PCR-based gene targeting of the inducible nitric oxide synthase ...

T r a n s g e n i c R e s e a r c h 5, 4 1 3 M 2 0 (1996)

PCR-based gene targeting of the inducible nitric oxide synthase (NOS2) locus in murine ES cells, a new and more cost-effective approach DAVID FANG,

A. RANDOLPH, J A M E S W. V E R B S K Y , LIPING RAZQALLAH HAKEM * and LARRY E. FIELDS*

YANG,

YIFU

Departments of Medicine and Pathology, Divisions of Cardiology and Biology and Biomedical Sciences, Washington University School of Medicine, St. Louis, MO 63110-1093, USA (Fax: 1 314 362 8957)

Received 14 July 1995; revised 1 November 1995; accepted 13 November 1995

Gene targeting by double homologous recombination in murine embryonic stem (ES) cells is a powerful tool used to study the cellular consequences of specific genetic mutations. A typical targeting construct consists of a neomycin phosphotransferase (neo) gene flanked by genomic DNA fragments that are homologous to sequences in the target chromosomal locus. Homologous DNA fragments are typically cloned from a murine genomic DNA library. Here we describe an alternative approach whereby the inducible nitric oxide synthase (NOS2) gene locus is partially mapped and homologous DNA sequences obtained using a long-range PCR method. A 7 kb NOS2 amplicon is used to construct a targeting vector where the neo gene is flanked by PCR-derived homologous DNA sequences. The vector also includes a thymidine kinase (tk) negative-selectable marker gene. Following transfection into ES cells, the PCR-based targeting vector undergoes efficient homologous recombination into the NOS2 locus. Thus, PCR-based gene targeting can be a valuable alternative to the conventional cloning approach. It expedites the acquisition of homologous genomic DNA sequences and simplifies the construction of targeting plasmids by making use of defined cloning sites. This approach should result in substantial time and cost savings for appropriate homologous recombination projects. Keywords: homologous recombination; inducible nitric oxide synthase; embryonic stem cell

Introduction Gene targeting is used to develop animal models of disease, explore mechanisms of biological development, and assess the functional importance of molecules. The subject of gene targeting has been reviewed recently (Robbins, 1993; Bronson and Smithies, 1994; Melton, 1994; Nathan and Xie, 1994a,b; Shastry, 1994; Routtenberg, 1995; Soriano, 1995). This technique makes use of double homologous recombination in ES cells between chromosomal DNA and homologous sequences in the targeting vector (Thomas and Capecchi, 1987; Hasty and Bradley, 1993; Ramirez-Solis et al., 1993). Transfected ES cells are screened and those with a correctly ~:Present address: Ontario Cancer Institute/Princess Margaret Hospital, Amgen Institute, Toronto, Canada. *To whom correspondenceshould-be addressed. 0962-8819 9 1996 Chapman & Hall

targeted allele can be injected into blastocysts to generate mice having a defined germline mutation. Conventional targeting constructs consist of the neo positive-selectable marker gene flanked by a total of at least 4 kb of homologous genomic DNA sequence (Thomas and Capecchi, 1987). The tk negative-selectable marker gene is commonly used in conventional targeting constructs (Mansour et al., 1990; McMahon and Bradley, 1990; Chisaka and Capecchi, 1991; Kitamura et al., 1991; Lee et al., 1992; Li et al., 1992; Tarakhovsky et al., 1994; Kuida et al., 1995). The frequency of homologous recombination varies as a function of insertional-versus replacement-vector design (Thomas and Capecchi, 1987), length of homologous DNA sequences (Thomas and Capecchi, 1987; Hasty et al., 1991; Deng and Capecchi, 1992), and degree of polymorphic variation between the vector and the chromosome (te Riele et al., 1992). In a conventional

Randolph et al.

414 targeting vector, DNA fragments are cloned from isogenic or non-isogenic murine genomic libraries. We present an alternative approach where homologous genomic DNA is obtained using a long-range PCR method (Cheng et al., 1994). The homologous amplicon is used to construct a targeting vector that is transfected into murine ES cells, where double homologous recombination takes place. PCR-based gene targeting was successfully applied to the mouse inducible nitric oxide synthase locus, generating ES cell lines in which the NOS2 gene is disrupted on one chromosome.

