Interlaboratory comparison: liver spontaneous mutant frequency from lambda/lacI transgenic mice (Big Blue) (II)

Fundamental and Molecular Mechanisms of Mutagenesis

ELSEVIER

Mutation

Research

327 (1995) 67-73

Interlaboratory comparison: liver spontaneous mutant frequency from lambda/k& transgenic mice (Big Blue@) (II) Robert

R. Young

a, Brenda

J. Rogers b,*, G. Scott Provost Donald L. Putman a

b, Jay M. Short

‘,

a Microbiological Associates, Inc., 9900 Blackwell Road, Rockville, MD 20850, USA b Stratagene, 11011 N. Torrey Pines Road, La Jolla, CA 92037, USA ’ Industrial BioCatalysis Inc., 505 Coast Blvd., La Jolla, CA, USA Received

24 May 1994; revision

received 15 August 1994; accepted

19 August

1994

Abstract Spontaneous mutant frequency in livers of two transgenic mouse strains, each carrying identical lambda shuttle vectors with a lad target gene, was evaluated by two laboratories. These studies investigated variability in spontaneous mutant frequency between animals and as a function of the number of phage screened. Liver DNA was independently isolated from 7-11 week old phage were screened for mutation at lacl for C57BL/6 liver spontaneous mutant frequency mutant frequency was 42 + 10 X 10e6 at one

C57BL/6 and B6C3Fl Big Blue@ transgenic mice. At least 500000 each animal using standardized assay procedures. In the two labs, the was 45 & 9 X 10m6 and 41 + 7 X 10e6. The B6C3Fl liver spontaneous lab and 43 f 12 X 10m6 and 41+ 8 X 10e6 in two trials at the second

lab. Mean mutant frequency data from both labs, calculated in increments of 100000 plaque forming units (pfu) scored for each mouse strain, show stabilized mean mutant frequency and standard deviation after approximately 200000-300000 pfu screened. The frequency of spontaneous lacl mutants was reproducible both within and between Keywords:

labs and was comparable Transgenic

mouse;

between

Mutation

the two transgenic

assay in vivo; Mutation

1. Introduction In vivo transgenic mutation assays are a new tool for investigating mutation within laboratory rodents following exposure to chemical, physical or biological agents. Several transgenic mouse lines have been recently developed for mutagenesis studies which utilize different mutation target genes carried within recoverable lambda shuttle

’ Corresponding 0071.

author. Tel. (619) 535-5400; Fax (619) 535-

0027-5107/95/$09.50 0 1995 Elsevier SSDI 0027-5107(94)00080-8

Science

mouse strains. analysis;

Spontaneous

mutation

vectors, such as 1uc.Z (Short et al., 1988; Gossen et al., 1989; Kohler et al., 1990) or fuel (Kohler et al., 1991a, b). In each case, transgenes are latent targets within each cell of the mouse that record mutagenic events in vivo that can subsequently be recovered and assayed in vitro for the presence of mutant forms of the target gene. We have investigated assay variability in the Big Blue@ Transgenic Mouse Mutagenesis Assay (Stratagene, La Jolla, CA) that has been developed for tissue-specific mutagenicity testing in vivo (Kohler et al., 1991a). Both C57BL/6 and

B.V. All rights reserved

68

R.R. Young et al. /Mutation

B6C3Fl transgenic mouse strains are available that contain approximately 40 concatenated lambda ZAP-like shuttle vectors, each of which contains one copy each of an E. coli lacl target gene and an a-1acZ reporter gene. Studies were designed to provide baseline information on variability and reproducibility of the Big Blue” Mutagenesis Assay. In a companion report, sources of variation in mutant phenotype detection are identified, and standardized study procedures recommended (Rogers et al., 1995). In this work, 10 different untreated C57BL/6 and B6C3Fl Big Blue@ mice were used at Microbiological Associates, Inc. (MA) and Stratagene (SCS) to measure animal to animal, strain to strain and laboratory to laboratory variation in liver spontaneous mutant frequency.

