JOURNAL oF BacterioLocy, Mar. 1976, p. 934-945 Copyright © 1976 American Society for Microbiology Effects of Chloramine on Bacillus subtilis Deoxyribonucleic Acid KATHERINE LU SHIH'* anp JOSHUA LEDERBERG Department of Genetics, Stanford University School of Medicine, Stanford, California 94305 Received for publication 21 November 1975 The lesions induced in Bacillus subtilis deoxyribonucleic acid (DNA) after treating bacterial cells (in vivo) and bacterial DNA (in vitro) with chloramine were studied biologically and physically. Single-strand breaks and a few double- strand scissions (at higher chloramine doses) accompanied loss of DNA-trans- forming activity in both kinds of treatments. Chloramine was about three times more efficient in vitro than in vivo in inducing DNA single-strand breaks. DNA was slowly chlorinated; the subsequent efficiency of producing DNA breaks was high. Chlorination of cells also reduced activity of endonucleases in cells; however, chlorinated DNA of both treatments was sensitized to cleavage by endonucleases. The procedure of extracting DNA from cells treated with chlora- mine induced further DNA degradation. Both treatments introduced a small fraction of alkali-sensitive lesions in DNA. DNA chlorinated in vitro showed further reduction in transforming activity as well as further degradation after incubation at 50 C for 5 h whereas DNA extracted from chloramine-treated cells Vol. 125, No. 3 Printed in U.S.A. did not show such a heat sensitivity. A great deal of work has been carried out to investigate the mode of chlorine disinfection, however, much less attention has been given to the mutagenic potentiality of this disinfectant. W. C. Boyle (Ph.D. thesis, California Institute of Technology, Pasadena, 1963) demonstrated the chemical reaction between chloramine (NH,Cl) and pyrimidines, and that the reactiv- ity of cytosine is higher than that of uracil and thymidine. The possibility that deoxyribonu- cleic acid (DNA) is the site of chlorination has been raised by Prat et al. (12, 13); they obtained an appreciable quantity of 5-chlorocytosine and 5-chlorouracil from an acid hydrolysate after sodium hypochlorite (NaOCl) treatment of bae- terial DNA and ribonucleic acid (RNA), and cytosine was chlorinated more than uracil and thymidine. Evidence of incorporation of chlo- rine into DNA was also obtained by Fisenstark and Riva (also working in this laboratory; un- published data); after treating calf thymus DNA with NaOCl, radioactive chlorine re- mained in acid-precipitable material. Hayatsu et al. (6) have reported a rapid decrease in the absorbance of ultraviolet light of both calf thy- mus DNA and yeast transfer RNA after NaOCl treatment. They also found that after treat- ment of cytosine an intermediate, 4N,5-dichlo- rocytosine, was generated which was readily ! Present address: Biological Adaptation Branch, Ames Research Center, National Aeronautics and Space Admin- istration, Moffett Field, Calif. 94035. converted into 5-chlorocytosine on treatment with hydrochloric acid. Further aspects of the action of hypochlorous acid (HOC]) on cytosine under different physiological conditions have been reported by Patton et al. (11). The inhibit- ing effect of chlorine on biological activity of DNA was observed by Hsu (7) using transform- ing DNA of Haemophilus influenza pretreated with chlorine. The preliminary results of Eisen- stark (personal communication) have shown that Bacillus subtilis DNA was also reduced in its transforming activity after treatment with chloramine and hypochlorous acid and that the sodium hypochlorite-treated DNA yielded bro- ken strands. The above studies have been performed by treating the nucleic acids and nucleic acid bases with chlorine. The present study addressed the question of whether chlorine treatment of living cells could involve