Materials and methods

PCR

The genomic organization of the human NOS2 locus (Chartrain et al., 1994) and sequences of human (Geller et al., 1993) and murine NOS2 cDNAs (Lowenstein et al., 1992; Lyons et al., 1992; Xie et al., 1992) were used as a basis for predicting where putative murine intron-exon boundaries might be. Oligonucleotide primers were obtained that bind putative exons 1 to 8, 12, 18 and 23 (Protein Chemistry Laboratory; Table 1 and Fig. 1). Genomic DNA from the RAW 264.7 murine macrophage cell line (H-2 a, ATCC) (Raschke et al., 1978) was used as a template. Conditions for PCR were as follows: 93 ~ 5 min; (93 ~ min, 68 ~ min) for 16 cycles; (93 ~ 1 min, 68~ plus 15 sec) for 12 cycles; 93 ~ 3 rain. We used the GeneAmp XL PCR system (Perkin Elmer) which contains a mixture of thermostable DNA

2 3 4 5 6 7 8 9101112M ~23.1

9.4

7kb~

"--6.6 "~ 4 . 4 ~2.3 ---2,0

--

Fig. 1. Mapping the murine NOS2 genomic locus using long-range PCR. Approximate genomic distance between exons was estimated for primer pairs that generated an anaplicon (lanes 1, 5, 6, 7, 8, and 9). See Table 1 and Methods for primer sequences, putative exon locations, and size of amplicons. The largest amplicon is about 7kb and includes putative exons 4 to 7 (lane 8). This amplicon was used to construct a gene targeting vector. (lane M) size markers.

polymerases (rTth and Vent). Additional primers are as follows: (lane 10) sense-5' acactacgtaCTTTATGCCACCAACAATGGCAAC3' and antisense-5' GGAGATAGGACATAGTTCAACATCTCC3' primers in putative exons 7 and 12, respectively; (lane 11) sense primer of lane 10 and anti-sens~ primer-5'GGATGCTGCTGAGGGCTCTGTTGAGG3', in putative exons 7 and 18; and (lane 12) sense primer of lane 10 and antisense primer-5'CTCAGGGAGCTGGAAGCCACTGACACTTCG3', in putative exons 7 and 23. Amplicons were sizefractionated on a 1% agarose gel.

Table 1. PCR-based estimation of inter-exon distances in the murine NOS2 gene locus

Gel Lane (Fig. 1)

1 2 3 4 5 6 7 8 9

Primer-pairs

5'gagaggcctAACTTCTCAGCCACCTTGG3'-sense 5'gagaggcctGCTGAGAACACACAAGG3'-antisense 5'ATGGCTTGCCCCTGGAAGT3' 5'GAACATTCTGTGCTGTCCCAGTGAGGAGCTGC3' 5'gagtcgcgaCCGCAGCTCCTCACTGGGACAGCACAGAATGTTC3' 5' GATGTGGCCTTGTGGTGAAG3 ' 5 'ATGGCTTGCCCCTGGAAGT3 ' 5'GATGTGGCCTTGTGGTGAAG3' 5'CTGGACAAGCTGCATGTGAC3' 5'CTGGTTGATGAACTCAATGGCATGAGGCAGGA3' 5'gagtcgcgaGCTCCTGCCTCATGCCATTGATTCATCAACCAG3' 5'tgtatcgatTTGGACCACTGGATCCTGCCGATGCAGCGAG3' 5'acactacgtaCTCGCTGCATCGGCAGGATCCAGTGGTC3' 5'GTTGCCATTGTTGGTGGCATAAAG3' 5'CTGGACAAGCTGCATGTGAC3' 5' GTTGCCATTGTTGGTGGCATAAAG3' 5'acactacgtaCTTTATGCCACCAACAATGGCAAC3' 5'gagatcgatGTGAACTCCAAGGTGGCAGCATCC3'

"Predictions based on human NOS2 genomie organization ~N.A., no amplieon; underlined, restriction sites; lower-case, added sequence

0.56

Putative exons a

Amplicon size (kb)

1-2

1.6

2-3

N.A. b

3-4

N.A.

2-4

N.A.

