On the origin of unusual mutations recovered from the lacI transgenic assay

Mutation Research 536 (2003) 1–6

On the origin of unusual mutations recovered from the lacI transgenic assay Johan G. de Boer∗ Centre for Biomedical Research, University of Victoria, P.O. Box 3020 STN CSC, Victoria BC, Canada V8W 3N5 Received 14 August 2002; received in revised form 3 December 2002; accepted 12 December 2002

Abstract Amongst approximately 25,000 mutants recovered from tissues of the lacI mouse and rat transgenic mutation assay, we identified seven mutants that carry changes that are unlike the majority of mutations that are normally recovered in these systems. The recovered mutants feature replacements and insertions of sequences that originate in the animal’s genome, in the bacteriophage lambda construct that harbors the lacI gene, and in the genome of the E. coli plating host. These mutants demonstrate that mutations resulting from diverse mechanisms, in addition to the normal point mutations, can be recovered. In addition, the data indicate that such mutations may often not be of animal origin. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Mutation mechanisms; lacI; Transgenic

1. Introduction The lacI transgenic assay is used to study mutation in diverse tissues of mice and rats. The recovery of these mutants allows us to examine mutational events and their mechanisms in animals. The assay can recover a wide range of mutation types, however, they are generally limited to small sequence alterations, such as point mutations, and small deletions and insertions. This limitation to small events is due to the nature of the reporter construct and the requirement of packaging a viable length bacteriophage lambda genome. In the course of sequencing many thousands of lacI genes from recovered mutants we discovered a number of mutants that have sequence alterations unlike the majority of recovered mutations. Previously, we reported the recovery of a mutant that featured an ∗ Tel.: +1-250-472-4079; fax: +1-250-472-4075. E-mail address: [email protected] (J.G. de Boer).

insertion of a rat B2 transposable sequence element into a lacI gene [1]. This finding demonstrated that the assay is capable of detecting unusual mutational events. Here we report several additional mutants in which DNA sequences from “elsewhere” have been inserted into a lacI gene sequence. The mutants harbor sequences that are derived from the animal genome, from the bacteriophage lambda genome, and from the SCS-8 bacterial host strain. 2. Materials and methods The mutants have been recovered during the course of various experiments using Big Blue lacI transgenic mice and rats, as well as the Rat2 cell line. DNA was isolated from tissues according to published protocols [2] and mutants recovered after packaging (Transpack, Stratagene) and plating on SCS-8 bacteria. PCR amplification of a lacI-containing fragment was done using the p1 and p8 primers, located at position −234

1383-5718/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S1383-5718(03)00013-5

2

J.G. de Boer / Mutation Research 536 (2003) 1–6

to −215, and 1354–1337, respectively, or with p2 and p9, located at −53 to −37, and 1187–1208, respectively. DNA sequencing was performed using Pharmacia ALF or LiCOR automated DNA sequencers. 3. Results The mutants that are reported here were recovered from lacI transgenic mice, rats, and the Rat2 cell line. They are listed in Table 1 and described further in detail. 3.1. Mutant A This mutant was described previously [1]. A modified rat B2 repeated sequence element was found inserted between position 77 and 78, in a mutant recovered from the rat cell line. This was the first demonstration of a transposable sequence element jumping into a mutation target gene. 3.2. Mutant B The PCR product of this mutant was approximately 1230 bp longer than that seen for normal lacI mutants. Sequencing of the mutant revealed an insertion between position 537 and 538, which duplicated the sequence from 537 to 546. The result is that the inserted sequence is now flanked on both ends by these 10 bp (5 -GCGTGGAGCA-3 ). Approximately, 160 bp from each end of the insert were sequenced and were found to be identical to the ends of the published IS10 sequence from E. coli [3], and has six base differences in the 3 -end with the one published by Lawley et al. Table 1 Mutants recovered from lacI transgenic mice, rats, and the Rat2 cell line Mutant

lacI system

Deleted (bp)

A

Rat2 cell line Mice Mice Rat Rat Mice Rat

–

135

Rat B2 element [1]

– 491 387 8 1100 2

∼1230 214 273 16 74 3

E. coli IS10 Lambda 5 arm Lambda 3 arm E. coli GlyS E. coli (?) Rat, activin C gene

B C D E F G

Inserted (bp)

