Simultaneous determination of two human urinary metabolites of N,N-dimethylformamide using gas chromatography–thermionic sensitive detection with mass spectrometric confirmation

Two human urinary metabolites of the industrial solvent N,N-dimethylformamide (DMF), N-hydroxymethyl-N-methylformamide (HMMF) and N-acetyl-S-(N-methylcarbamoyl)cysteine (AMCC), were assayed using a new analytical method (gas chromatography and thermionic sensitive detection). Clean-up of urine samples includes a liquid–liquid extraction step followed by a solid-phase extraction step to separate HMMF and AMCC from other urine components. During clean-up, AMCC is converted into ethyl-N-methylcarbamate (EMC), and during gas chromatography, HMMF is degraded in the injector to N-methylformamide (NMF). All the validation data necessary for a quantitative procedure are given. The method was applied to urine samples from workers exposed to DMF and from the general population. The results were confirmed by mass spectrometric determination. For this purpose a further liquid–liquid extraction step was introduced in the clean-up procedure. Background levels of AMCC in the general population were identified.     

Determination of perfluorodecalin and perfluoro-N-methylcyclohexylpiperidine in rat blood by gas chromatography–mass spectrometry

A gas chromatography–mass spectrometry method (SIM mode) was developed for the determination of perfluorodecalin (cis and trans isomers, 50% each) (FDC), and perfluoromethylcyclohexylpiperidine (3 isomers) (FMCP) in rat blood. The chromatographic separation was performed by injection in the split mode using a CP-select 624 CB capillary column. Analysis was performed by electronic impact ionization. The ions m/z 293 and m/z 181 were selected to quantify FDC and FMCP due to their abundance and to their specificity, respectively. The ion m/z 295 was selected to monitor internal standard. Before extraction, blood samples were stored at −30°C for at least 24 h in order to break the emulsion. The sample preparation procedure involved sample clean-up by liquid–liquid extraction. The bis(F-butyl)ethene was used as the internal standard. For each perfluorochemical compound multiple peaks were observed. The observed retention times were 1.78 and 1.87 min for FDC, and 2.28, 2.34, 2.48 and 2.56 min for FMCP. For each compound, two calibration curves were used; assays showed good linearity in the range 0.0195–0.78 and 0.78–7.8 mg/ml for FDC, and 0.00975–0.39 and 0.39–3.9 mg/ml for FMCP. Recoveries were 90 and 82% for the two compounds, respectively with a coefficient of variation <8%. Precision ranged from 0.07 to 15.6%, and accuracy was between 89.5 and 111.4%. The limits of quantification were 13 and 9 μg/ml for FDC and FMCP, respectively. This method has been used to determine the pharmacokinetic profile of these two perfluorochemical compounds in blood following administration of 1.3 g of FDC and 0.65 g of FMCP per kg body weight, in emulsion form, in rat.     

Determination of isoprostaglandin F2α type III in human urine by gas chromatography–electronic impact mass spectrometry. Comparison with enzyme immunoassay

F₂-Isoprostanes are stable lipid peroxidation products of arachidonic acid, the quantification of which provides an index of oxidative stress in vivo. We describe a method for analysing isoprostaglandin F_2α type III (15-F_2t-IsoP) in biological fluids. The method involves solid-phase extraction on octadecyl endcapped and aminopropyl cartridges. After conversion to trimethylsilyl ester trimethylsilyl ether derivatives, isoprostaglandin F_2α type III is analysed by mass spectrometry, operated in electronic impact selected ion monitoring mode. We have compared enzyme immunoassay (EIA; Cayman, Ann Arbor, MI, USA) to this method with 30 human urine aliquots following the same extraction procedure in order to determine the agreement between both methods. Isoprostaglandin F_2α type III concentrations determined with gas chromatography–mass spectrometry (GC–MS) did not agree with those determined with EIA. Our results suggest that GC–MS and EIA do not measure the same compounds. As a consequence, comparison of clinical results using GC–MS and EIA should be avoided.     

