Division of Chemistry

1. Division of Chemistry Robert J. Turesky, Ph.D. Division Director Ruth York Dwight Miller, Ph.D. Paul Siitonen Jack Lay, Jr., Ph.D. Analytical Chemical & Biomarkers NTP Coordinator Mass Spectrometry Shannon Snellings, Ph.D. Lee Holder Larry Rushing * J. Pat Freeman, Ph.D. Catharina Ang, Ph.D. Julian Leakey, Ph.D. Rick Beger, Ph.D. Tom Schmitt * Jon G. Wilkes, Ph.D. F. Evans Ph.D. Yanyan Cui, Ph.D. Robert Cecotti, Ph.D. Ronald Evans Dan Buzatu, Ph.D. Wenhong Luo, Ph.D. Willie Cooper Rick Holland Alex Shvartsburg, Ph.D. Theresa Gehring * Thomas Heinze Kenneth Roberts, Ph.D. Mark Billedeau Danny Nestorick * To be filled To be filled Eugene Hansen To be filled *Multiple activities

2. Division of Chemistry Mission Statement To utilize chemical research techniques including analytical chemistry, mass and NMR spectrometry, spectroscopic and computational methods to implement intradivisional, intercenter and FDA relevant research initiatives in toxicology, risk assessment, and regulatory compliance

3. Key Research Projects R. Beger, E07068, Spectrometric Data Activity Relationship (SDAR) Models for Compounds Binding to Receptors of Toxic Responses: Predictive Toxicology F. Evans, E07078, NMR spectroscopy of drug purity and public health implications J. Lay, E07005, Rapid identification of intact whole bacteria based upon spectral patterns using MALDI-TOF MS D. Miller, E06874, Fresh Tag SensorTM technology for product safety, quality, and rapid screening of explosives J. Wilkes, E06931, Rapid screening and identification of complex mixtures by pyrolysis-mass spectrometry with pattern recognition

4. Key Research Projects (cont.) C. Ang, E07056, Chemical characterization of selected medicinal botanical products J. Leakey and C. Ang, X00031, Impact of dietary supplements on woman’s health issues D. Buzatu E07077, Comparison of principal components analysis (PCA) and artificial neural networks (ANN) for the prediction of qualitative and quantitative biological end points from spectrometric data R. Turesky X……., Risk assessment of dietary contaminants (heteroyclic aromatic amines and mycotoxins)

5. National Toxicology Program Activities Summary Reports Dosage Form NTP Study Dosage Verification Method Development Dose Certification Stability Homogeneity

6. Surveillance Activities Rodent Diets Drinking Water Analyses Bedding

7. Active Collaborations with FDA Center for Veterinary Medicine • Amoxicillin • Erythromycin • Lincomycin • Sulfa Drugs All Projects Requiring Development of Determinative Methods that Achieve CVM Method Trial Ruggedness Testing Requirements for Reliability

8. Characterization of Bacteria by MALDI TOF/MS According to the CDC in 1999, as a direct result of microbial contamination of food there were: · 76,000,000 food-borne illnesses in the United States · 325,000 reported hospitalizations and · 5,000 deaths 64% of the deaths were from unknown organisms.

9. For V. p. MALDI gives signals that correlate well with regional outbreaks of seafood pathogens Strains from the Pacific Northwest are reproducible (see below) but significantly different from strains associated with the Gulf coast (over). V. parahaemolyticus -Washington State Vibrio Parahaemolyticus 10293 9477 p18-a8-vp10293 20 (1.889) Sb (49,1.00 ); Sm (SG, 1x10.00); Cm (20:21) 100 % 9087 9419 8910 0 m/z 8800 8900 9000 9100 9200 9300 9400 9500 9600 9700 9800 9900 10000 9478 Vibrio Parahaemolyticus 10290 p18-a5-vp10290 20 (1.895) Sb (49,1.00 ); Sm (SG, 1x10.00); Cm (20:28) 100 9088 9458 % 8911 9419 0 m/z 8800 8900 9000 9100 9200 9300 9400 9500 9600 9700 9800 9900 10000 m/z

