Transgenic Plants Having Increased Freezing and Choline Tolerance
The present invention relates to a transgenic monocotyledonous or dicotyledonous plant having increased freezing and choline tolerances, a method of enhancing freezing and choline tolerances in such a plant as well as use of a transgenic monocotyledonous or dicotyledonous plant for the production of glycine betaine or analogues thereof.
Technical Background
Plants are frequently being exposed to environmental stress such as cold, drought or salinity. Temperatures below 0°C cause severe disturbances to metabolism and photosynthesis as well as physical damage to structural parts e.g. cell membranes and cell walls. Such cell injuries often originate from the loss of water from the cytoplasm and the membranes as a result of ice crystal formation. The cellular and metabolic responses in organisms exposed to low temperatures include increased levels of sugar or soluble proteins such as antifreeze proteins and altered lipid membrane compositions.
Another common cellular adaptation mechanism is the accumulation of osmotically active, low molecular weight, non-toxic molecules. Collectively referred to as compatible solutes, these compounds benefit stressed cells in two ways: by acting as a cytoplasmic osmolyte, thereby facilitating water uptake and retention, and by protecting and stabilising macromolecules from damage induced by low temperatures and high ion concentrations .
The quaternary ammonium compound glycine betaine, a compatible solute, has been shown to accumulate in response to freezing, water deficit or high ion concentrations in a number of prokaryotic and eukaryotic organisms [1-3] . Certain plant species have an endogenous production of glycine
Z betaine such as sugar beet and spinach [4,5], but other important crops such as potato, rice and sorghum produce no or very low amounts of glycine betaine.
Betaine levels are frequently measured by one-dimensional proton NMR [30] These determinations are sensitive to small pH changes which cause changes in the shift of the betaine peak in the NMR spectra between different samples. Moreover, in plants which often show a complex metabolite pattern, other compounds can frequently be superimposed on the betaine peak. To allow a secure separation between betaine and interfering compounds in plants, a two-dimensional NMR method as described in the examples was developed.
Glycine betaine biosynthetic pathways are similar in higher plants and microorganisms. In both cases, glycine betaine is derived from a two step oxidation of choline to glycine betaine via the unstable intermediate, glycine betaine aldehyde. The substrate, choline, is ubiquitous in higher plants and in the conversion of choline to glycine betaine aldehyde, spinach uses the ferredoxin dependent choline monooxygenase [6] , while E. coli uses the membrane bound choline dehydrogenase (CDH) [7,8]. The second step in the glycine betaine biosynthetic pathway is catalysed by the NAD+ dependent glycine betaine aldehyde dehydrogenase (BADH) , which exhibits strong similarities in plants and bacteria [8] . These include structural homologies and the ability of both sugar beet and bacterial BADH to produce glycine betaine from glycine betaine aldehyde in transgenic tobacco [9-11] .
Bacterial CDH is one of the most useful enzymes for introducing this pathway into new species, since it is able to catalyse not only the oxidation of choline to glycine betaine aldehyde, but also the second step to glycine betaine. Bacterial CDH is also independent of soluble cofactors [7,12], a favourable trait in order to avoid
interference with other metabolic pathways in a transgenic organism.
Description of the invention
Thus, in order to increase the stress tolerance, particularly freezing and choline tolerance, and crop yield of monocotyledons or dicotyledons, a biosynthetic pathway for glycine betaine or analogues thereof has now been introduced into the genome of such plants.
According to the invention there is provided a transgenic monocotyledonous or dicotyledonous plant having freezing and choline tolerances, which plant has been transformed by insertion into its genome of a gene encoding an enzyme that affects the biosynthetic pathway for glycine betaine or analogues thereof together with a promoter for expression of the gene in said plant .
There is also provided a method of enhancing freezing and choline tolerances in a monocotyledonous or dicotyledonous plant, whereby the intracellular concentration of glycine betaine or analogues thereof is increased by insertion into the plant genome of the gene encoding an enzyme that affects the biosynthetic pathway for glycine betaine or analogues thereof together with a promoter for expression of the gene in said plant.
A further aspect of the invention is the use of a transgenic monocotyledonous or dicotyledonous plant for the production of glycine betaine or analogues thereof, which plant in its genome comprises a gene encoding an enzyme affecting the biosynthetic pathway for glycine betaine or analogues thereof together with a promoter for the expression of the gene in said plant.
Preferred enzymes are choline dehydrogenase and choline oxidase or mutants thereof having choline dehydrogenase or choline oxidase enzymatic activity respecively, choline dehydrogenase being the most preferable.
For example, the E. coli choline dehydrogenase and corresponding gene is known from [12; GenBank accession number x52905) ] and the Sinorhizobium meliloti choline dehydrogenase and corresponding gene is known from [18; GenBank accession number u39940)] .
Examples of the choline oxidase from Arthrobacter globiformis and Arthrobacter pascens ATCC 13346 as well as the genes encoding said enzymes are known from [19; GenBank accession number x84895] and [20] , respectively. Other bacterial and fungal choline oxidase genes which can be used in the production of the transgenic plants according to the invention are described below in the examples.
A further preferred enzyme of the present invention is choline monooxygenase or mutants thereof having choline monooxygenase enzymatic activity. For example, choline monooxygenase from Spinacia oleracea and the corresponding gene is known from [26; GenBank accession number u85780] . Furthermore, the gene encoding the choline monooxygenase from Beta vulgaris is known (GenBank accession number af023132) .
