Use of a Small Palindrome Genetic Marker to Investigate Mechanisms of Double-Strand-Break Repair in Mammalian Cells

Corresponding author: Mark D. Baker, Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, Ontario N1G 2W1, Canada., mbaker02{at}uoguelph.ca (E-mail)

Communicating editor: C. KOZAK

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We examined mechanisms of mammalian homologous recombination using a gene targeting assay in which the vector-borne region of homology to the chromosome bore small palindrome insertions that frequently escape mismatch repair when encompassed within heteroduplex DNA (hDNA). Our assay permitted the product(s) of each independent recombination event to be recovered for molecular analysis. The results revealed the following: (i) vector-borne double-strand break (DSB) processing usually did not yield a large double-strand gap (DSG); (ii) in 43% of the recombinants, the results were consistent with crossover at or near the DSB; and (iii) in the remaining recombinants, hDNA was an intermediate. The sectored (mixed) genotypes observed in 38% of the recombinants provided direct evidence for involvement of hDNA, while indirect evidence was obtained from the patterns of mismatch repair (MMR). Individual hDNA tracts were either long or short and asymmetric or symmetric on the one side of the DSB examined. Clonal analysis of the sectored recombinants revealed how vector-borne and chromosomal markers were linked in each strand of individual hDNA intermediates. As expected, vector-borne and chromosomal markers usually resided on opposite strands. However, in one recombinant, they were linked on the same strand. The results are discussed with particular reference to the double-strand-break repair (DSBR) model of recombination.

HOMOLOGOUS recombination is of fundamental importance in living organisms. It can increase genetic diversity and thus the robustness of the species, and it can repair errors of replication and other forms of DNA damage. On the other hand, homologous recombination is implicated as a mechanism of homozygotization of recessive oncogenes seen in some human cancers (LASKO et al. 1991 ). The important role homologous recombination plays in shaping the genome clearly underlines the necessity of learning more about its mechanisms.

Unfortunately, the investigation of homologous recombination mechanisms in mammalian cells is problematic. One important reason is that recombination is usually investigated in cultured, mitotically dividing cells with the result that not all products arising from individual recombination events are recovered. Other difficulties include a paucity of well-characterized mutants affecting the recombination process, problems associated with constructing appropriate strains to study recombination, and complications associated with the detection and monitoring of recombination. Consequently, much of our knowledge of homologous recombination mechanisms has come from studies in microbial systems where recombination can be investigated using powerful genetic approaches in combination with recombinant DNA technology (PETES et al. 1991 ).

Intrinsic to the study of recombination mechanisms is the ability to determine the distribution of heteroduplex DNA (hDNA) in a recombination event. In Saccharomyces cerevisiae, small palindromes are thought to form hairpins in hDNA, resulting in their frequent escape from mismatch repair (MMR; NAG et al. 1989 ; NAG and PETES 1991 ; WENG and NICKOLOFF 1998 ). As a consequence of semiconservative DNA replication, unrepaired mismatches generate genetically distinct molecules that, upon cell division, segregate to different daughter cells. These can be detected as sectored (mixed) colonies, whether of meiotic or mitotic origin (PETES et al. 1991 ). Examination of the segregation patterns of small palindromes revealed information about hDNA formation and, hence, recombination mechanisms (NAG et al. 1989 ; NAG and PETES 1990 ; PORTER et al. 1993 ; GILBERTSON and STAHL 1996 ; WENG and NICKOLOFF 1998 ). Here, we used a small palindrome as a genetic marker to investigate mammalian homologous recombination.

We have previously described a powerful system where the product(s) of homologous recombination within an individual mammalian cell are retained for molecular analysis (NG and BAKER 1999A ). Our assay utilizes a gene targeting approach in which repair of a double-strand break (DSB) introduced into the vector-borne region of homology to the chromosome results in targeted vector integration. In S. cerevisiae, several features of the gene targeting reaction parallel meiotic recombination and can be explained by the double-strand-break repair (DSBR) model for recombination (RESNICK 1976 ; SZOSTAK et al. 1983 ). To further exploit our system and, specifically, to investigate the presence and position of hDNA formed during homologous recombination, we inserted a 30-bp palindrome into three restriction enzyme sites spanning the vector-borne region of homology to the chromosome, destroying each endogenous site and replacing it with a diagnostic NotI site embedded in the palindrome. Several important features emerged from the study of the targeted recombinants. It was determined that DSB processing usually did not result in formation of a large double-strand gap (DSG) and that recombination often occurred by crossover at or near the DSB, while in some recombinants an hDNA intermediate was involved. For the sectored (hDNA) recombinants, the linkage of markers in each strand of the hDNA intermediate was determined and, as expected, vector-borne and chromosomal markers were usually present in opposite strands of the hDNA intermediate. The above results are discussed with particular reference to the DSBR model of recombination.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Hybridomas and culture conditions:
The wild-type murine hybridoma Sp6/HL bears the target for homologous recombination, the single copy of the trinitrophenyl (TNP)-specific chromosomal immunoglobulin µ gene. The origin of Sp6/HL and the methods used for cell culture have been described previously (KOHLER and SHULMAN 1980 ; KOHLER et al. 1982 ).

