Mutagenesis, Vol. 15, No. 1, 45-55,
January 2000
© 2000 UK Environmental Mutagen Society/Oxford University Press
Carboxylesterases, a key factor in evaluating potential genotoxicity of Trinem antibiotics
Genetic Toxicology Department, Pre-Clinical Safety Sciences,Medicine Safety Evaluation Division, Glaxo Wellcome Research and Development, Park Road, Ware, Herts SG12 0DP, UK and 1 Tossicologia Genetica, Glaxo Wellcome SpA, Via Alessandro Fleming 2, 37135 Verona, Italy
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Sanfetrinem cilexetil, a hexetil ester of a Trinem antibiotic, does not induce micronuclei in rat bone marrow cells or induce DNA repair synthesis in rat hepatocytes following oral dosing. However, in vitro chromosome damage and mutations are induced in mammalian cells lacking carboxylesterase activity (human lymphocytes and mouse lymphoma L5178Y cells). In cells possessing carboxylesterase activity (CHL cells), chromosome damage induced by Sanfetrinem cilexetil is not observed. Similarly, if induced rat liver preparations or non-induced preparations from rat or human intestinal cells are present during exposure, genotoxic activity is lost, even in those cells lacking carboxylesterase enzymes. Thus the lack of demonstrable genotoxicity in vivo, in the assays used, is likely to be due to hydrolysis of the parent molecule by non-specific carboxylesterases present within the intestinal epithelium. In turn this data indicates that a genotoxic hazard to humans under therapeutic conditions is unlikely.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Sanfetrinem cilexetil is a hexetil ester of a Trinem antibiotic originally under development for the treatment of a variety of bacterial infections, including those caused by penicillin-resistant strains.
In vitro it was found that the ester undergoes rapid hydrolysis in aqueous solution to a number of products, including the parent compound (Trinem), acetaldehyde and cyclohexanol. In the in vitro and in vivo environments, non-specific carboxylesterases present either in the exogenous metabolic activation systems or the intestinal epithelium, respectively, are also able to cause rapid hydrolysis of this ester (see Figure 1
). Mammalian carboxylesterases represent a multigene family, the products of which are localized in the endoplasmic reticulum of many tissues, e.g. liver, intestine and lung. Aldridge's definition of A-, B- and C-esterases depending on their interaction with organophosphorous compounds is still widely used (Aldridge, 1953). A-esterases metabolize these compounds, B-esterases are inhibited by them and C-esterases are unaffected by them (Lund-Pero et al., 1994). According to this classification the serine superfamily of esterases (acetylcholinesterase, butyrlcholinesterase and carboxylesterase) falls into the B-esterase group. It is becoming increasingly clear that esterases have a broad and overlapping substrate specificity towards amides and esters and a single esterolytic reaction is frequently mediated by several enzymes (Satoh and Hosokawa, 1998). Carboxylesterases are responsible for the hydrolysis of many exogenous compounds, the consequences of which include both inactivation of drugs and activation of pro-drugs. For example, human liver and plasma carboxylesterase activates lovastatin and conversion of a pro-drug of prostaglandin F2
; other substrates include salicylates, cocaine and steroids (Satoh and Hosokawa, 1998). Therefore, these enzymes are major determinants of the pharmacokinetic behaviour of most therapeutic agents containing ester or amide bonds.
|
During preclinical safety evaluation, a comprehensive battery of in vitro and in vivo genotoxicity tests were performed on Sanfetrinem cilexetil. The in vitro systems comprised the Ames test, the liquid preincubation (Yahagi) Ames assay, cytogenetic analysis of human peripheral lymphocytes (HPLA) and Chinese hamster lung (CHL) cells and the L5178Y mouse lymphoma tk locus assay (MLA). All of the tests were performed in the absence and presence of a rat liver exogenous metabolic activation system (S9 mix). Additional MLA and HPLA were performed in the presence of either non-induced rat intestinal S9 or human intestinal S9 or microsomal preparations.
The in vivo mutagenic potential of Sanfetrinem cilexetil was assessed using a rat bone marrow micronucleus test (MNT) and a rat liver unscheduled DNA synthesis (UDS) assay.
The parent Trinem antibiotic, from which the hexetil ester had been derived, had been tested previously in the standard battery of genotoxicity tests, including the Ames test, HPLA, MLA and rat MNT, and yielded uniformly negative results (data not shown).
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Materials
Sanfetrinem cilexetil was supplied by Direzione Richerche di Tecnica Farmaceutica (Glaxo SpA, Verona, Italy) (purity 96.7%). For the in vitro studies, Sanfetrinem cilexetil was dissolved in dimethyl formamide supplied by Aldrich Chemical Co. (UK) or ethanol supplied by Eurobase SpA (Italy) and for the in vivo studies 0.5% w/w K15 Methocel was used as the vehicle. Acetaldehyde was dissolved in sterile water for irrigation, supplied by Fresenius Health Care Group (UK) and cyclohexanol was dissolved in dimethyl sulphoxide (DMSO) supplied by Romil Chemicals Ltd (UK).
