- Email: [email protected]
False positive result for a peptide drug in the gene conversion assay with Saccharomyces cerevisiae strain D7
FUNDAMENTAL
AND
APPLIED
TOXICOLOGY
17,627-634
(1991)
False Positive Result for a Peptide Drug in the Gene Conversion with Saccharomyces cerevisiae Strain D7’
Assay
LINVAL R. DEPASS,* REBECCA L. CHAN,* D. R. JAGANNATH,~ AND IRWIN A. HEYMAN* *Syntex
Research,
Palo Alto,
California
Received
94304,
February
and THazleton
Biotechnologies,
12, 1991; accepted
May
Kensington,
Maryland
20895
29, 1991
False Positive Result for a Peptide Drug in the Gene Conversion Assay with Saccharomyces cerevisiae Strain D7. DEPASS, L. R., CHAN, R. L., JAGANNATH, D. R., AND HEYMAN, I. A. (199 1). Fundam. Appl. Toxicol. 17,627-634. A battery of mutagenicity tests was performed with nafarelin, an agonist analogue of luteinizing hormone releasing hormone (LHRH) containing tryptophan (Trp) and histidine (His). Included were the Ames assay and the gene conversion assay with yeast strain D7. Roth tests were negative without S9 activation, and the Ames test was negative with S9, but the yeast test was positive with S9 activation. Since the yeast test is based on conversion of cells to Trp independence, release of Trp by metabolism of the drug could account for the positive result. The test was repeated using Trp instead of the drug. The result was positive even at the lowest Trp concentration. In another experiment with the drug amino acid analysis of the incubation mixture revealed the presence of Trp but no detectable His. Since the Ames test is based on mutation to His-independent cells, these data are completely consistent with the negative result in the Ames test and the false positive result in the yeast test. These data suggest the need for caution in interpreting the results from mutagenicity assays with peptide drugs. 0 1991 society of Toxicology
The Ames Salmonella assay and the gene conversion assay with yeast cells have been used extensively in the testing of drugs and other chemicals for mutagenic activity (Ames et al., 1975; Zimmerman, 1975; Zimmerman et al., 1975). Whereas the Ames assay relies upon reversion to a histidine-requiring state, the yeast assay relies upon reversion to a tryptophan-requiring state. The results of these tests are generally acceptable to regulatory agencies as evidence of the presence or absence of genetic toxicity of a drug. These tests were performed as part of a standard mutagenicity battery to assess the potential genetic activity of nafhrelin, an agonistic analogue of luteinizing hormone-releasing hormone (LHRH). The structures of
native LHRH and nafarelin are presented in Fig. 1. The initial results indicated that nafarelin, [D-N~~(~)~]LHRH, was negative in both assays in the absence of metabolic activation and also negative in the Ames assay with activation. However, the result was positive in the yeast assay performed with rat liver S9 activation. Since the probability of true genetic toxicity seemed remote for a drug that differs from native LHRH at only one amino acid residue, a series of experiments were designed in an attempt to explain the “positive” result. We were particularly interested in the possibility that the positive result may have been an artifact of the test system caused by release of tryptophan (Trp) by hydrolysis of nafarelin. MATERIALS
’ Presented in part at the Annual Meeting of the Society of Toxicology, New Orleans, Louisiana March, 1986, and at the Annual Meeting of the Drug Safety Sub-Section of the Pharmaceutical Manufacturer’s Association, Nashville, Tennessee, November, 1986.
AND
METHODS
Chemicals. Nakuelin samples used in these experiments were synthesized and analyzed for purity by Syntex Re search. The sample used in the mutagenicity studies was 627
0272-0590191
$3.00
Copyri&t 0 1991 by the society of Toxicology. All rights of repmductioo in any form nserved.
DEPASS ET AL.
628
Determination of cytotoxicity and mutagenicity. To determine bacterial cytotoxicity and mutagenicity in the 1 2 3 4 5 6 7 8 9 10 Salmonella strains, 0.025 to 0.15 ml of a solution of the pGlu - HIS - Trp - Se; Tyr - Gly Leu - Arg Pm Gly NH2 test substance in DMSO was added to a sterile test tube containing 2.0 ml of overlay agar, 0.2 ml of an overnight (6) culture, and 0.5 ml of 0.2 M phosphate buffer, pH 7.4, in a 43 to 45°C water bath. This mixture was swirled gently 1 2 3 4 5 6 7 6 9 10 pGlu HIS Trp - Ser. Tyr - NH - 7” - CO - Leu Arg - Pro - Gly NH2 and then poured onto minimal agar plates. After the overlay agar had set, the plates were incubated at 37°C for CH2 approximately 2 days. For cytotoxicity, the number of ,a 0 0 colonies growing on the plates was counted and recorded. A reduction in the number of revertants, appearance of microcolonies, or clearing of the background lawn on the FIG. 1. Structures of (A) LHRH and (B) nafarelin ([Ddrug-treated plates as compared to the solvent control Nal6]LHRH), where D-Nal is 3-(2-naphthyl)-D-alanine. plates was considered as an indication of toxicity. For mutagenicity, the number of His+ revertant colonies growing on the plates was counted and recorded. 98.8% pure, whereas that which was used in the metabolism The activation assay was run concurrently with the studies was 97.2% pure. The positive control mutagens nonactivation assay.The only difference was the addition used in the mitotic gene conversion studies were ethyl- of 0.5 ml of S9 mix in place of 0.5 ml of phosphate buffer. methanesulfonate (EMS, Eastman Kodak, Rochester, NY) To determine cytotoxicity in the gene conversion assay, and sterigrnatocystin (Sigma Chemical Co., St. Louis, MO). 