The above mentioned genes can be cloned using standard techniques [22] and employed to transform and produce the transgenic plants of the present invention.
Plants that can be transformed in this way are any monocotyledonous or dicotyledonous plants. Examples are oil plants (e.g. soya, brassica, rape, etc) , sugar beet, cereals, grapes, corn, cotton, tomato, potato, rice, sorghum, tobacco, of which potato at present is preferred.
In a preferred embodiment of the invention, the E. coli betA gene encoding CDH has been introduced into potato (Solanum tuberosum Desiree) in order to enhance stress tolerance by intracellular accumulation of glycine betaine.
Brief Description of the Figures
Figure 1. Transcription analysis of the betA gene by detecting mRNA levels in plant leave homogenates . The mRNA was purified by oligo (dT) -cellulose chromatography and amplified with rTth polymerase yielding detectable amounts of cDNA. The cDNA corresponds to a 506 bp middle section of the betA gene, encoding CDH. Wild-type potato was used as a negative control.
Figure 2. A. BIRD-HMQC 1H{13C} NMR spectrum (500.1 MHz) of a transgenic Solanum extract (pH* 7.43, 298 K) ; tissue treated with 30 mM choline. Assignments: glycine betaine (S-404) , phosphocholine (S-403) , choline (S-402) ,proline-d (203b), n.i. (419) (referenced to DSS) .
B. Retransformed row (BIRD-HMQC) containing the glycine betaine resonance. C. An 14N NMR ID spectrum (36.1 MHz) of the extract. The expansion shows the trimethyl correlation region with glycine betaine (S-404) , choline, phosphocholine, and an unidentified trimethyl compound downfield of the choline signal (all referenced to nitrate at 0.00 ppm) .
Figure 3. Growth experiments of transgenic potato line CDH2 : 3 and a wild-type control in MS media supplemented with varying concentrations of choline, 0, 15 and 30 mM. Each data point corresponds to the mean value calculated from 9 plants and the bars correspond to the standard error.
Figure 4. Degree of freezing injuries of transgenic potato line CDH2:3 and a wild-type control. The freezing injuries were visually assessed two days after freezing for leaf
damages and after three weeks for ability to regenerate. Four on the scale corresponds to unaffected leaves or fully vital plant, and 0 corresponds to chlorosis of all leaves or a dead plant. A total of 40 plants of each line were planted and the freezing evaluations were performed in two independent experiments . The differences between the transgenic and wild- type line were significant at P<0.05. Two independently repeated experiments were performed.
Figure 5. Photo depicting the transgenic potato CDH2:3 and a wild-type control 4 weeks after being subjected to freezing (see Example 1) .
The invention is explained further in detail below in connection with a non-limiting examples related to transformation of potato and rice. From the data presented it is especially shown that: (i) the betA gene was correctly transcribed and expressed; (ii) bacterial CDH was enzymatically active and produced glycine betaine in potato and rice, and (iii) transgenic potato and rice expressing CDH exhibited enhanced freezing tolerance.
Example 1: Transgenic potato expressing a betA gene encoding choline dehydrogenase
The commonly cultivated Solanum tuberosum exhibits low resistance to frost. A system for expressing the betA gene in plants has been constructed and concomitantly transformed into potato utilising Agrobacterium tumefaciens mediated gene transfer.
Bacterial strains and plasmids
Escherichia coli strain TGI was cultivated and transformed according to Maniatis et al . [13]. Agrobacterium tumefaciens strain LBA 4404 was grown in LB medium and transformed with the aid of E. coli HB 101 harbouring the conjugative plasmid
pR 2013. Plant plasmid pCDH2 expressing the betA gene has previously been described [14] .
DNA sequencing and PCR analysis
The betA gene amplified by PCR from E. coli DNA and inserted into pUC 19 was sequenced on an ABI automatic sequencer using the terminator chemistry and two universal primers, pUC5 (5'-GTA AAA CGA CGG CCA GT-3') and pUC3 (5' -CAG GAA ACA GCT ATG AC-31), directed towards sequences upstream and downstream the multiple cloning site, respectively. Extraction of genomic potato DNA was performed as described by Edwards et al . [14] . PCR analyses of these samples were performed using primers 1 and 2 previously described by Lilius et al . [16] .
Plant material
The Solanum tuberosum (Desiree) plants were exposed to 16 h light at 25 °C and were either kept in soil or in sterile jars on MS media. The tobacco plants were transformed by Agrobacterium tumefaciens mediated gene transfer as described by Rogers et al . [17] . All experiments were performed using stable transformants of the F^ generation.
mRNA
Leaves from transgenic and wild-type plants grown for two weeks were used as source for extraction of mRNA. 100 mg leaf material from each plant was subjected to a Quickprep mRNA purification kit™ (Pharmacia) . 10 μg mRNA was extracted and parts of the betA transcript were amplified using rTth polymerase™ (Perkin Elmer) and the following primers: mRNAl 5'-CGC GTC CTA ACC TGA CCA TTC GTA CTC ACG-3 ' (SEQ ID N0:1) and mRNA2 5 ' -GCA TTC GAG CCG TTA TAG TTA ATC GC-3 ' (SEQ ID NO: 2) .
