Clones of cauliflower mosaic virus identified by molecular hybridization in turnip leaves

ULRICH MELCHER, CHARLES O. GARDNER, JR., and RICHARD C. ESSENBERG

Department of Biochemistry, Oklahoma State University, StiUwater, OK 74078, USA

(Received 20 January 1981; in revised form 26 June 1981 )

Key words: Nucleic acid hybridization, local lesions, viral DNA, turnip

Abstract: Mechanical inoculation of turnip leaves with cauliflower mosaic virus (CaMV) results after one to two weeks in the appearance on these leaves of local lesions. Local lesions were detected by hybridization of radioactive CaMV DNA with nucleic acid immobilized in leaf skeletons by solvent extraction, proteinase digestion, and alkali treatment. The pattern of lesions detected as dark circles on autoradiographs of the washed leaf skeletons was the same as that detected by staining of solvent-extracted leaves for starch. Starch lesions appeared as white areas against a dark purple back- ground. These lesions were first detected between 5 and 8 days after inoculation and grew in size until 10 days after inoculation. Lesions were also detected by staining solvent-extracted and proteinase digested leaves with ethidium bromide. The lesions appeared as dark areas in a bright fluorescent background, and were found in the same positions as the starch lesions.

Int roduct ion

DNA-containing viruses of higher plants have the potential of serving as vectors for the efficient introduct ion of foreign DNA into plants. The use of these viruses as vectors would be greatly simplified by a method to identify and propagate only those viral DNA molecules that contain a desired inserted nucleic acid sequence. Using turnip leaves inoculated with cauliflower mosaic virus (CaMV), we have devised a technique analogous to bacterial colony hybridizat ion [6] . Our technique makes use of the restriction of viral growth

in the inoculated leaves to small (less than 2 mm diameter) circular areas, designated local lesions [7] , and to the draining veins. The appearance of such local lesions on plant leaves in response to infection by virus was first demonstrated by Holmes [7], who also found that the infectious virus was restricted to those regions of the leaf containing local lesions [8]. Quan- titative studies of the infectivity of CaMV on turnip leaves suggest that each lesion results from a single infectious event so that the virus in any one lesion is a clone [14]. Although local lesions can in some plants including turnip [14] be detected as chlorotic areas, staining of starch [8] has been widely used to detect and quanti tate local lesions more reliably. More recently local lesions undetectable by starch staining have been detected by other treat- ments, such as heat shock which allows the collapse of otherwise invisible

63

Plant Molecular Biology 01, 63 - 73 (1981) 016 7-4421/81/0011-0063 SOL 65 © Dr IV. Junk Publishers, The Hague, Printed in The Netherlands.

64

lesions [3] and 14CO2 feeding [2], that detect other physiological aberrations produced in leaf tissue by infection with viruses. Here we used the molecular hybridization of radioactively-labeled viral DNA with denatured and immobilized leaf nucleic acid to detect directly the viral nucleic acid contained within lesions.

Materials and methods

Vints propagation and starch lesion detection

The CM4-184 isolate of CaMV was maintained in and isolated from turnips, Brassica rapa L. cv. Just Right, as previously described [4]. Virus was purified from infected leaves by the method of Hull, et al. [10] 5"weeks after inoculation. DNA was purified from virus as previously described [4]. Leaves were inoculated by gently rubbing suitable dilutions in 1% (w/v)K2 HPO4 of CaMV or of DNA extracted from CaMV or SV40 (Bethesda Research Laboratories) onto leaves dusted with carborundum. One or two circular holes were punched in the leaves to facilitate later identification. After 5 to 12 days further growth leaves were removed from the plants and submerged in 2-methoxyethanol at 37°C until the leaves turned white (1 to 4hr.). For the detection of starch lesions' [15], the leaves were removed from the 2- methoxyethanol and submerged in potassium triiodide solution [0.060% (w/v) KI and 0.006% (w/v) I2 ] until the leaves became dark purple, at which time they were photographed using transillumination. Lesion sizes were measured from photographic prints and were corrected for photographic enlargement. To demonstrate infectious particles in leaf regions exhibiting starch lesions about 2 mm square sections of potassium triiodide stained leaf tissue were excised using a razor blade. The tissue was ground with 0.5 rnl 1% (w/v) K2 HPO4 and a little carborundum using a mortar and pestle. This hom- ogenate was used directly to inoculate test plants.

Nick translation

CaMV DNA or SV40 DNA (180 rig) was nick translated [ 12] in 20#16.6 mM Tris-HC1 pH 7.4, 6.6 mM MgC12, 1 mM dithiothreitol, 50 mM KC1 containing 1.0nmoles each of dATP, dGTP and TTP and 0.15nmole of t~ [32p]-dCTP (6 to 30Ci/mmole; New England Nuclear). Incubation at 15 °C was initiated after addition of 80pg deoxyribonuclease I (Sigma) and 16 pg DNA polym- erase I (P-L Biochemicals). The DNA (0.2 to 1.5 x 107 dpm//ag) was phenol! extracted and ethanol precipitated with carrier DNA before use in hybrid- ization.

