Proc. Nat!. Acad. Sci. USA Vol. 88, pp. 5212-5216, June 1991 Botany
In vivo random B-glucuronidase gene fusions in Arabidopsis thaliana (plant gene expression/insertional mutagenesis/promoter cloning/transgenic plants/phloem-specific gene expression)
SUNEE KERTBUNDIT*, HENRI DE GREVE*, FRANCINE DEBOECK*, MARC VAN MONTAGU*t, AND JEAN-PIERRE HERNALSTEENS*4 *Laboratorium voor Genetische Virologie, Vrije Universiteit Brussel, B-1640 Sint-Genesius-Rode, Belgium; and tLaboratorium voor Genetica, Rijksuniversiteit Gent, B-9000 Gent, Belgium
Contributed by Marc Van Montagu, March 11, 1991
can be transformed efficiently by Agrobacterium. The small cruciferous plant Arabidopsis thaliana is an ideal model system for molecular genetics because of its small genome size, low repetitive DNA content, short generation time, and well-known genetics (9). Efficient Agrobacterium-mediated transformation of Arabidopsis has been described (10-13). The aim of the present work was to combine the advantages of T-DNA insertion mutagenesis and gus gene fusion technology. By transformation of Arabidopsis with the appropriate T-DNA vectors, gus fusions were obtained at a high frequency. Different types of organ- and tissue-specific GUS expression patterns were found. The usefulness of the system was further proved by the cloning of a DNA fragment mediating phloem-specific gene expression in homologous and heterologous systems.§
Vectors were constructed for the isolation of ABSTRACT random transcriptional and translational fi-glucuronidase gene fusions in plants. This system is based on the random integration of the transferred DNA (T-DNA) into the plant nuclear genome. The Escherichia coli .3-glucuronidase coding sequence without promoter, and also devoid of its ATG initiation site in the translational gene fusion vector, was inserted in the T-DNA with its 5' end at a distance of 4 base pairs from the right T-DNA border sequence. Transgenic plants can be selected by using a chimeric (P355-nptlI-3' ocs) kanamycinresistance gene present in the same T-DNA. Subsequent screening of these for 13-glucuronidase expression allows the identification of clones harboring a fusion of the fi-glucuronidase coding sequence with plant 5' regulatory sequences. After transformation of Arabidopsis thaia C24 root explants, .8-glucuronidase expression was detected in 54% and 1.6% of the plants transformed with the transcriptional and translational fusion vectors, respectively. Several different patterns of tissue-specific 3-glucuronidase expression were identified. The plant upstream sequence of a 8-glucuronidase fusion that is specifically expressed in the phloem of all organs was cloned and sequenced. After introduction in A. thalana C24 and Nicotiana tabacum SR1, this sequence mediates the same highly phloem-specific 8-glucuronidase expression pattern as in the original transgenic plant from which it was isolated. These data demonstrate that this system facilitates the isolation and analysis of plant DNA sequences mediating regulated gene expression.