Origin

[4] for Salmonella. The length of the insertion is consistent with an IS10 element. The consensus sequence for IS10 insertion sites is 5 -GCTNAGC-3 [5]. The insertion site sequence in lacI mutant B has only a vague resemblance to the consensus sequence, similar to the supF gene where IS10 insertions were found at sites that are also not very homologous to the consensus sequence [6]. 3.3. Mutants C and D Two mutants were recovered that feature a replacement of several hundred nucleotides with a sequence derived from bacteriophage lambda (Fig. 1). In mutant C, recovered from mouse, 491 bp, from position 301 to 791, were replaced with 219 bp derived from position 15 to 233 in the bacteriophage lambda genome, immediately following the COS site sequence. In mutant D, recovered from the rat 387 bp, from position 181 to 567, were replaced with 273 bp derived (in reverse orientation) from position 45,575 to 45,339 in lambda. The breakpoint sequences are shown in Fig. 1. Interestingly, both the 5 and the 3 breakpoints have similarities between the two mutants. Although there is only a single guanine nucleotide homology between the 5 endpoints of the lacI and the lambda sequence in both mutants, those endpoints in the lacI gene share the (shifted) consensus sequence 5 -CAACYGNGTGNCA-3 . This consensus sequence (in either orientation) only occurs at these two sites in the lacI gene. Also, at the corresponding breakpoints in lambda, the sequence 5 -GGTTTTC-3 was found in both mutants. The exact breakpoints, however, are different in relation to these consensus sequences. This indicates that the recombination step involves a certain sequence in lacI (5 -CAACYGNGTGNCA-3 ) and a certain sequence in lambda (5 -GGTTTTC-3 ). The sequence 5- GGTTTTC-3 is also found in the lacI at position 701–707, while the 5 -GAAAACC-3 sequence is found twice after the lacI stop codon, at position 1285–1291 and beyond 1350, outside of the amplified region. No deletions internal to lacI have ever been found that range from these positions to either 301 or 181. A single deletion was detected that ranged from position 181 to 646 inclusive, but the sequence at position 646 bears no resemblance to GGTTTTC, and has only a CG homology with position 181. The downstream endpoints in the lacI

J.G. de Boer / Mutation Research 536 (2003) 1–6

3

Fig. 1. (a) Recombination events between lacI and bacteriophage lambda sequences in mutants C and D. Breakpoint sequences are aligned. Underlined sequences are present in the mutant. (b) Consensus sequences for the breakpoints. Underlined sequences are present in the mutant while sequence homology is in bold.

sequence (position 791 and 567, respectively) have some sequence homology; mutant C has a 5 -GCG-3 sequence homology between lacI and lambda, while mutant D has the sequence 5 -GCAXATCG-3 shared between lacI and lambda. Both mutants C and D share 5 -GC-3 at these breakpoints.

3.4. Mutant E Mutant E featured a replacement of 8 bp sequence (5 -AGCGGCGT-3 ) from position 332 to 339 with a 16 bp sequence (5 -CCGGCCACTCCACCAG-3 ). This new sequence was found in the glyS gene of

4

J.G. de Boer / Mutation Research 536 (2003) 1–6

Fig. 2. Mutant E. Alignment between parts of the lacI sequence and the E. coli glyS gene. Sequence homology is underlined and the replaced sequence is in bold. The GlyS sequence is in reverse orientation.

E. coli, coding for glycyl tRNA synthetase ␤ subunit (Genbank accession number J01622). Upon examination of the flanking sequences a significant sequence homology was found between this site in the lacI gene and in the glyS gene. Fig. 2 shows the homology, which consists of 10 out of 11 bp at the 5 side and 7 out of 7 bp at the 3 side of the sequence in lacI. 3.5. Mutant F This mutant, which has been recovered two times from independent experiments, has 1099 bp in the lacI gene (position −81 to 1018) replaced by a new 74 bp sequence (Fig. 3). Flanking the two break points in lacI are two 6-bp palindromic sequences, CGTATT just 5 of position −81 (−92 → −87), and AATACG just 3 of position 1018 (1025 → 1030). Both sequences are 6 bp away from the breakpoints. The palindromes may allow the formation of a stem–loop structure in which the deleted sequence is contained in the loop. The newly formed sequence (at least from −235 to 1200, including upstream lacI sequences, the new insert, and downstream lacI sequences) is found as part of many expression and cloning vectors, including the pESP-4 expression vector (accession number AF126280) and the pYZ fission yeast expression vectors, indicating that a recombination event with a construct present in the host cell might have resulted in the replacement of the lacI sequence. The 74 bp