Detection of viral nucleic acid

Leaves from which pigments had been extracted with 2-methoxyethanol as described above were incubated 16h at 37°C in 20ml per leaf of 0.1 mM NAN3,0.1% (w/v) sodium dodecyl sulfate containing 0.1 mg/ml proteinase K

65

to release DNA from viral particles. After several rinses with water, leaves were immersed in 1/ag/ml ethidium bromide for 1 to 3 hr before observation by transillumination with ultraviolet light (maximum emission at 30Onto). Leaves were then treated successively at room temperature for 15 min. each with 1.5 M NaC1, 0.5 M NaOH and then 3.0 M NaC1, 0.5 M Tris-HC1 pH 7.0. They were then dipped in 2 x SSC (SSC =0.15M NaC1, 0.O15M sodium citrate) and hung to air dry, before treatment in a vacuum oven at 80°C for 2 hr. Leaves treated in this manner have been stored for up to'six months in the dark at room temperature without any apparent loss in the hybridizability of their DNA. For hybridization, dried leaves were carefully introduced into heat-sealable plastic bags (Seal N Save, Sears) containing 4 ml of hybridization solution [1] (2 x SSC, 0.02% polyvinylpyrrolidone, 0.02% Ficoll, 0.02% bovine serum albumin, 0.1% sodium dodecyl sulfate, 1 mg/ml heat denatured, sonicated calf thymus DNA, 1 x l0 s to 1 x 106 dpm a2P-labeled CaMV or SV40 DNA). As the leaf rehydrated it was made to lie flat within the bag. As many as three leaves were introduced into one bag. The bags were heat-sealed and immersed in water at 68°C for 24hr. Leaves were washed after hybrid- ization in 3 to 6 changes of 2 x SSC each at 68°C for 10 to 30min. They were then blotted dry and arranged on a piece of filter paper under plastic wrap (Handi Wrap). Autoradiography using Cronex Lightning Plus intensifying screens (Dupont) with XR-film (Kodak) was performed at 4°C for '1 to 3 days.

Results

Starch lesions

To establish that local lesions in turnip leaves inoculated with CaMV were similar to those observed in other virus-host interactions, we examined the number, size, and time of appearance of lesions detectable by staining starch with iodine. Under the particular conditions of growth and nutrition of turnip plants that were used, the lesions appeared as unstained circles in a deep purple background (Figure 1). Such lesions were not seen in leaves that were not inoculated or leaves that were inoculated with SV40DNA (Figure 1). Lesions were first observed 5 to 8 days after inoculation (Table 1). After 8 days there was no significant increase in the number of lesions per inoculated leaf. The average size of the lesions and the fraction of lesions greater than 0.5 mm in diameter increased between 8 and 10 days but failed to increase significantly thereafter. It was possible to demonstrate that starch lesions contained infectious virus. Small squares of tissue were excised from stained leaves avoiding major veins. When the squares of tissue were ground and rubbed on leaves of healthy plants, lesions developed in only those leaves inoculated with material from an area containing a starch lesion. The lesions are thus analogous to the chlorotic lesions caused by CaMV described by Tomlinson and Shepherd [14].

66

Figure 1. Lesions induced by CaMV detected by iodine staining. Top leaf inoculated with 2 ~g]ml CaMV 12 days prior to excision, extraction and iodine staining as described in methods. Middle leaf was similarly treated 12 days after inoculation with 1.0t~g/ml SV40 DNA. Bottom leaf was not inoculated. Actual diameter of punched holes is 2.5 mm

67

Table 1 Number and Size of CaMV-Induced Starch Lesions During Early Stages of Infection

Days Lesions Mean Large lesions after per lesion as percent inoculation leaf diameter of all lesions 1

(mm)

5 0 m _

8 45 0.21 -+ 0.135 4% 10 30 0.52 -+ 0.27 47% 12 37 0.57 -+ 0.26 57% 123 38 0.62 -+ 0.32 63%

Large lesions were scored as greater than 0.5 mm in diameter 2 Standard deviation 3 Inoculated with CaMV DNA

