Restriction mapping of recombinant λ DNA molecules using pulsed field gel electrophoresis

ANALYTICAL

BIOCHEMISTRY

19 1,

70-74 (1990)

Restriction Mapping of Recombinant X DNA Molecules Using Pulsed Field Gel Electrophoresis Bruce

L. Goode

Department

Received

April

and Stuart

of Biological

C. Feinstein

Sciences and Neuroscience

Research Institute,

University

of California,

Santa Barbara,

California

93106

2, 1990

We have integrated pulsed field gel electrophoresis with the partial digestion strategy of Smith and Birnstiel(lS76, Nucleic Acids Res. 3,2387-2398) to generate a rapid and accurate method of restriction endonuclease mapping recombinant X DNA molecules. Use of pulsed field gels dramatically improves the accuracy of size determination and resolution of DNA restriction fragments relative to standard agarose gels. Briefly, DNA is partially digested with restriction enzymes to varying extents and then hybridized with a radiolabeled oligonucleotide which anneals specifically to one of the X cohesive (cos) ends, effectively end labeling only those digestion products containing that cos end. In this study, we have used an oligonucleotide hybridizing to the right cos end. DNA is then fractionated by pulsed field gel electrophoresis, the gel dried down, and cos end containing fragments visualized by autoradiography. Fragment sizes indicate the distances from the labeled cos end to each restriction site for the particular restriction enzyme employed. This procedure requires only minimal quantities of DNA and is applicable to all vectors utilizing X cos ends. O1990AcademicPress,Inc.

Determination of an unambiguous restriction map for cloned DNA fragments inserted into either plasmid or phage vectors is an essential component of virtually all molecular biology endeavors. Plasmids, which are generally in the size range of 3-10 kbp, can be readily mapped by the procedure of Smith and Birnstiel (1). This procedure involves asymmetric end labeling of a linearized plasmid, partial restriction enzyme digestion, and fractionation of the products by standard agarose gel electrophoresis. Finally, the gel is dried onto filter paper or a membrane and the labeled DNA fragments visualized by autoradiography. Analysis of the resulting autoradiogram allows the generation of an accurate restriction map for each enzyme tested.

Unfortunately, this procedure has not been readily adapted to X vectors, primarily because fractionation of DNA fragments in the required size range for X DNA vectors (lo-40 kbp) is poor using standard agarose gel electrophoresis. Additionally, asymmetric end labeling of X DNA is not a frequently performed manipulation. Consequently, restriction map determination for fragments cloned in X vectors has been approached generally by the time-consuming and tedious task of fragment isolation and subcloning into plasmid vectors, followed by restriction mapping each subclone and establishing the relative orientation of each subcloned fragment in the original recombinant molecule. These strategies require the isolation of a significant quantity of phage DNA, which can be difficult with many X vectors. This situation is enough of a problem that some of the more recently constructed X vectors have been specifically designed with features to simplify restriction mapping (for example, see Ref. (2)). Useful though they may be for mapping clones derived from newly constructed libraries, these new vectors are of no use to the large number of X-based libraries already in use. Given the importance and utility of Xvectors in molecular biology, we sought to design a rapid, simple, and accurate method for restriction mapping these molecules. The central problem of poor fractionation and resolution of DNA fragments in the 10 to 40-kbp size range using standard agarose gel electrophoresis has been overcome by employing pulsed field agarose gel electrophoresis (3). The second issue, i.e., achieving asymmetric end labeling of X DNA, is readily overcome by hybridizing 32P-labeled oligonucleotides complementary to one of the X cohesive (cos)l ends of the vector DNA, as described by Rackwitz et al. (4). With this combination of procedures, it is now possible to accurately restriction map X clones in a matter of days with only a 1 Abbreviations used: dium dodecyl sulfate.

cos, cohesive;

70

DTT,

dithiothreidol;

0003-2697/90 All

SDS,

so-

$3.00

Copyright 0 1990 by Academic Press, Inc. rights of reproduction in any form reserved.

RESTRICTION

few micrograms of DNA. Below, dure and demonstrate its utility.