Determination of carboxylesterase activity
Carboxylesterase activity was determined following the method described by Hojring and Svensmark (1976). Briefly, determination was performed spectrophotometrically at 30°C with four substrates (-naphthyl acetate, -naphthyl butyrate, p-nitrophenol acetate and p-nitrophenol butyrate). Eserine was added to the reaction mixture to inhibit cholinesterases.
Samples (0.2 ml) were added to a prewarmed mixture of 1.5 ml substrate solution and 1.5 ml 0.2 M Tris–HCl buffer, pH 7, in a cuvette thermostatically maintained at 30°C. E348 was recorded as a function of time. The enzyme activity was calculated as:
 |
E348, where E348 is the difference in extinction at 348 nm of 0.1 mM solutions of p-nitrophenol and p-nitrophenyl acetate (or butyrate) in 0.1 M Tris–HCl buffer, pH 7.
Preparation of the exogenous metabolizing systems
Rat liver preparation.
All the standard in vitro assays were performed in the absence and presence of a rat liver exogenous metabolizing activation system (S9 mix). The post-mitochondrial fraction (S9) was derived from livers of male Han Wistar rats, aged 6–9 weeks (supplied by the Rodent Breeding Unit, Glaxo Wellcome Research and Development Ltd, Ware, UK). The animals were pretreated with a combination of sodium phenobarbitone (administered i.p. at 40 mg/kg) and ß-naphthoflavone (administered i.p. at 100 mg/kg) for 3 consecutive days prior to killing. The S9 fraction was prepared essentially as described by Maron and Ames (1983) and stored for up to 6 months at –80°C. The S9 fraction was used in conjunction with an NADPH-generating system, the composition of which has been described previously (Gatehouse and Delow, 1979). The amount of S9 fraction used varied within each assay. For the Salmonella/plate incorporation assay, a final concentration of 40 µl S9 fraction (equivalent to ~1.2 mg S9 protein)/plate was used, whilst in the MLA and HPLA final concentrations of 1 and 2% v/v were used, respectively.
Rat intestinal preparation. An intestinal post-mitochondrial fraction, obtained from the small intestine of Wistar rats, ~300 g in weight, was made by Laboratorio di Tossicologia Genetica (Glaxo Wellcome SpA), essentially as described by Chhabra et al. (1974). The intestinal preparations were stored for up to 6 months at –80°C. The S9 fraction was used in conjunction with an NADPH-generating system.
Human intestinal microsomal preparation. The small intestinal samples originated from a Caucasian female and were supplied by the International Institute for the Advancement of Medicine (Exton, PA). The intestinal preparation was stored for up to 6 months at –80°C. The fraction was used in conjunction with an NADPH-generating system (including the addition of glucose 6-phosphate dehydrogenase).
Microbial mutagenicity assays
Salmonella/plate incorporation assay.
The procedure was carried out essentially as described by Ames et al. (1975) and Maron and Ames (1983) using the test strains Salmonella typhimurium TA1535, TA1537, TA98 and TA100 and Escherichia coli strains WP2 (pKM101) and WP2 uvrA (pKM101). A maximum concentration of 40 µg/plate was used and five plates were prepared for each test concentration. After treatment the plates were incubated at 37°C for 72 h before being scored for histidine-independent colonies using an automatic colony counter. Statistical analysis was performed on untransformed data using Dunnett's t-test (Dunnett, 1955).
Liquid preincubation (Yahagi) assay. The procedure followed that of Yahagi et al. (1975). Sanfetrinem cilexetil was preincubated at 37°C for 30 min with either rat liver S9 mix or phosphate buffer and the test strain [S.typhimurium TA1535, TA1537, TA98 or TA100 or E.coli WP2 (pKM101) and WP2 uvrA (pKM101)], prior to addition of the overlay agar. Five replicate plates were prepared for each test concentration and these were incubated at 37°C for 72 h before being scored for histidine-independent colonies using an automatic colony counter.
In vitro cytogenetics analysis using human peripheral lymphocytes
Sanfetrinem cilexetil was evaluated for in vitro clastogenicity using human peripheral lymphocytes. Assays were carried out using three different exogenous metabolic activation systems, induced rat liver S9 mix (2% v/v), non-induced rat intestinal S9 mix (2% v/v) and non-induced human intestinal S9 mix (2% v/v).
The procedure was carried out according to the recommendations of the UKEMS Guidelines Committee (Scott et al., 1983). Duplicate whole blood cultures (phytohaemagglutinin-stimulated human peripheral blood, from healthy male donors) were prepared in Iscoves medium for each test concentration. Treatment took place at 48 h post-culture initiation and lasted for 24 h in the absence of S9 mix and for 3 h in the presence of S9 mix. The spindle poison colcemid (final concentration 0.4 µg/ml) was added 2 h prior to harvest, which took place 72 h after culture initiation. After metaphase preparation, the chromosomes were stained with Giemsa; a maximum of 400 metaphase cells for solvent and untreated control cultures and a maximum of 200 metaphase cells for each treatment group and positive control were analysed for chromosome damage. All the slides were coded and scored blind in random order.
