Logo of narLink to Publisher's site
Nucleic Acids Res. 2004; 32(6): e54.
Published online 2004 Apr 1. doi:  10.1093/nar/gnh052
PMCID: PMC390377

Molecular analysis of high-copy insertion sites in maize


High-copy transposon mutagenesis is an effective tool for creating gene disruptions in maize. In order to molecularly define transposon-induced disruptions on a genome-wide scale, we optimized TAIL-PCR to amplify genomic DNA flanking maize Robertson’s Mutator insertions. Sample sequencing from 43 Mutator stocks and the W22 inbred line identified 676 non-redundant insertions, and only a small fraction of the flanking sequences showed significant similarity to maize repetitive sequences. We further designed and tested 79 arbitrary primers to identify 12 primers that amplify all Mutator insertions within a DNA sample at 3.1-fold redundancy. Importantly, the products are of sufficient size to use as substrates or probes for hybridization-based identification of gene disruptions. Our adaptation simplifies previously published TAIL-PCR protocols and should be transferable to other high-copy insertional mutagens.


Insertional mutagenesis is commonly used to disrupt genes and to study their function. In model organisms, there are multiple projects to create collections of single-gene disruptions at a genome-wide scale (reviewed in 13). Since many species have evolved with multiple duplications of their genomes (4), single-gene disruptions often show no or subtle phenotypes due to genetic redundancies unique to the organism (5,6). In contrast, phenotypic mutants reveal the non-redundant or low-redundancy genes that are most informative about the biological function of a novel sequence. High-copy insertions are more efficient mutagens, allowing screens of smaller populations for phenotypes of interest (reviewed in 7). However, this creates the molecular problem of identifying specific gene disruptions from a high-copy background.

In both plants and animals, inverse or adapter-mediated PCRs are commonly used to recover high-copy insertion flanking sequences (812). These methods rely on restriction digests, ligations and affinity purification steps to identify insertion sites. Importantly, methods that rely on restriction digests can only amplify flanking sequences that have the restriction site within an appropriate distance of the insertion. In maize, inverse PCR can require prescreening by DNA gel blots in order to identify appropriate restriction enzymes and DNA fragments that can be cut from an agarose gel for amplification (13). In contrast, many adapter PCR methods utilize restriction enzymes with 4-bp recognition sequences resulting in short flanking PCR products (average sizes 150–250 bp). These products can then be displayed on polyacrylamide gels to identify flanking products for gel elution and sequencing (8,9). In model organisms with complete genome sequences, short sequences can be easily used to identify insertion sites by database searches (14). When these methods are applied to incompletely sequenced organisms, such as maize, short flanking sequences most likely will not be associated with known sequences (9).

In order to identify larger transposon flanking sequences, the Maize Gene Discovery Project developed RescueMu, which is a transgenic Robertson’s Mutator (Mu) transposable element (reviewed in 15). The RescueMu transgene includes a plasmid construct that allows genomic sequences associated with RescueMu insertions to be recovered by direct transformation of Escherichia coli. This system has been used to sequence a large number of somatic and heritable insertion sites (~65 000 clones), but it is limited to analyzing transposon insertion events that are caused by the transgene. RescueMu can only transpose when the transgene is crossed into actively transposing Mu genetic stocks. Mu transposons typically accumulate between 50 and 200 copies within these Mu-active plants (16). However, the RescueMu transgene is found at 1–2 copies when propagated within Mu-active stocks. Therefore RescueMu plasmids sample a small fraction (~5%) of the heritable transposon insertions found in the plant and are unlikely to be associated with phenotypic mutants.

In Arabidopsis, several groups have systematically sequenced insertion sites from T-DNA mutagenized populations, which typically insert at one to two loci per transformation event. Both adapter-mediated and thermal asymmetric interlaced PCR (TAIL-PCR) have been used to amplify T-DNA insertion sites for sequencing templates (14,17). TAIL-PCR requires fewer manipulations to amplify larger flanking sequences from T-DNA insertions (18). TAIL-PCR utilizes both nested specific primers and degenerate arbitrary primers as well as alternating high and low annealing temperature cycles to amplify sequences neighboring a known sequence. The small number of manipulations with TAIL-PCR allowed a high-throughput approach to identifying >99 000 T-DNA knockouts in Arabidopsis (17). This application of TAIL-PCR is equivalent to amplifying a single-copy transposable element using primers specific to terminal inverted repeat sequences. However, it is not clear a priori whether the TAIL protocol can be directly applied to more complex systems. The maize genome is significantly more complex than Arabidopsis with ~2.5 × 109 bp (19), and in complex pools of insertions TAIL reaction conditions could lead to non-specific amplification products. Furthermore, TAIL-PCR may only amplify a small fraction of the flanking sequences found in a complex pool.