Gene targeting vector:
The 13.4-kb enhancer-trap vector pCµpalEn^- was constructed for use in the gene targeting studies. It is similar to the enhancer-trap vector pCµEn^- described earlier (NG and BAKER 1998 ) in that it bears a 5.8-kb Cµ segment from the wild-type chromosomal µ gene inserted into a derivative of the vector pSV2neo in which the 372-bp NsiI/NdeI fragment encompassing the SV40 early region enhancer important in neo gene expression was removed. As shown previously (BAUTISTA and SHULMAN 1993 ; NG and BAKER 1998 ), the enhancer-trap feature enriches for gene targeting events at the chromosomal immunoglobulin µ locus. This occurs because the endogenous chromosomal µ gene supplies the enhancer (or equivalent) activity required for expression of the enhancerless selectable marker (neo) in targeted recombinants, whereas most sites of random vector integration in the hybridoma genome do not. The distinctive feature of pCµpalEn^- is that it contains a 30-bp palindrome (5' GTACTGTATGTGCGGCCGCACATACAGTAC 3') inserted at each of three positions within the otherwise wild-type Cµ region of homology. The palindrome was self-annealed and ligated into the SacI, AflII, and ApaI sites located within the Cµ region at genomic positions 503, 1053, and 1765 bp, respectively, according to the numbering presented in GOLDBERG et al. 1981 . To permit cohesive-end ligation, the three palindrome-containing oligonucleotides were each synthesized with the appropriate terminal nucleotides (MOBIX, McMaster University, Hamilton, Ontario). The SacI, AflII, and ApaI sites are located at distances of 645, 1199, and 1925 bp relative to the unique XbaI site of vector linearization (defined as position 0 bp; Fig 1A). Palindrome insertion destroys the endogenous restriction enzyme site at each position creating, in its place, the unique NotI site embedded within the palindrome (denoted in boldface type). To propagate plasmids, the palindrome-permissive Escherichia coli strain DL795 was used (kindly provided by David Leach). Vector construction and plasmid isolation were performed according to standard procedures (SAMBROOK et al. 1989 ).

View larger version (21K): In this window In a new window Download PPT slide   Figure 1. Gene targeting at the chromosomal immunoglobulin µ locus. (A) The 13.5-kb enhancer-trap vector pCµEn-pal bears a perfect 30-bp palindrome containing a NotI site (sequence presented in MATERIALS AND METHODS) that replaces the SacI, AflII, and ApaI restriction enzyme sites at the indicated positions in the vector-borne Cµ region (sites numbered relative to the XbaI site of vector linearization at position 0). Although not relevant to this study, an enhancerless Herpes Simplex Virus-1 thymidine kinase (tk) gene is also present in the vector backbone. (B) The structure of the haploid, recipient chromosomal immunoglobulin µ gene in the wild-type, Sp6/HL hybridoma showing the endogenous SacI, AflII, and ApaI Cµ region restriction enzyme sites. (C) The structure of the chromosomal µ gene following targeted integration of a single copy of the vector, pCµEn-pal. As detailed previously (NG and BAKER 1999A ), the 5' Cµ region can be specifically amplified to yield a 4765-bp product using primer pair AB9703/AB9745, while primer pair AB9703/AB9438 generates a specific 4621-bp product from the 3' Cµ region. The primer binding sites are as follows: primer AB9703 binds within the vector-borne Cµ region, 85 bp 5' of the XbaI site; primer AB9745 binds to the complementary DNA strand near the beginning of the vector-borne amp gene; primer AB9438 binds to the complementary DNA strand outside the vector-borne Cµ region, 4730 bp 3' of the XbaI site. In both B and C, the fragment sizes generated following digestion with EcoRI and hybridization with probe F, an 870-bp XbaI/BamHI Cµ region fragment, are presented. E, EcoRI; VH TNP, TNP-specific µ heavy-chain variable region; Cµ, µ gene constant region; neo, neomycin phosphotransferase gene; tk, HSV-1 thymidine kinase gene. The figures are not drawn to scale.

Gene targeting and transformant isolation:
pCµpalEn^- vector DNA (8.7 pmol) was linearized by XbaI digestion. The cut vector was then introduced into 2 x 10⁷ recipient Sp6/HL hybridomas by electroporation as described (BAKER et al. 1988 ). To isolate independent G418^R transformants, immediately following electroporation the Sp6/HL culture was resuspended in 333 ml of DMEM (Dulbecco's modified Eagle's medium containing 13% bovine calf serum and 5.3 x 10^-4 M 2-mercaptoethanol) and 0.1 ml aliquots (~6000 hybridomas) were distributed to 3333 individual wells of 96-well microtiter plates. Trypan blue staining revealed that hybridoma survival following electroporation averaged ~50%. Therefore, each culture well received ~3000 viable hybridomas. Two days postplating, each culture well received 0.1 ml of DMEM supplemented with G418 to achieve a final active concentration of 1.2 mg/ml. The culture plates were placed at 37° in a humidified, 7% CO₂ atmosphere to allow growth of the G418^R colonies.

Screening for targeted, G418^R recombinants:
To identify targeted G418^R recombinants, genomic DNA was prepared from each G418^R transformant according to the procedure of GROSS-BELLARD et al. 1973 and subjected to Southern analysis. Restriction enzymes were purchased from Bethesda Research Laboratories (Gaithersburg, MD), New England Biolabs (Beverly, MA), and Pharmacia Inc. (Piscataway, NJ) and used in accordance with the manufacturers' specifications. Gel electrophoresis, transfer of DNA onto nitrocellulose membrane, ³²P-labeled probe preparation, and hybridization were all performed according to standard procedures (SAMBROOK et al. 1989 ).

PCR analysis:
PCR was used to specifically amplify the 5' and 3' Cµ region in targeted G418^R recombinants. As described previously (NG and BAKER 1999A ), oligonucleotide primers AB9703 (5'-CTACTTGAGAAGCCAGGATCTAGG-3') and AB9745 (5'-ACCGGATCTTACCGCTGTTGAG-3') specifically amplify the 5' Cµ region while primers AB9703 and AB9438 (5'-GTACCATCAGACTGCACTGTTCCA-3') specifically amplify the 3' Cµ region. The position of each primer binding site in the recombinant µ gene is described in the legend to Fig 1. All primers were synthesized at MOBIX. The conditions used for PCR were described previously (NG and BAKER 1999A ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Experimental system:
To investigate homologous recombination in mammalian cells, a gene targeting assay was used. The vector pCµEn^-pal (Fig 1A) was linearized at the unique XbaI site within the 5.8-kb µ gene constant (Cµ) region segment in order to effect homologous recombination with the haploid, chromosomal Cµ region in the wild-type, murine hybridoma, Sp6/HL (Fig 1B). Linearization at XbaI provided 1.5 and 4.3 kb of overall homology to the Sp6/HL Cµ region, 5' and 3' of the DSB, respectively. Two distinctive features characterize pCµEn^-pal. First, the 372-bp NsiI/NdeI fragment encompassing the SV40 early region enhancer responsible for neo expression was removed from the pSV2neo backbone. Second, an identical 30-bp palindrome was inserted at the SacI, AflII, and ApaI sites within the vector-borne Cµ region, replacing each with the diagnostic NotI site embedded within the palindrome. In Fig 1A, the palindrome insertion sites are numbered relative to the XbaI site of vector linearization at position 0. The restriction enzyme site markers permit the contribution of vector-borne and chromosomal sequences in the recombinant product(s) to be determined. Palindrome insertion was planned so as to minimize disruption of the perfect sequence homology shared between the vector-borne and chromosomal Cµ regions. With the exception of the palindromes, the vector-borne and chromosomal Cµ regions were isogenic.