0.025 to 0.15 ml of a solution of the test substance in The positive control mutagens used in the Ames SalmoDMSO was added to a sterile 20-ml scintillation vial connella assayswere sodium azide, 2-nitrofluorene, 9-amitaining 0.3 ml of the indicator organism (1 to 2 X IO* noacridine, and 2-aminoanthracene obtained from Sigma. cells) and 2.65 ml of 0.2 M phosphate buffer, pH 7.4. This Stock solutions were prepared in dimethylsulfoxide mixture was incubated for approximately 18 hr at 30°C (DMSO, Sigma). in a rotary shaker. After incubation, cell survival was deDimethylaminoazobenzenesulfonyl chloride (DABStermined by adding 0.2 ml of 10e4 dilution of the suspenCl), used for the derivatization of amino acids, was ob- sion to 2 ml of molten overlay agar, which was then poured tained from AG Fluka, Switzerland. Amino acid standards, onto yeast complete plates. These plates were incubated NADP, and glucose 6-phosphate were purchased from for 2 to 3 days and scored for surviving population. Sigma. To determine gene conversion, the same procedure was Cell culture. The Salmonella typhimurium strains used followed except that after incubation, the surviving pop in these studies were originally obtained from Dr. Bruce ulation was estimated by plating an ahquot of a lO-4 diAmes, University of California at Berkeley. The following lution of each cell suspension in 0.15 M NaCl onto yeast strains were used: TA-1535, TA-1537, TA-1538, TA-98. complete plates (five plates per treatment). Mitotic gene and TA- 100. The Saccharomyces cerevisiae strain D7 used conversion was measured by plating 0.2 ml of undiluted in these assayswas originally constructed by Dr. F. K. samples onto minimal media supplemented with adenine Zimmerman (Zimmerman et al., 1975). The SalmoneZla and isoleucine (five plates per treatment level). The plates strains were grown overnight at 37°C in Oxoid Media No. were incubated at 30°C for 4 to 6 days and scored for 2 (nutrient broth). The minimal media plates for the se- survivors and convertants. lection of histidine revertants consisted of the Vogel Bonner The activation assay,containing 1.5 ml of S9 mix and Medium E (Vogel and Bonner, 1956) with 2% glucose and 1.15 ml of phosphate buffer (instead of 2.65 ml of buffer), 1.5% bactoagar. Log phase cultures of the yeast strain D7. was run concurrently with the nonactivation assay. grown in yeast complete broth, were washed twice in Amino acid analysis. Samples were derivatized with phosphate buffer and resuspended in the buffer at a density DABS-Cl and analyzed with a reversed-phase HPLC sysof 1 to 2 X lo* cells/ml and used in these studies. The tem according to a published procedure (Chang et al., 198 1, yeast complete media plates for scoring survivors and the 1982). The 18-hr incubation mixtures of nafarelin with yeast synthetic media plates without tryptophan for de- S9 mix (with or without yeast cells) were centrifuged. Altecting the convertants were prepared according to the iquots (30 ~1) of the supernatant were mixed with 20 ~1 methods of Zimmerman (1975). of a stock solution of phenylalanine (Phe, 300 nmol/ml). 100 ~1 of 0.2 M sodium bicarbonate buffer (pH 9.0), and Preparation of the rat liver homogenate reaction mixture 100 ~1 of DABS-Cl in acetone (6 nmol/rl). Phe was added (S9 mix). Aroclor 1254-induced rat liver homogenate (9000g soluble fraction, S9). purchased from Bionetics Lab as an internal standard for the quantitation of tryptophan Products (Charleston Heights, SC) was used as the exog- (Trp). The mixture was heated at 70°C for 10 min, dried enous source of mixed function oxidase enzyme system. under nitrogen, and redissolved in 400 pl of 50% ethano! The S9 mix was prepared in pH 7.4, 0.1 M sodium phos- in 0.04 M, pH 6.5. sodium phosphate. Aliquots (20 ~1) were analyzed on the HPLC column (detector at 444 nm) phate buffer containing 33 mM KCI, 8 mM MgC&, 5 mM within 1 hr after derivatization. glucose 6-phosphate. 4 mtvrNADP, and 10% (v/v) S9. (A)
FALSE POSITIVE TA-1535
0
GENE CONVERSION
TA-1537 -u--
1500
3000 Concentration
629
TA-100 . ..#--.
TA-1538 -*-
4500
DATA
6000
of Nafarelin
TA-98 + ..
7500
9000
10500
( pg/Plate)
FIG. 2. Results of the Ames activation assaywith nafarelin. Results were similar without activation.
The HPLC systemconsisted of a Varian LC 5060 liquid chromatograph equipped with a Varian UV-50 variable wavelength detector and a 5 pm Lichrosorb RP- 18 column (Rainin, Wobum, MA). A 3.2 mm X 5 cm precolumn packed with 37-53 Frn of pellicular silica gel (Whatman, Clifton, NJ) was placed before the injector as a saturation column. Solutions containing known quantities of Trp (O-6 nmol) were derivatized with DABS-Cl and analyzed on
the HPLC. A calibration curve was prepared from the Trp:Phe peak height ratios versus Trp concentrations. The amounts of Trp released were estimated from the Trp:Phe peak height ratios. Degradation ofnafarelin with S9 mix. A mixture of 2.7 ml of S9 mix. 0.3 ml of nafarelin stock solution (IO mg/ ml in DMSO), and 3 ml of 0.2 M sodium phosphate (pH 7.4) was incubated at 37°C. At various time points, 50-~1 aliquots of the incubation mixture were added to 400 pl
TABLE 1 RESULTS OF THE MITOTIC GENE CONVERSION ASSAY WITH NAFARELIN IN THE ABSENCE OF EXOGENOUS MAMMALIAN METABOLISM
Dose” DMSO EMS Nafarelin
100 &3 ml 1.O% 1.0 10.0 100.0 1000.0
5000.0 10000.0
Number of cells scored x104 811 317 778 931 843 837 553 526
’ Values are rg/ml unless otherwise indicated.
Percentage survivors
Total no. convertants
Convertants per 1OSsurvivors
100.0 39.1 95.9 114.8 103.9
345 1730 344 373 348 392 302 289
4.3 54.6 4.4 4.0 4.1
103.2
68.2 64.9
4.7
5.5 5.5
630
DEPASS TABLE
ET
AL 2
RESULTS OF THE MITOTIC GENE CONVERSION ASSAY WITH NAFARELIN IN THE PRESENCE OF EXOGENOUS MAMMALIAN METABOLISM
DMSO Sterigmatocystin Nafarelin
’ Values
are pg/ml
Dosea
Number of cells scored x104
100 pl/3 ml 5.0 1.0 10.0 100.0 1000.0 5000.0 10000.0
891 661 825 814 812 815 113 60
unless otherwise
The results of the Ames/Salmonella assay with nafarelin in the presence of S9 activation are presented in Fig. 2. The results demonstrated no increase in the number of revertants at any dose tested in the five strains. The results
TABLE
DMSO Sterigmatocystin Nafarelin
100 pl/3 ml 5.0 1.0 10.0 100.0 1000.0 5000.0 10000.0
a Values are &ml unless otherwise b Mean f SD of three experiments.