We have adapted and optimized TAIL-PCR to amplify genomic sequences associated with maize Mu insertion sites (MuTAIL). We have developed and characterized nested primers specific for the terminal inverted repeat (TIR) sequence of the Mu family of DNA transposons. Sample sequencing experiments from a large number of Mu lines indicate that MuTAIL faithfully amplifies germinal insertion flanking sequences in a variety of pooling strategies. In addition, we developed and tested 79 arbitrary primer sequences to select a set of 12 optimized primers for maize that improve the representation of all Mu insertions within a DNA sample. Hybridization and additional sequence analysis indicates that these 12 primers amplify 80–95% of the transposons found in a complex pool of insertions. In contrast with the RescueMu system, our MuTAIL method identifies native transposon insertion sites and is far more likely to identify an insertion event associated with a phenotype. Our adaptation should be transferable to other high-copy insertion systems or for sampling complex DNA pools from low-copy insertional mutagenesis populations.


Genetic stocks for sequencing

Mutator insertion sites were sequenced from 44 genotypes as follows: the color-converted W22 inbred, 13 diverse Mu-active lines, seven parental lines from the UniformMu population and 23 UniformMu transposon-inactive lines. In all cases, 5–15 plants were pooled for DNA sampling. Transposon activity was determined either by the presence of a mutable phenotype that reverts when transposase is expressed, or based on the genetic relationship of the individual stock to lines that produce novel mutations due to Mu transposons.

The 13 diverse Mu-active lines each segregated for a single visible mutant phenotype including a1-mum2 (a1-mum2 library), a dwarf mutation (d1 library), a gametophytic small kernel mutation (sk1 / 98F-79-2), two ectopic auricle mutations (eco, Wab), a big embryo mutation (be), two defective kernel mutations (407, 96-603) and five viviparous mutations (3286, 99F-263B, 92-74-2, 92-66-3 and 92-80). The majority of these lines were not genetically marked for somatic Mu activity, but the mutations arose in Mutator populations.

The UniformMu population is a Mutator line with the mutable bz1-mum9 anthocyanin biosynthetic gene introgressed into the W22 genetic background and will be described in detail elsewhere. The parental lines were selected to have somatic transposon activity based on reversion of the bz1-mum9 locus in the endosperm. A Mu-active plant shows sectors of purple color in a bronze-colored endosperm. The seven UniformMu parental libraries were 99F-245, 99F-240, 99F-235, 99F-232, 99F-220, 99F-217 and 99-060.

The UniformMu transposon-inactive lines did not show reversion of the bz1-mum9 locus, and all lines segregated for independent seed mutant phenotypes. The 23 inactive libraries included embryo defective (emb), rough endosperm (rgh), small kernel (smk), shrunken (sh) and other defective kernel (dek) mutants in the following libraries: 00-112-2 (emb), 00-113-1 (emb), 00-116-1 (rgh), 00-118-1 (smk), 00-119-1 (rgh), 00-122-1 (smk), 00-124-1 (smk), 00-129-1 (rgh), 00-132-1 (emb), 00-133 (dek), 00-136 (dek), 00-137 (dek), 00-140A (dek), 00-141S (rgh), 00-143B (rgh), 00-144 (emb), 00-145-2 (dek), 00-146B (dek), 00-147B (sh), 00-148 (smk) and 00-149 (dek).

Genetic stocks for hybridization

For the 10 locus hybridization assay, transposon insertion sites were analyzed from 10 divergent stocks that had Mu insertions in known genes. These stocks included the following alleles: rs2-mum2 (20), hcf106-mum3 (21), cr4 (22), vp1-mum1 (23), vp14-2274 (24), a2-mu1 (25), bz2-mu1 (26), bz1-mu3 (V. Walbot, personal communication) and a1-mum2 (described in 27). The bc96 stock has a Mu insertion in a single-copy sequence that was cloned from a lambda genomic library (B.C.Tan, personal communication).


