Plant Physiol. (1 976) 58, 100-106
Attempts to Detect Agrobacterium tumefaciens DNA in Crown Gall Tumor Tissue1 Received for publication November 10, 1975 and in revised form March 11, 1976
DONALD J. MERLO2 AND JOHN D. KEMP Plant Disease Resistance Research Unit, Agricultural Research Service, United States Department of Agriculture, and Department of Plant Pathology, University of Wisconsin, Madison, Wisconsin 53706 ance of neoplastic, nonself-limiting overgrowths on the stems of wound-inoculated plants. The wound requirement is absolute, and thus far, only the presence of live bacteria in the intercellular space of the wound site has been shown to incite tumors reproducibly. Once the transformation to tumor cells has been attained, presence of the bacteria is no longer required so that sterile crown gall tissue can be cultured indefinitely. The most striking feature of crown galls on plants is the rapid appearance and growth of unregulated, disorganized tissues. Polyploidy and polyteny are characteristic of many tumor tissues, but those of sunflower (Helianthus annuus) often remain diploid (1, 6, 19, 25). On some plant hosts, notably sunflower, Paris daisy, and chrysanthemum, the development of primary tumors at the wound site is followed by the appearance of secondary tumors at sites which may be several cm above or below the primary tumor (2). Such tumors are similar to the primary tumors, but do not contain A. tumefaciens (4, 41), thus greatly facilitating their establishment in sterile tissue culture. In initial isolations from plants, tumor tissues will grow on media lacking phytohormones (3), whereas healthy plant tissues will not grow unless hormones are supplied. Hormone independence is not, however, a unique feature of tumor cells for it is possible to induce hormone independence in culture by appropriate manipulation of healthy callus (13). Schilperoort and co-workers (26-29) first suggested that A. tumefaciens A6 DNA sequences were present in tobacco (Nicotiana tabacum L.) crown gall tumor cells. Evidence supporting Schilperoort's work came from Quetier and co-workers (24), who detected significant hybridization between A. tumefaciens DNA and tobacco tumor tissue DNA, and estimated that one complete bacterial genome was present per transformed plant cell. Subsequently, Srivastava (33) detected approximately twice as much homology between A. tumefaciens DNA and tocacco tumor tissue DNA as between bacterial DNA and normal tobacco callus DNA. After repeating the saturation hybridization experiments of Schilperoort (28), Eden et al. (10, 12) concluded that 1% of either tobacco tumor or normal callus DNA hybridized to phage or bacterial sequences. However, the thermal stability of the hybrids was very low and the complexes did not exhibit a sharp thermal transition profile. These workers concluded that neither A. tumefaciens DNA nor phage PS8 DNA could be detected in the DNA of five different lines of crown gall tumor cells, including a tumor line incited by A. tumefaciens A6. By DNA-DNA reassociation experiments, Drlica and Kado Crown gall disease incited by Agrobacterium tumefaciens (Smith and Townsend) Conn (31) is evidenced by the appear- (9) examined the DNA from periwinkle (Vinca rosea L.) crown gall tissues for the presence of A. tumefaciens B6 DNA and ' Research cooperative with the College of Agricultural and Life concluded that less than 0.02% of tumor DNA was homologous Sciences, University of Wisconsin, Madison, and the Agricultural Re- to A. tumefaciens DNA. If bacterial DNA was present at this level, it must average less than 0.2 bacterial genome equivalents search Service, United States Department of Agriculture. 2 Present address: Department of Microbiology and Immunology, per diploid periwinkle tumor cell. Chilton et al. (7) examined tobacco tumor DNA for the presence of A. tumefaciens and University of Washington, SC-42, Seattle, Wash. 98195
ABSTRACT Primary and secondary crown gafl tissue cultures were established from sunflower plants (Helianthus annuus, variety Mammoth Russian) wound-inoculated with Agrobacterium tumefaciens (Smith and Townsend) Conn strain B6. Growth rates of tumor tissues and habituated healthy sunflower stem section tissues on basal medium lacking auxin and cytokinin were compared to those of healthy sunflower stem section tissue grown on the same medium with added phytohormones. No difference was detected in the thermal denaturation midpoints (74.8 C) and melting profiles in 25 mm sodium phosphate (pH 6.8), or the buoyant densities in cesium chloride equilibrium centrifugation (1.687 g cm-3), between deoxyribonudeic acids (DNAs) isolated from crude nudear preparations of the four tissue types. No satellite DNA was observed in equilibrium centrifugation of unsheared plant DNAs. Heterologous DNA renaturation kinetic analyses were performed in 0.14 M sodium phosphate (pH 6.8) at 70 C. Thermal stability measurements of reassociated DNA revealed less than 1% of mismatched base pairs. Reannealing of sheared, denatured, radioactive A. tumefaciens B6 DNA (molecular weight, 325,000 daltons) in the presence of a 5400fold excess of sheared calf thymus, healthy tissue, or secondary sunflower crown gail DNA obeyed second order kinetics, with a CotI,2 of 2.8, identical to that observed when Bf DNA was reannealed in the absence of foreign DNA. Reannealing rates of B6 DNA in the presence of 5400-fold excesses of DNA from two lines of primary sunflower crown gaOl were increased 2.24- or 1.47-fold. Digestion of the tumor DNA preparations with pancreatic deoxyribonudease I until no detectable DNA remained, foOowed by restoration of solution viscosity by added calf thymus DNA, failed to remove the acceleration effect of the tumor DNA preparations. Reisolation of the reannealed nudeic acid formed in this experiment, and digestion with ribonudease A or deoxyribonudease I revealed that the double-stranded fraction was composed entirely of DNA-DNA duplexes, with no detectable DNA-RNA hybrids. The data indicate that tumor, but not healthy tissue DNA preparations contain some factor or factors (not DNA) which accelerate the reannealing of bacterial DNA. Sunflower tumor tissue DNAs, therefore, do not contain integrated A. tumefaciens DNA sequences in amounts greater than a random 'Is of the bacterial genome per diploid amount of plant DNA, or a complete bacterial genome per five diploid plant cell DNA equivalents. Further, the possibility of the presence of many copies of a specific portion greater than 5% of the bacterial genome is exduded.
