Mutagenicity testing of diethylene glycol monobutyl ether.

The mutagenic potential of diethylene glycol monobutyl ether (diEGBE) was examined with a Tier I battery of in vitro assays followed by a Tier II in vivo Drosophila sex-linked recessive lethal assay. The in vitro battery consisted of: the Salmonella mutagenicity test, the L5178Y mouse lymphoma test, a cytogenetics assay using Chinese hamster ovary cells and the unscheduled DNA synthesis (UDS) assay in rat hepatocytes. Results of the Salmonella mutagenicity test, the cytogenetics test, and the rat hepatocyte assay were negative at concentrations up to 20 microL/plate, 7.92 microL/mL, and 4.4 microL/mL, respectively. Toxicity was clearly demonstrated at all high doses. A weak, but dose-related increase in the mutation frequency (4-fold increase over the solvent control at 5.6 microL/mL with 12% survival) was obtained in the L5178Y lymphoma test in the absence of metabolic activation. Results of the mouse lymphoma assay were negative in the presence of the S-9 activation system. The significance of the mouse lymphoma assay were negative in the presence of the S-9 activation system. The significance of the mouse lymphoma assay results were assessed by performing the Tier II sex-linked recessive lethal assay in Drosophila in which the target tissue is maturing germinal cells. Both feeding (11,000 ppm for 3 days) and injection (0.3 microL of approximately 14,000 ppm solution) routes of administration were employed in the Drosophila assay. Approximately 11,000 individual crosses with an equal number of negative controls were performed for each route of administration. diEGBE produced no increase in recessive lethals under these conditions.(ABSTRACT TRUNCATED AT 250 WORDS)


Introduction
The mutagenic potential of diethylene glycol monobutyl ether (diEGBE) was examined with a Tier I battery of in vitro assays followed by a Tier II in vivo Drosophila sex-linked recessive lethal assay. The in vitro battery was designed to measure point mutations in prokaryotes (Salmonella mutagenicity assay) and eukaryotes (L5178Y mouse lymphoma assay). The battery also measures changes at the chromosome level (in vitro cytogenetics with CHO cells) and nonspecific DNA damage as measured by unscheduled DNA synthesis (UDS) in primary cultures of rat hepatocytes.

CH2-CH2-O-CH2-CH2
OH 0-CH2-CH2-CH2-CH3 We feel that these four in vitro tests provide an adequate measure of mutagenic potential. However, occasionally isolated positive results are obtained with the Tier I battery, and it is necessary to perform in vivo Tier II testing to add perspective to these results. Our Tier II battery is designed to detect similar types of mutations in germinal tissues. Thus, positives in either the Salmonella mutagenicity assay or the mouse lymphoma assay are resolved by performing the sex-linked recessive lethal assay in Drosophila. Positives in the cytogenetic or UDS assays are resolved by performing in vivo cytogenetic assays in bone marrow, or alkaline elution assays in germinal tissues of rodents, respectively. The results of the Tier I battery of mutagenicity assays and a Drosophila sex-linked recessive lethal assay on diEGBE are presented in this communication.