NMR Samples
Extracts for NMR analysis were prepared using the following procedure. Tissue of leaves and stem were frozen in liquid N2, perchloric acid was added, and the samples were crunched in a pre-chilled mortar. After thawing and further grinding, the mixture was centrifuged and the supernatant was neutralised using a concentrated KOH solution. After 30 minutes on ice the mixture was centrifuged, and the supernatant was treated with Chelex-100 (BioRad) at neutral pH to remove paramagnetic impurities. The sample was centrifuged again and the supernatant was freeze dried over night. The extracts were dissolved in D2O and 1 μl of a chemical shift reference standard was added (DSS or 2,2- dimethyl-2-silapentane-5-sulfonate, 112 mM in stock) . Finally a stock solution of inorganic phosphate was added to give a final concentration of 20 mM and the pH* (uncorrected reading in D2O) was adjusted to 7.40.
NMR spectroscopy
1H{13C} and 14N NMR spectroscopy (Bruker AMX2-500, 5 mm triple nucleus inverse probe) were used to identify and quantify the metabolites in the extracts. The 1H{13C} heteronuclear multiple quantum coherence NMR spectra with BIRD filtering (BIRD-HMQC) were acquired (at 500.13 MHz and 125.77 MHz) using TPPI and presaturation of water (Is presaturation) , using the spectral widths 5000 Hz (^H) , and 18000 Hz (13C) , at 298 K. The number of scans was 32 to 64, with 16 dummy scans once in the sequence, and 1024 points were used in 512 spectra. 2048 by 2048 points, and a cosine window was used for the data processing. The 1H 90° pulse width was c. 10.0 μs and the 13C 90° pulses were 8.5 μs .
GARP decoupling as used for 13C decoupling during the acquisition (using a pulse width of 65 μs) . The optimized BIRD water suppression delay used was 400 ms, and the delays
in the BIRD sequence were calculated using a scalar coupling constant of 144 Hz . Selected rows from the BIRD-HMQC spectra were extracted and retransformed to improve the digitization. The chemical shift reference was DSS (assigned to 0.00 ppm for and-1.8 ppm for 13C) . 14N NMR spectra with WALTZ-16 proton decoupling were acquired at 36.129 MHz using a spectral width of 7300 Hz and 1 Hz line broadening. The time between acquisitions was 1 s (using 12K data points) , and 57600 acquisitions were collected. The chemical shift reference in the 14N NMR spectra was nitrate, assigned to 0.00 ppm.
Evaluation of the choline tolerance
To assess the choline tolerance, potato seedlings of transgenic fine CDH2 : 3 and a wild-type line were planted in sterile containers containing 100 mL MS-medium, solidified by Phytagel™ (Sigma) , and a range of different choline concentrations (0 to 30 mM) . The seedlings were grown for 60 days before harvest. The fresh weights were determined and the metabolites were extracted for NMR analyses.
Evaluation of the freezing tolerance
Potato plants of transgenic line CDH2 : 3 and a wild-type line were planted in boxes containing soil and grown in a greenhouse at 20°C until roots were developed, approximately one week. At this stage the plants had reached a height of 7 cm. The boxes containing plants were placed in the freezing chamber at 2°C and the temperature was lowered by 1°C per hour to -4°C. This temperature was kept for 5 hours and was then raised by 1°C per hour from -4 to 2°C. The seedlings were again placed in a greenhouse at 20°C. The injuries of the leaves were visually assessed after two days according to a scale from 0 to 4 where 0 indicates that all leaves are affected by chlorosis, and 4 that the leaves are unaffected by the freezing procedure.
Furthermore, the ability of the seedling to recuperate from the freezing injuries were also investigated by visually assessing the vitality of the plants according to a scale from 0 to 4 two weeks after freezing of the potato plants. 0 corresponds to a completely dead plant and 4 to a plant of normal height without any signs of injuries.
Transcription analysis of the transgenic potato
The plant expression vector for betA , pCDH2 , was used to transform potato [15] . DNA was extracted from 15 randomly chosen kanamycin resistant transgenic potatoes . PCR was utilised as the initial screening method to select betA positive plants using primers mRNAl and CDH2. Approximately 90% of the transformed plants gave rise to the 1700 base pairs full-length betA fragment (data not shown) .
In order to select plants expressing high levels of CDH the transgenic plants were grown in MS media supplemented with 50 mM choline. Under these selective conditions only the transgenic potato line CDH2 : 3 was able to form roots and hence was chosen for further investigations . PolyA"1" mRNA was isolated from leaves of this potato line. Double stranded cDNA was obtained by amplifying the purified mRNA using rTth polymerase and two strongly hybridising polynucleotide primers (mRNA and 2) . The primers were directed towards a 506 base pairs fragment of the betA gene. As shown in Figure 1, the transgenic potato line CDH2 : 3 gave rise to the desired cDNA fragment, whereas the control lane containing wild-type potato was empty.
Determination of the glycine betaine concentrations using NMR spectroscopy
As shown in Figure 2A, the glycine betaine resonance is denoted S-404, choline S-402, phosphocholine S-403, and
proline 203b in the 1H{13C} BIRD-HMQC NMR spectra. The volume under the peaks were integrated using the Bruker program 'uxnmr' (940501) and compared with a standard sample to calculate the concentrations (reference containing c. 1.20 mM of glycine betaine, etc.; 130 mM KC1 ; 20 mM inorganic phosphate; 1 mM DSS; pH* 7.40) . The maximal glycine betaine levels in the transgenic and wild-type potato were 95 ± 4 and 2 ± 1 μmol/kg FW, respectively, both obtained when grown in soil .
