A model for testing recombinogenic sequences in the mouse germline
A model for testing recombinogenic sequences in the mouse germlineMary Ellen Moynahan, Ercan Akgün1 and Maria Jasin1,*
Department of Medicine and 1Cell Biology and Genetics Program, Memorial Sloan-Kettering Cancer Center and Cornell University Graduate School of Medical Sciences, 1275 York Ave, New York, NY 10021, USA
Received February 14, 1996;Revised and Accepted April 25, 1996
Homologous recombination is a conserved process of genetic exchange generated by homologous pairing of nucleotides. Species diversity and gene evolution are dependent on the outcomes of recombination during germ cell development, yet systems to study mammalian germline recombination, especially those with applications to human genetics, are not well developed. We report on a transgenic mouse system designed to study recombination within test sequences in the male germline utilizing an intron-interrupted lacZ reporter gene. [beta]-galactosidase positive sperm are detected and quantitated by flow cytometry using fluorogenic substrates. Examination of recombination within a 1.7 kb repeat of test sequences derived from the human glycophorin breakpoint cluster region detects approximately 0.04-0.09% fluorescent sperm. Confirmation that these sperm result from recombination in the germline comes from histochemical staining of testicular cells, examination of spliced mRNA, and PCR analysis of sorted sperm populations. The system is readily adaptable to studies of other sequences reported to have elevated levels of recombination, including those implicated in human genetic disease. Investigations of the molecular basis for genomic instability at specific chromosomal locations may yield important insights into mechanisms of chromosomal loss and rearrangements.
Genetic exchange during germ cell development in mammals determines the extent of species diversity through gene evolution, and in individuals determines heritable traits that have direct implications on health and disease. Recombination rates differ along the lengths of chromosomes and even at identical chromosomal regions between males and females. Overall, females exhibit a higher rate of recombination than males, although some chromosomal regions recombine more frequently in males than in females (1 ). Homologous recombination during meiosis plays a critical part in ensuring proper chromosome segregation through homolog pairing. The requirement for meiotic recombination for proper chromosome segregation is evidenced by the obligatory exchange that occurs in the short pseudoautosomal region in the X-Y chromosome pair (2 ).
Outcomes of homologous recombination include both reciprocal (crossing-over) and non-reciprocal (gene conversion) exchanges, processes that have been studied extensively in yeast (3 -5 ). The ability to score each meiotic product and the availability of recombination deficient yeast mutants have made advances possible (6 ). Recombination between alleles is at its highest during meiosis, when homologous chromosomes are paired at the first meiotic prophase. Although mitotic recombination is much lower than meiotic recombination, it plays a key part in the repair of damaged DNA (7 ).
The molecular mechanisms of both spontaneous and induced rearrangements have important implications for human genetic disease. The consequence of inappropriate intrachromosomal recombination has been recently highlighted by the inversions on the long arm of the human X chromosome at the factor VIII gene, leading to severe hemophilia, and at the IDS gene resulting in Hunter's syndrome. These inversions are thought to arise by intrachromosomal recombination in male germ cells (8 -10 ).
In cultured mammalian cells, intrachromosomal gene conversion and crossing-over, as well as gene targeting between a chromosome and exogenous homologous sequences, have been detected in studies utilizing selectable markers (11 ). The reliance on cultured cells to investigate recombination is clearly not ideal, especially when considering germline events. However, there is no currently available culture system to study germline events in vitro, although advances have recently been made in the culture of male germ cells (12 ). Thus, most studies of mammalian recombination processes have been limited to classical genetic approaches. More recently, gene conversion within lacZ transgenes has been documented in mouse spermatids (13 ).
The non-random nature of meiotic recombination has fueled the study of this process to determine the molecular mechanisms that influence the events. Specific sequences, chromosomal location, chromatin structure, upstream and downstream effectors and levels of transcription have been shown to affect recombination in yeast (14 -16 ). Regions with both suppressed and enhanced recombination have been identified at loci in yeast, Drosophila and mice (14 ,17 -19 ). Well defined `hot spots' include regions in the mouse H-2 complex, where sex-specific and haplotype-specific differences exist (20 ,21 ). Analysis of Huntington's disease (HD) pedigrees and elucidation of the location of the HD gene on human chromosome 4 suggest a human meiotic recombination hot spot (22 ). Further analysis by individual sperm typing confirmed an elevated recombination frequency in this region spanning 280 kb (23 ).
Several hotspots of recombination are contained in physically small distances which make them amenable for study. Examples include the extremely restricted region (ERR) in intron 2 of the RARA gene and a 1.2 kb fragment in intron 6 of the MLL gene in humans (24 ,25 ) and the E[beta] and A[beta]3/A[beta]2 hotspots in mice (20 ,26 ). Recombinogenic sequences have been inferred in evolutionarily conserved multigene families including globin, haptoglobin and glycophorin gene variants (27 -30 ). DNA sequence analysis reveals limited homology to recombination hotspots in prokaryotes and also predicts unstable DNA conformations such as tetrameric repeats, inversions and telomeric sequence repeats (26 ,31 -34 ). Functional testing of these sequences is lacking and their significance remains speculative.
The human glycophorin locus has been particularly fruitful for the identification of recombinants, as the locus encodes blood group antigens that are easily screened (29 ). The glycophorin genes are closely linked on chromosome 4 and are highly conserved from the leader sequence through exon 5, making it likely that they arose from gene duplication (35 ). Several variants have been identified that result from homologous recombination between the glycophorin A gene (GPA) and the glycophorin B gene (GPB) (36 ). The recombination breakpoints cluster in a conserved region of the genes involving intron 2 through intron 3.