Salmonella Mutagenicity Test
The plate incorporation mutagenicity assay with and without an Aroclor-induced Sprague-Dawley rat liver S-9 fraction was performed as described by Ames et al. (1). A preliminary toxicity assay was performed with strain TA100 in order to determine an appropriate dose range. Toxicity was estimated by examining the background lawn with a stero microscope over a wide range of doses. As toxicity increases, the density of the microcolonies decreases while their size increases with respect to the negative control. Toxicity was also estimated by plating approximately 300 viable cells of TA100 onto minimal medium. The cells were suspended in the top agar normally used except it contained 10 x histidine which permits normal growth. A decrease in the number of colonies appearing after 48 hr of incubation relative to the negative control is the measure of toxicity.
L5178Y Mouse Lymphoma Assay L5178Y TK+'-3.7.2C cells were obtained from Dr. Donald Clive, Burroughs-Wellcome Co., Research Triangle Park, NC. The cells were maintained and cleansed of spontaneous TK7--mutants biweekly as described by Clive et al. (2). Mutagenicity tests were performed with and without Aroclor-induced Sprague-Dawley rat liver S-9 activation (2). A 2-day expression period was used to detect mutation at the TK locus.
Unscheduled DNA Synthesis Assay in Primary Cultures of Rat Hepatocytes Primary cultures of rat hepatocytes were prepared by using the collagenase perfusion technique described by Williams et al. (3). Sprague-Dawley rats (Charles River) of approximately 200 g were used for each preparation. Hepatocyte yield averaged 2 x 108 viable cells. Overall viability averaged 91% on the basis of trypan blue exclusion. Hepatocytes were seeded onto 25 mm diameter sterile plastic coverslips in eight-well tissue culture plates containing Williams medium E supplemented with 10% (v/v) fetal bovine serum and penicillin and streptomycin. After a 1.5 to 2 hr attachment period (at 37°C in an atmosphere of 5% CO2 in air), unattached cells were removed by gently washing. Preliminary toxicity tests were performed on all materials before UDS assays were conducted. The test chemical in a wide range of doses was added to hepatocyte cultures along with 10 ,uCi/mL of 3H-thymidine (New England Nuclear, 20 Ci/mmole). After overnight incubation, cultures were washed and viability was determined on the basis of trypan blue exclusion. UDS assays were performed over a more narrow dose range.
After exposure to the test material and 3H-thymidine, the cultures were processed for light microscope autoradiography and UDS determination (4,5).
At least 50 randomly chosen, normal-appearing hepatocytes were scored for UDS by counting the darkened silver grains localized over the nucleus. All scoring was done with an Artek 880 electronic counter equipped with an auxiliary TV camera and Zeiss standard microscope with a 100 x oil emersion objective. Net incorporation of 3H-thymidine into a nucleus was determined by subtracting the mean grain count of three nucleus-sized areas adjacent to the nucleus.

In Vitro Cytogenetics Assay with CHO Cells
The methods used for the assay for induction of chromosome aberrations in CHO cells were a modification of those described by Natarajan et al. (6). CHO cells were obtained from the American TIype Culture Collection and were maintained in Ham's F12 medium (GIBCO) supplemented with 10% (v/v) fetal bovine serum (Reheis). In each assay, cells were treated in suspension with the test material or test material plus S-9 mix for 4 hr at 37°C with constant, gentle agitation. S-9 mix consisted of one part rat liver homogenate prepared from 200 to 300 g Sprague-Dawley rats induced 5 days prior to sacrifice with an intraperitoneal injection of Aroclor 1254 combined with 2 parts of a cofactor solution prepared in Ham's F-12 medium without serum. The cofactor solution contained glucose-6-phosphate (7.1 mM), MgC12 (7.1 mM) and NADP (5.7 mM). The cofactor solution was adjusted to pH 7.0 with NaOH or HCl. Doses for the cytogenetics assay were chosen on the basis of a preliminary toxicity assay. The toxicity assay consisted of a determination of the effects of a wide range of test material on the plating efficiency of CHO cells following a 4-hr exposure with and without S-9.
In the cytogenetic assay, the cells were washed following exposure to test material, seeded into T25 culture flasks, and incubated approximately 16 hr at 37°C. Cultures were then exposed to 2 ,ug/mL colcemid for 2 hr to arrest cells in metaphase. Mitotic cells were selectively dislodged by striking the side of the culture flask with the palm of the hand. Suspended cells were collected by centrifugation, resuspended in 0.075 M KCl and incubated 4 min at room temperature. Cells were again collected by centrifugation and fixed by slowly resuspending them in Carnoy's fixative (methanol:glacial acetic acid, 3:1). Cells were then centrifuged and resuspended in fresh fixative and stored overnight at 4°C. The following day, the fixed cells were centrifuged, resuspended in Carnoy's fixative, dropped on an inclined microscope slide, and dried. Slides were then coded and scored for chromosome number and aberrations. In each experiment the top three doses showing less than 90% relative toxicity were scored. A total of 50 metaphase spreads per dose were evaluated.