2. Materials

and methods

Transgenic animals Male Big Blue@ transgenic mice were received from Taconic Laboratory Animals and Services, Germantown, NY. Strains of Big Bluea mice used were the inbred C57BL/6 (Lambda LIZ: C57BL/6NTac[LIZ]) and the hybrid B6C3Fl (Lambda LIZ: C57BL/6NTac[LIZ] females X C3H/HeNTac males). Both strains of mice were confirmed to be hemizygous for the transgene through slot-blot hybridization analysis of tail genomic DNA using a 32P-labeled lambda DNA probe. The livers were collected from 7-11 week old mice, immediately frozen in liquid nitrogen and stored at 5 -70°C until analyzed. DNA isolation and vector packaging Genomic DNA was isolated using the methods described by Kohler et al. (1991a) with improvements recommended by Stratagene (Stratagene, 1992; Rogers et al, 1995). Samples of about 100 mg were taken randomly from frozen livers. Tissue was disaggregated in a dounce homogenizer in douncing buffer, digested in proteinase K and RNAse, extracted with phenol/chloroform, and precipitated with ethanol. DNA was dissolved in pH 7.5 TE buffer, diluted to a concentration of approximately 0.5~. DNA/p1 and stored at 4°C protected from light. TranspackTM in vitro pack-

Research

327 (1995) 67-73

aging extract (Stratagene) was used to recover the lambda shuttle vector from genomic DNA and to package the vector into empty phage particles creating infective phage. The packaging reaction was stopped by addition of SM buffer. Viable phage were either immediately screened in the mutation assay or 50 ~1 chloroform added and phage stored at 4°C for up to 5 days for later evaluation. h4uta tion assay At MA, packaged phage were adsorbed onto E. coli SCS-8 cells (Stratagene), mixed with top agarose (NZY medium with 0.7% SeaChem LE agarose, FMC Bioproducts, Rockland, ME) containing 2 mg 5-bromo-4-chloro-3-indolyl-/I-ogalactopyranoside (X-gal; Gold Biotechnology, Inc., St. Louis, MO) per ml and plated onto 25 x 25 cm square bioassay trays containing 200 ml NZY bottom agar (NZY medium with 1.5% agar). Similar procedures were used at Stratagene with the exception that top agarose was prepared with 0.7% Low EEO Agarose (Stratagene) and 1.5 mg/ml X-gal (Stratagene) and bioassay trays contained 250 ml NZY bottom agar. At Stratagene, the adsorbed phage mixture was sampled, diluted, and plated in duplicate or triplicate as plating efficiency plates to determine the total number of phage plated per packaging reaction kit. The number of plates per packaging reaction was adjusted to give a target density of 15000 plaques per plate (24 plaques/cm2). All experiments included control plates using a series of IacZ mutants that produce a spectrum of color intensity from very faint (CM-O) to intense dark blue (CM-3). All plates were incubated inverted for 16-18 h at 37°C. After incubation, plates were visually screened for blue, mutant plaques using a red enhancement screening filter. At MA, the number of plaques on each plate was determined by counting three representative areas, each equal to l/100 of the entire plate. At Stratagene, the total number of plaques was counted on plating efficiency plates to calculate total plaques screened. Putative mutant plaques were cored and replated with E. coli SCS-8 on X-gal containing medium to confirm phenotypic stability of the mutant.

R.R. Young et al./Mutation Research 327 (1995) 67-73

Mutant frequency for each packaging reaction was calculated by dividing the number of confirmed mutants (blue plaques) by the estimated total clear plaques. Samples were packaged and evaluated until at least 500000 plaques had been screened from the liver of each animal.

3. Results

69

spontaneous mutant frequency calculations were based on screening at least 500000 plaques per animal. Spontaneous mutant frequency data are summarized for both strains and both testing laboratories in Table 1. The spontaneous mutant frequency of the lacl target gene in liver DNA from Big Blue@ transgenic mice was between 41 x lop6 and 45 x lO-‘j for groups of nine or 10 animals screened to over 500000 plaques each. The observed 1acZ mutant frequencies were the same at the two labs for both strains.

Interlab and inter-strain comparison of liver spontaneous mutant frequency C57BL/6 and B6C3Fl Big Blue@ mice were screened for spontaneous IacZ mutations in the

Animal to animal variation in spontaneous mutant frequency

liver. Each laboratory evaluated liver DNA from nine or 10 different mice for each strain. Mean

Spontaneous mutant frequencies are reported by individual animal for each strain and labora-

A

E

5

E

s

40

4 t emm

h

z

2

40.

P

E

mm

3

n Y

llMA

IZMA

13MA 14MA 15MA 16MA 17MA l*MA

19MA 20MA

0 MEAN MF

ISCS PSCS sscs ‘scs

sscs sscs

7scs

sscs9scs1oscs

Animal Number

Animal Number

Animal Number

Animal Number

MEAN MF

Fig. 1. Spontaneous mutant frequency in livers from untreated 8-11 week old C57BL/6 and B6C3Fl male Big Bluem mice using standardized assay conditions. Nine or 10 different mice of each strain were analyzed by Microbiological Associates, Inc. (MA) and Stratagene (SCS). Bars represent final mutant frequency (X 10V6) for each animal after screening more than 500000 phage per animal. Mean MF represents mean mutant frequency for each group of animals. (A) C57BL/6 at MA. (B) C57BL/6 at SCS. (C) B6C3Fl at MA. (D) B6C3Fl at SCS. Results of two separate analyses of same DNA preparations are shown for B6C3Fl mice at scs.

