Reprinted from Proc. Nat. Acad, Sci. USA Vol. 72, No. 6, pp. 2207-2211, June 1975 Electrophoretic Separation of Bacillus subtilis Genes (EcoR1/agarose gel electrophoresis/genetic transformation /DNA/restriction nuclease) R. M. HARRIS-WARRICK, Y. ELKANA*, S. D. EHRLICH}, AND J. LEDERBERG Department of Genetics, Kennedy Laboratory for Molecular Medicine, Stanford University Contributed by Joshua Lederberg, March 26, 1975 ABSTRACT The cleavage of Bacillus subtilis DNA by EcoR, restriction endonuclease produces segments which retain various degrees of genetic transforming activity. The active segments analyzed thus far, range in size from 23 to 3 kilobases and can be partially separated by agarose gel electrophoresis. Various markers can thus be enriched from 30- to 60-fold. The discovery of site-specific endonucleases (DNA-restriction enzymes) has furnished a major technical advance in separat- ing genes. Site-specific cleavaget of a uniform population of DNA molecules has yielded an ensemble of segments, f unique in size and genetic composition. The utility of these for phys- ical and chemical methods of separation has already been demonstrated in several laboratories (1, 2) by the separation of DNA segments of small viruses and plasmids. The separation of genetically specific DNA from bacteria and higher organisms is complicated by the greater size of the genome and by the vulnerability of the chromosome to random shear during isolation; thus, the molecules subse- quently exposed to cleavage are not altogether homogeneous. Nonetheless, if these molecules are larger than the produced segments, effective assortment of genes into unique segments should still be possible, with only minor contamination due to shear. Cleavage of Hemophilus influenzae DNA by the ZH. parainfluenzae endodeoxyribonuclease, followed by size-sepa- ration of segments by neutral sucrose gradient centrifugation, has been reported (3). However, the enrichment of specific activity by this method is not significantly better than was previously achieved (4). We have found that enrichment and separation of specific genetic activity of bacterial DNA can be improved with cleavage by the EcoRy endonuclease followed by low-voltage agarose gel electrophoresis. We used transforming DNA from Bacillus subtilis for biological assay. The reduction of bio- logical activity seen after cleavage, is comparable to that caused by shear-paring of the DNA molecules. Cleayed DNA, after size-sorting by electrophoresis in agarose gels, gives a reproducible banding pattern. When the DNA in these gels is assayed for biological activity, peaks of activity for different genetic markers are observed in positions that correlate with Abbreviation: Kb, kilobase. * Permanent address: Department of Molecular Biology, Hebrew University, Hadassah Medical School, Jerusalem, Israel. Tt On leave from the Centre National de la Recherche Scientifique, Paris, France. {t Cleavage and segment will be used to describe interval-specific- DNA products, in contrast to fragments caused by breakage or shear). 2207 Medical School, Stanford, California 94305 the DNA banding pattern. Almost complete separation of several genes, and enrichments in specific transforming activity of 30- to 60-fold, have been attained. MATERIALS AND METHODS B. subtilis Strains used are described in Table 1. They are all derivatives of 8B168, widely used in different laboratories. DNA Preparation. The method described by Klotz and Zimm (5) was modified as follows. B. subtilis cells were grown in minimal Spizizen medium supplemented with 0.1% of hydrolyzed casem, 50 «g/ml of tryptophan, 200 yve/ml of deoxyadenosine, and 1 wCi/ml of (H]thymidine (NEN, diluted to a specific activity of 100 mCi/mmol). Cells were harvested in mid-exponential phase, washed with standard saline-citrate solution (0.15 M sodium chloride-0.015 M sodium citrate, pH 7), and resuspended in 1/100 volume of 20% sucrose in 0.05 M Tris-HCI—1 mM EDTA at pH 7.6; 10 mg/ml of lysozyme was added, and the suspension incu- bated at 37° for 10 min. Two volumes of 1% lauroy! sarcosy- late in 0.1 M EDTA at pH 9.6 were added, followed by addition of 10 mg/ml of Pronase. The solution was