JOURNAL
OF
Vol. 138, No. 3
BACTERIOLOGY, June 1979, p. 871-877
0021-9193/79/06-0871/07$02.00/0
Characterization of Azotobacter vinelandii Deoxyribonucleic Acid and Folded Chromosomest H. L. SADOFF,* B. SHIMEI,4 AND S. ELLIS§ Microbiology and Public Health, Michigan State University, East Lansing, Michigan 48824
Received for publication 9 April 1979
The properties of Azotobacter vinelandii deoxyribonucleic acid (DNA) and folded chromosomes were studied and compared to those of Escherichia coli as a standard. Based on melting temperature and buoyant density measurements, the guanosine + cytosine content of purifiedA. vinelandii DNA was 65%, whereas that of E. coli DNA was 50%. The results of renaturation studies showed that the unique DNA sequence lengths of the two organisms were similar with CotI12 values of 7.3 ± 0.4 mol.s/liter and 7.5 ± 0.3 mol.s/liter, respectively, for A. vinelandii and E. coli. Folded chromosomes of A. vinelandii sedimented in a centrifugal field at a rate identical to those derived from E. coli, 1,600 to 1,700S. Based on the DNA content per cell and the mass of a single genome, A. vinelandii contains at least 40 chromosomes per cell.
Azotobacter vinelandii is a large, aerobic, gram-negative, nitrogen-fixing bacterium. Under certain conditions the organism encysts, producing metabolically dormant cells which are analogous to endospores of the Bacillaceae in being more resistant than vegetative cells to deleterious physical and chemical agents (24). The cell-to-cyst transition can be induced in exponential-phase cultures of A. vinelandii by replacing the glucose or other carbohydrate in the growth medium with ,8-hydroxybutyrate or crotonate (9). Since the metabolic fate of these latter two compounds is reasonably understood, encystment is attractive as a model of cell differentiation with strong possibilities for studying the biochemical events in the induction process. In our studies of encystment in A. vinelandii, we have attempted over a 4-year period to isolate mutants blocked in fl-hydroxybutyrate metabolism or at specific morphological stages of development, or to isolate auxotrophs that could be useful in controlling protein or nucleic acid synthesis. To date, only one purine, one pyrimidine, and one sulfite reductase mutant have been isolated despite concerted efforts with a variety of mutagenic agents and procedures. On the other hand, nitrogenase-less (7, 15, 21, 22) and antibiotic-resistant mutants (15) have been relatively easy to isolate and transfer in a simple transformation system (2, 15, 16).
The mechanism(s) controlling the uniquely limited mutant range in A. vinelandii is unknown but could be related to the cells' high content of DNA. Cultures of this organism in midexponential growth contain 15 x 10-14 g of DNA per cell (19), but a reduction to 3.4 x 10-14 g of DNA per cell occurs during encystment. It is significant that even this minimal level of DNA per cell is equivalent to approximately 10 Escherichia coli genomes and could constitute considerable redundancy in the A. vinelandii genome, thus accounting for the difficulty in
obtaining mutants. There is little reason to expect that the genetic information contained in chromosomes of A.
vinelandii is much more complex than that of E. coli, since both organisms have approximately equal biochemical capabilities. That is, they are both capable of growing in a glucosemineral medium although A. vinelandii does fix atmospheric nitrogen. The greater DNA content per cell of A. vinelandii is in direct proportion to its approximate 10-fold-greater volume than E. coli. It is the nature of the array of DNA in A. vinelandii that could determine whether the difficulty in isolating mutants lies in mutagenesis of a large, multiply iterated nucleus or in the segregation after mutagenesis of one cell, all of whose many nuclei contain the same DNA modification (homokaryotic in the strictest sense). t Journal Article no. 8825 from the Michigan Agricultural We have therefore characterized the DNA of A. Experiment Station. : Present address: Franklin Pierce School of Law, Concord, vinelandii and in this report describe its unique sequence length and the size of its folded chroNH 03301. § Present address: Department of Microbiology, University mosome relative to E. coli. Our results suggest of Michigan, Ann Arbor, MI 48109. that A. vinelandii contains at least 40 chromo871
872
SADOFF, SHIMEI, AND ELLIS
somes per cell during active growth and that segregation of a homokaryon may be the most likely problem in mutant isolation. MATERIALS AND METHODS
J. BACTERIOL.
The enzyme-treated DNA was extracted with 1.5 volumes (30 ml) of freshly distilled, water-saturated phenol for 30 min at 4°C with shaking. The phenol and aqueous phases were separated by centrifugation, and the aqueous phase was extracted with phenol Strains and cultivation. A. vinelandii ATCC twice more as described above. The DNA was then 12837 was grown at 30°C in Burk nitrogen-free me- precipitated from the aqueous phase under ethanol, dium (27) containing 0.5% glucose (BBG). Cultures of dissolved in 0.lx SSC, reprecipitated, and finally dis50-ml volume were shaken in 500-ml Erlenmeyer solved in 4.5 ml of SSC buffer. flasks at 140 oscillations per min. For volumes up to 3 The final purification of the DNA was achieved by liters, an inoculum (10 volume percent) of exponen- density gradient centrifugation in CsCl. Water was tially growing cells was added to sterile medium in a added to the DNA solution to achieve a total weight New Brunswick fermentor. These cultures were aer- of 11.10 g. A 14.55-g sample of CsCl was added, resultated at 1 volume of air per min and stirred at 200 rpm. ing in a solution that was 56.71 weight percent CsCl; Sterile Dow-Corning Antifoam-B emulsion was added the density of the solution was 1.71 g/cm3. A 2.5-mi (0.1%, vol/vol) to control foaming. Cells were har- sample of the DNA-CsCl solution was dispensed into vested by centrifugation prior to the end of exponen- each of six polyallomer centrifuge tubes (0.5 by 2 tial growth, washed twice in 0.15 M NaCl-0.1 M EDTA inches [ca. 12.7 by 50.8 mm]), mineral oil was added to (pH 8.0; saline-EDTA), and then suspended in water fill the tubes, and these were spun at 36,000 rpm for 42 to a turbidity corresponding to an optical density at h at 150C to generate a density gradient. DNA was 620 nm (OD6w) of 10. Turbidities were measured in a easily recovered from the gradients by punching the Gilford spectrophotometer using 1-cm-light path cu- bottoms of the tubes and collecting drops of high vettes. E. coli K-12 was grown in nutrient broth viscosity. The collected fractions were dialyzed ex(Difco) with aeration at 37°C either in Erlenmeyer haustively against 0.012 M sodium phosphate buffer flasks or in a 3-liter fermentor as described above. (pH 7.0), which is equivalent in sodium ion content to Purification of DNA. Solutions and miscellaneous 0.1X SSC. equipment coming in contact with the DNA were Density of DNA. Guanine + cytosine (G+C) consterilized prior to their use in the purification. The tents of A. vinelandii and E. coli DNAs were initial cell lysis was a modification of the Marmur calculated from their densities obtained by isopycnic procedure (12). A 130-ml volume of concentrated cell sedimentation in CsCl gradients (20) in a Spinco suspension, OD620 = 10, was sedimented in a centri- SW50.1 rotor at 36,000 rpm for 42 h at 15°C. Samples fuge, and the cells were suspended in 32 ml of 0.01 M were prepared by mixing 4.85 g of optical grade CsCl Tris-0l M NaCl (pH 8.1) (buffer A) in a 500-ml with 0.5 ml of a solution of purified DNA (OD2D of Erlenmeyer flask. The following two solutions were 5.0) and sufficient water to bring the total weight of added: 8.1 ml of 0.12 M Tris-0.05 M EDTA-lysozyme the solution to 8.55 g (density of 1.71 g/cm3). The (0.4 mg/ml) at pH 8.1 and 41 ml of 1% Brij-58-0.4% resulting 5 ml of solution was dispensed equally into deoxycholate-2.0 M NaCl-0.01 M EDTA at pH 8.1. two polyallomer centrifuge tubes (0.5 by 2 inches, ca. The suspension was incubated at 37°C for 10 min to 12.7 by 50.8 mm), which were then filled with mineral effect cell lysis. Sufficient saline-EDTA was added to oil. Blank gradients were made by adding 4.85 g of bring the volume of the lysate to 130 ml, after which CsCl to 3.70 g of water, dispensing equally into two 13 ml of freihly prepared 6 M sodium perchlorate was polyallomer centrifuge tubes (0.5 by 2 inches, ca. 12.7 by 50.8 mm), and then filling them with mineral oil. added. The lysate was then shaken at 25°C for 15 mi with 150 ml of 3% isoamyl alcohol in