The enhancer-trap feature of pCµEn^-pal is very important because it enriches for gene targeting events at the chromosomal µ locus (BAUTISTA and SHULMAN 1993 ; NG and BAKER 1998 ). This occurs because the µ locus supplies the enhancer (or equivalent) activity required for expression of the enhancerless neo gene in targeted recombinants, whereas most positions of random vector integration in the hybridoma genome do not. The enrichment is sufficiently high that it permits targeted recombinants to be conveniently isolated under G418 selection by a plating procedure detailed in MATERIALS AND METHODS. As described below, the procedure ensured that each G418^R recombinant was derived from a single cell and that the product(s) of targeted vector integration were retained for molecular analysis.

Recombinant isolation and identification:
A total of 153 independent G418^R transformants were isolated from the plating procedure. Trypan blue staining revealed that of the 2 x 10⁷ hybridomas electroporated with pCµEn^-pal, ~50% survived. Thus, the enhancer-trap vector resulted in a frequency of stable transformants/cell of = 1.5 x 10^-5. Genomic DNA was prepared from each G418^R transformant, digested with EcoRI, and screened by Southern analysis. Using ³²P-labeled Cµ-specific probe F, the blots showed that 21 of the 153 G418^R transformants (~14%) bore the diagnostic 16.2- and 9.6-kb EcoRI fragments expected for the duplicate Cµ region structure generated by the targeted integration of one copy of pCµEn^-pal into the Sp6/HL chromosomal µ gene (Fig 1C). No other Cµ-hybridizing fragments were visible. The recovery of 21 targeted recombinants among the total of 3333 wells plated (see MATERIALS AND METHODS) is expected to follow the Poisson distribution and under these circumstances, the probability that the recombinants in a well actually derived from more than one independent recombinant is ~0.003.

Of the remaining G418^R transformants, 131 retained the 12.5-kb EcoRI endogenous µ gene fragment with evidence in many of one or more variable-sized fragments. These transformants likely represented cases of random integration of the targeting vector into the hybridoma genome. Examples of the µ gene structure in targeted recombinants and random transformants are presented in Fig 2. In the one remaining transformant the endogenous chromosomal µ gene was ~11 kb, suggesting that it had suffered a small deletion. It was not analyzed further.

View larger version (57K): In this window In a new window Download PPT slide   Figure 2. Analysis of µ gene structure in representative G418R transformants. Twenty micrograms of genomic DNA from the indicated G418R transformants was digested with EcoRI, electrophoresed through a 0.7% agarose gel, blotted to nitrocellulose, and hybridized with 32P-labeled Cµ-specific probe F (Fig 1C). The sizes (in kilobases) of the bands of interest are shown on the left of the blot while the sizes of relevant DNA marker bands (fragments of HindIII-digested DNA and the 1-kb DNA ladder; Bethesda Research Laboratories) are presented on the right. Included as controls are genomic DNAs from the wild-type Sp6/HL hybridoma in which the endogenous chromosomal µ gene is present on the 12.5-kb EcoRI fragment, the recombinant 49/9 (NG and BAKER 1999A ) in which targeted integration of a single copy of an enhancer-trap Cµ insertion vector has replaced the endogenous 12.5-kb EcoRI µ fragment with diagnostic fragments of 16.2 and 9.6 kb, and the mutant igm10 hybridoma (KOHLER and SHULMAN 1980 ; KOHLER et al. 1982 ), which has lost its chromosomal µ gene and serves as a negative control for probe specificity. G418R transformants 39 and 43 have lost the endogenous 12.5-kb EcoRI µ gene fragment and instead display the diagnostic 16.2- and 9.6-kb EcoRI fragments predicted for targeted vector integration (Fig 1C). In contrast, G418R transformants 33–36, 40–42, and 44 retain the 12.5-kb EcoRI µ gene fragment in addition to one or more variable size µ-hybridizing fragments suggesting random integration of the pCµEn-pal vector. Fragments less than ~8 kb were not retained in the gel.

In this study, the ~14% frequency of targeted recombinants obtained using a G418 concentration of 1.2 mg/ml was significantly higher (², P < 0.001) than the ~2% frequency obtained in an earlier gene targeting study where transformant selection was conducted at only 600 µg/ml G418 (NG and BAKER 1999A ). Control experiments revealed that the difference was due to a lower frequency of random transformants recovered at the higher G418 concentration (data not shown).

Palindrome insertions do not affect gene targeting efficiency:
Based on the number of hybridomas surviving electroporation, the absolute frequency of targeted vector integration was = 2.1 x 10^-6 events/cell. This was close to the absolute frequency of targeted vector integration reported previously (1.4 x 10^-6 events/cell; NG and BAKER 1999A ) for a similar enhancer-trap vector in which the chromosomal and vector-borne Cµ regions differed by six restriction enzyme site polymorphisms, of which two were at the same positions of 645 and 1199 bp relative to the DSB at XbaI as the palindrome insertions in this study. Further, an enhancer-trap insertion vector bearing the completely wild-type Cµ region targets the chromosomal µ locus with similar absolute frequency (1.74 x 10^-6; NG and BAKER 1998 ). These comparisons argue that the palindrome markers in pCµEn^-pal did not affect the frequency of recovering targeted recombinants.