The MuTAIL-PCR protocol was modified from Liu et al. (18), and the primers used in this study are shown in Table Table1.1. DNA was extracted from seedling and adult leaf samples by grinding 2 g of tissue in 5 ml of urea DNA extraction buffer (168 g urea, 25 ml of 5M NaCl, 20 ml of 1M Tris–HCl pH 8, 16 ml of 0.5M EDTA pH 8, 20 ml of sarkosine and 190 ml of H2O). Each extract was mixed with phenol–chloroform–isoamyl alcohol (25:24:1) for 15 min and the phases separated by centrifugation (10 min at 2000 g). The aqueous phase was transferred and mixed with 500 µl of 3M sodium acetate pH 5.2 and 4 ml isopropanol. The precipitated DNA was washed with 70% ethanol and resuspended in TE (10 mM Tris–HCl pH 8, 1 mM EDTA pH 8). Each DNA sample was diluted to ~50 ng/µl in water and 1–2 µl were added to a 20 µl primary MuTAIL-PCR. The primary MuTAIL-PCR utilized Invitrogen or GIBCO-BRL Taq polymerase and buffers, and was composed of 20 mM Tris–HCl pH 8.4, 50 mM KCl, 2 mM MgCl2, 200 µM of each dNTP, 1 µM arbitrary primer, 100 nM TIR6 primer and 1 U of Taq polymerase. The reaction was incubated using the thermocycling conditions shown in Table Table22 for primary MuTAIL-PCR. The primary reaction was then diluted 1:50 in water and 1 µl was added to a 20 µl secondary reaction. Water-only negative controls from the primary reaction were diluted and processed identically to reactions containing DNA templates. The secondary MuTAIL reaction was identical to the primary except that the TIR4 or TIR8 nested primer was used in place of the TIR6 primer and the reaction was thermocycled as shown for the secondary reaction in Table Table2.2. The TIR8 primer was composed of the TIR8.1, TIR8.2, TIR8.3 and TIR8.4 primers mixed in a 2:4:4:1 ratio, respectively, to adjust for the degeneracy of each primer synthesis. An independent reaction was completed for each arbitrary primer and template DNA combination.

Table 1.
Primers used in this study
Table 2.
MuTAIL-PCR thermocycling conditions

MuTAIL microlibrary sequencing

MuTAIL-PCR products for each genomic DNA sample were pooled from a maximum of six arbitrary primers. The products were size selected using a Sephacryl-400 (Promega) spin column following the manufacturer’s instructions, except that a 0.5 ml bed volume of resin was used for each column. This column reduced the concentration of products smaller than 500 bp. The size-selected products were then cloned into the pCR4-TOPO vector (Invitrogen) following the manufacturer’s instructions. Up to 96 colonies for each ligation were randomly selected for sequencing at the UF-ICBR genome core facility. Plasmid sequencing templates were prepared using Qiagen Turbo Miniprep purification kits on a Qiagen Biorobot 9600. The sequencing chemistry was Amersham ET-terminators. All sequencing reactions were separated on an Amersham Biosciences MegaBACE 1000 DNA sequencer.

Sequence assembly

The single-pass sequences were filtered for the pCR4-TOPO vector sequence and Mu-TIR sequences and assembled into contigs using PHRED, CrossMatch, a custom quality trimmer and PHRAP (28,29). The quality trimmer used a 50 bp sliding window that removed 5′ or 3′ bases if five or more bases in the window had a PHRED quality score of ≤10. The assembly was edited using Consed (30) and results from pairwise BLAST searches (31). The Mu-TIR sequences were added to singleton reads by the repeating base-calling, quality trimming and vector masking steps using PHRED, quality trimmer and CrossMatch without Mu-TIR sequences included in the vector file. For contigs, the Mu-TIR sequences were hand-edited based on the sequence chromatograms. The flanking sequences described here are available in the DDBJ/EMBL/GenBank GSS database, accession numbers BH417702-BH418377, CC050461-CC050722 and CG785063-CG786032.

Database searches

The non-redundant MuTAIL sequences were used to search the DDBJ/EMBL/GenBank GSS and EST databases using BLASTN and the DDBJ/EMBL/GenBank NR protein databases using BLASTX (31). Matches were considered significant at E-values <10–5 for NR, BLASTX searches, E-values <10–6 for EST, BLASTN searches and E-values <10–25 for GSS, BLASTN searches.

DNA gel blot analysis of Mu copy number

Approximately 15 µg of each genomic DNA sample was restriction digested in separate reactions with HindIII and XhoI. The digested DNA samples were size fractionated on 0.8% TBE agarose gels and blotted to Hybond-N nylon membranes (Amersham) using standard capillary transfer conditions. Hybridization probes were prepared by gel purification of Mu element specific fragments. The MuDR probe was the 1.5 kbp EcoRI/BamHI fragment from pBMP1.3 (27). The Mu1 probe was the 650 bp EcoRI/HindIII fragment from pA/B5 (32). The Mu3 probe was the 1 kbp EcoRI/HindIII fragment cloned by Oishi and Freeling (33). The Mu7 probe was the 1 kbp EcoRI fragment from pSBH6 (34). The Mu8 probe was the 800 bp EcoRI/HindIII fragment cloned by Fleenor et al. (35). All probe fragments were purified by agarose electrophoresis and extracted with Qiagen purification kits following the manufacturer’s instructions. The fragments were labeled with Amersham Ready-to-Go labeling kits. The blots were hybridized with Church hybridization buffer (1% BSA, 1 mM EDTA pH 8.0, 0.5 M Na2HPO4 pH 7.2 and 7% SDS) overnight at 65°C and washed once with 2× SSC, 0.1% SDS, and three times with 0.2× SSC, 0.1% SDS at 65°C for 20 min each wash. The probe–digest combinations that showed the best resolution of the Mu hybridizing bands were chosen for Figure Figure33.