10. The Gulf coast strains have a different spectrum in this mass region giving a marker ion near 9588. [The similar mass value does not mean the proteins are related!] 9478.8 Vibrio Parahaemolyticus 10290 100 9088.4 V. parahaemolyticus -Washington State 9458.0 % 8911.1 9419.2 9947.0 8766.3 0 m/z 8800 8900 9000 9100 9200 9300 9400 9500 9600 9700 9800 9900 10000 9587.5 Vibrio Parahaemolyticus 2030 Texas Outbreak 9090.5 100 9459.4 9422.0 % 8912.5 0 8800 8900 9000 9100 9200 9300 9400 9500 9600 9700 9800 9900 10000 m/z

11. Proteomics and Mass Spectrometry Acid resistance and protein biomarkers in bacteria can be monitored by MALDI TOF MS of intact cells. The ions below from are marker proteins (HdeA and HdeB) from the acid resistance gene. 9060 9735 9060 9735 S. flexneri E. coli

12. MILESTONES • MALDI can differentiate bacteria by genus, species, and strain: • J.O. Lay, Jr., “MALDI TOF Mass Spectrometry and Bacterial Taxonomy” Trends in Analytical Chemistry, 19, 507 (2000). • Specific Biomarkers for virulence can be detected by MALDI: • R.D. Holland, C.R. Duffy, F. Rafii, J.B. Sutherland, T.M. Heinze, C.L. Holder, K.J. Voorhees and J.O. Lay, Jr., “Identification of Bacterial Proteins Observed in MALDI TOF Mass Spectra from Whole Cells”, Anal. Chem.71:3226-3230 (1999). • Biomarker proteins can sometimes be detected in contaminated media without pre-MS culture steps: • R.D. Holland, F. Rafii, T.M. Heinze, J.B. Sutherland, K.J. Voorhees and J.O. Lay, Jr. “MALDI TOF/MS detection of bacterial biomarker proteins isolated from contaminated water, lettuce and cotton cloth” Rapid Communications in Mass Spectrometry, 14:911 (2000).

13. Future Experiments: • Correlation of toxicity and strain types with MALDI spectra • Development of more powerful MS methods (MALDI/FTMS) • More accurate assignment of biomarker (protein) identity. • Benefits to FDA include: • Differentiation of strains from more difficult Vibrio species • Detection of biomarkers associated with antibiotic resistance • Applications to FDA programs in bioterrorism, proteomics • and even characterization of other cell types, possibly malignant cells, by MS.

14. Metastable Atom Bombardment Time of Flight Mass Spectrometry (MAB/TOF/MS) an Alternative Approach to Bacterial I.D. GOALS and OBJECTIVES • Rapid chemotaxonomic strain-specific bacterial identification • Development of bacterial databases and search strategies • Applications to food/seafood borne bacteria, especially Vibrio species (CFSAN & ORA) • Development of patents for new methods • Identification of bacteria without a prior cell-culture step

15. KEY FINDINGS TO DATE • Demonstrated that a multiplicity of laboratory variables distort mass spectral fingerprints. • Patented a simple algorithm to correct for such method-related spectral changes. • The correction is more practical than using identical conditions. • (US Pat. App. No. 60/239,549 filed 10/10/2000)

16. 100 % 0 m/z 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 Matching a Reference Spectrum and an Experimental Spectrum from a Field Test Using MAB TOF MS 100 114 Bacillus globigii(bioagent simulant) reference spectrum 162 98 % 272 168 136 124 254 186 192 204 218 284 255 242 236 362 341 386 410 0 ARCA3_15 63 (0.105) Cm (40:121) 114 A spectrum of airborne particulate collected down-wind from an actual release of Bacillus globigii by the Canadian Military (rapid analysis and no culture step!) 162 272 168 254 126 146 284 218 182 136 191 223 204 362 316 m/z

17. FUTURE DIRECTIONS • Experimental • -test Py-MAB-TOF- MS (from Dephy, Montreal) at NCTR. For applications in rapid speciations • Assemble and validate a 200-sample spectral database using bacteria from CFSAN and ORA reference collections. • Computational • License the patent on using a spectral correction method to mitigate laboratory-based variations • Develop similar algorithms to transform spectra from environmental samples to their equivalent laboratory (data base) spectra

18. Exp.# E7080 "Fresh Tag" Fulton Fish Market Consumer version printable In the bag test Commercial version before after

19. Exp.# E7080 Indole test Grind 20 g shrimp in 50 mL toluene and 5 mL 5% TCA for 1 minute Centrifuge puree for 30 minutes at 3500 rpm and decant off toluene layer Filter extract through a 0.45 mm syringe filter into a beaker containing anhydrous Na2SO4 GC-MS Method Colorimetric Method shrimp Std.