4-5

1.8

5-6

4.1

6-7

1.1

4-7

7.0

7-8

1.3

415

PCR-basedgene targegng Plasmid construction A 7.0 kb amplicon spanning putative exons 4 to 7 was subctoned into the Bluescript SKII + plasmid (Stratagene), using Not I and Cla I restriction sites that were added to the primers (Scharf et al., 1986; Higuchi, 1989). The PCR-derived insert was mapped using restriction endonucleases and partially sequenced using the di-deoxy sequencing method (Sanger et al., 1977) and a Sequenase T7 DNA sequencing system (Amersham). Exon 6 was opened using a partial SacI digestion and the 1.7kb phosphoglycerokinase (pgk)-neo cassette inserted (Adra et al., 1987; Soriano et al., 1991; Tybulewicz et al., 1991). A 2.6kb pgk-tk cassette was inserted 5' of exon 4 (Tybulewicz et al., 1991). The long-arm spanned exons 4 to 6 (5.9 kb) and the short-arm spanned exons 6 to 7 (1.1 kb). The targeting vector was linearized at the NotI site. Cells E14 ES cells (Handyside et al., 1989) were co-cultured on G418-resistant murine embryonic fibroblasts (Martin and Evans, 1975; Doetschman et al., 1985; Handyside et al., 1989) in high glucose DMEM (4.5 g1-1, Gibco) supplemented with 15% foetal calf serum (Hyclone), glutamine (2 mM), non-essential amino acids (0.1 raM), penicillin (100 Uml-1), streptomycin (100 txgml-1), 2-mercaptoethanol (50 ~M), and leukemia inhibitory factor (1000 Um1-1, Gibco) at 6% CO2. Transfection and selection ES cells (2 • 10 7 cells) were electroporated using the BTX 600 electroporation system (340V, 100 ~LF, 0.4cm gap) and 50 ~tg of Not I-linearized targeting vector. Electroporated cells were plated onto 10 cm dishes and selection initiated 2 4 h later using 250 txgm1-1 G418 (Gibco) and 2 ~ ganciclovir (Syntex). Colonies were picked on day 6. Screening Genomic DNA was isolated from ES celt clones and initially evaluated using PCR analysis. A primer external to the short-arm but within exon 7 (5'GTTGCCATTGTTGGTGGCATAAAG3') and one specific for the pgkneo cassette were used (5'CGGTAGAATTATCGAATTCCTGCAGC3'). Accurate targeting events were determined by Southern blot analysis using a 0.3 kb Bam HI/ NcoI genomic probe located external to the short-arm (probe A, Fig. 3), and a 1.7 kb pgk-neo probe (probe B, Fig. 3). Genomic DNA was digested with Eco RI and/or HindIII, size-fractionated on a 0.7% agarose gel which was dried, pre-hybridized, and hybridized using established techniques for in-gel Southern blot analysis (Southem, 1975; Tsao et al., 1983; Lueders and Fewell, 1994). Gels were washed at 68 ~ in 4. • SSC/0.1% SDS followed by

0.1 • SDS for l h. Autoradiograms were digitized using an HP ScanJet IIc/ADF scanner (Hewlett-Packard), a PowerMac 7100/66, and Deskscan II software (Hewlett-Packard).

Results

PCR-based genomic mapping To determine where a pgk-neo selectable marker cassette could be placed in order to disrupt the NOS2 gene, a combination of Southern blot analysis and long-range PCR (Cheng et al., 1994) were used to map genomic regions of the murine NOS2 locus. Using the organization of the human NOS2 gene (Chartrain et al., 1994), the location of exons in murine cDNA sequence were predicted (Xie et al., 1992). The distance between putative exons was then mapped using genomic DNA template from the RAW 264.7 murine macrophage cell line, isolated as part of a related study. Long-range PCR was performed using primer-pairs listed in Table 1 and Fig. 1. A blend of thermostable DNA polymerases was used to generate amplicons containing an intron flanked by a portion of exons containing primer-binding sites (Fig. 1). The distance between exons 1 and 2 of the murine NOS2 gene is similar to that reported for the human locus (Chartrain et al., 1994). They are approximately 1.6 and 1.8kb, respectively (lane 1, Fig. 1). Distances between exons 2, 3, and 4 of the human NOS2 gene are approximately 1.1, 0.9, and 2.0kb, respectively. An amplicon derived using primers specific for putative exons 2 to 4 was not detected (lanes 2, 3, and 4; Fig. 1), although we generated a 20 kb amplicon using control plasmid and primers provided with the GeneAmp XL system (data not shown). The organization of exons 4 through 7 appears to be comparable between human and murine NOS2 genes. The distances are approximately 1.2, 6.5, and 1.7kb in the human NOS2 gene and approximately 1.8, 4.1 and 1.1 kb in the murine gene (lanes 5, 6 and 7; Fig. 1). To determine the capacity of this PCR system for generating genomic amplicons larger than 7 kb (lane 8, Fig. 1), primer-pairs from putative murine exons 7, 12, 18 and 23 were used because inter-exon distances are approximately 7.5, 15 and 22 kb in the human NOS2 gene (Chartrain et al., 1994). That genomic amplicons were not obtained for this region of the murine NOS2 locus, suggests the presence of differences in organization or the location of intron-exon boundaries compared to the human gene (lanes 10, 11 and 12; Fig. 1; see discussion). PCR-based genomic mapping was further validated by comparing PCR-generated data about inter-exon distances to data obtained using Southern blot analysis (Fig. 2). We digested murine genomic DNA with Eco RI, Hin dill, or Sac I and probed it with the 0.3 kb DNA fragment used

