Crystal structure of human type III 3alpha-hydroxysteroid dehydrogenase/bile acid binding protein complexed with NADP(+) and ursodeoxycholate

The crystal structure of human type III 3alpha-hydroxysteroid dehydrogenase (HSD)/bile acid binding protein (AKR1C2) complexed with NADP(+) and 3alpha,7beta-dihydroxy-5beta-cholanic acid (ursodeoxycholate) at 3.0 A resolution is presented. Thus, the three-dimensional structure has now been solved for a human HSD member of the aldo-keto reductase superfamily. AKR1C2 is implicated in the prostatic production of the potent androgen 5alpha-dihydrotestosterone and the hepatic transport of bile acids. It also catalyzes the formation of the neurosteroid 3alpha-hydroxy-5alpha-pregnan-20-one in the central nervous system, and its allosteric modulation by fluoxetine has been linked to the use of this drug for premenstrual dsyphoria. Like other members of the superfamily, AKR1C2 folds into an alpha/beta-barrel and binds NADP(+) in an extended conformation. The carboxylate of ursodeoxycholate binds to AKR1C2 in the oxyanion hole at the active site. More interestingly, the orientation of ursodeoxycholate is essentially "backwards" and "upside-down" from that observed for testosterone in the related rat 3alpha-HSD.NADP(+).testosterone ternary complex, where testosterone assumes the position of a 3-ketosteroid substrate. The orientation of ursodeoxycholate is thus similar to that expected of a 17beta-HSD substrate. The ternary structure explains the ability of AKR1C2 to catalyze 3alpha-, 17beta-, and 20alpha-HSD reactions. Comparison of the steroid binding pocket of AKR1C2 with that of rat 3alpha-HSD reveals significant differences in the positions of conserved and nonconserved loop residues, providing insights into the structural basis for the functional flexibility that is observed in all the human 3alpha-HSD isoforms but not in the rat isoform.

Steroid recognition and regulation of hormone action: crystal structure of testosterone and NADP+ bound to 3 alpha-hydroxysteroid/dihydrodiol dehydrogenase

BACKGROUND

Mammalian 3 alpha-hydroxysteroid dehydrogenases (3 alpha-HSDs) modulate the activities of steroid hormones by reversibly reducing their C3 ketone groups. In steroid target tissues, 3 alpha-HSDs act on 5 alpha-dihydrotestosterone, a potent male sex hormone (androgen) implicated in benign prostate hyperplasia and prostate cancer. Rat liver 3 alpha-HSD belongs to the aldo-keto reductase (AKR) superfamily and provides a model for mammalian 3 alpha-, 17 beta- and 20 alpha-HSDs, which share > 65% sequence identity. The determination of the structure of 3 alpha-HSD in complex with NADP+ and testosterone (a competitive inhibitor) will help to further our understanding of steroid recognition and hormone regulation by mammalian HSDs.

RESULTS

We have determined the 2.5 A resolution crystal structure of recombinant rat liver 3 alpha-HSD complexed with NADP+ and testosterone. The structure provides the first picture of an HSD ternary complex in the AKR superfamily, and is the only structure to date of testosterone bound to a protein. It reveals that the C3 ketone in testosterone, corresponding to the reactive group in a substrate, is poised above the nicotinamide ring which is involved in hydride transfer. In addition, the C3 ketone forms hydrogen bonds with two active-site residues implicated in catalysis (Tyr55 and His117).

CONCLUSIONS

The active-site arrangement observed in the 3 alpha-HSD ternary complex structure suggests that each positional-specific and stereospecific reaction catalyzed by an HSD requires a particular substrate orientation, the general features of which can be predicted. 3 alpha-HSDs are likely to bind substrates in a similar manner to the way in which testosterone is bound in the ternary complex, that is with the A ring of the steroid substrate in the active site and the beta face towards the nicotinamide ring to facilitate hydride transfer. In contrast, we predict that 17 beta-HSDs will bind substrates with the D ring of the steroid in the active site and with the alpha face towards the nicotinamide ring. The ability to bind substrates in only one or a few orientations could determine the positional-specificity and stereospecificity of each HSD. Residues lining the steroid-binding cavities are highly variable and may select these different orientations.

Rapid liquid chromatographic method to distinguish wild salmon from aquacultured salmon fed synthetic astaxanthin

Analytical methods are needed to determine the presence of color additives in fish. We report a liquid chromatographic (LC) method developed to identify the synthetic form of the color additive astaxanthin in salmon, based on differences in the relative ratios of the configurational isomers of astaxanthin. The distributions of configurational isomers of astaxanthin in the flesh of wild Atlantic and wild Pacific salmon are similar, but significantly different from that in aquacultured salmon. Astaxanthin is extracted from the flesh of salmon, passed through a silica gel Sep-Pak cartridge, and analyzed directly by LC on a Pirkle covalent L-leucine column. No derivatization of the astaxanthin is required-an important advantage of our approach, which is a modification of our previously described method. This method can be used to distinguish between aquacultured and wild salmon. The method has general applicability and can also be used to identify astaxanthins derived from other sources such as Phaffia yeast and Haematococcus pluvialis algae.