incubated at 50° until clarified, and 1.23 times (w/w) of CsCl was added. This solution was centrifuged in a Spinco rotor 40 for 48 hr at 36000 rpm. Centrifuge tubes were pierced, DN A-containing fractions pooled and dialyzed against 10 mM Tris-HCl at pH 8, 10 mM NaCl, 1 mM EDTA at pH 8 (TEN). The yield of DNA was approximately 0.5 mg/100 ml of culture, with a specific activity of about 20,000 epm/pe. EcoR, Enzyme Digestions. YN A (20-400 ug/ml) was ineu- bated with the enzyme (a gencrous gift of Drs. S. Cohen and P. Wensink) at 37° for 10-60 min in a solution containing 100 mM Tris-HCl at pH 7.6, 5 mM MgCls, and 20-100 mM NaCl (gelatin was sometimes added to 0.01%). Completeness of degradation was verified by sucrose gradient centrifuga- tions. Electrophoresis was performed in tubes (7, 1); the buffer contained 0.5 ug/ml of ethidium bromide. DNA was visu- alized and photographed under long-wavelength UV light (UVL-21, Ultraviolet Products, Inc., San Gabriel, Calif.). DNA was recovered from gels for transformation by finely mashing 1-5 mm thick slices and incubating the slurry in 0.2-1 ml of Spizizens minimal medium supplemented with 0.5% glucose and 0.02 M MeCl, at 37° for 2-3 hr. Close to 60% of DNA is found in the supernatant by that time. In some cases, a slice of gel containing DNA of interest was placed on top of a new gel, and the DNA was electrophoresed into the new gel for 10 min at 150 V, followed by 12-24 hr at 20 V (2 V/cm of gel). 2208 Genetics: Harris-Warrick eé al. n Anas A A \ VA ; IK yu Vy WM \l \ N \ \ Hp) WN " | Lhe y | "fh al Me™ ! Ni ‘ re y 7 hb v | 7 OPTICAL DENSITY sv40 “«—— DISTANCE TRAVELED Fic. 1. Electrophoretogram of EcoRr DNA in agarose gels. B. subtilis DNA (300 ug/ml) was cleaved with EcoR: endonuclease (see Materials and Methods). A sample containing 10 yl of diges- tion mixture and 5 ul 0.05% bromphenol blue in 30% sucrose was layered on a5 X 100 mm gel of 0.74% agarose in 0.089 M Tris, 0.002 M EDTA, 0.089 M borate buffer (TEB). The sample was run into the gel for 3 min at 150 V and thereafterat 2 V/cm of gel (20 V) for 24 hr. Photographs were made under long wavelength ultraviolet light, and these were scanned by the Joyce-Loebel Densitometer. The locations of DNA from bacteriophage T7 and simian virus 40 (SV40) are indicated for comparison. RESULTS Sedimentation and Electrophoresis of EcoR; Degraded DN A. Sucrose gradient centrifugation of B. subtilis DNA before and after degradation with KcoRrz restriction endonuclease showed a reduction of weight-average molecular weight from 80 to close to 8 million. The final value is close to that reported for other DNAs exposed to the same treatment (8). Electrophoresis of cleaved B. subtilis DNA in agarose gels results in the appearance of a number of bands in a repro- ducible pattern (Fig. 1A). A similar finding was reported for EcoRy-cleaved E. coli DNA (1). Densitometric measurement of gel fluorescence is shown in Fig. 1B. The bands observed in Fig. 1A do not correspond to unique chromosome segments since the sum of their sizes is only about one tenth that of the entire B. subtilis chromosomes, (2500 kilobase), (5); this TaBLE 1. Strains used: Recipient strains of donor strain SB1070 (genotype thyA thyB) Recipient Strains Genotype S8B130 aro hisB SB194. trypE 5SB326 gly met SB420 trypB inh-1 SB564 aroB trypE tyrA SB626 aroBtyrA hisA ura-1 SB863 aroB trypC tyrA hisA eys-1 leu str SB1023 aroB trypC hisB tyrA cys-t lys SB1035 ura-1 arg tryp leu SB1067 purA1é metBs argC4 leus All strains have been deposited in the B. subéil’s stock collection of the Department of Genetics, Stanford University. Pedigrees will be provided upon request. Proc. Nat. Acad. Sct. USA 72 (1975) RELATIVE DNA QUANTITY MOLECULAR SIZE (KILOBASES) Fic. 2. DNA patterns obtained by computer simulation of electrophoretic separation of B. subtilis DNA fragments (A) or segments (B). finding is corroborated by the high background fluorescence displayed in Fig. 18. It would indeed be surprising if the several hundreds of segments produced by the cleavage of B. subtilis genome were resolvable by our electrophoretic proce- dures and the observed banding was unexpected. Either a patterned distribution of