Potential impurities in drug substances: Compound-specific toxicology limits for 20 synthetic reagents and by-products, and a class-specific toxicology limit for alkyl bromides.

This paper provides compound-specific toxicology limits for 20 widely used synthetic reagents and common by-products that are potential impurities in drug substances. In addition, a 15 µg/day class-specific limit was developed for monofunctional alkyl bromides, aligning this with the class-specific limit previously defined for monofunctional alkyl chlorides. Both the compound- and class-specific toxicology limits assume a lifetime chronic exposure for the general population (including sensitive subpopulations) by all routes of exposure for pharmaceuticals. Inhalation-specific toxicology limits were also derived for acrolein, formaldehyde, and methyl bromide because of their localized toxicity via that route. Mode of action was an important consideration for a compound-specific toxicology limit. Acceptable intake (AI) calculations for certain mutagenic carcinogens assumed a linear dose-response for tumor induction, and permissible daily exposure (PDE) determination assumed a non-linear dose-response. Several compounds evaluated have been previously incorrectly assumed to be mutagenic, or to be mutagenic carcinogens, but the evidence reported here for such compounds indicates a lack of mutagenicity, and a non-mutagenic mode of action for tumor induction. For non-mutagens with insufficient data to develop a toxicology limit, the ICH Q3A qualification thresholds are recommended. The compound- and class-specific toxicology limits described here may be adjusted for an individual drug substance based on treatment duration, dosing schedule, severity of the disease and therapeutic indication.

Data-driven analysis approach for biomarker discovery using molecular-profiling technologies.

High-throughput molecular-profiling technologies provide rapid, efficient and systematic approaches to search for biomarkers. Supervised learning algorithms are naturally suited to analyse a large amount of data generated using these technologies in biomarker discovery efforts. The study demonstrates with two examples a data-driven analysis approach to analysis of large complicated datasets collected in high-throughput technologies in the context of biomarker discovery. The approach consists of two analytic steps: an initial unsupervised analysis to obtain accurate knowledge about sample clustering, followed by a second supervised analysis to identify a small set of putative biomarkers for further experimental characterization. By comparing the most widely applied clustering algorithms using a leukaemia DNA microarray dataset, it was established that principal component analysis-assisted projections of samples from a high-dimensional molecular feature space into a few low dimensional subspaces provides a more effective and accurate way to explore visually and identify data structures that confirm intended experimental effects based on expected group membership. A supervised analysis method, shrunken centroid algorithm, was chosen to take knowledge of sample clustering gained or confirmed by the first step of the analysis to identify a small set of molecules as candidate biomarkers for further experimentation. The approach was applied to two molecular-profiling studies. In the first study, PCA-assisted analysis of DNA microarray data revealed that discrete data structures exist in rat liver gene expression and correlated with blood clinical chemistry and liver pathological damage in response to a chemical toxicant diethylhexylphthalate, a peroxisome-proliferator-activator receptor agonist. Sixteen genes were then identified by shrunken centroid algorithm as the best candidate biomarkers for liver damage. Functional annotations of these genes revealed roles in acute phase response, lipid and fatty acid metabolism and they are functionally relevant to the observed toxicities. In the second study, 26 urine ions identified from a GC/MS spectrum, two of which were glucose fragment ions included as positive controls, showed robust changes with the development of diabetes in Zucker diabetic fatty rats. Further experiments are needed to define their chemical identities and establish functional relevancy to disease development.

Clustering of hepatotoxins based on mechanism of toxicity using gene expression profiles.