416

Randolph et al. exons

B

A E

H

S

EH E S HS

genomic locus

1

2

XH __]1! , :

3

!m/8, 0 -

,-,20 kb

9

targetedlocus

7

8

S '

]

E H ~ ,~'Jl"

H 1

1] I

i

I

!

,42

CE

S H ,F . .i

S

N ,~

N .4. CE ]pBSSKII I pgktk ~ I

.

.

S ~

H I

S

X H E S __~1I II II II#1 I9 II. . i . I . . 9

H

SH~ ~

targeting vector

2.2 kb

"

6

S /i

II

PCR arnplicon

5.6 kb

9

5 E ,

..... "

9

4

I

~

,'gKne~ ES

3t,probes

Fig. 2. Validation of PCR-based genomic mapping using in-gel Southern blot analysis, where genomic DNA fragment sizes are the same as those estimated by PCR analysis. (panel A) Murine genomic DNA (8 pg/lane) was digested with EcoRI (E), HindlII (H), or Sac I (S). (panel B) DNA was double-digested with Eco RI and Hind lI1 (EH), EcoRI and Sac I (ES), or Hind III and Sac I (HS). Samples underwent electrophoresis within the same gel and were hybridized with probe A (Fig. 3). Dots represent size markers as described in Fig. 1. to screen ES cells (probe A, Fig. 3). This probe was derived from a 1.3 kb amplicon spanning putative exons 7 and 8 (lane 9, Fig. 1). Probe A detected approximately 8.5 kb EcoRI, 20kb HindlII, and 4.0kb SacI fragments on Southern blot analysis (panel A, Fig. 2). The 8.5 kb distance predicted by PCR-mapping between the two Eco RI sites flanking putative exon 6 is the same length as the fragment detected by probe A on Southern blot analysis. We did further mapping of the region arotmd probe A by performing double digests of murine genomic DNA using EcoRI and Hind III, EcoRI and Sac I, or Hind lII and Sac I. The PCR-generated map predicts an approximately 5.6kb EcoRI/HindlII, and a 2.2kb Eco RI/Sac I fragment. These fragment sizes are the same as those determined by Southern blot analysis, further validating this approach. Targeting construct For PCR-based gene targeting, a 7 kb amplicon flanked by portions of exons 4 and 7 (lane 8, Fig. 1) was generated. The pgk-neo positive selectable marker gene was inserted into putative exon 6 of this amplicon, at a Sac I site (Fig. 3). This approach resulted in a 1.1 kb short-arm and a 5.9 kb long-arm of genomic DNA sequence homologous to the murine NOS2 chromosomal locus. The pgk-tk negative-selectable marker gene was placed 5' of exon 4. Exon 6 of the NOS2 gene was targeted for several reasons. Our ability to generate a 7 kb amplicon between exons 4 and 7 using existing primer-pairs initially prompted us to consider exons located within this region. There is currently no evidence for alternative splicing of murine NOS2 mRNA between exons 2 and 7 (Dinerman et al., 1993; Nathan and Xie, 1994a). The binding domains for calmodulin (Chartrain e t al., 1994), tetra-

~ 2 kb

E H ]1 /~-~

:

