Genotoxic and carcinogenic risk to humans of drug-nitrite interaction products

Novel study on N-nitrosamines as risk factors of cardiovascular diseases

Millions of people are exposed daily to N-nitrosamines from different environmental sources. The present study aims at investigating the role of N-nitrosamines in the alteration of homocysteine, lipid profile, oxidative stress, paraoxonase activity, antioxidant enzymes, and free radicals which are important risk factors for CVD. In addition, biomarkers of cardiovascular diseases such as creatine kinase MB activity (CK-MB) and lactate dehydrogenase (LDH) as well as protein expression of both glutathione peroxidase and glutathione S-transferase π isozyme were assayed after treatment of rats with 0.2 mg/kg body weight of N-nitrosodibutylamine (NDBA), N-nitrosoethylbutylamine (NEBA), N-nitrosobutylpropylamine (NBPA), N-nitrosodiethylamine (NDEA), N-nitrosodimethylamine (NDMA), and N-nitrosodiphenylamine (NDPA) as a daily dose for two weeks. LDL levels, paraoxonase activity, reduced glutathione levels, and glutathione reductase activities were increased, whereas HDL levels decreased after treatment of rats with most of N-nitrosamines compared to control group. Moreover, levels of free radicals and catalase activity increased, whereas protein expression of both glutathione peroxidase and glutathione S-transferase decreased after treatment of rats with some N-nitrosamines. The data showed that most N-nitrosamines increased CK-MB and LDH activities. It is concluded that N-nitrosamines increased levels of free radicals, and decreased the activity of antioxidant enzymes which may consequently increase the incidence of CVDs.

Prenatal nitrate intake from drinking water and selected birth defects in offspring of participants in the national birth defects prevention study

BACKGROUND

Previous studies of prenatal exposure to drinking-water nitrate and birth defects in offspring have not accounted for water consumption patterns or potential interaction with nitrosatable drugs.

OBJECTIVES

We examined the relation between prenatal exposure to drinking-water nitrate and selected birth defects, accounting for maternal water consumption patterns and nitrosatable drug exposure.

METHODS

With data from the National Birth Defects Prevention Study, we linked addresses of 3,300 case mothers and 1,121 control mothers from the Iowa and Texas sites to public water supplies and respective nitrate measurements. We assigned nitrate levels for bottled water from collection of representative samples and standard laboratory testing. Daily nitrate consumption was estimated from self-reported water consumption at home and work.

RESULTS

With the lowest tertile of nitrate intake around conception as the referent group, mothers of babies with spina bifida were 2.0 times more likely (95% CI: 1.3, 3.2) to ingest ≥ 5 mg nitrate daily from drinking water (vs. < 0.91 mg) than control mothers. During 1 month preconception through the first trimester, mothers of limb deficiency, cleft palate, and cleft lip cases were, respectively, 1.8 (95% CI: 1.1, 3.1), 1.9 (95% CI: 1.2, 3.1), and 1.8 (95% CI: 1.1, 3.1) times more likely than control mothers to ingest ≥ 5.42 mg of nitrate daily (vs. < 1.0 mg). Higher water nitrate intake did not increase associations between prenatal nitrosatable drug use and birth defects.

CONCLUSIONS

Higher water nitrate intake was associated with several birth defects in offspring, but did not strengthen associations between nitrosatable drugs and birth defects.

Prenatal exposure to nitrosatable drugs, vitamin C, and risk of selected birth defects

Nitrosatable drugs, such as secondary or tertiary amines and amides react with nitrite in an acidic environment to form N-nitroso compounds, teratogens in animal models. Vitamin C is a known nitrosation inhibitor.

METHODS

Using data from the National Birth Defects Prevention Study, we assessed nitrosatable drug exposure and vitamin C intake during the first trimester among 11,606 case-mothers of infants with oral clefts, limb deficiencies (LDs), or congenital heart defects and 6807 control-mothers of infants without major birth defects during 1997-2005. Daily intake of vitamin C was estimated from maternal interviews that elicited information about supplement use and dietary intake.