MAPPING

we describe

RECOMBINANT

the proce-

METHODS Cohesive end probe. A DNA oligomer 12 nucleotides long was synthesized on an ABI oligonucleotide synthesizer with the sequence complementary to the right cos end of X DNA (5GGGCGGCGACCT3’). The oligomer sample (1 ml) was deprotected, dried in a Speed Vat microcentrifuge (Savant), and resuspended in 0.1 ml TE (10 mM Tris-HCl, pH 7.6/l mM EDTA). A l-ml quantity of 100% ethanol was added and the sample dried again in a Speed Vat to remove traces of ammonia. The dried pellet was resuspended in 0.1 ml TE. The concentration of the oligomer was determined spectrophotometrically. Oligonucleotides were end labeled with T4 polynucleotide kinase and [-r-32P]ATP (10 mCi/ml; 3000 Ci/ mmol in Tricine). [32P]ATP (180 &i) was dried in a Speed Vat microcentrifuge and resuspended in 20 ~1 of kinase buffer (50 mM Tris-HCl, pH 7.8110 mM MgClJ 10 mM DTT/l mM spermidine) containing 24 ng (5 pmol) of oligomer DNA. To this mixture was added 2 ~1 (20 units) of T4 polynucleotide kinase, followed by incubation at 37°C for 1 h. Labeled oligonucleotides were separated from unincorporated nucleotides with a 1 ml Sephadex G-50 column. The recovered volume was adjusted to 24 ~1 with TE, bringing the oligonucleotide concentration to 1 nglpl. Restriction digestions. Partial restriction digestions of X DNA were performed as follows: 1.2 pg X DNA, in 40 ~1 of appropriate restriction enzyme buffer, was preincubated at 37°C for 3 min. Next, 1 unit of the restriction enzyme of choice was added. Aliquots of 10 ~1 were removed after 30 s, 2 min, and 5 min and added to tubes containing 2 ~1 of 0.5 M EDTA to end the digestion. After the removal at the 5-min time point, an additional 9 units of restriction enzyme was added to the remaining reaction volume and the tubes were incubated at 37°C for an additional hour. All samples were adjusted to 20 ~1 by addition of TE. Hybridization of labeled cos oligonucleotide to digested DNA. A 2-~1 sample of 1 M NaCl was added to each restriction digestion time point, followed by 1 ng (1 ~1) of end-labeled cos oligonucleotide. Reactions were incubated at 75°C for 2 min to denature annealed cos ends from the X vector. Next, reactions were incubated at 45’C for 30 min to allow oligonucleotide hybridization to the free cos ends. A ~-PI quantity of gel sample buffer (25% ficoll/lO mM EDTA/l% SDS/0.25% bromphenol blue/0.25% xylene cyanol) was added to each reaction, and samples were immediately loaded onto agarose gels for electrophoretic separation. Agarose gel electrophoresis. Standard loo-ml agarose gels (18 X 13 cm) consisted of 0.4% agarose in a

X DNA

BY

GEL

ELECTROPHORESIS

71

0.5X Tris borate buffer system (1X TBE = 90 mM TrisHCl, pH 8.3/90 mM boric acid/2 mM EDTA). Gels were run at 0.6 V/cm at room temperature. CHEF gels (100 ml) (Bio-Rad) consisted of 1.2% agarose in a 0.5~ TBE buffer system, and were run at 200 V on a 2.0- to 8.0-s pulsing ramp. Gel buffer was recirculated and maintained at 14°C. Following fractionation, the gels were dried onto Whatman 3MM paper and autoradiographed. Materials. Wild-type X DNA and restriction enzymes were purchased from Bethesda Research Laboratories. T4 polynucleotide kinase was purchased from International Biotechnologies, Inc. [y-32P]ATP was purchased from New England Nuclear, and the CHEF gel apparatus was purchased from Bio-Rad. RESULTS Comparison of Pulsed Field and Standard Agarose Gel Electrophoresis: Linearity of Migration and Size Determination of Fractionated DNA Fragments In order to adapt the procedure of Smith and Birnstiel(1) for DNA cloned into X vectors, it is necessary to resolve and accurately determine the size of DNA fragments in the range of lo-40 kbp. One approach to improve the resolution of large DNA fragments using standard agarose gel electrophoresis is to reduce the agarose concentration in the gel, as well as to reduce the voltage and increase the run time. The practical lower limit for agarose gel concentration is 0.4%, below which the gel can no longer be easily handled. However, the recent development of pulsed field gel electrophoresis technologies has allowed extremely large fragments of DNA (megabases) to be fractionated in sturdy, higher percentage agarose gels, in the 0.8-1.5% range (3). Given these considerations, we examined the utility of pulsed field electrophoresis to resolve and accurately determine the size of DNA fragments in the lo- to 40kbp size range. Initially, we compared the fractionation of a fulllength 44-kbp recombinant h DNA molecule (S. C. Feinstein, unpublished data) and HindIII-digested wild-type X DNA molecular weight standards, using standard 0.4% agarose gels and pulsed field 1.2% agarose gels. DNA sequencing has demonstrated that the HindIII-digested X DNA fragments range in size from 23.3 to 0.5 kbp. An additional fragment is often observed at 27.6 kbp when the Hind111 size standards are observed on ethidium bromide-stained gels, resulting from the annealing of the 23.3-kbp fragment and the 4.3-kbp fragment by virtue of their complementary cos ends. In order to directly compare the two electrophoretie systems (pulsed field vs standard), the 9.4-kbp fragment of HindIII-digested X DNA was allowed to migrate exactly 80 mm from the well in both gels. Figure 1