B

-,14 .~

Fig. 3. Schematic diagram for PCR-based gene targeting of the murine NOS2 locus. (exons, genomic locus) Genomic positions of putative exons were determined by the PCR strategy shown in Fig. 1. Distance between exons 2 to 4 is represented by interrupted lines. The distance between the second and third Hind III (H) sites was determined by Southern blot analysis to be approximately 20 kb. Other restriction sites are Cla I (C), Eco RI (E), Sac I (S), and Xba I (X). For the sake of clarity, only relevant restriction sites are shown, (PCR amplicon) Primers 1 and 2 are represented by arrow heads and were used to generate a 7kb amplicon, represented by the stippled bar. (targeting vector) This was constructed by cloning the 7kb amplicon into a Bluescript vector containing the pgk-tk selection cassette and inserting the pkg-neo selection cassette into the Sac I site of exon 6. The targeting vector was linearized using NotI and transfected into ES cells. (targeted locus) The map of the murine NOS2 locus after homologous recombination. Primers 3 and 4 (arrow heads) were used to screen for homologous recombination involving the short-arm. (probes) Homologous recombination was confirmed by Southern blot analysis using the 0.3 kb genomic probe A and the 1.7 kb pgkneo probe B. hydrobiopterin (Uvarov and Lyashenko, 1995), the r a v i n nucleotides FMN and FAD (Chartrain et al., 1994), and NADPH (Xie et al., 1994) are all 3' of exon 6. Part of the haem binding site appears to be encoded by exon 6 (Chen et al., 1994). Thus, if disruption of the murine NOS2 gene at exon 6 results in a stable but truncated protein, it should be non-functional. Gene targeting We then transfected the PCR-based targeting vector, linearized with NotI, into El4 ES cell using the electroporation method. Following a 24 h recovery period, cells were selected in media containing G418 and ganciclovir (GANC). Colonies resistant to G418 and GANC were picked, expanded, and Screened for homologous recombinants by PCR analysis using a primer-pair specific for a region 3' of the short-arm of the targeting vector and for the pgk-neo cassette (Table 1, Fig. 3). Clones found to be positive by PCR analysis were confirmed by ingel Southern blot analysis using probes A and B (Fig. 3). In addition to the germline 8.5 kb EcoRI fragment, a new 2.4 kb EcoRI fragment is detected by

417

PCR-based gene targeting

probe A when targeting vector and chromosomal DNA undergo homologous recombination (Table 2, lane A in Fig. 4A and 4B). Double-digestion with E e o R I and HindIII results in a 5.6kb germline fragment whereas the 2.4 kb E e o R I new fragment is unchanged (Table 2, Fig. 4B). Digestion with Hind III generates an approximately 20 kb germline fragment which is difficult to resolve from the 19 kb fragment of the targeted allele (Table 2, Fig. 4C). Probe B detects the transfected pgk-neo cassette and fragments of the endogenous p g k gene (Table 2). A 4.4kb E e o R I and an approximately 4.3kb HindIII fragment represent the endogenous p g k gene (Table 2, Fig. 5). DNA from a correctly targeted clone, ROH. 1, has the expected 8kb E c o R I and approximately 19kb Hind III fragments representing the targeted allele (Table 2, lane A of Fig. 5). In contrast, the results of probing DNA from clone ROH.2 with probe B are consistent with

A

B EcoRI

Hind lll

8.0 kb* 4.4kb

---

A B C

-

,-,19 k b *

-

4.3 kb

A B C

Fig. 5. Confirmation of homologous recombination using probe B. Genomic DNA was isolated from ES cell clones ROH.1 (lane A), ROH.2 (lane B), and El4.1 (lane C). Ten micrograms of DNA was digested with EcoI (panel A) or HindlII (panel B). The band representing the mutant allele is marked with an asterisk. Hybridizations were performed as in Fig. 4.

Table 2. The length of predicted restriction fragments of the murine NOS2 locus in germline and targeted ES cells Fragment Length (kb) probe A

probe B

Locus

Eco RI

Eco RI/ Hind III

Hind III

Eco RI

Hind III

Germline Targeted"

8.5 8.5 2.2

5.6

5.6 2.2

20 20 19

4.4a 8.0 4.4

4.3a 19 4.3

aEndogenouspgk gene

A

B EcoRI

C EcoRI/Hindlll

Hind III

__(,-.,20 kb ,-.,19 kb ~

8.5 kb

5.6 kb 2.2 kb*

2.2 kb*

ABC

ABC

ABC

Fig. 4. Homologous recombination is detected using Southern blot analysis of genomic DNA from transfected ES cell clones, hybridized with probe A. Genomic DNA was isolated from ES cell clones ROH.1 (lane A), ROH.2 (lane B), and El4.1 (lane C). Ten micrograms of DNA was digested with EcoRI (panel A), EcoRI and HindlII (panel B), or HindIII (panel C). The band representing the mutant allele that has undergone homologous recombination is marked with an asterisk. Hybridizations were performed simultaneously, using the same probe preparation.

Randolph et al.