Total no. convertants
Convertants per lo5 survivors
402 1339 381 383 360 568 532 304
4.5 111.0 4.6 4.4 4.4 7.0 31.2 50.7
indicated.
RESULTS
Dose”
SUlViVOrS
100.0 14.2 92.6 98.1 91.1 91.5 12.7 6.1
of ethanol and centrifuged. Aliquots (40 ~1) of the supernatants were injected onto a 5-pm Rainin Microsorb Cl 8 column. The decrease of peak height of nafarelin on the HPLC chromatogram (detector at 225 nm) was a measure of its extent of degradation.
RESULTS OF THE MITOTIC OF EXOGENOUS
Percentage
of the Ames/Salmonella assay in the absence of S9 activation enzymes were similar and are, therefore, not presented. The results of the mitotic gene conversion assay without activation are presented in Table 1. There was no increase in number of convertants with increasing concentration of nafarelin. The initial results of the mitotic gene conversion assay with activation are presented in Table 2. The data show increased numbers of convertants per lo5 survivors at the three highest concentrations of nafarelin. The num-
3
GENE CONVERSION ASSAY WITH NAFARELIN ~MMALIAN METABOJJSM USING WASHED
IN THE PRESENCE CELL.Y
Number of cells scoredb x104
Percentage survived
Total no. convertants’
1118 + 36Ok 1013 + 975 f 959 + 83Ok 502 f 216+
100.0 32.1 -t 3.9 90.7 + 5.3 87.3 + 1.5 86.1 + 3.6 75.0 f 12.5 44.3 + 7.3 19.4 k 3.6
301 + 33 4469 z!z 936 312+ 34 302+ 41 300* 55 425 + 134 385 + 35 307 + 115
indicated.
164 81 143 136 108 118 139 29
Convertants per IO5 survivom~ 2.76 124.38 3.15 3.18 3.20 5.18 8.12 14.51
f + k f f + f +
0.13 4.77 0.80 0.91 1.00 1.76 2.52 6.42
FALSE
POSITIVE
GENE TABLE
RESULTS
OF THE MITOTIC OF EXOGENOUS
Dose= DMSO
Sterigmatocystin Tryptophan
50.0 &ml 5.0
1.0 10.0 100.0 250.0 500.0
1000.0 L?Values are &ml
CONVERSION 4
GENE CONVERSIONASSAY WITH TRY~~OPHAN MAMMALIAN METABOLISM USING WASHED Number of cells scored x104
830 730 785 754 754 791 767 728
631
DATA
Percentage
survivors 100.0 88.0 94.6 90.8 90.8 95.3 92.4 87.7
IN THE PRESENCE
CELLS
Total no. convertants
Convertants per IO5 survivors
151 1 1174
I .82 55.63 14.96 16.10 19.73 18.94 17.00 4.66
406
1214 1488 1498 1304 339
unless otherwise indicated.
ber of cells scored and the percentage survivors The results of the experiment with washed were greatly reduced at the two highest con- cells are presented in Table 3. Again, there centrations. The degree of cytotoxicity at high was evidence of toxicity to the indicator strain, concentrations was greater than that observed especially at the two highest concentrations. in the absence of the S9 fraction. Note also A dose-related increase in the convertant frethat the total number of convertants was not quency was observed at the three highest consignificantly increased compared to the solvent centrations of nafarelin. However, the increase control. Nevertheless, because of the positive above solvent control was smaller than that dose response over three drug concentrations, observed in the initial experiment in which the result was interpreted as evidence of ge- the cells were not washed. Again, there was netic toxicity of nafarelin. no significant increase in the absolute number Because nafarelin differs from native LHRH of convertants at any concentration compared at only a single amino acid residue, there was to the solvent control value. To confirm the no reason to expect that this compound would apparent positive result with the washedalter cellular DNA so as to produce intragenic cells, this experiment was repeated twice. The recombination (gene conversion). Instead, we results were similar with all three replicates hypothesized that growth of the yeast colonies (Table 3). on Trp-free medium may have occurred by In order to ascertain if tryptophan was reuptake and metabolic utilization of Trp re- sponsible for the observed increase in converleased by hydrolysis of nafarelin. Transfer of sion frequency, an additional experiment in the drug or its degradation products may have which Trp was substituted for nafarelin was occurred when the suspended yeast cells were conducted. Again, one portion of the cells was plated onto minimal media. washed three to four times in phosphate buffer We tested the aforementioned hypothesis before being plated onto minimal media. A by repeating the experiment with one change separate aliquot of cell suspension was plated in procedure. After the incubation, the treated without being washed to evaluate the effeccell suspension was washed three to four times tiveness of the washing procedure in removing by centrifuging and suspending the cell pellet Trp. The highest concentration of Trp (1.0 in saline or in phosphate buffer. The washed mg/ml) was chosen to be approximately equal cells were then plated as before, and survivors to that which would be expected if all of the and convertants were scored. Trp were released by metabolic activity of
632
DEPASS ET AL.
Standards
0.25 mg/ml
0.5 mg/ml
Phe TOP
Pyle
JL.AJ.L -P
h
0.75 mgiml
!?,t
1.0 mg/ml
L, Phe
n
l-a--J-\ 0
16
24
32
Min
-L-Y--k 8
16
24
32
Min
FIG. 3. Amino acid analysis of the 1%hr incubation mixtures of nafarelin, yeastcells, and S9 mix. Aliquots of the incubation mixtures containing various initial concentrations (0 to 1.0 mg/ml) of nafarelin were derivatized with DABS-Cl and analyzed by gradient elution on a LiChrosorb RP- 18 HPLC column as previously described (Chang et al., 198 1). Detector wavelength was 444 nm.