RESULTS

With no reported use of nitrosatable drugs as the referent group, a lower odds ratio (OR) was observed for transverse LDs among births to mothers exposed to secondary amine drugs and daily vitamin C supplementation (adjusted odds ratio [aOR] 1.2, 95% confidence interval [CI] 0.83-1.8) compared with women taking these drugs and no supplementation (aOR 2.7, 95% CI 1.5-4.6). The OR for longitudinal LDs associated with secondary amine exposure was lower with daily dietary vitamin C intake ≥85 mg (aOR 1.2, 95% CI 0.68-2.0) compared with <85 mg (aOR 1.9, 95% CI 1.2-3.1). Daily vitamin C supplementation in combination with higher dietary vitamin C intake reduced associations between nitrosatable drug exposures and limb deficiencies and atrial septal defects not otherwise specified.

CONCLUSION

Prenatal dietary and vitamin C supplement intake may diminish the association between nitrosatable drug exposure during pregnancy and selected birth defects.

BRCA2 affects the efficiency of DNA double-strand break repair in response to N-nitroso compounds with differing carcinogenic potentials

The tumor suppressor gene breast cancer susceptibility gene 2 (BRCA2) is frequently mutated or epigenetically repressed in human cancer and has a significant role in the homologous recombination (HR) of DNA double-strand breaks (DSBs). Although N-nitrosodiethylamine (NDEA), N-nitrosodiethanolamine (NDELA) and N-nitrosodipropylamine (NDPA) have similar chemical structures and are able to induce DNA damage, they have varying carcinogenic risks. We hypothesized that the DNA damage repair pathways that are induced by these N-nitroso compounds (NOCs) may differ and that this may contribute to the genotoxic-carcinogenic effect of the NOCs. The present study aimed to characterize the formation of DSBs by NDEA, NDELA and NDPA and also to investigate whether BRCA2 is involved in the DNA damage response. The NOCs were observed to time-dependently induce DSBs and the expression of γ-H2AX in gastric cancer SGC7901 cells. It was observed that the DNA damage induced by NDEA, the most potent carcinogen, was not repaired as efficiently as that caused by NDELA or NDPA. The expression of BRCA2 and RAD51 was demonstrated to be inhibited by NDEA treatment but upregulated by NDELA or NDPA treatment. Furthermore, the knock down of BRCA2 expression impaired the DNA damage repair induced by NDELA or NDPA. The cells with this knock down exhibited an increased sensitivity to NDELA or NDPA treatment, but not to NDEA. These findings suggest that a BRCA2-mediated pathway contributes to differential DSB repair and sensitivity in response to NOC exposure and that it may be associated with the genotoxic-carcinogenic potential of NOCs.

Associations between red meat and risks for colon and rectal cancer depend on the type of red meat consumed

Cancer prevention guidelines recommend limiting intake of red meat and avoiding processed meat; however, few studies have been conducted on the effects of specific red meat subtypes on colon cancer or rectal cancer risk. The study aim was to evaluate associations between intake of red meat and its subtypes, processed meat, fish, and poultry and risk for colon cancer or rectal cancer in the Danish Diet, Cancer and Health cohort study. We also evaluated whether fish or poultry should replace red meat intake to prevent colon cancer or rectal cancer. During follow-up (13.4 y), 644 cases of colon cancer and 345 cases of rectal cancer occurred among 53,988 participants. Cox proportional hazards models were used to compute incidence rate ratio (IRRs) and 95% CIs. No associations were found between intake of red meat, processed meat, fish, or poultry and risk for colon cancer or rectal cancer. The risk associated with specific red meat subtypes depended on the animal of origin and cancer subsite; thus, the risk for colon cancer was significantly elevated for higher intake of lamb [IRR(per 5g/d) = 1.07 (95% CI: 1.02-1.13)], whereas the risk for rectal cancer was elevated for higher intake of pork [IRR(per 25g/d) = 1.18 (95% CI: 1.02-1.36)]. Substitution of fish for red meat was associated with a significantly lower risk for colon cancer [IRR(per 25g/d) = 0.89 (95% CI: 0.80-0.99)] but not rectal cancer. Substitution of poultry for red meat did not reduce either risk. This study suggests that the risks for colon cancer and potentially for rectal cancer differ according to the specific red meat subtype consumed.