MATERIALS AND METHODS Bacterial Strains and Vector Constructions. DNA manipulations were performed as described (14). E. coli JM101 (15) and K514 (16) were used for plasmid transformation. E. coli NM430 (17) and MC1061 (18) were used for growing bacteriophage EMBL4. Plasmid pUC8 (19) was used for subcloning. The pGV1030 was derived from pGV943 (20) by deleting the unique Pst I/Sca I fragment of pGV943, thereby inactivating the 8-lactamase gene. The binary vectors were mobilized (21) from E. coli into Agrobacterium tumefaciens strain EHA101 (22). Plant Materials and Transformation. A. thaliana C24 and Columbia were provided by M. Jacobs (Vrije Universiteit Brussel, Belgium) and G. Redei (University of Missouri, Columbia), respectively. Root explants were transformed as described (13). After selection on medium containing carbenicillin (500 mg/liter) and kanamycin (50 mg/liter) and solidified with 0.35% Phytagel, transgenic plants were identified by testing for nopaline synthase activity (23). Leaf discs of Nicotiana tabacum SR1 were transformed essentially as described by Horsch et al. (24). DNA Hybridization. The number of T-DNA copies in transgenic plants transformed by pGV1047 and pGV1048 was determined by DNA hybridization. DNA was extracted (25) and digested with the restriction enzymes EcoRI, EcoRV, Ase I, and Ssp I. Resulting fragments were subjected to electrophoresis in 0.8% agarose gel, transferred to Hybond-N membrane (Amersham), and hybridized with specific probes for left and right border fragments-respectively, the HindIll fragment of pGV0422 (26) containing the nos gene and the BamHI/EcoRI
Gene fusions are useful tools for studying the regulation of gene expression. Previous studies have shown that, in plants, random transcriptional and translational gene fusions can be mediated at a high frequency by transferred DNA (T-DNA) integration (1-3). This strategy allowed the isolation of the promoter sequences of genes expressed in different organs of the plants. In these studies, the neomycin phosphotransferase II (nptIl) was used as a reporter gene. This gene is an excellent selectable marker (4-6); however, quantification of NPTII is laborious and no histochemical technique for its detection is available. A more attractive alternative for the isolation of random gene fusions is the Escherichia coli f3-glucuronidase gene (gus). Its gene product (GUS) is very stable, can tolerate large amino-terminal additions, and was not detected in most higher plants (7). GUS can be assayed by sensitive and specific techniques (8). In particular, the histochemical staining using 5-bromo-4-chloro-3-indolyl 3-Dglucuronic acid (X-GlcA) facilitates the detailed analysis of tissue or cell-type specificity of the gene fusions. T-DNA-mediated gene fusions were reported to occur at high frequency, independently of genome size (3). Therefore, this method can in principle be applied to all plant species that
Abbreviations: T-DNA, transferred DNA; X-GlcA, 5-bromo-4chloro-3-indolyl 1-D-glucuronic acid. tTo whom reprint requests should be addressed at: Laboratorium voor Genetische Virologie, Vrije Universiteit Brussel, Paardenstraat 65, B-1640 Sint-Genesius-Rode, Belgium. §The sequence reported in this paper has been deposited in the GenBank data base (accession no. M64739).
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 5212
Proc. Natl. Acad. Sci. USA 88 (1991)
Botany: Kertbundit et al. fragment of pBI121 (7) containing the gus coding sequence and the 3' nos sequence. Probes were labeled with [a-32P]dCTP using the Amersham multiprime labeling kit (RPN.1601) and were separated from unincorporated nucleotides by a PD-10 column (Pharmacia) equilibrated with 10 mM Tris-HCl, pH 8/1 mM EDTA/100 mM NaCl. GUS Detection. f3-Glucuronidase activity was detected by staining hand-cut leaves, stems, roots, and flowers of each transgenic plant with X-GlcA (7). For localization of GUS activity, plant tissue was embedded in 5% agarose (Sigma; type VII) and cut into 40- to 150-pm-thick slices using a Vibroslice model 752 microtome. Sections were incubated in a substrate solution containing 1 mM X-GIcA, 0.1 mM K3Fe(CN)6, and 0.1 mM K4Fe(CN)6 in 50 mM phosphate buffer for 1-16 hr at 370C. Chlorophyll was removed as described by Benfey et al. (27). In transgenic plants strongly expressing GUS, histochemical GUS detection using X-GIcA can produce spurious results because of diffusion of the soluble indoxyl derivative of X-GIcA into adjacent GUS-negative cells. The use of the Vibroslice microtome to cut thin tissue sections, the K3Fe(CN)6/K4Fe(CN)6 catalyst (28), and the shorter incubation time improved the localization of GUS. DNA Sequencing. DNA sequences were determined by the technique of Sanger et al. (29) using the Boehringer pUC sequencing kit. T-DNA Rescue. For cloning of T-DNA and flanking plant DNA sequences from the transgenic Arabidopsis line At3001, total plant DNA was digested by EcoRI. Fragments of 10-14 kilobases (kb) were purified from a 0.8% agarose gel by electroelution (14) and ligated into EcoRI-digested A vector EMBL4 (17). After in vitro packaging using the Promega Packagene kit, resulting bacteriophages were plated onto E. coli MC1061. The resulting plaques were transferred to E. coli NM430 grown on LB medium containing 5-bromo-4-chloro-3-indolyl f-Dgalactopyranoside (40 ,ug/ml) and 0.5 mM isopropyl f-Dthiogalactopyranoside (14). The clones harboring the T-DNA produce blue lysis zones on this medium. Cloning of T-DNA Target Site. A Sau3AI genomic library of A. thaliana Landsberg in A Charon 35 (a kind gift of D. Jofuku) was screened by plaque hybridization (14) using the flanking plant DNA as a probe. From one of the positive plaques, the EcoRI fragment hybridizing with the probe was subcloned into pUC8 to generate pGV1063 and was subsequently sequenced.