72

GOODE

AND

100000~

q

1.2% CHEF Gel

*

0.4% Standard

Agarooe

of Migration

in Millimeters

Gal

D 0 f s z E E 5 E

10000~

2 6 5 r

loo0 -0

Distance

FIG. 1. Comparison of CHEF and standard gel electrophoresis for the fractionation of ZfindIII-digested X DNA Size Standards. HindIII-digested wild-type X DNA size standards and an undigested 44-kbp recombinant X DNA molecule were fractionated on either a 1.2% agarose CHEF gel or a 0.4% standard agarose gel. DNA bands were visualized by ethidium bromide staining.

is a plot of the distance migrated by each DNA fragment as a function of the logarithm of its size, a plot that should in principle be linear (5). Two points are especially noteworthy. First, the pulsed field gel plot shows essentially no deviation from linearity whereas the plot for the standard agarose gel is nonlinear in the higher molecular weight range. Second, the slope of the curve for the pulsed field gel is not as steep as the standard gel for fragments in the larger size range. Both of these features improve the accuracy with which sizes can be obtained for fragments in the relevant lo- to 40-kbp range. These data indicate that pulsed field gel electrophoresis should provide more accurate size determination than standard agarose gel electrophoresis in the loto 40-kbp size range.

Comparison of Puked Field and Standard Electrophoresis: Restriction Mapping Wild-Type X DNA

Agarose Gel

We next sought to test the relative merits of the two electrophoretic systems, standard 0.4% agarose and pulsed field 1.2% agarose, in the task of restriction mapping a X DNA molecule. Since the entire sequence and restriction map of wild-type X DNA (48.5 kbp) is known (6), it was used as a substrate for this comparison. Equal aliquots of wild-type h DNA were incubated with BamHI, BgZII, EcoRI, and HindIII, each in their respective digestion buffers, and aliquots removed at various times as described under Methods. The samples were then hybridized to the 32P-labeled cos-R oligonucleotide and each of these samples divided in half. One aliquot

FEINSTEIN

was fractionated on a 0.4% standard agarose gel and the other on a 1.2% pulsed field agarose gel. DNA fragment sizes were determined from the resulting autoradiograms. As an example, Fig. 2 displays the pulsed field gel autoradiogram. Table 1 compares these experimentally determined values with the actual sizes as determined by DNA sequence analysis (6). There are two key points to note. First, these data demonstrate that the pulsed field gel system is capable of resolving bands of similar but distinct size that fail to resolve in the standard gel system. For example, the 23.5i25.5 doublet generated by Hind111 digestion is resolved by the pulsed field gel but not by the standard gel. Thus, closely spaced restriction sites that are undetected using a standard gel can be readily detected with a pulsed field gel. Second, comparison of the observed and actual sizes of DNA fragments demonstrates that both gel systems have about the same average error of size determination for fragments less than 20 kbp in size. (Percentage error is defined as the difference between the calculated and actual size divided by the actual fragment size, multiplied by 100.) However, the pulsed field gel system is much more accurate than the standard gel system for DNA fragments between 20 and 44 kbp. In this size range, pulsed field gels yield fragment sizes with an average error of 3.2%, whereas standard gels have an average error of 10.5%. Most relevant to this project, these larger sizes are essential for restriction mapping of cloned DNA in Xbased vectors. The increased accuracy of pulsed field gels relative to standard gels is consistent with the implication of the data in Fig. 1, which indicate that the pulsed field system should provide more accurate and reliable size determinations. DISCUSSION The most frequently used procedures to map X phage DNA require the time-consuming and tedious purification of significant quantities of DNA, followed by fragment isolation, subcloning, mapping, and ordering. We have modified the much simpler plasmid restriction mapping procedure originally described by Smith and Birnstiel (l), allowing adaptation of the procedure for X vectors. The key to the presently described method is the use of pulsed field gel electrophoresis to replace standard gel electrophoresis. There are several advantages to this substitution. First, fragment migration using pulsed field gels is linear when plotted against the logarithm of size, in contrast to similar fractionation of DNA fragments on standard gels where migration is very nonlinear. This feature greatly improves the accuracy of fragment size determination. Second, the pulsed field system markedly improves resolution in the relevant lo- to 45-kbp size range compared to standard gels, making it possible to resolve closely spaced restriction sites. Finally, standard electrophoretic fractionation re-