Incomplete data sets: coping with inadequate databases

Three problems arise in handling numerical values in databases: bad data, missing data, and sloppy data. The effects of bad data are mitigated by using statistical subterfuges such as robust statistics or outlier removal. Missing data are replaced by creating a substitute through interpolation or by using statistics appropriate to unbalanced designs. Sloppy, semiquantitative data are relegated to innocuous positions by using nonparametric, rank, or attribute statistics. These techniques are illustrated by the telephone directory, a database of carcinogenicity test results, and a database of precision parameters derived from method performance (collaborative) studies.

U.S. Food and Drug Administration pesticide program: incidence/level monitoring of domestic and imported pears and tomatoes

In 1992-1993, the U.S. Food and Drug Administration (FDA) conducted a statistically based study of pesticide residues in domestic and imported pears and tomatoes. For pears, 710 domestic and 949 imported samples were collected and analyzed; 79% of the domestic and 72% of the imported samples had detectable residues. Thiabendazole, a fungicide with postharvest uses, was found with greatest frequency in both groups of pears. Four domestic and 12 imported samples contained violative residues, mainly of pesticides for which there are no U.S. tolerances on pears. The statistically weighted (by shipment size) violation rates for domestic and imported pears were 1.0 and 0.9%, respectively. For tomatoes, 1219 domestic and 144 imported samples were collected and analyzed; 84% of the domestic and 91% of the imported samples had detectable residues. Methamidophos, an insecticide, had the greatest frequency of occurrence in both groups of tomatoes. Thirty-three domestic and 5 imported samples were violative, nearly all the result of acephate use, for which there is no U.S. tolerance on tomatoes. The statistically weighted violation rates for domestic and imported tomatoes were 1.9 and 7.0%, respectively. The statistically weighted violation rates calculated for domestic and imported pears and domestic tomatoes in this study were lower than those observed under FDA's regulatory monitoring in recent years. The violation rate for imported tomatoes was somewhat higher under statistical monitoring than under regulatory monitoring. The results of the statistically based study show that, as in regulatory monitoring, the levels of pesticide residues found are generally well below U.S. tolerances.

Multifunctional column coupled with liquid chromatography for determination of aflatoxins B1, B2, G1, and G2 in corn, almonds, brazil nuts, peanuts, and pistachio nuts: collaborative study

An AOAC/IUPAC collaborative study was conducted to evaluate the effectiveness of a multifunctional column for the determination of aflatoxins. The test portion is extracted with acetonitrile-water (9 + 1), the extract is filtered, and the filtrate is passed through the column. The aflatoxins in the eluate are determined by reversed-phase liquid chromatography after derivatization with trifluoroacetic acid. Naturally contaminated corn, almonds, Brazil nuts, peanuts, and pistachio nuts spiked with total aflatoxins at 5, 10, 20, and 30 ng/g were sent to 12 collaborators in the United States, Denmark, France, Japan, and Switzerland. Eleven collaborators completed the study. Average recoveries of total aflatoxins for each spike level for the various commodities (excluding Brazil nuts at 5 ng/g) were 93, 97, 95, and 95%, respectively; the repeatability relative standard deviation (RSDr) ranged from 6.0 to 23.2% and the reproducibility relative standard deviation (RSDR) ranged from 12.0 to 69.4%. The multifunctional column coupled with a liquid chromatographic method for determination of aflatoxins in corn, almonds, Brazil nuts, peanuts, and pistachio nuts has been adopted first action by AOAC INTERNATIONAL.

Time required to achieve homogeneity of test substances added to dog feed

The homogeneity of test substances in a carrier (animal feed) is a critical factor in conducting long-term feeding studies in laboratory animals. A method for determining the adequate amount of mixing to achieve homogeneity by a mixer of the type described has been determined when 2 distinctly different compounds are added to ground dog feed. Nicotinic acid and butylated hydroxyanisole at a concentration of 1% were separately mixed with the dog feed for 15, 30, 45, 60, and 120 min to determine optimum mixing time. Test portions were taken from 4 different sampling sites at each time period and analyzed in duplicate for the added substance. Four batches were prepared and the results were aggregated. Very little interbatch variability was observed. The variance of the average values from the 4 sampling sites at each time period was calculated and used as a simple, crude, but effective numerical quantity to monitor the approach to homogeneity of the mixture.