RESULTS Vectors for Transcriptional and Translational 8-Glucuronidase Gene Fusion. Two vectors were constructed to isolate fusions between promoter regions of plant genes and the ,B-glucuronidase coding sequence. The transcriptional fusion vector pGV1047 (Fig. 1A) harbors the HindIII/EcoRI fragment of pGV1030, which contains a streptomycin-resistance (SmR) gene and the replica-
tion origins from pVS1 and ColEl (20). The 7.2-kb T-DNA of pGV1047 is flanked by synthetic left and right borders. It contains a chimeric kanamycin-resistance gene comprising the cauliflower mosaic virus P35S promoter (30), the neomycin phosphotransferase II (nptIl) coding sequence of TnS (4), and the 3' controlling sequence of the octopine synthase gene (3' ocs) (31) as a selectable marker. The T-DNA also contains the nopaline synthase (nos) gene (32) as a screenable marker, the bacterial suppressor gene supF, which can be used as a marker in E. coli for the isolation of recombinant clones, a Not I linker, and a promoter-less ,B-glucuronidase (gus) reporter gene (7). The ATG initiation site of the gus gene is located 4 base pairs (bp) to the right border of the T-DNA (Fig. 1B). The gus coding sequence is followed by the 3' controlling sequence of the nopaline synthase (3' nos) (32).
5213
A
snvsp pBR325
pVS1 Rep/Sta
B pGV1047 GACA
AATTACAACGGTATATATCCTGCC
Met Val Arg Pro Val Glu Thr Pro AGTC ATG GTC CGT CCT GTA GAA ACC CCA
In vivo random B-glucuronidase gene fusions in Arabidopsis thaliana (plant gene expression/insertional mutagenesis/promoter cloning/transgenic plants/phloem-specific gene expression)
SUNEE KERTBUNDIT*, HENRI DE GREVE*, FRANCINE DEBOECK*, MARC VAN MONTAGU*t, AND JEAN-PIERRE HERNALSTEENS*4 *Laboratorium voor Genetische Virologie, Vrije Universiteit Brussel, B-1640 Sint-Genesius-Rode, Belgium; and tLaboratorium voor Genetica, Rijksuniversiteit Gent, B-9000 Gent, Belgium
Contributed by Marc Van Montagu, March 11, 1991
can be transformed efficiently by Agrobacterium. The small cruciferous plant Arabidopsis thaliana is an ideal model system for molecular genetics because of its small genome size, low repetitive DNA content, short generation time, and well-known genetics (9). Efficient Agrobacterium-mediated transformation of Arabidopsis has been described (10-13). The aim of the present work was to combine the advantages of T-DNA insertion mutagenesis and gus gene fusion technology. By transformation of Arabidopsis with the appropriate T-DNA vectors, gus fusions were obtained at a high frequency. Different types of organ- and tissue-specific GUS expression patterns were found. The usefulness of the system was further proved by the cloning of a DNA fragment mediating phloem-specific gene expression in homologous and heterologous systems.§
Vectors were constructed for the isolation of ABSTRACT random transcriptional and translational fi-glucuronidase gene fusions in plants. This system is based on the random integration of the transferred DNA (T-DNA) into the plant nuclear genome. The Escherichia coli .3-glucuronidase coding sequence without promoter, and also devoid of its ATG initiation site in the translational gene fusion vector, was inserted in the T-DNA with its 5' end at a distance of 4 base pairs from the right T-DNA border sequence. Transgenic plants can be selected by using a chimeric (P355-nptlI-3' ocs) kanamycinresistance gene present in the same T-DNA. Subsequent screening of these for 13-glucuronidase expression allows the identification of clones harboring a fusion of the fi-glucuronidase coding sequence with plant 5' regulatory sequences. After transformation of Arabidopsis thaia C24 root explants, .8-glucuronidase expression was detected in 54% and 1.6% of the plants transformed with the transcriptional and translational fusion vectors, respectively. Several different patterns of tissue-specific 3-glucuronidase expression were identified. The plant upstream sequence of a 8-glucuronidase fusion that is specifically expressed in the phloem of all organs was cloned and sequenced. After introduction in A. thalana C24 and Nicotiana tabacum SR1, this sequence mediates the same highly phloem-specific 8-glucuronidase expression pattern as in the original transgenic plant from which it was isolated. These data demonstrate that this system facilitates the isolation and analysis of plant DNA sequences mediating regulated gene expression.