Quantitation of loperamide and N-demethyl-loperamide in human plasma using electrospray ionization with selected reaction ion monitoring liquid chromatography–mass spectrometry

We report here the development and validation of an LC–MS method for quantitation of loperamide (LOP) and its N-demethyl metabolite (DMLOP) in human plasma. O-Acetyl-loperamide (A-LOP) was synthesized by us for use as an internal standard in the assay. After addition of the internal standard, the compounds of interest were extracted with methyl tert.-butylether and separated by HPLC on a C₁₈ reversed-phase column using an acetonitrile–water gradient containing 20 mM ammonium acetate. The three compounds were well separated by HPLC and no interfering peaks were detected at the usual concentrations found in plasma. Analytes were quantitated using positive electrospray ionization in a triple quadrupole mass spectrometer operating in the MS–MS mode. Selected reaction monitoring was used to quantify LOP (m/z 477→266), DMLOP (m/z 463→252) and A-LOP (m/z 519→266) on ions formed by loss of the 4-(p-chlorophenyl)-4-hydroxy-piperidyl group upon low energy collision-induced dissociation. Calibration curves, which were linear over the range 1.04 to 41.7 pmol/ml (LOP) and 1.55 to 41.9 pmol/ml (DMLOP), were run contemporaneously with each batch of samples, along with low (4.2 pmol/ml), medium (16.7 pmol/ml) and high (33.4 pmol/ml) quality control samples. The lower limit of quantitation (LLQ) of LOP and DMLOP was about 0.25 pmol/ml in plasma. The extraction efficiency of LOP and DMLOP from human plasma was 72.3±1.50% (range: 70.7–73.7%) and 79.4±12.8% (64.9–88.8%), respectively. The intra- and inter-assay variability of LOP and DMLOP ranged from 2.1 to 14.5% for the low, medium and high quality control samples. The method has been used successfully to study loperamide pharmacokinetics in adult humans.     

Simultaneous determination of dimethylamphetamine and its metabolites in rat hair by gas chromatography–mass spectrometry

In order to study the disposition of dimethylamphetamine (DMAP) and its metabolites, DMAP N-oxide, methamphetamine (MA) and amphetamine (AP), from plasma to hair in rats, a simultaneous determination method for these compounds in biological samples using gas chromatography–mass spectrometry with selected ion monitoring (GC–MS-SIM) was developed. As DMAP N-oxide partially degrades to DMAP and MA during GC–MS analysis, it was necessary to avoid conditions which co-extract the N-oxide in the sample preparation so as to assure no contribution of artifactual products from DMAP N-oxide in the detection of the other compounds. For confirmation of the satisfactory separation of DMAP N-oxide from the others, the internal standards used for quantification were labeled with different numbers of deuterium atoms. Determination of unchanged DMAP was performed without any derivatization, that of DMAP N-oxide was carried out after conversion into trifluoroacetyl-MA by reaction with trifluoroacetic anhydride, and MA and AP were quantified after trifluoroacetyl-derivatization.After intraperitoneal administration of DMAP HCl to pigmented hairy rats (5 mg kg⁻¹ day⁻¹, 10 days, n=3), concentrations of DMAP and its metabolites in urine, plasma and hair were measured by GC–MS-SIM. The area under the concentration versus time curves (AUCs) of DMAP, DMAP N-oxide, MA and AP in the plasma were 397.2±97.5, 279.7±68.3, 18.4±1.2 and 15.9±2.2 μg min ml⁻¹, while their concentrations in the hair newly grown for 4 weeks after administration were 4.82±0.67. 0.45±0.09, 3.25±0.36 and 0.89±0.05 ng mg⁻¹, respectively. This fact suggested that the incorporation tendency of DMAP N-oxide from plasma into hair was distinctly low in comparison with the other compounds.     

Gas chromatographic–mass spectrometric determination of α-ketoisocaproic acid and [H7]α-ketoisocaproic acid in plasma after derivatization with N-phenyl-1,2-phenylenediamine