the modification and the destruction of the genetic material. MATERIALS AND METHODS Bacterial strains. All strains used in this article were derivatives of indole-requiring B. subtilis strain 168 (14). Cell preparation. Cells grown overnight in 5 ml of Penassay broth were centrifuged and suspended in 50 ml of Spizizen minimal medium {per liter con- tains: (NH,).SO,, 2 g; K,HPO,, 14 g; Na citrate, 1 g; and MgSO,-7H,0O, 0.2 g] supplemented with 0.5% glucose, 0.05% casein hydrolysate, and 25 wg of those nutrients required for growth of auxotrophic 934 Vou. 125, 1976 strains per ml. Thymine-requiring strains were used for DNA labeling experiments, and additional 0.35 to 0.4 mg of thymidine and 10 «Ci of “C per 0.19 umol of thymidine or 0.1 mCi of *H per 0.07 wmol of thymidine were added to 50 ml of growth medium. Cells were grown at 37 C until the start of station- ary phase and were then centrifuged and washed twice with 0.05 M phosphate buffer, pH 7, sus- pended in 12 ml] of phosphate buffer, and were either used immediately or stored in liquid nitrogen after adding dimethyl sulfoxide to a final concentration of 5%. DNA preparation. [HJDNA used for chloramine treatment was extracted from B. subtilis cells strain SB747 (hisB tryC thy) at log phase of growth. Cells were lysed in SSC (0.15 M NaCl plus 0.015 M sodium citrate) with 12% sucrose and 1 mg of lysozyme per ml at 37C for 30 min. After adding ribonu- clease (RNase), 50 g/ml, the suspension was incu- bated at 37 C for 3 h. The lysis was completed by addition of 25% sodium lauryl sulfate to a final concentration of 0.5% and incubated at 70 C for 20 min. Pronase was then added to the lysate (1 mg/ml final concentration) and the lysate was incubated at 37 C for 3 h. At the end of the incubation, 5 M NaClO, was added to the lysate to a final concen- tration of 1 M. An equal volume of chloroform and one-fifth volume of n-octanol were added and the lysate was shaken at room temperature for 30 min. After spinning the lysate at 27,000 x g for 15 min the top layer was precipitated by cold ethanol and the precipitate was dissolved in 0.1x SSC-0.02 M tris(hydroxymethyl)aminomethane (Tris)-hydrochlo- ride, pH 7.4. RNase and Pronase treatments were repeated after the addition of one-tenth volume of 10x SSC. The solvent was deproteinized with chloroform-n-octanol three more times. The cold- ethanol-precipitated DNA was again dissolved in the same buffer (10 ml) and was then dialyzed against two changes of 1 liter of the same buffer. DNA concentration was determined by absorption at 260 and 280 nm. Chloramine treatment. HOCIl was diluted in phosphate buffer to different concentrations. NH,Cl was formed by incubating eight parts of HOC] with one part of 0.1 M NH,Cl (in vivo treatment) or 1M NH,Cl (in vitro treatment) at 37 C for 1 h: HOC] + NH,Cl — NH,.Cl + H,O + HCl (1). The prepared cells or DNA was diluted 10 times into the NH.Cl solution to a final concentration of 10% to 4 x 10* cells/ml] or 11 to 22 xg of DNA/ml and treated for 30 min at 37 C. The reaction was stopped by adding 1 volume of 0.02 M (in vivo treatment) or 0.1 M (in vitro treatment) sodium thiosulfate (Na,S,O,). In each experiment, the highest dose of NH,Cl was added to the corresponding amount of thiosulfate before the addition of the cells or DNA and was used as control. Viable cell counts were scored on plates of nutri- ent agar (Difco). Transformation. Transforming DNA was pre- pared from the chloramine-treated and control cells as described, but without the 70 C treatment and without further treatments after the DNA was first dissolved in 0.1x SSC. The method of preparing the competent cells (SB202 aroB hisB trpC tyrA) and the EFFECTS OF CHLORAMINE 935 procedure for transformation were described by Stewart (15). The media used to select the aroBt, AisB*, trpC*, and