17 GATA 9(1): 17-25, 1992

Double-Minute Chromosomes as Megabase Cloning Vehicles P E T E R J. H A H N , LEANNA GIDDINGS, J O H N L O N G O , M I C H A E L J. L A N E , J A N E S C A L Z I , and J O H N H O Z I E R

Radiation-reduced chromosomes provide valuable reagents for cloning and mapping genes, but they require multiple rounds of x-ray deletion mutagenesis to excise unwanted chromosomal DNA while maintaining physical attachment of the desired DNA to functional host centromere and telomere sequences. This requirement for chromosomal rearrangements can result in undesirable xray induced chromosome chimeras where multiple noncontiguous chromosomal fragments are fused. We have developed a cloning system for maintaining large donor subchromosomal fragments of mammalian DNA in the megabase size range as acentric chromosome fragments (double-minutes) in cultured mouse cells. This strategy relies on randomly inserted selectable markers for donor fragment maintenance. As a test case, we have cloned random segments of Chinese hamster ovary (CHO) chromosomal DNA in mouse EMT-6 cells. This was done by cotransfecting plasmids pZIPNeo and pSV2dhfr into DHFRCHO cells followed by isolation of a Neo + DHFR + CHO donor colony and radiation-fusion-hybridization (RFH) to EMT-6 cells. We then selected for initial resistance to G418 and then to increasing levels of methotrexate (MTX ). Southern analysis of pulsed-field gel electrophoresis of rare-cutting restriction endonuclease digestions of DNA from five RFH isolates indicated that all five contain at least 600 kb of unrearranged CHO DNA. In situ hybridization with the plasmids pZIPNeo and pSV2dhfr to metaphase chromosomes of MTX-resistant hybrid EMT-6 lines indicated that these markers reside on double-minute chromosomes.

Introduction Cloning and physically mapping mammalian genes have been hampered by the difficulty in maintaining contiguous DNA sequences in size ranges between Fromthe Departmentsof Radiology(P.J.H., L.G., J.L.), and Medicine(M.J.L.), StateUniversityof New YorkHealthScience Center, Syracuse, New York; and the Department of Genetics (J.S., J.H.), FloridaInstituteof Technology,Melbourne,Florida, USA. Address correspondence to Dr. P.J. Hahn, State University of New YorkHealthScienceCenter, Syracuse,NY 13210, USA. Received20 September1991;revisedand accepted17January 1992.

whole chromosomes and cosmid or yeast artificial chromosome (YAC) cloning vehicles. This deficiency results in the requirement for subcloning and analyzing a very large number of relatively small fragments to have a reasonable chance for cloning any particular DNA segment [1]. Additionally, knowledge of the genetic map of a particular chromosome is of limited utility in the construction of physical maps around a particular locus because the physical distance between even closely linked genetic markers is very large relative to genomic DNA cloned in cosmid or YAC cloning vehicles. To overcome this problem, a number of chromosome reduction strategies have evolved that either reduce the size of chromosomal segments or translocate chromosome fragments of human chromosomes to rodent cell chromosomes via radiation-fusion-hybridization (RFH) [2-5]. These strategies depend on the high probability of integration of radiation-induced chromosome fragments into host cell chromosomes [3, 6] following fusion of lethally irradiated donor cells to normal host cells. However, an unwanted side effect is that other unrelated and noncontiguous fragments are also retained resulting in chimeric hybrid clones [3]. We have been developing large genomic segment (megabase) cloning vehicles based on double-minute chromosomes (DMs). These genomic segments are generally associated with genes amplified in vitro in drug-resistant cultured mammalian cells [7, 8] and in vivo amplified oncogenes in tumor cells [8-14]. They have many properties in common with bacterial plasmids that make them potentially attractive as megabase cloning vehicles (MCVs) for mammalian genomic DNA when coupled to RFH. Specifically, their episomal nature and small size relative to intact chromosomes means that they can be readily purified from host DNA [15], and when coupled to an amplifiable marker gene such as dihydrofolate reductase (DHFR), their copy number can be increased [14, 16]. The size of these episomal elements makes them potentially ideal for cloning large genomic fragments. The reported sizes vary with cell line and system used in measurement, but the sizes reported cover a very useful range. Pulsed-field gel electrophoresis (PFGE) has been used to measure the size of DMs (also called episomes) as small as 250 kb [10] and up to as large as 5 mb [17]. All of these structures have been shown to be circular. Using light-microscopic examination of metaphase chromosomes where the lower limit of resolution is 2-5 mb, numerous DMs have been observed in a variety