A method for determination of α-ketoisocaproic acid (KIC) and [4,5,5,5,6,6,6-²H₇]α-ketoisocaproic acid ([²H₇]KIC) in rat plasma was developed using gas chromatography–mass spectrometry-selected ion monitoring (GC–MS-SIM). [5,5,5-²H₃]α-Ketoisocaproic acid ([²H₃]KIC) was used as an analytical internal standard to account for losses associated with the extraction, derivatization and chromatography. The keto acids were extracted by cation-exchange chromatography using BondElut SCX cartridge and derivatized with N-phenyl-1,2-phenylenediamine to form N-phenylquinoxalinone derivatives. Quantitation was performed by SIM of the respective molecular ions at m/z 278, 281 and 285 for the derivatives of KIC, [²H₃]KIC and [²H₇]KIC on the electron impact method. The limit of detection was found to be 70 fmol per injection (S/N=3) and the limit of quantitation for [²H₇]KIC was around 50 nM in rat plasma. Endogenous KIC concentrations in 50 μl of rat plasma were measured with relative intra- and inter-day precision of 4.0% and 3.3%, respectively. The intra- and inter-day precision for [²H₇]KIC spiked to rat plasma in the range of 0.1 to 10 μM gave good reproducibility with relative standard deviation (RSD) of 6.5% and 5.4%, respectively. The intra- and inter-day relative errors (RE) for [²H₇]KIC were less than 6.4% and 3.8%, respectively. The method was applied to determine the plasma concentration of [²H₇]KIC after an intravenous administration of [²H₇]KIC in rat.     

Systematic analysis of acid, neutral and basic drugs in horse plasma by combination of solid-phase extraction, non-aqueous partitioning and gas chromatography–mass spectrometry

A sample preparation method for mass chromatographic detection of doping drugs from horse plasma is described. Bond Elut Certify (1 g/6 ml) is used for the extraction of 4 ml of horse plasma. Fractionation is performed with 6 ml of CHCl₃–Me₂CO (8:2) and 5 ml of 1% TEA–MeOH according to its property. Simple and effective clean-up based on non-aqueous partitioning is adopted to remove co-eluted contaminants in both acid and basic fractions. Two kinds of 1-(N,N-diisopropylamino)-n-alkanes are co-injected with the sample into the GC–MS system for the calculation of the retention index. Total recoveries of 107 drugs are examined. Some data of post administration plasma are presented. This procedure achieves sufficient recoveries and clean extracts for GC–MS analysis. The method is able to detect ng/ml drug levels in horse plasma.     

Determination of metabolites of pyrethroids in human urine using solid-phase extraction and gas chromatography–mass spectrometry

The described method permits the determination of the five most important metabolites of the pyrethroids permethrin, cypermethrin, deltamethrin, λ-cyhalothrin, fenvalerate, phenothrin and β-cyfluthrin in human urine in one run. The major urinary metabolites of these substances are cis-3-(2,2-dichlorovinyl)-2,2-dimethylcyclopropane-1-carboxylic acid (cis-Cl₂CA), trans-3-(2,2-dichlorovinyl)-2,2-dimethylcyclopropane-1-carboxylic acid (trans-Cl₂CA), cis-3-(2,2-dibromovinyl)-2,2-dimethylcyclopropane-1-carboxylic acid (Br₂CA), fluoro-3-phenoxybenzoic acid (F-PBA) and 3-phenoxybenzoic acid (3-PBA). After acidic hydrolysis to release the conjugated carboxylic acid metabolites, the analytes were separated from the matrix by means of solid-phase extraction using a reversed-phase column. The components of the eluate were converted to their methyl esters and extracted in hexane. Separation and quantitative analysis of the pyrethroid metabolites was carried out by capillary gas chromatography and mass selective detection. 2-Phenoxybenzoic acid served as an internal standard. The detection limits lay between 0.3 and 0.5 μg per litre urine. The relative standard deviations of the within-series imprecision were between 1% and 6%. The relative recovery rates ranged between 90% and 98%. Using this method we determined the elimination of pyrethroid metabolites in 24-h urine samples from eight pest controllers after indoor application of permethrin. The detected concentrations ranged from 1 to 70 μg g⁻¹ creatinine.     

Determination of extracellular and intracellular enrichments of [1-C]-α-ketoisovalerate using enzymatically converted [1-C]-valine standardization curves and gas chromatography–mass spectrometry

We report a validated method for the determination of extra- and intracellular [1-¹³C]-α-ketoisovalerate ([1-¹³C]-KIV) enrichments by gas chromatography–mass spectrometry. Standardization curves were prepared by enzymatic oxidation of [1-¹³C]-valine enriched standards of known composition. Slopes of [1-¹³C]-valine standardization curves (mean±SD: 0.99±0.02, n=5) and [1-¹³C]-KIV standardization curves (mean±SD: 0.98±0.01, n=7) were not significantly different. The method was applied for the determination of [1-¹³C]-KIV enrichments in plasma and tissues during [1-¹³C]-valine infusion in a piglet. [1-¹³C]-KIV enrichment could be determined±0.1 MPE (C.V. 1%), and extracellular [1-¹³C]-KIV enrichment was a reliable estimate of intracellular (skeletal muscle, bone growth plate) [1-¹³C]-KIV enrichment.     