418 a random insertion of the targeting construct. Approximately 4.3 kb E c o R I and 2.3 kb H i n d I I I fragments are detectable using probe B (Fig. 5A and 5B). No other bands are observed, suggesting that no additional copies of the targeting vector integrated into the genome of these ES cell clones. Examination of the 5'-end of the genomic locus external to the targeting vector revealed that a correct homologous recombination event had also occurred at the 5'-end (data not shown). We used as probes, a 2.9 kb Barn HI genomic fragment from between exons 5 and 6 and a 0.6 kb H i n d I I I / N c o I fragment from the 3'-end of the neo cassette. Southern blot analyses were performed using genomic DNA isolated from wildtype and targeted ES cells and digested with N c o I . The targeting frequency of PCR-based gene targeting is of a similar order-of-magnitude as conventional gene targeting (Reaume et al., 1995; Soriano, 1995). In the first 41 double-resistant clones screened, four correctly targeted clones were identified (Table 3). Thus, approximately one out of ten double-resistant clones were accurately targeted using the PCR-based gene targeting method.

Discussion Estimation of inter-exon distances using PCR and primerpairs specific for flanking exons appears feasible. This approach permits a more rapid estimation of genomic organizational structure and was applied to the murine NOS2 locus because such information was not available. A comparison of murine amplicon sizes to those predicted for the human NOS2 locus (Chartrain et al., 1994) provided our earliest indication of the potential usefulness of this approach. Evaluation of the murine NOS2 locus by in-gel Southern blot analysis provided further validation. Our ability to only generate amplicons of genomic regions spanning putative exons 1 to 2 and 4 to 7 may be due to sub-optimal primers, a need for alternative PCR conditions, use of primers located at unanticipated intron/ exon boundaries, use of primers that are actually located adjacent to each other within the same exon, or an actual intron size greater than the size-generating capacity of our current long-range PCR system. To further delineate the possibilities, we have obtained two murine P1 clones (Genome Systems) using primer-pairs internal to putative

exons 2 and 4. In contrast, organization of murine and human NOS2 genes appears to be similar for exons 1 to 2 and exons 4 to 8. Thus, in situations where genomic organization of the murine locus can be determined, PCR-based gene targeting should be a useful option. This is enhanced by the ability to add restriction sites where needed to simplify design of the targeting vector. The overall targeting frequency for the murine NOS2 locus using the PCR-based method is approximately 1 per 10 double-resistant colonies. This is comparable to values obtained previously using conventional targeting vectors (Reaume et al., 1995; Soriano, 1995). The targeting frequency should be further enhanced by using even longer-isogenic genomic amplicons. PCR-based gene targeting is substantially faster than the conventional approach in which DNA fragments are cloned from a murine genomic library. The development of long-range PCR systems has made PCR-based gene targeting feasible (Cheng et al., 1994). Now adequate lengths of homologous genomic DNA sequence can be obtained quickly. We used the GeneAmp XL system which consists of a mixture of rTth and Vent DNA polymerases that has an error-rate of 10 -4 to 10 -5 (Perkin Elmer technical services, 1995), a rate not great enough to prevent the occurrence of homologous recombination events at a suitable frequency. The PCR-based method should be applicable to singleexon as well as multiple-exon loci, by using PCR cloning of flanking sequences (Ochman et al., 1989). In conclusion, PCR-based gene targeting should result in significant time and cost savings for appropriate double homologous recombination experiments by bypassing the conventional genomic library cloning step. An ability to generate a defined amplicon also permits the addition of specific restriction endonuclease sites (Scharf et al., 1986; Higuchi, 1989), facilitating the subcloning step.

Acknowledgments We thank the Protein Chemistry Laboratory, Charlie Rodi, Bob Karr, Mark Currie, Stanley Korsmeyer, John Atkinson, Emil Unanue, and David Kipnis. This work was supported by a grant from Monsanto-Searle.

Table3. Frequency of homologous recombination using PCR-

References

based gene targeting

Adra, C.N., Boer, P.H. and McBurney, M.W. (1987) Cloning and expression of the mouse pgk-1 gene and the nucleotide sequence of its promoter. Gene 60, 65-74. Bronson, S.K. and Smithies, O. (1994) Altering mice by homologous recombination using embryonic stem cells. J Biol. Chem. 269, 27155-8. Chartrain, N.A., Geller, D.A., Koty, RR, Sitrin, N.E, Nussler, A.K., Hoffman, E.R, Billiar, T.R., Hutchinson, N.I. and Mudgett, J.S. (1994) Molecular cloning, structure, and chromosomal

No. of cells or clones

electroporated G418r, GANCr clones picked clones screened homologous recombinants aResistant colonies (r)