EcoR; cleavage sites along the chromosome, or a random clustering of sizes would lead to appearance of bands. To distinguish between these possibili- ties, a computer simulation of the cleavage was performed. It was postulated that shearing occurs at random during isolation, and that the cleavage sites are randomly located on the chromosome; i.e., the size of segments follows a discrete exponential distribution. Fig. 2 shows the distribution of fragments obtained by simulating the response of a 2500 kb molecule: (a) for random shear breakage to pieces of 10 kb (average size); (3) for random shear breakage to 120 kb (the size of B. subtilis DNA after isolation) followed by cleavage at fixed points, distributed randomly along the chromosome, to the 10 kb size. The first mode of breaking gives rise to a continuous distribution (Fig. 2A), the second to a pattern resembling the banding observed in the electrophoretogram of the HcoR, degraded DNA (Fig. 23). Consequently, the electrophoretic pattern of the cleavage products of B. subtilis (and, very likely Escherichia colt) chromosome can be ex- plained without postulating non-random intervals of EcoRy sites within the chromosome. The qualitative aspects of the simulation pattern do not change with assumed further shear of the chromosome as long as the weight-average molecular weight of the fragments before cleavage exceeds that expected for a 30 kb molecule; more extensive breakage progressively blurs the pattern. Transforming Actwity of HeoR, Degraded (Cleaved) B. subtilis DNA. The survival of different markers after KcoRy cleavage varies from close to 60% GnetBS) to below 0.1% (ys, gly) (Table 2). Residual activity does not change with either pro- longed incubation of DNA with the enzyme, or further addi- tion of enayine. This finding is difficult to relate to the simplest all-or-none madel of marker inactivation due to cleavage sites within the corresponding gene. An alternative model better fits our data: the survival of a marker is postulated to be a funetion of two variables, the length of the fragment carrying Proc. Nat. Acad. Sct. USA 72 (1975) TaBLe 2. Survival of biological activity of genetic markers after LicoR digestion Range of Survival survival Marker Recipient (%) (%) metBS SB1067 60 59-60 .5 leu SB1023 10 7-13 trypC-tyrA SB863 8.2 6-13 irypE SB194 9 trypC SB1023 il 5-15 hisB SB1023 1 5-15 tyrA SB1023 14 5-15 inh-1 SB420 19 hisA SB863 6.7 4.3-11.3 cys SB1023 2.2 0.7-3.6 ura-1 SB1035 4.5 2.7-6.9 ade1l6 $B1067 4.3 0.5-8 arol S$B130 10 aroB S8B1023 0.36 0.25-0.47 lys S8B1023 0.02 0-0.1 gly SB326 0 DNA from 8B1070 was degraded with EcoRx endonuclease as described in Materials and Methods. Completion of reaction was tested by determination of molecular size by neutral sucrose gra- dient centrifugation. Biological activity was measured by trans- formation at limiting DNA concentrations. Survival is expressed as percentage of transforming activity remaining after HcoRr digestion. Assays were conducted as in (6). the marker, and the distance of the marker from the site of action of the endonuclease; however, survival is indepen- dent of the configuration of the DNA termini. From it we can predict that: (a) transforming activity of DNA will be reduced by both shear and by cleavage to about the same level, provided that: (1) the average size of fragments produced by either shearing or enzymatic digestion is similar; (2) the number of cuts intro- duced by shearing into the marker studied is low; (3) the marker is not so close to the cleavage site as to be completely inactivated by the EcoR; cut. To test this prediction, we sheared B. subtilis DNA to a (weight-average) size of 13 kb TABLE 3. Separation of genetic markers by agarose gel electrophoresis Other markers trypC- Peak tyrA metBS hisA ura-1 trypC-tyrA _ <0.001 0.005 0.06 metBS 0.004 — 0.001 0.02 hisA <0.0003 0.04 — 0.014 ura-1 <0.0003 0.01 0.002 — DNA from 8B1070 was digested by EcoR; and 3 yg of DNA loaded on 0.73% agarose gels in TEB buffer. After an initial elec- trophoresis of 150 V for 3 min, the gels were electrophoresed at 20 V for 24 hr. Slices (1 mm) were cut and DNA extracted as de- scribed in Materials and Methods. Biological activity was deter- mined by transformation and peaks of activity for four markers were found (Fig. 4); in each case, the activity of other markers was determined for each peak. Separation of markers was calculated as the ratio Ra/Rn, where Rp and Ra are ratios of activities of peak to contaminating marker before and after electrophoresis, respectively. Electrophoretic Separation of Bacterial Genes 2209 4.0 T T T T, 7 T A " H 4 Ms on Hn 30K i on An t ‘} i us at aa 20k aan 0.2 r, 1! of t ” 17 mt sv40 x & vt b 7 j eo 5 zor Ki ii dois z 1.0 4 i 19 w i 1f pS 2 E el ea > Zz of u 2 t [¢ ug t s s =f 4 ! 