Microarray technology, which allows one to quantitate the expression of thousands of genes simultaneously, has begun to have a major impact on many different areas of drug discovery and development. The question remains of whether microarray analysis and gene expression signature profiles can be applied to the field of toxicology. To date, there are very few published studies showing the use of microarrays in toxicology and important questions remain regarding the predictability and accuracy of applying gene expression profiles to toxicology. To begin to address these questions, we have treated rats with 15 different known hepatotoxins, including allyl alcohol, amiodarone, Aroclor 1254, arsenic, carbamazepine, carbon tetrachloride, diethylnitrosamine, dimethylformamide, diquat, etoposide, indomethacin, methapyrilene, methotrexate, monocrotaline, and 3-methylcholanthrene. These agents cause a variety of hepatocellular injuries including necrosis, DNA damage, cirrhosis, hypertrophy, and hepatic carcinoma. Gene expression analysis was done on RNA from the livers of treated rats and was compared against vehicle-treated controls. The gene expression results were clustered and compared to the histopathology findings and clinical chemistry values. Our results show strong correlation between the histopathology, clinical chemistry, and gene expression profiles induced by the agents. In addition, genes were identified whose regulation correlated strongly with effects on clinical chemistry parameters. Overall, the results suggest that microarray assays may prove to be a highly sensitive technique for safety screening of drug candidates and for the classification of environmental toxins.

Microarray analysis of hepatotoxins in vitro reveals a correlation between gene expression profiles and mechanisms of toxicity.

A rate-limiting step that occurs in the drug discovery process is toxicological evaluation of new compounds. New techniques that use small amounts of the experimental compound and provide a high degree of predictivity would greatly improve this process. The field of microarray technology, which allows one to monitor thousands of gene expression changes simultaneously, is rapidly advancing and is already being applied to numerous areas in toxicology. However, it remains to be determined if compounds with similar toxic mechanisms produce similar changes in transcriptional expression. In addition, it must be determined if gene expression changes caused by an agent in vitro would reflect those produced in vivo. In order to address these questions, we treated rat hepatocytes with 15 known hepatoxins (carbon tetrachloride, allyl alcohol, aroclor 1254, methotrexate, diquat, carbamazepine, methapyrilene, arsenic, diethylnitrosamine, monocrotaline, dimethyl-formamide, amiodarone, indomethacin, etoposide, and 3-methylcholanthrene) and used microarray technology to characterize the compounds based on gene expression changes. Our results showed that gene expressional profiles for compounds with similar toxic mechanisms indeed formed clusters, suggesting a similar effect on transcription. There was not complete identity, however, indicating that each compound produced a unique signature. These results show that large-scale analysis of gene expression using microarray technology has promise as a diagnostic tool for toxicology.

Protein oxidation: examination of potential lipid-independent mechanisms for protein carbonyl formation.

Previous data indicated that diquat-mediated protein oxidation (protein carbonyl formation) occurs through multiple pathways, one of which is lipid dependent, and the other, lipid independent. Studies reported here investigated potential mechanisms of the lipid-independent pathway in greater detail, using bovine serum albumin as the target protein. One hypothesized mechanism of protein carbonyl formation involved diquat-dependent production of H2O2, which would then react with site-specifically bound ferrous iron as proposed by Stadtman and colleagues. This hypothesis was supported by the inhibitory effect of catalase on diquat-mediated protein carbonyl formation. However, exogenous H2O2 alone did not induce protein carbonyl formation. Hydroxyl radical-generating reactions may result from the H2O2-catalyzed oxidation of ferrous iron, which normally is bound to protein in the ferric state. Therefore, the possible reduction of site-specifically bound Fe3+ to Fe2+ by the diquat cation radical (which could then react with H2O2) was also investigated. The combination of H2O2 and an iron reductant, ascorbate, however, also failed to induce significant protein carbonyl formation. In a phospholipid-containing system, an ADP:Fe2+ complex induced both lipid peroxidation and protein carbonyl formation; both indices were largely inhibitable by antioxidants. There was no substantial ADP:Fe(2+)-dependent protein carbonyl formation in the absence of phospholipid under otherwise identical conditions. Based on the lipid requirement and antioxidant sensitivity, these data suggest that ADP:Fe(2+)-dependent protein carbonyl formation occurs through reaction of BSA with aldehydic lipid peroxidation products. The precise mechanism of diquat-mediated protein carbonyl formation remains unclear, but it appears not to be a function of H2O2 generation or diquat cation radical-dependent reduction of bound Fe3+.

