An isocratic separation of underivatized gentamicin components, 1H NMR assignment and protonation pattern

A simple method for the separation of the major components of commercial gentamicin sulfate (C-1, C-1a, C-2, C-2a) by high-performance liquid chromatography (HPLC) on an analytical and a semipreparative scale was developed. The method utilized ion-pair reversed-phase chromatography, isocratic elution with an aqueous solution containing 9% trifluoroacetic acid and 2.5% acetonitrile as the mobile phase at a flow rate of 2 and 9 mL/min for analytical and semipreparative columns, respectively. Detection was carried out at 213 nm without derivatization. The protonation pattern of the separated gentamicins was determined by potentiometry and 15N and 1H NMR. The full proton NMR assignment for gentamicin C-1 was obtained through the use of 1H 1D and 2D 1H-1H COSY measurements.     

Purification and characterization of subtilisin inhibitors 'Marinostatin' produced by marine Alteromonas sp.

The four novel peptide subtilisin inhibitors, which have been named Marinostatin B-1, B-2, C-1 and C-2, were purified from the culture supernatant fluid of a marine bacterium, Alteromonas sp. A combination of Diaion HP-20 and CM-cellulofine ion exchange column chromatographies, Sephadex G-25 gel filtration and preparative high performance liquid chromatography (HPLC) were employed. Final preparations gave a single peak on HPLC. The molecular weights by gel filtration of Marinostatin B-1 and C-1 were estimated to be 1500, and those of B-2 and C-2 were 1700. All the inhibitors were stable in acidic (pH range 4–6) but less stable in alkaline solutions (pH 10). The metal ions Co²⁺ and Fe²⁺ repressed the inhibitory activity by 30 and 20%, respectively. The four inhibitors had inhibitory effects on serine proteases like α-chymotrypsin.     

Regio- and stereoselective metabolism of 7,12-dimethylbenz[a]anthracene by Mycobacterium vanbaalenii PYR-1

The degradation of 7,12-dimethylbenz[a]anthracene (DMBA), a carcinogenic polycyclic aromatic hydrocarbon, by cultures of Mycobacterium vanbaalenii PYR-1 was studied. When M. vanbaalenii PYR-1 was grown in the presence of DMBA for 136 h, high-pressure liquid chromatography (HPLC) analysis showed the presence of four ethyl acetate-extractable compounds and unutilized substrate. Characterization of the metabolites by mass and nuclear magnetic resonance spectrometry indicated initial attack at the C-5 and C-6 positions and on the methyl group attached to C-7 of DMBA. The metabolites were identified as cis-5,6-dihydro-5,6-dihydroxy-7,12-dimethylbenz[a]anthracene (DMBA cis-5,6-dihydrodiol), trans-5,6-dihydro-5,6-dihydroxy-7,12-dimethylbenz[a]anthracene (DMBA trans-5,6-dihydrodiol), and 7-hydroxymethyl-12-methylbenz[a]anthracene, suggesting dioxygenation and monooxygenation reactions. Chiral stationary-phase HPLC analysis of the dihydrodiols showed that DMBA cis-5,6-dihydrodiol had 95% 5S,6R and 5% 5R,6S absolute stereochemistry. On the other hand, the DMBA trans-5,6-dihydrodiol was a 100% 5S,6S enantiomer. A minor photooxidation product, 7,12-epidioxy-7,12-dimethylbenz[a]anthracene, was also formed. The results demonstrate that M. vanbaalenii PYR-1 is highly regio- and stereoselective in the degradation of DMBA.     

