cycle leading to GSH synthesis. These data provide support for the efficacy of larger doses of OTZ in stimulating cysteine production for GSH synthesis in conditions where the body's GSH stores are low (12). This unique tracer may act as an adjunct to monitoring GSH stores in patients at risk for GSH depletion and perhaps in following the response to therapeutic agents designed to replete low GSH stores (5, 8).
By extending the application of the OTZ probe to therapy management, it should be possible, in a similar manner, to characterize individuals both according to their basal OTZ MRI and to their production capacity in response to a controlled insult such as that provided by APAP. Patients with a high basal MRI or a low stimulation score might be less likely to benefit from antioxidant and related therapies aimed at boosting the GSH pool and might be the most susceptible to harm from treatments such as chemotherapy, whose coincident biochemical effects include chronic GSH depletion. Patients with higher basal MRI scores or lower stimulation scores may sustain higher morbidity and mortality.
As shown in earlier studies, we did not find a change in plasma GSH levels in this study. The dose of OTZ administered has been shown not to influence free and total GSH concentrations in plasma but did increase GSH levels in the lymphocytes (19). A number of factors makes it unlikely that there would have been a detectable change in plasma GSH over the course of this study, including the size of the dose, the acute nature of the study, the health status of the volunteers, and the well-known compartmentation of GSH in the body (23).
The only significant response induced by the intravenous infusion of GSH was a rise in plasma cysteine levels, as previously reported (7). We previously concluded that essentially all of the exogenous GSH was metabolized with the release of free cysteine into plasma, accounting for the inability to influence endogenous GSH production but reflecting minimal changes in the hydrolysis of OTZ. Designing an experiment whereby GSH stores are in excess and therefore a reduction in OTZ hydrolysis might be anticipated presents a biological challenge. However, the potential for identifying individuals requiring enhanced synthesis of GSH and who might benefit from pharmacological efforts to increase GSH production is exciting. Future studies are needed to address the influence of bicarbonate kinetics on modeling of OTZ metabolism and the use of the tracer in a variety of clinical states. In addition, more rigorous approaches that assess the reciprocal rates of OTZ appearance in plasma and/or urine are needed. Separate investigations will also need to focus on the kinetics of oxoproline itself, for which OTZ is a nonrecycling surrogate. The approach used in this first study offers a convenient shortcut that is predicated on describing the metabolism of a xenobiotic, since OTZ has no endogenous pool but shares a pivotal enzymatic conversion step with the natural substrate. Hence, OTZ can be used as a surrogate of oxoproline, and quantifying its rate of hydrolysis becomes a means to assess the flow of precursor for GSH synthesis and potentially the status of GSH homeostasis.
We thank the volunteers for participation, the General Clinical Research Center staff for assistance, and Amy Prue for help with preparation of the manuscript.
Address for reprint requests and other correspondence: N. K. Fukagawa, Univ. of Vermont, College of Medicine, Given Bldg. Rm. C-207, Burlington, VT 05405-0068 (E-mail:).
This work was supported by National Institutes of Health Grants AG-00599 and RR-00109.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society