Peanut (Arachis hypogaea) lectin recognizes α-linked galactose, but not N-acetyl lactosamine in N-linked oligosaccharide terminals

Peanut (Arachis hypogaea) agglutinin (PNA) is extensively used as tumour marker as it strongly recognises the cancer specific T antigen (Galβ1→3GalNAc-), but not its sialylated version. However, an additional specificity towards Galβ1→4GlcNAc (LacNAc), which is not tumour specific, had been attributed to PNA. For correct interpretation of lectin histochemical results we examined PNA sugar specificity using naturally occurring or semi-synthetic glycoproteins, matrix-immobilised galactosides and lectin-binding tissue glycoproteins, rather than mono- or disaccharides as ligands. Dot-blots, transfer blots or polystyrene plate coatings of the soluble glycoconjugates were probed with horse-radish peroxidase (HRP) conjugates of PNA and other lectins of known specificity. Modifications of PNA-binding glycoproteins, including selective removal of O-linked oligosaccharides and treatment with glycosidases revealed that Galβ1→4GlcNAc (LacNAc) was ineffective while terminal α-linked galactose (TAG) as well as exposed T antigen (Galβ1→3 GalNAc-) was excellent as sugar moiety in glycoproteins for their recognition by PNA. When immobilised, melibiose was superior to lactose in PNA binding. Results were confirmed using TAG-specific human serum anti-α-galactoside antibody.     

Synthesis of N-monomethylagmatine, a decarboxylation product of N-monomethyl- -arginine

N^G-Monomethylagmatine, a decarboxylation product of N^G-monomethyl- -arginine, has been synthesized by reacting putrescine with N,S-dimethylthiopseudouronium iodide. The structural identity of the product was confirmed by proton NMR and mass spectroscopy, and its properties were determined on thin-layer and electrophoretic chromatography.     

DNA sequence dependence of guanine-O alkylation by the N-nitroso carcinogens N-methyl- and N-ethyl-N-nitrosourea

After intracellular in vitro exposure to the mutagenic and carcinogenic N-nitroso compounds N-methyl-N-nitrosourea (MeNU) or N-ethyl-N-nitrosourea (EtNU), respectively, the average relative amounts of the premutational lesion O⁶-alkylguanine represent about 6% and 8% of all alkylation products formed in genomic DNA. At the level of individual DNA molecules gunine-O⁶ alkylation does nor occur at random; rather, the probability of a substitution reaction at the nucleophilic O⁶ atom is influenced by nucleotide sequence, DNA conformation, and chromatin structure. In the present study, 5 different double-stranded polydeoxynucleotides and 15 double-stranded oligodeoxynucleotides (24-mers) were reacted with MeNU or EtNU in vitro under standardized conditions. Using a competitive radioimmunoassay in conjunction with an anti-(O⁶-2′-deoxyguanosine) monoclonal antibody, the frequency of guanine-O⁶ alkylation was found to be strongly dependent on the nature of the nucleotides flanking guanine on the 5t́ and 3′ sides. Thus, a 5′ neighboring guanine, followed by 5t́ adenine and 5′ cytosine, provided an up to 10-fold more ‘permissive’ condition for O⁶-alkylation of the central guanine than a 5′ thymine (with a 5-methylcytocine in the 5′ position being only slightly less inhibitory). Thymine and cytosine were more ‘permissive’ when placed 3′ in comparison with their affects in the 5′ flanking position.     