Aberrations were classified according to Savage (1976), into chromosome- and chromatid-type damage, with further subdivision into deletions and exchanges. Displaced and undisplaced fragments separated by a non-staining region equal to or greater than the width of the chromatid were scored as deletions. Non-staining regions of less than the chromatid width were scored as gaps. Prior to chromosomal analysis, the mitotic indices (MI) for each culture were estimated based upon 1000 lymphocytes. The maximum test concentration of Sanfetrinem cilexetil was 100 µg/ml, which caused an ~50% reduction in MI.
Each of the treatment groups (including positive control groups) were compared separately to the solvent control group using a one-sided individual comparison test. Each 2x2 contingency table was analysed using Fisher's exact test (Armitage, 1971). This analysis was carried out twice; firstly, using the numbers of all aberrant cells, including those containing chromatid and chromosome gaps and, secondly, on the same data but excluding cells containing exclusively gaps. Treatment groups are significantly different from the solvent control group if P < 0.05.
In vitro cytogenetics analysis using CHL cells
The protocol using this cell line was very similar to that used for the human peripheral lymphocytes. A single assay using continuous treatment in the absence of S9 mix for 24 h was performed. One hundred well-spread metaphases were scored for the highest and intermediate concentrations and the solvent control. Only 50 metaphases were scored from the cultures treated with the positive control. All the slides were coded and scored blind in random order.
Aberrations were classified according to Savage (1976), into chromosome- and chromatid-type damage, with further subdivision into deletions and exchanges. Displaced and undisplaced fragments separated by a non-staining region equal to or greater than the width of the chromatid were scored as deletions. Non-staining regions of less than the chromatid width were scored as gaps. Prior to chromosomal analysis, the MI for each culture were estimated based upon 1000 lymphocytes. The maximum test concentration of Sanfetrinem cilexetil was 12.5 µg/ml, which caused an ~50% reduction in MI. Statistical analysis was performed using Fisher's exact test (Armitage, 1971), as described in the previous section.
In vitro mammalian cell mutagenicity assay using mouse lymphoma L5178Y cells
A number of assays were performed to assess the mutagenicity of Sanfetrinem cilexetil in the standard MLA, using the same range of exogenous metabolic activation systems as those in the HPLA, specifically induced rat liver S9 mix (1% v/v), non-induced rat intestinal S9 mix (2% v/v) and non-induced human microsomes (2% v/v). The final concentration for all these samples was equivalent to 1 mg protein/ml. Further assays were carried out to assess the mutagenic effect after degradation of Sanfetrinem cilexetil and of its breakdown products acetaldehyde and cyclohexanol.
In the initial MLA, a range of concentrations were assessed. On the basis of these assays, two concentrations were selected, at which mutant induction had occurred in the absence of rat liver S9 mix. These two concentrations were then used for the assays using non-induced rat intestinal S9 mix and non-induced human microsomes.
The same standard protocol was employed for all of the assays (Cole et al., 1983). Duplicate cultures of mouse lymphoma L5178Y Tk+/– 3.7.2C cells in the exponential growth phase were treated at a cell density of 1.67x106 cells/ml in treatment medium [RPMI 1640 Glutamax 1, containing 3.0 mM L-glutamine, 25 mM HEPES, supplemented with 200 µg/ml sodium pyruvate, 50 µg/ml streptomycin sulphate, 50 IU/ml benzylpenicillin and 3% v/v heat-inactivated donor horse serum (heat inactivated at 56°C for 30 min)]. For the standard assays in the presence of an exogenous metabolic activation system the final concentration of rat liver S9 was 1% v/v and in the investigative assays 2% v/v non-induced rat or human intestinal preparations were used. Treatment was carried out for 3 h at 37°C in a shaking incubator.
For each treatment group, the cell titre was determined post-treatment. To determine post-treatment cloning efficiency cells were plated out into 96-well microtitre plates at a density of 1.6 cells/well (2 plates/treatment group) in RPMI 1640 containing 10% v/v heat-inactivated donor horse serum (CM10). The remaining cell populations were resuspended in CM10 at a density of 2x105 cells/ml. Where necessary, cultures were diluted back to 2x105 cells/ml ~24 h later.
Following the 48 h expression period each culture was plated in CM10 (at 1.6 cells/well) into 96-well microtitre plates (2 plates/treatment group) to determine post-expression time cloning efficiency (viability) and plated into 96-well microtitre plates in CM10 containing 4 µg/ml TFT to select for mutant colonies (2 plates/treatment group, each well containing 2x103 cells).
After 6–8 (cloning efficiency plates) or 11 (mutant plates) days, the colonies were stained with 2.5 mg/ml 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide in phosphate-buffered saline. The plates were scored for the number of colony-bearing wells in each microtitre plate using a Titertek mirror box. Only colonies of >50 cells were scored. On the mutant plates, colonies were sized by eye and recorded accordingly (i.e. small or large). Colonies that were <
of the well diameter were scored as small. Conversely, those > of the well diameter were scored as large.
All the calculations are standard (Clive and Spector, 1975), with the exception of the survival calculation, which takes into account any cell loss during the treatment period by reduction in post-treatment cell count (cell count factor) and the cloning efficiency (CE) for each treated culture.
 |
 |
 |
The statistical analysis procedure used was that for mammalian fluctuation tests as recommended by the UKEMS Sub-committee on Guidelines for Mutagenicity Testing (Robinson et al., 1989).