We report a unique model system designed to detect germline recombination in the mouse at the human glycophorin breakpoint cluster region. A direct repeat of glycophorin sequences is introduced within an intron in a lacZ reporter gene, disrupting its expression in the male germline. Crossing-over within the glycophorin repeat restores lacZ expression, giving rise to [beta]-galactosidase positive sperm which are detected by flow cytometry using fluorogenic substrates. Fluorescent sperm are detected as approximately 0.04-0.09% of the total sperm population. PCR analysis of a sorted sperm population and examination of spliced testes mRNA provides support at the molecular level for crossing-over.
The Escherichia coli lacZ gene has been used extensively as a reporter gene in mammalian tissue culture and in mice. We wanted to investigate crossing-over within sequences that were implicated in higher frequencies of mammalian recombination, yet utilize the versatility of the lacZ reporter gene to detect and analyze these events. In order to accomplish this, we introduced an intron with splice donor and acceptor sites into the lacZ gene to allow greater flexibility in the design of recombination substrates. A repeat of the recombination test sequence was then placed within the intron (Fig. 1 a). In order to disrupt lacZ gene expression, a second splice acceptor site and a transcription termination signal were placed between the repeated sequences. Thus, expression of an intact lacZ mRNA is dependent upon crossing-over within the test sequences, as either an intrachromatid cross-over (Fig. 1 a) or as a sister chromatid exchange (not shown).
The SV40 promoter was cloned upstream of the lacZ gene to monitor [beta]-galactosidase activity in tissue culture cells during sequential modification of the gene (A2, Fig. 2 ). In addition, the coding region of this lacZ gene and all of the modified genes described below have an SV40 nuclear localization signal, as we have found efficient retention of this modified form of [beta]-galactosidase in mature sperm.
Transgenic mice containing the recombination substrate were constructed by injecting the pHDH transgene (Fig. 1 b) into fertilized mouse eggs as a 10.5 kb Sphl-Kpnl fragment. Six founder animals were obtained from 34 G0 offspring as determined by Southern blot analysis of tail tip DNA. Restriction digestion with BamHl and hybridization to a 32P-labeled lacZ probe revealed two common patterns of transgene insertion into the genome, an 8.4 kb band and a 5.0 kb band (Fig. 3 ). The 8.4 kb BamHl fragment was indicative of an unrecombined transgene integration, whereas the 5.0 kb BamHl fragment was indicative of transgene recombination at the glycophorin repeats prior to integration. Although unrecombined transgenes predominate, recombined transgenes are a significant portion of the total. Linear DNA transfected into tissue culture cells by a variety of methods is known to be highly recombinogenic (11 ). Our data show that DNA injected into fertilized eggs is also recombinogenic. The recombinogenicity may result in the formation of large tandem arrays that are typically found at integration sites (43 ).
From the six founders, four (mice 10, 16, 102 and 103) and progeny derived from them have the fragments expected from both recombined and unrecombined transgenes (Fig. 3 b). Female founder 12 also contains both recombined and unrecombined transgenes (Fig. 3 c). However, there are two transgene integration sites in this founder as determined by breeding analysis. Two sublines, 12U and 12R, were generated from her progeny. Subline 12R contained the bulk of the transgene copies, some of which were recombined and some of which were unrecombined. We have used the lines with recombined transgenes as controls in subsequent experiments to monitor lacZ expression. Subline 12U contained only unrecombined transgenes, with a copy number of approximately four to six.
Line 18, in addition to subline 12U, was generated with unrecombined transgenes. In addition to the 8.4 kb unrecombined transgene fragment, there is a second larger band (Fig. 3 d). Founder mouse 18 is mosaic for the transgene integration, as it transmitted the transgene to fewer than 50% of its progeny, and tail tip DNA from the progeny show a stronger hybridization signal than DNA from mouse 18 itself. It is estimated that the unrecombined transgenes are present in two or three copies in the progeny from 18.
Mice from lines 12U and 18 are appropriate for studies of germline recombination events, as they have transgenes that have not recombined at the glycophorin sequences. During breeding, no transgene effect on fertility was noted. A total of 68 transgenic progeny were obtained from breeders in line 12U and 89 progeny from line 18 and none of these mice were found to have a recombined transgene. (A sample of the breeding analysis is shown in Fig. 3 d.) Thus, recombination in the germline between the 1.7 kb glycophorin repeats is less than 10-2.
In order to verify that the lacZ reporter gene was appropriately expressed, northern blot analysis was performed on RNA from testes of transgenic mice (Fig. 4 ). Transgenic mice from the previously described mouse line 5 (38 ), which contains an uninterrupted prm1IlacZ transgene, were used as a control. A 3.1 kb lacZ transcript is seen from testes of transgenic mice from this line using either a 5' lacZ probe (Fig. 4 ) or a 3' lacZ probe (data not shown).
Mice from lines 12U and 18 containing only unrecombined transgenes do not show the full length transcript by northern analysis. Instead, they have a much smaller transcript that is detected with the 5' lacZ probe (Fig. 4 ). The transcript does not hybridize to the 3' lacZ probe (data not shown). These results are consistent with initiation of expression at the prm1 promoter, followed by premature transcription termination at the globin termination signal. Thus, the locus is actively transcribing in both of these lines and is being efficiently terminated.
Mice from lines 12R and 10, containing both recombined and unrecombined transgenes, have both the full length lacZ transcript and the prematurely terminated one, consistent with expression from the recombined and unrecombined transgenes, respectively (Fig. 4 ). The full length transcript expressed from the recombined transgene is the same size as that from line 5, demonstrating appropriate splicing of the single glycophorin repeat at the Ad1 splice sites.