Determination of Cµ region marker patterns in the recombinants:
The PCR and gel analysis methods used to analyze the Cµ region marker patterns in the recombinants have been documented previously (NG and BAKER 1999A , NG and BAKER 1999B ). As shown in Fig 1C, the PCR assay utilizes two sets of primers to generate specific 4765- and 4621-bp products from the 5' and 3' Cµ regions, respectively. Digestion of the 5' and 3' Cµ region PCR products with restriction enzymes specific for the chromosomal (SacI, AflII, ApaI) and vector-borne palindrome (NotI) sites yields diagnostic Cµ region fragments (Fig 3) that can be analyzed by standard gel electrophoresis. This analysis was performed on all 21 recombinants (data not shown) and the results are presented in Fig 4.

View larger version (13K): In this window In a new window Download PPT slide   Figure 3. Restriction enzyme fragment sizes from the 5' and 3' Cµ region PCR products. Primer pair AB9703/AB9745, specific for the 5' Cµ region and pair AB9703/AB9438, specific for the 3' Cµ region are represented by the half arrows. Digestion of the 5' Cµ region PCR product with the various restriction enzymes generates the indicated fragment sizes. The same Cµ fragment sizes are produced from the 3' Cµ region, except for those marked by an asterisk, which are 144 bp smaller. In the case of digestion with NotI, the indicated fragment sizes would be produced only if each position in the 5' and 3' Cµ regions contained the site. In the event one or more Cµ positions bore the corresponding chromosomal restriction enzyme site instead, the sizes of diagnostic NotI fragments would change according to the sum of one or more of the adjacent fragments. The diagram is not drawn to scale.

View larger version (26K): In this window In a new window Download PPT slide   Figure 4. Analysis of Cµ region marker patterns in the recombinants. The Cµ region positions bearing the various chromosomal restriction enzyme site markers (SacI, AflII, and ApaI) are denoted by the open circles while those bearing the vector-borne palindrome NotI site are indicated by the solid circle. Sectored positions exhibiting a mixed cleavage pattern with NotI and one of the corresponding chromosomal restriction enzymes are denoted by half-solid circles. Details of how the individual Cµ marker patterns were established are presented in the text.

For most Cµ region positions in the recombinants, either the vector-borne or chromosomal marker was present. However, in eight recombinants (13, 39, 69, 80, 87, 92, 98, and 138), one or more Cµ positions demonstrated only partial susceptibility to digestion with the chromosome and vector-specific restriction enzymes. As indicated above, the Poisson analysis showed that each recombinant was highly likely to have been derived from a single cell deposited in the culture well. Thus, each sectored colony very likely originated from an individual targeted cell in which hDNA had not been completely repaired prior to DNA replication and cell division. Accordingly, the Cµ region marker patterns in the cell populations comprising each sectored recombinant are expected to reflect those configurations present in each strand of the hDNA intermediate in the parental recombinant. In recombinants 39, 69, and 80, a single sectored site was present in the 5' Cµ region at position 1925 and gel analysis revealed cleavage of approximately one-half the PCR product with the NotI or ApaI restriction enzymes while all other positions were completely sensitive to cleavage with the other enzymes tested. This suggested that the two distinct cell populations composing each of these recombinants were present in approximately equal proportion and that they differed only with respect to the 5' Cµ region marker at position 1925. However, recombinants 13, 87, 92, 98, and 138 were sectored for two or more Cµ region positions. For these recombinants, it was necessary to determine whether vector-borne and/or chromosomal markers resided on the same strand of the hDNA intermediate. Each recombinant was cloned at 0.1 cell/well and individual subclones (originating from a single cell) were saved. When checked, Southern analysis of EcoRI-digested genomic DNA confirmed that the overall chromosomal µ gene structure in the subclones was identical to that in the parental recombinant (data not shown). For each subclone, the 5' and 3' Cµ region was amplified, the PCR products were digested separately with the various diagnostic restriction enzymes, and the Cµ fragment sizes were analyzed by gel electrophoresis. The number of subclones analyzed and their Cµ region marker patterns are presented in brackets in Fig 4. This information defined the marker configurations in each of the corresponding parental recombinants. With the possible exception of sectored recombinants 98 and 138, the two cell populations making up each of the remaining sectored recombinants were present in about the expected 50% frequency. The apparent disparity in the frequencies associated with the subpopulations of recombinants 98 and 138 can be explained if, by chance, a daughter cell(s) from an early division of the parental recombinant cell failed to survive.

As indicated in Fig 4, the various marker positions in the 5' and 3' Cµ region PCR products demonstrated complete sensitivity to cleavage with the palindrome-specific NotI enzyme, the restriction enzyme diagnostic of the corresponding chromosomal site, or a mixed cleavage pattern with both enzymes. These cleavage patterns suggested that the palindromic sequence in the duplex DNA of the PCR product did not undergo cruciform extrusion at detectable frequency, as the resulting structure, bearing only a single strand of the NotI recognition sequence, would be insensitive to NotI cleavage. The same result was obtained when genomic DNA isolated from the recombinants was examined (data not shown). Lack of extrusion of the palindrome may explain why it was not deleted from the chromosomal µ locus during growth of the recombinants.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In this study, we investigated mechanisms associated with the targeted integration of transferred vector DNA into the chromosomal µ locus in a murine hybridoma. This was accomplished using an enhancer-trap insertion vector in which the Cµ region of homology contained three identical 30-bp palindrome genetic markers distinguishable from the corresponding sites in the hybridoma chromosomal Cµ region. In S. cerevisiae, small palindromes in hDNA are thought to form hairpins that are capable of avoiding MMR (NAG et al. 1989 ) and, consequently, are useful in the study of recombination because they yield sectored sites that identify the location of hDNA. As discussed below, several interesting features of the mammalian gene targeting reaction were revealed through the use of the palindrome genetic markers in this study.

