46. A simple screening method for transgenic rice tissue based on PCR

A. MAYER, G. B. ZONDAG and L.A.M. HENSGENS

MOLBAS Research Group, Department of Plant Molecular Biology, Leiden University, Wassenaarseweg 64, 2333AL Leiden, The Netherlands

Transformation of rice, like other cereals, is a difficult task (see for a review: Potrykus 1990). Most methods, such as protoplast electroporation (Shimamoto et al. 1989), polyethylene glycol (PEG) treatment (Zhang and Wu 1988; Hayashimoto et al. 1990) and particle bombardment involve much tissue culture work.

Methods for detecting transformed tissue employ phenotypic expression of resistance or reporter genes or are based on Southern analysis of purified rice DNA. The expression of the gene and tight correlation with physical data is considered to be proof of the transformed state (e.g. Potrykus) of the tissue or whole plant. Since expression of many genes (including chimeric) is under developmental control, actually transformed tissue may be considered as nontransgenic and disposed of (Benfey and Chua 1989; Meijer et al. 1991).

Most vectors used for transformation of rice harbor the hygromycin phosphotransferase (HPT) gene for selection and the beta-D-glucuronidase reporter gene. We have optimized the Polymerase Chain Reaction (PCR) method using three primer pairs in one reaction. By this multiplex PCR analysis we were able to detect an HPT-gene, a GUS-gene fragment and a fragment derived from the single copy gene Gos5 (de Pater et al. 1990). The latter was chosen as an internal control for the quality of the PCR analysis.

To minimize labour, an easy and siniple method for DNA isolation was developed.

Primer selection:

With help of the OLIGOTM program (copyright 1988-1990 Wojciech Rychlik) primer-pairs with high melting temperatures (T\m\), minimal self-complementarity and high specificity to the target DNA were selected. Tm and lenghts of the three primer pairs were chosen so that multiplex PCR was possible (see Table 1). The lengths of the expected GUS-, HPT- and Gos5-fragments were 683, 419 and 231 bp, respectively.

Using CsCl purified plasmid (construct HH271, containing the GUS-and HPT-gene) and genomic DNA isolated from rice tissue transgenic with HH271 (Meijer et al. 1991) and a genomic lambda clone carrying the Gos5-gene (de Pater et al. 1990), PCR amplification was optimized for the three primer sets seperately. When three individual primer pairs were annealed at 72 deg C, the background was minimal (data not shown).

Different dilutions of plasmid and phage DNA were used to optimize the concentrations of MgCl\2\, Tween 20, BSA, dNTP's, primers and the amount of AmpliTaq. With the optimized protocol, one femtogram of plasmid or phage

Table 1 - Primer sequences for detection of the GUS-, HPT- and Gos 5-gene. The position within the gene after the AUG start codon is given in nucleotides. Melting temperatures T\m\(deg C) were calculated using a primer concentration of 250 pM and a salt concentration of 50 mM. 

primer      primer sequence                       position       T\m\
                                                within gene     (deg C)
                                                 after AUG


GUS-upper   5'CAGCGAAGAGGCAGTCAACGGGGAA 3'           1124         67.7

GUS-lower   5'CATTGTTTGCCTCCCTGCTGCGGTT 3'           1807         67.4

HPT-upper   5'CGCACAATCCCACTATCCTTCGCAA 3'            -94         64.6

HPT-lower   5'GGCAGTTCGGTTTCAGGCAGGTCTT 3'            325         64.3

Gos5-upper  5'CGACCTCGAGGACATCGGCAACACC 3'            292         68.0

Ges5-lower  5'GCCGAGCAGCAGGAACTTGAGCAGG 3'            523         67.7
DNA was sufficient to obtain visible bands (data not shown). Hence a mixture of all 3 primer pairs was used in a multiplex PCR to study whether amplification of the three fragments could be performed in a single tube. Different quantities (10, 1 and 0.1 ng) of genomic DNA from HH271 transformed rice (cv. Taipei 309) cell suspension were tested (see Fig. 1). 

Optimized PCR-protocol:

                 100 uM     ddATP, ddCTP, ddGTP, ddTTP
                   2 mM     MgCl\2\
                  25 mM     KCl
                  20 mM     TRIS-HCl pH 8.3
                   0.5%     Tween 20 (Sigma) deionized, filter sterilized
                   0.05 mg  acetylated BSA, Promega, USA
                  10 uM     primers GUS-upper, GUS-lower
                  10 uM     primers HPT-upper, HPT-lower
                   2.5 uM   primers Gos5-upper, Gos5-lower
                   2 ul     DNA
                   x ul     H\2\O up to a volume of 90 ul at 85 deg C: 1 unit
                            AmpliTaqTM (Perkin Elmer Cetus)in buffer (20 mM
                            TRIS-HCl pH 8.3, 25 mM  KCl, 0.05 mg acetylated
                            BSA, 0.5% Tween 20, H\2\O up to a volume of 10 ul),
                            reaction volume 100ul overlaid with 70ul mineral
                            oil

PCR-program: (run in PREMTM III, LEM Scientific)
                    20 min. 98 deg C
                    cooling to 85 deg C
                    addition of 1 u AmpliTaq per reaction
                    45 cycles:  1 min. 93 deg C
                                2 min. 72 deg C
                                5 min. 72 deg C

Fig. 1. Influence of addition of 1 u AmpliTaq enzyme before (lanes B) and after (lanes A) the first denaturation step. The numbers above the slots refer to the amount of genomic rice (cv. T309) DNA, which was isolated from 3-5 g cell suspension transgenic for plasmid HH271 according to Hensgens and van OsRuygrok (1989). 45 PCR cycles were performed as following: One min. 93 deg C and 2 min. 72 deg C. Concentration of all 6 primers was 10 uM. Ten ul of the PCR reaction (100 ul) were loaded on a 2% agarose gel containing 0.5 ug/ml ethidium bromide, and the DNA fragments fractionated by electrophoresis.