Figure 3
Genomic DNA gel blots hybridized with Mu element specific probes. Genomic DNA that was used as the template for the MuTAIL sequencing libraries was digested with HindIII (A–D) or XhoI (E). Lanes 1–11 are the template DNA samples for libraries ...

DNA gel blot analysis of insertion clones

The 99F-220 and 99F-232 parental cultures contained F1 progeny from crosses of W22 by the Mu-active parents 99S-82-2 and 99S-102-2, respectively. Twelve F1 plants from each culture were pooled for MuTAIL-PCR, cloning and sequencing. Each of the F1 plants was self-pollinated, and 15 F2 progeny from each self-pollination were germinated. DNA was extracted for each self-pollination from F2 seedling leaf tissue, pooled by F1 individual. Approximately 15 µg of each genomic DNA sample was restriction digested in separate reactions with HindIII and SstI. The digested DNA samples were size fractionated on 0.8% TBE agarose gels, blotted and hybridized as described above. Probe fragments were purified by PCR amplification using the T3 and T7 primers or by agarose gel purification after plasmid digests with EcoRI. MuTAIL-PCR products were scored as polymorphic only if they showed polymorphisms with both digests.

Ten-locus hybridization assay

Genomic DNA was extracted from the genetic stocks rs2-mum2, hcf106-mum3, cr4, vp1-mum1, vp14-2274, a2-mu1, bz2-mu1, bz1-mu3, bc96 and a1-mum2. MuTAIL-PCR products were amplified from the 10 DNA templates using each of the arbitrary primers shown in Table Table1.1. The products and 1 ng of the cloned gene insert were separated on 1% TBE agarose gels and were blotted, hybridized and washed as described above. Positive signals were scored after an overnight exposure to X-ray film.

Mu-TIR analysis

The Mu-TIR sequences were identified using CrossMatch and a custom Java program. All Mu-TIR sequences analyzed passed the quality trimmer step in the sequence assembly. Identical Mu-TIR sequences were clustered using BLASTCLUST (31) (ftp://ftp.ncbi.nih.gov/blast), and the non-redundant Mu-TIR sequences were hand-aligned and clustered using DNAML in the Phylip software package (36). The Mu-TIR phylogenetic trees were constructed with a representative TIR sequence from each cluster.


MuTAIL amplifies transposon insertion sites

When used in MuTAIL, six of the published Arabidopsis arbitrary primers (18,37) yield diverse large products from Mu lines (Fig. (Fig.1A).1A). To determine if MuTAIL-PCR faithfully amplified genomic sequences associated with Mu insertion events, we cloned and sequenced MuTAIL-PCR products from 44 genotypes including 13 diverse actively transposing lines, 30 lines from the UniformMu population (see Materials and Methods) and the W22 color-converted inbred. The 30 UniformMu lines included seven somatic Mu-active parental lines and 23 seed mutant lines that did not display transposon activity as scored by somatic reversion of the anthocyanin kernel marker bz1-mum9. Each of the 44 genotypes were used to construct libraries of 96 MuTAIL-PCR clones. The clones were single-pass sequenced and the sequences were assembled. Initially, MuTAIL-PCR products were cloned directly from 12 of the genotypes including the W22 inbred, eight of the diverse Mu lines and three UniformMu parental lines. After vector masking and quality trimming, the average sequence length was 226 bp. In order to suppress cloning bias for small insert clones, the remaining 31 genotypes were cloned after passing the products over a size-exclusion column, increasing average sequence read length to 439 bp (Fig. (Fig.1B).1B). In total, there were 3341 passing reads that assembled into 676 non-redundant sequences with 253 contigs and 423 singletons.

Figure 1
(A) MuTAIL-PCR products from the leaf mutant, ectopic auricle (eco). Products were amplified from a pool of eco homozygous plants using the AD1, AD2, AD3, AD10, AD20 and W4 arbitrary primers (lanes 1–6). The AD10 and AD20 primers frequently gave ...