sequence includes the E. coli replication factor Y effector site (L strand) present at position 2353–2416 in pBR322. The 74 bp sequence is also found in pBluescript (∼918–1227) around the colE1 ori, but not in the genomic sequence of E. coli K12 (accession number NC 000913). This new sequence is not found on the lambda/LIZ construct, and the bacterial host SCS-8 is not reported to contain a plasmid. The tet resistance of this strain is located on an integrated Tn10 transposon (Stratagene catalogue). A PCR amplification performed on the SCS-8 cells using the p2 and p9 primers that are normally used to amplify the lacI gene from the lambda bacteriophage would be expected to yield a product of 400 bp, based on the presence of such a construct. However, the reaction yielded a PCR product of approximately 1250 bp, consistent with a normal lacI amplification product from lambda, but too large if the recovered construct in mutant F is present in E. coli. The phi80 marker in E. coli contains a lac operon (phi80dlac) [7]. This likely accounts for the approximately 1250 bp PCR fragment that can be recovered from SCS-8. However, it does not explain the origin of mutant F. In addition, amplification using either p9 or p8 as downstream primers indicated that the p8 site (further downstream than p9) is not present. This is consistent with the sequence found in cloning vectors such as pESP-4, where the homology ends at lacI position 1248, between the p8 and p9 primer sites. A PCR reaction using the p1 primer and a primer (called

Fig. 3. Mutant F. Vertical lines indicate the breakpoints of the replacement at position −81 and 1018, inclusive. Palindromic sequences are underlined and capitalized. The sequence in italics is the L-strand Y effector site.

J.G. de Boer / Mutation Research 536 (2003) 1–6

5

Fig. 4. Mutant G. Alignment between parts of the lacI sequence and the rat activin C gene. Sequence homology is underlined and the replaced sequence is in bold.

p74) situated in the 74 bp insert yields a 230 bp fragment from the mutant, and a 600 bp fragment from SCS-8 cells. We sequenced this fragment from the SCS-8 cells using the p74 primer. The sequence that we found is the second half of the lacI gene, going towards the 3 -end, rather than the 5 -end. At present we have no idea of the origin of the mutant sequence. 3.6. Mutant G The following mutant has a TG dinucleotide sequence at position 942–943 in the lacI gene replaced by the sequence CCT (Fig. 4). Upon a search of sequence databases the sequences 5 and 3 flanking the TG in the lacI gene were found to exist in the activin beta C gene of the rat (accession number AF140031), a 16-nucleotide sequence from position 688 to 673 (reverse orientation). This mutant was indeed recovered from rat, not mouse. In mice, there is a single nucleotide difference (position 675) in the homologous sequence flanking the replacement at the 3 site (in lacI), potentially reducing the likelihood of this event to occur. The sequence homology between the lacI and the activin beta C gene may have directed the exchange of nucleotides (Fig. 3). A similar mechanism, but involving a structure based on a quasipalindromic sequence, has been invoked in explaining the replacement of nucleotides in bacteriophage T4 [8].

4. Discussion The mutants that are described in this report were recovered from the lacI transgenic assay. Due to the nature of the system, it is not always possible to know whether the recovered mutations arose in the animal. Generally, however, the nature of the spectrum of spontaneous mutations recovered from the animal tissues clearly points to an animal origin. This is illustrated by the lack of bacteria-specific mutational hotspots [9]. Except in the case of mutant A and mu-

tant G, where the source sequence is present in the rat, the mutants described in this report appear to have derived their sequences from the bacterial host cell or the bacteriophage lambda vector that the lacI gene resides in. In case the sequence is present in lambda, the mutational event may have occurred in the animal or during plating. Several mechanisms have been proposed that can explain the acquisition of other sequences. Small changes have been found templated by nearby sequences through sequence homology [10]. Sequence similarities between two close or distant sequences can facilitate temporary alignment between these sequences, followed by strand extension or attempts to repair. Such a mechanism can be postulated to have resulted in the mutations in mutants E and G, and happened in the rat and the bacterial cell, respectively. Quasipalindromic or imperfect inverted repeat sequences, which can fold into imperfect stem–loop structures, can act as template for repair processes [8]. Repair of the stem converts the sequence of one side of the stem into the complement of the other side. One consequence of such mechanisms is the possibility of multiple mutations in close proximity. Palindrome sequences flanking a deleted DNA sequence, such as seen in mutant F have been suggested to be involved in deletion mutagenesis [11]. The recovery of mutants C and D reflects events that may have occurred in the animal tissues or in the bacterial host cell. The similarity between the two mutants is interesting, and illustrates that the mechanism that generates these events is frequent enough to allow such multiple recovery. An exploration of sequence databases did not reveal any likely sequence homologies for the target sequences. For example, none of the sites involved conform to a Chi recombination sequence (5 -GCTGGTGG-3 ), in fact, the Chi sequence does not occur in the pLIZ sequence or in the bacteriophage lambda sequence. The 3 -breakpoint in lambda in mutant D may have a partial topoisomerase II cleavage site of alternating purine/pyrimidine nucleotides (GCGCACAT) [12].