20. Exp.# E7080 Aldehydes & Sulfides test GAS PHASE TEST Solid phase purpal test (DEVELOPMENTAL) cod

21. Fresh Tag onpaper Exp.# E7081 post exposure Pre-exposure Component of Explosives Ammonium Nitrate 20 Hr. purge 40 mL/min AM UNK 5

22. In Collaboration with CFSANMethods Development for Bioactive Herbal Ingredients in Functional Foods Research Progress: Extraction and LC methods developed for 4 SJW components in tea powder, fortified drinks, puffs and snack bars Methods Developed for 5 phenolic compounds in echinacea capsules and tablets

23. Potential Toxicity of Herbal Constituents Investigators: J. Leakey, C.Ang, R. Cecotti, Y. Cui. Objectives: 1. To develop human cell-based assays to determine whether a test substance affects key enzymes involved in the metabolism of pharmaceuticals. 2. To use these assay systems to investigate potential drug-herb interactions between prescribed pharmaceuticals and dietary supplements.

24. PRELIMINARY FINDINGS • Developed methods for isolating hyperforin, the major active ingredient of St. John’s Wort. • Developed or procured battery of cell lines expressing major isoforms of human drug metabolizing enzymes: used in inhibition assays. • Established that constituents of Echinacea inhibit enzymes conjugating estrogens.

25. Future Work • Develop human hepatocyte-based assay systems for measuring drug metabolizing enzyme induction. • Investigate the metabolism of active ingredients of St. John’s Wort by human enzymes. • Apply inhibition and induction assays to other herbal products. • Apply gene array technology and ultimately proteomics to elucidate mechanisms of action. • Isolate and identify the inhibitory constituents of Echinacea and St. John’s Wort.

26. O H C H 3 H H H 200 150 100 50 0 H O Protocol E0706801: Relationship between Structure-Activity Relationship (SAR) and Spectrometric Data-Activity Relationship (SDAR) Modeling Spectra Structure SDAR/QSDAR SAR/QSAR Biological Activity

27. SDAR model of 108 compounds binding to the estrogen receptor using NMR and MS data. QSDAR model of 26 poly- chlorinated dibenzofurans binding to the aryl receptor using predicted NMR data. -3.5 0.98 -0.50 Component 2 Predicted Log EC50 -9.5 -0.48 1.1 Log EC50 -9.5 -3.5 Component 1 Success of SDAR and QSDAR Models

28. SDAR Publications and Patents • 13C NMR and EI Mass Spectrometric Data to Produce a Predictive Model of Estrogen Receptor Binding Toxicology and Applied Pharmacology. 169: 17-25, 2000. • Producing 13C NMR, Infrared Absorption and EI Mass Spectrometric Data Monodechlorination Models of Chlorobenzenes, Chlorophenols, and Chloroanilines J. Chem. Inf. Comput. Sci. 40:1449-1455, 2000. • Developing 13C NMR Quantitative Spectrometric Data-activity Relationship (QSDAR) Models to the Corticosteroid Binding Globulin. J. Comput.-Aided Molec. Design. • Models of Polychlorinated Dibenzodioxins, Dibenzofurans, and Biphenyls Binding Affinity to the Aryl Hydrocarbon Receptor Developed Using 13C NMR Data. J. Chem. Inf. Comput. Sci. • Patent Pending for “Methods for Predicting the Biological, Chemical, and Physical Properties of Molecules From Their Spectral Properties.”

29. Future Directions of SDAR • Protocol E0706801: “Continuing to develop SDAR models for the Ames test, neuraltoxicity (Neurotox), and other toxic endpoints” • Protocol E0706811: “Developing new strategies for spectrometric models of toxicity” (ROW) • Protocol E0708301: “Computational predictive system for rodent organ-specific carcinogenicity” (Biometry, CDER, ROW) • Producing hybrid spectrometric models that incorporate three-dimensional structural information directly into the SDAR model.

