Use of Vectorette and Subvectorette PCR to Isolate Transgene Flanking DNA

Downloaded from genome.cshlp.org on October 2, 2021 - Published by Cold Spring Harbor Laboratory Press

Maxine J. Allen, Andrew Collick, and Alec J. Jeffreys

Department of Genetics, University of Leicester, Leicester LE1 7RH, UK

Vectorette PCR permits the specific PCR allows specific amplification of structures and in the recovery of un- amplification of DNA segments flank- DNA fragments between two regions of known bacterial sequences. ~4-6~ In this ing a known DNA sequence. It en- known sequence to which primers can paper we detail its use in the systematic ables the application of the PCR be directed. Often, however, molecular isolation of 5'- and 3'-flanking junction where sequence information is only biological techniques require the isola- sequences around mouse transgenes of available for one primer site. We now tion and subsequent sequence analysis the minisatellite MS32. MS32 is a hu- show that vectorette PCR can be used of unknown regions of DNA. PCR can be man minisatellite composed of a tan- for the systematic mapping and re- used to retrieve this unknown sequence dem array of 29-bp repeat-unit sequence trieval of transgene flanking DNA. if it is flanking a region of known se- variants. The interspersion of these We also show that the sequence of quence. In this case, a primer can be de- sequence variants can be assayed in large vectorette PCR fragments can signed in either orientation from the the minisatellite variant repeat (MVR)- be obtained without cloning, by the known sequence. Such techniques as in- PCR ~7~, which utilizes repeat-unit-spe- production of subvectorette frag- verse PCR, panhandle PCR, capture PCR, cific primers. MS32 has a high muta- ments. and vectorette PCR have all been de- tion rate in humans, and knowledge of scribed as PCR methods for genome the flanking mouse sequences around walking.~ 1-4~ transgenes of this locus is important in The basic vectorette PCR strategy has ascertaining any positional effects on been described previously. ~s~ Briefly, a transgene mutation rate and process. vectorette-linker library is constructed It was therefore necessary to isolate at from digested DNA for use in the PCR. least one flanking sequence junction The duplex vectorette linker consists of around each of eight transgenes of the two annealed oligonucleotides of com- MS32 locus. Vectorette PCR was chosen plementary sequence flanking a nonho- as a swift method for isolating transgene mologous region resulting in a "bubble" flanking sequence products, obviating in the DNA. The vectorette primer has an the need for cloning of the junction identical sequence to the nonhomolo- regions. The double-stranded products gous region of the lower oligonucleotide obtained were sequenced using PCR- and therefore cannot anneal directly to based double-stranded sequencing the linker. The specificity of the PCR is methods. Most double-stranded PCR se- achieved by performing primer exten- quencing methods will only give 200- sion reactions from a sequence-specific 300 bp from each end of a fragment. primer to a vectorette linker in the adja- Therefore, to access full sequence infor- cent unsequenced DNA. This results in mation for the original vectorette frag- the synthesis of a strand complementary ment, subvectoretting procedures were to the vectorette linker, thus generating developed to provide a series of overlap- a target sequence for the vectorette-spe- ping contigs for sequencing. We also cific primer in subsequent PCR cycles. In show how the specificity of the vector- this way, a specific DNA fragment can be ette library can be increased by the use of amplified from a complex fragment mix- size-selected fragments and how par- ture and recovered for use in sequencing tially digested DNA can be used to give a reactions. series of flanking products, aiding struc- Previously, vectorette technology has tural analysis. In this paper vectorette been used to isolate terminal sequences analysis of three MS32 transgenic loci, from yeast artificial chromosome (YAC) called 110D, 102B, and 110A, is de- inserts, in the determination of exon scribed.