tyrA* transformants contain all amino acids (25 ug/ml) except phenylalanine, histi- dine, tryptophan, and tyrosine, respectively. The media selected for tyrosine transformants were sup- plemented with 25 ug of shikimic acid per ml. The viable cell counts and the colony counts in transformation were obtained by averaging colony counts of two plates. The minimum number of colo- nies per plate counted was 107 and the maximum standard deviation in colony counts was +12%. Sucrose gradients. Cell lysate (0.15 to 0.30 ug of DNA) or purified DNA (0.15 to 1.1 wg) was layered on top of a 5 to 20% linear sucrose gradient (neutral, 1M NaCl, 0.01 M Tris-hydrochloride, pH 8; alka- line, 0.8 M NaCl-0.2 M NaOH, pH 11.8). (°H]P22 DNA (a gift from V. Sgaramella) was used as stand- ard either in the same gradient with the sample or in a parallel gradient in the same centrifuge run. The gradients were centrifuged at 20 C for 1.5 to 4h at 35,000 to 40,000 rpm in the IEC model SB-405 rotor. Fractions were collected dropwise from holes drilled in the bottom of the tubes directly onto Whatman GF/C filter disks. The disks were dried and the *H or C activities were counted in a liquid scintillation counter. The molecular weights of DNA were calculated from the formula (5): D,/D, = (M,/M,)*. D, and D, are the distances sedimented by sample and stand- ard, and M, and M, are the respective molecular weights. The distances were based on the mean position of the material in the gradient and there- fore express average molecular weights only. Values of 27.0 x 10° and 13.5 x 108 for whole molecules of P22 DNA at neutral and alkaline gradients, respec- tively, were used as standard molecular weights. The a@ value is 0.35 for DNA in neutral gradients, 0.40 for DNA in alkaline gradients, and 0.549 for the denatured DNA in neutral gradients (18). DNA uptake by competent cells. A 0.9-ml amount of SB202 competent cell suspension was incubated with 0.1 ml of the []HJDNA solutions to be tested at the DNA concentration of 0.45 ug/ml. At the conclu- sion of a 30-min incubation period, 0.05 ml of a 1-mg/ ml of solution of deoxyribonuclease (DNase) (in 1 mg/ ml bovine serum albumin-0.1 M Tris, pH 7.0) was added to the mixture of cells and the suspension was incubated at 37 C for 30 min. As a control, separate portions of the DNA solutions were pretreated for 30 min at 37 C with 0.1 ml of 100 zg of DNase per ml (in 100 ug/ml of bovine serum albumin-10-' M MgCl.- 0.1 M Tris, pH 7.0) and then 0.9 ml of cell suspension was added. The difference between the ?H counts fixed when DNase treatment of the DNA followed or preceded incubation of the cell suspension with DNA was taken as representing the DNA taken up. After incubation the cells were either filtered through membrane filters (Millipore Corp.), washed with Spizizen solution and counted, or used for the determination of transformants. RESULTS Effects of chloramine treatment in vivo on DNA-transforming activity and DNA molecu- 936 LU SHIH AND LEDERBERG lar weight. The survival of B. subtilis popula- tions treated with increasing concentrations of chloramine and the transforming activity of DNA extracted from these cells are shown in Fig. 1. (To simplify the discussion, DNA ex- tracted from treated bacteria will be called V- DNA; DNA treated in vitro will be called T-DNA. “Chlorinated” means treated with chloramine. We do not have new data here on the introduction of chlorine atoms into DNA.) V-DNA does show a reduction in transforming activity for the markers tested, tyrA‘ and 10° poiiinl L 10-2 pol 10-3 vil 10-4} i 1 L fi 0 20 40 60 80 CELL SURVIVING FRACTION AND DNA TRANSFORMING ACTIVITY NHyCI DOSE (uM) Fic. 1. Cell surviving fraction and DNA-trans- forming activity of B. subtilis after chloramine treat- ment. Cells of strain 168 (trpC hisB* tyrA*) were treated with different doses of