2 o Ie oy i! 3 7 | i \ x ¥ i fi i Le 1 0 x < 20 s cd T T T T T T bo B se : \ ‘ 404 SE : \ % . \ * th \ ‘ rN \ . Loe A PN OSS 3, VO" 7°? ' \ %, : . NOs ose : Yc ! \ x ¢ i ~\ “ . 0 J i i> 1 I 0 2 3 a 5 6 7 8 DISTANCE FROM ORIGIN, em Fic. 3. Separation of genetic markers by agarose gel electro- phoresis. (A) DNA from SB1070 was treated with EcoR, endo- nuclease and electrophoresed in agarose gels as in Fig. 1. Trans- forming activity of DNA after electrophoresis was measured as described in Materials and Methods; different regions of the gel were screened for different genctic markers. Peaks from left to right correspond to trypC-tyrA metB5, hisA (left-hand scale), and ura-I (right-hand scale). (B) DNA was sheared to a weight average molecular weight, corresponding to that expected for a 10 kb molecule, by passage 10 times through a 30 & 1 hypoder- mic needle, then electrophoresed and tested for biological activity as above. Top, middle, and bottom line correspond to hisA, ura-1 (left-hand scale), and irypC-tyrA (right-hand scale), respec- tively. (close to that observed after cleavage). Assuming a Poisson distribution of cuts, one can show that 90% of targets of size 1.5 kb (the length of a gene coding for a polypeptide of 500 amino acids) are intact following shear. Activity of trypC, hisB, aroB, ancl lys were 11%, 6%, 6%, and 7%, respectively, of values before shearing. Cleavage reduces activity of the first two markers to a comparable level, while almost com- pletely inactivating the last two (Table 2). These data indicate that the decrease of transforming activity of irypC and hisB resulting from cleavage of the DNA can be explained by reduction in DNA size. The inactivation of areB and lys by EcoRy, by cleavage but not by shear, could be a consequence of their close proximity to a site of action of the enzyme and subsequent destruction at some stage during recombination. (6) Reduction of transforming activity by KcoRy will be correlated with the molecular size of the DNA before degrada- tion; however, markers close to the EcoRr cleavage site will be inactivated regardless of the starting DNA size. Two experiments were performed to verify this prediction: (1) DNA was separated by preparative sucrose gradient cen- 2210 Genetics: Harris-Warrick e¢ al. 415 (+ t ! a \ f ) ‘ \ o z-0l * [W/SLNVNHOSSNVEL “TRANSFORMANTS/mi x 1072 60 ab 40 1 \ ar l +20 4 Pon ~~ 1 \N “aA aN le™ ole eee he Se \—Jo 0 2 4a 6 8 10 DISTANCE FROM ORIGIN, cm Fic. 4. Coincident electrophoretic mobility of markers in the trypE-aroE region. DNA from 8B1070 was cleaved and electro- phoresed for 18 hr as in Fig. 1; 5 mm slices were assayed for bio- logical activity of four markers (shown on A and B for clarity): (A) , trypC-tyrA coselected (scale in left ordinate); --~-— , trypC (scale on right ordinate). (B) hisB (scale on right); _--, tyrA (scales on left ordinate). In other experiments, the region of gel corresponding to the peak was sliced into 1 mm slices, and the following markers were shown to migrate together as a single peak: trypE, trypC, trypA, hisB, tyrA, inh, and arok. trifugation into fractions larger and smaller than 35 kb; both were subsequently cleaved and the decrease in trypC bio- logical activity measured. As expected, the activity of the heavier fraction of DNA fell 10-fold; the lighter fraction only 2-fold. (2) DNA was sheared to a size of 15 kb; this DNA was subsequently cleaved and the drop in biological activity for the trypC and hisB markers measured. The same experi- Proc. Nat. Acad. Sci. USA 72 (1975) EcoR, SITE EcoR, SITE re a My o <= ree] w s Qa oa a