In vivo MNT in rat bone marrow (oral administration)
Solutions of Sanfetrinem cilexetil were administered to groups of 10 male and groups of 10 female Charles River Wistar rats by oral gavage at doses of 500, 1000 and 2000 mg/kg. Vehicle control groups received K15 Methocel at an oral dose of 0.5% w/w and positive control groups received an oral dose of cyclophosphamide at 15 mg/kg. Each treatment was given to two groups of animals: one group was killed by cervical dislocation 24 h post-dosing and the second group at 48 h post-dosing. Bone marrow smears were prepared, stained and 2000 immature (polychromatic) erythrocytes were examined from each animal for the presence of micronuclei. The proportion of polychromatic cells within the total erythrocyte population after each treatment was also determined by counting the number of polychromatic cells observed in fields containing a total of 500 erythrocytes. This was used as an indicator of cytotoxicity (Richold et al., 1990). Statistical analysis was performed using the likelihood ratio test (Amphlett and Delow, 1984).
In vivo rat liver UDS assay
Selection of the dose levels of Sanfetrinem cilexetil was based on the results of an acute oral toxicity study (data not shown). The top dose level of this study was the maximum dose of 2000 mg/kg body wt.
The assay was split into two experiments (experiments A and B), mainly because the practicalities of the assay limit the number of animals which can be processed at any one time. The second experiment serves as a check on inter-animal variation as it is generally accepted that data compiled over two or more experiments can be pooled at the end of the study to give an overall view (Butterworth et al., 1987).
Groups of 12 male Han Wistar rats received oral doses of 1500 or 2000 mg/kg. Approximately 2 or 12 h later (12 h only for the positive control, 2-acetylaminofluorine at an oral dose of 7.5 mg/kg) the rats were killed and their hepatocytes isolated by a two-stage perfusion (Kennelly et al., 1993).
Three slides were prepared for each animal and, where possible, 100 morphologically normal cells were scored in blind fashion over at least two of the slides using an image analysis system. Slides were scored for four or five rats from each of the treatment, negative and positive control groups ~2 and 12 h after dosing. All slides were selected prior to coding and selection was based solely on slide quality.
For each slide the number of cells scored was recorded and for each cell the nuclear area, nuclear grain count and cytoplasmic grain count were measured. Using this data the net grain (nuclear grain count – cytoplasmic grain count, N – C) and percentage of cells in repair (net grain 
5) were calculated. The mean and standard deviation of the nuclear area, nuclear and cytoplasmic grain counts and net grain were then calculated for each animal and the mean and standard error calculated for each group.
The criteria for a positive response are given in the UKEMS guidelines (Kennelly et al., 1993). Briefly, an animal or group is considered positive (in repair) when the mean net grain is 0 and the percentage of cells in repair is 20.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Negative genotoxicity data (bacterial and in vivo data)(Tables I–V
)Bacterial mutation assays. In the standard Ames and the liquid preincubation (Yahagi) Ames assays, negative results were observed up to 40 µg Sanfetrinem cilexetil/plate in the absence of induced rat liver S9 mix and up to 15 µg Sanfetrinem cilexetil/plate in the presence of induced rat liver S9 mix for strains TA1535, TA1537, TA98, TA100 and WP2uvrA (pKM101). Non-reproducible increases of <2-fold were obtained at isolated concentrations in strains TA1357 and TA98 in the absence of S9 mix. In the Ames test with WP2 (pKM101), small statistically significant reproducible increases were observed in the presence of rat liver S9 mix at the highest two non-toxic concentrations. However, these values fall within the historical range (35–160 revertant colonies/plate) and so are not considered to be biologically significant.
|
|
|
|
|
In vivo genotoxicity assays. Negative results were obtained in the in vivo rat bone marrow MNT. Following oral administration of Sanfetrinem cilexetil at 500, 1000 or 2000 mg/kg, there was no significant increase in micronuclei at either time point (24 and 48 h post-dosing). In addition, no biologically significant dose-related change in erythroblast proliferation rate was apparent. The positive control, cyclophosphamide at 15 mg/kg, induced a clear positive effect, thus validating the test (Table VI
).
|
Negative results were also obtained in the in vivo rat liver UDS assay following oral administration of Sanfetrinem cilexetil at 1500 or 2000 mg/kg. UDS was assessed in hepatocytes isolated 2–4 or 12–14 h after dosing. There was no evidence of increased UDS in the treated animals at either time point. The positive control, 2-acetylaminofluorine (7.5 mg/kg) induced a clear positive effect, thus validating the assay (Table VII
).
|
In vitro cytogenetics analysis using human peripheral lymphocytes (Table VIII
)In the absence of induced rat liver S9 mix, Sanfetrinem cilexetil induced a reproducible, dose-related statistically significant increase in aberration frequency (including or excluding gaps) compared with the solvent control, at test concentrations of 25 and 100 µg/ml. At these concentrations the mitotic indices were reduced by 15–28 and ~50%, respectively. Untreated and solvent controls exhibited low spontaneous aberration frequencies and the positive control induced a high level of damage.
|
In the presence of induced rat liver S9 mix (2% v/v), no increases in aberration frequency were observed at the test concentrations selected for metaphase analysis (6.25, 25 and 100 µg/ml). Furthermore, there was only a slight reduction in MI compared with the reduction in the absence of induced rat liver S9 mix.