© 1992ElsevierSciencePublishingCo., Inc. 655 Avenueof the Americas,New York, NY 10010 1050-3862/92/$5.00

18

GATA 9(1): 17-25, 1992

of cells both from drug-resistant tissue culture cells [11, 18] and from tumor cells [19]. The sizes therefore range from 250 kb up to perhaps 15-19 mb. The size range is important because it approximates the desired 5 centimorgan (that is, - 5 mb) resolution of the human genetic map, enabling the use of these markers to screen directly for genetically linked, but physically distant, genes. We are testing the use of DMs as MCVs by initially cotransfecting two plasmids beating the marker genes for G418 and MTX resistance into the CHO genome, followed by RFH rescue in mouse EMT-6 cells of the CHO chromosome fragments marked with these selectable genes. Selection for the markers results in survival of EMT-6 host cells harboring large genomic regions of CHO DNA flanking the selectable marker. Since the donor chromosomal DNA is never out of the protective cellular environment, sheafing and rearrangement effects common to YAC and other large genomic segment cloning systems are avoided. We report here that we have successfully cloned >600 kb of unrearanged CHO DNA as DMs in mouse EMT-6 cells by using this technique. Materials and Methods

Cells, Cell Cultures, and Plasmids Mouse EMT-6 cells were originally from Bob Kalman at Stanford, and were cultured as previously described [18], except they were cultured in Hams FI2 medium. The D H F R - C H O DG44 cells [20] were provided by L. Chasin at Columbia University. The plasmids pSV2dhfr [21 ] and pZIPNeo [22] were used in the cotransfection.

Transfection and Radiation-FusionHybridization Conditions The cotransfection of pSV2dhfr and pZIPNeo to the D H F R - CHO cells were performed exactly as described in Current Protocols in Molecular Biology [23]. Single colonies were isolated and one was chosen to be used as a donor. The RFH protocol [24] was employed as previously modified [6]. Ten single colonies were isolated. Five of these colonies were grown and analyzed by PFGE, and two of these five were analyzed by in situ hybridization.

Methotrexate Selection and Dihydrofolate Reductase Gene Amplification Selection for resistance to methotrexate (MTX) was essentially as described [17, 18]. Cells were first

P.J. Hahn et al.

isolated in 0.15 IxM MTX and then subjected to stepwise doublings in MTX concentration. When cells were able to grow well at one MTX concentration, that concentration was doubled. The process was continued for 5 RFH isolates until some developed the ability to grow in 6.4 I~M MTX.

Plug Formation Confluent cultures of cells were trypsinized, washed, resuspended in 0.75% low-melt agarose (FMC) at 37°C and poured into 200 I~1 molds (7 x 3 x 10 mm) (referred to hereafter as "plugs") at a cell concentration of 106 cells per plug. The agarose plugs were treated as previously described [ 18, 25].

Restriction Enzyme Digestion and Gel Electrophoresis Prior to electrophoresis, plugs were first redigested overnight with fresh 0.5 M EDTA, pH 8.0, 1% sarcosyl, and 0.1% proteinase K (ESP) at 55°C and washed as described [17]. Thin slices of plugs were then digested in Not I or Mlu I restriction enzymes over night at 37°C as described by the manufacturer. The digested slices were then placed into the wells of a gel (1% agarose in 1 × TBE [ 0.1 M Tris base, 0.1 M boric acid and 2 mM EDTA]) and electrophoresed in a contour-clamped homogeneous electric field (CHEF) apparatus [26] at 125 V. The initial pulse time of 60 s was ramped with 4-s increments for 60 cycles, with a final pulse time of 300 s at the end of 60 cycles. The cycles were repeated continuously for 60 h. For the Hind IlI enzyme digestion, slices from the same set of plugs were first melted at 65°C then cooled to 37°C. The digestion was carried out in the molten agarose at 37°C 2 h. The melted agarose was then pipetted into the wells of an agarose gel (0.8%) and electrophoresed with 35 V for 18 h at room temperature in a horizontal gel electrophoresis apparatus.