一株能水解巴卡亭ⅢC-10位乙酰基的成团泛菌的筛选和鉴定

10-脱乙酰巴卡亭Ⅲ是合成紫杉醇和多烯紫杉醇的前体。以巴卡亭Ⅲ为底物,结合TLC、HPLC、HPLC-MS分析方法,通过设计专门的筛选方法筛选产酶菌株,得到一株巴卡亭ⅢC-10位脱乙酰基酶产生菌株Z1-56。以形态特征、生理生化特征、16S rDNA序列分析作为菌株的鉴定手段,Z1-56被鉴定为成团泛菌(Pantoea agglomerans),首次发现成团泛菌产生巴卡亭ⅢC-10-脱乙酰酶。     

Determination of yohimbine and its two hydroxylated metabolites in humans by high-performance liquid chromatography and mass spectral analysis

The existence of at least two metabolites of yohimbine (YO) in humans is demonstrated. Combined high-performance liquid chromatographic (HPLC), NMR and mass spectral analyses permitted them to be identified as hydroxylated metabolites at the C-10 and C-11 positions. A normal-phase HPLC method allowing the simultaneous determination of YO and its main metabolite, 11-hydroxyyohimbine (11-OHYO), in biological samples is described. This assay was performed using a LiChrosorb Si 60 column and a mobile phase consisting of 0.02 M sodium acetate (pH 5)—methanol (5:95, v/v) at a flow-rate of 1 ml/min. Detection was achieved by a fluorimetric method (excitation at 280 nm and emission at 320 nm). The extraction yields of YO, 10-OHYO and 11-OHYO from plasma were 91.8, 45.3 and 17.8%, respectively, and their respective within-day reproducibilities were 3.8, 1.4 and 5.9%. The between-day reproducibility for YO at the concentrations of 1 and 10 ng/ml were 8.9 and 6.4%, respectively. The accuracy of the method for YO at concentrations of 1 and 10 ng/ml were 5.1 and 2.3%, respectively. The limits of determination of YO, 10-OHYO and 11-OHYO were 0.1, 0.5 and 1 ng/ml, respectively. The method was used in bioavailability study of YO following oral and intravenous administration in humans.     

HPLC methods for detection of uniconazole-P in soils and plant tissues

High-performance liquid chromatographic (HPLC) methods have been developed for the detection of uniconazole-P [(E)-1-(4-chlorophenyl)-4,4,-dimethyl-2-(1,2,4-triazol-1-yl)-1-penten-3-ol; XE-1019; the active ingredient in Prunit and Sumagic] in soil and plant tissue samples. Methanolic extracts of soil and plant samples were dried to the aqueous phase, the pH adjusted to 11, and partitioned against methylene chloride. The methylene chloride phases were washed with pH 11 water and then passed through C-18 solid phase extraction (SPE) columns. The soil extracts were then dried and the residues taken up in 1 ml acetonitrile of which 20 l were injected directly onto a C-18 reverse phase analytical column for HPLC analysis. Plant tissue extracts were purified by partitioning and passing through a sequence of Florisil/C-18/Florisil SPE columns before HPLC analysis. Recovery of uniconazole-P was 70% from soils and 40% from plant tissues. Quantitative detection of 10 parts per billion (ppb) uniconazole-P in plant tissues and soil samples was feasible following these procedures. The soil cleanup procedures were also used to detect uniconazole-P in leachates collected from container-grown plants.     

Carbohydrate structure of erythropoietin expressed in Chinese hamster ovary cells by a human erythropoietin cDNA

The proper glycosylation of erythropoietin is essential for its function in vivo. Human erythropoietins were isolated from Chinese hamster ovary cells transfected with a human erythropoietin cDNA and from human urine. Carbohydrate chains attached to these proteins were isolated and fractionated by anion-exchange high performance liquid chromatography (HPLC) and HPLC employing a Lichrosorb-NH2 column. The structures of fractionated saccharides were analyzed by fast atom bombardment-mass spectrometry and methylation analysis before and after treatment with specific exoglycosidases. Both erythropoietins were found to contain one O-linked oligosaccharide/mol of the proteins, and its major component was elucidated to be NeuNAc alpha 2----3Gal beta 1----3(NeuNAc alpha 2----6)GalNAcOH (where NeuNAc represents N-acetylneuraminic acid) in both proteins. The N-linked saccharides of recombinant erythropoietin were found to consist of biantennary (1.4% of the total saccharides), triantennary (10%), triantennary with one N-acetyllactosaminyl repeat (3.5%), tetraantennary (31.8%), and tetraantennary with one (32.1%), two (16.5%), or three (4.7%) N-acetyllactosaminyl repeats. All of these saccharides are sialylated by 2----3-linkages. Tetraantennary with or without polylactosaminyl units are mainly present as disialosyl or trisialosyl forms, and these structures exhibit the following unique features. alpha 2----3-Linked sialic acid and N-acetyllactosaminyl repeats are selectively present in the side chains attached to C-6 and C-2 of 2,6-substituted alpha-mannose and C-4 of 2,4-substituted alpha-mannose. We have also shown that the carbohydrate moiety of urinary erythropoietin is indistinguishable from recombinant erythropoietin except for a slight difference in sialylation, providing the evidence that recombinant erythropoietin is valuable for biological as well as clinical use.     