In vitro cytogenetics analysis using CHL cells (Table IX)
In the cytogenetics assay using CHL cells following treatment for 24 h in the absence of S9 mix only, Sanfetrinem cilexetil was analysed at 6.25 and 12.5 µg/ml. At these concentrations the MI was reduced by 39 and 50%, respectively. No clastogenic activity was observed at either test concentration. The positive control, mitomycin C at 0.05 µg/ml, produced a clear positive response, thus validating the assay.
|
In vitro mammalian cell mutagenicity assay using mouse lymphoma L5178Y Tk+/– 3.7.2C cells (Tables X and XI
)In the cytotoxicity assays a wide concentration range of Sanfetrinem cilexetil was assessed: cells survived concentrations of 100 µg/ml producing relative survival values of 3 and 31% in the absence and presence of induced rat liver S9 mix (1% v/v), respectively (data not shown). Six concentrations were selected for assessment, in the concentration ranges 10–60 and 50–175 µg/ml in the absence and presence of induced rat liver S9 mix, respectively.
|
|
In the first assay, mutant frequency was assessed at concentrations of Sanfetrinem cilexetil up to 60 and 150 µg/ml in the absence and presence of induced rat liver S9 mix, respectively. At the highest concentration tested the relative survival values were 13 and 32%, respectively. In the absence of induced rat liver S9 mix, dose-related, statistically significant increases in mutant frequency were observed at 40 and 60 µg/ml. At 60 µg/ml, the mutant frequency was ~5-fold higher than the solvent control. In the presence of induced rat liver S9 mix statistically significant increases in mutant frequency were also observed, but at higher concentrations
80 µg/ml. The maximum fold increase was 5.5 higher than the solvent control. These increases were confirmed in a second assay using a slightly modified concentration range. Further MLA and HPLA were undertaken to investigate the potential mutagenicity and clastogenicity of this compound in the presence of more relevant exogenous metabolic activation systems. The in vitro effects of Sanfetrinem cilexetil in the absence of induced rat liver S9 mix were considered to be of little biological relevance to the oral administration of Sanfetrinem cilexetil to humans, since the molecule is rapidly hydrolysed in vivo by non-specific carboxylesterases. As it is known that carboxylesterases reside in the intestine, the alternative exogenous metabolic activation systems employed were non-induced rat and human intestinal S9 or microsomal preparations.
In vitro cytogenetics analysis using human peripheral lymphocytes in the presence of non-induced rat and human intestinal preparations (Fig. 2)
The same concentrations of Sanfetrinem cilexetil were selected as previously tested, as acceptable cytotoxicity and significant increases in aberration frequency had been observed in the absence of induced rat liver S9 mix at 100 µg/ml. As anticipated, at 50 and 100 µg/ml in the absence of an exogenous metabolizing system, Sanfetrinem cilexetil markedly reduced MI by 40 and 50%, respectively, and significant increases in chromosome aberration frequency were observed at these two concentrations. In the presence of either exogenous metabolizing system (non-induced rat or human intestinal preparations) the ester was less cytotoxic to the cells and no statistically significant increase in chromosome aberration frequency occurred at any of the concentrations tested. These results indicate that the clastogenic effects observed are due to the intact ester, which is detoxified by rapid hydrolysis by the carboxylesterases present at 0.14 and 0.78 U/mg protein in the rat and human S9 fractions, respectively. This is considerably higher than the carboxylesterase activity (>30-fold) found in human peripheral lymphocytes (0.004 U/mg protein).
|
In vitro mammalian cell mutagenicity assay using mouse lymphoma L5178Y cells in the presence of non-induced rat and human intestinal preparations (Table XII
)Two concentrations of Sanfetrinem cilexetil, 64 and 80 µg/ml, were selected from the previous assays as these had reproducibly induced significant increases in mutant frequency at acceptable levels of cytotoxicity, in the absence of induced rat liver S9 mix. As anticipated, in the absence of an exogenous metabolizing activation system, at both concentrations Sanfetrinem cilexetil markedly reduced relative survival and significant increases in mutant frequency were observed.
|
However, at the same two concentrations in the presence of either the rat or human non-induced intestinal metabolizing system, relative survival was increased, i.e. the ester was less cytotoxic to the cells. In the presence of non-induced human microsomes (2% v/v), relative survival was reduced to 57% at 80 µg/ml and at the same concentration in the presence of non-induced rat intestinal S9 (2% v/v), relative survival was reduced to 76%. In the presence of either intestinal preparation, no statistically significant increase in mutant frequency occurred at the two concentrations tested. These results indicate that the mutagenic effects observed are due to the intact ester and that detoxification of the ester occurs by the action of the carboxylesterases found within the intestinal fractions.