R.R. Young et al. /Mutation Research 327 (1995) 67-73

70 Table 1 lnterlab and interstrain tant frequency Strain

a

Laboratory

comparison

of liver spontaneous

b

mu-

Mutant frequency (mean f SD) ’

C57BL/6

MA (10 mice) SCS (9 mice)

45+9x10-” 41*7x10-6

B6C3Fl

MA (10 mice) SCS trial 3 (10 mice) SCS trial 4 (10 mice)

42+ 10x lo-” 43+12x10-” 41*8x10-6

mice. Untreated animals 8-11 a Big Blue@ male transgenic weeks old. h Lab identification: MA. Microbiological Associates, Inc.; SCS, Stratagene. ’ Mean and standard deviation of mutant frequency from nine or 10 animals screened to greater than 500000 phage each.

tory in Fig. 1. Spontaneous mutant frequencies in C57BL/6 male mice liver ranged from 31 X 10d6 to 61 x 1O-7 for 10 mice at MA (Fig. 1A) and from 34 x 10e6 to 52 x 10e6 for nine different mice at SCS (Fig. 1B). The mean and standard deviation of the spontaneous mutant frequency in C57BL/6 liver DNA was 45 & 9 X lop6 at MA and 41 f 7 X 10e6 at SCS. In 10 male B6C3Fl mice, the liver spontaneous mutant frequency ranged from 22 x lOAh to 56 x lop6 at MA (Fig. 10 and from 29 X lo-” to 67 x 10e6 and 28 x 10e6 to 53 X 10e6 in two trials of the same DNA at SCS (Fig. lD>. The mean and standard deviation of the spontaneous mutant frequency in B6C3Fl mice was 42 * 10 X lop6 at MA and 43 + 12 x lo-” and 41 + 8 X 10-h in the two trials at SCS. Reproducibility of mutation data

The same liver DNA from 10 B6C3Fl male Big Blue@ mice was evaluated in two independent repeats to determine reproducibility of the mutation assay (Fig. 1D). The spontaneous mutant frequency from the 10 animals was 43 + 12 x lop6 initially and was 41 + 8 X 10e6 in the repeat trial. Effect of number of plaques screened on mutant frequency

The effect of the number of phage screened on mutant frequency was investigated by determin-

ing the mutant frequency of each animal after increments of 100 000 plaques had been screened, up to 500000 plaques per animal. Mutant frequencies from both laboratories were combined for each strain of mouse. Mean mutant frequency and standard deviation of the mean were plotted as increments of 100000 phage screened for C57BL/6 mice (Fig. 2A) and B6C3Fl mice (Fig. 2B). For both strains, mean mutant frequency remained stable between 100000 and 500 000 phage screened per animal. The standard deviation stabilized after approximately 200 OOO300 000 plaques were screened. Confirmation of mutant phenotype

All blue mutant plaques and suspect blue areas from B6C3Fl experiments at MA were cored, eluted, sterilized with chloroform and replated with X-gal to confirm the phenotypic stability of the blue plaque coIor. Of 248 blue plaques considered initially to be mutants, 246 maintained the mutant phenotype upon confirmation plating. Two questionable blue areas not originally counted as mutants were also cored and plated for confirmation. One questionable blue area did confirm as a mutant of very weak intensity and one blue area did not confirm as a mutant. 4. Discussion We have reported on studies to investigate sources of variability in the lambda 1acZ transgenie mouse mutagenesis assay. These studies are part of an ongoing project to develop and validate an in vivo transgenic mutation assay. This work has also contributed to two of the goals of a series of workshops on transgenic assay statistics sponsored by Procter & Gamble: to develop standard protocols for the transgenic mutation assays and to use the standardized protocol to generate a large data base of mutant frequencies in liver DNA from untreated transgenic mice for statistical study and analysis (Gorelick and Thompson, 1994). This two stage project first developed a standardized study protocol using optimized procedures and then used the standardized protocol to measure liver spontaneous mutant frequencies in

R.R. Young et al. /Mutation

01

1

I

100

200

Plaques Screened

1

300

I

I

400

500

Per Animal (thousands)

B

T

,-

I_

I

I

I

1

I

100

200

300

400

500

Plaques Screened

Per Animal (thousands)