Isolation and partial structural characterization of an equine fibrinogen CNBr fragment that exhibits immunologic cross-reactivity with an A alpha-chain cross-linking region of human fibrinogen

Immunochemical studies of equine fibrinogen were conducted to characterize the structural basis for the immunologic cross-reactivity observed between human and equine A alpha chains when employing an antiserum to the 26K, human cyanogen bromide (CNBr) fragment, A alpha 241-476 (CNBr VIII). A 38K, equine CNBr fragment that reacts with this antiserum was isolated from CNBr-digested equine fibrinogen by Sephadex G-100 gel filtration. It was further purified by sequential hydrophobic chromatography on phenyl-Sepharose CL-4B, followed by reversed-phased (C-8) high-performance liquid chromatography (HPLC). NH2-Terminal analysis of the purified fragment, designated EqA alpha CNBr, identified one major sequence whose first three residues, E-L-E, were identical with those of human CNBr VIII. Tryptic and staphylococcal protease digests of the equine fragment were resolved by reversed-phase HPLC (C-4, C-18), and the separated components were characterized by amino acid analysis and automated Edman degradation. A total of 34 tryptic and 20 staph protease peptides yielded sequence information that permitted the alignment of 271 equine residues with residues A alpha 241-517 from the COOH-terminal two-thirds of the human A alpha chain so that 63% of the possible matches were identical. Other features of interest included (1) an amino acid substitution in which the methionine residue at A alpha 476 in the human A alpha chain was replaced by a valine residue, thus accounting, in part, for the larger EqA alpha CNBr fragment obtained from the equine molecule, and (2) a region of striking homology in which 36 successive residues, corresponding to A alpha 428-464 in the human A alpha chain, were identical in both species. These findings, together with available structural data for the COOH-terminal portion of the rat and bovine A alpha chains, indicate that the region corresponding to (human) A alpha 240-517 represents a conserved portion of the fibrinogen molecule. This may, in turn, explain the difficulties encountered when trying to raise monoclonal antibodies to cross-linking regions that are contained within the COOH-terminal two-thirds of the human A alpha chain.     

High performance liquid chromatography preparation of the molecular species of GM1 and GD1a gangliosides with homogeneous long chain base composition

A semi-preparative, analytical high performance liquid chromatographic (HPLC) procedure is described for the isolation of molecular species of GM1 and GD1a gangliosides containing a single long chain base, C18 or C20 sphingosine, C18 or C20 sphinganine, each in its natural erythro or unnatural threo form. The threo forms were obtained from 2,3-dichloro-5,6-dicyanobenzoquinone/NaBH4 -treated gangliosides. The ganglioside molecular species separated by HPLC were analyzed for carbohydrate, fatty acid, and long chain base composition. In particular, long chain bases were submitted to gas-liquid chromatographic-mass spectrometric analyses as their trimethylsilyl (TMS) or N-acetyl-TMS derivatives, and chain length, presence or absence of C4-C5 double bond, and C-3 steric configuration were ascertained. The final preparations of individual molecular species of GM1 and GD1a gangliosides were more than 99% homogeneous in their saccharide moiety, contained a single long chain base (homogeneity higher than 99%), and had a fatty acid composition primarily of stearic acid (92 to 97%). All the individual molecular species of GM1 and GD1a gangliosides were also prepared in radioactive form by selective tritiation at C-3 of the long chain base. Their specific radioactivity ranged from 1.3 to 1.45 Ci/mmol. The availability of these molecular species of gangliosides is expected to facilitate studies aimed at ascertaining the role played by the hydrophobic portion in the functional behavior of gangliosides.     