Physiol. Vol. 58,
BACTERIAL DNA IN TUMOR TISSUE
phage PS8 DNA and concluded that tobacco crown gall tumor DNA did not contain as much as one entire bacterial genome per three diploid tumor cells, or two entire phage genomes per diploid tumor cell. More recently, Patillon (23) examined the reassociation kinetics of radioactive A. tumefaciens B6 DNA in the presence of an excess of moderately redundant or unique sequence DNA from Scorzonera hispanica crown gall tumor tissues. From the increase in the reannealing rate of the radioactive probe DNA, he concluded that 1.18 bacterial genome equivalents were present per DNA equivalent of a Scorzonera tumor cell. In view of these conflicting data obtained with other plant species, we established several lines of healthy, habituated healthy, primary crown gall, and secondary crown gall sunflower tissues in axenic culture. DNA was extracted from the four tissue types and compared by the techniques of thermal denaturation, cesium chloride equilibrium centrifugation, and DNA-DNA hybridization. MATERIALS AND METHODS
ton4 tissue disrupter in an equal volume (w/v) of nuclei isolation buffer (26) containing 67 mm sodium phosphate (pH 7.2), 0.3M sucrose, 1.8 mm magnesium sulfate, 2 mm disodium EDTA, and 4% Triton X-100. Nuclei were pelleted by centrifugation at 14,700g and the pellet was further homogenized in an equal volume (w/v) of predigested pronase (26) and shaken at 28 C for 1.5 hr. After adjusting the mixture to 0.5M in sodium perchlorate, it was incubated an additional1 hr. The preparation was extracted with an equal volume of chloroform-octanol (10:1), phases separated by centrifugation, and the nucleic acids precipitated from the aqueous phase by addition of 2 volumes of 95% ethanol. Precipitated nucleic acids were collected by centrifugation at 14,000g and washed three times with three pellet volumes of 70% ethanol. The washed pellet was dissolved in a minimal volume of 0.1 x SSC, RNase A was added to a final concentration of 50 /ig/ml, and the mixture was incubated at 28 C for 1.5 hr. Protein was removed, the DNA was precipitated and washed as above, and the final washed pellet was dissolved in 25 mm sodium phosphate (pH 6.8) (12.5 mm Na2HPO4 and 12.5 mM
Isolation of Tissue Cultures. Sunflower plants (Helianthus The DNA was further purified by passage through 5 g (dry annuus, variety Mammoth Russian) were grown in the green- weight) of hydroxylapatite (Bio-Gel HTP) packed in a heated house for 3 weeks (15 to 30 cm tall) and then were either (60 C) glass column (2.2 x 4.5 cm) equipped with a sintered inoculated as described previously (16) with A. tumefaciens glass bed support. The DNA preparation was adsorbed to the (Smith and Townsend) Conn strain B6 or used to establish column in 25 mm sodium phosphate and the column was washed healthy sunflower callus cultures. with this buffer until the effluent absorbance at 254 nm was Two to 6 weeks after inoculation, two primary sunflower negligible. DNA was then eluted by a linear gradient of sodium crown galls3 (PSCG-4 and PSCG-1 1) were removed from their phosphate from 25 mm to 0.3M at a flow rate of 1.6 ml/min. respective plants, washed, and surface-sterilized with 95% Double-stranded DNA, which eluted from the column between ethanol and 6.4% sodium hypochlorite. The central core of the 0.20 and 0.25M sodium phosphate, was collected. crown gall was aseptically removed, again surface-sterilized, A. tumefaciens strain B6 cells were grown on a medium conwashed with basal medium containing antibiotics, and cut into 1- taining per liter: 3 g K2HPO4, 1.15 g NaH2PO4 H20, 1 g mm cubes. The cubes were planted on Linsmaier and Skoog's NH4CI, 0.3 g MgSO4-7H2O. 0.15 g KCI, 12 mg CaCl2 2H20, medium (20) lacking hormones (basal medium), but containing 2.5 mg FeSO4- 7H2O, 2.5 g glucose, and 0.25 g casein hydrolyper liter: 50 mg Neomycin sulfate; 80 mg penicillin G; 50 mg sate. Radioactively labeled A. tumefaciens DNA was prepared streptomycin sulfate; and 50,000 units Polymixin B sulfate. All from cells grown in 20 ml of medium at 28 C supplemented with 3 antibiotics were cold sterilized and aseptically added to sterile mCi of (methyl- [3HJthymidine, 43 Ci/mmol, New medium. After 2 weeks, tissue pieces were transferred to basal England[3H]thymidine added to the culture in 0.3-mCi aliquots each Nuclear) medium without antibiotics and grown at 27 C in the dark. hr beginning in early log phase of growth. Virulence of the cells Secondary sunflower crown galls developed one or two nodes used to inoculate the labeling medium was confirmed for each above the primary gall and were apparently free of bacteria. DNA preparation. DNA was purified from frozen cell pellets These galls were removed from the plant, surface-sterilized by after lysing at 40 C for 5 hr in 0.3 