We have previously utilized flow cytometry of transgenic mouse sperm to detect [beta]-galactosidase expression in the mouse germline (38 ). Although sperm are not transcriptionally active, enough [beta]-galactosidase is retained from expression during the previous haploid spermatid stage to detect protein activity. A number of fluorogenic substrates for [beta]-galactosidase are available that produce cellular fluorescence. Although we had previously used C12FDG (5-dodecanoyl-aminofluorescein di-[beta]-D-galactopyranoside), we found that C8FDG (5-octanoyl-aminofluorescein di-[beta]-D-galacto- pyranoside) staining with a chloroquine pre-incubation gives a higher level of cellular fluorescence for positive sperm and lower background staining of negative sperm.
Sperm were collected from the caudal epididymus of 2-4 month old adult males that were hemizygous for the transgene. They were stained with C8FDG and analyzed by flow cytometry. Sperm from nontransgenic mice showed little background fluorescence, whereas sperm from males of lines 12R and 10, both of which contain recombined transgenes, stained positive for [beta]-galactosidase activity (Fig. 5 ; data not shown). We have also shown that the other mouse lines containing recombined transgenes have [beta]-galactosidase activity in sperm (data not shown).
Testicular and epididymal [beta]-galactosidase activity has been verified for most of these lines by ONPG assays of tissue extracts (data not shown). These analyses demonstrate that the glycophorin segment embedded within the Adl intron does not interfere with [beta]-galactosidase expression. In general, mice containing the recombined transgenes were found to express [beta]-galactosidase less well than our previously reported prm1/lacZ transgene (38 ). The lower level of activity is likely due to the presence of SV40 sequences, rather than prm1 sequences, 3' to the lacZ gene, as intronless lacZ transgenes containing SV40 3' sequences also express less well than the previously reported prm1/lacZ transgene (data not shown). The 3' prm1 sequences have been identified to play a part in posttranscriptional expression (37 ).
To examine recombination events occurring within the male germline, sperm from transgenic mouse lines which contain unrecombined transgenes were analyzed for [beta]-galactosidase activity. Overall, sperm from lines 18 and 12U gave a profile very similar to that of nontransgenic mice, such that the mean fluorescence intensity of the population is almost identical (Fig. 5 ; Table 1 ). This concurs with the lack of observed recombinants during breeding.
To view a small number of sperm with increased fluorescence better, the y-axis sperm counts were maximally set to 26 (Fig. 5 ). A small percentage of sperm from lines 18 and 12U show greater fluorescence over the bulk sperm population. Only 0.02-0.03% of sperm from control nontransgenic mice have this level of fluorescence. By contrast, between 0.07 and 0.11% positive sperm for mice from lines 18 and 12U are observed (Table 1 ). For an analysis of approximately 40 000 sperm, this corresponds to about 30-40 sperm from lines 18 and 12U versus 10 sperm from the nontransgenic mice. No significant difference in cross-over frequency was detected between line 18, which has two to three transgene copies, and line 12U, which has four to six copies. Simplistically, a one to one concordance between [beta]-galactosidase positive sperm and recombination would mean that 0.04-0.09% of sperm had a recombined transgene. Our analysis of the two mouse lines have consistently given similar results.
The flow cytometric detection of [beta]-galactosidase positive sperm implies that [beta]-galactosidase positive spermatids should be found in the testes. In order to confirm this, X-gal staining of testicular cells was performed on testes sections and testicular cell preparations. X-gal staining of testes sections from lines 12U and 18 revealed occasional individual blue cells and small clusters of blue cells near the seminiferous tubule lumen, consistent with the staining of haploid spermatids (Fig. 6 a-e). Nontransgenic testes did not exhibit any such staining (data not shown). As expected, control animals containing recombined lacZ transgenes revealed testes with most of the tubules containing blue cells ringed around the seminiferous tubule lumen (Fig. 6 f). As compared with control prm1/lacZ lines (38 ), the intensity of staining was diminished in line 10, consistent with the lower level of RNA produced (Fig. 4 ).
To quantitate the testicular staining, enriched spermatid cell populations were prepared (44 ) from testes of line 18 mice and stained with X-gal. Blue cells were counted using microscopy. Spermatids were identified by morphology and made up the majority of cells in the preparation. In one experiment, 0.035% of the cells were deep blue (38 blue spermatids/107 520 total), whereas no blue spermatids were found in 167 040 cells from a nontransgenic mouse. By contrast, testes from a 12R mouse containing a recombined transgene gave 56.75% blue cells. A similar result was obtained in a second experiment. These results are similar to those obtained by flow cytometry of sperm.
Fluorescent staining of sperm and X-gal staining of spermatids have demonstrated that [beta]-galactosidase is present in a small percentage of cells in transgenic mice containing the unrecombined transgene. The presence of [beta]-galactosidase activity predicts that properly spliced lacZ mRNA is being produced in a small proportion of the cells after recombination. To test this, PCR analysis was performed on reverse transcribed RNA isolated from testes (Fig. 7 a). A PCR product of the correct size was obtained from RNA from line 12U (Fig. 7 b). As expected, there was less product from line 12U than from lines 16 and 10 which contain already recombined transgenes. With fewer amplification cycles and Southern blotting to detect the product, this differential appears even greater (data not shown).