Alterations in the cardiopulmonary effects and pharmacokinetics of a bisphosphonate drug by a cytochrome P-450 inhibitor in conscious rats.

U-91502, a bisphosphonate for arthritic inflammation treatment, was evaluated for its parental toxicity. The objective was to differentiate between the parent drug and a reactive metabolite(s) as the proximate cause of the toxic effects using two methods. The first method was to block the metabolism of U-91502 with a broad-spectrum cytochrome P-450 inhibitor, 1-aminobenzotriazole (ABT), to increase its toxicity. The second method was to scavenge any electrophilic intermediates of U-91502 with supplemental nucleophiles, L-methionine (LM) and N-acetylcysteine (NaLc) to decrease its toxicity. Two groups of rats each were given an i.v. injection of saline or ABT followed by an i.v. infusion of U-91502 at a constant dose rate. A third group was given two oral doses of LM followed by a co-infusion of U-91502 and NaLc. The breathing rate (BR) and electrocardiogram (ECG) of the rats were monitored. Blood samples were taken at specified time points for plasma drug concentration analyses (PDC) and pharmacokinetics determination. Each rat was infused until its BR was depressed by approximately 30% from the rates prior to injection of saline or ABT, or the second oral dose of LM. Thereafter, half of the rats in each group were sacrificed immediately and the remaining half at 180 min post infusion. All infused rats, except for those of the co-infusion group, and a group of untreated rats were analyzed for hepatic non-protein sulfhydryl for indication of glutathione depletion. The results indicated that ABT pretreatment expedited the elevation of PDC to a critical level that caused BR and then heart rate (HR) depression and ECG alterations. There was no unusual depletion of glutathione. The maximum concentration and the area under the curve were significantly increased while the total clearance was significantly reduced. Consequently, the postinfusion PDC remained high and the BR and HR depressions persisted. LM and NaLc did not alleviate the toxicity or alter the pharmacokinetics of U-91502. It was concluded that the toxic effects of U-91502 were due mainly to the parent drug and not the metabolites.

Cyclophosphamide is neuroprotective in a gerbil model of transient severe focal cerebral ischemia: correlation with effects of tirilazad mesylate (U-74006F).

Using a gerbil model of severe, temporary focal ischemia (3 h unilateral carotid occlusion), preliminary experiments identified an involvement of neutrophils in the reperfusion injury to the ischemic hemisphere. The present experiments were designed to (1) quantitate the temporal accumulation of neutrophils in the gerbil model, (2) determine if cyclophosphamide-induced neutropenia provided cytoprotection to the ischemic hemisphere, and (3) attempt to correlate the cytoprotective efficacy of tirilazad mesylate with possible effects on postischemic neutrophil accumulation. Following 3 h of unilateral carotid occlusion, animals were collected at increasing times of reperfusion and the CA1 region of the hippocampus and the lateral cortex were assessed for postischemic neuronal damage using a semiquantitative index (N.D.I.) of 0 (no damage) to 4 (>75% neuronal loss). The extent of neutrophil accumulation was determined by counting intensely cytochrome oxidase-positive cells. Minimal neuronal death was evident after 2 h of reperfusion, mean N.D.I. = 0.36. However, between 2 and 4 h of reperfusion, neuronal death did not increase. By 6 h of reperfusion, the neuronal death began to proceed at an accelerated rate, N.D.I. = 0.78. By 12 h, the N.D.I. reached 3.20. The accelerated neuronal death coincided with parenchymal invasion of neutrophils. Cyclophosphamide administration delayed neuronal death in the hippocampus, but exhibited a more sustained protective effect in the lateral cortex. Administration of tirilazad mesylate also resulted in a significant reduction in neutrophil accumulation and significant neuronal protection in both brain areas. Thus, in this gerbil model of transient, but prolonged focal cerebral ischemia, neutrophils appear to play an active role in the reperfusion injury to brain tissue. Our experiments confirm the previously demonstrated neuroprotective efficacy of tirilazad mesylate in this model and provide evidence for a similar protective effect of cyclophosphamide. Although other effects of this antioxidant are also thought to contribute to the overall efficacy, the data are consistent with the hypothesis that one mechanism by which tirilazad acts involves limiting the ability of neutrophils to participate in the reperfusion phase of ischemic cerebral injury.