PFGE of Whole DMs Plug samples were exposed to 40 Gy x-irradiation to linearize the DMs. Slices of these samples were loaded into the wells of an agarose gel (1.0%) and electrophoresed in a CHEF apparatus for 12 days at 50 V with pulse times ramped to separate roughly 0.27.0 mb.

© 1992 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010

19

GATA 9(1): 17-25, 1992

Double-MinuteMCVs

Southern and in situ Hybridization Conditions Following electrophoresis, the gels were stained with ethidium bromide, photographed, and Southern blotted onto nylon membranes (Amersham). The membranes were hybridized with p32-1abeled pSV2dhfr to detect the DHFR-containing DMs or the DHFRcontaining restriction endonuclease digestion products [27]. The probes were labeled using the random primer DNA-labeling system from Gibco BRL. To locate the fragments containing the cotransfected sequences that had been transferred from CHO to mouse EMT-6 via RFH, biotinylated probes were prepared by random priming the plasmids used originally in the cotransfections. Aged metaphase chromosomes on slides were treated with RNase A (100 ~zg/ml in 0.3 M sodium chloride-0.3 M sodium citrate (2 × SSC) for 1 h at 37°C. The slides were dehydrated in an increasing ethanol series (room temperature 70% and 95%) for 3 min each, followed by denaturation in 70% formamide (BRL redistilled, nucleic acid grade) 2 x SSC (pH 7.0). The chromosomes were immediately immersed in an ice-cold ethanol series (70% and 95%) for 3 min each and then incubated with proteinase K (0.06 ~g/ml, 20 mM riffs_ HC1, and 2 mM CaCI2, pH 7.0) for 7.5 min at 37°C. The slides were redehydrated in an increasing ethanol series (room temperature 70% and 95%) for 3 min and dried using an air jet. Carrier DNA (nonbiotinylated sheared salmon sperm DNA) at 10 Ixg/ml was added to a 50-ng/ml mixture of each biotinylated probe DNA (pZIPNeo-pSV2dhfr), denatured at 75°C for 5 min, and immediately chilled on ice. Chilled hybridization cocktail (50% formamide, 2 x SSC, 10% dextran sulfate, 1 × Denhardt's solution, 40 mM phosphate buffer, and 1% sodium dodecyl sulfate) was added to the probe DNAs and the resulting hybridization mix was kept on ice until the time of application to the slides. Each slide received 50 ill of mix followed by a coverslip and was incubated at 37°C overnight in a moist chamber. Signal detection involved two posthybridization washes in 50% formamide-2 x SSC (pH 7.0) at 42°C. The slides were washed twice in 2 x SSC followed by a wash in 1 x BN buffer (0.1 M sodium bicarbonate and 0.05% Nonidet P-40, pH8.0). The slides were incubated with blocking solution (1 x BN buffer containing 0.02% sodium axide and 1% nonfat dry milk) followed by incubation with fluorescein isothyocyanate (FITC)-conjugated avidin DCS (5 Ixg/ml in blocking solution) for 20-45 min at 37°c in a moist chamber. The slides were rinsed twice in 1 × BN buffer ant 42°C (2-5 min) and amplified with anti-avidin DCS

solution (12 ixg/ml in I x BN buffer containing 6.25% goat serum) for 20-45 rain at 37°C in a moist chamber. The 1 × BN buffer rinses were repeated twice, followed by another layer of FITC avidin-DCS. The slides were rinsed in 1 x BN, and a thin layer of antifade solution containing 0.25 p,g/ml of propidium iodide was applied to counterstain the chromosomes. The slides were scanned with an Olympus BH2 fluorescence microscope. The fluorescein and propidium iodide were excited at 450-490 nm. Results