DSB processing usually does not involve extensive DSG formation:
To effect gene targeting, a DSB was introduced into the insertion vector at the unique XbaI site within the Cµ region of homology. As suggested from earlier studies (SHULMAN et al. 1990 ; HASTY et al. 1992 ; JIANG et al. 1992 ; PFEIFFER et al. 1994 ; RICHARD et al. 1997 ), a possible consequence of transferring the cut vector to the hybridoma cells is that some degradation of the DNA ends might occur resulting in formation of a DSG. However, in the majority (18/21) of recombinants, the first palindrome located 645 bp from the vector-borne DSB was present in at least one of the two Cµ regions. Therefore, if a DSG was formed at all, it must have been small, as its 3' border could not have exceeded 645 bp from the DSB.

In contrast, the Cµ region marker patterns in three recombinants (38, 39, and 69) are consistent with the possibility of some DSG formation. That is, a gap may have removed vector-borne palindromes at positions 645 and 1199 in recombinants 39 and 69 and from all three positions in recombinant 38 followed by gap repair using the chromosomal sequence as template.

With regard to the possibility of replacement of nucleotides lost from the cut vector ends, our previous studies using similar gene targeting vectors revealed that the XbaI cut site was always restored in targeted recombinants (NG and BAKER 1998 , NG and BAKER 1999A , NG and BAKER 1999B ). This suggests that the vector-borne DSB initiates gene targeting by providing homologous ends that synapse with the corresponding regions of the chromosome to permit recombination and proper DSB repair.

Crossover at or near the DSB:
The absence of extensive DSG formation provided the opportunity for the crossover event that integrated the vector to occur at or near the DSB at XbaI. In the event this occurred, all positions in the 5' and 3' Cµ region would contain the vector-borne palindrome and corresponding chromosomal restriction enzyme sites, respectively. Examination of the recombinants revealed exactly this marker pattern in nine of the 21 recombinants (3, 5, 28, 29, 56, 84, 107, 122, and 125; Fig 4).

Formation of hDNA in recombination:
The Cµ region marker patterns in the remaining 12 recombinants suggested the involvement of an hDNA intermediate in recombination. Direct evidence for hDNA formation was obtained from the position of sectored sites in 8 of the 21 recombinants (38%; 13, 39, 69, 80, 87, 92, 98, and 138). As each recombinant was highly likely to have originated from a single cell deposited in the culture well, sectored recombinants very likely arose as a result of the vector-borne palindrome forming a small hairpin when mismatched with the corresponding chromosomal restriction enzyme site in hDNA, a structure which then avoided MMR (NAG et al. 1989 ). The high frequency of colony sectoring was significantly different (², P = 0.02) from the much lower frequency (~6%) observed in an earlier gene targeting study (NG and BAKER 1999A ) in which the vector-borne Cµ region of homology bore six restriction enzyme site markers rather than palindromes and where two of the markers were at the same Cµ positions as the palindromes used here. This suggests that the hybridoma is normally MMR proficient, as simple mismatches involving restriction enzyme sites undergo efficient repair prior to DNA replication and cell division.

MMR of hDNA:
While the position of sectored sites offered the best evidence for hDNA, in other cases the evidence was indirect and based on the pattern of MMR toward either the vector-borne palindrome or the chromosomal sequence. To determine in which recombinants MMR had probably occurred, we considered only those in which the first palindrome at position 645 was retained in at least one of the Cµ regions and where the Cµ region marker patterns could not be explained on the basis of crossover at, or near, the vector-borne DSB. This ensured that formation and repair of a DSG was not a factor in determining the marker patterns and further, that any departure from the simple crossover pattern was likely due to MMR. Accordingly, in at least one of the Cµ regions in recombinants 13, 43, 80, 89, 92, 138, and 144, the results suggested that one or more mismatches had been repaired toward either the vector-borne palindrome or the corresponding chromosomal sequence. Recombinants 39 and 69 may represent cases where a DSG removed palindromes at Cµ positions 645 and 1199 leaving the third marker at position 1925 available for inclusion in hDNA. As suggested from Fig 4, one mismatch may have been corrected to the chromosomal sequence while the other was unrepaired. Examination revealed evidence for very short MMR tracts (<554 nucleotides) in some of the recombinants. For example, in the 5' Cµ region of recombinants 13 and 92, a single marker at position 1199 underwent MMR toward the vector-borne palindrome whereas adjacent markers either were left unrepaired or were repaired toward the chromosomal sequence. The possibility of biased short tract repair of palindrome loops has been suggested from a previous study (TAGHIAN et al. 1998). On the other hand, several recombinants in our study revealed evidence for much longer MMR tracts. For example, in the 5' and/or 3' Cµ in recombinants 13, 43, 80, 89, 92, 138, and 144, two or more adjacent markers were often repaired in the same direction, and in some cases the tract spanned all Cµ region markers over a distance of 1.9 kb. While consistent with MMR, an alternative explanation for adjacent chromosomal markers extending from the DSB is that of 5'3' exonucleolytic processing of the DSB to yield 3'-OH single-stranded tails (SUN et al. 1991 ; HUANG and SYMINGTON 1993 ; HENDERSON and SIMONS 1997 ), whereupon a palindrome(s) deleted from that vector strand would be replaced with those sequences from the chromosomal template by DNA repair synthesis.

The mechanisms by which the palindrome in hDNA avoids MMR (or, alternatively, is repaired) are not understood. In S. cerevisiae, it has been suggested that palindromes can uncouple the ATPase and mismatch binding activities of the Msh2p-Msh6p MMR complex, resulting in failure to repair the mismatch (ALANI 1996 ). Residual repair of palindrome loops might occur if Msh2p-Msh6p binding showed an incomplete response to ATP or, alternatively, repair of mismatches involving palindromes might involve the activity of MMR-related genes or other genes (WENG and NICKOLOFF 1998 ).