4:71-759 by Cold Spring Harbor Laboratory Press ISSN 1054-9803/94 $5.00 PCR Methods and Applications 71 Downloaded from genome.cshlp.org on October 2, 2021 - Published by Cold Spring Harbor Laboratory Press

MATERIALS AND METHODS Vectorette PCR Amplification Winship (8) using either the vectorette or transgene primer. Vectorette Production Hot start PCR was performed to prevent the annealing of the top- and bottom- The sequences of the vectorette top and strand vectorette oligonucleotides and Subvectoretting of Vectorette PCR bottom strands are detailed in Figure 1. any subsequent template-directed exten- Products Oligonucleotides were synthesized in sion from the free 5' end of the top vec- the Department of Biochemistry, Uni- The full sequence of vectorette PCR torette strand, which would not have li- versity of Leicester, on an Applied Bio- products >500 bp in length was ob- gated to the 3' end of the template top systems 380B DNA synthesizer, using re- tained by the production of subvector- strand. This obviated the need for 5' agents supplied by Cruachem. Ten ette fragments. Full or partial digestions phosphorylation of the top strand of the micrograms of vectorette top strand and with enzymes cleaving within the initial vectorette oligonucleotide. Ten nano- 10 I~g of vectorette bottom strand were vectorette product were performed on grams of vectorette genomic DNA was annealed at 60~ for 30 min in 20 I~l of 200-300 ng of initial product. The di- diluted to 10 I~l with water, denatured at 0.1 M NaC1, 10 mM MgC12, and 10 mM gested DNA was phenol-extracted and 96~ for 3 min, and then held at 80~ Tris-HC1 (pH 7.6) to produce a complete recovered by ethanol precipitation. Ten microliters of 2x concentrated PCR vectorette linker. Neither vectorette oli- Twenty nanograms of blunt-ended vec- buffer, primers, and Taq poymerase gonucleotide was 5' phosphorylated. torette linker was annealed to 10- to 15- (Perkin-Elmer Cetus AmpliTaq) were ng DNA fragments in a 20-1~1 reaction us- added, to give a final concentration of 1 ing the previous ligation conditions. ~l.M transgene primer, 1 ~M vectorette DNA Digestion, Preparation, and Reamplifications were performed on 1 i~l primer, and 0.06 U/l~l of Taq polymerase. Ligation of a 1:20 dilution of this subvectorette PCR buffer was as described previ- library for 25 cycles using the second Approximately 2 t~g of genomic DNA di- ously. (7) Amplification was at 96~ for vectorette primer plus either the trans- gested with a six-cutter restriction en- 1.2 min, 68~ for 1 min, 70~ for 3 min gene primer or the first vectorette zyme was ligated to 0.2 p.g of annealed for 30 cycles on a Perkin-Elmer Cetus primer. PCR products were repurified by vectorette linker in 20 t~l of 50 mM NaC1, thermal cycler. Vectorette PCR products agarose gel electrophoresis and electroe- 20 mM Tris-HCl (pH 7.6), 5 mM DTT, and were detected by agarose gel elec- lution followed by ethanol precipitation. 0.1 mM ATP, with 2 units of T4 ligase trophoresis followed by Southern blot (GIBCO BRL) overnight at 15~ for sticky hybridization or ethidium bromide end ligations and 4~ for blunt-ended li- staining and visualisation by long-wave- RESULTS gations. The amount of vectorette linker length UV. Appropriate DNA fragments The structure of the MS32 allele used in in the reaction was increased to 2 I~g for were recovered by electroelution onto transgenesis is detailed in Fig. 2. It has 71 total genomic DNA ligations cleaved dialysis membrane and reamplified with repeat units of a- or t-type, each repeat with a restriction endonuclease recog- vectorette and transgene primers for 25- being 29 bp long. Standard restriction di- nizing a 4-bp cleavage site. Control liga- 30 cycles depending on yield. PCR prod- gestion and Southern blot analysis were tions without DNA were also prepared, ucts were repurified by agarose gel elec- performed on mice positive for trans- and ligation was monitored by removal trophoresis and electroelution followed gene insertion. This identified appropri- of a 3-1~1 aliquot and addition of 0.5 I~g by ethanol precipitation. of k DNA cleaved with HindIII and then ate restriction sites around the transgene agarose electrophoresis after ligation. Li- for use in vectorette PCR to recover gations ("vectorette libraries") were di- mouse flanking DNA sequences. Sequencing luted with 5 mM Tris-HC1 (pH 7.6) to 5 i~g/ml of genomic DNA and stored at Fifty to 150 ng of double-stranded PCR Transgene 1 IOD