Because of the large number of sperm available from one transgenic animal, it is possible to detect rare events and pursue molecular analysis of them following FACS sorting. To confirm the recombination event predicted by flow cytometry analysis of sperm, DNA analysis was performed on enriched populations of [beta]-galactosidase positive sperm. PCR has previously been used to confirm certain recombination events in single sperm (23 ). However, we expected that single sperm analysis with repeated sequences would lead to PCR artifacts, as amplified intermediates from an unrecombined glycophorin repeat have the potential to form `recombinants' containing a single glycophorin segment during reannealing steps in vitro. Thus, we sought to enrich for populations of [beta]-galactosidase positive sperm and then size-select the sperm DNA to identify the recombination product.
After staining with the fluorogenic C8FDG substrate, FACS sorting was performed in separate experiments on sperm from a line 12U mouse and a line 18 mouse. Sperm falling in the upper 0.6-0.9% of the fluorescence range were sorted, a calculated enrichment of 110-170-fold. Of 1.4 * 107 sperm, 1.2 * 105 sperm were collected from a size-gated cell population (+ sort; Fig. 8 ). Sperm were also collected from the bulk unstained population for use as a negative control (- sort; Fig. 8 ). Given the additional steps involved in sperm DNA isolation, a purified population was not attempted.
We have developed a lacZ system for the testing of recombinogenic sequences in the germline of male mice. In this model system, a direct repeat of the recombination test sequences, separated by a transcription terminator, disrupts expression of the lacZ reporter gene. Recombination within the repeat restores lacZ expression, resulting in [beta]-galactosidase positive sperm. Positive sperm are quantitated by flow cytometry after incubation with fluorogenic substrates.
Our test sequences are 1.7 kb in length and are derived from the human glycophorin B gene (GPB). Recombination to restore lacZ expression can occur by unequal crossing-over within the glycophorin repeat. With the FACS assay, we detect approximately 0.04-0.09% fluorescent sperm. This number is similar to the number of blue spermatids detected by X-gal staining of testicular cell preparations. Molecular confirmation of recombination is supported by examination of testes RNA, in which the appropriately spliced RNA is detected by RT-PCR, and from PCR analysis of sorted sperm populations. Absolute quantitation may require further molecular studies, considering that cytoplasmic sharing of [beta]-galactosidase can occur between adjacent spermatids.
We chose the human GPB gene sequences as the first test of this model system because it contains a number of recombination breakpoints (29 ). The previously mapped recombinants are derived from both gene conversion and unequal crossing-over between the GPB and GPA genes. Recombinants of both types are detected because they give rise to variant cell surface glycophorin proteins on erythrocytes which are readily identified in studies of human populations by serologic testing (45 ). We used an identical repeat of the GPB breakpoint cluster region, despite 97% regional homology to GPA, because it has been reported that small heterologies can decrease the frequency of recombination in mammalian tissue culture experiments (46 -48 ). Therefore, in this original study we felt that the use of identical repeats would optimize conditions for crossing-over.
Comparison of the recombination frequencies within the GPB repeat obtained in this study and other available data is imperfect. Human studies reveal that glycophorin variants are rarely detected in Western countries. However, in non-Western countries, particular variants are found in a relatively high proportion of the population. For example, the Sta variant, which is derived from a crossover in intron 3, has an incidence of 6.5% in Japan. Similarly, the MiIII variant, which is derived from a gene conversion with breakpoints in pseudoexon 3 and intron 3, has an incidence of 10% in Thailand (29 ). In mouse studies, the interchromosomal recombination frequency at the 2 kb E[beta] hotspot is 0.1%, whereas non-hotspot sequences recombine at 0.02% over several hundred kilobases (33 ). This suggests that a measure of 0.04-0.09% recombinants may reflect enhanced recombination in the GPB sequence, although the relationship of intrachromosomal recombination to interchromosomal events is not clearly established.
Within the glycophorin breakpoint cluster region are a number of sequence elements speculated to affect recombination in this and other systems. These include six nearby repeats of sequences related to the phage lambda chi site, short inverted repeats, polypurine tracts and regions of high AT content (32 ). The 1.7 kb GPB homology unit can be modified to test the role of these elements in recombination. Replacement of the sequences from the breakpoint cluster region with neutral sequences (e.g. from a non-recombinogenic intron) will address this question. The exact role of hotspot sequences in recombination has not been formally elucidated. It remains to be determined whether these sequences enhance the frequency of recombination in an interval or rather direct the resolution of recombination intermediates to the interval.
Our experimental design is the second instance in which a lacZ-based assay has been used to monitor germline recombination events in the mouse. Previously, lacZ gene conversion substrates were introduced into transgenics and recombination was quantitated by X-gal staining of spermatids (13 ). In this case, recombination was estimated to be at least 10-fold higher, possibly due to the type of recombination, gene conversion rather than crossing-over, being measured. The different assay system, flow cytometry of sperm versus X-gal staining of spermatids, is not expected to account for the difference, as we have confirmed our flow cytometric results with X-gal staining of spermatids. We have noted a difference, however, in the X-gal staining pattern of spermatids in testes sections. Others have reported blue cells in large clusters near the seminiferous tubule lumen (13 ) and we have also observed such a staining pattern in mice containing a gene conversion substrate (E. Akgün and M. Jasin, in preparation). Large clusters of positive cells strongly suggest the occurrence of mitotic events early in germ cell development. By contrast, the small clusters of blue cells and the individual blue cells obtained with our crossover substrate suggest that recombination occurred later in germ cell development, perhaps during meiosis. From these initial reports, it is difficult to conclude that the altered staining pattern is due to a different timing of gene conversion and unequal crossing-over or to other factors.