Fig. 2. Determination of the optimal volume of DNA extract for amplification. Various volumes of extraxt (5, 2, 0.5 ul) were added as indicated to the PCR reaction mixture and amplified with 1U AmpliTaq added in the hot-start procedure run for 45 cycles: One min. at 93 deg C and 2 min. at 72 deg C. Primer concentrations were 10 uM for the primer pairs of the GUS and HPT genes and 2.5 uM for Gos5. 10 ul of the 100 ul reaction volume was loaded on a 2% agarose gel containing 0.5 ug/ml ethidium bromide and electrophorized.

One unit of AmpliTaq was added either at room temperature before the first denaturation step of 93 deg C or at 85 deg C after an initial denaturation step of 1 minute at 93 deg C (hot-start). In both cases no previous heat-treatment at 98 deg C was made.

For the amplification of all 3 fragments using the hotstart procedure 0.1 ng of genomic DNA was sufficient (Figure 1). The size of the rice genome is estimated to be approximately 9 X 108 bp (Bennett and Smith 1976). Therefore, 0.1 ng rice DNA is equivalent to 100 genome copies, thus proving the sensitivity of the reaction.

Using the sensitive PCR protocol several DNA isolation methods were tested. It appeared that low amounts of EDTA, SDS, cysteine and beta-mercaptoethanol did not interfere with the PCR amplification. A reliable, efficient and fast method for preparing samples was found to be: 

          50 mg of fresh plant tissue (root, callus or green tissue), in
          Potter tube (Kontes Scientific Glassware/Instruments, Vineland New
          Jersey)
          200 ul of isolation buffer: 200 mM TRIS-HCl pH 8.3  20 mM Na-EDTA
          500 mM NaCl   0.1% cysteine    20 mM beta-mercaptoethanol
          0.2% sodium-dodecyl sulfate (SDS)   100 ug/ml proteinase K
          homogenization for 1 minute with a pestle (Kontes Scientific
          Glassware/Instruments) in an electric driven Potters homogenizer
          (B.  Braun, Melsungen, FRG)
          incubation for 6 hours at 65 deg C
          1 hour at 105 deg C
          spinning down in Eppendorf centrifuge for 1 min.
          supernatant was used for PCR analysis.
None of the compounds in the extraction buffer inhibited the activity of AmpliTaq in the used concentrations. In this way a sample was prepared from 50 mg of transgenic callus transformed with HH271 (Meijer et al. 1991). Different amounts of extract were tested (Fig. 2).

It can be seen from Figure 2 that 2 ul of extract was sufficient to yield clear bands of the 3 fragments. When it is estimated that 1% of the 5 million rice callus cells (50 mg) released their genomic DNA, this quantity should be equivalent to 500 genome copies. This number is in the same order of magnitude as the numbers found employing plasmid or purified genomic DNA.

Both the optimized PCR and the DNA extraction method therefore fulfill the requirement of being simple and quick and allow the screening of large numbers of samples like calli or parts of rice plants. Nevertheless the PCR analysis is no proof for stable integration of a foreign gene into the genome.

As we found optimum temperatures for annealing and denaturation to be critical, optimization of temperatures for each type of thermocycler should be carried out. Using a thermosensor we found rather large (2-4 deg C) differences in different PCR apparatus at the same temperature settings. Secondly, as optimal amplification of fragments in the PCR reaction depends on the concentration of template DNA, the absence of amplified fragments does not exclude its presence. It can be present in too high or too low a concentration. Therefore, false negatives can not always easily be discriminated from truely nontransformed tissue. As a routine we further incorporate negative controls to check contamination reagents for the PCR reaction, DNA isolation and rice tissue culture medium since the PCR is known to be very sensitive to contaminating DNA. It is known that using 45 cycles increases the risk of amplifying contaminating DNA molecules. We have found that using 35 cycles minimizes this risk and hardly any false positives have been found.

References

Benfey, P. N. and N. H. Chua, 1989. Regulated genes in transgenic plants. Science 244: 174-181.

Bennet, M. D. and J. B. Smith, 1976. Nuclear DNA amounts in angiosperms. Proc. R. Soc. London B274: 227-274.

Hayashimoto, A., Z. Li and N. Mural, 1990. A PEG-mediated protoplast transformation system for production of fertile transgenic rice plants. Plant Physiol. 93: 857-863.

Hensgens L.A.M. and P. E. van Os-Ruygrok, 1989. Isolation of RNA and DNA from different rice tissues. RGN 6: 163-168.

Meijer, E. G. M., R. A. Schilperoort, S. Rueb, P. E. van Os-Ruygrok and L.A.M. Hensgens, 1991. Transgenic rice cell lines and plants: expression of transferred chimeric genes. Plant Mol. Biol. 16: 807-820.

de Pater, S., L.A.M. Hensgens and R. A. Schilperoort, 1990. Structure and expression of a light-inducible rice gene. Plant Mol. Biol. 15: 399-406.

Potrykus, I., 1990. Gene transfer to cereals: an assessment. Bio/Technology 8: 535-542.

Shimamoto, K., R. Terada, T. Izawa and H. Fujimoto, 1989. Fertile transgenic rice plants rege- nerated from transformed protoplasts. Nature 338: 274-276.

Zhang, W., and R. Wu, 1988. Efficient regeneration of transgenic plants from rice protoplasts and correctly regulated expression of the foreign gene in the plants. Theor. Appl. Genet. 76: 845-840.