Drosophila Sex-Linked Recessive Lethal Assay
The Drosophila sex-linked recessive lethal assay has been described in detail by Wurgler et al. (7). In this study wild-type Canton S males were dosed either by feeding in a 5% sucrose vehicle buffered to pH 6.8 with phosphate buffer or by injection with a 0.7% NaCl solution.
Feeding was done in shell vials containing a glass fiber filter (Whatman g F/A) saturated with 0.2 mL of the feeding solution. One-day-old males were placed in the vials and transferred to new treatment vials daily for 3 days. Each treatment vial contained 0.2 mL of freshly prepared feeding solution. Negative control flies were fed by the same procedure on 5% sucrose buffered to pH 6.8 with phosphate buffer. Dimethylnitrosamine (25 ppm in the feeding and 500 ppm in the injection study) was the positive control compound for this study.
Injection was done with a glass needle attached by polyethylene tubing to a Hamilton 5-,uL syringe. The plunger was moved by a micrometer so that approximately 0.3 ,uL was injected into the body of the etherized fly. A single injection was performed, and the flies were allowed to recover on regular culture medium for 24 hr before mating.
Following dosing, the wild-type Canton S males were individually mated for 3 days with three virgin Basc females, whose X chromosomes carry inversions and are marked with apricot eye and bar eye. The females were 3 to 10 days old when mated. After 3 days, each male was transferred to three new virgin females. After 2 days the males were again transferred to vials with three new virgin females. Two days later the males were discarded. Each group of females was allowed to lay eggs for 4 days and then they were discarded. This type of mating scheme examines only postmeiotic germ cells corresponding roughly to mature sperm (brood 1), spermatid (brood 2), and spermatocyte (brood 3).
The cultures from each brood were maintained for approximately 10 days, then brothers and sisters were mated individually. Approximately 33 F1 matings were performed from each brood for each male. Thus, 99 chromosomes were examined from each dosed male.
The F2 generation was scored when the cultures were 11 to 15 days old for the absence of the wild-type X chromosome in the male. Suspected lethal cases were retested by mating the heterozygous female with the males from the F2 cross and observing the F3 generation.
The recessive lethal data were evaluated by comparing the overall mutation frequency of each treated group with the overall mutation frequency of its concurrent negative controls, by applying the Kastenbaum-Bowman test (8).
When individual P1 males produced more than one lethal among his progeny, the group of lethals is referred to as a "multiple." A multiple may result from many independent mutations in the post-micotic germ cells (as is the case with a potent mutagen) or from a single spontaneous mutation in a gonial cell, which then multiplies to produce a clusterof germ cells, each carrying an identified recessive lethal mutation. Thus, each multiple must be subjected to analysis to determine whether or not it could be considered a "cluster." The statistical test is based on a cumulative Poisson distribution as described by Owen (9). Given the number of lethals in the multiple, the average number of lethals per male, and the number of fertile males, the probability for independent events is calculated. If a multiple is concluded to be a cluster, those data are removed from the data.

Results
Salmonella Mutagenicity Test Table 1 shows the results of the toxicity assay for strain TA100. At 20 ,uL/plate, diEGBE produced a slight depression in both the viable counts and the background lawn. Thus, the assay was performed at a dose which did produce noticeable toxicity. Table 2 shows the results of the Salmonella mutagenicity test. Dose-related responses were not obtained with any of the strains either in the presence or absence of the S-9 activation system. All positive controls produced at least a 5-fold increase over the spontaneous mutation rate.
L5178Y Mouse Lymphoma Assay diEGBE was tested for the induction of forward mutations in L5178Y mouse lymphoma cells over a dose aViable counts are determined by plating an appropriately diluted culture of TA100 onto minimal medium. The overlay agar contains lOx histidine which allows the cells to grow normally. bSubjective evaluation of the density of the background bacterial lawn when viewed through a stereo microscope at approximately 40 x magnification. A 1 respresents a normal bacterial lawn. A 2 represents a slightly reduced bacterial lawn: 15-25% thinning of the microcolonies which is accompanied by an increase in size of the microcolonies compared to the negative control.  TA1537  TA1538  TA98  With S-9 No S-9 With S-9 No S-9 With S-9 No S-9 With S-9 i 181 ± 10 6 ± 2 10 ± 5 12 ± 5 16 ± 7 26 ± 2 40 ± 3 177 ± 7 6  9-Aminoacridine, 75p.Lg 735 ± 163 aTests were performed in the absence of the S-9 metabolic activation system or in the presence of the S-9 metabolic activation system. Each number is the average number of revertant colonies on three plates with the standard deviation. range of 0.42 [uL/mL to 5.6 iuL/mL in the absence of metabolic activation and over a dose range of 0.56 ,iL/mL to 7.5 ,uL/mL in the presence of metabolic activation. The maximal toxicity achieved was 88% kill in the absence of metabolic activation and 42% kill in the presence of the activation system. The next higher dose (10 ,uL/mL) in the latter assay produced 100% toxicity. Thus, diEGBE was tested throughout a range of toxicities (Table 3). When the cells were plated in selective medium, a weak, but dose-related response was obtained in the absence of metabolic activation. diEGBE was negative in this assay in the presence of the metabolic activation system (Table 3).