the S9 mix at the highest concentration of nafarelin used in the previous experiments (10 w/ml). The results of the experiment with Trp using washed cells are presented in Table 4. Increased numbers of “convertants” compared to the solvent control were observed even at the lowest concentration of 1.O pg/ml. The results with unwashed cells (not presented) were similar except that the cells formed a “lawn” at the highest concentration of Trp. In contrast, there was a reduction in number of “convertants” at the highest concentration
relative to that observed at lower concentrations, when the cells were washed. Aliquots of the 18-hr incubation mixtures containing 0 to 1.Omg/ml of nafarelin, S9 mix, and yeast cells were derivatized with DABSCl. The HPLC chromatograms of the derivatized mixtures are shown in Fig. 3. The amount of Trp liberated in the incubation mixture increased with increasing concentration of nafarelin. The quantities of Trp released were calculated from the Trp:Phe peak height ratios. The percentage of total Trp released ranged from 19 to 27% and was inde-
FALSE POSITIVE 0 Hr
GENE CONVERSION 0.25 Hr
!L
I 1. i A-
2 Hr sample spiked with 4-10 standard peptide
4-10
Naf
Naf
6
6
= Nafarelin
Min
L 2
Naf
5 Hr
S-10
4-10
4
2 Hr
Naf
Naf
2
633
DATA
4
Naf
6
6
Min
2
4
6
6
Min
RG. 4 A. time-course experiment showing formation of 4- 10 and 5-10 peptide fragments of nafarelin alter incubation with S9 mix. Aliquots of the incubation mixture were deproteinized with ethanol and centrifuged. Portions of the supematant were injected on a Microsorb C 18 HPLC column. The mobile phase contained 32% of acetonitrile and 68% of 0.1 M phosphate (pH 2.6). Detector wavelength was 225 nm.
pendent of the concentration of nafarelin. In another incubation mixture containing S9 mix and 0.5 mg/ml of nafarelin but without yeast cells, Trp was also detected. The amount of Trp released was 17.7% of the total amount in nafarelin. Thus, the S9 mix alone was capable of degrading nafarelin to the extent that free Trp was released. To determine the mechanism of release of Trp, aliquots of the incubation mixture (after deproteination with ethanol) were analyzed on the HPLC at 225 nm. Two hours after incubation, the nafarelin peak was smaller (Fig. 4), whereas a new peak that coeluted with the 410 heptapeptide, Ser-Tyr-D-Nal-Leu-ArgPro-Gly-NH2 , appeared. Upon further incubation, another peak that coeluted with the 510 hexapeptide, Tyr-D-Nal-Leu-Arg-ProGly-NH2, appeared.
DISCUSSION In the course of performing a standard battery of mutagenicity tests with the decapeptide drug nafarelin, a positive result was obtained in the mitotic gene conversion test, but only in the presence of metabolic activation. Since true genetic toxicity seemed unlikely in a drug that differs from native LHRH at a single amino acid residue, we designed a series of experiments which ultimately demonstrated that the positive result may have been an artifact of the test system. Since Trp would not have been expected to induce gene conversion in yeast cells, these data supported the hypothesis that Trp was transferred to the minimal medium plates in sufficient quantities to support cell growth,
634
DEPASS
even when the cells were washed prior to plating. Although these results suggested that the presence of Trp in the incubation medium could have produced a false positive result in this assay, there was a need for direct evidence that metabolism of nafarelin to yield Trp actually occurred under these assay conditions. To address this issue, amino acid analysis was performed on 1%hr incubation mixtures. The data showed that Trp was released in the incubation mixture as a result of hydrolysis of nafarelin by the S9 mix. The results of the time-course experiment (Fig. 4) suggested that the incubation mixture contained chymotrypsin-type activity which preferentially cleaved nafarelin at the Trp-Ser bond to give the 4- 10 peptide and possibly the l-3 (pGlu-His-Trp) fragment as well. An endopeptidase in the incubation medium may then have cleaved the His-Trp bond in the l3 tripeptide to release Trp. It is noteworthy that His was not present in the 18-hr incubation medium, as shown in Fig. 3. The absence of His was probably due to the pyroglutamic acid residue bound to His. A peptide bond involving a pyroglutamic acid residue requires a specific enzyme, pyroglutamyl aminopeptidase, for its cleavage (Tsuru et al., 1984). This enzyme may not have been present in the S9 fraction, thereby leaving the pGlu-His bond intact. The absence of His may explain the negative result in the Ames activation assay, since hydrolysis of nafarelin to yield His may have resulted in another false positive result. Thyrotropin releasing hormone (pGlu-HisPro-NH*) was also negative in the Salmonella reverse mutation assay, perhaps for the same reason (Allen and Panfili, 1986).
ET
AL.
These results suggest the need for caution in interpreting data from tests such as the Ames and mitotic gene conversion assays performed on protein or peptide drugs. This problem may be avoided completely by simply not testing such compounds in assaysthat depend on the mutation or conversion of organisms to amino acid prototrophs. Although this approach seems reasonable, it may not be acceptable to some regulatory agencies that require a standard battery of mutagenicity tests on all new drug candidates. If positive results are obtained in such cases, they should not be viewed as positive evidence of mutagenicity, but rather as equivocal data, the significance of which requires further investigation. REFERENCES ALLEN, J. S., AND PANFILI, J. (1986). Ames Salmonella/ mammalian-microsome testing of peptides and peptide synthesis reagents. Mutation Rex 170,23-29. AMES, B. N., MCCANN, J., AND YAMASAKI, E. (1975). Methods for detecting carcinogens and mutagens with the Salmonella/mammalian microsome mutagenicity test. Mutation Rex 31, 347-364. CHANG, J., KNECHT, R., AND BRAWN, D. G. (1981). Amino acid analysis at the picomole level. Biochem. J. 199,.547-555.
CHANG, J.. KNECHT, R., AND BRAUN, D. G. (1982). A complete separation of dimethylaminoazobenzenesulfonyl-amino acids. Biochem. J. 203, 8034306. TSURU, D.. NAKAMURA, T.. YOSHIMOTO, T., AND FuJIWARA, K. (1984). Pyroglutamyl-peptidase from bacillus amyloliquefaciens. Biochem. Biophys. Acta 791, 117-122. VOGEL, H. J., AND BONNER. D. M. (1956). Acetylomithinase of E. coli: Partial purification and some prop erties. J. Biol. Chem. 218, 97-106. ZIMMERMAN, F. K., KERN, R., AND RASENBERGER,H. (1975). A yeast strain for simultaneous detection of induced mitotic crossing over, mitotic gene conversion and reverse mutation. Mutation Rex 28, 38 l-388. Z~RMAN, F. K. ( 1975). Procedures used in the induction of mitotic recombination and mutation in the yeastSaccharomyces cerevisiae. Mutation Res. 31,71-86.