chloroform. The Upon the completion of each run, the tubes were aqueous phase was separated by centrifugation, de- punched, and fractions were collected in either 2-drop canted, and extracted twice more at the same condi- amounts from DNA-containing gradients or 5-drop amounts from blank gradients. The fractions from the tions with the same solvent. DNA was precipitated by overlaying the aqueous tubes containing DNA were diluted by the addition of phase with 2 volumes of 95% ethanol at 4°C and 0.5 ml of water, their absorbances were read at 260 winding the DNA onto glass rods. The crude DNA nm, and DNA content was plotted versus gradient was dissolved in 55 ml of 10-fold-diluted standard volume. The refractive index of each CsCl gradient saline-citrate buffer, 0.1x SSC (SSC is 0.15 M NaCl- fraction was measured in a Bausch and Lomb precision 0.015 M sodium citrate). A 5.5-ml volume of 1Ox SSC refractometer, and the densities calculated from these was added to the DNA in solution, and it was once data were also plotted versus gradient volumes, thus again precipitated from ethanol. The DNA was then permitting DNA densities to be read directly. Thermal denaturation of DNA. G+C contents of suspended in 15 ml of 0.lx SSC and incubated for 2 h at 250C, after which it was reacted with 50.ug of sterile DNAs were calculated from their "melting temperapancreatic RNase (Worthington Biochemicals Corp.) tures" (Tm) (13). Purified DNAs of A. vinelandii and per ml for 1 h at 37°C. The RNase had been dissolved E. coli were dialyzed against sodium phosphate buffers in sterile water, boiled for 5 min, and incubated at (pH 7) ranging from 0.012 to 0.024 M and de-aerated 370C for 1 h prior to its addition to the DNA solution. under vacuum. These solutions, whose OD2w0 ranged The DNA was then reacted for 4 h with 100 1&g of from 0.40 to 0.70, were loaded into the cell of a Gilford Pronase (Sigma Chemical Co.) per ml which had been Thermo-programmer, a component of a Gilford repreincubated for 2 h at 37°C. One-tenth volume of cording spectrophotometer, and their absorbance was lOx SSC was then added to the DNA-enzyme reaction read at 25°C (A25c). It was necessary to avoid using mixture, and it was stored at 0°C for 16 h. citrate-containing buffers because that reagent de-
VOL. 138, 1979
A. VINELANDII DNA AND CHROMOSOMES
stroyed the silver-soldered joints in the thermal cell. The DNA was heated at 1°C/min until its temperature reached 500C; thereafter the heating rate was 0.50C/ min until absorbance measurements (no further hyperchromic shift) indicated that thermal denaturation of the DNA was complete. Corrections were made for thermal expansion of DNA solutions, A., and Tm values (temperature at half denaturation) were obtained from plots of A.,/A25c versus temperature. G+C contents were calculated by G+C = 2.44 (Tm 81.5 - 16.6 log M) (11), where M was the molar concentration of cations present. Thermal renaturation of DNA. Renaturation kinetics were used to calculate the unique sequence length of A. vinelandii DNA relative to that of E. coli (26). To shear the DNA, 0.75 ml of purified DNA, OD260 of 5.0, was added to 4.25 ml of 2 M NaCl, and the solution was passed twice through a French pressure cell operated at 1,050 kg/cm2. This solution was then dialyzed against renaturation buffer, which was usually 0.12 M sodium phosphate (pH 7.0). Where appropriate to enhance the renaturation rate, higher buffer concentrations were used. The DNA solution was degassed under vacuum and loaded into a cell of the Gilford Thermo-programmer, and its absorbance was measured at 260 nm and 250C, A.. (native). The cell temperature was then rAised to 99.90C and held at this temperature to denature the DNA. Its absorbance was measured, Ao (denatured). Renaturation was conducted at Tm 250C, which was 720C forA. vinelandii DNA and 660C for E. coli DNA, and was monitored by absorbance measurements, A, which were continuously recorded. The term (Ao- A.)/(A - A.), the reciprocal of the fraction of denatured DNA remaining (26), was plotted as a function of time for the secondorder reaction. The rate constant k was determined by dividing the slope of the resulting straight line by the initial molar concentrations of nucleotides in the DNA. Renaturation rates are affected by the concentration of cations present during the process. In these studies the concentration of phosphate buffer varied with various DNA preparations. Therefore, all rate constants were corrected, using established concentrationrate relationships, to values that would have been obtained in 0.12 M phosphate buffer (3). Cot1/2, the parameter related to unique sequence length, was the reciprocal of k. Labeling of DNA. A. vinelandii will not incorporate exogenous thymidine or thymine with or without added deoxyadenosine. The following procedure, which had been demonstrated to be effective in introducing radioactive label into both RNA and DNA (10), was therefore followed to label DNA of A. vinelandii. The uracil auxotroph, strain 21U-, was grown aerobically at 300C in BBG medium containing 10 ug of uracil per ml. A 0.1-ml sample of exponentially growing cells was then used to inoculate 4.9 ml of BBG medium containing 10 ,uCi of [2-'4C]uracil (50,uCi/Mmol) plus 20Oug of cytosine per ml. E. coli was grown aerobically at 370C in 5 ml of M9 medium (1) supplemented with 10 MCi of tritiated thymidine ([methyl-3H]thymidine, 2 Ci/mmol). Preparation of folded chromosomes. Folded chromosomes of A. vinelandii and E. coli were prepared by adaptation of a method (23) that was a
873
modification of published procedures (25, 28). A 4.0ml sample of a radioactively labeled, exponentialy growing culture, ODemo of 0.5, plus 200 ,ug of chloramphenicol in 0.2 ml of water, were pipetted into a chiUed centrifuge tube, which was then spun at 12,000 x g at 5°C for 10 min. The resulting cell pellet was suspended in 0.2 ml of a buffer containing 0.01 M Tris (pH 8.15)0.01 M sodium azide-10% (wt/vol) sucrose-1.0 M NaCl plus 0.05 ml of a buffer containing 0.12 M Tris (pH 8.15)-0.05 M EDTA-0.4 mg of lysozyme per ml, and chiled in ice for 3 min. A 0.25-ml portion of a solution composed of 1% Sarkosyl NL-30-2.0 M NaCl-0.0125 M EDTA (solution C) was added, and the mixture was incubated at 250C for 10 min. A 0.2-ml volume of a solution containing 0.12 M Tris (pH 8.15)-0.05 M EDTA-20.0 mg of lysozyme per ml was then added, and the mixture was incubated 10 min at 250C. Final lysis was achieved by the addition of 2 ml of solution C at 250C, and the lysate was then chiUed in ice. A 0.1-ml sample of the lysate was placed on top of a 10 to 30% sucrose density gradient prepared in 0.01 M Tris (pH 8.15)-i M NaCl-0.001 M EDTA-0.001 M ,Bmercaptoethanol-0.5% Sarkosyl NL-30. A 0.05-ml portion of a suspension of 3H-labeled T4 bacteriophage, which served as a sedimentation marker, was placed on an identical gradient. These were spun at 16,500 rpm for 25 min in an SW50 rotor of a Spinco preparative ultracentrifuge. Fractions of 15 drops each were colected after punching the tube bottoms. For bacteriophage and materials derived from E. coli, 2 ml of cold 5% trichloroacetic acid was added to each fraction. These were incubated in the cold for 30 min, filtered, rinsed with 2.5 ml of water followed by 2.5 ml of 95% ethanol, dried, and then counted in a Packard 3000 scintiBation spectrometer. Fractions derived from A. vinelandii were initially treated as foUlows. KOH (2 N; 0.3 ml) was added to each sample of the gradient, and the mixture was incubated at 370C for 16 h to hydrolyze RNA. The DNA was then precipitated by the addition of 2.2 ml of cold 10% trichloroacetic acid, and after cooling on ice for 30 min the samples were filtered, washed, dried, and counted for radioactivity as above. It was noted that sedimentation patterns of folded chromosomes (discerned by the distribution of radioactivity in the gradient) were the same whether RNA had been fit hydrolyzed or the fractions had been precipitated directly and counted without the hydrolysis step. The significant difference observed was that the removal of RNA produced a 25 to 30% reduction in radioactivity of the peak corresponding to the folded chromosome. This is the extent of loss that would be effected if the folded bacterial genome of A. vinelandlii were similar to that of E. coli in containing 30% RNA (29). In consideration of the above, the direct precipitation of gradient fractions with 2 ml of 5% trichloroacetic acid was employed when studying the sedimentation of the A. vinelandii folded chromosome. The counts in the peak fractions represented the sum of their DNA and RNA contents.