sodium chloride, 0.1 M the procedures outlined for the primary galls, and cultured EDTA (pH 8), 2% SDS, and 0.02% predigested pronase. directly on basal medium without antibiotics. Subsequent steps were similar to those described for the purifiA culture of healthy SSS was established from pith cells cation of plant DNA. Specific radioactivity of the DNA excised from near a leaf axil. This tissue was maintained on basal preparations ranged from 0.8 to 2 x 106 dpm/,ug DNA. medium supplemented with 50 uM naphthaleneacetic acid and Physical Measurements. Thermal melting profiles of DNA 50 /LM kinetin. were determined by monitoring the absorbance of DNA soluAfter several transfers of SSS tissue on hormone medium, tions at 257 nm as the temperature was raised at 1.5 C/min. approximately 100 small pieces of the tissue were transplanted Absorbance was recorded by a Gilford model 240 recording to basal medium. Although most of these pieces showed no spectrophotometer equipped with a thermostated cell chamber. subsequent growth, a small percentage produced small amounts Temperatures were measured by a thermocouple located in the of new, viable tissue. These tissues pieces were transferred to cuvette chamber, and recorded on the same chart as the absorbfresh basal medium to establish a line of HSSS. ance. Compensation for volume changes was made by comparGrowth Rates of Tissue Cultures. Growth rates of the tissues ing the absorbance of the DNA solution to the absorbance of a were compared under standard culture conditions. Equally sized uracil solution of approximately the same absorbance. tissue pieces were planted, six pieces/plate, on the appropriate the techBuoyant density (p) measurements were made by and medium, with only one tissue type/plate. At intervals after the nique Blair of Meselson et al. (22), as modified by Wells initial planting, tissue from three plates was weighed in tared (39), in a Spinco model E ultracentrifuge equipped with uv aluminum dishes. The pieces were then dried to constant dry optics. All samples were run with a marker of Escherichia coli weight in a 55 C oven. DNA and all the values were based on an E. coli DNA DNA Extraction. Plant tissues were homogenized with a Poly- density of 1.7035 density (11, 37). Shearing of DNA. DNA was sheared with a Bronwill Biosonik 3Abbreviations: PSCG: primary sunflower crown galls; SSS: sunflower stem section; SSCG: secondary sunflower crown galls; HSSS: 4Mention of companies or commercial products does not imply rechabituated sunflower stem section tissue; Tm: thermal midpoints; GC: ommendation or endorsement by the United States Department of guanosine plus cytosine. Agriculture over others not mentioned.
MERLO AND KEMP
Plant Physiol. Vol. 58, 1976
II sonicator equipped with a 4-mm microtip, and a maximum power output of 125 w at 25 kHz. Shearing (power setting of 37%) was performed in bursts of 30 sec, for a total of 8 min. Nitrogen purges of 15 sec were performed between each burst. Aliquots of each sheared DNA preparation were analyzed by sedimentation velocity centrifugation (35) to determine the average mol wt (236-440 thousand daltons). In analysis of the DNA-reannealing kinetics, all curves have been normalized to an A. tumefaciens DNA mol wt of 325,000 by rearrangement of the equation of Wetmur and Davidson (40): where Cot,,,,rm =(mol wt'/570) (Cot,X,
= molecular weight of the Bi DNA in the particular exper-
normalized Cot value
iment experimental Cot value
Reannealing Procedures. Approximately 1 ,ug of sheared radioactive A. tumefaciens DNA was mixed with a 5400-fold excess of sheared foreign (calf thymus or plant tissue) DNA and sealed in 100-,ul capillary tubes (30-80 /.I/tube). The DNA was denatured at 105 C for 5 min, and reannealed in 0.14 M sodium phosphate buffer (pH 6.8) at 70 C (25 C below the thermal denaturation temperature of A. tumefaciens DNA in this buffer). At appropriate intervals, the contents of a capillary tube were diluted into water, and the amounts of single- or doublestranded DNA present were determined by hydroxylapatite column chromatography at 60 C (18). Columns were packed with 0.75 g (dry weight) of Bio-Gel HTP hydroxylapatite in a waterjacketed column (0.9 x 15 cm) equipped with a sintered glass bed support. The DNA was adsorbed to the column and eluted by a linear gradient of sodium phosphate as described above. Elution of DNA was determined by measuring absorbance of each 1-ml fraction at 257 nm. and counting an aliquot of each fraction in a liquid scintillation counter as previously described (15). Results are plotted as percentage of single-stranded radioactive bacterial DNA versus Cot (calculated from the product of the original concentration of radioactive bacterial DNA and the time of incubation of the particular aliquot) (5, 18).