Detection of recombinants by flow cytometry of fluorescent sperm has advantages over X-gal staining of spermatids. Using flow cytometry, recombinants can be measured easily from an individual mouse epididymus from which a pure sperm population can be obtained. As many as 105 sperm can be analyzed within a few minutes by flow cytometry and 107 sperm can be sorted in a few hours, whereas counting blue spermatids requires lengthy microscopic examination. In the future, flow cytometric sorting of sperm, followed by in vitro fertilization, may make it possible to create progeny with defined genotypes. Such an approach would assist recombination analysis and may have applications to other studies requiring genome manipulations.
We expect that our intron-interrupted recombination substrate has general utility to measure recombination within other non-lacZ sequences. In addition to the GPB sequence, we have cloned a recombination hotspot defined in the intergenic region of the A[beta]3-A[beta]2 histocompatibility genes (34 ) into the lacZ intron and have found that the construct expresses high levels of [beta]-galactosidase in cultured cells (P. Rouet and M. Jasin, unpublished results). Other known recombination hotspots or specific sequences implicated in recombinogenic regions could also be tested. Refinements of our system include a gene targeting approach to compare chromosomal location effects and to detect interchromosomal cross-overs, which are expected to be at their highest level during meiosis. Once mice defective for genes involved in meiotic recombination have been derived, the lacZ recombination substrates can be examined in these genetic backgrounds.
Plasmid a-2 was previously described (38 ). It consists of a 3.1 kb fragment of the E.colilacZ gene containing a simian virus 40 (SV40) nuclear localization signal inserted into the mouse protamine 1 gene (prm1). Upstream of lacZ there are 1.7 kb of prm1 promoter sequences and 0.1 kb of prm1 5' untranslated sequences. Downstream of lacZ is the remainder of the prm1 gene, including the intron and 3' untranslated sequences. To test for lacZ expression in tissue culture cells, the StuI-SpeI fragment of the prm1 promoter (49 ) was replaced with the 401 bp NaeI-AvrII promoter/enhancer fragment of SV40, creating plasmid A2. All other constructs were similarly modified to test for lacZ expression in tissue culture.
The prm1 intron, 3' untranslated region, and polyadenylation signal from plasmid A2 was replaced with an SV40 polyadenylation signal to create plasmid A4. The SV40 poly(A) site was contained on a 134 bp HpaI-KpnI fragment derived from the plasmid Tag, which has the SV40 early region cloned into pUC. This fragment was cloned into the NaeI-KpnI sites of plasmid a-2. The HindIII site from this plasmid was destroyed by Klenow polymerase treatment. To place the intron within the lacZ gene, a 234 bp sequence from the first intron of the tripartite leader segment of the adenovirus (Ad1) major late transcription unit (39 ) was inserted into a unique Bsu36I site in the lacZ gene, creating plasmid A4Ad1. The intron sequence was PCR amplified from pBSAd1 (gift from M. Konarska, Rockefeller University) with primers 5'GGCCTCAGGAAAAAAAAGGGACAGG and 5'CT CCCTGAGGTGAGTACTCCCTCTCA which contain Bsu36I restriction sites. The cloning step conserved the splice donor and acceptor sites. The glycophorin homology unit was PCR amplified from intron 2 to intron 3 of the human glycophorin B (GPB) gene and was generously supplied by O. Blumenfeld and C-H. Huang (Albert Einstein College of Medicine) (36 ). It was amplified on two fragments, a 5' HindIII-XbaI fragment using primers GPB-P15' (5'CTCAAGCTTGGCTCCGAAA- GATTTTTGTG) and G - P25' (5'CTAACCTTCCCTCTCAGTCCA) and a 3' XbaI-HindIII fragment using primers GPB-P13' (5'CTCAAGCTTCCAAATAAGAAAGACATGTGC) and GP- P23' (5'CCAAAG CCCA TATAG- CACC). The amplified fragments were subcloned into pUC18 at HindIII-XbaI sites, creating pGPB5' and pGPB3'. The HindIII-XbaI fragments from pGPB5' and pGPB3' were inserted into the unique HindIII site in the Ad1 intron of plasmid A4AdI, creating plasmid pH, which contained one 1.7 kb homology unit. The splice acceptor site from the adenovirus E1A gene and the mouse [beta]maj-globin gene transcription termination sequence (41 ) were PCR amplified from plasmid DEF (gift from E. Falck-Pedersen, Cornell Medical School) with primers 5'GCTCTAGAGCGGCCGCAGAGTGGTGGT TTGGTG and 5'GGAGGTGTGTTAGAAGCACCGG. The 1.8 kb amplified fragment was subcloned as a NotI-BglII fragment into the NotI-BamHI sites of pBS (Stratagene) yielding plasmid pBSDEF. The 1.7 kb GPB homology sequence was subcloned 5' to the DEF unit of pBSDEF, creating pBSHD. For this, the GPB segment was cleaved from pH with HindIII and the pBSDEF was cleaved with NotI and both were treated with Klenow polymerase to blunt the DNA ends. To create plasmid pHD, the 3.4 kb SacII-HindIII fragment containing GPB-DEF from pBSHD replaced a 30 bp SacII-HindIII fragment of the Ad1 intron in plasmid pA4Ad1. A second 1.7 kb homology unit from pH was cloned 3' to the transcription termination signal at a HindIII site in plasmid pHD finalizing the intrachromosomal recombination construct pHDH.