RESULTS In the present study E. density. Buoyant coli DNA was used as a standard in evaluating
874
SADOFF, SHIMEI, AND ELLIS
J. BACTERIOL.
the properties of A. vinelandii DNA in its ag1.4 . gregated (folded chromosome) and purified state. The extended extraction process described in Materials and Methods, which was necessary to purify A. vinelandii DNA, was also successfully applied to the preparation of E. coli DNA. r These DNAs had average buoyant densities of 13 1.723 and 1.709 g/cm3, respectively, based on REL Ano EC AV three determinations for each DNA species. The Acorr G+C contents calculated from these data (20) A25C were 64.2% for A. vinelandii and 50% for E. coli. Thermal denaturation. The melting tem12 perature, Tm, of DNA is dependent on both its G+C content and the concentration of cations in which it is suspended. The higher the ionic m 1Tm concentration in the suspending buffer, within l77.2 184.3 limits, the higher the melting temperature (11). / For this reason, the denaturation characteristics 1.1 J of A. vinelandii DNA were always studied in buffers whose sodium ion concentration was less than 0.075 M (0.05 M sodium phosphate, pH 7.0) to keep the Tm below 900C and the upper limit of the melting temperature within the 99.9°C range of the heating apparatus. 1.0C 8 7 75 80 85 90 95 Typical denaturation data for both E. coli and A. vinelandii DNA are presented in Fig. 1. The TEMPERATURE (C) midpoint of each transition, Tm, is shown. HyFIG. 1. Melting curves for purified DNA from E. perchromic shifts of 1.32 (E. coli) and 1.33 (A. coli (; EC) and A. vinelandii (0; AV) conducted in vinelandii) were calculated from the initial and' 0.020 and 0.024 M sodium phosphate buffer (pH 7.0), final absorbances at 260 nm (corrected temper- re8pectively. The data are plotted in relative absorbature differences). The G+C molar composi ance units (corrected for expansion ofDNA solutions) temperature. (Ad ad final tions, calculated (11) from Tm values of 77.20C versus at solutionsbsorb DNA 260 mn ofInitial (correspective 4.3C, espetivly,wer 51%forE. oliances andand 84.30C, respectively, were 51% for E. coli rected for temperature difference) were: for EC DNA, DNA (in 0.02 M sodium phosphate, pH 7.0) and Ai 1.016, Af - 1.337; for AV DNA, Ai 1.020, Af65.2% for A. vinelandii DNA (in 0.024 M sodium 1.360. phosphate, pH 7.0). DNA renaturation. Renaturation kinetics were examined to establish the unique sequence renaturation. The COt1/2 of 7.4 mol- s/liter oblength or nuclear complexity of A. vinelandii tained for E. coli DNA in 0.12 M phosphate DNA relative to that of E. coli. Typical second- buffer (pH 7.0) corresponded well to published order plots of the progress of the renaturation values (4, 6) and thus established the validity of process in both A. vinelandii and E. coli DNAs the shearing, melting, and renaturation proceare presented in Fig. 2. These data were taken dures employed. The average value of k obtained at the time intervals indicated from continuous from five renaturation studies of E. coli DNA plots of decreasing OD20 which occurred during was 0.134 ± 0.005 liter/mol. s (corrected; 3) in renaturation. Initial concentrations of denatured phosphate buffers of 0.12 to 0.16 M correspondDNA were calculated by CO = Ao/7,200 for E. ing to an average Cotl,2 of 7.5 + 0.3 mol. s/liter. coli DNA (11) and C0 = Ao/7,080 for A. vinelanThe k value for the renaturation of A. vinedii (calculated on the basis of 65% G+C), where landii DNA in 0.18 M phosphate buffer (pH 7) 7,200 and 7,080 are the respective molar absorb- was 0.273 liter/mol.s which, upon correction for ance constants. The second-order velocity con- cation concentration equivalent to 0.12 M phosstant of 0.135 liter/mol -s (corrected to the cation phate buffer (3), was 0.137 liter/mol - s. The corconcentration in 0.12 M phosphate buffer; 3) is responding Cot1/2 was 7.3 mol. s/liter. The averthe reciprocal of the renaturation parameter for age value of k obtained from six renaturation E. coli usually quoted with regard to unique studies of A. vinelandii DNA was 0.138 ± 0.007 sequence length of DNA, Cotl/2. This latter value liter/mol. s (corrected; 3) in 0.12 to 0.18 M phosis the product of initial denatured DNA concen- phate buffers, corresponding to an average Cotl/2 tration and the timne required to achieve 50% of 7.3 ± 0.4 mol. s/liter. Renaturation data sug-
-
I
VOL. 138, 1979
A. VINELANDII DNA AND CHROMOSOMES
1A4
I
I 1
I
AV 1.3 AO-Aco
A -Am
EC
1.21.2 _,/0
/
1.1 r
1.1
_
0 is
I
20
3
I
Ug
511
1
MINUTES FIG. 2. Second-order renaturation plots forE. coli (0; EC) and A. vinelandii (0; AV) DNA. The fraction of denatured DNA remaining, (Ao - A.)/(A A.) (see text), was plotted versus time for E. coli DNA at 66°C and A. vinelandii DNA at 72°C.
875
chromosomes. Both sedimentation values are in the range (1,600-1,700S) reported for the average E. coli folded chromosome devoid of membrane (17). Membrane-bound chromosomes whose sedimentation rate was 3,200S (25) have been reported in E. coli, but efforts to make similar preparations in A. vinelandii by lysis of cells at C40 were unsuccessful. Because of the similarity in sedimentation rates of the folded chromosomes of the two species prepared by the same procedure, we assumed they were of similar size and tested them as follows. Suspensions of A. vinelandii and E. coli folded chromosomes were prepared, mixed, and then spun together in a sucrose density gradient. Figure 3 is a plot of radioactivity versus fraction number for one such experiment. Only one peak of radioactivity was obtained with lysates of either E. coli or A. vinelandii, and the individual peaks were recovered in the same gradient fraction. The data suggest very strongly that folded chromosomes of exponentially growing A. vinelandii and E. coli have very similar sedi:mentation velocities and are therefore of
similarsize.
255 AV(21U-) l 20
(2-.dC)-URACIL
4
gest that the DNAs of E. coli and A. vinelandii are very similar in their
unique sequence lengths.
t Because of its high content per cell, the DNA ofA. vinelandii could exist as many single folded x1 _33 chromosomes like those of E. coli, as fewer 0zI KJ multiply iterated structures of two or more unique sequence lengths of DNA, or as mixtures a. J of the two classes. Our approach to differentiat- 0 2* ing between these possibilities was to prepare a X C iEC folded chromosomes from A. vinelandii whose / (6_3H)-THYMADNE DNA and RNA had been labeled with 14C and to compare their sedimentation velocities (and 1 hence relative size) with preparations derived from E. coli whose DNA had been labeled with tritium. If the folded chromosome of A. vinelandii contained more than one unique sequence of . DNA, its sedimentation velocity would be 1N R iO greater than the 2,200S value observed for the ATION propER F S E. coli folded chromosome in its presumably FIG. 3. Sedimentation properties of folded chrofully replicated state (28). The average sedimentation rate (five deter- mosomes of E. coli (0; EC) and A. vinelandii (0; The E. coli DNA was labeled with 3H, and the minations) E. coli coli folded folded chromosomeswasAV). Of of E. chromosomes was minations) vinelandii RNA and DNA were labeled with "C. 1,600S upon comparison with that of the T4 A. Folded chromosomes were prepared, mixed, spun in bacteriophage marker (1,025S). Similar studies a sucrose density gradient, and sampled. Peak radioconducted with A. vinelandii preparations activity representing the folded chromosomes of both yielded an average value of 1,700S for its folded cell types was found in a single fraction.