NHCl. After treat- ment, an equal amount of SB156 cells (trpC* hisB tyrA) was added into each sample. DNA was ex- tracted from the mixed cells and served as donor to the recipient cells, SB202 (aroB trpC hisB tyrA). The trpC* transformants were used as internal controls. The ratios of hisB* or tyrA* transformants to trpC* transformants were the indices to the DNA activity of the 168 cells. Concentration of DNA was 0.2 to 0.25 pgiml during transformation. Symbols: (A) surviv- ing fraction; (O) hisB*-transforming activity; (@) tyrA*-transforming activity. J. BACTERIOL. hisB*. Further studies showed some impair- ment of uptake of V-DNA, compared to control, but not enough to account for the total reduc- tion in genetic activity (Table 1). It seems that although part of the V-DNA can be taken up by the competent cells, the lesions induced pro- hibit the DNA from participating further. It was also found, from the same experiment, that the cotransfer of the linkage group (aroB~ hisB*+ tyrA*) is reduced in V-DNA. This indi- cates that reduction in biological activity of DNA was accompanied by the introduction of DNA strand breaks. Physical evidence that chloramine treatment of cells leads to single-strand breaks in DNA is obtained by zone sedimentation. Figure 2 shows that alkaline-denatured V-DNA is de- graded from 65 to 44 or 14 mega D (mega D, 10° average molecular weight unless otherwise stated). In neutral gradients the shift is only from 140 to 137 or 105 mega D. These data point out that many single-strand breaks introduced in DNA by chloramine treatment are exposed by denaturation with alkali; further double- strand scissions occur at higher chloramine doses. In this experiment, 56 and 112 uM chloramine treatments give nonadditive results by producing 0.5 and 3.6 breaks per 65 mega D single-strand DNA, respectively. [To simplify the calculation, it is assumed that DNA breaks evenly, i.e., hits = (average molecular weight of control/average molecular weight of treated) — 1]. In a separate experiment, by treating the cells from the same culture with 28, 56, and 84 uM chloramine, 0, 0.6, and 2.0 breaks per 65 mega D single-strand DNA, respectively, were induced. If the two experiments are combined, the chloramine effect seems to show an initial lag, but may be more nearly additive in the range from 56 to 112 4M. It is presumed that some chloramine is taken up by other struc- tures in the cells before it can reach the DNA molecules in sufficient concentration to induce the breaks. The induction of DNA breaks is also time dependent. Figure 3 demonstrates that the longer the treatment, the slower the sedimen- tation of DNA. The number of DNA breaks per single-strand 66 mega D molecule induced are 1.3, 2.0, and 3.1 for 10-, 20-, and 40-min treat- ment, respectively. Whether the addition of sodium thiosulfate also stops further degradation of DNA was in- vestigated. After 10 min of chloramine treat- ment at 37 C, sodium thiosulfate was added to the chlorinated bacterial samples. DNA of the chlorinated cells incubated further at 37 C has a molecular weight of 26 mega D compared with 29 mega D of sample kept in ice. Sodium thio- VoL. 125, 1976 EFFECTS OF CHLORAMINE 937 TaBLe 1. Effects of chloramine treatment in vivo on DNA-transforming activity and DNA uptake* Single marker transforming activity Linked markers co- (%) transforming activity (%) NH,Cl (uM) Cell survival (%) aroB* aroB* DNA uptake + . : hisBt hisB* aroB -hisBt tyrA tyrA"/ tyrA*/ tyrA* hisB~ 0 100 100 100 100 77.50 66.50 100 56 32.50 69.42 67.77 _ 66.67 69.50 61.50 91.01 112 0.00026 12.60 9.82 11.82 58.00 50.50 53.72 “ (HIDNA was extracted from SB566 cells (¢rpC thy) after NH.Cl treatment in vivo. SB202 competent cell suspension was used to test transforming activity and uptake of the [“HIDNA solutions. Cotransformation was performed by replicating 200 hisB* and 200 tyrA* transformants from each treatment to phenylalanine-, histidine-, and tyrosine-deficient media. 