Experimental Design Our cloning strategy for large genomic segments is outlined in Figure 1. Our approach involved randomly inserting linked Neo and DHFR genes into mammalian chromosomal locations by cotransfecting pZIPNeo and pSV2dhfr into a D H F R - C H O DG44 cell line. This was followed by fragmentation of the chromosomes with lethal x-irradiation and fusion of the fragments to mouse EMT-6 cells, which we had previously determined to be capable of harboring DMs in the desired size range [17, 18]. Since the two genes are linked, successful transfer is indicated by the ability to grown in the presence of G418 due to Neo resistance, and the copy number is increased by selection for increasing levels of resistance of MTX, thus screening for DHFR activity (the ability to grow in the absence of thymidine, hypoxanthine, or glycine). In this study, one of the surviving cotransfected isolates that had received both functional genes was chosen to serve as a donor to test whether we could clone the genomic region surrounding the cotransfected genes. To clone these genomic segments as MCVs in EMT-6 mouse cells, the CHO donor cells were subjected to 100 Gy x-irradiation, fused to the mouse cells, and plated into media containing G418 to select for fusion transfer of the donor CHO genomic fragments containing the Neo gene. Five of the resulting colonies were selected for PFGE analysis of the Not I and Mlu I digestion patterns and for horizontal gel electrophoresis of the Hind III digestion patterns of the chromosomal DNA.

Southern Hybridization of Not I, Mlu I Digested CHO Donor and Mouse Fusion Isolates Southern analysis of Mlu I (Figure 2a) and Not I (Figure 2b) digests of DNA from cotransfected CHO donor probed with pSV2dhfr indicates that the CHO

© 1992 ElsevierSciencePublishingCo., Inc. 655 Avenueof the Americas,New York, NY 10010

20 GATA 9(1): 17-25, 1992

P.J. Hahn et al.

Step 1 INSERTION

DONOR (CliO DG44DHFR-)

Neo+ DHFR+ (Cotransfection)

f

Q

RECIPIENT (EMT-6 Mousecell)

Step 2 X-RAY, FUSION (tO Recipient)

/

Chromosomes with Centromeres

Fisure 1. Cloning strategy for generating donor Chinese hamster ovary (CHO) double-minute chromosomes for use as megabase cloning vehicles in mouse EMT-6 cells.

Step 3 SELECTION AMPLIFICATION

N °V/ MTX

/ h

Step 4 ANALYSIS (PFGE, Chromosome painting)

donor contains at least two polymorphic copies (or alleles) of the cointegrations of the two plasmids. The initial donor contained at least three Not I bands (0.25, 0.44, and 0.60 mb) and two Mlu I bands (0.25 and 0.60 mb). To determine whether the multiple bands were due to multiple integration of the cotransfected plasmids in the CHO donor cell, or to mixed populations, we isolated five subisolates of this cotransfected CHO donor. Four donor subisolates had an Mlu I band corresponding to the major band - 0 . 6 0 mb present in the donor population (two of which are shown in Figure 2a), and the fifth donor subisolate (also shown in Figure 2a) had an Mlu I

band of 0.25 mb similar to the minor band in the donor. This suggests that these polymorphisms are due to a mixed population in the donor CHO cells where the majority have one allele and the remainder have the other. Since the two alleles are very similar in fine structure (apparent in the Hind III digestion patterns in Figure 6 discussed below), it is likely that these alleles represent methylation variants in a minor subset of the donor population, and not rearrangements that occurred during the RFH process, although they may be associated with events related to the cotransfection process. We have cloned both of these alleles in the EMT-

© 1992 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010

21 GATA 9(1): 17-25, 1992

Double-Minute MCVs

o

o ro

~ *,9,o

~ C a

"--

,~

o

o

(~

U)

i It)

o

~

i I~,

1 t~ O..

~o ..Q t-"

0.6

0.6

c~

0.44

.o_

0.25

~

.o_ .m

E

0

0.25

Figure 2. Southern blots. (a) Mlu 1-digested CHO donor and three subisolates of the donor, the sensitive fusion isolates 5 and 7, and 1.6-1xM methotrexate (MTX)-resistant fusion isolates 5 and 7. (b) Not I digests of EMT-6, the Chinese hamster ovary (CHO) donor, and sensitive and resistant fusion isolates 5 and 7. The membranes were probed with pSV2dhfr, which hybridizes with the majority of the plasmid DMA used in the cotransfection as well as the endogenous dihydrofolate reductase (DHFR) gene. The donor contains a major Mlu I band at 0.6 mb and a minor band at 0.25 mb. Most of the subisolates of the donor and most of the fusion isolates contain a single Mlu I band at 0.6 nab. However, one of the subisolates of the donor line and one of the fusion isolates contain the minor 0.25-mb band. EMT-6 cells have a single Not I band containing the DHFR gene at --1 million base pairs and a larger Mlu I band (not shown in Figure 2a), whereas the DHFR - C H O donor has no endogenous DHFR sequences.