Our analysis of PCR-amplified and genomic Cµ regions suggests that the palindrome does not undergo cruciform extrusion in duplex DNA and is stably maintained during hybridoma growth. This agrees with studies in prokaryotes, where small palindromes of a few hundred base pairs or less are, in general, also replicated faithfully (LEACH 1994 ). In contrast, deletions are often associated with larger palindromes and may result from cruciform extrusion, posing a significant threat to genome stability (LEACH 1994 ; COLLICK et al. 1996 ; AKGUN et al. 1997 ).

Features of heteroduplex DNA formation:
The evidence above suggests that several recombinants provided support for hDNA beginning near the DSB and then extending outward into the Cµ region on at least the one side of the DSB examined. In some recombinants hDNA tracts were extensive, spanning all three Cµ region markers over a distance of at least 1.9 kb. This was clearly suggested by the position of sectored sites in recombinants 13, 92, and 138 as well as from the pattern of MMR in the Cµ regions of some of the recombinants referred to above. Formation of hDNA tracts of this size is consistent with meiotic recombination in S. cerevisiae where tracts often spanned a distance of 1.8 kb (DETLOFF and PETES 1992 ). In some recombinants the hDNA tracts were much shorter. For example, the sectored site at position 645 in both Cµ regions of recombinant 98 and at positions 645 and 1199 in both Cµ regions of recombinant 87, followed by vector-borne and chromosomal markers in the 5' and 3' Cµ regions, respectively, was consistent with short hDNA tracts initiating from the DSB prior to the single reciprocal crossover event that integrated the vector.

Formation of hDNA could be asymmetric or symmetric on the one side of the DSB. Asymmetric hDNA was suggested by the differences in the position of sectored sites between the 5' and 3' Cµ regions of recombinants 13 and 138 and also from the greater extent of sectoring in the 3' Cµ region of recombinant 138. In recombinants 87, 98, and 138, symmetric hDNA formation was suggested by unrepaired mismatches in equivalent positions in both the 5' and 3' Cµ regions. Although based on the repair patterns of simple restriction enzyme site mismatches, our previous studies also suggested that hDNA tracts formed during gene targeting could be extensive with evidence of both asymmetry and symmetry in their formation (NG and BAKER 1999A ).

Recombination models:
The presence and position of hDNA in the recombinants is consistent with recombinant generation according to the DSBR model of recombination (SZOSTAK et al. 1983 ) and its later revision (SUN et al. 1991 ). The vector-borne DSB may have been processed to form 3'-OH, single-stranded tails by 5'3' exonuclease activity (Fig 5A). Although not illustrated, this may have resulted in removal of a palindrome(s) from one vector strand. Assimilation of a 3'-OH tail into the intact homologous chromosomal duplex would provide the opportunity for replacement of any palindrome(s) deleted from the strand by DNA repair synthesis. Following D-loop formation and annealing of the second 3'-OH tail, asymmetric hDNA would have the opportunity to form near the DSB and extend into the Cµ region as suggested in several recombinants. The evidence for symmetric hDNA in some recombinants would be consistent with Holliday junction formation and branch migration (Fig 5B). Opposite sense cleavage of the Holliday junctions would result in targeted vector integration (Fig 5C).

View larger version (30K): In this window In a new window Download PPT slide   Figure 5. DSBR model for recombination. A–C depict a possible outcome of pCµEn-pal vector integration into the recipient hybridoma chromosomal µ gene according to the modified yeast DSBR model (SUN et al. 1991 ). The bracketed side panels (A'–D') illustrate a possible DSBR mechanism involved in generating recombinant 13. For clarity, the diagrams focus on only the Cµ region, and exons 1–4 (indicated in Fig 1A) are omitted. The vector-borne palindrome containing the NotI site is denoted by the solid circle while the corresponding restriction enzyme sites in the chromosomal Cµ region are denoted by the open circle. Hairpin formation following palindrome inclusion within hDNA is indicated by the lollipop symbol. The asterisk denotes a 3'-OH end formed by DSB processing. The integrated pSV2neo vector sequences are represented by the heavy lines. For further details, see DISCUSSION.

As depicted in Fig 5C, in each strand of the hDNA intermediate (prior to MMR) vector-borne markers are expected to reside in one Cµ region linked to the corresponding chromosomal markers in the other Cµ region. In fact, as revealed in Fig 4, the subcloning analysis of recombinants 87, 92, 98, and 138 revealed precisely the expected marker linkage if the individual strands of the hDNA intermediate were configured in this way. However, the mismatched sites in recombinant 13 display a different linkage pattern, one in which the 5' and 3' Cµ regions in one population of daughter cells contained the chromosomal ApaI and SacI markers, respectively, while the corresponding Cµ positions in the other daughter population bore the vector-borne NotI site. A similar linkage pattern was also seen for a single recombinant (27-1) in a previous study (NG and BAKER 1999A ). Two mechanisms might explain the linkage pattern in the latter recombinants. One mechanism invokes two recombination events. The first event, vector integration, generates the duplicate Cµ region structure in the recombinant µ locus, while the second results in reversal of the expected marker linkage pattern. The second event may have been either intrachromosomal homologous recombination between the recombinant Cµ regions or a second gene targeting event resulting in gene conversion in the absence of vector integration. A second mechanism is presented in Fig 5 (bracketed side panels A'–D'). In A', DSB processing removes a single NotI palindrome from one vector strand. Following strand invasion, repair synthesis replaces the deleted NotI site with the corresponding chromosomal sequence and incomplete MMR generates the intermediate in B'. Reverse migration of the right-hand Holliday junction (WHITBY et al. 1993 ) occurs as shown in C', reforming the NotI/ApaI mismatch at Cµ position 1925 bp. Opposite sense cleavage of Holliday junctions results in vector integration. If the mismatches are not complete, the recombinant 13 Cµ region marker pattern is formed (D'). Markers escaping MMR are enclosed by a rectangle.