30. O N N H O H H N HO N H C C H N H N N 3 2 N H O H N C O O H N O H C O H O 3 N O H N H N N H C C H 3 3 O N T NHSO - 3 N N H C N CH N H 3 3 2 N N N N H C C H 3 3 N Risk Assessment, Interspecies Extrapolaton and Predictive Toxicology with Biomarkers and Computational Chemisry ? ? Interspecies extrapolation DNA A T A T G C C G DNA adduct QSAR SDAR QSDAR in vitro Metabolites Structure & activity Spectra & activity Computational Chemistry

31. Protocol E07077.01: Comparison of Principal Components Analysis (PCA) and Artificial Neural Networks (ANN) for the Prediction of Qualitative and Quantitative Biological End Points from Spectrometric Data Chemical Spectrum Artificial Neural Network Predicted Biological End Point

32. Success of Quantitative Spectral Data Activity Relationship Artificial Neural Network Model (QSDAR-ANN) QSDAR-ANN model results of 28 Poly-chlorinated Biphenyl, Dioxin, and Furan Toxic Equivalence Factors (TEFs) using predicted 13C NMR spectra.

33. Publications : Predicting Toxic Equivalent Factors from NMR Spectra for Dioxins Furans and PCBs Using Principle Components Analysis and Artificial Neural Networks, Environmental Health Perspectives, manuscript in preparation (2001). Future Directions: • Currently developing a quantum mechanical parameter based neural network model for the prediction of TEFs for the dioxins and dioxin-like compounds. • Development of an internet parallel distributed neural network to allow for the handling of large data sets as well as increasing the efficiency of the neural network.

34. A New Approach to the NMR Spectroscopy of Drug Purity and the Public Health Implications (E070781) Objectives: · Determine properties and develop procedures for use of NMR spectrometer at the NCTR under high dynamic range conditions. · Develop concepts and methodology for application of NMR spectroscopy to investigation of very-low-level impurities in drugs using results on genistein as a model

35. MS Instrumentation Available (or Planned) for Proteomics Instrument Application LC Triple Quadrupole /MS ESI MW determination for isolated proteins confirmation of MW for peptides/small proteins Quadrupole TOF MS MALDI and LC/ESI for sequencing especially for tagged proteins in measurement or relative levels of expression SELDI MALDI of affinity surfaces rapid screening of dirty samples for end-point specific proteins {Offsite} MALDI TOF MS {at UAF} MW determination for proteins and digests MALDI FTMF {at UAF} more accurate mass assignments and analysis of whole cells

36. Mass Spectrometry Applications in FDA Research Initiatives Allergenicity Bacteria Taxonomy/Speciation Bioterrorism Drug Purity (Chemicals and Recombinant Proteins) Ion Mobility MS (Protein conformation, configuration) Microbial metabolism (biotransformation of drugs, contaminants, and antibiotics/resistance) Proteomics Quality Assurance and Compliance Rapid through-put Analysis Redox Status (Vitamins, Lipids, Proteins, DNA) Risk Assessment (Biomarkers, DNA- and Protein Adducts, DNA Damage, Metabolites)

37. Development of a New Tandem Instrumental Approach to the Detection of Prions: HPLC/IMS/MS Chromatographic HPLC Mixture Resolution (liquid-phase) (protein) prion level) time(min) Mobility Based IMS Prion Separation (gas-phase) (folding) changes) time (ms) MS Based MS Prion Detection (high-vacuum) m/z (mass confirmation)

38. NMR Spectroscopy Applications in FDA Research Initiatives Computational Chemistry Metabolomics Drug Purity Proteomics LC-NMR-MS

39. In vivo NMR for Detection of Biomarkers and the Intermediates of Metabolic Pathways Example: Downs Syndrome 13CbH2- Serine 515N-THF Met 5,10-13CH2-THF HCN NMR experiment can monitor 5,10-13CH2-THF and515N-13CH3-THF compounds Homo- Cysteine 515N-13CH3- THF • Cost to Upgrade NMR ~ $350,000 or New NMR ~ $550,000 • Cost of labeled compounds ~ $15,000/year

40. Areas Supported in FY-2001 • Ethinyl Estradiol on Bone Growth in Rats • Erythromycin from Farmed Animals • Malachite Green/Leuco Malachite Green in Mice • Retinyl Palmitate: Isolation & Detection • DNA Adducts of Tamoxifen • Dietary Supplements & Herbals: Identification of Bioactive Ingredients

41. Areas Supported in FY-2001(continued) • Endocrine Disrupters: Genistein & Daidzein • Phytoestrogen Conversion to Estrogenic Compounds: Genistein & Daidzein • Fluoroquinolone Biotransformation by Fungi • Microbial Degradation of Drugs & Feed Additives in Aquaculture • Antihistamine Drugs in Neonatal Mouse Cells