Separation and structural analysis of some saponins from Quillaja saponaria Molina

A fraction of saponins from Quillaja saponaria Molina, QH-B, was fractionated by consecutive separations on three different reverse-phase HPLC systems. Eight compounds were isolated and the structures of these were elucidated mainly by sugar analysis and NMR spectroscopy. The structures consisted of a quillaic acid substituted with two different trisaccharides at C-3, beta-D-Galp-(1-->2)-[alpha-L-Rhap-(1-->3)]-beta-D-GlcpA and beta-D-Galp-(1-->2)-[beta-D-Xylp-(1-->3)]-beta-D-GlcpA, and a tetra- or pentasaccharide at C-28, beta-D-Xylp-(1-->4)-[beta-D-Glcp-(1-->3)]-alpha-L-Rhap-(1--> 2)-beta-D-Fucp and beta-D-Apif-(1-->3)-beta-D-Xylp-(1-->4)-[beta-D-Glcp-(1-->3) ]-alpha-L- Rhap-(1-->2)-beta-D-Fucp. These compounds were further substituted with an acyl group either at O-3 or O-4 of the fucose residue, which is the sugar linked to C-28 of the quillaic acid.     

Glucose Catabolism in Rhizobium japonicum

Glucose catabolism in Rhizobium japonicum ATCC 10324 was investigated by the radiorespirometric method and by assaying for key enzymes of the major energy-yielding pathways. Specifically labeled glucose gave the following results for resting cells, with values expressed as per cent (14)CO(2) evolution: C-1=59%, C-2=51%, C-3=45%, C-4=59%, and C-6=43%. These values indicate that glucose was degraded by the Entner-Doudoroff pathway alone. Cells which grew in glucose-yeast extract-salts medium gave essentially the same pattern except for retardation of the C-6 carbon. The rates were: C-1=54%, C-2=42%, C-3=51%, C-4=59%, and C-6=32%. Hexokinase, glucose-6-phosphate dehydrogenase, transketolase, and an enzyme system which produces pyruvate from 6-phosphogluconate were found to be present in these cells. No 6-phosphogluconate dehydrogenase was detected. Oxidation of specifically labeled pyruvate gave the following (14)CO(2) evolution pattern: C-1=78%, C-2=48%, and C-3=37%; the pattern from acetate was C-1=73%; and C-2=56%. Oxidation of glutamate showed the preferential rate of (14)CO(2) evolution to be C-1 > C-2=C-5 > C-3, 4, whereas a higher yield of (14)CO(2) was obtained from the C-1 and C-4 carbons of succinate than from the C-2 and C-3 carbons. These data are consistent with the operation of the Entner-Doudoroff pathway and tricarboxylic acid cycle as the catabolic pathways of glucose oxidation in R. japonicum.     

Purification and characterization of a heat-stable enterotoxin of Vibrio mimicus

A heat-stable enterotoxin produced by Vibrio mimicus (VM-ST) was studied. VM-ST was purified from a culture supernatant of V. mimicus strain AQ-0915 by ammonium sulfate fractionation, hydroxyapatite treatment, ethanol extraction, column chromatography on both SP-Sephadex C-50 and DEAE-Sephadex A-25, and HPLC, and the recovery rate was about 15%. Purified VM-ST was heat-stable. VM-ST activity was cross-neutralized by anti-STh antiserum. The amino acid composition of the purified VM-ST was determined 17 amino acid residues in the following sequence: Ile-Asp-Cys-Cys-Glu-Ile-Cys-Cys-Asn-Pro-Ala-Cys-Phe-Gly-Cys-Leu-Asn. This composition and sequence were identical to those of V. cholerae non-O1-ST. These results clearly demonstrate the production of a characteristic VM-ST by V. mimicus.     