The nested Mu-TIR specific primers were designed to amplify 33 bp of sequence from the Mu element in addition to the genomic sequence flanking the insertion. A correctly amplified Mu flanking sequence would be expected to include 33 bp of Mu-TIR sequence. We analyzed the 676 non-redundant sequences and identified 354 Mu-TIR sequences. Because most of the non-redundant flanking sequences derive from single-sequence reads and average MuTAIL product sizes are larger than average read lengths, we would expect that many would not include the full sequence of the MuTAIL-PCR product. Therefore we further examined a subset of complete PCR product sequences within the 676 non- redundant flanks for Mu-TIRs. Of a set of 215 full-length MuTAIL product sequences, 88% had Mu-TIR sequences (189 sequences), suggesting that MuTAIL products are primarily transposon flanking sequences.

BLAST searches revealed that the majority (406/676) of the MuTAIL sequences have significant similarity to sequences within the DDBJ/EMBL/GenBank GSS database. These matches resulted from 100 bp or greater identity with maize genomic sequences that are predominantly from methyl-filtered and high cot clones of the Maize Genomics Consortium sequencing project (38,39). In addition, 35% of the non-redundant sequences (233/676) have similarity to sequences in the EST database, while only 13% of these sequences (91/676) show similarity in BLASTX searches of the NR protein database. Importantly, only 22 of the 676 sequences show similarity to retrotransposons that comprise the bulk of maize repetitive DNA (40), indicating that the MuTAIL-PCR products are highly enriched for low-copy sequences. This result is consistent with the insertion bias of Mu elements for low-copy sequences (41). Interestingly, eight of the 22 MuTAIL sequences similar to retrotransposons have clear Mu-TIR sequences, suggesting that these are bona fide transposon insertion sites into repetitive DNA. Consistent with this result, a low frequency of RescueMu sequences were also found to be inserted in retrotransposons (15).

Each library had a characteristic number of non-redundant sequences that correlated with genotype and Mu activity state. Transposon-active lines yielded a larger number of non-redundant sequences (Fig. (Fig.2).2). Interestingly, the libraries had a similar number of shared sequences that were found in all lines, suggesting that there are common ancestral Mu insertions found in a large percentage of Mutator lines. This result is consistent with the common origin of active Mutator populations from Don Robertson’s stocks, first reported in 1978 (42). The increased number of insertions found in libraries from transposon-active lines were due to a higher percentage of unique transposon flanking sequences. Libraries from UniformMu transposon-active lines had 57% unique sequences, while transposon-inactive libraries averaged only 36% unique sequences.

Figure 2
Average number of non-redundant sequences found in the MuTAIL-PCR microlibraries. White bars indicate average numbers of sequences that were shared between two or more genotypes within the set of 676 non-redundant sequences. Gray bars show the mean number ...

To determine the relative copy number of Mu elements in the diverse and UniformMu lines, we tested the W22 inbred and 10 of the 43 genotypes by DNA gel blot (Fig. (Fig.3).3). The diverse Mu lines show a large number of unique hybridizing fragments with all the Mu element specific probes tested. The UniformMu lines show a higher percentage of shared bands with many of these bands shared with the W22 inbred. Interestingly, the W22 inbred shows a low number of Mu1, Mu7 and Mu8 hybridizing fragments (Fig. (Fig.3B,3B, D and E, lane 1), while having a relatively large number of MuDR-related and Mu3 elements (Fig. (Fig.3A3A and C, lane 1). Based on these and other DNA gel blot data (not shown), we estimate that the W22 inbred contains 20–30 Mu elements. Both the UniformMu mutant and parental lines show increased numbers of Mu elements relative to the W22 inbred, but exact copy numbers are difficult to determine due to the complexity of the hybridizing bands. These hybridization results indicate that the number of unique MuTAIL clones correlates with the complexity of the transposon insertions within the DNA template. The higher percentage of unique flanking sequences we observed in transposon-active lines could be caused by a larger diversity of germinal insertions, by somatic insertion events or by a mix of both types of events.

