Supplementary material
Methods S1 Supplementary Experimental Procedures
For the construction of modified Amy8 SRC/GARC containing an extra G box, the G box was inserted into the region between the GC box and the GARE of Amy8 SRC/GARC at two different positions in a two-stage PCR. In the first PCR, GboxF1 or GboxF2 primer was used as the forward primer and 35SR primer was used as the reverse primer with p8SRC/GARC-4 (Chen et al., 2006) serving as the DNA template. The PCR product was subsequently used as the reverse megaprimer in a second PCR with T3 primer (Strategene) as the forward primer and p8SRC/GARC-4 as the DNA template. The DNA fragment containing the extra G box adjacent to either 3’ of the GC box or 5’ of the GARE of Amy8 SRC/GARCand flanked by ApaI and PstI sites was then sub-cloned into p35mA-Luc, generating p8SRC/GARC-6 and p8SRC/GARC-9, respectively.
For the construction of modified Amy8 SRC/GARC containing two copies of the TA box, the TA box in Amy8 SRC/GARC was duplicated in a two-stage PCR as described above, 2TAF and 35SR primers were used to generate p8SRC/GARC-8.
For the construction of modified Amy8 SRC/GARCcontaining both G box and the duplicated TA box, Amy8 SRC/GARCwas modified by PCR using GboxF1/GbpxF2 and 35SR primers as the forward and reverse primers, respectively, and p8SRC/GARC-8 as the DNA template in the first PCR. The PCR product was then used as the reverse megaprimer in the second PCR with T3 primer as the forward primer and p8SRC/GARC-8 as the DNA template. The DNA fragment containing the G box and duplicated TA box of Amy8 SRC/GARC was flanked by ApaI and PstI sites and subsequently sub-cloned into p35mA-Luc, generating p8SRC/GARC-7 and p8SRC/GARC-10, respectively.
For the construction of modifiedAmy8 SRC/GARC containing two copies of the G box and the duplicated TA box, Amy8 SRC/GARCwas modified by PCR using GboxF1 and 35SR primer as the forward and reverse primers, respectively, and p8SRC/GARC-10 as the DNA template in the first PCR. The PCR product was then used as the reverse megaprimer in the second PCR with T3 primer as the forward primer and p8SRC/GARC-10 as the DNA template. The DNA fragment containing two copies of the G box and the duplicated TA box of Amy8 SRC/GARC was flanked by ApaI and PstI sites and then sub-cloned into p35mA-Luc, generating p8SRC/GARC-11.
For the construction of plasmid carrying modified Amy8 SRS/GARS with the G box and the duplicated TA box in the Amy8 promoter, the 5’-flanking region (-1060 to +127) of Amy8 was excised from pAG8 (Chan et al., 1993) by HindIII and NcoI and the restriction fragment was sub-cloned into p35mA-Luc, yielding pAmy8-Luc. The TA box in Amy8 was duplicated in a two-stage PCR as described above using 2TAF and Luc1 primers as the forward and reverse primers, respectively, and pAmy8-Luc as the DNA template in the first PCR. The PCR product was used as the reverse megaprimer in the second PCR with T3 primer as the forward primer. The PCR fragment containing the duplicated TA box of Amy8 promoter was then sub-cloned into p35mA-Luc in between HindIII and NcoI, generating pAmy8(+2TA)-Luc. The G box and the duplicated TA box in Amy8 promoter was modified by PCR using GboxF1 and Luc1 primer as the forward and reverse primers, respectively, and pAmy8(+2TA)-Luc as the DNA template in the first PCR. The PCR product was then used as the reverse megaprimer in the second PCR with T3 primer as the forward primer and pAmy8(+2TA)-Luc as the DNA template. The DNA fragment containing the G box and the duplicated TA box of Amy8 promoter was then sub-cloned into p35mA-Luc at HindIII and NcoI, generating pAmy8(+G+2TA)-Luc.
For the construction of plasmid containing the Amy3 promoter-signal peptide sequences-Luc chimericgene, the Amy3 5’ region (including ~1.0-kb 5’-flanking sequences,92-bp 5’ untranslated region, and90-bpsignal peptide sequences ) was PCR-amplified by primers Amy3F and Amy3R using p3G-132II (Lu et al., 1998) as the DNA template. The DNA fragment containing Amy3 promoter and signal peptide and with HindIII and NcoI sites at both ends was then sub-cloned into p35mA-Luc, generating pAmy3-Luc.