876
SADOFF, SHIMEI, AND ELLIS
DISCUSSION The high DNA content in cells ofAzotobacter species is well established (14, 19, 30). Early efforts using nuclear staining suggested that during exponential growth these bacteria contained multi-lobate nuclei (5, 18). However, direct staining methods for nuclear analysis are fraught with artifacts and are difficult to interpret. Thus, little has been known of the array of nuclei in these cells. In our studies we purified both A. vinelandii and E. coli DNA and used the latter material as a standard for the procedures we have employed in analyzing the genomic complement of A. vinelandii. This was particularly important in renaturation studies, where the extent of shearing of DNA and cation concentration have significant effects on the results of each experiment. The cell lysis and DNA extraction process outlined in Materials and Methods was a consequence of our need for pure DNA and our inability to obtain it from A. vinelandii by established techniques (12). Impurities affect both melting and renaturation characteristics of DNA (3). The final step, isopycnic centrifugation in a CsCl gradient, was required to separate low levels of RNA and carbohydrate from the DNA. The close correspondence between G+C contents calculated from the Tm and buoyant density data suggests that A. vinelandii DNA contains very few or no unusual bases. The average G+C of 65% noted in these studies agrees with accepted values (8) for A. vinelandii and Azotobacter chroococcum, a closely related organism.
Both denaturation and renaturation kinetics of DNA were studied spectrophotometrically because of the ease of observation and because of the precision of temperature control achieved in the thermal programmer, which was an integral part of the spectrophotometer. Typical secondorder kinetics were obtained in renaturation reactions for the initial 60 min of the reactions (Fig. 2) (10 to 15% renaturation). The reaction deviates from second-order kinetics when observed over a prolonged time, but initial rates are adequate for determinations of Cot1/2 values (6, 26). Presumptive evidence for the successful preparation of E. coli folded chromosomes included cell lysis without a marked increase in the viscosity of the lysate and the occurrence of DNAcontaining particles sedimenting at 1,600S (25, 28). When treated with RNase, cell lysates became viscous, and no discrete concentration of DNA sedimenting at 1,600S could be detected. A single procedure was effective in the preparation of folded chromosomes of both E. coli and
J. BACTERIOL.
A. vinelandii (23). Their identical sedimentation properties (Fig. 3), as well as their very similar Cot1/2 values and, hence, unique sequence lengths, suggest that the chromosomes of these two organisms are of the same size. Assuming the mass of the E. coli chromosome to be 3.5 x lo-15 g and knowing that exponentially growing A. vinelandii contains 15 x 10-14 g of DNA per cell (19), we conclude that this large bacterium must contain at least 40 genomic equivalents of DNA per cell. This large number of chromosomes may result in difficulties in the segregation of homocaryotic cells following mutation in this organism. ACKNOWLEDGMENTS We thank Loren Snyder for the gift of 3H-labeled T4 bacteriophage. This study was supported by Public Health Service grant AI-01863 from the National Institute of Allergy and Infectious Diseases.
LITERATURE CITED 1. Anderson, E. H. 1946. Growth requirements of vimsresistant mutants of Escherichia coli strain "B". Proc. Natl. Acad. Sci. U.S.A. 32:120-128. 2. Bishop, P. E., and W. J. Brill. 1977. Genetic analysis of Azotobacter vielandii strains unable to fx nitrogen. J. Bacteriol. 130:954-956. 3. Britten, R. J., D. E. Graham, and B. R. Neufeld. 1974. Analysis of repeating DNA sequences by reasociation. Methods Enzymol. 29:363-418. 4. Britten, R. J., and D. E. Kohne. 1968. Repeated sequences in DNA. Science 161:529-540. 5. Dondero, N. C., and ML R. Zelle. 1963. Observations on the formation and behavior of conjugation cells and large bodies in Azotobacter agile. Science 118:34-36. 6. Gills, M., J. De Ley, and M. De Cleene. 1970. The determination of molecular weight of bacterial genome from renaturation rates. Eur. J. Biochem. 12:143-153. 7. Green, M., M. Alexander, and P. W. Wilson. 1953. Mutants of the Azotobacter unable to use N2. J. Bacteriol. ":623-624. 8. Johnstone, D. B. 1974. Genus I. Azotobacter Beijerinck 1901, 567, p. 254-255. In R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology. The Williams & Wilkins Co., Baltimore. 9. Lin, L P., and H. L Sadoff. 1968. Encystment and polymer production by Azotobacter vinelandii in the presence of 8-hydroxybutyrate. J. Bacteriol. 95:23362343. 10. Loperfido, B., and H. L Sadoff. 1973. Gennination of Azotobacter vinelandii cysts: sequence of macromolecular synthesis and nitrogen fixation. J. Bacteriol. 113: 841-846. 11. Mandel, M., and J. Marmur. 1968. Use of ultraviolet absorbance-temperature profile for detennining the guanine plus cytosine content of DNA. Methods Enzymol. 12:195-206. 12. Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from microorganism. J. Mol. Biol. 3: 208-218. 13. Marmur, J., and P. Doty. 1962. Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. J. Mol. Biol. 5:109-118. 14. Miiler, H. P, and H. Kern. 1967. Strahlenresisteng, Gehalt, und Base nsetzung der DNA einiger strahlendurzierter Mutaten von Azotobacter chroococcum. Z. Naturforwch. 22:1330-1336. 15. Page, W. J., and H. L. Sadoff. 1976. Physiological factors
VOL. 138, 1979
16.
17.
18. 19. 20.
21.
22.
A. VINELANDII DNA AND CHROMOSOMES
affecting transformation of Azotobacter vinelandii. J. Bacteriol. 125:1080-1087. Page, W. J., and H. L. Sadoff. 1976. Control of transformation competence in Azotobacter vinelandii by nitrogen catabolite derepression. J. Bacteriol. 125:10881095. Pettijohn, D. E., and R. Hecht. 1973. RNA molecules bound to the folded bacterial genome stabilize DNA folds and segregate domains of supercoiling. Cold Spring Harbor Symp. Quant. Biol. 38:31-41. Pochon, J., Y. T. Tchan, and T. L Wang. 1948. Recherches sur le cycle morphologique et l'appareil nucleaire des Azotobacter. Ann. Inst. Pasteur 74:182-188. Sadoff, H. L, E. Berke, and B. Loperfido. 1971. Physiological studies of encystment in Azotobacter vinelandii. J. Bacteriol. 105:185-189. Schildkraut, C. L, J. Marmur, and P. Doty. 1962. Determination of the base composition of deoxyribonucleic acid from its buoyant density in CsCl. J. Mol. Biol. 4:430-443. Shah, V. K., L C. Davis, J. K. Gordon, W. H. OrmeJohnson, and W. J. Brill. 1973. Nitrogenase. III. Nitrogenaseless mutants of Azotobacter vinelandu: activities, cross-reactions and EPR spectra. Biochim. Biophys. Acta 292:246-255. Shah, V. K., L C. Davis, M. Stieghorst, and W. J. Brifl. 1974. Mutant of Azotobacter vinelandii that
23.
24. 25.
26. 27. 28. 29. 30.
877
hyperproduces nitrogenase component II. J. Bacteriol. 117:917-919. Sirotkin, K., J. Wei, and L Snyder. 1977. T-4 bacteriophage-coded RNA polymerase subunit blocks host transcription and unfolds the host chromosome. Nature (London) 265:28-31. Socolofsky, M. D., and 0. Wyss. 1962. Resistance of the Azotobacter cyst. J. Bacteriol. 84:119-124. Stonigton, 0. G., and D. E. Pettijohn. 1971. The folded genome of Escherichia coli isolated in a proteinDNA-RNA complex. Proc. Natl. Acad. Sci. U.S.A. 68: 6-9. Wetmur, J. G., and N. Davidson. 1968. Kinetics of renaturation of DNA. J. Mol. Biol. 31:349-370. Wilson, P. W., and S. G. Knight. 1952. Experiments in bacterial physiology. Burgess Publishing Co., Minneapolis. Worcel, A., and E. Burgi. 1972. On the structure of the folded chromosome of Escherichia coli. J. Mol. Biol. 71:127-147. Worcel, A., E. Burgi, J. Robinton, and C. L. Carlson. 1973. Studies on the folded chromosome of Escherichia coli. Cold Spring Harbor Symp. Quant. Biol. 38:43-51. Zaitseva, G. N., I. A. Khmel, and A. N. Belozerskii. 1961. Biochemical changes in synchronous cultures of Azotobacter vinelandii. Dokl. Akad. Nauk. S.S.R. 141: 740-743.
