Vector design and construction
To search for developmentally regulated expression elements to drive cre specific expression in reproductive tissues, a set of putative meiosis-related Arabidopsis genes were manually identified from GenBank, including Arabidopsis CDC45. The maize and rice CDC45 homologue sequences were identified by performing BLAST searches in the GenBank genomic sequences against the Arabidopsis CDC45 protein sequence. There were two versions of maize and rice CDC45 promoters identified from GenBank by searching Arabidopsis for CDC45 protein homologues. Two promoters from monocot CDC45 genes, one from maize at chromosome 3 and one from rice at chromosome 11 (Table 1), appear to be restricted in reproductive tissue, and were used to drive the cre expression in pMON138232 and pMON243847, respectively. The relevant expression elements of these genes were cloned by PCR. The corresponding species genomic DNA was used as a template for PCR amplification using Q5® Hot Start High-Fidelity DNA polymerase (NEB Cat. No. M0493) according to the manufacturer instructions. The corresponding genomic regions of these expression elements in GenBank and the primers used for amplifying them are listed in Table 1.
Table 1
Primers and DNA sources used for PCR amplification of cre autoexcision promoters and 3’ UTRs
Expression element | Forward primer | Reverse primer | Size (bp) | GenBank No. / region |
P-At.CDC45 | 5’ ctaatacaaaggtgcatgagtagtagtaactg 3’ | 5’ ttccgtgaaattgaatcacccagaagg 3’ | 1030 | CP002686.1 (9143262..9144291) |
T-At.CDC45 | 5’ catagtctcattgttcttcgattcagtg 3’ | 5’ cacgagcttcaggtcataactctgg 3’ | 734 | CP002686.1 (9146083..9146816) |
P-Gm.RSP1 | 5’ aaataatatataaaaatattacaaaaatc 3’ | 5’ tgaagcaaagtggttagagatgagaatg 3’ | 720 | NC_016091 (44628717..44629436) |
P-Zm.CDC45-1 | 5’ agccacatgcagtgaattctatactcg 3’ | 5’ tgcctcatcaatcagctaggtcggatc 3’ | 2000 | CM007649.1 (234291301..234293300) |
P-Os.CDC45-1 | 5’ acatacatctgtctagattcattaatat 3’ | 5’ tggcgcatcaatcgaagtggtgaattgg 3’ | 1957 | AP014967.1 (1304410..1306330) |
The soybean RSP1 promoter (P-Gm.RSP1) was initially nominated as a disease responsive gene promoter (Resistance Sensitive Protein 1) based on RNAseq data generated in-house. Upon testing this promoter to express a gusA transgene in soybean, we found that it was active at background levels in multiple tissues, except roots and top of hypocotyls (data not shown). Even though it does not seem to fit the developmental regulation pattern of other promoters that we tested, it was included to avoid a decrease in TF because of its low-level expression in most tissues. The same promoter comes from a gene that belongs to the BURP domain-containing protein family and was reported to be expressed in roots and hypocotyls, and is inducible by ABA, salt, and drought treatments (Gm04.3 gene, Xu et al. 2010). pMON263552 with the RSP1 promoter (Fig. 1) was constructed to test SMG autoexcision in soybean.
The 1.2 kp λ phage segment corresponding to GenBank accession No. J02459.1, region 21042 to 22237, was synthesized in Bio Basic Inc. (Markham, ON, Canada) and used in pMON243107 as a spacer sequence. The 754 bp of Arabidopsis AtpE intron corresponding to GenBank accession No. LR699765.1, region 17923873 to 19152863, was amplified by PCR and cloned into pMON291996 to test enhance P-Gm.RSP1 expression. The 804 bp of maize DnaK intron, which was previously annotated as ZmHSP70 intron and disclosed in GenBank Accession No. KX640115.1, was cloned after P-Zm.CDC45-1 in pMON138232 to enhance expression.
The cre coding sequence used is as previously described and is interrupted by 189 bp IV2 intron from the potato ST-LS1 gene (Vancanneyt et al. 1990; Zhang et al. 2003). The promoters described above were used to drive expression of the cre coding sequence. For 3’UTR, T-At.CDC45 (Table 1) was used for soybean, cotton, and canola, and Agrobacterium nos transcription terminator (Depicker et al. 1982) was used for corn.
All dicotyledonous transformation vectors were built on ori pRi vector backbone with kanamycin resistance gene, and the maize transformation vectors were on RK2 oriV replicon with spectinomycin selection (Ye et al., 2011). The right and left borders sequences were described previously (Ye et al., 2008). The dicotyledonous gusA cassette driven by the CaMV 35S promoter and terminated by the Agrobacterium nos transcription terminator was described (Vancanneyt et al. 1990, Ye et al. 2008). The gusA cassette in maize vector pMON138232 was driven by the rice actin1 (Os.Act1) promoter with an additional 333 bp CaMV enhancer sequence in front of the Os.Act1 first intron (McElroy et al. 1990). The gusA cassette in maize vector pMON243847 was driven by a 2181 bp rice tublin-3 (TubA-3) promoter (GenBank accession No. MH931401). In dicotyledonous transformation, the aadA gene with the chloroplast target sequence ctp2 (Chen et al. 2014) driven by Arabidopsis actin 7 (At.Act7) promoter (GenBank accession No. JN400384) and terminated by the Agrobacterium nos transcription terminator was used with spectinomycin for plant selection. In maize transformation, the cp4 epsps coding sequence with the Os.Act1 promoter and the Agrobacterium nos terminator was used for glyphosate selection as described previously (Ye et al. 2011). In some vector designs, the splA (sucrose phosphorylase-like gene, GenBank accession No. AE007871, region 153761..155218) gene derived from Agrobacterium Ti plasmid driven by enhanced USP88 (eUSP88) promoter (Bäumlein et al., 1991; Wang et al. 2006) and terminated by the nos transcription terminator in three dicotyledonous constructs was included to reduce R1 seed screening due to seed abortion phenotype, similar to the approach taken for 2 T-DNA transformation (Fig S1). Some vector designs were simplified to omit this negative selection. The SMG and cre genes were flanked by lox sites for autoexcision. The genetic elements and the T-DNA structure of all binary vectors are depicted in Fig. 1.