MuTAIL amplifies germinal insertion sites

To determine if the unique sequences are amplification products from somatic or germinal insertion events, we tested eight single-copy probes by genomic DNA gel blot. The probes were selected from the unique flanking sequences of the 99F-220 and 99F-232 UniformMu parental lines. These two MuTAIL libraries were generated from a pool of 12 F1 siblings from crosses of W22 by transposon-active male parents. Germinal insertions in the F1 plants are expected to be inherited in the F2 progeny. We extracted and pooled DNA from 15 F2 seedlings from each of the 12 F1 siblings in the 99F-220 and 99F-232 lineages. The F2 DNA pools were tested for polymorphisms by DNA gel blot (e.g. Fig. Fig.4).4). Four of the unique sequence probes segregated for polymorphisms in a 1:1 ratio consistent with germinal insertions inherited from the F0 transposon-active parent. Two of the probes had polymorphisms found in single F2 families, consistent with novel germinal insertions within a single F1 plant, while the remaining two probes did not show any polymorphisms, indicating that these flanking sequences probably derived from somatic insertion events. These data suggest that the increased number of unique sequences found in the transposon-active libraries are partially due to amplification of somatic insertions. Importantly, the majority of the transposon flanking sequences amplified by MuTAIL (six of eight tested) represent germinal insertion events even in Mu-active backgrounds, and somatic insertions should be readily suppressed by selecting Mu-inactive lines for template DNA.

Figure 4
DNA gel blots probed with cloned MuTAIL-PCR products. (A) DNA gel blot of HindIII digested genomic DNA extracted from W22 (lane 1) and F2 progeny of the 99F-232 UniformMu parental pool (lanes 2–13). The blot is probed with the insert from MuTAIL ...

Representation of MuTAIL amplified products

The sequencing and DNA gel blot results above indicate that TAIL-PCR faithfully amplified diverse transposon flanking sequences in complex pools of insertion sites. However, these data do not quantify the efficiency of MuTAIL for amplifying all of the transposon insertion sites found within a DNA sample. To estimate the population size of insertions recovered by MuTAIL-PCR, we analyzed the frequency of matching host site duplications in the Mu-TIR flanking sequences.

Mu transposons create a 9 bp target site duplication during transposition to form a 9 bp direct repeat flanking each Mu-insertion site (43). During MuTAIL-PCR, each transposon can independently produce a left and a right flanking sequence. We analyzed the 9-bp sequences adjacent to the Mu-TIRs identified in our sequences. These predicted 9-bp host-site duplications were used to identify pairs that are reverse complement matches. A reverse complement match indicates sequences left and right of the same insertion site were recovered. The frequency of reverse complement matches estimates the probability of recovering both the left and right side of the transposon insertion, or f(matches) = p(L) × p(R). Assuming that MuTAIL-PCR, cloning and sequencing randomly samples the flanking sequences, each flanking sequence has the same unknown probability of being cloned and sequenced. Thus, p(L) = p(R) and the frequency of reverse complement matches equals the square of the frequency of recovering any one Mu-flanking sequence. For example, assuming MuTAIL recovers 50% of the flanking sequences within a genome, the probability of recovering two flanking sequences with matching reverse complement host-site duplications equals the probability of recovering the left flanking sequence (0.5) multiplied by the probability of recovering the right flanking sequence (0.5), or f(matches) = (0.5) × (0.5) = (0.5)2 = 0.25.

In order to estimate the probability of recovering a flanking sequence, we identify the non-redundant 9-bp duplications within a genotype and determine the frequency of reverse complement matches. If the MuTAIL sequences from a genotype contain 16 duplications and four of these duplications have a reverse complement match, we would estimate that 50% of the transposon flanking sequences within the line were sampled, or (4 matches/16 duplications) = 0.25 = (0.5)2. Thus the square root of the frequency of matching duplications estimates the probability of cloning and sequencing a single flanking sequence from the genomes sampled.

In our sequencing study with the W22 inbred and 43 Mu lines, we identified an average of 16 Mu-TIR sequences per line, but only 16 lines had one or more pairs of matching duplications. However, the MuTAIL products from each line were not exhaustively sampled, because we only recovered ~80 sequences for each line. The entire collection of MuTAIL sequences is a more representative sample of the MuTAIL-PCR products from all 44 lines. The set of 354 unique TIR-containing sequences included 28 matching duplications, giving an estimate of 28% representation of all flanking sequences. We can further calculate the number of transposon insertions that would be represented with at least one flanking sequence as the probability of identifying either the left or the right flanking sequence minus the frequency of recovering both sides of a transposon insertion site, or p(L) + p(R) – p(L) × p(R). In practice, this estimate is calculated from the same frequency of matching duplications as [2 × (matches/duplications)1/2 – (matches/duplications)]. By this measure, the collection of 354 sequences with Mu-TIRs represents 48% of the total population of transposon insertions sampled by MuTAIL-PCR within the 44 genotypes.

