Dominant Suppressible Mutations

The knottedl (knl) locus of maize is defined by a series of dominant mutations affecting leaf development. We recovered 10 additional mutant alleles in lines containing active Mutator transposable elements. Nine of these alleles contain Mu1 or Mu8 elements inserted within a 310-bp region of the knl third intron. All five Mu8 insertions are in the same orientation whereas both orientations of Mu1 were recovered. Northern analysis showed that ectopic expression of knl within developing leaves is correlated with the mutant phenotype for the four alleles analyzed. Transcript size was not altered. The effect of Mu activity, as measured by the extent of Mu element methylation or by the presence of the autonomous MuDR element, was investigated for two alleles. Knl-mum2, containing a Mu8element, and Knl-mum7, containing a Mu1 element, required Mu activity for the knotted phenotype. We examined the effect of Mu activity on ectopic knl expression in Knl-mum2 and found that the transcript was present in leaves of Mu active individuals only. We discuss possible mechanisms by which Mu activity could condition knl gene expression. T RANSPOSABLE elements are powerful mutagens for generating novel alleles of a locus (COEN et al. 1986; SCHWARZ-SOMMER and SAEDLER 1987; CORCES and GEWR 1991). In addition to insertions that simply block gene function, transposons may alter the timing or tissue specificity of gene action when they insert into regulatory regions (COEN et al. 1989; CHEN et al. 1987; BRADLEY et al. 1993). Transposons may provide alternate transcription start sequences (BARKAN and MARTIENSSEN 1991) or alternative splicing sites (WEIL and WESSLER 1990; ORTIZ and STROMMER 1990), thereby modlfylng the mRNA sequence. Excision of a transposon may generate additional stable alleles due to imprecise excision (SOMMER et al. 1988; COEN et al. 1986). Transposable elements may also alter the regulation of the gene due to sequences contained within the transposon (MASON et al. 1987; MARTIENSSEN et al. 1990; TANDA and CORCES 1991). Mutator (Mu) transposable elements were first recognized by DON ROBERTSON (1978) in a line of corn that exhibited an unusually high mutation rate. Genetic and molecular experiments have shown that the high rate of mutation results from the transposition and insertion of numerous Mu elements [for review, see CHANDLER and WEMAN (1992) and BENNETZEN et al. (1993)l. Most lines of corn contain several families of distinct Mu elements that are lated by their 220-bp terminal inverted repeats; these families are distinguished by the sequences between the repeats. The majority of Mu elements are non-autonomous; their transposition requires 94501, ' Present address: Pharmacia LKB Biotechnology, Alameda, California Genetics 138 1275-1285 (December, 1994) the presence of an autonomous element referred to as MuDR [previously called MuR, MuA and Mu9 (CHOMET et al. 1991; QIN et al. 1991; HERSHBERGER et al. 1991)l. While MuDR has inverted repeats that are similar to non-autonomous elements, it is distinguished by internal unique sequences that encode trans-acting factors needed for transposition. In addition to its presumed role in regulating Mu transposition, MuDR can also influence the expression of genes that contain nonautonomous Mu element insertions. Mu activity ( i . e., the presence of MuDR) can be determined by a number of methods. A somatic reversion assay takes advantage of a non-autonomous Mu element inserted in a color gene. The element excises in the presence of Mu activity to restore color gene function, thereby producing revertant sectors (WALBOT 1986; CHOMET et al. 1991; BROWN and SUNDARESAN 1992). A molecular assay distinguishes active from inactive elements based on the fact that the inactive elements are methylated (CHANDLER and WALBOT 1986; BENNETZEN 1987). Southern analysis using a MuDR probe can distinguish the autonomous element from related defective