AND
APPLIED
TOXICOLOGY
17,627-634
(1991)
False Positive Result for a Peptide Drug in the Gene Conversion with Saccharomyces cerevisiae Strain D7’
Assay
LINVAL R. DEPASS,* REBECCA L. CHAN,* D. R. JAGANNATH,~ AND IRWIN A. HEYMAN* *Syntex
Research,
Palo Alto,
California
Received
94304,
February
and THazleton
Biotechnologies,
12, 1991; accepted
May
Kensington,
Maryland
20895
29, 1991
False Positive Result for a Peptide Drug in the Gene Conversion Assay with Saccharomyces cerevisiae Strain D7. DEPASS, L. R., CHAN, R. L., JAGANNATH, D. R., AND HEYMAN, I. A. (199 1). Fundam. Appl. Toxicol. 17,627-634. A battery of mutagenicity tests was performed with nafarelin, an agonist analogue of luteinizing hormone releasing hormone (LHRH) containing tryptophan (Trp) and histidine (His). Included were the Ames assay and the gene conversion assay with yeast strain D7. Roth tests were negative without S9 activation, and the Ames test was negative with S9, but the yeast test was positive with S9 activation. Since the yeast test is based on conversion of cells to Trp independence, release of Trp by metabolism of the drug could account for the positive result. The test was repeated using Trp instead of the drug. The result was positive even at the lowest Trp concentration. In another experiment with the drug amino acid analysis of the incubation mixture revealed the presence of Trp but no detectable His. Since the Ames test is based on mutation to His-independent cells, these data are completely consistent with the negative result in the Ames test and the false positive result in the yeast test. These data suggest the need for caution in interpreting the results from mutagenicity assays with peptide drugs. 0 1991 society of Toxicology
The Ames Salmonella assay and the gene conversion assay with yeast cells have been used extensively in the testing of drugs and other chemicals for mutagenic activity (Ames et al., 1975; Zimmerman, 1975; Zimmerman et al., 1975). Whereas the Ames assay relies upon reversion to a histidine-requiring state, the yeast assay relies upon reversion to a tryptophan-requiring state. The results of these tests are generally acceptable to regulatory agencies as evidence of the presence or absence of genetic toxicity of a drug. These tests were performed as part of a standard mutagenicity battery to assess the potential genetic activity of nafhrelin, an agonistic analogue of luteinizing hormone-releasing hormone (LHRH). The structures of
native LHRH and nafarelin are presented in Fig. 1. The initial results indicated that nafarelin, [D-N~~(~)~]LHRH, was negative in both assays in the absence of metabolic activation and also negative in the Ames assay with activation. However, the result was positive in the yeast assay performed with rat liver S9 activation. Since the probability of true genetic toxicity seemed remote for a drug that differs from native LHRH at only one amino acid residue, a series of experiments were designed in an attempt to explain the “positive” result. We were particularly interested in the possibility that the positive result may have been an artifact of the test system caused by release of tryptophan (Trp) by hydrolysis of nafarelin. MATERIALS
’ Presented in part at the Annual Meeting of the Society of Toxicology, New Orleans, Louisiana March, 1986, and at the Annual Meeting of the Drug Safety Sub-Section of the Pharmaceutical Manufacturer’s Association, Nashville, Tennessee, November, 1986.
AND
METHODS
Chemicals. Nakuelin samples used in these experiments were synthesized and analyzed for purity by Syntex Re search. The sample used in the mutagenicity studies was 627
0272-0590191
$3.00
Copyri&t 0 1991 by the society of Toxicology. All rights of repmductioo in any form nserved.
DEPASS ET AL.
628
Determination of cytotoxicity and mutagenicity. To determine bacterial cytotoxicity and mutagenicity in the 1 2 3 4 5 6 7 8 9 10 Salmonella strains, 0.025 to 0.15 ml of a solution of the pGlu - HIS - Trp - Se; Tyr - Gly Leu - Arg Pm Gly NH2 test substance in DMSO was added to a sterile test tube containing 2.0 ml of overlay agar, 0.2 ml of an overnight (6) culture, and 0.5 ml of 0.2 M phosphate buffer, pH 7.4, in a 43 to 45°C water bath. This mixture was swirled gently 1 2 3 4 5 6 7 6 9 10 pGlu HIS Trp - Ser. Tyr - NH - 7” - CO - Leu Arg - Pro - Gly NH2 and then poured onto minimal agar plates. After the overlay agar had set, the plates were incubated at 37°C for CH2 approximately 2 days. For cytotoxicity, the number of ,a 0 0 colonies growing on the plates was counted and recorded. A reduction in the number of revertants, appearance of microcolonies, or clearing of the background lawn on the FIG. 1. Structures of (A) LHRH and (B) nafarelin ([Ddrug-treated plates as compared to the solvent control Nal6]LHRH), where D-Nal is 3-(2-naphthyl)-D-alanine. plates was considered as an indication of toxicity. For mutagenicity, the number of His+ revertant colonies growing on the plates was counted and recorded. 98.8% pure, whereas that which was used in the metabolism The activation assay was run concurrently with the studies was 97.2% pure. The positive control mutagens nonactivation assay.The only difference was the addition used in the mitotic gene conversion studies were ethyl- of 0.5 ml of S9 mix in place of 0.5 ml of phosphate buffer. methanesulfonate (EMS, Eastman Kodak, Rochester, NY) To determine cytotoxicity in the gene conversion assay, and sterigrnatocystin (Sigma Chemical Co., St. Louis, MO). 