Although the Cµ region marker patterns in the recombinants have been interpreted within the context of the DSBR model, it is possible that targeted vector integration occurred by invasion of the chromosome by a single, 3'-OH DNA strand as proposed in one-sided invasion (OSI) models of homologous recombination (BELMAAZA and CHARTRAND 1994 ). However, OSI can result in loss or gain of genetic information and these types of alterations were not normally observed in the insertion type targeting events reported in this or our previous studies (BAKER et al. 1988 ; NG and BAKER 1998 , NG and BAKER 1999A ).

To summarize, our gene targeting studies revealed the following: (i) the vector-borne DSB was usually not processed to a large DSG, (ii) a substantial fraction of recombinants (43%) were consistent with crossover at or near the DSB, and (iii) hDNA was an intermediate in generating the remaining recombinants. Individual hDNA tracts were either long or short and could form asymmetrically or symmetrically on the one side of the DSB examined. Clonal analysis of sectored recombinants was informative in establishing the linkage of vector-borne and chromosomal markers in each strand of individual hDNA intermediates.

	ACKNOWLEDGMENTS

We thank Leah Read and Erin Wever for providing expert technical assistance and David Leach for kindly supplying E. coli strain DL795. Marc Shulman and Philip Ng provided helpful comments on the manuscript. This research was supported by an operating grant from the Medical Research Council (MRC) of Canada (MT-14416) to M.D.B. and an MRC postdoctoral fellowship to J.L.

Manuscript received May 6, 1999; Accepted for publication November 10, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AKGÜN, E., J. ZAHN, S. BAUMES, G. BROWN, and F. LIANG et al., 1997  Palindrome resolution and recombination in the mammalian germ line. Mol. Cell. Biol. 17:5559-5570[Abstract].

ALANI, E., 1996  The Saccharomyces cerevisiae Msh2 and Msh6 proteins form a complex that specifically binds to duplex oligonucleotides containing mismatched DNA base pairs. Mol. Cell. Biol. 16:5604-5616[Abstract].

BAKER, M. D., N. PENNELL, L. BOSNOYAN, and M. J. SHULMAN, 1988  Homologous recombination can restore normal immunoglobulin production in a mutant hybridoma cell line. Proc. Natl. Acad. Sci. USA 85:6432-6436[Abstract/Free Full Text].

BAUTISTA, D. and M. J. SHULMAN, 1993  A hit-and-run system for introducing mutations into the immunoglobulin heavy chain locus of hybridoma cells by homologous recombination. J. Immunol. 151:1950-1958[Abstract].

BELMAAZA, A. and P. CHARTRAND, 1994  One-sided invasion events in homologous recombination events at double-strand breaks. Mutat. Res. 314:199-208[Medline].

COLLICK, A., J. DREW, J. PENBERTH, P. BOIS, and J. LUCKETT et al., 1996  Instability of long inverted repeats within mouse transgenes. EMBO J. 15:1163-1171[Medline].

DETLOFF, P. and T. D. PETES, 1992  Measurements of excision repair tracts formed during meiotic recombination in Saccharomyces cerevisiae.. Mol. Cell. Biol. 12:1805-1814[Abstract/Free Full Text].

GILBERTSON, L. A. and F. W. STAHL, 1996  A test of the double-strand break repair model for meiotic recombination in Saccharomyces cerevisiae.. Genetics 144:27-41[Abstract].

GOLDBERG, G. I., E. F. VANIN, A. M. ZROLKA, and F. R. BLATTNER, 1981  Sequence of the gene for the constant region of the µ chain of Balb/c mouse. Gene 15:33-42[Medline].

GROSS-BELLARD, M., P. QUDET, and P. CHAMBON, 1973  Isolation of high-molecular weight DNA from mammalian cells. Eur. J. Biochem. 36:32-38[Medline].

HASTY, P., J. RIVERA-PÉREZ, and A. BRADLEY, 1992  The role and fate of DNA ends for homologous recombination in embryonic stem cells. Mol. Cell. Biol. 12:2464-2474[Abstract/Free Full Text].

HENDERSON, G. and J. P. SIMONS, 1997  Processing of DNA prior to illegitimate recombination in mouse cells. Mol. Cell. Biol. 17:3779-3785[Abstract].

HUANG, K. N. and L. S. SYMINGTON, 1993  A 5'-3' exonuclease from Saccharomyces cerevisiae is required for in vitro recombination between linear DNA molecules with overlapping homology. Mol. Cell. Biol. 13:3125-3134[Abstract/Free Full Text].

JIANG, L., A. CONNOR, and M. J. SHULMAN, 1992  Effects of mutation position on frequency of marker rescue by homologous recombination. Mol. Cell. Biol. 12:3609-3613[Abstract/Free Full Text].

KÖHLER, G. and M. J. SHULMAN, 1980  Immunoglobulin M mutants. Eur. J. Immunol. 10:467-476.

KÖHLER, G., M. J. POTASH, H. LEHRACH, and M. J. SHULMAN, 1982  Deletions in immunoglobulin mu chains. EMBO J. 1:555-563[Medline].

LASKO, D., W. CAVENEE, and M. NORDENSKJÖLD, 1991  Loss of constitutional heterozygosity in human cancer. Annu. Rev. Genet. 25:281-314[Medline].

LEACH, D. R. F., 1994  Long DNA palindromes, cruciform structures, genetic instability and secondary structure repair. Bioessays 16:893-900[Medline].

NAG, D. K. and T. D. PETES, 1990  Genetic evidence for preferential strand transfer during meiotic recombination in yeast. Genetics 125:753-761[Abstract].

NAG, D. K. and T. D. PETES, 1991  Seven base-pair inverted repeats in DNA form stable hairpins in vivo in S. cerevisiae.. Genetics 129:669-673[Abstract].