RESULTS Soon after isolation, the SSS, HSSS, and PSCG tissues were indistinguishable in color and tissue morphology, both from one another and from the SSCG line which had been in culture for about 3 years. After1 year. only the SSS and the HSSS tissues became loose and friable (Fig. 1). These loose tissues were easily disrupted simply by gently shaking the Petri plate. Thetwo types of crown gall tissues (PSCG and SSCG). however, have retained their original morphology and were considerably more difficult to disrupt. During DNA extraction, the SSS and HSSS tissue homogenates turned a dark brown color almost immediately upon disruption, whereas the PSCG and SSCG homogenates browned much more slowly. In culture, the healthy tissues grew faster than the tumor tissues during the first 10 days after transfer to fresh medium. Fresh or dry weight doubling times for healthy and tumor tissues were 4 days and 6.5 days, respectively. After 2 weeks, the growth rates of healthy tissues decreased, resulting in an increase of the doubling time to about 8 days. whereas the growth rates of the tumor tissues remained unchanged.
of DNA. DNA purified from healthy
tumor tissues had a typical nucleic acid absorption spectrum, with a maximum absorbance at 257 nm and a minimum at 230 to 235 nm. The four tissue types yielded 9.5 to 13 gg of DNA/g of
tissue homogenized. The thermal denaturation
profiles for HSSS, SSS,
FIG. 1. Primary and secondary sunflower crown galls (PSCG and SSCG, respectively) and habituated sunflower stem section (HSSS) tissues were grown on basal medium without phytohormones. Sunflower stem section (SSS) tissue was grown on the same medium with added auxin and cytokinin. Cultureswere established on these plates by aseptically planting 3-mm cubes of tissue approximately 4 weeks earlier. A: PSCG; B: SSCG; C: SSS; D: HSSS.
SSCG DNA preparation in 0.038 M sodium phosphate (pH 6.8) were nearly superimposable and had thermal midpoints of 77.5 C. Three of these profiles are shown in Figure 2. Using the equation of Mandel and Marmur (21). the molar percentage of guanosine plus cytosine (GC content) of all three plant DNA preparations was calculated to be 41.4%. A. tumefaciens DNA had a Tm of 83.5 C in 0.03 M sodium phosphate. This corresponds to a GC content of 59.4%, a value in accord with that previously reported (32). No significant differences were found in the buoyant densities of the various types of plant DNA (1.687-1.688 g cm-:') but they were considerably different from the buoyant density of A. tumefaciens DNA (1.713 g cm-:'). From the equation of Kemp and Sutton (17), the GC contents of these DNAs were found to confirm the previous values calculated from Tm determinations. The presence of large amounts of high GC content A. tumefaciens DNA in preparations of sunflower tumor DNA might be detected by either a significant difference between the buoyant densities of tumor and healthy tissue DNA, or by the appearance of a satellite DNA more dense than the main band tumor DNA. No significant differences in the buoyant densities of any of the plant DNAs were detected nor was any satellite DNA observed in the analytical ultracentrifuge. Reassociation of denatured, double-stranded DNA from organisms with genes in only a single copy/genome (e.g. bacterial DNA) has been shown to follow second order reaction kinetics (5, 40). This is illustrated by the reannealing of denatured A. tumefaciens DNA (Fig. 3). Tritium-labeledA.tumefaciens DNA was added to a100-fold excess of nonlabeled A. tumefaciens DNA and reannealing was followed by monitoring both the absorbance and the radioactivity of the column fractions. The percentage of single-stranded DNA ranged from 97% at Cot 0, to 4% at Cot 100. The experimentally determined points coincide with the line drawn for a theoretical second order reaction with a of 2.8 and the percentage of single-stranded DNA as calculated from absorbance and radioactivity measurements of the column fractions are nearly identical.
Plant Physiol. Vol. 58, 1976
BACTERIAL DNA IN TUMOR TISSUE
The reannealing kinetics of radioactive A. tumefaciens DNA in the presence of a 5400-fold excess of calf thymus DNA (Fig. 5A) reveals that a large excess of nonrelated DNA has no effect ;ooo 35+-I00f on the reannealing kinetics of the bacterial DNA. Neither was a significant change observed in the reannealing rate of A. tumefaL° 00 ciens DNA in the presence of a 5400-fold excess of healthy I- 30± .' sunflower tissue DNA (Fig. SB), nor in the presence of the same fold excess of secondary sunflower crown gall DNA (Fig. 5C). u 25-41 Therefore, no sequence homology existed between A. tumefa0. ciens DNA and DNA from tissues derived from noninfected 20-+ plants or from secondary tumors. 6 In contrast, when A. tumefaciens DNA was reannealed in the I- 15 presence of a 5400-fold excess of PSCG-4 DNA, the bacterial -tz DNA-reannealing rate increased slightly, but significantly (Fig. uJV Ur I 6). The Cot112 of 1.9 indicates a 1.47-fold shift from the standard [email protected]
|lgOE ~~~~~I 10 -+curve. In a reconstruction experiment, enough nonlabeled A. tumefaciens DNA was added to the tumor DNA preparation to increase the reannealing rate of the radioactive bacterial DNA 5by a factor of 5.27. The observed increase, 5.19-fold (Fig. 6), demonstrated that A. tumefaciens DNA can be detected if pres0 ent in small amounts. The acceleration of reannealing of A. 50 60 30 40 70 80 90 100 tumefaciens DNA by the PSCG-4 DNA, therefore, implied the contribution of 0.47 ,ug of bacterial sequence DNA by 5.4 mg of TEMPERATURE (C) PSCG-4 tumor DNA. FIG. 2. Thermal denaturation profiles of SSS (A), SSCG (EO), and When radioactive A. tumefaciens DNA was reassociated in the PSCG (0) DNAs in 0.038 M sodium phosphate, pH 6.8. presence of a 5400-fold excess of DNA from PSCG-1 1 tissue, the reannealing rate of the bacterial DNA increased even more 100 l Ii I I (Fig. 7). The Cot1,2 of the best fit second order reaction curve was 1.25, a shift of 2.24-fold from the standard curve. This shift *0 implied the contribution of 1.24 ,ug of bacterial sequence DNA 80by 5.4 mg of PSCG-11 tumor DNA. In order to determine cr whether the accelerated reannealing rate was indeed due to bacterial DNA in the PSCG DNA preparations, a DNA diges60tion-replacement experiment was performed as follows. Either sheared PSCG-4 (Fig. 6) or PSCG-l1 (Fig. 7) DNA was diin 2 mm sodium phosphate at room temperature for 1 hr gested z \o with 40 gg/ml of pancreatic DNase I. This digestion was found 40to convert all of the plant DNA preparation to nucleotides and 11 I I 1 11 1 40, -r%lpI
Cot FIG. 3. Reannealing of sheared A. tumefaciens DNA in 0.14 M sodium phosphate at 70 C. Cot values in this and subsequent Cot curves were normalized to a bacterial DNA mol wt of 325,000 daltons as described in text. Per cent of single-stranded DNA calculated from the absorbance (@) or radioactivity of [3H]DNA (0) in fractions from hydroxylapatite columns. Solid line represents a theoretical second order reaction curve with Cot112 = 2.8.