Genotoxicity and carcinogenicity studies of bronchodilators and antiasthma drugs

This survey is a compendium of genotoxicity and carcinogenicity information of bronchodilators and antiasthma drugs. Data from 46 marketed drugs were collected. Of these 46 drugs, 25 (54.3%) did not have retrievable genotoxicity or carcinogenicity data. The remaining 21 (45.7%) had at least one genotoxicity or carcinogenicity test result. Of these 21 drugs, 10 had at least one positive finding: three tested positive in at least one genotoxicity assay, eight in at least one carcinogenicity assay, and one of them gave positive results in both genotoxicity assay and carcinogenicity assay. Concerning the predictivity of genetic toxicology findings for the result(s) of long-term carcinogenesis assays, 15 drugs had both genotoxicity and carcinogenicity data: seven of them (46.6%) were neither genotoxic nor carcinogenic, 6 (40.0%) were carcinogenic in at least one sex of mice or rats but tested negative in genotoxicity assays, 1 (6.7%) tested positive in genotoxicity assay but was non-carcinogenic, and 1 (6.7%) gave positive responses in both genotoxicity and carcinogenicity assay. Only 11 (23.9%) of the 46 drugs considered had all data required by current guidelines for testing of pharmaceuticals, but a large fraction of them were developed and marketed prior to the present regulatory climate.

Nitrosatable drug exposure during the first trimester of pregnancy and selected congenital malformations

BACKGROUND

Nitrosatable drugs can react with nitrite in the stomach to form N-nitroso compounds, and results from animal studies suggest that N-nitroso compounds are teratogens. With data from the National Birth Defects Prevention Study, the relation between prenatal exposure to nitrosatable drugs and limb deficiencies, oral cleft, and heart malformations in offspring was examined.

METHODS

Maternal reports of drugs taken during the first trimester of pregnancy were classified with respect to nitrosatability for mothers of 741 babies with limb deficiencies, 2774 with oral cleft malformations, 8091 with congenital heart malformations, and 6807 without major congenital malformations. Nitrite intake was estimated from maternal responses to a food frequency questionnaire.

RESULTS

Isolated transverse limb deficiencies and atrioventricular septal defects were associated with secondary amine drug exposures (adjusted odds ratios [aORs], 1.51; 95% confidence limit [CI], 1.11-2.06 and aOR, 1.97; 95% CI, 1.19-3.26, respectively). Tertiary amines were associated with hypoplastic left heart syndrome (aOR, 1.50; 95% CI, 1.10-2.04) and single ventricle (aOR, 1.61; 95% CI, 1.06-2.45). These two malformations were also significantly associated with amide drugs. For several malformations, the strongest associations with nitrosatable drug use occurred among mothers with the highest estimated dietary nitrite intake, especially for secondary amines and atrioventricular septal defects (highest tertile of nitrite, aOR, 3.30; 95% CI, 1.44-7.58).

CONCLUSION

Prenatal exposure to nitrosatable drugs may be associated with several congenital malformations, especially with higher nitrite intake. The possible interaction between nitrosatable drugs and dietary nitrite on risk of congenital malformations warrants further attention.