Diquat-dependent protein carbonyl formation. Identification of lipid-dependent and lipid-independent pathways.

In a previous report on diquat-dependent oxidative damage in rat hepatic microsomes, protein oxidation, as measured by protein carbonyl (PC) formation, was observed in addition to lipid peroxidation (LP). Both phenomena were antioxidant sensitive. Inhibition of PC formation was somewhat surprising given the proposed mechanism of metal-catalyzed protein oxidation. Studies reported here examined diquat-dependent PC formation in greater detail. In rat hepatic microsomes, diquat-dependent thiobarbituric acid-reactive substances (TBARS) and PC formation were time and concentration dependent. In this system, LP was inhibited completely by U-74006F or U-78517G, whereas PC formation was inhibited only partially by these antioxidants. In an essentially lipid-free system consisting of purified rat hepatic cytochrome P450 reductase, BSA and an NADPH-generating system, PC formation was also observed, but was not antioxidant-sensitive. Under these conditions, minimal diquat-dependent TBARS formation was observed. The observation of relative antioxidant insensitivity is consistent with H2O2 (generated during the diquat redox cycle) catalyzing protein oxidation via a site-specific, metal-catalyzed mechanism. Thus, different pathways would appear to be involved in diquat-dependent PC formation in lipid-containing and lipid-free systems. Carbon tetrachloride induces LP following reductive activation to the trichloromethyl free radical, a pathway not directly involving H2O2 generation. In the microsomal system, CCl4 induced TBARS and PC formation, both of which were completely inhibitable by antioxidants. Taken together, these data suggest that diquat induces PC formation by lipid-dependent (antioxidant-sensitive) and lipid-independent (antioxidant-insensitive) pathways. In microsomes, both pathways contribute to diquat-dependent PC formation. Data for the lipid-independent pathway are consistent with the mechanism of metal-catalyzed protein oxidation proposed by Stadtman and colleagues (reviewed in Free Radic Biol Med 9: 315-325, 1990), while the lipid-dependent pathway is likely secondary to LP itself--via a Michael-type addition reaction between hydroxyalkenals and protein sulfhydryl groups, amino groups or other protein nucleophiles. The latter pathway is also responsible for carbon tetrachloride-dependent PC formation. Additional studies are in progress to further characterize the lipid-independent mechanism.

S-(1,2-dichlorovinyl)-L-cysteine-induced nephrotoxicity in the New Zealand white rabbit: characterization of proteinuria and examination of the potential role of oxidative injury.

S-(1,2-dichlorovinyl)-L-cysteine (DCVC)-induced nephrotoxicity in vivo was investigated in New Zealand White rabbits. A primary emphasis in these studies was further characterization of DCVC-induced nephrotoxicity using a variety of serum and urinary analytes, including sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Additionally, the role of oxidative injury was assessed to address the dichotomy between reports indicating that such a mechanism is important in vivo and those indicating that such mechanisms do not contribute substantially to the mechanism of effects observed in vitro. Urine was collected prior to and at 8 and 24 hr after iv administration of DCVC. Serum was collected 15 min prior to and 24 hr after DCVC administration. Rabbits were euthanized 24 hr post-DCVC administration, and kidneys were fixed in formalin and further processed for light microscopic examination. DCVC (10 mg/kg, iv) induced a 45-50-fold increase in total urinary protein excretion, a 10-15-fold increase in urinary N-acetyl-beta-D-glucosaminidase concentration, plus a marked glucosuria by 24 hr postadministration. Additionally, DCVC increased serum creatinine levels by about 2-fold, with a trend toward increased blood urea nitrogen. SDS-PAGE analysis of rabbit urine confirmed the clinical finding of marked proteinuria in DCVC-treated animals, which in contrast to previously reported data was due to the presence of both low and high molecular weight proteins. Antioxidants had no significant effect on DCVC-dependent renal injury, nor was there evidence for DCVC-induced lipid peroxidation, as measured by either thiobarbituric acid-reactive substances or a commercial assay for malondialdehyde and hydroxalkenals.(ABSTRACT TRUNCATED AT 250 WORDS)

Antioxidant-dependent inhibition of diquat-induced toxicity in vivo.