Enzymatic synthesis of N-acetylglucosaminobioses by reverse hydrolysis: characterisation and application of the library of fungal β-N-acetylhexosaminidases

The regioselectivity of 20 extracellular β-N-acetylhexosaminidases of fungal origin was screened in the reverse hydrolysis with 2-acetamido-2-deoxy- -glucopyranose. Most of the enzymes used yielded 2-acetamido-2-deoxy-β- -glucopyranosyl-(1→4)-2-acetamido-2-deoxy- -glucopyranose (3) and 2-acetamido-2-deoxy-β- -glucopyranosyl-(1→6)-2-acetamido-2-deoxy- -glucopyranose (4). So far unknown product of enzymatic condensation, 2-acetamido-2-deoxy-β- -glucopyranosyl-(1→3)-2-acetamido-2-deoxy- -glucopyranose (2) was synthesised using the β-N-acetylhexosaminidases from Penicillium funiculosum CCF 1994, P. funiculosum CCF 2325 and Aspergillus tamarii CCF 1665. Addition of salts ((NH₄)₂SO₄ or MgSO₄ (0.1–1.0 M)) to the reaction increased the yields and also enhanced the β-N-acetylhexosaminidase regioselectivity.     

Analysis of urinary N-acetyl-S-(N-methylcarbamoyl)cysteine, the mercapturic acid derived from N,N-dimethylformamide

Human biotransformation of the industrial solvent N,N-dimethylformamide gives raise to N-acetyl-S-(N-methylcarbamoyl)cysteine (AMCC) which has the longest half-life (about 23 h) among urinary metabolites of N,N-dimethylformamide. It could be used for monitoring industrial exposure over several workdays, by measuring it in urine samples collected at the end of the working week. This is consistent with the suggestions of the American Conference of Governmental Industrial Hygienists, which established a limit of 40 mg/l for the year 2000. An easy, cheap and user-friendly method has been developed for determination of urinary AMCC. Unlike currently available methods, it requires neither a time-consuming preparation phase nor gas chromatographic analysis with a nitrogen-phosphorus or mass detector. The method uses high-performance liquid chromatography (HPLC), with an UV detector at 436 nm. A 10-μl volume of urine is added to a carbonate–hydrogen carbonate buffer and mixed with a dabsyl chloride solution in acetonitrile. The reaction between AMCC and the reagent is performed at 70°C for 10 min. The ‘dabsylated’ product is stable for at least 12 h. After brief centrifugation, the solution is ready for HPLC analysis using a C₁₈ column (250×4.6 mm, 5 μm). The method is sensitive (detection limit 1.8 mg/l) and specific. It identified urinary AMCC in urine of 40 subjects not exposed to N,N-dimethylformamide with a median concentration of 3.9 mg/l. In urine samples from 20 workers exposed to N,N-dimethylformamide (5–40.8 mg/m³), AMCC concentrations ranged from 16 to 170 mg/l. Industrial toxicology laboratories with limited instrumentation will be able to use it in the biological monitoring of workers exposed to N,N-dimethylformamide.     

Identification of low molecular weight aromatic compounds by gas chromatography–mass spectrometry (GC–MS) from kraft lignin degradation by three Bacillus sp.

Three bacterial strains identified as Paenibacillus sp., Aneurinibacillus aneurinilyticus and Bacillus sp. have been shown to decolourise kraft lignin in 6 days of incubation. The release of low molecular weight aromatic compounds by these bacterial strains during degradation of kraft lignin was analysed by GC–MS analysis. The total ion chromatograph (TIC) of ethyl acetate extract from kraft lignin sample inoculated by Paenibacillus sp. was similar to control except some minor changes in the chromatographic profile indicating incapability of this bacterium to modify kraft lignin. On the other hand the TIC of ethyl acetate extract from kraft lignin inoculated by A. aneurinilyticus and Bacillus sp. caused formation of several aromatic lignin-related compound that were not present in the extract of control. The compounds identified in extract of the sample degraded by A. aneurinilyticus were guaiacol, acetoguiacone, gallic acid and ferulic acid while t-cinnamic acid, 3,4,5 trimethoxy benzaldehyde, and ferulic acid by Bacillus sp. indicating oxidization of coniferylic (G units) and sinapylic (S units) alcohol of lignin polymer. Differences between the identified compounds from different bacterial treatment were strain-specific. Among the identified aromatic compounds, ferulic acid and 3,4,5-trimethoxy benzaldehyde could be useful to the industry of preservatives, aromas and perfumes.     