Ecdysone metabolism in Pieris brassicae during the feeding last larval instar

Ecdysone metabolism in Pieris brassicae during the feeding last larval stage was investigated by using ³H-labeled ecdysteroid injections followed by high-performance liquid chromatographic (HPLC

1 Abbreviations: 3DE = 3-dehydroecdysone; 3D20E = 3-dehydro-20-hydroxyecdysone; 2026E = 20,26-dihydroxyecdysone; E = ecdysone; Eoic = ecdysonoic acid; 2026E′ = 3-epi-20,26-dihydroxyecdysone; E′ = 3-epiecdysone; E′oic = 3-epiecdysonoic acid; E′8P = 3-epiecdysone 3-phosphate; 20E′ = 3-epi-20-hydroxyecdysone; 20E′3P = 3-epi-20-hydroxyecdysone 3-phosphate; FT = Fourier transform; HPLC = high-performance liquid chromatography; 20E = 20-hydroxyecdysone; 20Eoic = 20-hydroxyecdysonoic acid; NMR = nuclear magnetic resonance; NP-HPLC = normal phase HPLC; RP-HPLC = reverse phase HPLC; TFA = trifluoroacetic acid; Tris = tris(hydroxymethyl)-aminomethane.

) analysis of metabolites. Metabolites were generally identified by comigration with available references in different HPLC systems. Analysis of compounds for which no reference was available required a large-scale preparation and purification for their identification by ¹H nuclear magnetic resonance spectrometry. The metabolic reactions affect the ecdysone molecule at C-3, C-20, and C-26, leading to molecules which are modified at one, two, or three of these positions. At C-20, hydroxylation leads to 20-hydroxyecdysteroids. At C-26, hydroxylation leads to 26-hydroxyecdysteroids which can be further converted into 26-oic derivatives (ecdysonoic acids) by oxidation. At C-3, there are several possibilities: there may be oxidation into 3-dehydroecdysteroids, or epimerization possibly followed by phosphate conjugation. Thus, injected 20-hydroxyecdysone was converted principally into 20-hydroxyecdysonoic acid, 3-dehydro-20-hydroxyecdysone, and 3-epi-20-hydroxyecdysone 3-phosphate. Labelled ecdysone mainly gave the same metabolites doubled by a homologous series lacking the 20-hydroxyl group.     

Structures of Dimers Produced from Methyl Linoleate during Initial Stage of Autoxidation

A structural investigation on the main fraction of dimers formed during the induction period of autoxidation in methyl linoleate (ML) was carried out. The dimeric fraction (A₂), which was isolated from the autoxidized ML (POV= 18) by various Chromatographic techniques and gave a single spot on TLC, was further separated into four major components (components 1–4) by high performance liquid chromatography (HPLC). The mean molecular weights of these components were found to be 643–655 and component 4 gave the parent peak 652 on an FD-mass spectrum which corresponded to 2 × ML + 4O. The reduced products of each component with stannous chloride were identified in common as methyl 9- or 13-hydroxy octadecadienoate and methyl 9,13-dihydroxy octadecenoate by GC-MS. These results show that all of these dimers contained a peroxide bridge linking between a pair of MLs across C-9 or C-13 on one of the MLs and C-9 or C-13 on the other, with a hydroperoxy group.     

Lipase-catalyzed acetylation of N-acetylneuraminic acid derivative

A facile preparation of triacetylated derivative of 2-phenylthioglycoside of N-acetylneuraminic acid (4) was achieved by treatment with lipase PS in vinyl acetate. The major product 4 has a free hydroxyl group at C-7. Results of time-course HPLC analysis indicate that the reactivity of the hydroxyl groups under this condition is in the following order; C-9 > C-4 > C-8 > C-7.     

