We have been analyzing the sequences of the autonomous transposon Ac required for transposition in transgenic tobacco plants. We recently reported (Baker et al., MNL 1986; Baker et al., EMBO J. 6:1547, 1987) a phenotypic assay which allowed the selection of Ac excision events in tobacco. This assay involved the construction of an NPTII gene inactivated by the insertion, in vitro, of a copy of Ac. This inactivated gene was inserted into the T-DNA of A. tumefaciens and transferred to tobacco protoplasts. The presence of transformed tobacco calli resistant to the antibiotic kanamycin indicated Ac excision from the NPTII gene, restoring activity of the gene. We have now used this assay to study Ac transposition by inserting a series of in vitro constructed derivatives within the NPTII gene and determining whether they could excise. A preliminary report of these experiments was given by Coupland et al., MNL 1986.
The first type of Ac derivatives have suffered deletions within the DNA which encodes the long (600-700 b) untranslated leader of the Ac transcript. Two of these deletions extended from Ac coordinates (we number Ac from the BamHI end, i.e. reversed coordinates compared to Muller-Neumann et al., Mol. Gen. Genet. 198:19, 1984) 356 bp to 920 bp (in plasmid pKU31) and from coordinates 245 bp to 736 bp (in plasmid pKU33). Both of these derivatives yielded KmR calli at a frequency equal to or higher than that of Ac itself (in plasmid pKU3). As the deletions remove almost all of the untranslated leader, we conclude that this unusual feature of the Ac transcript is not required for expression of a product required for transposition. Indeed the deletions remove approximately 14% of the Ac element, which still transposes autonomously.
However, a deletion which removes the DNA encoding approximately 27 amino acids from the 5'-end of the open reading frame (in plasmid pKU9) prevents transposition. However, this deleted element does transpose in tobacco cells already containing Ac, indicating the deletion derivative could be complemented by an intact Ac, so the deletion must remove sequences required for expression of a trans-acting Ac product.
In order to confirm that the 3.5 kb Ac transcript encodes a product required for transposition, we cloned the open reading frame (ORF) of the transcript, present in a cDNA clone constructed from the partial clones of Kunze et al. (EMBO J. 6:1555-1563, 1987), between a plant promoter and a plant polyadenylation signal. The combined cDNA clone contained the entire ORF and 160 bp of the leader. The cDNA was cloned in the sense (in pKU38) and the nonsense (in pKU39) orientations. These constructions were then transferred to tobacco protoplasts which had previously been transformed with a T-DNA (from plasmid pGV3850::pKU4) which contained an Ac D element incapable of autonomous transposition stably inserted within an NPTII gene.
The Ac D element had lost the entire internal HindIII fragment of Ac. We have three lines of evidence to indicate that expression of the ORF in tobacco protoplasts allowed excision of the Ac D element: (i) A low number of KmR calli were recovered after transformation of the pGV3850::pKU4 transformed protoplasts with pKU38, but not after transformation with pKU39. KmR ought to occur only after excision of Ac D. (ii) The T-DNAs of pKU38 and pKU39 carry a Hygromycin resistance gene as well as the Ac ORF HyR calli which had inherited the T-DNA of pKU38 or pKU39 were selected and then tested for NPTII expression by an in situ gel assay. None of 6 calli containing the pKU39 T-DNA expressed NPTII. Four of six calli containing the pKU38 T-DNA expressed NPTII. Again the presence of the ORF in the sense orientation results in NPTII expression. (iii) Southern analysis of DNA isolated from one HyR NPTII- callus transformed with the pKU38 T-DNA and the pGV3850::pKU4 DNA revealed a 2.9 kb EcoRI-HindIII fragment which hybridized to an NPTII gene probe. This fragment is expected only after excision of the Ac D element from the NPTII gene. This confirmed that NPTII expression in the callus was a consequence of Ac D excision.
The ORF must therefore encode a product required for Ac excision, and it becomes increasingly likely that this product is the only Ac product required for transposition, that is the transposase.
In addition we have tried to identify the sequences required at the ends of the Ac element for transposition. This analysis is most advanced at the end encoding the 5'-end of the transcript. The deletion of Ac sequences between coordinates bp 44 and bp 92 (in plasmid pKU19) and between bp 75 and bp 181 (in plasmid pKU28) prevent transposition even if Ac is present in the same cell. We conclude that these sequences are required for recognition of the ends of the element by an Ac encoded and/or host proteins prior to excision. An active end of Ac clearly constitutes more than 75 bp, as pKU19 will not transpose, but less than 245 bp, as pKU33 excises autonomously We have constructed a series of deletions progressing outwards from 245 bp towards the end of the element and have tested their ability to excise in tobacco protoplasts containing Ac. These experiments are in progress, but for full activity more than approximately 150 bp are apparently required at the 5'-end of the Ac element.
A similar series of deletions at the 3'-end are currently being transformed into tobacco protoplasts.
In addition we have constructed a plasmid which contains an Ac element in which the 3'-end of the element has been replaced by approximately 480 bp from the 5'-end. This element therefore contains a functional 5'-end at both ends. This element was unable to transpose in tobacco protoplasts containing Ac.
These experiments pose two problems: how does the transposase or other host proteins recognize, but distinguish between, the two ends of Ac, and secondly how does Ds1 transpose when it only contains very short sequences of less than 20 bp at the ends which are directly homologous to the ends of Ac? We are currently pursuing these problems.
George Coupland, Christiane Plum, Barbara Baker, Jozef Schell and Peter Starlinger
Return to the MNL 62 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page