The fidelity of base-pairing in the reannealed DNA product determined by comparison with the thermal denaturation profiles of sheared, native A. tumefaciens DNA (Fig. 4). When the observed Tms for the reannealed DNA preparations shown in Figure 4 were normalized to the same sodium phosphate concentration (0.03 M) used to reanneal the native, sheared DNA, all Tms were within 0.5 C (82.5-83). Ullman and McCarthy (36) reported a decrease of 2.9 to 3.5 C in the Tm of reannealed DNA for every 1 % of mismatched base pairs. Therefore, the stringency of our reannealing conditions was such that less than 1 % of the base pairs formed during reannealing were not perfect matches. The fraction of DNA which eluted from hydroxylapatite columns as single-stranded DNA apparently contains few base pairs conferring helical structure, because the melting profile has no extended sharp transition. was
FIG. 4. Thermal denaturation profiles of sheared A. tumefaciens DNA (mol wt, 318,000) in 0.03 M NaPO4 (A); and of reannealed DNA products. Single-stranded DNA at Cot = 2.01 in 0.05 M NaPO4 (X); double-stranded DNA at Cot = 2.01 in 0.055 M NaPO4 (E); doublestranded DNA at Cot = 101 in 0.033 M NaPO4 (0).
Plant Physiol. Vol. 58, 1976
MERLO AND KEMP 100 a
rate (open circles in Figs. 6 and 7). A control experiment in which 6 ,ug nonradioactive A. tumefaciens DNA was added to the tumor DNA before DNase digestion revealed no additional acceleration of the reannealing rate of the radioactive probe (points identical to open circles in Fig. 6). This result would be expected if the DNA had been totally digested. Although our DNA preparations were extensively treated with ribonuclease during isolation, and were freed of residual RNA by elution from hydroxylapatite columns, the possibility of the formation of a [3H]DNA-RNA hybrid during reannealing reactions could not be dismissed. Reannealed [3HJDNA, prepared in the digestion-replacement experiment with a PSCG-I 1 DNA preparation (Fig. 7), was concentrated and dialyzed to a
80 60 -4
T10 1 *0
z 05 40
1&J -J 0
O _ 0.01
Cot FIG. 6. Reannealing of [3H]DNA from A. tumefaciens in the presence of a 5400-fold excess of PSCG-4 DNA preparations. Conditions same as for Figure 3; dotted line same as solid line in Figure 3. [3H]DNA
reannealed in presence of PSCG-4 DNA (O), or in presence of DNasetreated PSCG-4 DNA with added calf thymus DNA (0). Reconstruction experiment included nonradioactive bacterial DNA at 4.27 times the amount of [3H]DNA (A).
Cot FIG. 5. Points represent the reannealing of sheared [3HJDNA from A. tumefaciens (2.5-2.9 ,ug/ml) in the presence of a 5400-fold excess of sheared foreign DNA. Solid line represents a theoretical second order reaction curve with CotlI2 = 2.8. Conditions same as Figure 3. A: excess of calf thymus DNA; B: excess of SSS DNA (0) or HSSS DNA (0); C: excess of SSCG DNA.
oligonucleotides. Enzyme activity was eliminated by heating to 99 C for 5 min, and the viscosity of the solution was restored by adding sheared calf thymus DNA in an amount equal to the original quantity of plant DNA. Tritium-labeled A. tumefaciens DNA was added, and reannealing was performed as described previously. Digestion of the DNA from either PSCG preparation failed to eliminate the acceleration in the DNA-reannealing
Cot FIG. 7. Reannealing of [3HJDNA from A. tumefaciens in the presence of a 5400-fold excess of PSCG-1 1 DNA preparations. Conditions same as for Figure 3; dotted line represents the standard curve with Cot112 = 2.8. [3HJDNA reannealed in presence of PSCG-1 1 DNA (0), or in presence of DNase-treated PSCG-1 1 DNA with added calf thymus DNA (0).
BACTERIAL DNA IN TUMOR TISSUE
sodium phosphate concentration of 6.5 mm. Half of the preparation was treated with 60 ,ug/ml of pancreatic ribonuclease A for 1.5 hr at 21 C; the other half was adjusted to 1 mm MgSO4, and incubated with 60 of pancreatic deoxyribonuclease I. The two preparations were then chromatographed on analytical hydroxylapatite columns. Treatment with deoxyribonuclease converted all of the radioactive material into a form which eluted from hydroxylapatite in 25 mm sodium phosphate, indicating the conversion of double-stranded DNA to nucleotides and oligonucleotides. Ribonuclease treatment, however, had no effect on the chromatographic behavior of the duplexes.