800 600 ; 400 | DNA BREAKS/65.2 mega D jeo—e a 28 84 112 NH2CI DOSE iM) 200 T 800+ 600 - 400+ 200+ cl4 activity (COUNTS PER 5 MINUTES) FRACTION NUMBER Fic. 2. Degradation of DNA in vivo with different chloramine doses. The ['*C]thymidine-labeled cells SB566 (trpC thy) were treated with NHC and lysed by lysozyme (1 mg/ml) in SSC at 37 C for 30 min at cell concentrations less than 4 x 10°/ml. The lysates were added to 0.05% SLS, incubated at 37 C until clear, and then layered on 5 to 20% sucrose gradients for zone centrifugation. Average molecular weights for 0-(—O-), 56- (@), 112- (A) uM NH,Cl treatments were 65, 44, and 14 mega D, respectively, for alkaline gradients and 140, 137, and 105 mega D, respectively, for neutral gradients. (top) Alkaline gradients, insert shows DNA breaks per single strand 65 mega D molecule as a function of NH,CI doses; (bottom) neutral gradients; {- -O- -), P22 DNA. sulfate stops most of the DNA degradation in Effect of chloramine treatment in vitro on chlorinated cells; however, small portion of DNA molecular weight. The reaction of chlora- degradation is present after neutralization mine with DNA (in vitro) was also studied. The (Fig. 3). sedimentation patterns of T-DNA resemble 938 LU SHIH AND LEDERBERG J. BACTERIOL. T | J } 1,000 ——7~—7—5 : 800} 3 5 600 E 0 10 20 30 40 2 6 TIME OF 112 uM = CHLORAMINE = 400 + TREATMENT (MIN) | w ec w = 200/- Qa Eg 2 = 1,000 ° oO — 800 > 7 2 600} - oO 200+ 4 E = 100+ . 4 OQ ©“ £ ° < FEF gS R00 > ¢ 0 4 } | I t | | _ 0 3 7 11 15 19 23 27 31 FRACTION NUMBER Fic. 7. Degradation of in vitro chlorinated DNA by endonucleases. SB747 DNA samples treated and untreated with 640 »M NH,Cl were sedimented in alkaline sucrose gradients after incubating with cell extract from strain 168 for 30 min at 37 C. Cell extract were prepared as described in Fig. 6. Value of 12 mega D for control DNA (— @—) was used as standard molecular weight. Symbols: (— O—} chlorinated DNA, 7.0 mega D, (- -@- -) control DNA + cell extract, 4.3 mega D; (- -O- -) chlorinated DNA + cell extract, 2.7 mega D. VoL. 125, 1976 lower temperatures (2), was then used in com- parison with NaOH. Control and T-DNA dena- tured either with alkali or formamide were neutralized and sedimented in neutral gra- dients. Figure 8 shows that both alkali and formamide treatments expose T-DNA degrada- tion. Whereas formamide-denatured and al- kali-denatured control DNA have insignifi- cantly different sedimentation rates (13 versus EFFECTS OF CHLORAMINE 941 is consistently faster sedimenting than that of alkali-denatured T-DNA (10.5 versus 8.5 mega D or 81% versus 71% of the respective controls). V-DNA (extracted from SB566 cells treated with 56 4M chloramine) gave similar results: the formamide-denatured V-DNA and alkali- denatured V-DNA reduced their respective molecular weights to 61 and 48% of their re- spective controls. DNA whether chlorinated in 12 mega D), formamide-denatured T-DNA vitro or in vivo shows a number of alkali-sensi- 600 -— T T T T T T (a) NaOH a en, TREATMENT 400 200 | 4 0 L “ ! l ! | ! 600 T T T TT T (b) HCONH, ee 400; TREATMENT. / . 4 200 0 1,000 T T T T H? ACTIVITY (COUNTS PER 5 MINUTES) 800 4 600 7 400 7 a 200 ye 0 L 0 3 7 11 15 19 23 2731 FRACTION NUMBER Fic. 8. Degradation of in vitro chlorinated DNA denatured by sodium hydroxide and formamide. SB747 DNA (55 pg/ml} treated with 3200 uM NHC was sedimented in neutral sucrose gradients after denatura- tion. (a) DNA was denatured at a concentration of 27.5 ug/ml in 0.1 M NaOH without added salt. After 5 min at room temperature, the solution was neutralized with one-tenth volume of 1.1 M HCL0O.2 M Tris. (b) DNA was denatured by incubating in 90% formamide at 37 C for 30 min