Furthermore, the 1H chemical shift of the row containing the glycine betaine peak is displayed in Figure 2B.
14N NMR spectroscopy was used to verify the presence of glycine betaine in the extracts (Figure 2C) .
Evaluation of the choline tolerance in growth experiments
Wild-type and transgenic potato grew equally well in MS medium without added choline. After increasing the choline concentration to 15 mM, the growth was severely retarded, particularly of the wild-type potato. As shown in Figure 3, the wild-type potato exhibited a 50% decrease in fresh weight whereas the transgenic potato only displayed a 15% decrease. However, after increasing the choline concentration to 30 mM the differences in fresh weight were not significant.
Evaluation of the freezing tolerance
Evaluation of the freezing tolerance was performed in a freezing chamber according to the protocol described above. The degree of leaves affected by chlorocis after freezing was 27% lower in the transgenic potato compared to the wild-type when evaluated two days after freezing (Figure 4) . Moreover, the vitality two weeks after freezing was also evaluated. As indicated in Figure 4, the transgenic plants were 35% more vital than the wild-type controls. In Figure 5, the potato
plants are shown four weeks after subjecting them to freezing. Seven out of ten of the transgenic plants survived the freezing procedure whereas only 2 wild-type plants survived.
It has now been demonstrated that by increasing the intracellular glycine betaine concentration, enhanced freezing tolerance can be achieved. Glycine betaine is accumulated in winter and spring barley as a response to cold temperatures in concentrations up to 30 and 80 μmol/g dry weight, respectively [1] . The maximal glycine betaine concentration achieved by expressing CDH in transgenic potato according to the invention was 95 μmol/kg fresh weight (approximately corresponding to 5 μmol/g dry weight) which is a 50-fold higher than values obtained in wild-type potato grown under the same conditions. In contrast to most other glycine betaine measurements performed, these values were obtained by extracting the metabolites from whole plants, i.e. both stem and leaf material. Furthermore, since the cytoplasmic space only occupies 5% of the total cell volume, it is plausible that the effective glycine betaine concentration is approximately 20 times higher.
CDH is one out of a couple of enzyme systems which can be utilised for genetically introducing a glycine betaine biosynthesis pathway into plants. Other enzymes, such as choline oxidase from Arthrobacter, are also contemplated to further increase the intracellular glycine betaine concentration and enhance the freezing tolerance.
Example 2 : Transgenic rice expressing a betA gene encoding choline dehydrogenase
This example illustrates the transformation of rice using particle bombardment with a polynucleotide that directs the expression of E. coli choline dehydrogenase in plant cells.
Plasmid construction
Plasmid pLB4 was constructed as a derivative of pUC that contains the betA gene under the regulation of the CaMV 35S promoter and a nos polyadenylation sequence was constructed. Additionally, the stop codon of the betA gene was altered using PCR mediated mutagenesis from TAA to TGA (a more frequently used stop codon in plants) . The betA gene was isolated from E. coli according to Example 1. The PCR primers were designed with homology to the 5'- and 3 ' -end of the CDH gene except for the start codon which was altered from TTG to ATG. The 5 '-end contained an Xbal site and the 3 ' -end a Sad site. For the selection of transgenic rice tissue, the aph IV gene from E. coli which confers resistance to hygromycin was included in the transformation construct. This marker gene was under the control of the CaMV 35S promoter and the CaMV polyadenylation sequence.
A DNA fragment containing the betA gene and the hygromycin resistance gene (and lacking the ampicillin resistance gene) was isolated using known methods for use in posterior bombardment of plant embryos .
Rice transformation
Immature Oryza sativa embryos of the Japonica variety Taipei 309 were aseptically isolated 10-14 days after pollination from greenhouse plants and plated scutulum site up on solid MS medium [16] containing 3% sucrose, 2 mg/1 2,4-dichlorophenoxyacetic acid (2,4-D) and 50 mg/1 cefotaxime (MSI) . After 4-6 days (28°C, darkness) embryos were transferred to solid MS medium containing 10% sucrose, 2 mg/1 2,4-D and 50 mg/1 cefotaxime (MS2) and subjected within 1 hour to microprojectile bombardment with a particle inflow gun [22]. The DNA
fragment containing the betA and aph IV genes (5 μg) was precipitated on 1-3 mm gold particles (Aldrich) as described previously [23]. Gold particles (400 mg per bombardment) were accelerated to the target with a particle inflow gun [24] at a pressure of 6 bar. Embryos were placed 16 cm below the syringe filter. Twenty-four hours post-bombardment embryos were subjected to selection on solid media (MSI medium containing 20 mg/1 hygromycin B) and incubated at 28°C in the dark.
After one week embryos were transferred to a liquid selection media, R2 medium [25] supplemented with: 3% sucrose, 1 mg/1 thiamine, 1 mg/1 2,4-D, 50 mg/1 cefotaxime and 20 mg/1 hygromycin B. The embryos were incubated with shaking at 28°C in the dark and subcultured weekly. Developing calli were isolated 3 to 6 weeks later, and transferred to a callus increasing media (R2 medium supplemented with: 6% sucrose, MS vitamins, 100 mg/1 inositol, 2 mg/1 2,4- D, 50 mg/1 cefotaxime and 20 mg/1 hygromycin B) . The calli were incubated in this media at 28°C in the dark and subcultured weekly.