- 20oc. product was sequenced by the method of Restriction mapping showed this to be a single-copy insertion of the transgene construct with the loss of 10 repeat units Vecl top -64nt - AvaII degenerate but with full 5' and 3' MS32 flanking

5'-ga/tcaggctggagatgtagcagattgagatattcgttatagtttacctatcccgaccgagcatg-3 ' sequences. There was a lack of common four- and six-cutter restriction sites in Vecl top -64nt - Mbo I the flanking mouse DNA, but the en- 5'-gatcaggctggagatgtagcagattgagatattcgttatagtttacctatcccgaccgagcatg-3' llllIlllllllllllIl llllillllllllIIflll zyme AvaII cleaved frequently and was Vecl bottom -60nt 3'-....tccgacctctacatcgtccttggattgggagatctggtcacggatagggctggctcgtac-5' utilized in the preparation of a vectorette Vecl primer -26nt 3' -cttggattgggagatctggtcacgga- 5' library. A size fraction of AvaII-digested

Vec3 top - 64nt - Blunt ended 110D positive mouse genomic DNA, known to contain the transgene frag- 5 ' -gatcatgccctgacccaggctaagtat tttaact t taaccact t tgct ccgacagtcccattgc-3 ' lllIlllflllltllllllJllll llllIJJfllllJllill ment, was recovered by electroelution Vec3 bottom -64nt 3 ' -ctagtacgggactgggtccgattctcggttcgaaaccgggagtcaggaggctgtcagggtaacg-5 ' Vec3 primer -24nt 3 ' -tcggttcgaaaccgggagtcagg- 5 ' from agarose and used in the prepara- FIGURE 1 The oligonucleotides used in production of the vectorette units. One vectorette top tion of the vectorette library. Because strand and one vectorette bottom strand are annealed to produce a complete vectorette linker. Avail cleaves a degenerate G' GA/TCC site, Vectorette linkers for different restriction fragments are produced by varying the vectorette top a 50% mixture of GAC and GTC 5' over- strand sequence. hanging Vecl top sequences was used in

72 PCR Methods and Applications Downloaded from genome.cshlp.org on October 2, 2021 - Published by Cold Spring Harbor Laboratory Press

products in the initial PCR. Sequencing the 5' and 3' junction fragments confirmed the unusual lack of restriction sites in the flanking DNA that was attrib- utable to the -y satellite nature of the sequence.