0.025 to 0.15 ml of a solution of the test substance in The positive control mutagens used in the Ames SalmoDMSO was added to a sterile 20-ml scintillation vial connella assayswere sodium azide, 2-nitrofluorene, 9-amitaining 0.3 ml of the indicator organism (1 to 2 X IO* noacridine, and 2-aminoanthracene obtained from Sigma. cells) and 2.65 ml of 0.2 M phosphate buffer, pH 7.4. This Stock solutions were prepared in dimethylsulfoxide mixture was incubated for approximately 18 hr at 30°C (DMSO, Sigma). in a rotary shaker. After incubation, cell survival was deDimethylaminoazobenzenesulfonyl chloride (DABStermined by adding 0.2 ml of 10e4 dilution of the suspenCl), used for the derivatization of amino acids, was ob- sion to 2 ml of molten overlay agar, which was then poured tained from AG Fluka, Switzerland. Amino acid standards, onto yeast complete plates. These plates were incubated NADP, and glucose 6-phosphate were purchased from for 2 to 3 days and scored for surviving population. Sigma. To determine gene conversion, the same procedure was Cell culture. The Salmonella typhimurium strains used followed except that after incubation, the surviving pop in these studies were originally obtained from Dr. Bruce ulation was estimated by plating an ahquot of a lO-4 diAmes, University of California at Berkeley. The following lution of each cell suspension in 0.15 M NaCl onto yeast strains were used: TA-1535, TA-1537, TA-1538, TA-98. complete plates (five plates per treatment). Mitotic gene and TA- 100. The Saccharomyces cerevisiae strain D7 used conversion was measured by plating 0.2 ml of undiluted in these assayswas originally constructed by Dr. F. K. samples onto minimal media supplemented with adenine Zimmerman (Zimmerman et al., 1975). The SalmoneZla and isoleucine (five plates per treatment level). The plates strains were grown overnight at 37°C in Oxoid Media No. were incubated at 30°C for 4 to 6 days and scored for 2 (nutrient broth). The minimal media plates for the se- survivors and convertants. lection of histidine revertants consisted of the Vogel Bonner The activation assay,containing 1.5 ml of S9 mix and Medium E (Vogel and Bonner, 1956) with 2% glucose and 1.15 ml of phosphate buffer (instead of 2.65 ml of buffer), 1.5% bactoagar. Log phase cultures of the yeast strain D7. was run concurrently with the nonactivation assay. grown in yeast complete broth, were washed twice in Amino acid analysis. Samples were derivatized with phosphate buffer and resuspended in the buffer at a density DABS-Cl and analyzed with a reversed-phase HPLC sysof 1 to 2 X lo* cells/ml and used in these studies. The tem according to a published procedure (Chang et al., 198 1, yeast complete media plates for scoring survivors and the 1982). The 18-hr incubation mixtures of nafarelin with yeast synthetic media plates without tryptophan for de- S9 mix (with or without yeast cells) were centrifuged. Altecting the convertants were prepared according to the iquots (30 ~1) of the supernatant were mixed with 20 ~1 methods of Zimmerman (1975). of a stock solution of phenylalanine (Phe, 300 nmol/ml). 100 ~1 of 0.2 M sodium bicarbonate buffer (pH 9.0), and Preparation of the rat liver homogenate reaction mixture 100 ~1 of DABS-Cl in acetone (6 nmol/rl). Phe was added (S9 mix). Aroclor 1254-induced rat liver homogenate (9000g soluble fraction, S9). purchased from Bionetics Lab as an internal standard for the quantitation of tryptophan Products (Charleston Heights, SC) was used as the exog- (Trp). The mixture was heated at 70°C for 10 min, dried enous source of mixed function oxidase enzyme system. under nitrogen, and redissolved in 400 pl of 50% ethano! The S9 mix was prepared in pH 7.4, 0.1 M sodium phos- in 0.04 M, pH 6.5. sodium phosphate. Aliquots (20 ~1) were analyzed on the HPLC column (detector at 444 nm) phate buffer containing 33 mM KCI, 8 mM MgC&, 5 mM within 1 hr after derivatization. glucose 6-phosphate. 4 mtvrNADP, and 10% (v/v) S9. (A)
FALSE POSITIVE TA-1535
0
GENE CONVERSION
TA-1537 -u--
1500
3000 Concentration
629
TA-100 . ..#--.
TA-1538 -*-
4500
DATA
6000
of Nafarelin
TA-98 + ..
7500
9000
10500
( pg/Plate)
FIG. 2. Results of the Ames activation assaywith nafarelin. Results were similar without activation.
The HPLC systemconsisted of a Varian LC 5060 liquid chromatograph equipped with a Varian UV-50 variable wavelength detector and a 5 pm Lichrosorb RP- 18 column (Rainin, Wobum, MA). A 3.2 mm X 5 cm precolumn packed with 37-53 Frn of pellicular silica gel (Whatman, Clifton, NJ) was placed before the injector as a saturation column. Solutions containing known quantities of Trp (O-6 nmol) were derivatized with DABS-Cl and analyzed on
the HPLC. A calibration curve was prepared from the Trp:Phe peak height ratios versus Trp concentrations. The amounts of Trp released were estimated from the Trp:Phe peak height ratios. Degradation ofnafarelin with S9 mix. A mixture of 2.7 ml of S9 mix. 0.3 ml of nafarelin stock solution (IO mg/ ml in DMSO), and 3 ml of 0.2 M sodium phosphate (pH 7.4) was incubated at 37°C. At various time points, 50-~1 aliquots of the incubation mixture were added to 400 pl
TABLE 1 RESULTS OF THE MITOTIC GENE CONVERSION ASSAY WITH NAFARELIN IN THE ABSENCE OF EXOGENOUS MAMMALIAN METABOLISM
Dose” DMSO EMS Nafarelin
100 &3 ml 1.O% 1.0 10.0 100.0 1000.0
5000.0 10000.0
Number of cells scored x104 811 317 778 931 843 837 553 526
’ Values are rg/ml unless otherwise indicated.
Percentage survivors
Total no. convertants
Convertants per 1OSsurvivors
100.0 39.1 95.9 114.8 103.9
345 1730 344 373 348 392 302 289
4.3 54.6 4.4 4.0 4.1
103.2
68.2 64.9
4.7
5.5 5.5
630
DEPASS TABLE
ET
AL 2
RESULTS OF THE MITOTIC GENE CONVERSION ASSAY WITH NAFARELIN IN THE PRESENCE OF EXOGENOUS MAMMALIAN METABOLISM
DMSO Sterigmatocystin Nafarelin
’ Values
are pg/ml
Dosea
Number of cells scored x104
100 pl/3 ml 5.0 1.0 10.0 100.0 1000.0 5000.0 10000.0
891 661 825 814 812 815 113 60
unless otherwise
The results of the Ames/Salmonella assay with nafarelin in the presence of S9 activation are presented in Fig. 2. The results demonstrated no increase in the number of revertants at any dose tested in the five strains. The results
TABLE
DMSO Sterigmatocystin Nafarelin
100 pl/3 ml 5.0 1.0 10.0 100.0 1000.0 5000.0 10000.0
a Values are &ml unless otherwise b Mean f SD of three experiments.