Resistant calli were transferred to solid R2 regeneration media supplemented with 2% sucrose, 3% sorbitol, 20 mg/l hygromycin B, 1 mg/1 zeatin, 0.5 mg/1 indole-3 -acetic acid (IAA) , MS vitamins and 0.65% agarose . The callus tissue was maintained at 28°C with 12 h of light in order to enhance shoot formation. The calli were then subcultured every 3 weeks until shoots had reached a length of 2-3 cm. They were transferred to half-strength MS rooting medium without hormones, supplemented with 1.5% sucrose and 0.3% gelrite (Sigma) . After 2-4 weeks of cultivation, plantlets were transferred directly to the greenhouse and planted in soil. Plantlets were grown in 7 liter aquaculture pots with fertilizer enriched earth, 3 plants per pot, (day: 12
h, 28°C, 80% humidity; night: 12h, 21°C, 60% humidity) until they flowered and set seeds.
Analysis of the transgenic rice
To check the presence of the transgene, complexity of insertion (s) and number of copies present, Southern blot analysis was performed A PCR-amplified, DIG-labeled (Boehringer) 300-bp fragment of the coding region of the betA gene was used as a probe.
Enzymatic measurements of choline dehydrogenase activity were performed according to [7] and demonstrated that the transformed plants expressed the betA transgene.
Seeds from the RQ plants were collected and R^ plants were grown. Exposure to subzero temperatures as described above of the Rj_ generation plants showed a higher freezing and choline tolerance as compared to control plants without the transgene .
Example 3 Transgenic plants expressing a gene encoding choline oxidase
A gene encoding choline oxidase can be employed in a manner similar to that in the above described protocols using the choline dehydrogenase gene in order to arrive at the transgenic plants of the present invention.
For example, the nucleotide sequence of Arthrobacter globiformis choline oxidase is known [19]. The gene was cloned using the published sequence (GenBank accession number x84895) and PCR using the following oligonucleotides:
5 ' -GACCACGGCACGGAGAACTAGATGCACATCGACAACATCGAGAAC-3 ' (5 ' -end oligonucleotide) and
\6
5 ' -TGGCCGATATCGCTTAGGCGAGGGCCGCGCTCAGCTCGGC-3 ' (3' -end oligonucleotide).
The primers are homologous to the 5' -and 3' -end of the choline oxidase gene. The 5' -end oligonucleotide hybridizes to position 339-385 and the 3 ' -end oligonucleotide hybridizes to position 1968-2017, when the nucleotide sequence numbering in GenBank sequence databank (accession number x84895) is used. The primer hybridizing to 5' -end contains an extra Xbal restriction site and the 3 ' -end primer an EcoR V site.
Chromosomal DNA was isolated using the following procedure. The Arthrobacter globiformis strain ATCC 4336 was grown as described previously [21]. The cells were disrupted by French press at 150 Mpa in a 10 mM Tris-HCl buffer containing 100 mM NaCl and 1 mM EDTA. The DNA was purified using standard CsCl- ethidium bromide centrifugation [22]. The purified DNA was used as the template for PCR reactions. The amplification was performed in following conditions: 30 cycles of 1 min at 94°C for denaturation, 3.5 min at 32 °C annealing and 2.5 min at 72°C for synthesis. The amplified DNA was cloned into pUC 19 vector by using the Xbal and EcoR V restriction sites in the primers. Sequencing and restriction analysis confirmed that the cloned fragment contained the choline oxidase gene.
Said cloned choline oxidase fragment is then employed as described in Example 1 or 2 to arrive at transgenic plants of the present invention.
Genes encoding choline oxidase can also be isolated other bacterial and fungal sources and employed to construct the transgenic plants of the present invention.
For example, the choline oxidase gene from Arthrobacter pascens (for example, A. pascens ATCC 13346) and Arthrobacter aurescens (for example, A. aurescens ATCC
t?
13344) can be obtained by culturing these organisms by using the growth conditions recommended by the ATCC or in the growth medium described in [21] .
Additionally, Brevibacterium album (for example, FERM-P 3777, NRRL B-11,046 or ATCC 15111 or FERM-P 3778, NRRL B- 11047 or ATCC 15112) or Corynebacterium murisepticum (for example, FERM-P 3779, NRRL B-11049 or ATCC 21374) can also be used to isolate a gene encoding for choline oxidase. Cultivation of these organisms is performed as described in [27] .
Furthermore, Alcaligenes sp . such as Alcaligenes sp . FERM P- 4105 can also be used as a source for a gene encoding the choline oxidase gene. These bacteria were grown using a specific alcoholic compound as the inducer under conditions described in [28] .
Chromosomal DNA isolation from bacteria
Chromosomal DNA can be isolated from the above mentioned bacteria by using standard isolation procedures optimized for bacteria [29] . The cells may alternatively be lysed by lysozyme treatment followed by freezing and thawing the samples at least five times [20] . The chromosomal DNA was further purified by CsCl density gradient centrifugation according to published protocols [22] .