elements based on size of the hybridizing bands (CHOMET et al. 1991; D. LISCH, P. CHOMET and M. FREELING, manuscript submitted for publication). Finally, the phenotypes of certain Mu-induced mutations that are dependent upon the presence of Mu activity can also provide a reliable, independent marker for the presence of Mu activity (MARTIENSSEN et aL 1990; CHOMET et al. 1991; LOWE et al. 1992; MARTIENSSEN and BARON 1994). The knottedl (knl) gene was cloned by transposon tagging (HAKE et al. 1989a). The KNl protein contains 1276 B. Greene, R. Walko and S. Hake a homeodomain, suggesting that KN1 functions as a transcriptional regulator (VOLLBRECHT et al. 1991). KN1 is localized to nuclei of the apical meristem and a subset of cells in the stem; it is not found in wild-type leaf primordia (SMITH et al. 1992; JACKSON et al. 1994). The dominant K n l mutations affect cell differentiation specifically along the lateral veins of the leaf blade ( GELINAS et al. 1969; FREELINC and HAKE 1985). These cells adopt fates of cells normally found in more basal positions on the leaf (HAKE 1992). Ectopic KNl expression within developing leaves is correlated with the dominant phenotype (SMITH et al. 1992). The original allele, K n l 0 , is a tandem duplication of 17 kb (VEIT et al. 1990). Another allele, Knl -2Fl1 , results from the insertion of a Ds2 element (HAKE et al. 1989a), a non-autonomous element of the (Activator) Ac family [see FEDOROFF (1989) for review]. A third allele, K n l -N (gift of G. NEUFFER, University of Missouri) contains an rDt element in the fourth intron (HAKE 1992). In this report, we describe the isolation of nine Mucontaining K n l mutations. Significantly, the M u elements have all inserted in a discrete region of one of the introns. The insertions do not alter the size of the knl transcript, but result in ectopic expression of the transcript in leaves of mutant plants. At least two of these dominant alleles require Mu activity for expression of the knotted phenotype. We discuss mechanisms whereby suppressible insertions can produce dominant phenotypes. MATERIALS AND METHODS Genetic stocks: The K n l mutants arose in Mutator lines maintained collectively by S. HAKE and M. FREELING (HAKE et al. 1989b). Allelism was suggested by linkage of the knotted phenotype to a d h l , which is 1 map unit distal to knl on chromosome 1L. ADHl genotypes were determined by assaying scutellar slices (FREELING and SCHWARTZ 1973). Our use of nomenclature is according to the recommendations of the maize nomenclature committee (Maize Genetics Cooperation Newsletter 67: 171-73). The name and symbol of a gene locus is represented in lower case italics ( kn l ) , the dominant alleles are designated with the first letter of the symbol capitalized ( K n l ) . The gene product, either RNA or protein, is KN1. The dominant K n l mutations were found in Mu screens, each screen containing approximately 1500 families of 20 kernels each, or in outcross populations to generate seed for screening (HAKE et al. 198913). Knl-mum1 was found in the 1987 screen. Knl-mum2 was found in the outcross population of approximately 5000 Mu plants in 1988. Knl-mum3 was found in the 1989 screen; only tissue was taken, no seed was harvested. Knl-mum4 (90*620), Knl-mum5 (90*428) and Knl-mum6 (90*1234) were foundin the 1990 screen (seedwas obtained only from Knl-mum5) . Knl -mum7 (MF11333-1) and Knl-mum9 (MFl1333-2) were isolated from openpollinated Mutator stocks (5000 seeds planted). Knl-mum8 was a gift from N. SHEPARD (E.I. DuPont, Delaware). K n l mum10 (91*1433) was found in the 1991 screen. Ash bz, a deletion of both bronze1 ( b z l ) and shrunken1 ( s h l ) ( s h bz X 3 ) , was the primary non-Mu line used in our crosses (containing zero to a few inactive elements). It was a gift of J. MOTTINGER (University of Rhode Island). bz-Mum9 is a mutable allele of bzl that contains a M u 1 element (BROWN et al. 1989) inserted at the single 