COS1 cells were grown and maintained in DME-high glucose medium supplemented with 10% fetal calf serum and grown at 37oC. Cells were transfected with 20 mg plasmid DNA by calcium phosphate coprecipitation, followed by a glycerol shock 24 h posttransfection. Forty-eight hours posttransfection, cells were washed with phosphate-buffered saline (PBS, pH 7.3), fixed with 2% formaldehyde for 5 min at room temperature, and washed twice with PBS. Staining for lacZ expression was with 5-bromo-4-chloro-3-indolyl-[beta]-D-galactopyranoside (X-gal) as previously reported (38 ). Cell extracts were prepared by multiple freeze-thaw cycles as described (50 ). Protein concentration was determined by Bio-Rad assay as per manufacturer's recommendations. Enzymatic activity was quantitatively determined by o-nitrophenyl-[beta]-D-galactopyranoside (ONPG) activity, as described (50 ).
The SphI-KpnI fragment from pHDH was injected into the pronuclei of fertilized (C57BL/6*CBA/Ca) F2 mouse eggs as previously described (51 ). Mouse lines were maintained by breeding with F1 mice. Mice containing the transgene were identified by extraction of genomic tail tip DNA, restriction enzyme digestions and Southern blot analysis by standard methods (50 ). Total mRNA was prepared by tissue homogenization in liquid N2, shearing through a 23 gauge needle in guanidine isothiocyanate buffer, and cesium chloride centrifugation. Northern analysis was by standard techniques (50 ).
Tissue from transgenic animals was embedded in OCT (optimal-cutting-temperature compound, Miles) and frozen at -70oC. Cryostat sections were 7 [mu]m thick. Testicular cell preparations were in EKRB (Krebs-Ringer bicarbonate buffer) as described (44 ). Sections and cells were fixed and stained as described (38 ).
For FACS analysis and sorting, sperm were obtained from the caudal epididymis of adult male mice and stained as described (38 ) using 33 [mu]M 5-octanoylaminofluorescein di-[beta]-D-galacto- pyranoside (C8FDG, Molecular Probes) with 10 mM chloroquine. Incubations were for 30 min at 37oC. Prior to FACS analysis, 400 ml of PBS was added. Samples for FACS analysis were run on a Becton-Dickinson FACScan and analyzed using Lysis II software. The samples for FACS sorting were run on a Becton-Dickinson FACStar Plus.
Extraction of sperm DNA required the addition of 50 mM DTT to the lysis buffer. Samples were digested with BamHI and run on a 1% low-melting temperature agarose gel. The gel area containing the 5.0 kb fragments, as measured by HindIII-digested lamda DNA, was excised. Gel slices were melted at 65oC for 15 min and 10-20 ml of melted agarose/DNA was used for PCR analysis. The lacZ primers were 5'TCGCTACCTGGAGAGACGCG and 5'ACTGCTGCCAGGCGCTGATG which amplified a 550 bp fragment from the 3' end of the gene. PCR products were analyzed by Southern blot analysis using a 3' lacZ gene probe. The glycophorin primers were GP2-5' and GP2-3' which amplified a 260 bp fragment. PCR products were run on a 1.5% agarose gel and stained with ethidium bromide.
Total testes RNA (4 [mu]g) was treated with DNase I and reverse transcribed with random primers and Superscript II (H-) according to manufacturer's protocol (Clontech). RT-PCR was performed with primers RT-P1 (5'AGCGAAGAGGCCCGCACCGATCG) and RT-P2 (5'ATGCCGCTCATCCGCCACATATC) which span the lacZ intron insertion site.
We would like to thank Magda Konarska and Erik Falck-Pedersen for materials, Petronio Zalamea, Tom Delohery, Katia Manova, and Peter Romanienko for technical assistance and Olga Blumenfeld and Cheng-Han Huang who kindly provided helpful discussions and glycophorin sequences. This work was supported by grant CA09512 from the National Cancer Institute (M.E.M.), grants from the Beckman Foundation and the Pew Charitable Trusts (M.J.), and a cancer center core grant (P30-CA-08748-26). M.J. is recipient of the Frederick R. Adler Chair.
1 Donis-Keller, H., Green, P., Helms, C., Cartinhour, S., Weiffenbach, B., Stephens, K., Keith, T. P., Bowden, D. W., Smith, D. R., Lander, E. S., Botstein, D., Akots, G., Rediker, K. S., Gravius, T., Brown, V. A., Rising, M. B., Parker, C., Powers, J. A., Watt, D. E., Kauffman, E. R., Bricker, A., Phipps, P., Muller-Kahle, H., Fulton, T. R., Ng, S., Schumm, J.W., Braman, J.C., Knowlton, R. G., Barker, D. F., Crooks, S. M., Lincoln, S. E., Daly, M. J. & Abrahamson, J. (1987) A genetic linkage map of the human genome. Cell51, 319-37.MEDLINE Abstract
2 Rouyer, F., Simmler, M. C., Johnsson, C., Vergnaud, G., Cooke, H. J. & Weissenbach, J. (1986) A gradient of sex linkage in the pseudoautosomal region of the human sex chromosomes. Nature319, 291-5.MEDLINE Abstract
3 Holliday, R. (1964) A mechanism for gene conversion in fungi. Genet. Res. 5, 282-304.
4 Meselson, M. & Radding, C. (1975) A general model for genetic recombination. Proc. Natl Acad. Sci. USA72, 358-361.MEDLINE Abstract
5 Szostak, J. W., Orr-Weaver, T. L., Rothstein, R. J. & Stahl, F. W. (1983) The double-strand-break repair model for recombination. Cell33, 25-35.MEDLINE Abstract
6 Petes, T. D., Malone, R. E. & Symington, L. S. (1991) In The Molecular and Cellular Biology of the Yeast Saccharomyces: Genome dynamics, protein synthesis and energetics., ed. Broach, J. R. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, pp. 407-521.