The abilities of two experimental antioxidants (U-74006F and U-78517G), as well as the model antioxidant, diphenyl-p-phenylenediamine (DPPD), to protect against diquat-induced toxicity in male Fischer-344 rats were examined. Both experimental compounds afforded near complete protection against diquat-induced hepatotoxicity, as measured by clinical chemistry and histopathological indices. When observed, diquat-induced nephrotoxicity was also inhibited. Minimal protection was afforded by the model compound, DPPD. In follow-up studies with U-78517G, no effect on diquat-induced biliary excretion of oxidized glutathione was observed, suggesting that a shift in the thiol:disulfide ratio is not responsible for diquat-induced hepatotoxicity. These data are consistent with those from previous in vitro studies in our laboratory and are in agreement with studies by others which suggest that lipid peroxidation is an important event in diquat-induced hepatotoxicity in vivo. The antioxidant effects were largely route-independent as either oral pre-treatment alone (200 mg/kg, 24 h before diquat), intravenous pre-treatment alone (6 mg/kg, 5 min before diquat) or the combination of both treatments produced a similar degree of protection. While pre-treatment with antioxidants was quite effective, no significant U-78517G-dependent inhibition of toxicity was observed when administration was delayed by as little as 10 min post diquat. These latter data suggest that initiation of diquat-induced hepatotoxicity is rapid and that these compounds would therefore be unlikely to have clinical utility in the treatment of diquat intoxication.

Inhibition of diquat-induced lipid peroxidation and toxicity in precision-cut rat liver slices by novel antioxidants.

The ability of the novel antioxidants U-74,006F and U-78,517G and a known antioxidant (N,N'-diphenyl-p-phenylenediamine, (DPPD)) to inhibit chemically induced (diquat dibromide) oxidative stress was examined in precision-cut liver slices. Previous studies in rat liver microsomes demonstrated the ability of these antioxidants to inhibit lipid peroxidation without preventing redox cycling of diquat. Diquat (1 mM) initiated lipid peroxidation in liver slices prepared from F344 rats. A 30-min preincubation with antioxidants inhibited formation of thiobarbituric acid reactive substances to control levels; ethane evolution, when elevated, was also inhibited by antioxidants. The toxicity of diquat (100 microM-3 mM) was evaluated in liver slices; 1 and 3 mM diquat caused decreases in intracellular K+ and intracellular LDH. Preincubation with antioxidants substantially decreased the toxicity of diquat as indicated by K+ and LDH. Diquat significantly decreased total glutathione levels in the slices; the antioxidants did not significantly inhibit this diquat-dependent effect. In summary, diquat, a compound which undergoes redox cycling and produces oxidative stress, was shown to produce lipid peroxidation, glutathione depletion, and toxicity in liver slices. Two experimental antioxidants, a 21-aminosteroid (U-74,006F) and a trolox-amine (U-78,517G) as well as a known antioxidant (DPPD) were shown to be effective in preventing lipid peroxidation and reducing the subsequent toxicity.

Diquat-induced oxidative damage in hepatic microsomes: effects of antioxidants.