3' intron exon junction (DOSEFF et al. 1991). Itwas agiftfrom D. ROBERTSON (Iowa State University). Probes: The k n l genomic fragment (H2) was derived from the Knl-2F11 genomic clone as previously described (HAKE et al. 1989a). The entire k n l cDNA, pOC5 (VOLLBRECHT et al. 1991) was used as a probe for RNA blot hybridization. The ubiquitin probe, a 700-bp Pstl/SacI fragment from pskUBI, was a gift from P. QUAIL (University of California, Berkeley) (CHRISTENSEN et al. 1992). The Mul-specific probe was pA/B5 (CHANDLER et al. 1986), the Mu8-specific probe was an internal 550-bp PvuII/Ps~I fragment subcloned from the Mu8 element inKnl-muml.TheMu*probewasa1.3-kbEcoRI/BamHIfragment of MuDR-1 (CHOMET et al. 1991). Tissue dissections: Two tissue fractions were collected for RNA extraction, meristem enriched (ME) and leaf ( L ) fractions. To isolate ME and L fractions, 14day seedlings (approximately three leaves visible) were harvested by cutting at the seed. The oldest six leaves were discarded, and the next three leafprimordia (plastochrons 4,5 and 6, lengths of 0.5-2.5 cm) were collected, pooled and used for leaf RNA isolations. The meristem-enriched fraction included the leafbases, stem, meristem and the most recently initiated leaf primordia. Tissue from six to eight individuals was pooled for each sample. RNA isolation and northern blots: RNA extractions were performed as described previously (SMITH et al. 1992). Ten micrograms of total RNA in a volume of 5 pl were glyoxylated with 10 pl dimethyl sulfoxide, 2 pl 1OX running buffer (0.1 M NaPO,, pH 7.0), and 3 pl deionized glyoxal (ethandial, Sigma) and incubated at 50" for 60 min. Samples were electrophoresed through a 1 % agarose gel in 1 X running buffer and trans ferred to Nytran (Schleicher & Shuell) as described elsewhere (SAMBROOK et al. 1989). Deglyoxalization was performed by baking the filter at 80" for 2 hr followed by a brief wash in 200 mM Tris, pH 8.5. Hybridization was performed overnight as described (SMITH et al. 1992). DNA analysis: DNA isolations were done by grinding 1-2 g of seedling leaf tissue under liquid nitrogen and transferring the powder into a microcentrifuge tube containing 500 pl of urea extraction buffer (7 M urea, 30 mM NaCl, 5 mM Tris, 2 mM EDTA, 1% Sarcosyl, pH 8.0). Samples were vortexed gently, followed by extraction with pheno1:chloroform:isoamyl alcohol (25:24:1). Nucleic acids were precipitated from the aqueous phase with isopropanol following the addition of 0.1 volume of 4.4 M ammonium acetate. DNA samples were digested with the indicated enzymes according to the manufacturer's instructions, fractionated on 0.8% agarose gels, transferred to Nytran nylon membranes (Schleicher & Shuell) and W cross-linked to immobilize the digested DNA. The filters were hybridized at 68" in a Robbins Scientific Hybridization Oven according to the manufacturer's suggestions. Genomic DNA from the Mu8-containing K n l alleles was polymerase chain reaction (PCR)-amplified using Promega Taq DNA polymerase after digestion with PvuII. The primers were: PJ2 5' GATCGATTCCA'ITTGGAATG 3', contained within the k n l third intron, and Mu15A-5 5' GTCATCGTC CAGAACTCGGA 3', designed from internal Mu8 sequences (FLEENOR et al. 1990). Genomic DNA from the Mulcontaining alleles was first digested with BstEII prior to amplification. The M u end primer, MuE2 5' GCGAA'ITCCATAATGGCAATTATCTC 3', a gift from B. K. GRUISSEM and M. FEELING, was used in combination with PJ2. Conditions were 25 cycles of 1 min, 94"; 1 min, 50"; and 1.5 min, 72", in the buffer as suggested by the manufacturer, except 2 mM dGTP and 2 mM 7 deaza-dGTP (Sigma) were included in reactions Mutator-Induced K n l Mutations 1277 with the Mu15A-5 primer. PCR products corresponding to ne end of the Mu insertion site were either directly sequenced or were filled in with Klenow and ligated into the EcoRV site of pBluescript SK' (Stratagene). Sequencing was according to CHI et at. (1988).