Fig. 2. The effect of number of phage screened per animal on observed spontaneous mutant frequency. At least 500000 phage were screened per animal. The mutant frequency was calculated at each 100000 phage increment for each animal. The mean mutant frequency and standard deviation of mean was calculated for all animals at each 100000 phage increment. (A) Data combined from 19 C57BL/6 mice: 10 from MA and nine from SCS. (B) Data combined from 20 B6C3Fl mice: 10 from MA and 10 from SCS (trial 4).

transgenic mice. For the initial phase, the in vitro procedures used to isolate DNA, recover vectors, package and screen for mutant phage were studied. The multi-step procedure for isolating and screening transgenes for mutants was shown to

Research 327 (1995) 67-73

71

cause assay variability if certain variables were not controlled for. An optimized study protocol was developed and has been reported separately (Rogers et al., 1995). The recommended procedure monitors the sensitivity of the assay for the detection of Zucl mutants, minimizes sources of technical variability, and provides for internal color control mutants to standardize assay performance between trials and laboratories. In this work we have used the standardized protocol to measure the spontaneous mutant frequency in livers of two transgenic mouse strains, each carrying identical IucI target genes in lambda shuttle vectors. Parallel studies were performed at the two laboratories with each laboratory using different animals. Spontaneous mutant frequency data have been collected and reported to allow comparison between strains, animals and laboratories. The data also permit analysis of how the number of plaques screened per animal affects observed mutant frequency. The results showed that there was little difference in spontaneous mutant frequencies in livers of both C57BL/6 and B6C3Fl strains of Big Blue@ mice with mutant frequencies ranging between 41 X 10M6 and 45 X 10e6. Other authors have reported C57BL/6 Big Blue@ male liver spontaneous mutant frequencies of 19 X 1O-6 (Kohler et al., 1991b), 27 X lop6 (Stephard et al., 1993) and from 28 + 11 X lop6 to 48 k 31 x 1O-6 (Mirsalis et al., 1993). B6C3Fl male liver spontaneous mutant frequencies have been reported to be 29 f 18 X lop6 to 46 & 21 X 10e6 (Mirsalis et al., 1993). The range of values reported may represent true variation in spontaneous mutant frequency but also may result from evaluating small numbers of either animals or plaques or performing the assay without optimized assay conditions. Work reported by Rogers et al. (1995) demonstrated how lacl mutants with low levels of P-galactosidase production could be missed under suboptimal screening conditions. When different animals were evaluated at MA and SCS, similar patterns of animal to animal variation in spontaneous mutant frequency were observed in both strains. Maximum variation in mutant frequency between 1.6-fold and 2.2-fold for animals in the various groups. Once optimum

72

R.R. Young et al. /Mutation

assay procedures are implemented in a laboratory and sufficient phage are screened for each tissue, inherent variation in spontaneous mutant frequency between Big Blue@ mice may become the major source of assay variability. This is consistent with findings by Piegorsch et al. (1994). Control of experimental variables is required to obtain reproducible results with mutation assays. Individual animal and group mutant frequencies were in close agreement between the two replicate trials performed on the same B6C3Fl liver DNA within a single lab. These results demonstrate that with standardized procedures the assay can reproducibly detect mutant fuel phage in DNA samples and that variability due to technical variables can be successfully controlled. The number of animals analyzed and the number of phage screened per tissue are two important variables in the design and cost of a transgenie assay. When the mutant frequency data were combined from both labs in increments of 100000 phage screened, a similar pattern was observed in both the C57BL/6 and B6C3Fl data sets. Mean mutant frequency remained stable at slightly higher than 40 X lo-” when between 100000 and 500 000 phage were screened per animal. The standard deviation decreased as the number of phage screened increased. The standard deviation stabilized after 200 000-300 000 phage had been screened. This suggests that no appreciable improvement in the determination of the mutant frequency could be expected by screening more than 200000-300 000 phage per animal. Callahan and Short (1995) have reported that mean mutant frequency of the B6C3Fl liver data from Stratagene stabilized after 300000 phage had been screened per animal. Assuming one treated and one control group, 300000 plaques screened per tissue and equal variance between groups of animals, a doubling in mutant frequency could be detected with five B6C3Fl mice (Callahan and Short, 1995). Additional sensitivity to detect less than a doubling in mutant frequency, with other factors remaining equal, would require additional animals per treatment group. Over 99% (246/248) of the blue plaques ini-