Total no. convertants
Convertants per lo5 survivors
402 1339 381 383 360 568 532 304
4.5 111.0 4.6 4.4 4.4 7.0 31.2 50.7
indicated.
RESULTS
Dose”
SUlViVOrS
100.0 14.2 92.6 98.1 91.1 91.5 12.7 6.1
of ethanol and centrifuged. Aliquots (40 ~1) of the supernatants were injected onto a 5-pm Rainin Microsorb Cl 8 column. The decrease of peak height of nafarelin on the HPLC chromatogram (detector at 225 nm) was a measure of its extent of degradation.
RESULTS OF THE MITOTIC OF EXOGENOUS
Percentage
of the Ames/Salmonella assay in the absence of S9 activation enzymes were similar and are, therefore, not presented. The results of the mitotic gene conversion assay without activation are presented in Table 1. There was no increase in number of convertants with increasing concentration of nafarelin. The initial results of the mitotic gene conversion assay with activation are presented in Table 2. The data show increased numbers of convertants per lo5 survivors at the three highest concentrations of nafarelin. The num-
3
GENE CONVERSION ASSAY WITH NAFARELIN ~MMALIAN METABOJJSM USING WASHED
IN THE PRESENCE CELL.Y
Number of cells scoredb x104
Percentage survived
Total no. convertants’
1118 + 36Ok 1013 + 975 f 959 + 83Ok 502 f 216+
100.0 32.1 -t 3.9 90.7 + 5.3 87.3 + 1.5 86.1 + 3.6 75.0 f 12.5 44.3 + 7.3 19.4 k 3.6
301 + 33 4469 z!z 936 312+ 34 302+ 41 300* 55 425 + 134 385 + 35 307 + 115
indicated.
164 81 143 136 108 118 139 29
Convertants per IO5 survivom~ 2.76 124.38 3.15 3.18 3.20 5.18 8.12 14.51
f + k f f + f +
0.13 4.77 0.80 0.91 1.00 1.76 2.52 6.42
FALSE
POSITIVE
GENE TABLE
RESULTS
OF THE MITOTIC OF EXOGENOUS
Dose= DMSO
Sterigmatocystin Tryptophan
50.0 &ml 5.0
1.0 10.0 100.0 250.0 500.0
1000.0 L?Values are &ml
CONVERSION 4
GENE CONVERSIONASSAY WITH TRY~~OPHAN MAMMALIAN METABOLISM USING WASHED Number of cells scored x104
830 730 785 754 754 791 767 728
631
DATA
Percentage
survivors 100.0 88.0 94.6 90.8 90.8 95.3 92.4 87.7
IN THE PRESENCE
CELLS
Total no. convertants
Convertants per IO5 survivors
151 1 1174
I .82 55.63 14.96 16.10 19.73 18.94 17.00 4.66
406
1214 1488 1498 1304 339
unless otherwise indicated.
ber of cells scored and the percentage survivors The results of the experiment with washed were greatly reduced at the two highest con- cells are presented in Table 3. Again, there centrations. The degree of cytotoxicity at high was evidence of toxicity to the indicator strain, concentrations was greater than that observed especially at the two highest concentrations. in the absence of the S9 fraction. Note also A dose-related increase in the convertant frethat the total number of convertants was not quency was observed at the three highest consignificantly increased compared to the solvent centrations of nafarelin. However, the increase control. Nevertheless, because of the positive above solvent control was smaller than that dose response over three drug concentrations, observed in the initial experiment in which the result was interpreted as evidence of ge- the cells were not washed. Again, there was netic toxicity of nafarelin. no significant increase in the absolute number Because nafarelin differs from native LHRH of convertants at any concentration compared at only a single amino acid residue, there was to the solvent control value. To confirm the no reason to expect that this compound would apparent positive result with the washedalter cellular DNA so as to produce intragenic cells, this experiment was repeated twice. The recombination (gene conversion). Instead, we results were similar with all three replicates hypothesized that growth of the yeast colonies (Table 3). on Trp-free medium may have occurred by In order to ascertain if tryptophan was reuptake and metabolic utilization of Trp re- sponsible for the observed increase in converleased by hydrolysis of nafarelin. Transfer of sion frequency, an additional experiment in the drug or its degradation products may have which Trp was substituted for nafarelin was occurred when the suspended yeast cells were conducted. Again, one portion of the cells was plated onto minimal media. washed three to four times in phosphate buffer We tested the aforementioned hypothesis before being plated onto minimal media. A by repeating the experiment with one change separate aliquot of cell suspension was plated in procedure. After the incubation, the treated without being washed to evaluate the effeccell suspension was washed three to four times tiveness of the washing procedure in removing by centrifuging and suspending the cell pellet Trp. The highest concentration of Trp (1.0 in saline or in phosphate buffer. The washed mg/ml) was chosen to be approximately equal cells were then plated as before, and survivors to that which would be expected if all of the and convertants were scored. Trp were released by metabolic activity of
632
DEPASS ET AL.
Standards
0.25 mg/ml
0.5 mg/ml
Phe TOP
Pyle
JL.AJ.L -P
h
0.75 mgiml
!?,t
1.0 mg/ml
L, Phe
n
l-a--J-\ 0
16
24
32
Min
-L-Y--k 8
16
24
32
Min
FIG. 3. Amino acid analysis of the 1%hr incubation mixtures of nafarelin, yeastcells, and S9 mix. Aliquots of the incubation mixtures containing various initial concentrations (0 to 1.0 mg/ml) of nafarelin were derivatized with DABS-Cl and analyzed by gradient elution on a LiChrosorb RP- 18 HPLC column as previously described (Chang et al., 198 1). Detector wavelength was 444 nm.