Ethidium bromide lesions

Lesions were also detected by staining leaves with ethidium bromide. Since lesions were the site of viral growth, it was thought they might contain more nucleic acid and thus be detectable as brightly fluorescing areas when stained with ethidium bromide and observed with ultraviolet light. Leaves 8 to 12 days after inoculation with CaMV were therefore treated with 2- methoxyethanol to remove pigments and stained with ethidium bromide either with or without a prior t reatment with a proteolytic enzyme. When observed under ultraviolet light, uninoculated leaves showed a uniform bright orange fluorescence (Figure 2), wh ich was not seen without addition of ethidium bromide. A similar appearance was noted when leaves previously

inoculated with SV40 DNA were stained with ethidium bromide. Inoculated leaves had numerous small areas where little or no ethidium bromide fluorescence was observed. The size of these areas corresponded approximately to the size of the starch lesions. The positions of the ethidium bromide lesions corresponded to the positions of the starch lesions (Figure 2), indicating that both techniques detect the same set of virus-induced lesions. The contrast in fluorescence between lesions and surrounding tissue was greater when the leaf had been pretreated with protease.

Hybridization

To detect viral nucleic acid within the lesions, 2-methoxyethanol extracted leaves were treated overnight with proteinase K, with alkali for 15 min and then neutralized with Tris buffer. The leaf skeletons were dried and sub- sequently immersed in a hybridizat ion solution containing heat-denatured nick-translated CaMV DNA. After hybridizat ion and washing the leaves were autoradiographed. Skeletons of uninoculated leaves and of leaves inoculated with DNA of the mammalian virus, SV40, (Figure 3) showed a weak uniform background level o f radioactivity. Autoradiographs of skeletons of leaves inoculated with CaMV DNA showed numerous circular areas of blackening.

68

Figure 2. Lesions induced by CaMV detected by ethidium bromide staining. An uninoculated leaf (top) was treated as described in methods and stained with ethidium bromide. Another leaf had been inoculated with 1.0 t~g/ml DNA of CaMV 10 days prior to excision and stained with iodine (bottom) prior to ethidium bromide staining (center). Actual diameter of punched holes is 2.5 mm

69

Figure 3. Autoradiographic detection of lotions induced by CaMV. Leaves were extracted, stained with iodine and then further treated as described in methods to produce leaf skeletons. These were hybidized with 32 P-DNA of CaMV, washed and submitted to autoradiography. From top to bottom: autoradiograph of a skeleton of an uninoculated leaf; autoradiograph of a skeleton of a leaf inoculated with 2 zg/ml CaMV 12 days prior to excision; iodine staining pattern of this leaf; autoradiograph of a leaf inoculated with 1.0/zg/ml SV 40 DNA. Actual diameter of punched holes is 2.5 mm

70

There was an exact correspondence between the areas of hybridization and the starch lesions (Figure 3). However, approximately 10% of the starch lesions did not have corresponding spots on the autoradiographs. There was no correspondence between the autoradiographic density of lesions and size of starch lesions. Many of the areas of hybridization were more dense around the circumference of the circular lesions than in the interior of these circles

(Figure 4). When leaves were inoculated with CaMV DNA 12 days before harvesting there were, in addition, areas of hybridization in many of the veins (Figure 4). Considerably less hybridization to nucleic acid in veins was seen when intact virus was used inoculate leaves (Figure 3). To exclude the possibility that lesions in leaf skeletons contain substances that bind DNA non-specifically, skeletons of leaves that had been inoculated with CaMV DNA were hybridized with radioactive SV40 DNA. On the auto- radiograph of these leaves (Figure 5) only a weak background level of radioactivity was detected.

Figure 4. Autoradiographic appearance of lesions induced by 1.0/zg/ml DNA of CaMV, applied 12 days prior to excision. Leaf was treated as in Figure 3. Note the less dense appearance of the central region of the lesions and the dark regions in the veins. Actual lesion diameter is 1.2 mm

Discussion

The present results indicate that viral nucleic acid sequences may be readily detected within inoculated leaves of turnip by molecular hybridization. The viral nucleic acid is located in small rings or circles in the leaf tissue. The observation of rings suggests that after initial infection of a cell through wounding in the presence of virus particles or viral DNA and subsequent early spread to neighboring cells, the host ceils mount a near suicide response which degrades the viral nucleic acid along with starch and chlorophyll, but does not damage the cells' integrity. The ring would thus correspond to cells

71

Figure 5. Autoradiographic appearance of a leaf inoculated with 1.0 ~g/ml CaMV DNA 12 days prior to excision. Leaf skeleton was hybridized with 35 P-DNA from SV40

that have been recently infected and have not yet managed to respond in this manner. The decreased ethidium bromide fluorescence in the lesions may be due either to a destruction of nucleic acid in the cells comprising the lesion or to the local production of compounds that interfere with the fluorescence. Eventually virus enters the veins and establishes secondary centers of infection in the veins. Virus particles must also be transported down the veins to the growing point from which systemically infected leaves arise. That hybridization to the vein DNA was not seen with virus-inoculated leaves may indicate that at 12 days these leaves are at an earlier stage of infection, perhaps due to the extra time required to remove the virus coat.