To prepare the constructs for the Agrobacteria-mediated transformation, plasmids p8SRC/GARC-4, p8SRC/GARC-6, p8SRC/GARC-7, p8SRC/GARC-8, and p8SRC/GARC-11were linearized with SacII, blunt-ended, and inserted into the blunt-end binary vector pSMY1H (Ho et al., 2000), which contains the CaMV35S promoter, hpt coding region, and tumor morphology large gene (tml) terminator fusion gene, to yield pAT4 , pAT6, pAT7 , pAT8, and pAT11. Plasmids pAmy8-Luc, pAmy8(+G+2TA)-Luc, and pAmy3-Luc were digested with HindIII and the restriction fragments were sub-cloned into the corresponding site on pSMY1H to generate pA8, pA8(+G+2TA), and pA3.
Two expression vectors were designed to express recombinant hEGF in rice cells under the control of the modifiedAmy8 promoter. Plasmid pUC-hEGF, kindly provided by Dr. Wen-Hsing Chang (Chiayi University, Chiayi), contained a synthetic sequence of the hEGF gene (53 amino acids). A duplicated c-Myc epitope tag was amplified with the mycF and T3 primers using p2cmyc (Chen et al., 2006) as DNA template. The DNA fragment containing the duplicated c-Myc tag was digested with BglII and BamHI and then sub-cloned into pAHC18, generating pUbi-cmyc. The hEGF gene was excised from pUC-hEGF by BamHI and inserted into pUbi-cmyc at BamHI site to generate pUbi-cmyc-hEGF. The Amy8(+G+2TA) promoter was amplified with the T3 and Amy8BR primers using pAmy8(+G+2TA)-Luc as DNA template and the PCR fragment was flanked with HindIII and BglII sites and sub-cloned into pAHC18, generating pAmy8(+G+2TA)-Ubi(in)-Luc. The Amy8(+G+2TA)-Ubi(in)fusion region was amplified from pAmy8(+G+2TA)-Ubi(in)-Luc by PCR with the T3 and A8UbiR primers. A 113-bp DNA fragment containing the signal peptide sequence of Amy3 gene was amplified with the A3spF and Luc1 primers using pAmy3-Luc as DNA template. The DNA fragment containing the Amy8(+G+2TA)-Ubi(in)with HindIII and SpeI sites at both ends and the DNA fragment containing Amy3 signal peptide regionswith SpeI and NcoI sites at both ends were then sequentially ligated into pUbi-cmyc-hEGF pre-digested with HindIII and NcoI to generate pAmy8(+G+2TA)-SChEGF. The hEGF-KDEL fusion gene containing hEGF and the four-amino-acid endoplasmic reticulum retention signal peptide KDEL at its C-terminus was amplified from pUC-hEGF with the primers for hEGFA and hEGFB. The DNA fragment containing hEGF-KDEL was digested with BamHI and sub-cloned into the same site in pAmy8(+G+2TA)-SChEGF to yield pAmy8(+G+2TA)-SChEGF-KDEL. pAmy8(+G+2TA)-SChEGF and pAmy8(+G+2TA)-SChEGF-KDEL were then linearized with HindIII and inserted into the corresponding site on the binary vector pSMY1H, yielding pA8(+G+2TA)-SChEGF and pA8(+G+2TA)-SChEGF-KDEL.
For high-level production of recombinant hEGF in transgenic rice, the codon-optimized hEGF gene (three copies) was synthesized chemically based on the codon-usage preference in rice genes (MDBio, Inc.) and inserted into pBluscript II SK vector (Strategene) to generate pBS-H3xshEGF-KDEL. The synthetic H3xshEGF-KDEL fusion gene containing three copies of shEGF flanked by hexa-histidine tag (6xHis) at its N-terminus and the KDEL at its C-terminus, and a factor Xa cleavage site (IEGR) between His-tag and shEGF was excised from pBS-H3xshEGF-KDEL by NcoI and BamHI and the restriction fragment was sub-cloned into the same site in pAmy8(+G+2TA)-SChEGF to yield pAmy8(+G+2TA)-SH3xshEGF-KDEL.pAmy8(+G+2TA)-SH3xshEGF-KDEL was then linearized with HindIII and inserted into the corresponding site on the binary vector pSMY1H, generating pA8(+G+2TA)-SH3xshEGF-KDEL.
1