OF
Vol. 138, No. 3
BACTERIOLOGY, June 1979, p. 871-877
0021-9193/79/06-0871/07$02.00/0
Characterization of Azotobacter vinelandii Deoxyribonucleic Acid and Folded Chromosomest H. L. SADOFF,* B. SHIMEI,4 AND S. ELLIS§ Microbiology and Public Health, Michigan State University, East Lansing, Michigan 48824
Received for publication 9 April 1979
The properties of Azotobacter vinelandii deoxyribonucleic acid (DNA) and folded chromosomes were studied and compared to those of Escherichia coli as a standard. Based on melting temperature and buoyant density measurements, the guanosine + cytosine content of purifiedA. vinelandii DNA was 65%, whereas that of E. coli DNA was 50%. The results of renaturation studies showed that the unique DNA sequence lengths of the two organisms were similar with CotI12 values of 7.3 ± 0.4 mol.s/liter and 7.5 ± 0.3 mol.s/liter, respectively, for A. vinelandii and E. coli. Folded chromosomes of A. vinelandii sedimented in a centrifugal field at a rate identical to those derived from E. coli, 1,600 to 1,700S. Based on the DNA content per cell and the mass of a single genome, A. vinelandii contains at least 40 chromosomes per cell.
Azotobacter vinelandii is a large, aerobic, gram-negative, nitrogen-fixing bacterium. Under certain conditions the organism encysts, producing metabolically dormant cells which are analogous to endospores of the Bacillaceae in being more resistant than vegetative cells to deleterious physical and chemical agents (24). The cell-to-cyst transition can be induced in exponential-phase cultures of A. vinelandii by replacing the glucose or other carbohydrate in the growth medium with ,8-hydroxybutyrate or crotonate (9). Since the metabolic fate of these latter two compounds is reasonably understood, encystment is attractive as a model of cell differentiation with strong possibilities for studying the biochemical events in the induction process. In our studies of encystment in A. vinelandii, we have attempted over a 4-year period to isolate mutants blocked in fl-hydroxybutyrate metabolism or at specific morphological stages of development, or to isolate auxotrophs that could be useful in controlling protein or nucleic acid synthesis. To date, only one purine, one pyrimidine, and one sulfite reductase mutant have been isolated despite concerted efforts with a variety of mutagenic agents and procedures. On the other hand, nitrogenase-less (7, 15, 21, 22) and antibiotic-resistant mutants (15) have been relatively easy to isolate and transfer in a simple transformation system (2, 15, 16).
The mechanism(s) controlling the uniquely limited mutant range in A. vinelandii is unknown but could be related to the cells' high content of DNA. Cultures of this organism in midexponential growth contain 15 x 10-14 g of DNA per cell (19), but a reduction to 3.4 x 10-14 g of DNA per cell occurs during encystment. It is significant that even this minimal level of DNA per cell is equivalent to approximately 10 Escherichia coli genomes and could constitute considerable redundancy in the A. vinelandii genome, thus accounting for the difficulty in
obtaining mutants. There is little reason to expect that the genetic information contained in chromosomes of A.
vinelandii is much more complex than that of E. coli, since both organisms have approximately equal biochemical capabilities. That is, they are both capable of growing in a glucosemineral medium although A. vinelandii does fix atmospheric nitrogen. The greater DNA content per cell of A. vinelandii is in direct proportion to its approximate 10-fold-greater volume than E. coli. It is the nature of the array of DNA in A. vinelandii that could determine whether the difficulty in isolating mutants lies in mutagenesis of a large, multiply iterated nucleus or in the segregation after mutagenesis of one cell, all of whose many nuclei contain the same DNA modification (homokaryotic in the strictest sense). t Journal Article no. 8825 from the Michigan Agricultural We have therefore characterized the DNA of A. Experiment Station. : Present address: Franklin Pierce School of Law, Concord, vinelandii and in this report describe its unique sequence length and the size of its folded chroNH 03301. § Present address: Department of Microbiology, University mosome relative to E. coli. Our results suggest of Michigan, Ann Arbor, MI 48109. that A. vinelandii contains at least 40 chromo871
872
SADOFF, SHIMEI, AND ELLIS
somes per cell during active growth and that segregation of a homokaryon may be the most likely problem in mutant isolation. MATERIALS AND METHODS
J. BACTERIOL.
The enzyme-treated DNA was extracted with 1.5 volumes (30 ml) of freshly distilled, water-saturated phenol for 30 min at 4°C with shaking. The phenol and aqueous phases were separated by centrifugation, and the aqueous phase was extracted with phenol Strains and cultivation. A. vinelandii ATCC twice more as described above. The DNA was then 12837 was grown at 30°C in Burk nitrogen-free me- precipitated from the aqueous phase under ethanol, dium (27) containing 0.5% glucose (BBG). Cultures of dissolved in 0.lx SSC, reprecipitated, and finally dis50-ml volume were shaken in 500-ml Erlenmeyer solved in 4.5 ml of SSC buffer. flasks at 140 oscillations per min. For volumes up to 3 The final purification of the DNA was achieved by liters, an inoculum (10 volume percent) of exponen- density gradient centrifugation in CsCl. Water was tially growing cells was added to sterile medium in a added to the DNA solution to achieve a total weight New Brunswick fermentor. These cultures were aer- of 11.10 g. A 14.55-g sample of CsCl was added, resultated at 1 volume of air per min and stirred at 200 rpm. ing in a solution that was 56.71 weight percent CsCl; Sterile Dow-Corning Antifoam-B emulsion was added the density of the solution was 1.71 g/cm3. A 2.5-mi (0.1%, vol/vol) to control foaming. Cells were har- sample of the DNA-CsCl solution was dispensed into vested by centrifugation prior to the end of exponen- each of six polyallomer centrifuge tubes (0.5 by 2 tial growth, washed twice in 0.15 M NaCl-0.1 M EDTA inches [ca. 12.7 by 50.8 mm]), mineral oil was added to (pH 8.0; saline-EDTA), and then suspended in water fill the tubes, and these were spun at 36,000 rpm for 42 to a turbidity corresponding to an optical density at h at 150C to generate a density gradient. DNA was 620 nm (OD6w) of 10. Turbidities were measured in a easily recovered from the gradients by punching the Gilford spectrophotometer using 1-cm-light path cu- bottoms of the tubes and collecting drops of high vettes. E. coli K-12 was grown in nutrient broth viscosity. The collected fractions were dialyzed ex(Difco) with aeration at 37°C either in Erlenmeyer haustively against 0.012 M sodium phosphate buffer flasks or in a 3-liter fermentor as described above. (pH 7.0), which is equivalent in sodium ion content to Purification of DNA. Solutions and miscellaneous 0.1X SSC. equipment coming in contact with the DNA were Density of DNA. Guanine + cytosine (G+C) consterilized prior to their use in the purification. The tents of A. vinelandii and E. coli DNAs were initial cell lysis was a modification of the Marmur calculated from their densities obtained by isopycnic procedure (12). A 130-ml volume of concentrated cell sedimentation in CsCl gradients (20) in a Spinco suspension, OD620 = 10, was sedimented in a centri- SW50.1 rotor at 36,000 rpm for 42 h at 15°C. Samples fuge, and the cells were suspended in 32 ml of 0.01 M were prepared by mixing 4.85 g of optical grade CsCl Tris-0l M NaCl (pH 8.1) (buffer A) in a 500-ml with 0.5 ml of a solution of purified DNA (OD2D of Erlenmeyer flask. The following two solutions were 5.0) and sufficient water to bring the total weight of added: 8.1 ml of 0.12 M Tris-0.05 M EDTA-lysozyme the solution to 8.55 g (density of 1.71 g/cm3). The (0.4 mg/ml) at pH 8.1 and 41 ml of 1% Brij-58-0.4% resulting 5 ml of solution was dispensed equally into deoxycholate-2.0 M NaCl-0.01 M EDTA at pH 8.1. two polyallomer centrifuge tubes (0.5 by 2 inches, ca. The suspension was incubated at 37°C for 10 min to 12.7 by 50.8 mm), which were then filled with mineral effect cell lysis. Sufficient saline-EDTA was added to oil. Blank gradients were made by adding 4.85 g of bring the volume of the lysate to 130 ml, after which CsCl to 3.70 g of water, dispensing equally into two 13 ml of freihly prepared 6 M sodium perchlorate was polyallomer centrifuge tubes (0.5 by 2 inches, ca. 12.7 by 50.8 mm), and then filling them with mineral oil. added. The lysate was then shaken at 25°C for 15 mi with 150 ml of 3% isoamyl alcohol in