6 RFH isolates as MCVs. Four of the five RFH isolates (one of which, fusion isolate 7, is shown in Figure 2a) are identical and contain both the 0.25 mb Not I band (Figure 2b) and the 0.60 mb Mlu I band (Figure 2a). the fifth, fusion isolate 5, contains the 0.60 mb Not I band (Figure 2b) and the 0.25 mb Mlu I band (Figure 2a).

ferred markers are carried on structures resembling DMs (Figure 3), although somewhat heterogeneous in size. This is consistent with our observations of natural DMs in MTX-resistant populations of EMT6 cells where several size classes can be present in the same cells [17]. Therefore we have successfully cloned the CHO DNA in MCVs.

In situ Hybridization with pSV2dhfr and pZIPNeo

Stability of the Neo Gene in the Absence of Selection

To determine whether we had cloned CHO DNA as MCVs, we first performed in situ hybridization of metaphase chromosomes from the EMT-6 RFH fusion isolates by using the plasmids cotransfected into the CHO donor as probes. We expected that the biotinylated probes pSV2dhfr and pZIPNeo would be effective for in situ hybridization because we estimated that 430 kb of plasmid sequences had been integrated at the site of cotransfection (see Figure 5). Analysis using these probes indicates that the trans-

To confirm that the transferred DNA remained on selectable DMs rather than being attached to functional host centromeres, we cultured the RFH isolates in the absence of any selecting agent and assayed the cultures monthly for relative plating efficiency in G418. Two hundred cells were plated in triplicate into selective media containing G418, or into nonselective media, and the number of colonies that developed with time were compared. Figure 4 shows that all RFH isolates lost most of the Neo resistance genes

© 1992 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010

22 GATA 9(1): 17-25, 1992

P.J. Hahn et al.

Figure 3. Cytogenetic in situ hybridization of mixed pZIPNeo-pSV2dhfr probes to megabase cloning vehicles (MCVs) (radiation-fusion-hybridization fusion 5). MCVs in this cell line (large arrows) show signal hybridizing to several different size classes. The small arrow shows a nonhybridizing double-minute chromosome. Because the distinction between true signal and artifacts is difficult when analyzing small chromosomal structures such as MCVs by in situ hybridization, three criteria were used to ensure true signal: (a) the structure has a doublet appearance typical of a double-minute chromosomal fragment, (b) the color of the signal must be yellow, reflecting the green (fluorescein) signal component and the underlying red (propidium iodide) chromosomal stain, and (c) when destained for fluorescence and restained with Giemsa, obvious chromosomal material must be found in the position of the signal.

23 GATA 9(1): 17-25, 1992

Double-Minute MCVs

under nonselective conditions over the course of the study. In contrast, markers associated with centromeres were stable: the CHO donor had a 96% relative plating efficiency in G418 following 3 months in culture in the absence of G418 (data not shown). These results are also in contrast to the results in Hahn et al. [6], where RFH transfer of fragments from G418-resistant CHO cells to G418-sensitive CHO isolates resulted in isolates that could be maintained for 6 months in the absence of G418 without loss of G418 resistance.