In addition to the above mentioned bacterial strains, the following fungi have been shown to express a choline oxidase enzyme which can be employed to obtain the transgenic plants of the present invention:
Fusarium (for example, Fusarium oxysporum (AKU 3702) ; Fusarium anguioides (AKU 3703) ,- Fusarium solani (AKU 3704) ; Fusarium bulbigenum (AKU 3705) , and Fusarium caucasicum (AKU 3707) ) .
Additional choline oxidase genes which can be used according to the invention include the fungal choline oxidase genes from Gibberella ( for example, Gibberella fujikuroi (AKU 3802, AKU 3804) and Cylindrocarpon (for example, Cylindrocarpon didynum M-l [31] .
These fungal strains were grown on media suitable for cultivation of fungi, e.g. potato dextrose broth (Difco) or bacto potato malt broth (Difco) at 26-30°C in shake flasks rotating at 200 rpm. The cultivation time was 5-7 days.
Preparation of cDNA from fungi
Mycelium obtained from cultivation was frozen and ground into fine powder in liquid nitrogen. Total RNA was isolated according to [32] . The poly (A) + RNA fraction was isolated by oligo(dT) -cellulose chromatography [33] and purified by sucrose gradient centrifugation (5-20 % sucrose, 22,000 rev. /min, 16 h, Kontron TST 28.17 rotor) . The cDNA synthesis was carried out using a cDNA synthesis kit (Stratagene, USA) .
Construction of the libraries from the organisms
The same methodology was used for constructing the libraries from all organisms listed above.
The DNA isolated from the organisms (or cDNA synthesized from fungal RNA) was partially digested with a restriction enzyme (e.g. EcoRI) . Different restriction enzymes have to be tested for the DNAs from different organisms in order to obtain even DNA particle size distibution. Size fractionation of the cDNA was performed by Sepharose CL-4B chromatography (Pharmacia) to remove the small molecular weight fragments [29] . The digested chromosomal DNA was size fractionated using sucrose gradient centrifugation [22] . The
DNA was ligated to Lambda ZapII vector and packaged into Lambda particles using Lambda ZAPII cloning kit (Stratagene, San Diego, USA) according to lhe recommendations provided by the manufacturer.
Screening of the libraries
Hetereologous hybridization protocol was used for screening the libraries. Plaques of the gene bank were transferred onto a nitrocellulose membrane and hybridized with a PCR fragment containing the entire A. globiformis choline oxidase gene. The probe was prepared by PCR using the primers designed for cloning of the gene (see Example 3 and SEQ ID NO: 3 and 4) as the template. The PCR fragments were radioactively labelled with rediprime DNA labelling system (Amersham, UK) . Hybridization was carried out at 42 °C in 50 mM Tris-HCl, pH 7.5, 10 mM EDTA, 1 M NaCl, 05.% SDS, 0.1% sodium pyrophosphate, 10 x Denhardt's, 100 μg herring sperm DNA and 125 μg/ml polyA. The filters were washed first with low stringency conditions at 37°C using 3x SSC, 0.5% SDS and 10% sodium pyrophosphate [22] . The filters were exposed on X-ray film. The washing temperature was then raised in 5°C intervals up to temperature in which the background became invisible and only few positive plaques were obtained. The positive plaques were purified and the DNA was isolated for Southern hybridization to check the size of the cloned fragment. Positive lambda clones were cored and excised with EXAssist helper phage (Stratagene) to obtain phagemids . The phagemids obtained were transformed to SOLR E. coli host cells (Stratagene) and plasmid DNA was purified with Qiagen Plasmid Kit and used in the analysis of insert DNA and DNA sequencing.
The ORF of the novel choline oxidase genes was identified on the basis of sequence homology with the A. globiformis gene.
Expression constructions were made as described in the Example 3 using PCR.
Example 4. Transgenic plants expressing a gene encoding choline monooxygenase
The nucleotide sequences of two choline monooxygenase genes are known. The gene has been isolated from Spinacia oleracea [26, GenBank accession number u85780] and from Beta vulgaris [GenBank accession number af023123] . The published sequences were used for cloning of the genes from the cDNA by PCR.
Total RNA was isolated from leaves of spinach or sugar beet plants salinized in 150-200 mM NaCl [34] as described [35]. An additional carbodydrate precipitation step was added to the protocol as described earlier [26]. The poly (A) + RNA fraction was isolated by oligo (dT) -cellulose chromatography [33]. The cDNA synthesis was carried out using a cDNA synthesis kit (Stratagene, USA) . The cDNA was used as the template for the PCR reactions.
Following oligonucleotides were used for the cloning of the Spinacia oleracea choline monooxygenase gene:
5 ' -ACA AAA AGG AAG TGT TGA GCT CGT TAA TGA TGG CAG CAA GCG CAA GCG-3' (5 '-end oligonucleotide)
5'-GAA AAT GCA CGG GAT ATC TGA GGT AC-3' (3 ' -end oligonucleotide) .
The primers are homologous to the 5 '-and 3 ' -end of the choline monooxygenase gene. The 5 '-end oligonucleotide hybridizes to position 31-78 and the 3 '-end oligonucleotide hybridizes to position 1501-1526, when the nucleotide sequence numbering in GenBank sequence databank (accession number sou858780) is used. The primer hybridizing to 5 ' -end
contains a new Sad restriction site. The 3 ' -end primer contains a EcoRV site.