7 Jasin, M. (1996) Genetic manipulations of genomes with rare cutting endonucleases. Trends in Genetics 12, 224-228.MEDLINE Abstract
8 Bondeson, M. L., Dahl, N., Malmgren, H., Kleijer, W. J., Tonnesen, T., Carlberg, B. M. & Pettersson, U. (1995) Inversion of the IDS gene resulting from recombination with IDS-related sequences is a common cause of the Hunter syndrome. Hum. Mol. Genet.4, 615-621.MEDLINE Abstract
9 Naylor, J. A., Buck, D., Green, P., Williamson, H., Bentley, D. & Giannelli, F. (1995) Investigation of the Factor VIII intron 22 repeated region (Int22h) and the associated inversion junctions. Hum. Mol. Genet.4, 1217-1224.MEDLINE Abstract
10 Rossiter, J. P., Young, M., Kimberland, M. L., Hutter, P., Ketterling, R. P., Gitschier, J., Horst, J., Morris, M. A., Schaid, D. J., deMoerloose, P., Sommer, S., Kazazian, H. H. & Antonarakis, S. E. (1994) Factor VIII gene inversions causing severe hemophilia A originate almost exclusively in male germ cells. Hum. Mol. Genet.3, 1035-1039.MEDLINE Abstract
11 Bollag, R. J., Waldman, A. S. & Liskay, R. M. (1989) Homologous recombination in mammalian cells. Ann. Rev. Genet.23, 199-225.MEDLINE Abstract
12 Rassoulzadegan, M., Paquis-Flucklinger, V., Bertino, B., Sage, J., Jasin, M., Miyagawa, K., van Heyningen, V., Besmer, P. & Cuzin, F. (1993) Transmeiotic differentiation of male germ cells in culture. Cell75, 997-1006.MEDLINE Abstract
13 Murti, J. R., Bumbulis, M. & Schimenti, J. C. (1992) High-frequency germ line gene conversion in transgenic mice. Mol. Cell. Biol.12, 2545-2552.MEDLINE Abstract
14 Haber, J. E. (1992) Exploring the pathways of homologous recombination. Curr. Opin. Cell Biol.4, 401-412.MEDLINE Abstract
15 Wu, T. C. & Lichten, M. (1994) Meiosis-induced double-strand break sites determined by yeast chromatin structure. Science263, 515-518.MEDLINE Abstract
16 Thomas, B. J. & Rothstein, R. (1989) Elevated recombination rates in transcriptionally active DNA. Cell56, 619-630.MEDLINE Abstract
17 Dorer, D. R. & Christensen, A. C. (1989) A recombinational hotspot at the triplo-lethal locus of Drosophila melanogaster. Genetics122, 397-401.MEDLINE Abstract
18 Williams, S. M., Kennison, J. A., Robbins, L. G. & Strobeck, C. (1989) Reciprocal recombination and the evolution of the ribosomal gene family of Drosophila melanogaster. Genetics122, 617-624.MEDLINE Abstract
19 Steinmetz, M., Uematsu, Y. & Fisher-Lindahl, K. (1987) Hotspots of homologous recombination in mammalian genomics. Trends in Genetics3, 7-10.
20 Shiroishi, T., Sagai, T., Hanzawa, N., Gotoh, H. & Moriwaki, K. (1991) Genetic control of sex-dependent meiotic recombination in the major histocompatibility complex of the mouse. EMBO J. 10, 681-686.MEDLINE Abstract
21 Steinmetz, M., Stephan, D. & Fischer Lindahl, K. (1986) Gene organization and recombinational hotspots in the murine major histocompatibility complex. Cell44, 895-904.MEDLINE Abstract
22 Allitto, B. A., MacDonald, M. E., Bucan, M., Richards, J., Romano, D., Whaley, W. L., Falcone, B., Ianazzi, J., Wexler, N. S., Wasmuth, J. J. et al. (1991) Increased recombination adjacent to the Huntington disease-linked D4S10 marker. Genomics9, 104-112.MEDLINE Abstract
23 Hubert, R., MacDonald, M., Gusella, J. & Arnheim, N. (1994) High resolution localization of recombination hot spots using sperm typing. Nature Genet.7, 420-424.MEDLINE Abstract
24 Felix, C. A., Lange, B. J., Hosler, M. R., Fertala, J. & Bjornsti, M. A. (1995) Chromosome band 11q23 translocation breakpoints are DNA topoisomerase II cleavage sites. Cancer Res.55, 4287-4292.MEDLINE Abstract
25 Tashiro, S., Kotomura, N., Tanaka, K., Suzuki, K., Kyo, T., Dohy, H., Niwa, O. & Kamada, N. (1994) Identification of illegitimate recombination hot spot of the retinoic acid receptor alpha gene involved in 15;17 chromosomal translocation of acute promyelocytic leukemia. Oncogene9, 1939-1945.MEDLINE Abstract
26 Zimmerer, E. J. & Passmore, H. C. (1991) Structural and genetic properties of the Eb recombinational hotspot in the mouse. Immunogenetics33, 132-140.MEDLINE Abstract
27 Metzenberg, B. A., Wurzer, G., Huisman, J. H. T. & Smithies, O. (1991) Homology requirements for unequal crossing over in humans. Genetics128, 143-161.MEDLINE Abstract
28 Maeda, N. & Smithies, O. (1986) The evolution of multigene families: human haptoglobin genes. Ann. Rev. Genet.20, 81-108.MEDLINE Abstract
29 Huang, C. H., Johe, K. K., Seifter, S. & Blumenfeld, O. O. (1991) Biochemistry and molecular biology of MNSs blood group antigens. Baillières Clin. Haematol.4, 821-848.MEDLINE Abstract