Carboxylesterase activity
As illustrated in Table XIII, carboxylesterase levels were measured in the human peripheral lymphocyte, mouse lymphoma L5178Y and CHL cell systems and with all of the exogenous metabolizing activation systems used during the assessment of Sanfetrinem cilexetil. Negligible quantities were found in the human peripheral lymphocyte and mouse lymphocyte preparations, however, 4- to 5-fold more carboxylesterase activity was found in CHL cells. Higher levels of carboxylesterase activity were observed with the different metabolic activation systems, especially those derived from the rat. The rat liver and intestinal preparations contained between 90- and 200-fold higher levels of activity than that found in the human peripheral lymphocyte and mouse lymphoma preparations. Also, the activity was 35- to 56-fold higher in the human intestinal preparations than that found in the human peripheral lymphocyte and mouse lymphoma preparations.
|
Sanfetrinem degradation under in vitro test conditions
It has been demonstrated that the degradation of Sanfetrinem cilexetil is rapid under the test conditions of HPLA (Table XIV
). All the incubations contained 100 µg/ml Sanfetrinem cilexetil, which was the highest concentration tested for chromosome damage. In Iscoves medium alone, the concentration of Sanfetrinem cilexetil decreased by half after 30 min incubation at 37°C and after 3 h the level had fallen below the level of detection. A similar degradation pattern was also observed in an incubation mixture of 10% heat-inactivated human blood and phosphate buffer (0.1 M, pH 7.4) in the absence of induced rat liver S9 mix. The blood had been heat inactivated by heating to 56°C for 30 min to inactivate any esterases present in the donor blood. Under similar conditions but in the presence of induced rat liver S9 mix (2% v/v) or non-induced rat intestinal S9 mix, hydrolysis was much more efficient. The ester concentration was reduced to ~50% within 5 min, after 30 min to <20% and below the level of detection after 1 h incubation at 37°C. Hydrolysis of Sanfetrinem cilexetil was much more rapid in the presence of the induced liver fraction.
|
A similar effect was obtained when human intestinal preparations were assessed under the MLA test conditions. At the end of a 3 h incubation, only 3–4% Sanfetrinem cilexetil remained (data not shown).
Investigations into the potential mutagenicity of key degradation products using the MLA
As Sanfetrinem cilexetil also degrades in aqueous buffer under the conditions of the MLA, two further assays were performed in the absence of S9 mix to assess whether the increases in mutant frequency were due entirely to the effect of intact Sanfetrinem cilexetil or the result of a contribution from one or more degradation products. The standard protocol for the MLA was followed.
In the first, freshly prepared Sanfetrinem cilexetil at 64 and 80 µg/ml was tested for mutagenicity in comparison with the same concentrations of Sanfetrinem cilexetil that had been preincubated for 12 h at 37°C in treatment medium. The results show that freshly prepared Sanfetrinem cilexetil was cytotoxic, reducing relative survival to 18% at 80 µg/ml, and induced a 5-fold increase in mutant frequency. Following preincubation of the same concentrations of Sanfetrinem cilexetil, no significant reduction in relative survival and no increase in mutant frequency were observed, indicating that the products of degradation were not detectably mutagenic (Table XV).
|
Treatment medium containing the preincubated ester was analysed for the presence of Sanfetrinem cilexetil. After 12 h, intact ester could not be detected in the treatment medium initially containing 64 µg/ml and only 0.6% intact ester remained in the medium, which initially contained 80 µg/ml (data not shown).
In the second assay the mutagenic potential of the known hydrolysis products acetaldehyde and cyclohexanol was assessed, both alone and in combination, and the results compared with those obtained for Sanfetrinem cilexetil alone. Selection of the two concentrations of acetaldehyde (6.2 and 50 µg/ml) and cyclohexanol (15.5 and 2500 µg/ml) was based on data generated in earlier assays. The lower concentrations were based on earlier chemical analysis of the degradation of 64 µg/ml Sanfetrinem cilexetil within the test system (data not shown). The higher concentration significantly increased mutant frequency, under the same incubation conditions (data not shown). When the two were combined, the concentration ratio had been previously determined to represent the likely concentrations present after total degradation of Sanfetrinem cilexetil (at 64 µg/ml) within the test system (data not shown).
Mutagenic effects were observed at the highest concentrations of both acetaldehyde and cyclohexanol. Relative survival was not reduced by acetaldehyde at either concentration, but was reduced to 29% by cyclohexanol at 2500 µg/ml. Sanfetrinem cilexetil at 64 µg/ml reduced relative survival to 40% and induced a significant increase in mutant frequency (4-fold increase when compared with the solvent control). At 6.2 and 15.5 µg/ml, acetaldehyde and cyclohexanol were not mutagenic. Furthermore, when combined in the correct ratio no increase in mutant frequency was observed. Therefore it was not possible, under the test conditions, to reproduce the quantitative mutagenic effects induced by 64 µg/ml Sanfetrinem cilexetil alone using a combination of acetaldehyde and cyclohexanol at concentrations known to be present after total hydrolysis of this amount of drug. The validity of this assay may be questioned since acetaldehyde is a highly volatile and reactive molecule. It is probable that loss of this chemical through volatilization and non-specific protein binding within the test system could sequester acetaldehyde and therefore reduce exposure of cellular DNA. These results indicate that the mutagenic activity observed in this assay could be assigned in large part to the intact ester and not to the genotoxic hydrolysis products (Table XVI).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The intact ester Sanfetrinem cilexetil was devoid of in vitro mutagenic activity in a range of microbial tests and in vivo in the rat MNT and liver UDS assay. However, positive results were obtained in two in vitro mammalian assays in the absence of an exogenous metabolizing system. In the presence of an exogenous metabolizing system, a positive response was observed in the MLA, but higher concentrations were required compared with the results obtained after direct treatment with the drug. No significant effects were seen in the presence of an exogenous metabolizing system in the HPLA.