The amplification was performed under identical conditions as described in Example 3. The amplified DNA was cloned into pUC 19 vector by using the Sad and EcoRV restriction sites. Sequencing and restriction analysis confirmed that the cloned fragment contained the choline monooxygenase gene.
Th following oligonucleotides were used for the cloning of the Beta vulgaris choline monooxygenase gene:
5' -CGG CAC GAG AAA ATC GAT TAA CAA TGG CAG CAA GTG CTA CAA CC-3' (5 '-end oligonucleotide)
5' -GGA GCA TAG GCT GAT ATC ACT GCA AAG TTT CAT GCA ACC AAC-3' (3 '-end oligonucleotide)
The 5 '-end oligonucleotide hybridizes to position 14-57 and the 3' -end oligonucleotide hybridizes to position 1352-1453, when the nucleotide sequence numbering in GenBank sequence databank (accession number af023132) is used. The primer hybridizing to 5 '-end contains a new Clal restriction site and the 3 ' -end primer a EcoRV site. The amplification was performed under identical conditions as described in Example 3. The amplified DNA was ligated into pUC19 vector by using the Clal and EcoRV restriction sites. Sequencing and restriction analysis confirmed that the cloned fragment contained the choline monooxygenase gene.
Said cloned choline monooxygenase fragments are then employed as described in Example 1 or 2 to arrive at transgenic plants of the present invention.
Two major functions of glycine betaine in regard to water stress protection have been proposed. The ability of glycine betaine to adjust the osmotic potential plays a large role
when the accumulated intracellular glycine betaine concentration is high. A second function where glycine betaine, at low concentrations, protects and stabilises membranes and macromolecules has been proposed by Coughlan and Heber [17] after studying the cryoprotective properties of glycine betaine in a spinach thylakoid membrane model system. The glycine betaine molecules can interact with carboxyl groups of membrane proteins and protect the membrane and/or stabilise the surrounding water layer. It is plausible to assume that these cryoprotective properties of glycine betaine also protect the cell membranes in vivo and is, at least partially, responsible for the enhanced freezing tolerance displayed by the transgenic potatoes . The low glycine betaine concentrations obtained in the present transgenes suggest only minor effects on the cytoplasmic water retention capacity.
REFERENCES
[1] Kishitani, S., Watanabe, K., Yasuda, S., Arakawa, K. and Takabe, T. (1994), Plant, Cell and Environment 17, 89- 95.
[2] Hanson, A.D. and Burnet, M. (1994) in: Biochemical and Cellular Mechanisms of Stress Tolerance in Plants, Vol. H 86, pp. 291-302 (Cherry, J.H. , Ed.) Springer Verlag (Berlin, Heidelberg)
[3] Rhodes, D. and Hanson, A.D. (1993) Anna. Rev. Physiol. Plant Mol. Biol. 44, 357-384.
[4] Weretilnyk, E., Bednarek, S., McCue, K.F. Rhodes D. and Hanson, A. (1989) , Plants 178, 342-352.
[5] McCue, K. and Hanson, A. (1992) Aust. J. Plant Physiol. 19, 555-564.
[6] Brouquisse, R. , Weigel, P., Rhodes, D., Yocum, S.F. and Hanson, A. (1989), Plant Physiol. 90, 322-329.
[7] Landfald, B. and Strom, A. (1986) J. Bacteriol . 165, 849-855.
[8] Lamark, T., Rokenes, T.P., McDougall, J. and Strom, A.R. (1995) J. Bacteriol. 178, 1655-1662.
[9] Boyd, L., Adam, L., Pelcher, L.E., McHugen, A., Hirji, R. and Selvaraj , G. (1991) Gene 103, 45-52.
[10] Rathinasabapathi, B., McCue, K.F., Gage, D.A. and Hanson, A. (1994) Planta 193, 155-162.
[11] Holmstrόm, K.-O., Welin, B., Mandal, A.,
Kristiansdottir, I., Teeri, T., Lamark, T., Strom, A.R. and Palva, E.T. (1994) Plant J. 6, 749-758.
[12] Lamark, T., Kassen, I., Eshoo, M.W. , Falkenberg, P.,
McDougall, J. and Strom, A.R. (1991) Mol. Microbiol . 5, 1049-1064.
[13] Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
[14] Edwards, K. , Johnstone, C. and Thompson, C. (1991) Nucleic Acids Res. 19, 349.
[15] Lilius, G., Holmberg, N. and Bulow, L. (1996) Bio/Technology 14, 177-180.
[16] Murashige, T. and Skoog, F. (1962) Physiol. Plant 15, 473-497.
[17] Coughlan, S. and Heber, U. (1982) Planta 156, 62-69.
[18] Pocard, J.-A. et al . (1997) Microbiology 143, 1369- 1379.
[19] Deshnium, P., Los, D.A., Hayashi, H. , Mustardy, L.,
Murata, N. (1995) Transformation of Synechococcus with a gene for choline oxidase enhances tolerance to salt stress. Plant. Mol. Biol. 29, 897-907.
[20] Razwadowski, K.L. et al . (1991) J. Bacteriol. 173(2), 472-478.
[21] Ikuta, S., Matuura, K., Misaki, H. , Horiuti, Y. (1977) Oxidative pathway of choline to betaine in the soluble fraction prepared by Arthrobacter globiformis . J. Biol. Chem. 82, 157-163.