Reactions of amine and peroxynitrite: evidence for hydroxylation as predominant reaction and new insight into the modulation of CO2

Peroxynitrite is related to numerous diseases including cardiovascular diseases, inflammation, and cancer. In order to expand the understanding for the toxicology of peroxynitrite in biological system, the reactions of amine (morpholine as a probe) with peroxynitrite and the modulation of CO2 were investigated by using DFT methods. The results strongly indicate that the hydroxylation of amine by peroxynitrous acid ONOOH, which was previously overlooked by most studies, is predominant relative to the widely reported nitration and nitrosation in the absence of CO2. The product N-hydroxylamine is proposed to be mainly generated via nonradical pathway (two-electron oxidation). The modulation of CO2 exhibits two main functions: (1) inhibition of hydroxylation due to the promoted consumption of peroxynitrite via fast reaction of CO2 with ONOO¯ to form ONOOCO2¯; (2) dual effect (catalysis and inhibition) of CO2 toward nitration and nitrosation. As a new insight, amine does react with CO2 and produce inert amine carbamate R2NCOO¯. This reaction has the potential to compete with the reaction of CO2 and ONOO¯, which leads to inhibition of nitration and nitrosation. The concentration of CO2 could be a critical factor determining the final effect, catalysis or inhibition. As a new finding, HCO3¯ is probably an effective catalyst for the reaction of amine and CO2. Moreover, further studies on how the different types of the amine might affect the outcome of the reactions would be an interesting topic.

Studies on genotoxicity and carcinogenicity of antibacterial, antiviral, antimalarial and antifungal drugs

This review provides a compendium of retrievable results of genotoxicity and animal carcinogenicity studies performed of antibacterial, antiviral, antimalarial and antifungal drugs of long-term or intermittent frequent use. Of the 48 drugs considered, 9 (18.75%) do not have retrievable data, whereas the other 39 (81.25%) have at least one genotoxicity or carcinogenicity tests result. Of these 39 drugs, 24 tested positive in at least one genotoxicity assay and 19 in at least one carcinogenicity assay; 14 of them gave a positive response in both at least one genotoxicity assay and at least one carcinogenicity assay. Concerning the predictivity of genetic toxicology findings for the results of long-term carcinogenesis assays, of 23 drugs with both genotoxicity and carcinogenicity data: 2 (8.7%) were neither genotoxic nor carcinogenic, 2 (8.7%) tested positive in at least one genotoxicity assay but were non-carcinogenic, 4 (17.4%) tested negative in genotoxicity assays but were carcinogenic, and 15 (65.2%) gave a positive response in at least one genotoxicity assay and in at least one carcinogenicity assay. Only 18 (37.5%) of the 48 drugs examined had all data required by present guidelines for testing of pharmaceuticals, but a fraction of them (49%) were developed and marketed prior to the present regulatory climate. In the absence of compelling indications, the prescription of the 19 drugs that are animal carcinogens should be avoided.

Reactive impurities in excipients: profiling, identification and mitigation of drug-excipient incompatibility

Reactive impurities in pharmaceutical excipients could cause drug product instability, leading to decreased product performance, loss in potency, and/or formation of potentially toxic degradants. The levels of reactive impurities in excipients may vary between lots and vendors. Screening of excipients for these impurities and a thorough understanding of their potential interaction with drug candidates during early formulation development ensure robust drug product development. In this review paper, excipient impurities are categorized into six major classes, including reducing sugars, aldehydes, peroxides, metals, nitrate/nitrite, and organic acids. The sources of generation, the analytical method for detection, the stability of impurities upon storage and processing, and the potential reactions with drug candidates of these impurities are reviewed. Specific examples of drug-excipient impurity interaction from internal research and literature are provided. Mitigation strategies and corrective measures are also discussed.