2 • 107 826a 240 41 4

P C R - b a s e d g e n e targeting

localization of the human inducible nitric oxide synthase gene. J. Biol. Chem. 269, 6765-72. Chen, P.F., Tsai, A.L. and Wu, K.K. (1994) Cysteine 184 of endothelial nitric oxide synthase is involved in heme coordination and catalytic activity. J. Biol. Chem. 269, 25062-6. Cheng, S., Fockler, C., Barnes, W,M. and Higuchi, R. (1994) Effective amplification of long targets from cloned inserts and human genomic DNA. Proc. NatI Acad. Sci. USA 91, 5695-9. Chisaka, O. and Capecchi, M.R. (1991) Regionally restricted developmental defects resulting from targeted. Nature 350, 473-9. Deng, C. and Capecchi, M.R. (1992) Reexamination of gene targeting frequency as a function of the extent of homology between the targeting vector and the target locus. Mol. Cell. Biol. 12, 3365-71. Dinerman, J.L., Lowenstein, C.J. and Snyder, S.H. (1993) Molecular mechanisms of nitric oxide regulation. Potential relevance to cardiovascular disease. Circ. Res. 73, 217-22. Doetschman, T.C., Eistetter, H., Katz, M., Schmidt, W. and Kemler, R. (1985) The in vitro development of blastocyst-derived embryonic stem cell lines: formation of visceral yolk sac, blood islands and myocardium. J. Embryol. Exp. Morph. 87, 27-45. Geller, D.A., Lowenstein, C.J., Shapiro, R.A., Nussler, A.K., Di Silvio, M., Wang, S.C., Nakayama, D.K., Simmons, R.L., Snyder, S.H. and Billiar, T.R. (1993) Molecular cloning and expression of inducible nitric oxide synthase from human hepatocytes. Proc. Natl Acad. Sci. USA 90, 3491-5. Handyside, A.H., O'Neill, G.T.O., Jones, M. and Hooper, M.L. (1989) Use of BRL-conditioned medium in combination with feeder layers to isolate a diploid embryonal stem cell line. Roux's Arch. Dev. BioL 198, 48-55. Hasty, P. and Bradley, A. (1993) Gene targeting vectors for mammalian cells. In: Joyner, A.L. ed., Gene Targeting: a Practical Approach, pp. 1-31. Oxford: IRL Press at Oxford University Press. Hasty, R, Rivera-Perez, J. and Bradley, A. (1991) The length of homology required for gene targeting in embryonic stem cells. Mol. Cell. Biol. 11, 5586-5591. Higuchi, R. (1989) Using PCR to engineer DNA. Irr: Erlich, H.A. ed., PCR Technology: Principles and Applications for DNA Amplifications, pp. 61-70. New York: Stockton. Kitamura, D., Roes, J., Kuhn, R. and Rajewsky, K. (1991) A B cell-deficient mouse by targeted disruption of the membrane exon. Nature 350, 423-6. Kuida, K., Lippke, J.A., Ku, G., Harding, M.W., Livingston, D.J., Su, M.S.S. and Flavell, R.A. (1995) Altered cytokine export and apoptosis in mice deficient in interleuldn-l-beta converting enzyme. Science 267, 2000-3. Lee, K.E, Li, E., Huber, L.J., Landis, S.C., Sharpe, A.H., Chao, M.V. and Jaenisch, R. (1992) Targeted mutation of the gene encoding the low affinity NGF receptor. Cell 69, 737-49. Li, E., Bestor, T.H. and Jaenisch, R. (1992) Targeted mutation of the DNA methyltransferase gene results in. Cell 69, 915-26. Lowenstein, C.J., Glatt, C.S., Bredt, D.S. and Snyder, S.H. (1992) Cloned and expressed macrophage nitric oxide synthase contrasts with the brain enzyme. Proc. Natl Acad. Sci. USA 89, 6711-5. Lueders, K.K. and Fewell, J.W. (1994) Hybridization of DNA in