Evaluation of immunotoxicity in a subchronic feeding study of triphenyl phosphate

Triphenyl phosphate (TPP), a potential food contaminant, was fed to weanling Spartan Sprague-Dawley rats at dose levels of 0, 0.25, 0.5, 0.75, and 1.0% for 120 days. The immunotoxicity evaluation, planned as a minimum testing model in a subchronic study design as well as to provide information on TPP, was performed along with the routine testing of a separate group of animals. Traditional measures were made of growth and food consumption, total protein analysis, electrophoretic analyses of serum proteins, lymphoid organ weights in relation to growth, and histopathology, with expanded immunohistochemical evaluation of B- and T- lymphocyte regions in spleen, thymus, and lymph nodes, using immunoperoxidase staining. Assessment was made of the humoral response to a T-lymphocyte-dependent antigen, sheep red blood cells, and was begun at midterm of the feeding period for the primary response followed by secondary and tertiary booster immunizations at 3-week intervals. The kinetics of the responses were measured by hemolysin assay of relative antibody titers at days 3, 4, 5, and 6 postinjection. No significant effects on the responses were noted for either sex at any of the dose levels tested. The only effects noted were a decreased rate of growth at high levels of TPP and increases in the levels of alpha- and beta-globulins suggestive of increased hepatic activity.

Outliers in collaborative studies: coping with uncertainty

An overview is given of the options available in detecting and dealing with outliers in collaborative studies. The fundamental points of agreement and disagreement are highlighted. The common sense approach of just looking at the data is emphasized. The importance is stressed of making a harmonized choice of outlier treatments, even though such a choice may not be optimal for all circumstances.

Splenotoxicity associated with splenic sarcomas in rats fed high doses of D & C Red No. 9 or aniline hydrochloride

A histopathologic review of F344 rat spleens from the National Toxicology Program-National Cancer Institute bioassays of barium salt of 5-chloro-2-(2-hydroxy-1-naphthalenyl)-azo-4-methylbenzenesulfonic acid [(D & C Red No. 9) CAS: 516-00-21] and aniline HCI (CAS: 142-04-1) was conducted to assess splenotoxic changes associated with splenic sarcomas induced by these aromatic amines. Four splenic changes--fatty metamorphosis (FM), splenic fibrosis (FIB), capsule hyperplasia (CH), and hemorrhage--were markedly increased in incidence and severity in males treated with high doses of either D & C Red No. 9 or aniline HCI. Females treated with high doses of either of these compounds showed similar but less severe changes. FIB and FM showed strong group correlations with tumor incidence (r greater than or equal to 0.87). All groups that demonstrated FM also demonstrated splenic sarcomas; groups without the FM lesions did not exhibit splenic tumors. The morphologic similarity of the FIB and CH lesions to the induced splenic sarcomas suggests that these lesions are preneoplastic. Moreover, the treatment-related splenic lesions appear to be precursors of the induced splenic sarcomas. Carcinogenicity studies with serial sacrifices at varying intervals will be required for experimental verification of these conclusions. A schema, based on the findings of the study, suggests a hypothetical pathway for the progression of the treatment-related splenic lesions from onset to tumor formation.

Performance characteristics of methods of analysis used for regulatory purposes. I. Drug dosage forms. D. High pressure liquid chromatographic methods

Precision parameters of high pressure liquid chromatographic methods approved by AOAC for the analysis of drug dosage forms were recalculated on a consistent statistical basis, using the computer program "FDACHEMIST." Eleven collaborative studies of 12 compounds in 66 dosage forms analyzed by an average of 9 laboratories per study, with a total of 1150 determinations, were reviewed. For the approved methods and methods awaiting approval (9 studies, 11 compounds, 54 dosage forms, and 959 determinations), the average repeatability relative standard deviation (within-laboratory; RSDo) was 1.0%; reproducibility relative standard deviation (among-laboratories, including within-; RSDx) was 2.5%; the ratio RSDo/RSDx was an unusually low 0.40, with an average outlier rate of 0.6% of the reported values. The line of best fit for RSDx plotted against - log concentration increases with decreasing concentration, extending approximately from RSDx = 2% at 100% concentration to RSDx = 3.6% at 0.01% concentration, a change in RSDx of about 0.4% for each 10-fold decrease in concentration, independent of analyte and matrix.

Determination of lead and cadmium in foods by anodic stripping voltammetry: II. Collaborative study

A dry ash anodic stripping voltammetric method for determining lead and cadmium in foods was collaboratively studied by 20 laboratories. The food commodities studied were strained green beans, beef (baby food), fish (mackerel), infant formula (milk base), apple juice, and cereal (wheat farina). Each collaborator analyzed 3 commodities, each consisting of 2 duplicate lead and cadmium fortification levels, for a total of 4 samples for each commodity. The low fortification levels ranged from 0.03 to 0.08 ppm for cadmium and from 0.05 to 0.15 ppm for lead. The high fortification levels ranged from 0.12 to 0.28 ppm for cadmium and from 0.24 to 0.45 ppm for lead. Each commodity was analyzed by 10 collaborators. The average overall reproducibilities of the low level fortifications were 24% for lead and 21% for cadmium; for the high level fortifications, average overall reproducibilities were 18% for lead and 16% for cadmium. The average accuracies of the collaborative results as measured by comparison to reference values were 96 and 97% for cadmium and lead, respectively. This method has been adopted official first action.