A high-performance liquid chromatographic method to measure sphingosine 1-phosphate and related compounds from sphingosine kinase assays and other biological samples

Sphingosine 1-phosphate is an intermediate of sphingosine catabolism as well as a potent signaling compound. Conditions were established for the extraction and analysis of sphingosine 1-phosphate and other sphingoid base 1-phosphates from in vitro sphingosine kinase assays and other biological samples. The sphingoid base 1-phosphates were extracted in high yield (85%) using small C-18 reverse-phase columns (LiChroprep RP-18). After the extracts were treated with 0.1 N KOH to remove glycerolipids, the sphingoid base 1-phosphates were converted to fluorescent o-phthalaldehyde derivatives that were separated by HPLC using C-18 columns with a mobile phase of methanol:10 mM potassium phosphate (pH 7.2):1 M tetrabutylammonium dihydrogen phosphate (in water) (83:16:1, v/v/v). The o-phthalaldehyde derivative of sphingosine 1-phosphate was reasonably stable (t(1/2) > or = 18 h) when EDTA was present and could be detected in picomole amounts. The HPLC retention time of the sphingoid base 1-phosphates could be shifted by adjusting the mobile phase to pH 5.5, which is useful in separating overlapping compounds (such as sphingosine 1-phosphate and 4-D-hydroxysphinganine) and in confirming the identity of sphingoid base 1-phosphates in biological samples. The extraction procedure and HPLC method facilitated assays of sphingosine kinase with different sphingoid bases as substrates and/or inhibitors and enabled the quantitation of sphingoid base 1-phosphates in human plasma, serum, and platelets as well as in strains of Saccharomyces cerevisae with mutations in sphingolipid metabolism.     

Chiral HPLC separation and CD spectra of the C-2 diastereomers of naringin in grapefruit during maturation

Naringin is the chief flavanone glycoside of grapefruit (Citrus paradisi). It is responsible for part of the bitter taste of the fruit and can cause the inhibition of some cytochrome P450s. The direct separation of (2R)- and (2S)-naringin in the albedo of grapefruits was obtained in normal phase HPLC mode using Chiralcel OD as chiral stationary phase and n-hexane/ethanol with 0.1% of TFA as mobile phase. Chiralpak AD was almost ineffective in the separation. This procedure was used to evaluate the stereochemistry at C-2 during maturation of the grapefruit. The CD curves of (2R)- and (2S)-naringin isolated by semipreparative chiral HPLC were determined and the elution order of the chromatographic peaks was related to the absolute C-2 configuration. Partial resolution of the C-2 diastereomers of narirutin was obtained on Chiralpak AD.     

Calix[4]arene methylenebisphosphonic acids as inhibitors of fibrin polymerization

Calix[4]arenes bearing two or four methylenebisphosphonic acid groups at the macrocyclic upper rim have been studied with respect to their effects on fibrin polymerization. The most potent inhibitor proved to be calix[4]arene tetrakis-methylene-bis-phosphonic acid (C-192), in which case the maximum rate of fibrin polymerization in the fibrinogen + thrombin reaction decreased by 50% at concentrations of 0.52 × 10(-6) M (IC(50)). At this concentration, the molar ratio of the compound to fibrinogen was 1.7 : 1. For the case of desAABB fibrin polymerization, the IC(50) was 1.26 × 10(-6) M at a molar ratio of C-192 to fibrin monomer of 4 : 1. Dipropoxycalix[4]arene bis-methylene-bis-phosphonic acid (C-98) inhibited fibrin desAABB polymerization with an IC(50) = 1.31 × 10(-4) M. We hypothesized that C-192 blocks fibrin formation by combining with polymerization site 'A' (Aα17-19), which ordinarily initiates protofibril formation in a 'knob-hole' manner. This suggestion was confirmed by an HPLC assay, which showed a host-guest inclusion complex of C-192 with the synthetic peptide Gly-Pro-Arg-Pro, an analogue of site 'A'. Further confirmation that the inhibitor was acting at the initial step of the reaction was obtained by electron microscopy, with no evidence of protofibril formation being evident. Calixarene C-192 also doubled both the prothrombin time and the activated partial thromboplastin time in normal human blood plasma at concentrations of 7.13 × 10(-5) M and 1.10 × 10(-5) M, respectively. These experiments demonstrate that C-192 is a specific inhibitor of fibrin polymerization and blood coagulation and can be used for the design of a new class of antithrombotic agents.     