This sequence-based estimate of efficiency relies on the assumption that each Mu insertion site has the same random probability of being cloned and sequenced. In order to obtain an independent estimate of the fraction of all insertions represented solely by MuTAIL-PCR, we developed a hybridization assay using 10 genetic stocks that have Mu transposon insertions in cloned maize genes. These genetic stocks were not used for cloning and sequencing. Rather, MuTAIL was used to amplify transposon flanking sequences from each mutant DNA sample, and the products were blotted and hybridized with their respective cloned genes. The efficiency of a MuTAIL reaction is estimated as the fraction of these genes that produce a hybridization signal. A hybridization signal is detected if either the left or right side of the known insertion site is amplified, because all the probes used in this assay spanned the known insertion sites. Figure Figure55 (lanes A–F) shows a 10-locus assay with each of the six arbitrary primers developed for Arabidopsis. Each arbitrary primer amplifies one to three of the 10 loci, and in combination the set of six primers amplify 50% of the loci tested. When we repeated this hybridization assay, we found that some loci are amplified consistently with specific arbitrary primers, while others were amplified stochastically. In four replicate 10-locus hybridization assays, only three of the 10 loci (a2, cr4 and bc96) were amplified consistently and at high redundancy (summarized in Table Table3).3). In contrast, the vp1, a1 and bz2 loci were amplified at a lower redundancy, and usually were amplified by a single arbitrary primer in a given replicate assay. Importantly, four of the 10 loci were never amplified, suggesting that a subset of maize genes do not amplify with this adaptation of TAIL-PCR.

Figure 5
Ten-locus hybridization assay. DNA gel blots of MuTAIL-PCR products amplified from each of 10 DNA templates containing known Mu-tagged loci were probed with their respective gene sequences. All signals were scored after an overnight exposure. Lanes A–F ...
Table 3.
Known loci amplified with Arabidopsis arbitrary primers in MuTAIL-PCR

Both the 9 bp duplication analysis and the independent 10-locus hybridization assays suggest that 30–50% of the Mu insertion sites are represented by MuTAIL-PCR using the existing set of arbitrary primers. Convergence of these estimates indicates that the assumptions underlying the matching duplications analysis are valid. Therefore identifying matching target site duplications provides a useful estimate of the proportion of Mu-flanking sequences recovered by random sequencing of cloned MuTAIL-PCR products. Both the hybridization and matching duplication estimates of amplification and recovery indicate that this initial MuTAIL protocol is not optimal for amplifying maize genic sequences.

Optimization of MuTAIL-PCR

In order to improve MuTAIL-PCR, we addressed two potential sources of bias that may limit the efficiency of MuTAIL: (i) the Mu-specific primers for the primary and secondary MuTAIL reactions have significant overlap and may cause some bias in Mu element classes that are amplified; (ii) the arbitrary primers originally developed for Arabidopsis may not amplify all maize genic sequences robustly.

To optimize the efficiency of the Mu-specific primers, we replaced the nested TIR4 primer, which was originally selected to avoid a degenerate CT-rich region 33–21 bp 5′ from the end of the Mu-TIR. Owing to this design constraint, the TIR4 primer has a 20 bp overlap with the primary MuTAIL specific primer (TIR6). To reduce the overlap between TIR6 and TIR4, we designed a series of four degenerate primers that share 15 bp with TIR6 and have a greater potential specificity for known Mu-transposable elements. This new secondary MuTAIL primer, TIR8, was designed to retain 29 bp of downstream Mu-TIR sequence in order to confirm that the amplified products are authentic Mu-flanking sequences.

To develop improved arbitrary primers for maize, new primers were designed with higher GC content and tested with the 10-locus hybridization assay. In addition, we developed an algorithm to select degenerate sequences that are enriched in maize genes. The algorithm consisted of a random sequence generator that developed 2 million 15-base oligomers with 128-fold degeneracy. These oligomers were then compared with a random sample of 1 million 15-base sequences derived from the maize Unigene EST assemblies (http://www.maizegdb.org/zmdb.php/). The Unigenes consisted of 13 000 sequences that total 8 Mbp of genic sequence with all vector and adapter sequences removed. The 100 most over- represented oligomers were examined for sequence similarities, and a diverse subset were selected to be synthesized and tested. In total, we screened 79 arbitrary primers to identify a set of 12 primers that were tested in four replicate 10-locus assays (Table (Table4).4). It is important to note that none of these arbitrary primers were designed to specifically amplify any of the 10 loci used in the hybridization assay. Similar to our initial hybridization results, we found that the set of 12 optimized arbitrary primers amplified the loci in a stochastic manner. However, the optimized primer set consistently amplified 80% of the loci in contrast with 30% with the Arabidopsis arbitrary primers (Fig. (Fig.5,5, lanes G–R). When the four replicate experiments are considered together (Table (Table4),4), the 10 loci were amplified with 3.125-fold redundancy [125 signals/(10 loci × 4 replicates)]. Theoretically, these hybridization data indicate that the 12 optimized primers amplify 95.6% of the Mu insertion sites within a genome [1 – (1/e3.125) = 95.6% chance of amplification].