chloroform. The Upon the completion of each run, the tubes were aqueous phase was separated by centrifugation, de- punched, and fractions were collected in either 2-drop canted, and extracted twice more at the same condi- amounts from DNA-containing gradients or 5-drop amounts from blank gradients. The fractions from the tions with the same solvent. DNA was precipitated by overlaying the aqueous tubes containing DNA were diluted by the addition of phase with 2 volumes of 95% ethanol at 4°C and 0.5 ml of water, their absorbances were read at 260 winding the DNA onto glass rods. The crude DNA nm, and DNA content was plotted versus gradient was dissolved in 55 ml of 10-fold-diluted standard volume. The refractive index of each CsCl gradient saline-citrate buffer, 0.1x SSC (SSC is 0.15 M NaCl- fraction was measured in a Bausch and Lomb precision 0.015 M sodium citrate). A 5.5-ml volume of 1Ox SSC refractometer, and the densities calculated from these was added to the DNA in solution, and it was once data were also plotted versus gradient volumes, thus again precipitated from ethanol. The DNA was then permitting DNA densities to be read directly. Thermal denaturation of DNA. G+C contents of suspended in 15 ml of 0.lx SSC and incubated for 2 h at 250C, after which it was reacted with 50.ug of sterile DNAs were calculated from their "melting temperapancreatic RNase (Worthington Biochemicals Corp.) tures" (Tm) (13). Purified DNAs of A. vinelandii and per ml for 1 h at 37°C. The RNase had been dissolved E. coli were dialyzed against sodium phosphate buffers in sterile water, boiled for 5 min, and incubated at (pH 7) ranging from 0.012 to 0.024 M and de-aerated 370C for 1 h prior to its addition to the DNA solution. under vacuum. These solutions, whose OD2w0 ranged The DNA was then reacted for 4 h with 100 1&g of from 0.40 to 0.70, were loaded into the cell of a Gilford Pronase (Sigma Chemical Co.) per ml which had been Thermo-programmer, a component of a Gilford repreincubated for 2 h at 37°C. One-tenth volume of cording spectrophotometer, and their absorbance was lOx SSC was then added to the DNA-enzyme reaction read at 25°C (A25c). It was necessary to avoid using mixture, and it was stored at 0°C for 16 h. citrate-containing buffers because that reagent de-
VOL. 138, 1979
A. VINELANDII DNA AND CHROMOSOMES
stroyed the silver-soldered joints in the thermal cell. The DNA was heated at 1°C/min until its temperature reached 500C; thereafter the heating rate was 0.50C/ min until absorbance measurements (no further hyperchromic shift) indicated that thermal denaturation of the DNA was complete. Corrections were made for thermal expansion of DNA solutions, A., and Tm values (temperature at half denaturation) were obtained from plots of A.,/A25c versus temperature. G+C contents were calculated by G+C = 2.44 (Tm 81.5 - 16.6 log M) (11), where M was the molar concentration of cations present. Thermal renaturation of DNA. Renaturation kinetics were used to calculate the unique sequence length of A. vinelandii DNA relative to that of E. coli (26). To shear the DNA, 0.75 ml of purified DNA, OD260 of 5.0, was added to 4.25 ml of 2 M NaCl, and the solution was passed twice through a French pressure cell operated at 1,050 kg/cm2. This solution was then dialyzed against renaturation buffer, which was usually 0.12 M sodium phosphate (pH 7.0). Where appropriate to enhance the renaturation rate, higher buffer concentrations were used. The DNA solution was degassed under vacuum and loaded into a cell of the Gilford Thermo-programmer, and its absorbance was measured at 260 nm and 250C, A.. (native). The cell temperature was then rAised to 99.90C and held at this temperature to denature the DNA. Its absorbance was measured, Ao (denatured). Renaturation was conducted at Tm 250C, which was 720C forA. vinelandii DNA and 660C for E. coli DNA, and was monitored by absorbance measurements, A, which were continuously recorded. The term (Ao- A.)/(A - A.), the reciprocal of the fraction of denatured DNA remaining (26), was plotted as a function of time for the secondorder reaction. The rate constant k was determined by dividing the slope of the resulting straight line by the initial molar concentrations of nucleotides in the DNA. Renaturation rates are affected by the concentration of cations present during the process. In these studies the concentration of phosphate buffer varied with various DNA preparations. Therefore, all rate constants were corrected, using established concentrationrate relationships, to values that would have been obtained in 0.12 M phosphate buffer (3). Cot1/2, the parameter related to unique sequence length, was the reciprocal of k. Labeling of DNA. A. vinelandii will not incorporate exogenous thymidine or thymine with or without added deoxyadenosine. The following procedure, which had been demonstrated to be effective in introducing radioactive label into both RNA and DNA (10), was therefore followed to label DNA of A. vinelandii. The uracil auxotroph, strain 21U-, was grown aerobically at 300C in BBG medium containing 10 ug of uracil per ml. A 0.1-ml sample of exponentially growing cells was then used to inoculate 4.9 ml of BBG medium containing 10 ,uCi of [2-'4C]uracil (50,uCi/Mmol) plus 20Oug of cytosine per ml. E. coli was grown aerobically at 370C in 5 ml of M9 medium (1) supplemented with 10 MCi of tritiated thymidine ([methyl-3H]thymidine, 2 Ci/mmol). Preparation of folded chromosomes. Folded chromosomes of A. vinelandii and E. coli were prepared by adaptation of a method (23) that was a
873
modification of published procedures (25, 28). A 4.0ml sample of a radioactively labeled, exponentialy growing culture, ODemo of 0.5, plus 200 ,ug of chloramphenicol in 0.2 ml of water, were pipetted into a chiUed centrifuge tube, which was then spun at 12,000 x g at 5°C for 10 min. The resulting cell pellet was suspended in 0.2 ml of a buffer containing 0.01 M Tris (pH 8.15)0.01 M sodium azide-10% (wt/vol) sucrose-1.0 M NaCl plus 0.05 ml of a buffer containing 0.12 M Tris (pH 8.15)-0.05 M EDTA-0.4 mg of lysozyme per ml, and chiled in ice for 3 min. A 0.25-ml portion of a solution composed of 1% Sarkosyl NL-30-2.0 M NaCl-0.0125 M EDTA (solution C) was added, and the mixture was incubated at 250C for 10 min. A 0.2-ml volume of a solution containing 0.12 M Tris (pH 8.15)-0.05 M EDTA-20.0 mg of lysozyme per ml was then added, and the mixture was incubated 10 min at 250C. Final lysis was achieved by the addition of 2 ml of solution C at 250C, and the lysate was then chiUed in ice. A 0.1-ml sample of the lysate was placed on top of a 10 to 30% sucrose density gradient prepared in 0.01 M Tris (pH 8.15)-i M NaCl-0.001 M EDTA-0.001 M ,Bmercaptoethanol-0.5% Sarkosyl NL-30. A 0.05-ml portion of a suspension of 3H-labeled T4 bacteriophage, which served as a sedimentation marker, was placed on an identical gradient. These were spun at 16,500 rpm for 25 min in an SW50 rotor of a Spinco preparative ultracentrifuge. Fractions of 15 drops each were colected after punching the tube bottoms. For bacteriophage and materials derived from E. coli, 2 ml of cold 5% trichloroacetic acid was added to each fraction. These were incubated in the cold for 30 min, filtered, rinsed with 2.5 ml of water followed by 2.5 ml of 95% ethanol, dried, and then counted in a Packard 3000 scintiBation spectrometer. Fractions derived from A. vinelandii were initially treated as foUlows. KOH (2 N; 0.3 ml) was added to each sample of the gradient, and the mixture was incubated at 370C for 16 h to hydrolyze RNA. The DNA was then precipitated by the addition of 2.2 ml of cold 10% trichloroacetic acid, and after cooling on ice for 30 min the samples were filtered, washed, dried, and counted for radioactivity as above. It was noted that sedimentation patterns of folded chromosomes (discerned by the distribution of radioactivity in the gradient) were the same whether RNA had been fit hydrolyzed or the fractions had been precipitated directly and counted without the hydrolysis step. The significant difference observed was that the removal of RNA produced a 25 to 30% reduction in radioactivity of the peak corresponding to the folded chromosome. This is the extent of loss that would be effected if the folded bacterial genome of A. vinelandlii were similar to that of E. coli in containing 30% RNA (29). In consideration of the above, the direct precipitation of gradient fractions with 2 ml of 5% trichloroacetic acid was employed when studying the sedimentation of the A. vinelandii folded chromosome. The counts in the peak fractions represented the sum of their DNA and RNA contents.