Nitrosatable drug exposure during early pregnancy and neural tube defects in offspring: National Birth Defects Prevention Study

Nitrosatable drugs, such as secondary or tertiary amines and amides, form N-nitroso compounds in the presence of nitrite. Various N-nitroso compounds have been associated with neural tube defects in animal models. Using data from the National Birth Defects Prevention Study, the authors examined nitrosatable drug exposure 1 month before and 1 month after conception in 1,223 case mothers with neural tube defect-affected pregnancies and 6,807 control mothers who delivered babies without major congenital anomalies from 1997 to 2005. Nitrite intakes were estimated from mothers' responses to a food frequency questionnaire. After adjustment for maternal race/ethnicity, educational level, and folic acid supplementation, case women were more likely than were control women to have taken tertiary amines (odds ratio = 1.60, 95% confidence interval (CI): 1.31, 1.95). This association was strongest with anencephalic births (odds ratio = 1.96, 95% CI: 1.40, 2.73); odds ratios associated with tertiary amines from the lowest tertile of nitrite intake to the highest tertile were 1.16 (95% CI: 0.59, 2.29), 2.19 (95% CI: 1.25, 3.86), and 2.51 (95% CI: 1.45, 4.37), respectively. Odds ratios for anencephaly with nitrosatable drug exposure were reduced among women who also took daily vitamin supplements that contained vitamin C. Prenatal exposure to nitrosatable drugs may increase the risk of neural tube defects, especially in conjunction with a mother's higher dietary intake of nitrites, but vitamin C might modulate this association.

Interaction of 5-aminosalicylic acid with nitrous acid: formation of the diazonium derivative and nitric oxide release

The reaction of 5-aminosalicylic acid (5-ASA) with nitrous acid has been studied at low pH under conditions that simulate a gastric environment. The course of the reaction was followed by UV–visible and fluorescence spectroscopy and the products were analyzed by high performance liquid chromatography (HPLC) with UV–visible and mass spectroscopic detectors. In addition, the formation of nitric oxide (NO) was estimated electrochemically. 5-ASA was readily consumed in a process catalyzed by chloride and thiocyanate, whose rate is first order in 5-ASA and second order in nitrous acid. 2-Hydroxy-5-diazonium benzoic acid (diazonium derivative) and NO were detected as products of the reaction. From the NO formation profiles, it is concluded that NO is produced as a minor product in a process parallel to the path that leads to generation of the diazonium derivative. While the formation of NO could be beneficial for the protection of the stomach, the generation of the diazonium derivative could be considered a potentially toxic process.

Prevalence and patterns of nitrosatable drug use among U.S. women during early pregnancy

BACKGROUND

Experimental evidence indicates that certain drugs, that are secondary or tertiary amines or amides, form N-nitroso compounds in the presence of nitrite in an acidic environment. Nitrosatable drugs have been associated with birth defects in a few epidemiologic studies. This study describes the prevalence and patterns of nitrosatable drug use among U.S. women during early pregnancy and examines maternal factors associated with such use.

METHODS

Data were analyzed from the National Birth Defects Prevention Study and included 6807 mothers who gave birth to babies without major congenital malformations during 1997 to 2005. Information was collected by telephone interview about medication use, demographic factors, and maternal health. Drugs taken during the first trimester were classified according to nitrosatability, amine and amide functional groups, and primary indication of use.

RESULTS

Approximately 24% of the women took one or more nitrosatable drugs during the first trimester, including 12.4%, 12.2%, and 7.6% who respectively took secondary amines, tertiary amines, or amides. Five of the ten most commonly taken drugs were available over the counter. Women who were non-Hispanic white (29.5%), with 1 year or more college education (27.3%) or 40 years or older (28.8%) had the highest prevalence of use. Supplemental vitamin C, an inhibitor of nitrosation, was not taken by 41.6% and 19.3% of nitrosatable drug users during the first and second months of pregnancy, respectively.

CONCLUSIONS

In this U.S. population, ingestion of drugs classified as nitrosatable was common during the first trimester of pregnancy, especially among non-Hispanic white, more educated, and older mothers.