Transgene 102B Restriction mapping showed this single- copy transgene to have intact 3' MS32 human flanking DNA but no 5' flanking FIGURE 2 The structure of the MS32 transgene construct. The MS32 allele consists of 71 repeat sequences. The mouse flanking DNA se- units with 212 bp of 5' MS32-flanking DNA and 214 bp of 3'-flanking DNA. The total length of the construct is 2525 bp. The allele was PCR-amplified from human DNA using primers 32A and quences again showed an unusual lack 32D with attached 20-bp synthetic linkers E1 and E2 (hatched). Diagnostic restriction sites in the of most restriction sites other than for linkers and flanking DNA are PstI (P), HindII (H), and EcoRI (E). The positions of the MS32 Avail. AvaII was again chosen to con- primers are indicated by arrows. struct a vectorette library using partially digested 102B genomic DNA. Vectorette PCR amplifications were used as a mapping tool to generate products extending the preparation of the vectorette linker quencing, the size fraction containing out to a number of AvaII sites in the 5'- for this library. The first amplifications the correct product was recovered by and 3'-flanking DNA (Fig. 4A, B). Ampli- from this library were across the MS32 electroelution and amplified with a fications with the 32E and Vecl primers array, and products were verified by nested MS32 primer. This heminesting gave four bands, representing products diagnostic restriction analysis and strategy was utilized to specifically re- extending to four flanking AvaII sites. E1 Southern hybridization (Fig. 3). Before cover the correct 5' and 3' products from and Vecl primer amplification only gave reamplification of the product for se- the mixture of correct and Vecl/Vecl products extending to the first three of these AvaII sites, presumably owing to the favored amplification of shorter products. Mouse -y satellite DNA has a 234-bp repeating sequence. ~9~ Avail cleaves at a variant position in some -y satellite repeat units producing a 234- or 468-bp periodicity of digestion. This periodicity is seen in the flanking Avail restriction map determined from vectorette PCR products, suggesting strongly that the flanking mouse DNA is gamma satellite. As there was no 5' human MS32 flanking DNA, a heminested strategy for size reduction of the initial 5' vectorette PCR products (Vecl/E1 and Vecl/32E; Fig. 4B) was developed using MVR-PCR. For this reaction the Vecl primer was used as the flanking primer, and the repeat unit-specific primers 32-TAG-A or 32-TAG-T were used as the transgene primers to produce a series of products extending from the closest AvaII site in the 5'-flanking mouse DNA to each a- or t-type repeat unit in the transgene MS32 array (Fig. 4C). The first repeat unit in the array is an a- type repeat, and the appropriate band from the MVR-PCR (indicated by an arrow in Fig. 4C) was electroeluted and amplified for sequenc- FIGURE 3 The structure of the 110D transgene and associated vectorette products. (A) (A) AvaII ing, which confirmed the ~/satellite se- digestion site to which AvalI degenerate linkers were attached; (E) EcoRI; (P) PstI; (H) HindlII. (B) quence of the mouse DNA. Southern blot of the initial vectorette PCR products, Vec1/32E and Vecl/320, probed with MS32 The 3'-flanking mouse DNA was re- repeat-unit probe. covered by vectorette PCR using primer

PCR Methods and Applications 73 Downloaded from genome.cshlp.org on October 2, 2021 - Published by Cold Spring Harbor Laboratory Press

amplification using primer 32OR directed toward the 5'-flanking mouse DNA produced a 1.34-kb product detectable by ethidium bromide staining fol- lowing agarose gel electrophoresis. This was too long for full double-stranded PCR sequencing. Restriction mapping of this product revealed a number of blunt- end restriction sites to which a universal blunt-end novel sequence vectorette (Vec3) could be annealed (Fig. 5). The 1.34-kb DNA fragment was either fully digested with AluI or HincII or partially digested with RsaI (because there were two sites to access with Vec3) and phenol-extracted and ethanol-precipitated, before the novel Vec3 was ligated. These subvectorette libraries were reamplified using 32OR and Vec3, and the products were recovered for sequencing. Se- quence analysis of the original vectorette products and the four subvectorette products produced a complete sequence contig of the 5'-mouse flanking DNA, which was shown to be a unique sequence.