at a concentration 2.75 pg/ml without added salt. (c) DNA was not denatured. All of the samples were then dialyzed twice against 0.05 M phosphate buffer, pH 7.0. Average molecular weight of alkali-denatured control DNA (12 mega D) was used as a marker. The average molecular weight for formamide-denatured control DNA was 13 mega D and the average molecular weights for alkali and formamide denatured-chlorinated DNA were 8.5 and 10.5 mega D, respectively. The native chlorinated DNA does not show a detectable change in average molecular weight compared to the native control DNA. Symbols: (©) control DNA; (®@) chlorinated DNA. 942 LU SHIH AND LEDERBERG tive prebreaks which are not seen in untreated DNA. Heat-sensitive sites induced in DNA by chloramine treatment in vitro. The thermal stability of the chlorinated DNA was also tested. Different DNA samples were incubated at 50 C for 5 h and then assayed for transform- ing activity (16). The results show that V-DNA does not become thermolabile for tyrA~-trans- forming activity (Table 2), whereas T-DNA appears to be thermolabile for single as well as linked markers (Fig. 9). In a separate experi- ment, further degradation of T-DNA after heat- ing was indeed observed (Fig. 10). The average molecular weight of T-DNA was reduced to 65% whereas the control DNA remained unchanged after heat treatment. Degradation of DNA of chloramine-treated cells by DNA extraction procedure. Could the extraction procedure degrade V-DNA further? After cells were treated with 56 4M chlora- mine, both unextracted (as described in Fig. 2) and extracted DNA samples were studied for single-strand breakage. Compared to their respective control, the average molecular weight of single-strand DNA was reduced to 68% without extraction (Fig. 2), whereas the average molecular weight after extraction was reduced to 51% (Fig. 11). Thus, chlorinated DNA is a labile heteroge- neous substance that is sensitized to cleavage by endonucleases and contains single-strand breaks, double-strand scissions, alkali-sensi- tive sites, thermolabile lesions (T-DNA), and possibly other defects not yet identified by the present study. DISCUSSION The reduction in transformation caused by the introduction of defects into B. subtilis DNA TABLE 2. Effects of heat treatment on DNA from chloramine-treated cells* 7At-transforming activity (%) ty NH,Cl (4M) Not heated Heated 0 100° 95.76 56 22.32 22.93 80 18.94 21.25 « DNA was extracted from SB566 cells after chlor- amine treatment in vivo. DNA samples were di- vided into two parts; one part was kept at 5 C and the other part was incubated in a 50 C water bath for 5 h, and then transformation was performed. DNA concentration was 0.42 ug/ml during transforma- tion. ® The absolute number was designated as 100%. The other results were compared to this. J. BACTERIOL. via chloramine treatment in vivo can be attrib- uted to the failure of the extracted DNA mole- cule to enter the competent cell as well as the impairment of the process subsequent to entry. By studying the transformation of pneumococ- cus DNA treated with DNase I, Lerman and Tolmach (9, 10) proposed that the decline in DNA uptake reflects decline in the average molecular weight of the DNA molecules. It is not clear, however, what the correlation be- tween the reduction of DNA uptake and the reduction of the average molecular weight of the DNA molecule is. In chloramine treatment, whereas the average molecular weight of DNA was reduced to 72%, its uptake by the compe- tent cells was only 54%. Other lesions besides DNA strand breaks are possibly involved in the impairment of DNA uptake by distortion or modification of the configuration of the DNA molecules. The difference between DNA uptake and transforming activity of chlorinated DNA also resembles that observed in B. subtilis- transforming DNA