419 dried gels provides increased sensitivity compared with hybridization to blots. Biotechniques 16, 66-7. Lyons, C.R., Orloff, G.L and Cunningham, J.M. (1992) Molecular cloning and functional expression of an inducible nitric oxide synthase from a routine macrophage cell line. J BiOl. Chem. 267, 6370-4. Mansour, S.L., Thomas, K.R., Deng, C.X. and Capecchi, M.R. (1990) Introduction of a lacZ reporter gene into the mouse int-2 locus by homologous recombination. Proc, Natl Acad. ScL USA 87, 7688-92. Martin, G.R. and Evans, M.J. (1975) The formation of embryoid bodies in vitro by homogeneous embryonal carcinoma cell cultures derived from irolated single cells. In: Sherman, M.I. and Solter, D. eds., Teratomas and Differentiation, pp. 169187. New York: Academic Press. McMahon, A.P. and Bradley, A. (1990) The Wnt-1 (int-1) protooncogene is required for development of a large region of the mouse brain. Cell 62, 1073-85. Melton, D.W. (1994) Gene targeting in the mouse. BioEssays 16, 633-38. Nathan, C. and Xie, Q.W. (1994a) Regulation of biosynthesis of nitric oxide. J. BioL Chem. 269, 13725-8. Nathan, C. and Xie, Q.W. (1994b) Nitric oxide synthases: roles, tolls, and controls. Cells 78, 915-8. Ochman, H., Ajioka, J.W., Garza, D. and Hartl, D.L. (1989) Inverse polymerase chain reaction. In: Erlich, H.A. ed., PCR Technology: Principles and Applications for DNA Amplifications', pp. 105-11. New York: Stockton.

Ramirez-Solis, R., Davis, A. and Bradley, A. (1993) Gene targeting in embryonic stem cells. In: Wasserman, P.M. and DePamphilis, M.L. eds., Guide to Techniques in Mouse Development, pp. 855-878. New York: Academic Press. Raschke, W.C., Baird, S., Ralph, P. and Nakoinz, I. (1978) Functional macrophage cell lines transformed by Abelson leukemia virus. Cell 15, 261-7. Reaume, A.G., Desousa, P.A., Kulkarni, S., Langille, B.L., Zhu, D.G., Davies, T.C., Juneja, S.C., Kidder, G.M. and Rossant, J. (1995) Cardiac malformation in neonatal mice lacking cormexin43. Science 267, 1831-4. Riele, H. te., Maandag, E.R. and Berns, A. (1992) Highly efficient gene targeting in embryonic stem cells through homologous recombination with isogenic DNA constructs. Proc. Natl Acad. Sci. USA 89, 5128-32. Robbins, L (1993) Gene targeting. The precise manipulation of the mammalian genome. Circ. Res. 73, 3-9. Routtenberg, A. (1995) Knockout mouse fault lines. Nature 374, 314-5. Sanger, E, Nicklen, S. and Coulson, A.R. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl Acad. Sci. USA 74, 5463-7. Scharf, S.J., Horn, G.T. and Erlich, H.A, (1986) Direct cloning and sequence analysis of enzymatically amplified genomic sequences. Science 233, 1076--8. Shastry, B.S. (1994) More to learn from gene knockouts. Mol. Ceil Biochem. 136, 171-82. Soriano, R (1995) Gene targeting in ES cells. Annu. Rev. NeUrosci. 18, 1-18. Soriano, P., Montgomery, C., Geske, R. and Bradley, A, (1991) Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693-702.

420 Southern, E.M. (1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol. Biol. 98, 503-17. Tarakhovsky, A., Muller, W. and Rajewsky, K. (1994) Lymphocyte populations and immune responses in CD5-deficient mice. Eur. J Immunol. 24, 1678-84. Thomas, K.R. and Capecchi, M.R. (1987) Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell 51, 503-12. Tsao, S.G., Brunk, C.E and Pearlman, R.E. (1983) Hybridization of nucleic acids directly in agarose gels. Anal Biochem. 131, 365-72. Tybulewicz, V.L., Crawford, C.E., Jackson, P.K., Bronson, R.T. and Mulligan, R.C. (1991) Neonatal lethality and lymphopenia in

Randolph et al.

mice with a homozygous disruption of the c-abl protooncogene. Cell 65, 1153-63. Uvarov, V.Y. and Lyashenko, A.A. (1995) The identification of the pterin-binding domain in the nitric oxide synthase's sequence. Biochem. Biophys. Res. Commun. 206, 736-41. Xie Q.W., Cho, H.J., Calaycay, J., Mumford, R.A., Swiderek, K.M., Lee, T.D., Ding, A., Troso, T. and Nathan, C. (1992) Cloning and characterization of inducible nitric oxide synthase from mouse macrophages. Science 256, 225-8. Xie, Q.W., Cho, H., Kashiwabara, Y., Baum, M., Weidner, J.R., Elliston, K., Mumford, R. and Nathan, C. (1994) Carboxyl terminus of inducible nitric oxide synthase. Contribution to NADPH binding and enzymatic activity. J Biol. Chem. 269, 28500-5.