RESTRICTION 1

2

MAPPING 3

4

5

RECOMBINANT 6

7

8

9

10

X DNA 11

12

13

BY 14

GEL 15

73

ELECTROPHORESIS

16 17

18

19

-48.5

Kbp

-21.6 -23.1

Kbp Kbp

- 9.4 Kbp - 6.5 Kbp

- 4.5 Kbp

FIG. 2. 1.2% CHEF gel fractionation of asymmetrically labeled, partially digested wild-type X DNA. h DNA was end labeled, digested, and separated on a 1.2% CHEF gel as described under the Methods section. Lanes 1, 10, and 19 are HindIII-digested end-labeled X DNA size standards. (Precise position on the autoradiogram was determined from a lighter exposure.) Lanes 2-5, increasing extent of digestion with BarnHI. Lanes 6-9, increasing extent of digestion with EcoRI. Lanes 11-14, increasing extent of digestion with BglII. Lanes 15-18, increasing extent of digestion with HindIII.

quires use of a 0.4% agarose gel, difficult to handle, whereas the uses a sturdy and easily handled the procedure described here,

TABLE Comparison of CHEF Determinations

Enzyme

which is very fragile and pulsed field alternative 1.2% agarose gel. With it is possible to obtain

1

and Standard Gel Electrophoresis with Known Sizes of Restriction Digested X DNA Calculated/actual

DNA CHEF

Size

size (kbp)

gel

BamHI BglII EcoRI Hind111

15.0/14.0; 20.0/20.5; 25.0/26.2; 44.0143.0 9.7/9.7; X/9.8; 10.5110.4; 13.502.8; 25/O/26.1; X148.1 lO.Oi9.3; 18.0116.8; 23.0122.4; 27.0127.3 ll.O/ll.o; X/11.1; 12.0/11.7; 21.0/21.1; 24.0/23.5; 26.0125.5

BamHI B&II EcoRI Hind111

14.004.0; 10.0/9.7; 10.0/9.3; 10.5/11.0;

Standard gel 20.5/20.5; 30.0/26.2; X143.0 X/9.8; 11.0/10.4; 13.0112.8; 24.0/26.1; 17.0/16.8; 24.0/22.4; 32.0127.3 X/11.1; 11.0/11.7; 20.5121.1; X123.5;

Note. Data was taken from sizes were taken from Daniels Daniels et al., but not observed

autoradiograms (6). X, fragments in the analysis.

X148.1 27.0/25.5

(not shown). Known known to exist from

accurate and reliable restriction maps of recombinant X DNA molecules in a matter of a few days using only a few micrograms of isolated X DNA. In some respects, it is now even easier to restriction map X DNA molecules than plasmids. For example, in order to restriction map plasmids by the procedure of Smith and Birnstiel (l), preliminary restriction digestions must be performed on the plasmid in order to identify enzymes with single recognition sites and their approximate positions must be determined. Such preliminary data is unnecessary when using the cos oligonucleotide end-labeling strategy. Additionally, it should also be noted that restriction maps can be generated independently from both ends of X DNA, by using both cos-L and cos-R labeled oligonucleotides in parallel experiments. The two resulting maps should corroborate one another and further refine the accuracy of the final map. In summary, the utilization of pulsed field gel electrophoresis makes it possible to readily restriction map all X-based vectors using a modification of the commonly used procedure of Smith and Birnstiel(1). It should also be noted that this technology will be equally applicable to other vectors making use of X cohesive ends, including cosmids and X transducing phage. Finally, consider-

74

GOODE

AND

ing the ever-increasing use of large DNA fragments (such as those likely to be involved in the “human genome project”), the combination of asymmetric end labeling by oligonucleotide hybridization and pulsed field gel electrophoresis may have even wider applicability in the near future.

ACKNOWLEDGMENTS We are grateful to Terri Burgess and Jim Cooper for valuable comments on the manuscript and to John Carbon and Louise Clarke for helpful discussions regarding the feasibility of this project. This work was supported by grants to S.C.F. from the Muscular Dystrophy Association and the National Institutes of Health (ROl NS24387).

FEINSTEIN

REFERENCES

1. Smith,

H. O., and Birnstiel,

M.

L. (1976)

Nucleic

Acids Res. 3,

2387-2398.

2. Frischauf,

A. M., Lehrach,

H., Poustka,

A., and Murray,

N. (1983)

J. Mol. Biol. 170, 827-842. 3. Schwartz, 4. Rackwitz,

D. C., and Cantor,

C. R. (1984)

Cell 37,67-75.

H. R., Zehetner, G., Murialdo, H., Delius, J. H., Poustka, A., Frischauf, A., and Lehrach, H. (1985) 259-266. Zwann, J. (1967) Anal. Biochem. 21,155-168.

5. 6. Daniels,

H., Chai, Gene 40,

D. L., Schroeder, J. L., Szyhalski, W., Coulson, A. R., Hong, G. F., Hill, D. F., Petersen, G. B., and Blattner, F. R. (1983) Appendix II: Complete Annotated Lambda Sequence in Lambda Z1, (Hendrix, R. W., Roberts, J. W., Stalhl, F. W., and Weisberg, R. A., Eds.), Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.