inactivated with other agents —nitric acid, dimethylsulfate, hydroxy]- amine, ultraviolet light, etc. (4). It is conceiv- able that chlorinated DNA, in addition to strand breaks, contains other lesions that do not interfere with DNA uptake but impair the integration or even the eventual expression of the integrated DNA in the competent cells (3). Heating for 5 h at 50 C further decreased the transformation activity of the T-DNA. The deg- radation of these heat-sensitive lesions was shown to be a major factor contributing to the further inactivation of the chlorinated DNA. These heat-sensitive lesions, however, allow the unheated DNA to retain its transforming activity. The seeming heat insensitivity of the DNA extracted from chloramine-treated cells, as compared to in vitro-treated DNA, could be explained in different ways. As the process of DNA extraction does induce more breaks in V- DNA, the thermolabile sites could be broken during extraction procedure. This speculation, however, is not yet proven. Other possibilities could be that the heat-sensitive lesions were never introduced in vivo, or that the lesions induced could be either immediately repaired or cleaved in vivo. DNA degradation continues only slowly in chlorinated cells after sodium thiosulfate is added. The direct reaction of chloramine with DNA is the most plausible explanation for the DNA breaks induced in vivo. However, we can- not exclude a dynamic equilibrium of endonu- cleolytic breaks and enzymatic repair, the lat- ter being the target of the chloramine. We have Vou. 125, 1976 EFFECTS OF CHLORAMINE 943 a > _ NX - r \ yA > L ‘ a = \ J oO L a * q \ | oO — Zz r . = & cw ‘ 7 Oo L ‘ . Le \ \ ” \ \ a \ ‘ t \ \ a ‘ \ - \ \ 10— F (a) ab (b) ‘ | | l l ] 0 640 1280 19200 640 1280 1920 NH5Cl (uM) Fic. 9. Effects of chloramine treatment and heat treatment in vitro on DNA-transforming activity. Trans- forming DNA of strain 168 (trpC) was treated with different doses of NH Cl at a concentration of 2.9 ugiml and then DNA samples were dialyzed twice against 0.05 M phosphate buffer after the addition of sodium thiosulfate. Heat treatment and transformation were performed as described in Table 2. DNA concentration was less than 0.145 pg/ml during transformation. Cotransfer efficiency of each treatment was determined by replicating 200 hisB* or tyrA* transformants onto phenylalanine-, histidine-, and tyrosine-deficient media. The single marker activity and the linked marker activity of unheated control DNA were used as the standard and were designated as 10°. Symbols: (a) (@) hisB*; (O) hisB* after heating; (W@) (aroB* hisB* tyrAt)/hisB*; (C) (aroB~ hisB* tyrA*}/hisB* after heating. (b) (@) tyrA*; (CO) tyrA* after heating; (MD (aroB* hisBt tyrA*)/ tyrA*; (CO) (aroB* hisB* tyrA+)ityrA* after heating. 2 ! T T T = 800 4 iw 2 6o0t 4 2 5 @ 400+ 4 7 5 200 |- 4 S < 95 31 ~* x FRACTION NUMBER Fig. 10. Degradation of in vitro chlorinated DNA by heat. [H]DNA purified from SB747 cells was treated with 640 pM NH.Cl at a concentration of 2.7 ugiml. The control and chlorinated DNA samples (1 ml) were dialyzed twice against 1 liter of 0.05 M phosphate buffer, pH 7.0, after the addition of sodium thiosulfate. After dialyzing, heat treatment was performed to a part of each sample as described in Table 2. Transformation and sedimentation in alkaline sucrose gradients were then performed. DNA concentration was less than 0.27 yg/ ml during transformation. Symbols: (—O—) control DNA, 12 mega D, 100% tyrA*-transforming activity, (--O--) heated contral DNA, 12 mega D, 99% tyrA*-transforming activity; (—@®—) chlorinated DNA, 5.0 mega D, 62% tyrA*-transforming activity; (- -@- -) heated chlorinated DNA, 4.1 mega D, 34% tyrA*- transforming activity. 944 LU SHIH AND LEDERBERG J. BACTERIOL. | | | | | P22 DNA Z 1,000 T = “ 800- ” _ s 2 600; 2 z 400+ 2 & 200+