DISCUSSION In this paper we discuss the use of vectorette PCR as a method of systemati- FIGURE 4 The structure of the 102B transgene and associated vectorette PCR products. (A) (E) EcoRI; (P) PstI; (M) MnlI; (A) AvalI. Avail vectorette linkers were attached to DNA partially digested cally isolating mouse DNA adjacent to with Avail, to produce vectorette PCR products extending to Avail sites in both the 5'- and transgene insertions as a prelude to DNA 3'-mouse flanking DNA (a, b, and c represent the Vecl/E1 products detailed in B). Primer 32- sequence analysis and show that vector- TAG-A is MS32 repeat-unit-specific. (B) Vectorette products extending from the 5' junction of the ettes provide a highly flexible approach transgene. Using initial primer combinations Vecl/32E and Vecl/E1, there are three products to the mapping and recovery of se- extending out to three AvaII sites in the flanking DNA. There is also a product extending to a quences from complex genomes. In this fourth Avail site detectable in the Vec1/32E amplification (not detailed in A). (C) The MVR-PCR study we were able to isolate and se- pattern visible by ethidium bromide staining when the electroeluted band from the Vec1/32E quence at least one junction from each reaction was amplified with Vec1/32-TAG-A (triplicate reactions). Duplicate amplifications on of eight different transgenic inserts, the same electroeluted product using the 32-TAG-T repeat-unit-specific primer were also performed to generate the corresponding set of MVR-PCR products extending to each t-type repeat without the need for cloning via ge- unit. (F) ~X174 cleaved with HaeIII. An arrow indicates the smallest product from the Vec1/32- nomic DNA libraries. Junction frag- TAG-A reaction that was electroeluted and used in sequencing reactions. (D) Vectorette products ments up to 3.7 kb in length were suc- extending 3' of the transgene insertion. A Southern blot of amplification products with Vecl/ cessfully recovered by vectorette PCR. 32ER was probed with an MS32 3'-flanking probe (141 bp of the 3'-flanking DNA). Three vec- The ability to recover specific vector- torette products were amplified extending to Avail sites in the 3'-flanking mouse DNA. The ette PCR products depends on the com- smallest product is cleaved by MnlI, which cuts in the MS32 flanking human DNA. plexity of genomic DNA, the frequency of restriction sites in the DNA used to prepare the vectorette libraries, and the specificity of vectorette and transgene 32ER in the human 3'-flanking DNA Transgene 110A primers. The latter was optimized by de- (Fig. 4D). Junction products were veri- signing primers with a relatively high fied by Southern blot hybridization with Restriction mapping identified this GC content (54--58%), which permitted a human 3'-flanking sequence probe transgene as a complex multicopy array the use of a high annealing temperature and digestion with MnlI, which has a di- of the original construct. The enzyme (67~ during PCR. In some cases, speci- agnostic site in the 3'-flanking DNA. A MboI does not cleave in the original con- ficity was sufficiently high to enable the heminested strategy using the reverse struct but cleaves in the mouse DNA 1.3 correct vectorette PCR products to be primer of 32A (Fig. 1) with Vecl was used kb upstream of the 5' end of the multi- identified after the first round of PCR by to recover this fragment for sequencing, copy array. A complete MboI digest was agarose gel electrophoresis and staining which showed that the mouse sequence performed on the genomic DNA, and with ethidium bromide. In other cases, was ~/satellite. MboI vectorettes were ligated. The initial particularly if the vectorette library was

74 PCR Methods and Applications Downloaded from genome.cshlp.org on October 2, 2021 - Published by Cold Spring Harbor Laboratory Press

International Research Scholars award from the Howard Hughes Medical Insti- tute and in part by grants from the Med- ical Research Council and the Royal So- ciety.