DISCUSSION Crown gall disease is particularly suited for tissue culture study because tumor tissues can be grown in the absence of the inciting organism and will maintain their transformed state and autonomous growth. We have established, in culture, three types of sunflower tissues (PSCG, SSCG, and HSSS) all of which are capable of indefinite growth on hormone-free media. Physiological differences between control (HSSS) and tumor tissues which may be unrelated to transformation were minimized by isolating these tissues at about the same time from plants which were genetically and physiologically similar. These tissues, therefore. comprise an excellent model system for examination of biochemical and physiological events related to transformation to the tumor state. We detected some differences between healthy and tumor tissues. These included slight differences in growth rates, tissue morphology, and browning upon homogenization, but it is unclear whether they are related to the mechanism of tumorigeneSiS.
Characterization of the DNA extracted from the various tissue types revealed no significant differences on a gross physical basis. Therefore, the possibility of the inclusion of large amounts of bacterial DNA into the sunflower genome as part of the mechanism of tumorigenesis seems unlikely. Other workers have observed the presence of a satellite DNA when tumor, but not healthy, sheared DNA was centrifuged to equilibrium in CsCl (24. 34). No satellite DNA was observed when our sunflower DNA preparations were centrifuged either in the presence or absence of the internal standard E. coli DNA. However, we examined only native, unsheared DNA (mol wt about 6 x 106 daltons). Bacterial genomes, present in tumor DNA as covalently integrated segments of only 10 genes or less, would be expected to band with the bulk of the plant DNA and increase its average buoyant density. The absence of an increase in the buoyant density of tumor DNA indicates that large amounts of integrated small pieces were not present. The absence of satellite DNA indicates that the tumor DNA preparations were not contaminated with large amounts of nonintegrated foreign DNA or foreign DNA integrated in pieces larger than 10 genes. DNA renaturation experiments revealed no shift in the reannealing curve of A. tumefaciens DNA in the presence of large excesses of healthy tissue DNA. However, our preliminary results obtained with PSCG-1 1 DNA preparations (Fig. 7) suggested a contribution of 1.24,g of bacterial DNA by the addition of 5400 A.g of tumor DNA to the reaction mixture. Since the genome size of A. tumefaciens (3.7 x 109 daltons) (8) comprises 0.038% of the amount of DNA present in a diploid , g of bacterial sunflower cell (9.64 x 1012 daltons) (1), 1.24 DNA/5400 ,g of plant DNA would represent the presence of a random 0.37 of the bacterial genome/diploid amount of plant DNA, or an entire bacterial genome/three diploid plant cells. Similarly, PSCG-4 DNA (Fig. 6) was suspected to contain the equivalent of a random 0.20 of the bacterial genome in each diploid plant cell, or an entire bacterial genome in five diploid plantcells. Asimilarshiftin the reannealing curve of A. tumefa-
ciens DNA observed by Patillon (23) suggested the presence of an entire bacterial genome per Scorzonera tumor cell. Chilton et al. (7) observed no shift in the reannealing curve of A. tumefaciens DNA and concluded that tobacco tumor DNA contained no bacterial DNA. We previously demonstrated that tobacco tumor DNA, but not healthy tobacco tissue DNA, contains a non-DNA factor that can shift the reannealing curves of A. tumefaciens DNA (14). The DNase digestion experiments confirmed the presence of such a factor in sunflower. These experiments indicate that the accelerated reannealing rate of A. tumefaciens DNA in the presence of both types of PSCG DNA preparations may be due to such factors. The possibility that pancreatic deoxyribonuclease-resistant bacterial DNA sequences are present in tumor DNA preparations seems most unlikely, since added nonradioactive A. tumefaciens DNA was eliminated by DNase treatment. Interpretation of DNA-reannealing experiments relies on an accurate determination of the ratio of single- to double-stranded DNA. We defined reannealed double-stranded DNA as that material which eluted from hydroxylapatite at highest sodium phosphate concentrations. The thermal stability, hyperchromicity, and sharpness of the thermal transition profile (Fig. 4) of this material indicate that it is a double helical nucleic acid with fewer than 1% mismatched base pairs. The DNase sensitivity of reannealed duplexes confirms the identity of double-stranded DNA. Sedimentation velocity determinations of the reannealed double-stranded DNA suggested no change in mol wt after incubation at 70 C for 6 days (data not shown). Thus, little thermal degradation or aggregation had occurred. The possibility that residual RNA in our tumor DNA preparations caused the accelerated reannealing rate of A. tumefaciens DNA through formation of DNA-RNA hybrids was eliminated by examination of the nuclease sensitivity of the reannealed duplexes. Siebke and Ekren (30) found that treatment of DNARNA hybrid with ribonuclease in 10 mm KCI buffer destroyed the complex and resulted in the release of single-stranded DNA. The conditions of our ribonuclease digestion seemed sufficient to degrade any DNA-RNA complexes present in the reannealed DNA preparation. The increase in reannealing rate induced by the preparation was approximately 2-fold. Thus, assuming that the RNA was complementary to both of the original DNA strands, about half of the radioactivity would have been released as single-stranded DNA if an actual DNA-RNA hybrid had been present. Since no single-stranded DNA was present after the ribonuclease treatment, our reannealed product was a DNADNA complex. In conclusion, these experiments indicate that sunflower crown gall tissue cultures do not contain integrated segments of A. tumefaciens DNA in amounts greater than about one complete bacterial genome per five diploid plant cell DNA equivalents. Also, the results of the experiments exclude the presence of the specific portion (greater than 5%) of the bacterial genome. The presence of a specific portion smaller than 5% of the bacterial genome (e.g. the plasmid of Zaenen et al. [421, and Watson et al. ) would not have been detected. Therefore, our conclusions agree with those of others (7, 9, 10, 12). Patillon's report (23) of the presence of A. tumefaciens sequences in Scorzonera appears to be inconclusive since no reannealing reactions were reported in the presence of healthy tissue DNA, nor were any physical characterizations performed on the reannealed DNA. Perhaps Scorzonera tumor DNA contains a non-DNA-accelerating factor similar to those we found in tobacco tumor tissue (14) and in the sunflower tumors described in this paper. Tumorigenesis in crown gall, therefore, seems not to require the integration of a complete bacterial genome into the host cell genome. In fact, it may not require integration of any bacterial DNA.