The technique of detecting viral nucleic acid in leaves by hybridization with radioactive DNA will be important in the utilization of CaMV as a vector for the introduction of foreign DNA into a turnip plant. The detection of nucleic acid in lesions is specific for nucleotide sequence, since lesions are not detected when a DNA unrelated in sequence is used as a probe. Therefore, a lesion containing viral DNA that has foreign DNA inserted within it should be detectable by using as a probe radioactive DNA complementary to the insert. Digestion of CaMV DNA with restriction enzymes that cleave only once does not destroy the infectivity of the DNA [9, 11 ]. The method is suitable not only for the detection of such viral clones but also for their propagation since the lesions are sufficiently large that a portion of them can be excised prior to treatment for hybridization for subsequent inoculation on other plants. With this method, the number of possibte clones which must be propagated and have their DNA isolated and examined [5] is greatly reduced. The use of leaf skeleton hybridization also overcomes a possible difficulty inherent in methods of constructing recombinant CaMV DNA molecules which use cloning of plasmids in bacterial cells. Since in these latter approaches bacterial clones containing inserted DNA are identified while the CaMV DNA is in

72

bacterial cells as part of a plasmid DNA molecule, it is not possible to determine at the same time as insert-bearing clones are being identified whether the CaMV DNA port ion is infectious to plants. Natural variations in the viral DNA populat ion and sequence changes resulting from the processes of construction of recombinant DNA molecules and of cloning in bacteria may result in recombinant CaMV DNA molecules that cannot infect plants. This difficulty may be the reason that some laboratories [13] but not others [9] find that viral DNA cloned in bacteria is not capable of infecting plants when excised from the plasmid DNA. Leaf skeleton hybridizat ion using as probe DNA complementary to the insert simultaneously selects for infectivity

and the presence of inserted DNA.

Acknowledgments

We thank Drs. T. Ulrich and J. Widholm for introducing us to the iodine staining method. This research was supported by the USDA-SEA Competitive Grants Program, by a Presidential Challenge Grant from Oklahoma State University and by the Oklahoma Agricultural Experiment Station of which

this is journal article J-3942.

References

1. Denhaxdt DT (1966) A membrane filter technique for the detection of eDNA. Biochem Biophys Res Comm 23, 641-646

2. Doke N and Hirai T (1970) Radioautographic studies on the photosynthetic CO 2 fixation in virus-infected leaves. Phytopath 60,988-991

3. Foster JA, and Ross AF (1975) The detection of symptomless virus-infected tissue in inoculated tobacco leaves. Phytopath 65,600-610

4. Gardner CO Jr, Melcher U, Shockey MW and Essenberg RC (1980) Restriction enzyme cleavage maps of the DNA of two cauliflower mosaic virus isolates. Virol 103,250-254

5. Gardner RC and Shepherd RJ (1980) A procedure for rapid isolation and analysis of cauliflower mosaic virus DNA. Virol 106,150-161

6. Grunstein M and Hogness DS (1975) Colony hybridization: a method for the isolation of cloned DNAs that contain a specific gene. Proc Natl Acad Sci 72, 3961-3965

7. Holmes FO (1929) Local lesions in tobacco mosaic Bot Gaz 87, 39-55 8. Holmes FO (1931) Local lesions of mosaic in Nicotiana tabacum L. Contr. Boyce

Thompson 3,163-172 9. Howell SH, Walker LL and Dudley RK (1980) Cloned cauliflower mosaic virus

DNA infects turnips (Brassica rapa). Science 208, 1265-1267 10. Hull R, Shepherd RJ and Harvey JD (1976) Cauliflower mosaic virus: an improved

purification procedure and some properties of the virus particles. J Gen Virol 31, 93-100

11. Lebeurier G, Hirth L, Hohn T and Hohn B (1980) Infectivities of native and cloned DNA of cauliflower mosaic virus. Gene 12,139-146

12. Mackey JK, Brackmann KH, Green MR and Green M (1977) Preparation and characterization of highly radioactive in vitro labeled adenovirus DNA and DNA restriction fragments. Biochemistry 16, 4478-4483

13. Szeto WW, Hamer DH, Carlson PS and Thomas CA Jr (1977) Cloning of cauliflower mosaic virus DNA in Escherichia coli. Science 196,210-212

73

14.

15.

Tomlinson JA and Shepherd RJ (1978)Studies on mutagenesis and cross-protection of cauliflower mosaic virus. Ann appl Bio190,223-231 Ulrich TH (1980) Effect of liposome encapsulation of cauliflower mosaic vires and its DNA on plant infectivity and culture of turnip (Brassica rapa L) protoplasts, their uptake and expression of exogenous DNA. Ph D thesis. University of Illinois. 85 P.