Subchronic effects of dipyrone on the fish species Rhamdia quelen

The use of the non-steroidal anti-inflammatory drugs such as dipyrone is so widespread that this drug and its metabolites have been detected in effluents and surface water. This study aimed to evaluate the potential toxic effects of dipyrone on the aquatic environment, using a native fish species, Rhamdia quelen. Fish were exposed to three concentrations of dipyrone, 0.5, 5 and 50 μg/L, in the water for 15 days, and hematological, biochemical, genetic and morphological biomarkers were evaluated. The glutathione S-transferase activity decreased in the highest concentration in relation to the control group. In addition, hematocrit, red blood cells and thrombocyte counts were decreased in all three exposed groups in relation to the control group. The comet assay showed DNA damage at the lowest concentration of dipyrone and significant kidney damage. Those results suggest that a constant exposure of aquatic organisms to dipyrone presents potential toxic effects.

Risk of non-Hodgkin lymphoma and nitrate and nitrite from the diet in Connecticut women

It has been estimated that 65,980 individuals were diagnosed with non-Hodgkin lymphoma (NHL) and 19,500 died from NHL in the United States in 2009. Although established risk factors such as immunodeficiency and viral infections may be responsible for a portion of the cases, the majority of NHL cases remain unexplained. Dietary nitrate and nitrite intake are exposures of particular interest for NHL risk as they are precursors in the endogenous formation of N-nitroso compounds, which cause lymphomas in animal studies. We investigated NHL risk overall and by histologic type in relation to dietary nitrate and nitrite intake in a population-based case-control study of 1,304 women in Connecticut. Nitrate and nitrite intake were assessed using a 120-item food frequency questionnaire. We found no association between risk of NHL overall and dietary nitrate and a slightly increased risk of NHL with higher dietary nitrite intake (highest vs. lowest intake quartile OR = 1.4; 95% CI: 0.9-2.2). When we evaluated intake by subtype, a significant positive trend was observed for follicular lymphoma and nitrate (p-trend = 0.04) and nitrite (p-trend < 0.01) with an over twofold risk in the highest nitrite intake quartile (OR = 2.3; 95% CI: 1.1-4.9). An increased risk in the highest quartile of nitrite intake was also observed for T-cell lymphoma (OR = 3.4; 95% CI: 1.0-11.9). Animal products containing nitrite were more strongly associated with risk of follicular lymphoma; whereas, both animal and plant sources of nitrite were associated with elevated ORs for T-cell lymphoma. Our results confirm a previous finding for nitrite intake and NHL risk and highlight the importance of evaluating histologic type. We conclude that these results should be replicated in a larger study with data on drinking water as well as dietary sources of nitrate intake.

Genotoxic and carcinogenic effects of gastrointestinal drugs

This review provides a compendium of retrievable results of genotoxicity and carcinogenicity assays performed on marketed gastrointestinal drugs. Of the 71 drugs considered, 38 (53.5%) do not have retrievable data, whereas the other 33 (46.5%) have at least one genotoxicity or carcinogenicity test result. Of these 33 drugs, 15 tested positive in at least one genotoxicity assay and 13 in at least one carcinogenicity assay; 8 of them gave a positive response in both at least one genotoxicity assay and at least one carcinogenicity assay. Concerning the predictivity of genetic toxicology findings for the result(s) of long-term carcinogenesis assays, of 21 drugs with both genotoxicity and carcinogenicity data: 6 (28.6%) are neither genotoxic nor carcinogenic, 2 (9.5%) tested positive in at least one genotoxicity assay but were non-carcinogenic, 5 (23.8%) tested negative in genotoxicity assays but were carcinogenic and 8 (38.1%) gave a positive response in at least one genotoxicity assay and in at least one carcinogenicity assay. Only 12 (16.9%) of the 71 drugs examined have all data required by present guidelines for testing of pharmaceuticals, but a large fraction of them were developed and marketed prior the present regulatory climate.