the S9 mix at the highest concentration of nafarelin used in the previous experiments (10 w/ml). The results of the experiment with Trp using washed cells are presented in Table 4. Increased numbers of “convertants” compared to the solvent control were observed even at the lowest concentration of 1.O pg/ml. The results with unwashed cells (not presented) were similar except that the cells formed a “lawn” at the highest concentration of Trp. In contrast, there was a reduction in number of “convertants” at the highest concentration
relative to that observed at lower concentrations, when the cells were washed. Aliquots of the 18-hr incubation mixtures containing 0 to 1.Omg/ml of nafarelin, S9 mix, and yeast cells were derivatized with DABSCl. The HPLC chromatograms of the derivatized mixtures are shown in Fig. 3. The amount of Trp liberated in the incubation mixture increased with increasing concentration of nafarelin. The quantities of Trp released were calculated from the Trp:Phe peak height ratios. The percentage of total Trp released ranged from 19 to 27% and was inde-
FALSE POSITIVE 0 Hr
GENE CONVERSION 0.25 Hr
!L
I 1. i A-
2 Hr sample spiked with 4-10 standard peptide
4-10
Naf
Naf
6
6
= Nafarelin
Min
L 2
Naf
5 Hr
S-10
4-10
4
2 Hr
Naf
Naf
2
633
DATA
4
Naf
6
6
Min
2
4
6
6
Min
RG. 4 A. time-course experiment showing formation of 4- 10 and 5-10 peptide fragments of nafarelin alter incubation with S9 mix. Aliquots of the incubation mixture were deproteinized with ethanol and centrifuged. Portions of the supematant were injected on a Microsorb C 18 HPLC column. The mobile phase contained 32% of acetonitrile and 68% of 0.1 M phosphate (pH 2.6). Detector wavelength was 225 nm.
pendent of the concentration of nafarelin. In another incubation mixture containing S9 mix and 0.5 mg/ml of nafarelin but without yeast cells, Trp was also detected. The amount of Trp released was 17.7% of the total amount in nafarelin. Thus, the S9 mix alone was capable of degrading nafarelin to the extent that free Trp was released. To determine the mechanism of release of Trp, aliquots of the incubation mixture (after deproteination with ethanol) were analyzed on the HPLC at 225 nm. Two hours after incubation, the nafarelin peak was smaller (Fig. 4), whereas a new peak that coeluted with the 410 heptapeptide, Ser-Tyr-D-Nal-Leu-ArgPro-Gly-NH2 , appeared. Upon further incubation, another peak that coeluted with the 510 hexapeptide, Tyr-D-Nal-Leu-Arg-ProGly-NH2, appeared.
DISCUSSION In the course of performing a standard battery of mutagenicity tests with the decapeptide drug nafarelin, a positive result was obtained in the mitotic gene conversion test, but only in the presence of metabolic activation. Since true genetic toxicity seemed unlikely in a drug that differs from native LHRH at a single amino acid residue, we designed a series of experiments which ultimately demonstrated that the positive result may have been an artifact of the test system. Since Trp would not have been expected to induce gene conversion in yeast cells, these data supported the hypothesis that Trp was transferred to the minimal medium plates in sufficient quantities to support cell growth,
634
DEPASS
even when the cells were washed prior to plating. Although these results suggested that the presence of Trp in the incubation medium could have produced a false positive result in this assay, there was a need for direct evidence that metabolism of nafarelin to yield Trp actually occurred under these assay conditions. To address this issue, amino acid analysis was performed on 1%hr incubation mixtures. The data showed that Trp was released in the incubation mixture as a result of hydrolysis of nafarelin by the S9 mix. The results of the time-course experiment (Fig. 4) suggested that the incubation mixture contained chymotrypsin-type activity which preferentially cleaved nafarelin at the Trp-Ser bond to give the 4- 10 peptide and possibly the l-3 (pGlu-His-Trp) fragment as well. An endopeptidase in the incubation medium may then have cleaved the His-Trp bond in the l3 tripeptide to release Trp. It is noteworthy that His was not present in the 18-hr incubation medium, as shown in Fig. 3. The absence of His was probably due to the pyroglutamic acid residue bound to His. A peptide bond involving a pyroglutamic acid residue requires a specific enzyme, pyroglutamyl aminopeptidase, for its cleavage (Tsuru et al., 1984). This enzyme may not have been present in the S9 fraction, thereby leaving the pGlu-His bond intact. The absence of His may explain the negative result in the Ames activation assay, since hydrolysis of nafarelin to yield His may have resulted in another false positive result. Thyrotropin releasing hormone (pGlu-HisPro-NH*) was also negative in the Salmonella reverse mutation assay, perhaps for the same reason (Allen and Panfili, 1986).
ET
AL.
These results suggest the need for caution in interpreting data from tests such as the Ames and mitotic gene conversion assays performed on protein or peptide drugs. This problem may be avoided completely by simply not testing such compounds in assaysthat depend on the mutation or conversion of organisms to amino acid prototrophs. Although this approach seems reasonable, it may not be acceptable to some regulatory agencies that require a standard battery of mutagenicity tests on all new drug candidates. If positive results are obtained in such cases, they should not be viewed as positive evidence of mutagenicity, but rather as equivocal data, the significance of which requires further investigation. REFERENCES ALLEN, J. S., AND PANFILI, J. (1986). Ames Salmonella/ mammalian-microsome testing of peptides and peptide synthesis reagents. Mutation Rex 170,23-29. AMES, B. N., MCCANN, J., AND YAMASAKI, E. (1975). Methods for detecting carcinogens and mutagens with the Salmonella/mammalian microsome mutagenicity test. Mutation Rex 31, 347-364. CHANG, J., KNECHT, R., AND BRAWN, D. G. (1981). Amino acid analysis at the picomole level. Biochem. J. 199,.547-555.
CHANG, J.. KNECHT, R., AND BRAUN, D. G. (1982). A complete separation of dimethylaminoazobenzenesulfonyl-amino acids. Biochem. J. 203, 8034306. TSURU, D.. NAKAMURA, T.. YOSHIMOTO, T., AND FuJIWARA, K. (1984). Pyroglutamyl-peptidase from bacillus amyloliquefaciens. Biochem. Biophys. Acta 791, 117-122. VOGEL, H. J., AND BONNER. D. M. (1956). Acetylomithinase of E. coli: Partial purification and some prop erties. J. Biol. Chem. 218, 97-106. ZIMMERMAN, F. K., KERN, R., AND RASENBERGER,H. (1975). A yeast strain for simultaneous detection of induced mitotic crossing over, mitotic gene conversion and reverse mutation. Mutation Rex 28, 38 l-388. Z~RMAN, F. K. ( 1975). Procedures used in the induction of mitotic recombination and mutation in the yeastSaccharomyces cerevisiae. Mutation Res. 31,71-86.