Production of a co-polyester of 3-hydroxybutyric acid and 3-hydroxyvaleric acid from succinic acid by Rhodococcus ruber: biosynthetic considerations

The biosynthesis of the 3-hydroxyvalerate (3HV) monomer of polyhydroxyalkanoate by Rhodococcus ruber from succinic acid was investigated using nuclear magnetic resonance analysis. Polymer produced from [2,3-¹³C]- and [1,4-¹³C]succinate showed that the C-1-C-2 and C-4-C-5 fragments of 3HV were derived from carbons 2 and 3 of succinate, essentially without bond cleavage, and carbon 3 of 3HV was derived from a carboxyl carbon of succinate. Using [1,2-¹³C]succinate it was demonstrated that the C-1-C-2 bond of succinate was cleaved during polymer biosynthesis. Methylmalonyl-coenzyme A (CoA) mutase activity was detected in cell-free extracts of R. ruber by enzyme assay and HPLC analysis of reaction products. A pathway, involving the known methylmalonyl-CoA pathway for propionate formation in Propionibacteria, followed by the established pathway for PHA biosynthesis from propionyl-CoA and acetyl-CoA, is proposed for the biosynthesis of 3HV from succinate by R. ruber. Correspondence to: A. J. Anderson     

Regio and stereospecific DNA adduct formation in mouse lung at N6 and N7 position of adenine and guanine after 1,3 butadiene inhalation exposure

Butadiene monoepoxide (BMO) alkylated guanine N7 and adenine N 6 adducts were prepared and enriched by solid phase extraction and HPLC. The purified adducts were analysed by a modified 32P-postlabelling assay, which utilized one dimensional TLC chromatography and a subsequent HPLC analysis with UV and radioactivity detectors. In vitro with Ct-DNA the formation of N7-dGMP and N 6-dAMP adducts were linear at a concentration range of 44 to 870 nmol of BMO per mg DNA at physiological pH. N7- dGMP and N 6-dAMP adducts were formed in a ratio of 200:1. In dGMP and in dAMP 48 % and 86 % of adducts were covalently bound to the C-2 carbon of BMO. CD-1 mice were inhalation exposed to butadiene for 5 days and 6 h per day. The N7-dGMP adduct level in lung samples of animals exposed to 200, 500 and 1300 ppm was 2.8 +/- 0.9 fmol, 11 +/- 2.0 fmol and 30 +/- 6.7 fmol in 10 mug DNA, respectively. The level of N 6-dAMP adducts in lung samples after 500 ppm and 1300 ppm exposure was 0.09 +/- 0.06 fmol and 0.11 +/- 0.05 fmol in 10 mug DNA. At 200 ppm the adduct level was below the detection limit. A sub-group of animals exposed to 1300 ppm was killed 3 weeks after the last exposure. N7-dGMP adducts were not detected but the level of N 6-dAMP adducts was not affected. N7-dGMP adducts were formed in a clear stereospecific manner in vivo. S -BMO adducts were the main product and represented 77 % (n = 4, SD = 2%) of total BMO adducts. No clear conclusion can be drawn about the enantiospecific DNA binding at the N 6 position of dAMP, because of the poor separation of the enantiomers. However, we could separate regioisomeric adducts which indicated that C-2 adducts represented 69 +/- 3 % of the total N 6 adducts formed in mice lung DNA. This observation is supported by the data derived from in vitro DNA experiments but is different to our previously published data, which indicates the 2:1 (C-1:C-2) ratio in regioisomer formation in nucleotides or nucleosides. We suggest that the data presented in this communication indicate a different mechanism between nucleotides and DNA in BMO-derived adduct formation- Dimroth rearrangement dominates in nucleotides, but in double stranded DNA a direct alkylation is probably the major mechanism of adduct formation.