MERLO Ak],ND KEMP
Acknowledgments -The authors acknowledge the expert photographic assistance of S. Vicen. LITERATURE CITED 1. BAETCKE. K. P.. A. H. SPARROW. C. H. NAUMAN. AND S. S. SCHWEMMER. 1967. The relationship of DNA content to nuclear and chromosome volumes and to radiosensitivitv (LD5,,). Proc. Nat. Acad. Sci. U. S. A. 58: 533-540. 2. BRAUN. A. C. 1941. Development of secondary tumors and tumor strands in the crown gall of sunflowers. Phytopathology 31: 135-149. 3. BRAUN. A. C. 1956. The activation of two growth-substance systems accompanying the conversion of normal to tumor cells in crown gall. Cancer Res. 16: 53-56. 4. BRAUN, A. C. AND P. R. WHITE. 1943. Bacteriological sterility of tissues derived from secondary crown-gall tumors. Phytopathology 33: 85-110)). S. BRITTEN. R. J. AND D. E. KOHNE. 1968. Repeated sequences in DNA. Science 161: _529540. 6. BUTCHER. D. N. 1973. The origins, characteristics. and culture of plant tumour cells. In H. E. Street. ed. Plant Tissue and Cell Culture. Botanical Monographs. Vol. 11. Blackwell Scientific Publications. London. pp. 356-391. 7. CHILTON. M. D.. T. C. CURRIER. S. K. FARRAND. A. J. BENDICH. M. P. GORDON. AND E. W. NESTER. 1974. Agrobacterium tumefac-iens DNA and PS8 bacteriophage DNA not detected in crown gall tumors. Proc. Nat. Acad. Sci. U. S. A. 71: 3672-3676. 8. DELEY. J.. R. TYTGAT. J. DFSMEDT. AND M. MICHIELS. 1973. Thermal stability of DNADNA hybrids within the genus Agrobacterium. J. Gen. Microbiol. 78: 241-252. 9. DRLICA. K. A. AND C. 1. KADO. 1974. Quantitative estimation of Agrobacterium tumefaciens DNA in crown gall tumor cells. Proc. Nat. Acad. Sci. U. S. A. 71: 3677-3681. 1(). EDEN. F. C.. S. K. FARRAND. J. S. POWELL. A. J. BENDICH. M. D. CHILTON. E. W. NESTER, AND M. P. GORDON. 1974. Attempts to detect deoxyribonucleic acid from Agrobacterium tumefaciens and bacteriophage PS8 in crown gall tumors by complementary ribonucleic acid/deoxyribonucleic acid-filter hybridization. J. Bacteriol. 119: 547553. I1. ERIKSON. R. L. AND W. SZYBALSKI. 1964. The Cs. SO, equilibrium density gradient and its application for the study of T-even phage DNA: glucosylation and replication. Virology 22: 1 1 1-124. 12. FARRAND. S. K.. F. C. EDEN. AND M. D. CHILTON. 1975. Attempts to detect Agrobacierium tumefaciens and bacteriophage PS8 DNA in crown gall tumors by DNA DNA-filter hybridization. Biochim. Biophys. Acta 390): 264-275. 13. GAUTHERET. R. J. 1946. Comparison entre l'actions de l acide indole-acetique et celle du Phytomonas tumefacien.s sur la croissance des tissus segetaux. C. R. Seance. Soc. Biol.
140: 169-17 1. 14. KEMP. J. D. AND D. J. MERLO. 1975. Promotion of DNA renaturation by a non-DNA factor in crown gall tumor DNA preparation. Biochem. Biophys. Res. Commun. 67: 1522-
1526. 15. KEMP, J. D. AND D. W. SUTTON. 1971. Protein metabolism in cultured plant tissues. Calculation of an absolute rate of protein synthesis. accumulation. and degradation in tobacco callus in vivo. Biochemistry 10: 8 1-88. 16. KEMP. J. D. AND D. W. SUTTON. 1975. Protein metabolism in cultured plant tissues. IV. Rates of protein synthesis and degradation in different portions of various tobacco tissues. Physiol. Plant 35: 45-52. 17. KEMP. J. D. AND D. W. SUTTON. 1976. A chemical and physical method for determining the complete base composition of plant RNA and DNA. Biochim. Biophys. Acta: 148-156. 18. KOHNE. D. E. AND R. J. BRITrEN. 1971. Hydroxyapatite techniques for nucleic acid reassociation. In: G. L. Cantoni and D. R. Davies. eds.. Pr c. Nucleic Acid Res.. Vol. 2. Harper and Row. New York. pp. 500-512. 19. KupILA. S. 1958. Anatomical and cytological comparison of the development of crown gall in three host species. Ann. Bot. Soc. Zool. Bot. Fennicae Vanamo' 30: 1-89. 20. LINSMAIER. E. M. AND F. SKOOG. 1965. Organic growth factor requirements of tobacco tissue cultures. Physiol. Plant. 18: 10)0-127.