The ability of the redox cycling compound, diquat, to induce lipid peroxidation and oxidative damage was investigated using hepatic microsomes. Antioxidants, with demonstrated efficacy in physical models of oxidative stress, were examined in a diquat model. Diquat (10 microM-3 mM) induced lipid peroxidation (TBARS) in hepatic microsomes prepared from Fischer 344 rats. Diquat (1 mM) also increased protein carbonyl formation, NADPH oxidation and superoxide anion radical production (acetylated cytochrome c reduction). The novel antioxidants U-74,006F, U-78,517G and the known antioxidant, DPPD, decreased diquat-induced lipid peroxidation to levels below that of the control. These antioxidants also decreased protein carbonyl formation caused by diquat. U-74,006F and U-78,517G reduced NADPH oxidation slightly; although this inhibition was statistically significant, the biological significance is questionable. DPPD had no effect on this parameter. U-78,517G inhibited the reduction of acetylated cytochrome c slightly, whereas the other antioxidants had little effect. Thus overall, the increase in NADPH oxidation and the production of superoxide anion by redox cycling of diquat were not substantially affected by antioxidants. Neither did the test compounds show evidence of activity as iron chelators. This leads to the suggestion that antioxidants are preventing diquat-induced oxidative damage by scavenging lipid peroxyl radicals and preventing the propagation of the lipid peroxidation process.

Differential mutagenicity of two dihydrophenalene congeners: examination of formaldehyde generation and reactive intermediate formation as possible mechanisms.

Compounds in the dihydrophenalene series are currently under investigation as potential antipsychotic agents. The mutagenicity of two compounds in this series was evaluated in several strains in the Ames Salmonella (2,3-dihydro-N,N-dimethyl-1H-phenalen-2-amine:HCI) was less mutagenic than its monomethyl analogue, U-64,273A. Two hypothesis-the of formaldehyde and release of formaldehyde and the formation of macromolecular reactive intermediates--were evaluated as possible mechanisms for the observed mutagenicity. Formaldehyde release during biotransformation of U-65,556A but not U-64,273 was demonstrated, as measured by trapping with the Nash reagent. Thus, formaldehyde release does not correlate with the mutagenic potency of these compounds. Covalent binding of U-65,556A-[3H] equivalents to rat hepatic protein was observed, but binding to DNA, which is considered the more critical target molecule, was not observed. These data suggest that reactive intermediate formation does not explain the mutagenicity of U-65,556A in the Ames Salmonella assay. Follow-up studies were conducted to assess the possible contribution of tritium exchange to the observed covalent binding to protein by quantitatively recovering 3H2O from incubations containing U-65,556A-[3H] and rat liver microsomes. Data indicate that enzyme-dependent formation of 3H2O does occur but that this phenomenon does not account for U-65,556A-[3H]-derived, protein-bound tritium in covalent binding studies.

Inhibition of carbon tetrachloride-induced lipid peroxidation by novel antioxidants in rat hepatic microsomes: dissociation from hepatoprotective effects in vivo.

The ability of two novel antioxidants, U-74,006F and U-78,517G, as well as the known antioxidant N,N'-diphenyl-p-phenylenediamine to inhibit lipid peroxidation induced by carbon tetrachloride (CCl4) was investigated in Aroclor 1254-induced rat hepatic microsomes. All three compounds completely inhibited lipid peroxidation in microsomes as measured by the formation of thiobarbituric acid reactive substances (TBARS). Inhibition of lipid peroxidation was not a function of decreased bioactivation of CCl4, as the compounds did not substantially inhibit benzphetamine N-demethylase activity or covalent binding of [14-C]CCl4 to lipid or protein. Parallel studies examined the hepatoprotective effects of the compounds in vivo. Rats were pretreated with antioxidant or vehicle prior to administration of CCl4 (300 or 600 microL/kg i.p.). Sera were collected 24 h postadministration of CCl4 and analyzed for alanine aminotransferase (ALT) and alkaline phosphatase (ALP) activities and total bilirubin. Administration of CCl4 produced elevations in ALT, moderate changes in bilirubin, and no change in ALP activities. Histological examination of CCl4-treated livers revealed lipidosis and centrilobular necrosis. The antioxidants partially improved the clinical chemistry parameters, but had minimal effects on the histological lesion. In contrast to the complete inhibition of lipid peroxidation observed in the in vitro studies, none of the antioxidants markedly protected against CCl4-induced toxicity in vivo.