UDS in Rat Hepatocytes
Results of the UDS assay in rat hepatocytes are shown in Table 4. diEGBE did not induce UDS in these cells through a suitable range of toxicities.

In Vitro Cytogenetics Assay in CHO Cells
Results of the toxicity assay for the in vitro cytogenetics assay in the presence and absence of a S-9 metabolic activation system are shown in Table 5. Both assays were performed over a dose range that produced toxicity. However, in the presence of the activation system the toxicity curve appears much steeper.
Cytogenetic analyses of the three highest doses showing less than 90% toxicity are shown in Table 6. diEGBE was not clastogenic to these cells.

Sex-Linked Recessive Lethal Assay in Drosophila
Preliminary toxicity assays indicated that 14,000 ppm was a suitable dose level for the injection portion of the assay (average of 36% mortality 24 hr after the dose was administered). The preliminary toxicity assay for the feeding experiment was complicated apparently by the humidity in the dosing room. Table 7 shows the results of five preliminary toxicity assays, some performed under dry room conditions and some performed under humid conditions. These data clearly show that at least for diEGBE the relative humidity greatly affects the dose tolerated by the flies. It is believed that the flies simply ingest less under humid conditions (Ruby Valencia, personal communication). Based on these assays 11,000 ppm was selected for the feeding studies.
Results of the injection and feeding portions of the sex-linked recessive lethal assay are shown in Tables 8  and 9, respectively. Approximately 11,000 chromosomes were examined in the injection and feeding studies with an equal number examined in the concurrent controls. Thus, a total of 44,000 chromosomes were examined in this study. diEGBE did not induce sex-linked recessive lethal mutations in this assay. Table 5. Results of the toxicity assay on diEGBE for the in vitro cytogenetics assay with CHO cells in the presence and absence of a S-9 metabolic activation system.  bChromatid gaps are recorded, but included in the aberrations per cell calculation. cEach aberration type is given an arbitrary weight of one except pulverized and > 10 aberrations which are weighted as 10. dTested with S-9 metabolic activation system.

Conclusions
The results of three of the in vitro assays, the Salmonella mutagenicity test, the cytogenetics assay in CHO cells, and the UDS assay in rat hepatocytes, were straightforward negatives. Each assay was performed either at the limits of toxicity to the cells or at the limit of compound that could be added to the test system. Decreasing doses from this level provided an adequate number of data points to ensure that any possible dose-response would have been detected.
The mouse lymphoma assay presented conflicting results. In the absence of the S-9 metabolic activation system, a dose-related increase was obtained with the highest response being approximately 5-fold above the background level. In the presence of the S-9 metabolic activation system the compound was negative. It should be emphasized here that the response obtained in the absence of S-9 is relatively weak on a molar basis. One ,umole of diEGBE induced 1.1 mutations/106 survivors/hr of exposure at 10% survival compared to 812 for methyl methanesulfonate or 1.8 x 106 for N-methyl-N-nitro-N-nitrosoguanidine at the same relative toxicity (2).
The negative results in the presence of the S-9 activation system indicates that the mutagenic compound may be metabolized to a nonmutagenic compound. Such inactivation suggests that the compound would be metabolized in an in vivo test system and would be negative.
The sex-linked recessive lethal assay was run to  aMultiples may be derived from spontaneous mutations in gonia cells. Since these cells continue to divide, mutations occur in multiples. Each multiple is analyzed statistically (9), and if found significant the data derived from that P1 male are removed.
bThe percent lethals are high because of a multiple. These data are not used in the final calculations. 3.69 5(4) 2(6) 1 (9) aMultiples may be derived from spontaneous mutations in gonia cells. Since these cells continue to divide, mutations occur in multiples.
Each multiple is analyzed statistically (9), and if found significant the data derived from that P1 male are removed. bThe percent lethals are high because of a multiple. These data are not used in the final calculations.
resolve the question of mutagenic activity in an in vivo system. Since this assay measures the transmittance of mutation in germinal tissue to future generations, it provides information as to the potential genetic burden of a mutagen. More than 44,000 chromosomes were examined in the two tests and the results were conclusively negative. Statistical evaluation of these data indicate 95% confidence that diEGBE would produce illl less than a 0.2% increase in the mutation frequency. A 0.2% increase in the mutation frequency is the maximal power of the sex-linked recessive lethal assay.
Based on the three negative in vitro assays, a very weak positive in the mouse lymphoma assay and a conclusive negative in the in vivo assay, we conclude that diEGBE does not present any genetic risk to humans.