Plant Physiol. Vol. 58, 1976
21. MANDEL, M. AND J. MARMUR. 1968. Use of ultraviolet absorbance-temperature profile for determining the guanine plus cytosine content of DNA. Methods Enzymol. 12: 195-206. 22. MESELSON. M.. F. W. STAHL. AND J. VINOGRAD. 1957. Equilibrium sedimentation of macromolecules in density gradients. Proc. Nat. Acad. Sci. U. S. A. 43: 581-583. 23. PATILLON. M. 1974. Quantitative evaluation by reassociation kinetics of Agrobacterium DNA sequences residing in the genome of a crown-gall tissue culture. J. Exp. Bot. 25: 860-870. 24. QUETIER. F., T. HUGUET, AND E. GUILLE. 1969. Induction of crown gall: partial homology between tumor-cell DNA. bacterial DNA, and the G+C-rich DNA of stressed normal cells. Biochem. Biophys. Res. Commun. 34: 128-133. 25. RASCH. E., H. SwIFT. AND R. M. KLEIN. 1959. Nucleo-protein changes in plant tumor growth. J. Biophys. Biochem. Cytol. 6: 11-34. 26. SCHILPEROORT. R. A. 1969. Investigations on plant tumors. Crown gall. On the biochemistry of tumor induction by Agrobacterium tumefaciens. Ph.D. thesis. University of Leiden. The Netherlands. 27. SCHILPEROORT. R. A. 1972. Integration of Agrobacterium tumefaciens DNA in the genome of crown gall tumor cells and its expression. In: H. P. Geesteranus. ed., Proc. Third Intl. Conf. Plant Pathogenic Bacteria. Wageningen. Centre Agr. Publishing and Documentation. Wageningen. pp. 223-238. 28. SCHILPEROORT. R. A.. N. J. VAN SITTERT. AND J. SCHELL. 1973. The presence of both phage PS8 and Agrobacteriun tumefaciens A, DNA base sequences in A,-induced sterile crown gall tissue cultured in vitro. Eur. J. Biochem. 33: 1-7. 29. SCHILPEROORT. R. A.. H. VELDSTRA. S. 0. WARNAAR. G. MULDER. AND J. A. COHEN. 1967. Formation of complexes between DNA isolated from tobacco crown gall tumours and RNA complementary to Agrobacterium twnefaciens DNA. Biochim. Biophys. Acta 145: 523-525. 30. SIEBKE. J. C. AND T. EKREN. 197(0. Chromatography of RNA-DNA complexes on hydroxyapatite. Eur. J. Biochem. 12: 380-386. 31. SMITH. E. F.. AND C. 0. TOWNSEND. 1907. A plant tumor of bacterial origin. Science 25: 67 1-673. 32. SOBER. H. A. (ed.). 1970). Handbook of Biochemistry. Selected Data for Molecular Biology. Chemical Rubber Co.. Cleveland. Ohio. 33. SRIVASTAVA. B. I. S. 1970). DNA:DNA hybridization studies between bacterial DNA. crown gall tumour cell DNA. and normal cell DNA. Life Sci. 9 (II): 889-892. 34. SRIVASTAVA, B. 1. S. AND K. C. CHADHA. 1970. Liberation of Agrobacterium tumefaciens DNA from the crown gall tumor cell DNA by shearing. Biochem. Biophys. Res. Commun. 40: 968-972. 35. STUDIER. F. W. 1965. Sedimentation studies of the size and shape of DNA. J. Mol. Biol. I 1: 373-390. 36. ULLMAN. J. S. AND B. J. MCCARTHY. 1973. The relationship between mismatched base pairs and the thermal stability of DNA duplexes. 1. Effects of depurination and chain scission. Biochim. Biophys. Acta 294: 40)5-415. 37. VINOGRAD. J.. J. MORRIS. N. DAVIDSON. AND W. F. DOVE. JR. 1963. The buoyant behavior of viral and bacterial DNA in alkaline CsCl. Proc. Nat. Acad. Sci. U. S. A. 49: 12-17. 38. WATSON. B.. T. C. CURRIER. M. P. GORDON, M. D. CHILTON. AND E. W. NESTER. 1975. Plasmid required for virulence of Agrobacterium turnefaciens. J. Bacteriol. 123: 255-264. 39. WELLS. R. D. AND J. E. BLAIR. 1967. Studies on polynucleotides. LXXI. Sedimentation and buoyant density studies of some DNA-like polymers with repeating nucleotide sequences. J. Mol. Biol. 27: 273-288. 40. WETMUR. J. G. AND N. DAVIDSON. 1968. Kinetics of renaturation of DNA. J. Mol. Biol. 31: 349-370. 41. WHITE. P. R. AND A. C. BRAUN. 1941. Crown gall production by bacteria-free tumor tissues. Science 94: 239-241. 42. ZAENEN. I.. N. VAN LAREBEKE. H. TEUCHY, M. VAN MONTAGU. AND J. SCHELL. 1974. Supercoiled circular DNA in crown-gall inducing Agrobacterium strains. J. Mol. Biol. 86: 1(09-127.