Some novel N-fatty acyl derivatives of a microbial galactosaminan

Novel seven N-fatty acyl derivatives (degree of substitution 0.78–0.96) of a microbial galactosaminan were prepared in 59–79% yields by its reaction with fatty acid anhydrides in aqueous acetic acid-methanol. N-Acetyl and N-propionyl derivatives were soluble in water, aqueous 2% sodium hydroxide, and aqueous 2% acetic acid, but N-higher fatty acyl (>C6) derivatives were insoluble. Gel was not formed in these react ions     

Determination of LSD and N-demethyl-LSD in urine by liquid chromatography coupled to electrospray ionization mass spectrometry

A sensitive and highly specific method for the determination of LSD and N-demethyl-LSD in urine, using combined liquid chromatography and mass spectrometry (LC-MS) with electrospray ionization, has been developed. Extrelut-3 extraction cartridges were used for a basic sample clean-up. Elution was obtained by toluene-diethyl ether (60:40, v/v). A Nucleosil C₁₈ (150×1 mm I.D.) reversed-phase column was used for the chromatographic separation, together with a mixture of 2 mM ammonium formate buffer (pH 3) and acetonitrile (70:30, v/v) as mobile phase. Recoveries were 93 and 80%, detection limits 0.025 and 0.035 ng/ml for LSD and N-demethyl-LSD, respectively. Intra-assay precision, studied at four concentrations, was better than 9% at the ng/ml range and better than 14% at 0.10 ng/ml for both compounds. Limits of quantitation were 0.05 and 0.10 ng/ml for LSD and N-demethyl-LSD, respectively. Reproducibility was good and linearity excellent for LSD in the range from 0.05 to 20 ng/ml (r>0.9999, N=7).     

Determination of 2-hydroxypropyl-γ-cyclodextrin in plasma of cynomolgus monkeys after oral administration by gas chromatography–mass spectrometry

This report describes a specific and highly sensitive gas chromatography–mass spectrometry (GC–MS) assay for the analysis of industrially produced 2-hydroxypropyl-γ-cyclodextrin, a heterogeneous mixture of homologues and isomers, in plasma of cynomolgus monkeys. Instead of analyzing the polysaccharide mixture as a whole, in a first step the HP-γ-cyclodextrin mixture, together with the internal standard (2,6-di-O-methyl-β-cyclodextrin), was deuteromethylated, and in a second step hydrolyzed with hydrochloric acid to the respective monosaccharides. The resulting reaction mixture was trimethylsilylated to 1,4-bis(O-trimethylsilyl)-2,3-bis-O-deuteromethyl-6-O-2′-deuteromethoxypropylglucose (representative for HP-γ-CD) and 1,4-bis-(O-trimethylsilyl)-bis-2,6-O-methyl-3-O-deuteromethylglucose (representative for the internal standard), respectively, and analyzed by GC–MS. The limit of quantification of this assay was 20 nmol/l.     

Identification and quantification of sophorolipid analogs using ultra-fast liquid chromatography–mass spectrometry

An ultra-fast liquid chromatographic method combined with atmospheric pressure chemical ionization mass detection (UHPLC/APCI-MS) has been developed for the separation and quantification of sophorolipid analogs produced by the yeast Candida bombicola. The sophorolipid mixture was produced by growing the yeast in the presence of glucose and oleic acid under higher aeration. It was found that more than 95% of the analogs are lactonic sophorolipids and all the produced sophorolipids produced were either mono- or di-acetylated. Also observed was a sophorolipid analog with a tri-unsaturated fatty acid, which has not been reported previously.     

A three-phase liquid chromatographic method for δC analysis of amino acids from biological protein hydrolysates using liquid chromatography–isotope ratio mass spectrometry

We report a three-phase chromatographic method for the separation and analysis of δ¹³C values of underivatized amino acids from biological proteins (keratin, collagen, and casein) using liquid chromatography–isotope ratio mass spectrometry (LC–IRMS). Both precision and accuracy of δ¹³C values for standard amino acid mixtures over the range of approximately 8 to 1320 ng of carbon per amino acid on the column were assessed. The precision of δ¹³C values of amino acids was found to be better at higher concentrations, whereas accuracy improved at lower concentrations. The optimal performance for this method was achieved with between 80 and 660 ng of carbon of each amino acid on the column. At amino acid amounts lower than 20 ng of carbon on the column, precision and accuracy may become compromised. The application of this new three-phase chromatographic technique will allow the analysis of δ¹³C of amino acids to be carried out as a routine method and benefit fields of research such as biomedicine, forensics, ecology, nutrition, and palaeodiet reconstruction in archaeology.