NAG, D. K., M. A. WHITE, and T. D. PETES, 1989  Palindromic sequences in heteroduplex DNA inhibit mismatch repair in yeast. Nature 340:318-320[Medline].

NG, P. and M. D. BAKER, 1998  High efficiency, site-specific modification of the chromosomal immunoglobulin locus by gene targeting. J. Immunol. Meth. 214:81-96[Medline].

NG, P. and M. D. BAKER, 1999a  Mechanisms of double-strand-break repair during gene targeting in mammalian cells. Genetics 151:1127-1141[Abstract/Free Full Text].

NG, P. and M. D. BAKER, 1999b  The molecular basis of multiple vector insertion by gene targeting in mammalian cells. Genetics 151:1143-1155[Abstract/Free Full Text].

PETES, T. D., R. E. MALONE and L. S. SYMINGTON, 1991 Recombination in yeast, pp. 407–521 in The Molecular and Cellular Biology of the Yeast Saccharomyces: Genome Dynamics, Protein Synthesis and Energetics, edited by J. R. BROACH, J. R. PRINGLE and E. W. JONES. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

PFEIFFER, P., S. THODE, J. HANCKE, and W. VIELMETTER, 1994  Mechanisms of overlap formation in nonhomologous end joining. Mol. Cell. Biol. 14:888-895[Abstract/Free Full Text].

PORTER, S. E., M. A. WHITE, and T. D. PETES, 1993  Genetic evidence that the meiotic recombination hotspot at the HIS4 locus of Saccharomyces cerevisiae does not represent a site for a symmetrically processed double-strand break. Genetics 134:5-19[Abstract].

RESNICK, M. A., 1976  The repair of double-strand breaks in DNA: a model involving recombination. J. Theor. Biol. 59:97-106[Medline].

RICHARD, M., N. GUSEW, A. BELMAAZA, and P. CHARTRAND, 1997  Homologous junctions formed between vector and human genomic LINE-1 elements as a result of one-sided invasion. Somatic Cell Mol. Genet. 23:75-81[Medline].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual, Ed. 2., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SHULMAN, M. J., L. NISSEN, and C. COLLINS, 1990  Homologous recombination in hybridoma cells: dependence on time and fragment length. Mol. Cell. Biol. 10:4466-4472[Abstract/Free Full Text].

SUN, H., D. TRECO, and J. W. SZOSTAK, 1991  Extensive 3'-overhanging, single-stranded DNA associated with meiosis-specific double strand breaks at the ARG4 recombination initiation site. Cell 64:1155-1161[Medline].

SZOSTAK, J. W., T. L. ORR-WEAVER, R. J. ROTHSTEIN, and F. W. STAHL, 1983  The double-strand-break repair model for recombination. Cell 33:25-35[Medline].

TAGHIAN, D. G., H. HOUGH, and J. A. NICKOLOFF, 1988  Biased short tract repair of palindromic loop mismatches in mammalian cells. Genetics 148:1257-1268[Abstract/Free Full Text].

WENG, Y. and J. A. NICKOLOFF, 1998  Evidence for independent mismatch repair processing on opposite sides of a double-strand break in Saccharomyces cerevisiae.. Genetics 148:59-70[Abstract/Free Full Text].

WHITBY, M. C., L. RYDER, and R. G. LLOYD, 1993  Reverse branch migration of Holliday junctions by RecG protein: a new mechanism for resolution of intermediates of recombination and DNA repair. Cell 75:341-350[Medline].

This article has been cited by other articles:

				R. D. McCulloch and M. D. Baker Analysis of one-sided marker segregation patterns resulting from Mammalian gene targeting. Genetics, March 1, 2006; 172(3): 1767 - 1781. [Abstract] [Full Text] [PDF]

				E. C. Birmingham, S. A. Lee, R. D. McCulloch, and M. D. Baker Testing Predictions of the Double-Strand Break Repair Model Relating to Crossing Over in Mammalian Cells Genetics, November 1, 2004; 168(3): 1539 - 1555. [Abstract] [Full Text] [PDF]

				L. R. Read, S. J. Raynard, A. Ruksc, and M. D. Baker Gene repeat expansion and contraction by spontaneous intrachromosomal homologous recombination in mammalian cells Nucleic Acids Res., February 20, 2004; 32(3): 1184 - 1196. [Abstract] [Full Text] [PDF]

				S. J. Raynard and M. D. Baker Incorporation of Large Heterologies Into Heteroduplex DNA During Double-Strand-Break Repair in Mouse Cells Genetics, October 1, 2002; 162(2): 977 - 985. [Abstract] [Full Text] [PDF]

				S. J. Raynard, L. R. Read, and M. D. Baker Evidence for the Murine IgH {micro} Locus Acting as a Hot Spot for Intrachromosomal Homologous Recombination J. Immunol., March 1, 2002; 168(5): 2332 - 2339. [Abstract] [Full Text] [PDF]

				M. D. Baker and E. C. Birmingham Evidence for Biased Holliday Junction Cleavage and Mismatch Repair Directed by Junction Cuts during Double-Strand-Break Repair in Mammalian Cells Mol. Cell. Biol., May 15, 2001; 21(10): 3425 - 3435. [Abstract] [Full Text]

				B. Elliott and M. Jasin Repair of Double-Strand Breaks by Homologous Recombination in Mismatch Repair-Defective Mammalian Cells Mol. Cell. Biol., April 15, 2001; 21(8): 2671 - 2682. [Abstract] [Full Text]

				J. Li, L. R. Read, and M. D. Baker The Mechanism of Mammalian Gene Replacement Is Consistent with the Formation of Long Regions of Heteroduplex DNA Associated with Two Crossing-Over Events Mol. Cell. Biol., January 15, 2001; 21(2): 501 - 510. [Abstract] [Full Text]

				J. Li and M. D. Baker Mechanisms Involved in Targeted Gene Replacement in Mammalian Cells Genetics, October 1, 2000; 156(2): 809 - 821. [Abstract] [Full Text]