[22] Sambrook, J. , Fritsch, E.F., Maniatis, T. (1989)
Molecular cloning. A laboratory manual. Cold Spring Harbor Laboratory, NY.
[23] Vain et al . (1993) Plant Cell Rep. 12, 84-88.
[24] Finer et al . (1992) Plant Cell Rep. 11 :323-328.
[25] Ohira et al . (1973) Plant Cell Physiol. 14, 113-121.
[26] Rathinasabapathi, B. et al . (1997) Proc. Natl. Acad. Sci. (USA) 94(7), 3454-3458.
[27] US 4,245,050.
[28] JP 87,190,078.
[29] Ausubel et al . (1994) Current Protocols in Molecular Biology. Vol. 1. Wiley and Sons Inc. USA) .
[30] Wall, J.S. et al . (1960) Anal. Chem. 32, 8760-877.
[31] Tani, Y. , Mori, N. , Ogata, K. and Yamada H. (1979) Agrig Biol. Chem. 43, 805-820.
[32] Chirgwin, J., Przybyla A., MacDonal, R.J. and Rutter, W.J. (1979) Biochem J. 18, 5294-5299.
[33] Aviv and Leder (1972) Proc. Natl. Acad. Sci. USA 69, 1408-1412.
[34] Burnet, M. , Lafontaine, P.J., Hanson A.D. (1995) Plant Physiol. 108, 581-588.
[35] Hall, T.C., Ma, Y. , Buchbinder, B.U. Pyne, J.W., Sun, S.M. Bliss, F.A. (1978) Proc. Natl. Acad. Sci. USA 75, 3196-3200.
SEQUENCE LISTING
(1) GENERAL INFORMATION:
(i) APPLICANT:
(A) NAME: Leif Buelo
(B) STREET: Borgaslingan 6
(C) CITY: Lund
(E) COUNTRY: Sweden
(F) POSTAL CODE (ZIP) : 22472
(ii) TITLE OF INVENTION: Transgenic plants having increased freezing and choline tolerance
(iii) NUMBER OF SEQUENCES: 8
(iv) COMPUTER READABLE FORM:
(A) MEDIUM TYPE: Floppy disk
(B) COMPUTER: IBM PC compatible
(C) OPERATING SYSTEM: PC-DOS/MS-DOS
(D) SOFTWARE: Patentln Release #1.0, Version #1.30 (EPO)
(2) INFORMATION FOR SEQ ID NO: 1:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 30 base pairs
(B) TYPE: nucleic acid
(C) STRANDEDNESS: single
(D) TOPOLOGY: linear
(ii) MOLECULE TYPE: other nucleic acid
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Escherichia coli
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 1: CGCGTCCTAA CCTGACCATT CGTACTCACG 30
(2) INFORMATION FOR SEQ ID NO: 2:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 26 base pairs
(B) TYPE: nucleic acid
(C) STRANDEDNESS: single
(D) TOPOLOGY: linear
(ii) MOLECULE TYPE: other nucleic acid
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Escherichia coli
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 2: GCATTCGAGC CGTTATAGTT AATCGC 26
(2) INFORMATION FOR SEQ ID NO: 3:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 45 base pairs
(B) TYPE: nucleic acid
(C) STRANDEDNESS: single
(D) TOPOLOGY: linear
(ii) MOLECULE TYPE: other nucleic acid
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Arthrobacter globiformis
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 3: GACCACGGCA CGGAGAACTA GATGCACATC GACAACATCG AGAAC 45 (2) INFORMATION FOR SEQ ID NO: 4:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 40 base pairs
(B) TYPE: nucleic acid
(C) STRANDEDNESS: single
(D) TOPOLOGY: linear
(ii) MOLECULE TYPE: other nucleic acid
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Arthrobacter globiformis
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 4: TGGCCGATAT CGCTTAGGCG AGGGCCGCGC TCAGCTCGGC 40
(2) INFORMATION FOR SEQ ID NO: 5:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 48 base pairs
(B) TYPE: nucleic acid
(C) STRANDEDNESS: single
(D) TOPOLOGY: linear
(ii) MOLECULE TYPE: other nucleic acid
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Spinacia oleracea
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 5:
ACAAAAAGGA AGTGTTGAGC TCGTTAATGA TGGCAGCAAG CGCAAGCG 48
(2) INFORMATION FOR SEQ ID NO: 6:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 26 base pairs
(B) TYPE: nucleic acid
(C) STRANDEDNESS: single
(D) TOPOLOGY: linear
(ii) MOLECULE TYPE: other nucleic acid
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Spinacia oleracea
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 6: GAAAATGCAC GGGATATCTG AGGTAC 26
(2) INFORMATION FOR SEQ ID NO: 7:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 44 base pairs
(B) TYPE: nucleic acid
(C) STRANDEDNESS: single
(D) TOPOLOGY: linear
(ii) MOLECULE TYPE: other nucleic acid
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Beta vulgaris
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 7:
CGGCACGAGA AAATCGATTA ACAATGGCAG CAAGTGCTAC AACC 44
(2) INFORMATION FOR SEQ ID NO: 8:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 42 base pairs
(B) TYPE: nucleic acid
(C) STRANDEDNESS: single
(D) TOPOLOGY: linear
(ii) MOLECULE TYPE: other nucleic acid
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Beta vulgaris
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 8:
GGAGCATAGG CTGATATCAC TGCAAAGTTT CATGCAACCA AC 42