MATERIALS AND METHODS Bacterial Strains and Vector Constructions. DNA manipulations were performed as described (14). E. coli JM101 (15) and K514 (16) were used for plasmid transformation. E. coli NM430 (17) and MC1061 (18) were used for growing bacteriophage EMBL4. Plasmid pUC8 (19) was used for subcloning. The pGV1030 was derived from pGV943 (20) by deleting the unique Pst I/Sca I fragment of pGV943, thereby inactivating the 8-lactamase gene. The binary vectors were mobilized (21) from E. coli into Agrobacterium tumefaciens strain EHA101 (22). Plant Materials and Transformation. A. thaliana C24 and Columbia were provided by M. Jacobs (Vrije Universiteit Brussel, Belgium) and G. Redei (University of Missouri, Columbia), respectively. Root explants were transformed as described (13). After selection on medium containing carbenicillin (500 mg/liter) and kanamycin (50 mg/liter) and solidified with 0.35% Phytagel, transgenic plants were identified by testing for nopaline synthase activity (23). Leaf discs of Nicotiana tabacum SR1 were transformed essentially as described by Horsch et al. (24). DNA Hybridization. The number of T-DNA copies in transgenic plants transformed by pGV1047 and pGV1048 was determined by DNA hybridization. DNA was extracted (25) and digested with the restriction enzymes EcoRI, EcoRV, Ase I, and Ssp I. Resulting fragments were subjected to electrophoresis in 0.8% agarose gel, transferred to Hybond-N membrane (Amersham), and hybridized with specific probes for left and right border fragments-respectively, the HindIll fragment of pGV0422 (26) containing the nos gene and the BamHI/EcoRI
Gene fusions are useful tools for studying the regulation of gene expression. Previous studies have shown that, in plants, random transcriptional and translational gene fusions can be mediated at a high frequency by transferred DNA (T-DNA) integration (1-3). This strategy allowed the isolation of the promoter sequences of genes expressed in different organs of the plants. In these studies, the neomycin phosphotransferase II (nptIl) was used as a reporter gene. This gene is an excellent selectable marker (4-6); however, quantification of NPTII is laborious and no histochemical technique for its detection is available. A more attractive alternative for the isolation of random gene fusions is the Escherichia coli f3-glucuronidase gene (gus). Its gene product (GUS) is very stable, can tolerate large amino-terminal additions, and was not detected in most higher plants (7). GUS can be assayed by sensitive and specific techniques (8). In particular, the histochemical staining using 5-bromo-4-chloro-3-indolyl 3-Dglucuronic acid (X-GlcA) facilitates the detailed analysis of tissue or cell-type specificity of the gene fusions. T-DNA-mediated gene fusions were reported to occur at high frequency, independently of genome size (3). Therefore, this method can in principle be applied to all plant species that
Abbreviations: T-DNA, transferred DNA; X-GlcA, 5-bromo-4chloro-3-indolyl 1-D-glucuronic acid. tTo whom reprint requests should be addressed at: Laboratorium voor Genetische Virologie, Vrije Universiteit Brussel, Paardenstraat 65, B-1640 Sint-Genesius-Rode, Belgium. §The sequence reported in this paper has been deposited in the GenBank data base (accession no. M64739).