P F G E of the CHO Fragment DMs in EMT-6 Cells To determine the total size of the DNA to which the markers were attached, and whether they were attached to MCVs small enough to enter a pulsed-field gel, x-irradiated samples were electrophoresed under conditions that would resolve MCVs between 0.2 and 7 mb. We included as a control on this gel an EMT-6 line resistant to 2.4 ~M MTX that was previously determined to contain amplified endogenous DHFR genes on DMs of --1 and --2.5 mb in size. Since we were attempting to amplify the gene copy number of the introduced chromosomal DNA, samples were analyzed both from the sensitive RFH isolates originally isolated after G418 selection and from the MTX-resistant populations derived from these cells. The RFH isolates were originally isolated in G418 and then plated into 0.15 txM MTX. Once they were resistant to 0.15 ixM MTX, MTX alone was used to maintain selection. The MTX concentration was doubled in a stepwise fashion until some of the cells reached a resistance level of 6.4 fxM MTX. Figure 4 shows that in the MTX-sensitive RFH isolates 5 and 7, the markers were attached to DNA too large to enter the gel, indicating that they were at least 10 mg in size. This is consistent with the cytogenetic observation of markers on acentric chromosome fragments since the smallest chromosomal fragment discernible by light microscopy is 2-5 mb. In the 6.4-1xM MTX-resistant isolate 5, the hybridization signal was initially associated with amplified MCVs or DMs slightly smaller than the 1-mb DM shown in the resistant control EMT-6 cells (Figure 5), but further analysis (see below) showed that it was due to amplification of a DM containing the endogenous DHFR gene.

D H F R Gene Copy Number and Fine Structure of R F H Isolates To determine the copy number of the amplified DHFR genes and whether the introduced markers or the

endogenous DHFR gene were being amplified, Southern analysis was performed on DNA from the MTX-resistant isolates that had been digested with the frequent cutting restriction endonuclease Hind III and electrophoresed under conditions that would resolve fragments of k=

r,-

20'

1

2

Months cultured in the absence of G418

Figure 4. Stability of the introduced markers in the fusion isolates in the absence of selection. The five fusion isolates were grown for 3 months in the absence of the selection agent G418 used originally to select for successful radiation-fusion-hybridization isolates. At monthly intervals, they were tested for the relative number of cells capable of giving rise to colonies in the presence of G418.

associated shearing and so forth prior to introduction into the host. An added benefit is that many donor chromosome fragments should express donor proteins which may aid in mapping studies. MCVs have several important advantages over other RFH schemes. First, since nonselected DNA is readily lost, the chimera problem associated with use of host centromeres for fragment retention in other mammalian cell host systems is avoided. Secondly, since MCVs are physically much smaller than chromosomes, they can be readily purified from host chromosomal material [ 15]. However, the niche they would fill would be the one occupied by reduced chromosomes and somatic hybrid panels. The entire human genome could be subdivided into one or two thousand MCVs, each of which could be physically mapped onto the human genome by using in situ hybridization and blotted onto a series of membranes for screening against

Figure 5. Pulsed-field gel electrophoresis of circular double-minute chromosomes (DMs) linearized with 40 Gy. DNA from EMT6 cells sensitive and resistant to methotrexate (MTX), Chinese hamster ovary donor cells, and the fusion isolates sensitive and resistant to MTX was embedded in agarose plugs and subjected to 40-Gy x-rays to linearize circular DMs and electrophoresed under conditions that would separate molecules between 0.2 and 7 mb. The DNA in the gels was then blotted onto nylon membranes, probed with pSV2dhfr, and autoradiographed. Two positive controls are included. The first is an anonymous human YAC of - 1 0 0 kb in size, and the second is a 2.4-~M MTXresistant mouse EMT-6 cell line that harbors DMs of - 1 and - 2 . 5 mb.

molecular probes. This could effectively reduce the complexity of the human genome to a series of Escherichia coli-sized genomes--a size much easier to deal with. Probably the best use would be in conjunction with YACs or cosmids to generate a series of subchromosomal libraries, each representing a defined megabase chromosomal region. A 2-mb MCV could be subcloned into two hundred 50-kb cosmids or fifty 200-kb YACs with five fold redundancy. The reasons for the preferential amplification of the endogenous DHFR gene are not clear. The transfected DHFR gene was readily amplified in the CHO donor line by using the same MTX selection protocol (data not shown). Once transferred to the mouse genome, it seemed to be more actively transcribed than the endogenous gene since a single copy conferred resistance to 1.6 ~M MTX, whereas multiple copies of the endogenous gene are required to achieve that level of MTX resistance [17]. Therefore, other factors being equal, amplification should have favored the introduced construct over the endogenous one. Additionally, in CHO cells, cotransfected DHFR genes are more readily amplified than the endogenous CHO ones [28]. We are presently constructing a DHFR-