30 Rearden, A., Huan, P., Dubnicoff, T., Kudo, S. & Fukuda, M. (1990) Identification of the crossing-over point of a hybrid gene encoding human glycophorin variant Sta. J. Biol. Chem.265, 9259-9263.MEDLINE Abstract
31 Ashley, T. & Ward, D. C. (1993) A `hotspot' of recombination coincides with an interstitial telomeric sequence in the Armenian hamster. Cytogenet. Cell Genet.62, 169-171.MEDLINE Abstract
32 Huang, C. H. & Blumenfeld, O. O. (1991) Multiple origins of the human glycophorin Sta gene. Identification of hot spots for independent unequal homologous recombinations. J. Biol. Chem.266, 23306-23314.MEDLINE Abstract
33 Uematsu, Y., Kiefer, H., Schulze, R., Fischer-Lindahl, K. & Steinmetz, M. (1986) Molecular characterization of a meiotic recombinational hotspot enhancing homologous equal crossing-over. EMBO J.5, 2123-2129.MEDLINE Abstract
34 Shiroishi, T., Hanzawa, N., Sagai, T., Ishiura, M., Gojobori, T., Steinmetz, M. & Moriwaki, K. (1990) Recombinational hotspot specific to female meiosis in the mouse major histocompatibility complex. Immunogenetics31, 79-88.MEDLINE Abstract
35 Kudo, S. & Fukuda, M. (1989) Structural organization of glycophorin A and B genes: Glycophorin B gene evolved by homologous recombination at Alu repeat sequences. Proc. Natl Acad. Sci. USA86, 4619-4623.MEDLINE Abstract
36 Huang, C. H. & Blumenfeld, O. O. (1991) Identification of recombination events resulting in three hybrid genes encoding human MiV, MiV(J.L.), and Sta glycophorins. Blood77, 1813-1820.MEDLINE Abstract
37 Braun, R. E., Peschon, J. J., Behringer, R. R., Brinster, R. L. & Palmiter, R. D. (1989) Protamine 3'-untranslated sequences regulate temporal translational control and subcellular localization of growth hormone in spermatids of transgenic mice. Genes Devel. 3, 793-802.MEDLINE Abstract
38 Jasin, M. & Zalamea, P. (1992) Analysis of Escherichia coli beta-galactosidase expression in transgenic mice by flow cytometry of sperm. Proc. Natl Acad. Sci. USA89, 10681-10685.MEDLINE Abstract
39 Konarska, M. M. & Sharp, P. A. (1987) Interactions between small nuclear ribonucleoprotein particles in formation of spliceosomes. Cell49, 763-774.MEDLINE Abstract
40 Xuan, S., Baptista, C. A., Balas, G., Tao, W., Soares, V. C. & Lai, E. (1995) Winged helix transcription factor BF-1 is essential for the development of the cerebral hemispheres. Neuron14, 1141-1152.MEDLINE Abstract
41 Logan, J., Falck-Pedersen, E., Darnell, J. E., Jr. & Shenk, T. (1987) A poly(A) addition site and a downstream termination region are required for efficient cessation of transcription by RNA polymerase II in the mouse beta maj-globin gene. Proc. Natl Acad. Sci. USA84, 8306-8310.MEDLINE Abstract
42 Tantravahi, J., Alvira, M. & Falck-Pedersen, E. (1993) Characterization of the mouse beta maj globin transcription termination region: a spacing sequence is required between the poly(A) signal sequence and multiple downstream termination elements. Mol. Cell. Biol.13, 578-587.MEDLINE Abstract
43 Jaenisch, R. (1988) Transgenic animals. Science240, 1468-74.MEDLINE Abstract
44 Romrell, L. J., Bellve, A. R. & Fawcett, D. W. (1976) Separation of mouse spermatogenic cells by sedimentation velocity. A morphological characterization. Devel. Biol.49, 119-131.
45 Grant, S. G. & Bigbee, W. L. (1993) In vivo somatic mutation and segregation at the human glycophorin A (GPA) locus: phenotypic variation encompassing both gene-specific and chromosomal mechanisms. Mutat. Res.288, 163-72.MEDLINE Abstract
46 te Riele, H., Maandag, E. R. & Berns, A. (1992) Highly efficient gene targeting in embryonic stem cells through homologous recombination with isogenic DNA constructs. Proc. Natl Acad. Sci. USA89, 5128-32.MEDLINE Abstract
47 Liskay, R. M., Letsou, A. & Stachelek, J. L. (1987) Homology requirement for efficient gene conversion between duplicated chromosomal sequences in mammalian cells. Genetics115, 161-117.MEDLINE Abstract
48 Waldman, A. S. & Liskay, R. M. (1988) Dependence of intrachromosomal recombination in mammalian cells on uninterrupted homology. Mol. Cell. Biol.8, 5350-5357.MEDLINE Abstract
49 Peschon, J. J. (1987) Spermatid-specific expression of protamine 1 in transgenic mice. Proc. Natl Acad. Sci. USA84, 5316-5319.MEDLINE Abstract
50 Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor.
51 Hogan, B., Costantini, F. & Lacy, E. (1986) Manipulating the Mouse Embryo. Cold Spring Harbor Laboratory, Cold Spring Harbor.
*To whom correspondence should be addressed
This page is maintained by OUP admin. Last updated Thu Oct 31 15:25:15 GMT 1996. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