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 5212
Proc. Natl. Acad. Sci. USA 88 (1991)
Botany: Kertbundit et al. fragment of pBI121 (7) containing the gus coding sequence and the 3' nos sequence. Probes were labeled with [a-32P]dCTP using the Amersham multiprime labeling kit (RPN.1601) and were separated from unincorporated nucleotides by a PD-10 column (Pharmacia) equilibrated with 10 mM Tris-HCl, pH 8/1 mM EDTA/100 mM NaCl. GUS Detection. f3-Glucuronidase activity was detected by staining hand-cut leaves, stems, roots, and flowers of each transgenic plant with X-GlcA (7). For localization of GUS activity, plant tissue was embedded in 5% agarose (Sigma; type VII) and cut into 40- to 150-pm-thick slices using a Vibroslice model 752 microtome. Sections were incubated in a substrate solution containing 1 mM X-GIcA, 0.1 mM K3Fe(CN)6, and 0.1 mM K4Fe(CN)6 in 50 mM phosphate buffer for 1-16 hr at 370C. Chlorophyll was removed as described by Benfey et al. (27). In transgenic plants strongly expressing GUS, histochemical GUS detection using X-GIcA can produce spurious results because of diffusion of the soluble indoxyl derivative of X-GIcA into adjacent GUS-negative cells. The use of the Vibroslice microtome to cut thin tissue sections, the K3Fe(CN)6/K4Fe(CN)6 catalyst (28), and the shorter incubation time improved the localization of GUS. DNA Sequencing. DNA sequences were determined by the technique of Sanger et al. (29) using the Boehringer pUC sequencing kit. T-DNA Rescue. For cloning of T-DNA and flanking plant DNA sequences from the transgenic Arabidopsis line At3001, total plant DNA was digested by EcoRI. Fragments of 10-14 kilobases (kb) were purified from a 0.8% agarose gel by electroelution (14) and ligated into EcoRI-digested A vector EMBL4 (17). After in vitro packaging using the Promega Packagene kit, resulting bacteriophages were plated onto E. coli MC1061. The resulting plaques were transferred to E. coli NM430 grown on LB medium containing 5-bromo-4-chloro-3-indolyl f-Dgalactopyranoside (40 ,ug/ml) and 0.5 mM isopropyl f-Dthiogalactopyranoside (14). The clones harboring the T-DNA produce blue lysis zones on this medium. Cloning of T-DNA Target Site. A Sau3AI genomic library of A. thaliana Landsberg in A Charon 35 (a kind gift of D. Jofuku) was screened by plaque hybridization (14) using the flanking plant DNA as a probe. From one of the positive plaques, the EcoRI fragment hybridizing with the probe was subcloned into pUC8 to generate pGV1063 and was subsequently sequenced.
RESULTS Vectors for Transcriptional and Translational 8-Glucuronidase Gene Fusion. Two vectors were constructed to isolate fusions between promoter regions of plant genes and the ,B-glucuronidase coding sequence. The transcriptional fusion vector pGV1047 (Fig. 1A) harbors the HindIII/EcoRI fragment of pGV1030, which contains a streptomycin-resistance (SmR) gene and the replica-
tion origins from pVS1 and ColEl (20). The 7.2-kb T-DNA of pGV1047 is flanked by synthetic left and right borders. It contains a chimeric kanamycin-resistance gene comprising the cauliflower mosaic virus P35S promoter (30), the neomycin phosphotransferase II (nptIl) coding sequence of TnS (4), and the 3' controlling sequence of the octopine synthase gene (3' ocs) (31) as a selectable marker. The T-DNA also contains the nopaline synthase (nos) gene (32) as a screenable marker, the bacterial suppressor gene supF, which can be used as a marker in E. coli for the isolation of recombinant clones, a Not I linker, and a promoter-less ,B-glucuronidase (gus) reporter gene (7). The ATG initiation site of the gus gene is located 4 base pairs (bp) to the right border of the T-DNA (Fig. 1B). The gus coding sequence is followed by the 3' controlling sequence of the nopaline synthase (3' nos) (32).
5213
A
snvsp pBR325
pVS1 Rep/Sta
B pGV1047 GACA
AATTACAACGGTATATATCCTGCC
Met Val Arg Pro Val Glu Thr Pro AGTC ATG GTC CGT CCT GTA GAA ACC CCA