6

J.G. de Boer / Mutation Research 536 (2003) 1–6

This report illustrates a number of the mechanisms that can generate unusual mutational events in the lacI transgenic rodent system. The recovery of a bacterial IS10 element, a sequence from the E. coli GlyS gene, and the 74 bp sequence, indicate that mutations indeed can arise in the plating host bacteria. They can be recognized because of the unusual nature and the particular DNA sequence of the mutation. However, it cannot be excluded that the presence of bacterial DNA or animal DNA during packaging may provide the substrate for a recombination reaction. The mutants that we have found, however, are rare, and have been recovered among more than 25,000 mutants that were sequenced in this laboratory. The low frequency of occurrence shows that these events are not a concern in normal mutational studies, but do indicate that mutations can be recovered from the bacterial host cell. Other, simpler alterations cannot be recognized as originating in the plating host cell, but may contribute to the mutational spectra. While these normally go unnoticed, the nature of the mutations in this report may reveal the tip of a (probably small) iceberg.

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

Acknowledgements [10]

The support of the National Cancer Institute of Canada and the American Institute for Cancer Research is gratefully acknowledged.

[11]

References

[12]

[1] J.G. De Boer, A. Suri, D. Walsh, B.W. Glickman, The recovery of a B2 insertion in the lacI gene of a rat cell

line containing a lambda/LIZ shuttle vector, Mutat. Res. 375 (1997) 79–82. R.R. Young, B.J. Rogers, G.S. Provost, J.M. Short, D.L. Putman, Interlaboratory comparison: liver spontaneous mutant frequency from lambda/lacI transgenic mice (Big Blue) (II), Mutat. Res. 327 (1995) 67–73. G. Bogosian, K. Bilyeu, J.P. O’Neil, Genome rearrangements by residual IS10 elements in strains of E. coli K-12 which had undergone Tn10 mutagenesis and fusaric acid selection, Gene 133 (1993) 17–22. T.D. Lawley, V. Burland, D.E. Taylor, Analysis of the complete nucleotide sequence of the tetracycline-resistance transposon Tn10, Plasmid 43 (2000) 235–239. S.M. Halling, N. Kleckner, A symmetrical six-base-pair target site sequence determines Tn10 insertion specificity, Cell 28 (1982) 155–163. H. Rodriguez, E.T. Snow, U. Bhat, E.L. Loechler, An E. coli plasmid-based, mutational system in which supF mutants are selectable: insertion elements dominate the spontaneous spectra, Mutat. Res. 270 (1992) 219–231. J.R. Beckwith, E.R. Signer, W. Epstein, Transposition of the Lac region of E. coli, Cold Spring Harb. Symp. Quant. Biol. 31 (1966) 393–401. J.G. De Boer, L.S. Ripley, Demonstration of the production of frameshift and base-substitution mutations by quasipalindromic DNA sequences, Proc. Natl. Acad. Sci. U.S.A 81 (1984) 5528–5531. J.G. De Boer, B.W. Glickman, The lacI gene as a target for mutation in transgenic rodents and E. coli, Genetics 148 (1998) 1441–1451. G.B. Golding, B.W. Glickman, Sequence-directed mutagenesis: evidence from a phylogenetic history of human alphainterferon genes, Proc. Natl. Acad. Sci. U.S.A 82 (1985) 8577–8581. B.W. Glickman, L.S. Ripley, Structural intermediates of deletion mutagenesis: a role for palindromic DNA, Proc. Natl. Acad. Sci. U.S.A 81 (1984) 512–516. J.R. Spitzner, I.K. Chung, M.T. Muller, Eukaryotic topoisomerase II preferentially cleaves alternating purine-pyrimidine repeats, Nucleic Acids Res. 18 (1990) 1–11.