Chromosomal aberrations induced by imipramine and desipramine in mouse

Imipramine (IMI) and desipramine (DES) are two drugs widely used for the treatment of depression as well as for other diseases. In the present study, we determined their capacity to induce chromosomal aberrations in mouse bone marrow cells. Three doses of each compound were tested and their results were compared with the frequency of chromosomal aberrations obtained in a control group as well as with a group treated with cyclophosphamide. Our results showed a significant increase in chromosome damage with the doses tested for each compound: 7, 20, and 60 mg/kg in the case of IMI, and 2, 20, and 60 mg/kg as regards DES. This last drug induced stronger chromosomal damage than IMI. Our results agree with previous studies regarding the induction of micronuclei and sister chromatid exchanges by the drugs in mouse and suggest caution with respect to their use in long-term treatments.

Formation of an adduct by clenbuterol, a beta-adrenoceptor agonist drug, and serum albumin in human saliva at the acidic pH of the stomach: evidence for an aryl radical-based process

Clenbuterol (CLB) is an antiasthmatic drug used also illegally as a lean muscle mass enhancer in both humans and animals. CLB and amine-related drugs in general are nitrosatable, thus raising concerns regarding possible genotoxic/carcinogenic activity. Oral administration of CLB raises the issue of its possible transformation by salivary nitrite at the acidic pH of gastric juice. In acidic human saliva CLB was rapidly transformed to the CLB arenediazonium ion. This suggests a reaction of CLB with salivary nitrite, as confirmed in aerobic HNO(2) solution by a drastic decrease in nitric oxide, nitrite, and nitrate. In human saliva, both glutathione and ascorbic acid were able to inhibit CLB arenediazonium formation and to react with preformed CLB arenediazonium. The effect of ascorbic acid is particularly pertinent because this vitamin is actively concentrated within the gastric juice. EPR spin trapping experiments showed that preformed CLB arenediazonium ion was reduced to the aryl radical by ascorbic acid, glutathione, and serum albumin, the major protein of saliva. As demonstrated by anti-CLB antibodies and MS, the CLB-albumin interaction leads to the formation of a covalent drug-protein adduct, with a preference for Tyr-rich regions. This study highlights the possible hazards associated with the use/abuse of this drug.

Use of cell viability assay data improves the prediction accuracy of conventional quantitative structure-activity relationship models of animal carcinogenicity

BACKGROUND

To develop efficient approaches for rapid evaluation of chemical toxicity and human health risk of environmental compounds, the National Toxicology Program (NTP) in collaboration with the National Center for Chemical Genomics has initiated a project on high-throughput screening (HTS) of environmental chemicals. The first HTS results for a set of 1,408 compounds tested for their effects on cell viability in six different cell lines have recently become available via PubChem.

OBJECTIVES

We have explored these data in terms of their utility for predicting adverse health effects of the environmental agents.

METHODS AND RESULTS

Initially, the classification k nearest neighbor (kNN) quantitative structure-activity relationship (QSAR) modeling method was applied to the HTS data only, for a curated data set of 384 compounds. The resulting models had prediction accuracies for training, test (containing 275 compounds together), and external validation (109 compounds) sets as high as 89%, 71%, and 74%, respectively. We then asked if HTS results could be of value in predicting rodent carcinogenicity. We identified 383 compounds for which data were available from both the Berkeley Carcinogenic Potency Database and NTP-HTS studies. We found that compounds classified by HTS as "actives" in at least one cell line were likely to be rodent carcinogens (sensitivity 77%); however, HTS "inactives" were far less informative (specificity 46%). Using chemical descriptors only, kNN QSAR modeling resulted in 62.3% prediction accuracy for rodent carcinogenicity applied to this data set. Importantly, the prediction accuracy of the model was significantly improved (72.7%) when chemical descriptors were augmented by HTS data, which were regarded as biological descriptors.

CONCLUSIONS

Our studies suggest that combining NTP-HTS profiles with conventional chemical descriptors could considerably improve the predictive power of computational approaches in toxicology.