RESULTS In the present study E. density. Buoyant coli DNA was used as a standard in evaluating
874
SADOFF, SHIMEI, AND ELLIS
J. BACTERIOL.
the properties of A. vinelandii DNA in its ag1.4 . gregated (folded chromosome) and purified state. The extended extraction process described in Materials and Methods, which was necessary to purify A. vinelandii DNA, was also successfully applied to the preparation of E. coli DNA. r These DNAs had average buoyant densities of 13 1.723 and 1.709 g/cm3, respectively, based on REL Ano EC AV three determinations for each DNA species. The Acorr G+C contents calculated from these data (20) A25C were 64.2% for A. vinelandii and 50% for E. coli. Thermal denaturation. The melting tem12 perature, Tm, of DNA is dependent on both its G+C content and the concentration of cations in which it is suspended. The higher the ionic m 1Tm concentration in the suspending buffer, within l77.2 184.3 limits, the higher the melting temperature (11). / For this reason, the denaturation characteristics 1.1 J of A. vinelandii DNA were always studied in buffers whose sodium ion concentration was less than 0.075 M (0.05 M sodium phosphate, pH 7.0) to keep the Tm below 900C and the upper limit of the melting temperature within the 99.9°C range of the heating apparatus. 1.0C 8 7 75 80 85 90 95 Typical denaturation data for both E. coli and A. vinelandii DNA are presented in Fig. 1. The TEMPERATURE (C) midpoint of each transition, Tm, is shown. HyFIG. 1. Melting curves for purified DNA from E. perchromic shifts of 1.32 (E. coli) and 1.33 (A. coli (; EC) and A. vinelandii (0; AV) conducted in vinelandii) were calculated from the initial and' 0.020 and 0.024 M sodium phosphate buffer (pH 7.0), final absorbances at 260 nm (corrected temper- re8pectively. The data are plotted in relative absorbature differences). The G+C molar composi ance units (corrected for expansion ofDNA solutions) temperature. (Ad ad final tions, calculated (11) from Tm values of 77.20C versus at solutionsbsorb DNA 260 mn ofInitial (correspective 4.3C, espetivly,wer 51%forE. oliances andand 84.30C, respectively, were 51% for E. coli rected for temperature difference) were: for EC DNA, DNA (in 0.02 M sodium phosphate, pH 7.0) and Ai 1.016, Af - 1.337; for AV DNA, Ai 1.020, Af65.2% for A. vinelandii DNA (in 0.024 M sodium 1.360. phosphate, pH 7.0). DNA renaturation. Renaturation kinetics were examined to establish the unique sequence renaturation. The COt1/2 of 7.4 mol- s/liter oblength or nuclear complexity of A. vinelandii tained for E. coli DNA in 0.12 M phosphate DNA relative to that of E. coli. Typical second- buffer (pH 7.0) corresponded well to published order plots of the progress of the renaturation values (4, 6) and thus established the validity of process in both A. vinelandii and E. coli DNAs the shearing, melting, and renaturation proceare presented in Fig. 2. These data were taken dures employed. The average value of k obtained at the time intervals indicated from continuous from five renaturation studies of E. coli DNA plots of decreasing OD20 which occurred during was 0.134 ± 0.005 liter/mol. s (corrected; 3) in renaturation. Initial concentrations of denatured phosphate buffers of 0.12 to 0.16 M correspondDNA were calculated by CO = Ao/7,200 for E. ing to an average Cotl,2 of 7.5 + 0.3 mol. s/liter. coli DNA (11) and C0 = Ao/7,080 for A. vinelanThe k value for the renaturation of A. vinedii (calculated on the basis of 65% G+C), where landii DNA in 0.18 M phosphate buffer (pH 7) 7,200 and 7,080 are the respective molar absorb- was 0.273 liter/mol.s which, upon correction for ance constants. The second-order velocity con- cation concentration equivalent to 0.12 M phosstant of 0.135 liter/mol -s (corrected to the cation phate buffer (3), was 0.137 liter/mol - s. The corconcentration in 0.12 M phosphate buffer; 3) is responding Cot1/2 was 7.3 mol. s/liter. The averthe reciprocal of the renaturation parameter for age value of k obtained from six renaturation E. coli usually quoted with regard to unique studies of A. vinelandii DNA was 0.138 ± 0.007 sequence length of DNA, Cotl/2. This latter value liter/mol. s (corrected; 3) in 0.12 to 0.18 M phosis the product of initial denatured DNA concen- phate buffers, corresponding to an average Cotl/2 tration and the timne required to achieve 50% of 7.3 ± 0.4 mol. s/liter. Renaturation data sug-
-
I
VOL. 138, 1979
A. VINELANDII DNA AND CHROMOSOMES
1A4
I
I 1
I
AV 1.3 AO-Aco
A -Am
EC
1.21.2 _,/0
/
1.1 r
1.1
_
0 is
I
20
3
I
Ug
511
1
MINUTES FIG. 2. Second-order renaturation plots forE. coli (0; EC) and A. vinelandii (0; AV) DNA. The fraction of denatured DNA remaining, (Ao - A.)/(A A.) (see text), was plotted versus time for E. coli DNA at 66°C and A. vinelandii DNA at 72°C.
875
chromosomes. Both sedimentation values are in the range (1,600-1,700S) reported for the average E. coli folded chromosome devoid of membrane (17). Membrane-bound chromosomes whose sedimentation rate was 3,200S (25) have been reported in E. coli, but efforts to make similar preparations in A. vinelandii by lysis of cells at C40 were unsuccessful. Because of the similarity in sedimentation rates of the folded chromosomes of the two species prepared by the same procedure, we assumed they were of similar size and tested them as follows. Suspensions of A. vinelandii and E. coli folded chromosomes were prepared, mixed, and then spun together in a sucrose density gradient. Figure 3 is a plot of radioactivity versus fraction number for one such experiment. Only one peak of radioactivity was obtained with lysates of either E. coli or A. vinelandii, and the individual peaks were recovered in the same gradient fraction. The data suggest very strongly that folded chromosomes of exponentially growing A. vinelandii and E. coli have very similar sedi:mentation velocities and are therefore of
similarsize.
255 AV(21U-) l 20
(2-.dC)-URACIL
4
gest that the DNAs of E. coli and A. vinelandii are very similar in their
unique sequence lengths.
t Because of its high content per cell, the DNA ofA. vinelandii could exist as many single folded x1 _33 chromosomes like those of E. coli, as fewer 0zI KJ multiply iterated structures of two or more unique sequence lengths of DNA, or as mixtures a. J of the two classes. Our approach to differentiat- 0 2* ing between these possibilities was to prepare a X C iEC folded chromosomes from A. vinelandii whose / (6_3H)-THYMADNE DNA and RNA had been labeled with 14C and to compare their sedimentation velocities (and 1 hence relative size) with preparations derived from E. coli whose DNA had been labeled with tritium. If the folded chromosome of A. vinelandii contained more than one unique sequence of . DNA, its sedimentation velocity would be 1N R iO greater than the 2,200S value observed for the ATION propER F S E. coli folded chromosome in its presumably FIG. 3. Sedimentation properties of folded chrofully replicated state (28). The average sedimentation rate (five deter- mosomes of E. coli (0; EC) and A. vinelandii (0; The E. coli DNA was labeled with 3H, and the minations) E. coli coli folded folded chromosomeswasAV). Of of E. chromosomes was minations) vinelandii RNA and DNA were labeled with "C. 1,600S upon comparison with that of the T4 A. Folded chromosomes were prepared, mixed, spun in bacteriophage marker (1,025S). Similar studies a sucrose density gradient, and sampled. Peak radioconducted with A. vinelandii preparations activity representing the folded chromosomes of both yielded an average value of 1,700S for its folded cell types was found in a single fraction.