REFERENCES 1. Triglia, T., M.G. Peterson, and D.J. Kemp. 1988. A procedure for in vitro amplification of DNA segments that lie outside the FIGURE 5 The structure of the 5' region of the multicopy transminisatellite 110A and associated boundaries of known sequence. Nucleic vectorette and subvectorette PCR products. The initial Vecl product used an MboI vectorette Acids Res. 16: 8186. linker and the subvectorette was produced using a blunt-ended Vec3 linker ligated to sites for 2. Jones, D.H. and S.C. Winisforter. 1992. AluI (Al), RsaI (R), and HinclI (Hc). Sequence specific generation of a DNA panhandle permits PCR amplification of unknown flanking DNA. Nucleic Acids Res. 20: 595-600. constructed from genomic DNA cleaved flanking restriction site to each a- or 3. Lagerstrom, M., J. Parik, H. Malmgren, J. with frequently cutting restriction en- t-type repeat unit within the minisatel- Stewart, U. Peterson, and U. Landegren. zymes (e.g., MboI), relatively complex lite array. This makes it possible to deter- 1991. Capture PCR: Efficient amplifica- sets of PCR products were frequently ob- mine transgene structures in the absence tion of DNA fragments adjacent to a tained. Two approaches were used to cir- of suitable flanking human DNA and known sequence in human and YAC DNA. PCR Methods Applic. 1: 111-119. cumvent this problem. First was knowl- gives direct information on the distance 4. Riley, J., R. Butler, D. Ogilvie, R. Finniear, edge of the size of the restriction between the flanking mouse restriction D. Jenner, S. Powell, R. Anand, J.C. Smith, fragment containing the transgene en- site and the start of the MS32 repeat ar- and A.F. Markham. 1990. A novel method abled restricted genomic DNA to be size- ray. for the isolation of terminal sequence fractionated by agarose gel electrophore- Vectorette PCR can also be used as a from yeast artificial chromosome (YAC) sis, followed by the recovery of a size tool to aid sequence analysis of large clones. Nucleic Acids Res. 18: 2887-2890. fraction containing the transgene frag- PCR products by subvectoretting. Our 5. Arnold, C. and I.J. Hodgson. 1991. Vector- ment. This type of fractionation was per- general strategy is to partially digest ette PCR: A novel approach to genome formed in the recovery of transgene large PCR products with frequently cut- walking. PCR Methods Applic. 1: 39-42. 110D flanking DNA to reduce the com- ting restriction enzymes that yield blunt- 6. Roberts, G.R., A.J. Coffrey, M. Bobrow, and D. Bentley. 1992. Determination of plexity of the vectorette library 10- to end DNA fragments, followed by liga- the exon structure of the distal portion of 100-fold and to substantially improve tion of a second vectorette and specific the Dystrophin gene by vectorette PCR. the specificity of the vectorette PCR. The amplification of a set of DNA fragments Genomics 13: 942-950. second approach was to perform a first extending from one of the original prim- 7. Jeffreys, A.J., A. Macleod, K. Tamaki, D.L. round of vectorette PCR amplification, ers (the first vectorette primer or a trans- Neil, and D.G. Monkton. 1991. Minisatel- then to detect products by Southern blot gene primer) to each of the blunt-end lite repeat coding as a digital approach to hybridization, recover by gel fraction- sites capped with the second vectorette. DNA typing. Nature 354: 204-209. ation, and reamplify using a second In this way, sets of nested deletions can 8. Winship, P.R. 1989. An improved method nested transgene primer; this heminest- be rapidly generated and used for se- for directly sequencing PCR amplified ing strategy can also simplify the profile quence contig construction. We would material using dimethyl sulphoxide. Nu- cleic Acids Res. 17: 1266. of vectorette PCR products. note, however, that in our experience, 9. Hastie, N.D. 1989. Highly repeated DNA The use of partially digested genomic blunt-end vectorettes work very ineffi- families in the genome of Mus musculus. DNA in vectorette libraries also provides ciently in PCR amplifications from pri- In Genetic variants and strains of the labo- a general method of mapping restriction mary genomic vectorette libraries and ratory mouse, 2nd ed. (ed. M.F. Lyon and sites adjacent to a transgene, as shown do not provide a useful means for isolat- A.G. Searle), pp. 559-574. Oxford Univer- for transgene 102B. Such mapping data ing transgene/flanking DNA junctions sity Press, Oxford, UK. does not only provide clues about the from genomic DNA. nature of the mouse flanking DNA, but it can also generate primary vectorette Received May 17, 1994; accepted in ACKNOWLEDGMENTS products extending for any required dis- revised form July 12, 1994. tance into the flanking DNA, without We thank Debra Langton for oligonucle- the need for detailed restriction map- otide synthesis and Annette Macleod for ping prior to vectorette PCR. Similarly, helpful discussions on the vectorette vectorette PCR can also be used to ex- PCR technique. Production of the trans- plore the structure of the transgene it- genic mice was performed by Michael self. For example, it is possible to per- Norris and Sheila Barton at the Agricul- form MVR-PCR directly on primary tural Food and Research Council Insti- vectorette libraries (data not shown) to tute (Babraham, Cambridge, UK). The create PCR products extending from a work of A.J.J. was supported in part by an

PCR Methods and Applications 75 Downloaded from genome.cshlp.org on October 2, 2021 - Published by Cold Spring Harbor Laboratory Press

Use of vectorette and subvectorette PCR to isolate transgene flanking DNA.

M J Allen, A Collick and A J Jeffreys

Genome Res. 1994 4: 71-75

License

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the Service top right corner of the article or click here.

To subscribe to Genome Research go to: https://genome.cshlp.org/subscriptions