During the assessment of Sanfetrinem cilexetil, it has been shown that under the conditions of the in vitro mammalian assays the ester can undergo hydrolysis both in aqueous solution and enzymatically. It is also known that the hydrolysis products include the genotoxins acetaldehyde and cyclohexanol.
The positive results obtained in the in vitro mammalian assays were initially assumed to have arisen from the formation of these two genotoxins. Both of these products can induce chromosome damage in human lymphocytes (Collin, 1971; Böhlke et al., 1983). In human peripheral lymphocytes, sister chromatid exchanges were induced by ethanol only when alcohol dehydrogenase was present and the number was reduced in the additional presence of aldehyde dehydrogenase (Obe et al., 1986). These data are consistent with acetaldehyde being the genotoxic metabolite of ethanol. However, under the conditions of the in vitro mammalian assays, neither hydrolysis product is formed in sufficient concentration to exert any detectable genotoxic effects within these test systems. Consequently, it is likely that the observed genotoxicity is largely due to the intact ester.
Detection of genotoxic activity of the ester was dependent upon the enzymatic status of the mammalian indicator cells used within the in vitro assays and, in particular, on the absence or presence of non-specific carboxylesterases. In human lymphocytes and L5178Y cells, the levels of intracellular carboxylesterases were very low (~0.004 U/mg protein). Consequently, clastogenic and mutagenic effects were readily detected in these cells. However, when CHL cells (a rodent cell line containing 5-fold higher levels of carboxylesterase activity) were used as the target cell for measuring cytogenetic damage, the clastogenicity of Sanfetrinem cilexetil was not detectable. From the point of view of hazard identification, when evaluating the genotoxic potential of drugs within this particular class it is suggested that the CHL cell system provides a more accurate representation of the carboxylesterase activity likely to be present in vivo.
In a similar manner the addition of induced rat liver S9 (containing up to 200-fold higher levels of carboxylesterase activity) eliminated the clastogenic effects in human lymphocytes and considerably reduced the mutagenic activity in L5178Y cells. This was also achieved using non-induced rat or human intestinal preparations, which contain high levels of this class of enzyme.
In each case it is presumed that the genotoxic activity is lost due to rapid hydrolysis of the intact ester by non-specific carboxylesterases to the free acid and the breakdown products of the ester group (in particular acetaldehyde and cyclohexanol) to concentrations that were not detectable as clastogenic. As the intestinal epithelium is the primary tissue likely to be in contact with the intact ester following oral administration, the negative results obtained using intestinal preparations provide a good measure of reassurance on the safety of the ester in vivo. Any inherent genotoxicity associated with the intact ester would be rapidly eliminated by enzymatic hydrolysis. Rapid hydrolysis of the ester during transit through the gastrointestinal tract has been confirmed by the presence of <2% intact ester in the faeces of treated rats (data not shown).
The likely exposure to the hydrolysis products, acetaldehyde and cyclohexanol, following therapeutic administration to humans is negligible. In the case of acetaldehyde, the anticipated exposure would be similar to consuming a glass of wine. Carbonyl compounds are among the most volatile substances in alcoholic beverages. Acetaldehyde is the principal carbonyl compound in beer and has been found at similar ranges (0.1–16.4 mg/l) in US, German and Norwegian beers (IARC, 1988). Acetaldehyde constitutes >90% of the total aldehyde content of wines, occurring at 50–100 mg/l (Nykänen and Suomalainen, 1983).
Furthermore, the negative results obtained in vivo provide additional reassurance on the safety of Sanfetrinem cilexetil and its hydrolysis products, should any reach the liver or bone marrow.
It is concluded that the genotoxic activity observed within two in vitro mammalian assays is of no biological relevance to the intended therapeutic use of this antibiotic as an orally administered drug. Examination of intestinal tissue, using either the in vivo Comet assay or mouse transgenic assays, would provide further confirmation of this conclusion.
|
Notes |
|---|
2 To whom correspondence should be addressed. Tel: +44 1920 882497; Fax: +44 1920 882679; Email: jo3639{at}glaxowellcome.co.uk
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
-
Aldridge,W.N. (1953) Serum esterases I. Two types of esterases (A and B) hydrolysing p-nitrophenol acetate, proprionate and butyrate and a method for their determination. Biochem. J., 53, 110–117.[Web of Science][Medline]
Ames,B.N., McCann,J. and Yamasaki,E. (1975) Methods of detecting carcinogens and mutagens with Salmonella/mammalian microsomal mutagenicity test. Mutat. Res., 31, 347–354.[Web of Science][Medline]
Amphlett,G.E. and Delow,G.F. (1984) Statistical analysis of the micronucleus test. Mutat. Res., 128, 161–166.[Web of Science][Medline]
Armitage,P. (1971) Statistical Methods in Medical Research, 1st Edn. Stockwell Scientific, London, UK.