© 1992 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010

25 GATA 9(1): 17-25, 1992

Double-Minute MCVs

F5 - sensitive

F5 - R1600

F5 - R3200

F5 - R6400

Donor

EMT-6

F1 - s e n s i t i v e

F1 - R 1 6 0 0

.Ix

O'~

t,D

Kilobase pairs Fisure 6. Southern hybridization of Hind llI-digested DNA from the fusion isolates 5 and 1 sensitive and resistant to 1600 nM MTX (1.6 IxM). Hind III digests of donor Chinese hamster ovary and EMT-6 DNA are also present. The nylon membranes were probed with pSV2dhfr, which hybridizes with the cotransfected plasmids present in the donor and the fusion isolates, and the endogenous mouse dhfr gene present in EMT-6 and the fusion isolates.

m o u s e E M T - 6 c e l l line that s h o u l d a l l o w us to address this point.

This grant was supported by National Cancer Institute grant CA46880 and a grant from Genmap Inc.

References 1. Lander ES, Waterman MS: Genomics 2:231-239, 1988 2. Cox CA, Pritchard CA, Uglum E, Cox DR, Casher D, Kobori L, Mekyers RM: Genomics 4:397-408, 1989

3. Benham F, Hart K, Crolla H, Bobrow M, Francavilla M, Goodfellow PN: Genomics 4:509-517, 1989 4. Siden TS, Hoglund M, Rohme D: Somatic Cell Mol Genet 15:245-254, 1989 5. Goodfellow PJ, Povey S, Nevanlinna HA, Goodfellow PN: Somatic Cell Mol Genet 16:163-171, 1990 6. Hahn P, Morgan WF, Painter RB: Somatic Cell Mol Genet 13:597-608, 1987 7. Van Devanter DR, Piaskowski VD, Casper JT, Douglass EC, Von Hoff DD: J Natl Cancer Inst 82:1815-1821, 1990 8. Humphrey PA, Wong AJ, Vogelstein B, Friedman HS, Werner MH, Bigner DD, Bigner SH: Cancer Res 48:2231-2238, 1988 9. Lavialle C, Modjtahedi N, Cassingena R, Brison O: Oncogene 3:335-339, 1988 10. Carroll SM, De Rose ML, Gaudray P, Moore CM, Needham-Vandervanter DR, Von Hoff DD, Wahl GM: Mol Cell Biol 8:1525-1533, 1988 11. Ruiz JC, Choi KH, Von Hoff DD, Robinson IB, Wahl GM: Mol Cell Biol 9:109-115, 1989 12. Borst P, Van der Bliek AM, Van der Velde-Koerts T, Hes E: J Cell Biochem 34:247-258, 1987 13. Bigner SH, Friedman HS, Vogelstein B, Oakes WJ, Bigner DD: Cancer Res 50:2347-2350, 1990 14. Wahl GM: Cancer Res 49:1333-1340, 1989 15. Barker PE, Stubblefield E: J Cell Biol 83: 663-667, 1979 16. Stark GR, Debatisse M, Giulotto E, Wahl GM: Cell 57:901908, 1989 17. Hahn P, Nevaldine B, Longo J (in press) 18. Hahn P, Nevaldine B, Morgan WF; Somatic Cell Mol Genet 16:423-423, 1990 19. Bigner SH, Mark J, Bigner DD: Cancer Genet Cytogenet 47:141-154, 1990 20. Urlaub B, Kas E, Carothers AM, Chasin LA: Cell 33:405412, 1983 21. Subramani S, Mulligan R, Berg P: Mol Cell Biol 1:854864, 1981 22. Cepko CL, Roberts BE, Mulligan RC: Cell 37:1053-1062, 1984 23. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K (eds): Current Protocols in Molecular Biology 1:9.9.1-9.9.6, 1987 24. Cirullo RE, Dana S, Wasmuth JJ: Mol Cell Biol 3:892-902, 1983 25. Ahn SY, Nevaldine B, Hahn J: Int J Radiat Biol 59:661665, 1991 26. Chu G, Volrath D, Davis RW: Science 234:1582-1585, 1986 27. Church GM, Gilbert W: Proc Natl Acad Sci USA 81:19911995, 1984 28. Choo KH, Filby G, Greco S, Lau YF, Kan YW: Gene 46:277-286, 1986

© 1992 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010