Standard cloning procedures were applied for all binary vector construction (Sambrook et al. 1989). For seamless fusion between a promoter and cre elements, the hot fusion cloning protocol was used with PCR products bearing 20–25 bp element junction overlaps (Fu et al. 2014).
Agrobacterium preparation and plant transformation
A single binary vector was transfected into a nopaline type of Agrobacterium tumefaciens strain by electroporation as described previously (Ye et al. 2008). The ABI strain containing gentamicin and kanamycin resistance was used for maize vector transfection using spectinomycin for Agrobacterium selection. The AB30 strain (Ye et al. 2016), which is derived from ABI with deletion of kanamycin resistance gene, was used for soybean and canola binary vector transfection which contains kanamycin-resistant gene in the vector backbone for Agrobacterium selection. The AB33 strain, derived from AB30 with VirGI77V mutation (Ye et al. 2016) was used for cotton binary vector transfection with kanamycin for Agrobacterium selection (Chen et al. 2014).
For soybean (Glycine max) transformation, the dry meristem explants from the cultivar A3555 were mechanically excised (Calabotta et al. 2013). The explants were imbibed for 30 min in inoculation buffer, inoculated with Agrobacterium AB30 suspension containing corresponding binary vectors at OD600 = 0.3 and sonicated for 20 seconds (Ye et al. 2008). The explant co-culture, plant regeneration and growth in green house were described previously (Martinell et al. 2002; Ye et al. 2008), except that 150 mg/L spectinomycin instead of glyphosate was used for selection during shoot elongation.
For cotton transformation, the dry meristem explants from cotton cultivar DP393 seeds were excised mechanically (Dersch et al. 2015). The explants were imbibed in inoculation buffer for 30 min, inoculated with AB33 strain containing binary vectors, and co-cultured for 3–5 days. Plant regeneration was obtained with 150 mg/L spectinomycin selection, which was described in detail previously (Chen et al. 2014).
Canola hypocotyls explants from canola (Brassica napus L.) cultivar Ebony were used for canola transformation (Radke et al. 1992; Ye et al. 2011). Spectinomycin at 100 mg/L instead of glyphosate in the regeneration media was used to recover transgenic canola shoots.
The immature embryos of maize elite cultivar LH244 were used for generating maize transgenic plants with glyphosate selection as previous described (Sidorov & Duncan, 2009).
Molecular analyses
R0 regenerants were analyzed for transgene copy number and vector backbone presence or absence by TaqMan® technology (Applied Biosystems). Leaf samples were collected for DNA extraction (Dellaporta et al. 1983). For dicotyledonous transgenic plant analysis, the gusA gene as a GOI, the aadA, as well as cre were analyzed for copy number. For maize transgenic plants, the gusA, cp4 epsps and cre were analyzed for copy number. The T-DNA left border (LB) was also detected in all constructs for T-DNA intactness. The TaqMan® detection probes of the gusA, cp4 epsps, LB, and the backbone oriRi in dicotyledonous vectors or RK2 oriV in maize vectors were described previously (Ye et al. 2011).
The primers 5’- AGCTAAGCGCGAACTGCAAT-3’ (forward) and 5’- GGCTCGAAGATACCTGCAAGA-3’ (reverse) amplifying the aadA gene in the dicotyledonous binary vectors, and further detected by minor grove binding (MGB) TaqMan® probe 6FAM-TGGAGAATGGCAGCGCAATGACA, were used for the dicotyledonous selectable marker gene copy number assay. The primers 5’-CAAGTGACAGCAATGCTGTTTCA-3’ (forward) and 5’-GTCGAAATCAGTGCGTTCGAA-3’ (reverse) amplifying a cre fragment, and the TaqMan® probe 6FAM-CGGTGAACGTGCAAAA were used for cre cassette presence.
R1 plants are defined as the progeny produced from self-pollinating the R0 plant, i.e. the primary transformant derived from tissue culture. For R1 progeny screening, leaf samples from the green house grown plantlets were collected for DNA extraction. The GOI (gusA), marker gene (aadA for dicotyledonous, cp4 epsps for maize) and cre gene were assayed for copy number with TaqMan® analysis. The GOI TaqMan® detection positive, but marker and cre TaqMan® detection negative plants were counted as MF lines, and a subset of these marker free lines were partially verified by Southern blot with DIG-labeled probes (Ye et al. 2011; Chen et al. 2014). In general, for a population of 100 R1 plants, we project a total 75 R1 plants that are positive for the GOI, either as hemizygotes or homozygotes, and 25 null plants, assuming Mendelian segregation of a single locus (1:3 transmission; 1 null : 2 hemizygous : 1 homozygous transgene segregation). The R1 MF frequency is calculated as percentage of the projected transgenic R1 plants. If all 75 of these hemizygous and homozygous R1 plants are negative for the SMG, the calculated marker gene excision frequency would be 100%.