Böhlke,J.-U., Singh,S. and Goeddo,H.W. (1983) Cytogenetic effects of acetaldehyde in lymphocytes of Germans and Japanese. SCE, clastogenic activity and cell cycle delay. Hum. Genet., 63, 285–289.[Web of Science][Medline]
Butterworth,B.E., Ashby,J., Bermudez,E., Casciano,D., Mirsalis,J., Probst,G. and Williams,G.M. (1987) A protocol and guide for in vivo rat hepatocyte DNA repair assay. Mutat. Res., 189, 123–133.[Web of Science][Medline]
Chhabra,R.S., Pohl,R.J. and Fouts,J.R. (1974) A comparative study of xenobiotic-metabolising enzymes in liver and intestine of various animal species. Drug Metab. Dispos., 2, 443–447.[Abstract]
Clive,D. and Spector,J.F.S. (1975) Laboratory procedure for assessing specific locus mutations at the TK locus in cultured L5178Y mouse lymphoma cells. Mutat. Res., 31, 17–29.[Web of Science][Medline]
Cole,J., Arlett,C.F., Green,M.H.L., Lowe,J. and Muriel,W. (1983) A comparison of the agar cloning and microtitration techniques for assaying cell survival and mutation frequency in L5178Y mouse lymphoma cells. Mutat. Res., 111, 371–386.[Web of Science][Medline]
Collin,J.P. (1971) Effect cytogenetique du cyclamate de soude, de la cyclohexanone et du cyclohexanol. Le Diabete, 19, 215–221.[Medline]
Dunnett,C.W. (1955) A multiple comparison procedure for comparing several treatments with a control. J. Am. Statist. Assoc., 50, 1096–1121.[Web of Science]
Gatehouse,D.G. and Delow,G.F. (1979) The development of a `Microtitre' fluctuation test for the detection of indirect mutagens and its use in the evaluation of mixed-enzyme induction of the liver. Mutat. Res., 60, 239–252.[Web of Science][Medline]
Hojring,N. and Svensmark,O. (1976) Carboxylesterases with different substrate specificity in human brain extracts. J. Neurochem., 27, 523–528.[Web of Science]
IARC (1988) IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Alcohol Drinking. IARC Scientific Publications no. 44, IARC, Lyon.
Kennelly,J.C., Lefevre,P.A., Dean,S.W., Waters,R., Burlinson,B., de G Mitchell,I., Ashby,J. and Benford,D.J. (1993) In vivo rat liver UDS assay. In Report of the UKEMS Sub-committee on Guidelines for Mutagenicity Testing. Part II Revised. UKEMS, University Press, Cambridge, pp. 52–77.
Lund-Pero,M., Jeppson,B., Arkneklo-Nobin,B., Sjögren,H.O., Holmgren,K. and Pero,R.W. (1994) Non-specific steroidal esterase activity and distribution in human and other mammalian tissues. Clin. Chim. Acta, 224, 9–20.[Web of Science][Medline]
Maron,D. and Ames,B.N. (1983) Revised methods for the Salmonella mutagenicity test. Mutat. Res., 113, 173–215.[Web of Science][Medline]
Nykänen,L. and Suomalainen,H. (1983) Aroma of Beer, Wine and Distilled Alcoholic Beverages. Reidel, Dordrecht, The Netherlands.
Obe,G., Jonas,R. and Schmidt,S. (1986) Metabolism of ethanol in vitro produces a compound which induces sister-chromatid exchanges in human peripheral lymphocytes in vitro: acetaldehyde not ethanol is mutagenic. Mutat. Res., 174, 47–51.[Web of Science][Medline]
Richold,M., Ashby,J., Bootman,J., Chandley,A., Gatehouse,D.G. and Henderson,L. (1990) In vivo cytogenetics assays. In Report of the UKEMS Sub-committee on Guidelines for Mutagenicity Testing. Part I Revised. UKEMS, University Press, Cambridge, pp. 115–137.
Robinson,W.D., Green,M.H.L., Cole,J., Garner,R.C., Healy,M.J.R. and Gatehouse,D.G. (1989) Statistical evaluation of bacterial/mammalian fluctuation tests. In Report of the UKEMS Sub-Committee on Guidelines for Mutagenicity Testing. Part III. UKEMS, University Press, Cambridge, pp. 102–140.
Satoh,T. and Hosokawa,M. (1998) The mammalian carboxylesterases: from molecules to functions. Annu. Rev. Pharmacol. Toxicol., 38, 257–288.[Web of Science][Medline]
Savage,J.R.K. (1976) Classification and relationships of induced chromosomal structural changes. J. Med. Genet., 13, 103–122.
Scott,D., Danford,N., Dean,B.J., Kirkland,D. and Richardson,C.R. (1983) In vitro chromosome aberration assays. In Report of the UKEMS Sub-committee on Guidelines for Mutagenicity Testing. Part I. UKEMS, University Press, Cambridge, pp. 41–64.
Yahagi,T., Degawa,M., Seino,Y., Matsushima,T., Nagao,M., Sugimura,T. and Hashimoto,Y. (1975) Mutagenicity of carcinogenic azo dyes and their derivatives. Cancer Lett., 1, 91–96.[Web of Science][Medline]
Received on May 13, 1999; accepted on September 10, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||