876
SADOFF, SHIMEI, AND ELLIS
DISCUSSION The high DNA content in cells ofAzotobacter species is well established (14, 19, 30). Early efforts using nuclear staining suggested that during exponential growth these bacteria contained multi-lobate nuclei (5, 18). However, direct staining methods for nuclear analysis are fraught with artifacts and are difficult to interpret. Thus, little has been known of the array of nuclei in these cells. In our studies we purified both A. vinelandii and E. coli DNA and used the latter material as a standard for the procedures we have employed in analyzing the genomic complement of A. vinelandii. This was particularly important in renaturation studies, where the extent of shearing of DNA and cation concentration have significant effects on the results of each experiment. The cell lysis and DNA extraction process outlined in Materials and Methods was a consequence of our need for pure DNA and our inability to obtain it from A. vinelandii by established techniques (12). Impurities affect both melting and renaturation characteristics of DNA (3). The final step, isopycnic centrifugation in a CsCl gradient, was required to separate low levels of RNA and carbohydrate from the DNA. The close correspondence between G+C contents calculated from the Tm and buoyant density data suggests that A. vinelandii DNA contains very few or no unusual bases. The average G+C of 65% noted in these studies agrees with accepted values (8) for A. vinelandii and Azotobacter chroococcum, a closely related organism.
Both denaturation and renaturation kinetics of DNA were studied spectrophotometrically because of the ease of observation and because of the precision of temperature control achieved in the thermal programmer, which was an integral part of the spectrophotometer. Typical secondorder kinetics were obtained in renaturation reactions for the initial 60 min of the reactions (Fig. 2) (10 to 15% renaturation). The reaction deviates from second-order kinetics when observed over a prolonged time, but initial rates are adequate for determinations of Cot1/2 values (6, 26). Presumptive evidence for the successful preparation of E. coli folded chromosomes included cell lysis without a marked increase in the viscosity of the lysate and the occurrence of DNAcontaining particles sedimenting at 1,600S (25, 28). When treated with RNase, cell lysates became viscous, and no discrete concentration of DNA sedimenting at 1,600S could be detected. A single procedure was effective in the preparation of folded chromosomes of both E. coli and
J. BACTERIOL.
A. vinelandii (23). Their identical sedimentation properties (Fig. 3), as well as their very similar Cot1/2 values and, hence, unique sequence lengths, suggest that the chromosomes of these two organisms are of the same size. Assuming the mass of the E. coli chromosome to be 3.5 x lo-15 g and knowing that exponentially growing A. vinelandii contains 15 x 10-14 g of DNA per cell (19), we conclude that this large bacterium must contain at least 40 genomic equivalents of DNA per cell. This large number of chromosomes may result in difficulties in the segregation of homocaryotic cells following mutation in this organism. ACKNOWLEDGMENTS We thank Loren Snyder for the gift of 3H-labeled T4 bacteriophage. This study was supported by Public Health Service grant AI-01863 from the National Institute of Allergy and Infectious Diseases.
LITERATURE CITED 1. Anderson, E. H. 1946. Growth requirements of vimsresistant mutants of Escherichia coli strain "B". Proc. Natl. Acad. Sci. U.S.A. 32:120-128. 2. Bishop, P. E., and W. J. Brill. 1977. Genetic analysis of Azotobacter vielandii strains unable to fx nitrogen. J. Bacteriol. 130:954-956. 3. Britten, R. J., D. E. Graham, and B. R. Neufeld. 1974. Analysis of repeating DNA sequences by reasociation. Methods Enzymol. 29:363-418. 4. Britten, R. J., and D. E. Kohne. 1968. Repeated sequences in DNA. Science 161:529-540. 5. Dondero, N. C., and ML R. Zelle. 1963. Observations on the formation and behavior of conjugation cells and large bodies in Azotobacter agile. Science 118:34-36. 6. Gills, M., J. De Ley, and M. De Cleene. 1970. The determination of molecular weight of bacterial genome from renaturation rates. Eur. J. Biochem. 12:143-153. 7. Green, M., M. Alexander, and P. W. Wilson. 1953. Mutants of the Azotobacter unable to use N2. J. Bacteriol. ":623-624. 8. Johnstone, D. B. 1974. Genus I. Azotobacter Beijerinck 1901, 567, p. 254-255. In R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology. The Williams & Wilkins Co., Baltimore. 9. Lin, L P., and H. L Sadoff. 1968. Encystment and polymer production by Azotobacter vinelandii in the presence of 8-hydroxybutyrate. J. Bacteriol. 95:23362343. 10. Loperfido, B., and H. L Sadoff. 1973. Gennination of Azotobacter vinelandii cysts: sequence of macromolecular synthesis and nitrogen fixation. J. Bacteriol. 113: 841-846. 11. Mandel, M., and J. Marmur. 1968. Use of ultraviolet absorbance-temperature profile for detennining the guanine plus cytosine content of DNA. Methods Enzymol. 12:195-206. 12. Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from microorganism. J. Mol. Biol. 3: 208-218. 13. Marmur, J., and P. Doty. 1962. Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. J. Mol. Biol. 5:109-118. 14. Miiler, H. P, and H. Kern. 1967. Strahlenresisteng, Gehalt, und Base nsetzung der DNA einiger strahlendurzierter Mutaten von Azotobacter chroococcum. Z. Naturforwch. 22:1330-1336. 15. Page, W. J., and H. L. Sadoff. 1976. Physiological factors
VOL. 138, 1979
16.
17.
18. 19. 20.
21.
22.
A. VINELANDII DNA AND CHROMOSOMES
affecting transformation of Azotobacter vinelandii. J. Bacteriol. 125:1080-1087. Page, W. J., and H. L. Sadoff. 1976. Control of transformation competence in Azotobacter vinelandii by nitrogen catabolite derepression. J. Bacteriol. 125:10881095. Pettijohn, D. E., and R. Hecht. 1973. RNA molecules bound to the folded bacterial genome stabilize DNA folds and segregate domains of supercoiling. Cold Spring Harbor Symp. Quant. Biol. 38:31-41. Pochon, J., Y. T. Tchan, and T. L Wang. 1948. Recherches sur le cycle morphologique et l'appareil nucleaire des Azotobacter. Ann. Inst. Pasteur 74:182-188. Sadoff, H. L, E. Berke, and B. Loperfido. 1971. Physiological studies of encystment in Azotobacter vinelandii. J. Bacteriol. 105:185-189. Schildkraut, C. L, J. Marmur, and P. Doty. 1962. Determination of the base composition of deoxyribonucleic acid from its buoyant density in CsCl. J. Mol. Biol. 4:430-443. Shah, V. K., L C. Davis, J. K. Gordon, W. H. OrmeJohnson, and W. J. Brill. 1973. Nitrogenase. III. Nitrogenaseless mutants of Azotobacter vinelandu: activities, cross-reactions and EPR spectra. Biochim. Biophys. Acta 292:246-255. Shah, V. K., L C. Davis, M. Stieghorst, and W. J. Brifl. 1974. Mutant of Azotobacter vinelandii that
23.
24. 25.
26. 27. 28. 29. 30.
877
hyperproduces nitrogenase component II. J. Bacteriol. 117:917-919. Sirotkin, K., J. Wei, and L Snyder. 1977. T-4 bacteriophage-coded RNA polymerase subunit blocks host transcription and unfolds the host chromosome. Nature (London) 265:28-31. Socolofsky, M. D., and 0. Wyss. 1962. Resistance of the Azotobacter cyst. J. Bacteriol. 84:119-124. Stonigton, 0. G., and D. E. Pettijohn. 1971. The folded genome of Escherichia coli isolated in a proteinDNA-RNA complex. Proc. Natl. Acad. Sci. U.S.A. 68: 6-9. Wetmur, J. G., and N. Davidson. 1968. Kinetics of renaturation of DNA. J. Mol. Biol. 31:349-370. Wilson, P. W., and S. G. Knight. 1952. Experiments in bacterial physiology. Burgess Publishing Co., Minneapolis. Worcel, A., and E. Burgi. 1972. On the structure of the folded chromosome of Escherichia coli. J. Mol. Biol. 71:127-147. Worcel, A., E. Burgi, J. Robinton, and C. L. Carlson. 1973. Studies on the folded chromosome of Escherichia coli. Cold Spring Harbor Symp. Quant. Biol. 38:43-51. Zaitseva, G. N., I. A. Khmel, and A. N. Belozerskii. 1961. Biochemical changes in synchronous cultures of Azotobacter vinelandii. Dokl. Akad. Nauk. S.S.R. 141: 740-743.