Research 327 (1995) 67-73

tially scored as mutants produced blue plaques upon retesting and were confirmed as mutants. Two faint blue plaques originally counted as mutants did not confirm for unknown reasons. The one questionable blue area that did confirm as a mutant was cored from a diffuse blue area without obvious plaque morphology that was in a rivulet of condensation. The questionable blue area that did not confirm as a mutant was recorded as being a suspect bacterial colony. The results underscore the potential to occasionally miss weak (light blue) mutants under certain conditions or to score non-mutant plaques as mutants. Overall, there was a high degree of accuracy in visually scoring plates for mutant colonies. In subsequent work with C57BL/6 mice, only light blue putative mutants or questionable areas were plated out for confirmation. When similar experimental procedures were used, nearly identical spontaneous mutant frequencies were obtained at MA and Stratagene in both C57BL/6 and B6C3Fl male transgenic mice in the lambda/lucZ mouse mutagenesis assay. Companion work by Rogers et al. (1995) demonstrated the need to optimize technical procedures to obtain reproducible spontaneous mutant frequencies upon repeat screening of the same DNA. Our work demonstrated that once optimized technical procedures were used, identical results could be obtained in different laboratories for both strains of transgenic mice. There appears to be no observable difference in spontaneous mutant frequency of the ZucZ target gene in the two strains. Acknowledgements This work was supported by the National Institutes of Environmental Health Sciences under Contract No. NOl-ES-95252. We thank A. Penn Ritter for her expert technical assistance with the mutagenicity assays. References Callahan, J. and J.M. Short (1995) Statistical analysis transgenic lambda/lacl mutagenicity testing assay, mitted.

of a Sub-

R.R. Young et al. /Mutation Research 327 (1995) 67-73 Gorelick, N.J. and E.D. Thompson (1994) Overview of the workshops on statistical analysis of mutation data from transgenic mice, Environ. Mol. Mutagen., 23, 12-16. Gossen, J.A., W.J.F. de Leeuw, C.H.T. Tan, E.C. Zwarthoff, F. Berends, P.H.M. Lehman, D.L. Knook and J. Vijg (1989) Efficient rescue of integrated shuttle vectors from transgenic mice: a model for studying mutations in vivo, Proc. Nat]. Acad. Sci. USA, 86, 7971-7975. Kohler, S.W., G.S. Provost, P.L. Kretz, M.J. Dycaico, J.A. Sorge and J.M. Short (1990) Development of a short term in vivo mutagenesis assay: The effects of methylation on recovery of a lambda phage shuttle vector from transgenic mice, Nucleic Acids Res., 18, 3007-3013. Kohler, S.W., G.S. Provost, A. Fieck, P.L. Kretz, W.O. Bullock, J.A. Sorge, D.L. Putman and J.M. Short (1991a) Spectra of spontaneous and mutagen-induced mutations in the lucl gene in transgenic mice, Proc. Natl. Acad. Sci. USA, 88, 7958-7962. Kohler, S.W., G.S. Provost, A. Fieck, P.L. Kretz, W.O. Bullock, D.L. Putman, J.A. Sorge and J.M. Short (1991b) Analysis of spontaneous and induced mutations in transgenie mice using a lambda ZAP/l& shuttle vector, Environ. Mol. Mutagen., 18, 316-321. Mirsalis, J.C., G.S. Provost, C.D. Matthews, R.T. Hamner,

73

J.E. Schindler, K.G. O’Loughlin, J.T. MacGregor and J.M. Short (1993) Introduction of hepatic mutations in lacl transgenic mice, Mutagenesis, 8, 265-271. Piegorsch, W.W., A.C. Lockhart, B.M. Margolin, K.R. Tindall, N.J. Gorelick, J.M. Short, G.J. Carr, E.D. Thompson and M.D. Shelby (1994) Sources of variability in data from a lad transgenic mouse mutation assay, Environ. Mol. Mutagen., 23, 17-31. Rogers, B.J., G.S. Provost, R.R. Young, D.L. Putman and J.M. Short (1995) Intralaboratory optimization and standardization of mutant screening conditions used for a lambda/ lacl transgenic mouse mutagenesis assay. (I). Mutation Res., 327, 57-66. Shephard, S.E., C. Sengstag, W.K. Lutz and C. Schlatter (1993) Mutations in liver DNA of lacl transgenic mice (Big Blue) following subchronic exposure to 2-acetylaminofluorene, Mutation Res., 302, 91-96. Short, J.M., SW. Kohler, W.D. Huse and J. Sorge (1988) A transgenic model for the identification of genetic lesions, Fed. Proc., 8515a. Stratagene, La Jolla, CA (1992) Instruction Manual, Big Blue@ Transgenic Mouse Mutagenesis Assay System, Revision August 15, 38 pages.