Table 4.
Known loci amplified with optimized arbitrary primers in MuTAIL-PCR

To determine if optimized MuTAIL improved sampling efficiency of Mu elements in general, we sequenced MuTAIL-PCR products generated using the optimized protocol. We measured the frequency of matching 9 bp duplications from three genotypes that were not included in the 10-locus hybridization assay. The TIR8 primer was used in combination with the 12 optimized arbitrary primers to amplify, clone and sequence from the W22 inbred, a defective kernel mutant (00S-119-01) and a small kernel mutant (98F-79-06). None of these genotypes were used to select the optimized primers, but all three genotypes were included in the original sequencing study. As shown in Table Table5,5, comparisons of the initial and optimized primer sets reveal a marked increase in the relative frequency of matching 9 bp duplications recovered using the optimized protocol. The proportion of matching duplications indicates that we sampled 67–86% of the Mu insertion sites within the genotypes we tested. The frequency of matching duplications can also be used to estimate the population of Mu elements that were sampled by MuTAIL, cloning and sequencing. Importantly, the copy number estimates based on the sequenced insertion sites correlate with the expected number of insertions within each line. The non-Mutator inbred W22 had the smallest number of Mu-TIRs and the largest proportion of matching duplications, and is estimated to have 21 Mu insertions based on MuTAIL sequencing data. Based on DNA gel blot data (Fig. (Fig.33 and data not shown), we estimate that the W22 genome contains between 20 and 30 Mu-related sequences. Similarly, the transposon-inactive Mu stock 00S-119-01 had the greatest number of Mu-TIRs and the smallest proportion of matching duplications and is expected to have the highest Mu copy number based on its pedigree. The 98F-79-06 line is expected to have an intermediate Mu-copy number, because it was selected to be transposon-inactive and then was crossed to the W22 inbred.

Table 5.
Analysis of Mu insertion site duplications recovered by MuTAIL-PCR

Optimized MuTAIL amplifies diverse classes of Mu-TIRs

To determine if optimized MuTAIL had an effect on the classes of Mu elements recovered by MuTAIL, we further examined the Mu-TIR sequences from the 44 genotypes amplified with the initial protocol and the three genotypes that we amplified using the optimized protocol. Both groups of Mu-TIR sequences were clustered using maximum likelihood methods (36) and classified according to the closest published Mu-TIR (Fig. (Fig.6A6A and B). Because the TIR8 primer is shifted 6 bp towards the end of the TIR, some resolution of the Mu element classes was lost for TIR8 amplified products. Interestingly, we found a small percentage of sequences in both samples that appear related to Mu elements but are divergent 33–17 bp 5′ from the end of the TIR. The percentage of divergent Mu elements recovered did not change with our optimized MuTAIL protocol. These results suggest that there are divergent Mutator-like transposons within maize and are consistent with observations of MuLEs found in other plant species (4446). However, we did find an enrichment of Mu7 and Mu8 TIR sequences relative to Mu1 and Mu4 sequences in the products amplified with the TIR4 nested primer (Fig. (Fig.6C).6C). Thus it is possible that the optimized MuTAIL protocol presented here more accurately represents all Mu element insertions.

Figure 6
Comparison of Mu-TIR sequences amplified with the TIR4 and TIR8 nested primers. (A) Maximum likelihood phylogenetic tree of Mu-TIR sequences recovered with the TIR4 primer. The tree is based on the nine non-redundant 33 bp terminal ends of the 5′ ...


We have shown that an optimized MuTAIL-PCR protocol is highly effective for molecular analysis of high-copy Mu-transposable elements in the maize genome. We have sequenced from a diverse set of genotypes to show that the method can be applied to DNA samples with a wide range of copy numbers, which include samples containing pooled DNA from up to 15 highly Mu-active individuals. Hybridization and sequencing experiments indicate that our optimized MuTAIL protocol accurately amplifies the vast majority of Mu insertion sites within a complex pool for the following reasons. First, 88% of the non-redundant full-length sequences have authentic Mu-TIRs. Secondly, only 3.3% of the sequences derive from obvious maize repeat sequences, which is consistent with the known bias of Mu to insert in non-repetitive sequences. Thirdly, the diversity of sequences recovered correlates with total Mu-copy number within template DNA samples; lower-copy genotypes yield fewer non-redundant sequences than higher-copy samples. Fourthly, optimized MuTAIL amplifies flanking sequences from all genes in a set of 10 known Mu-disrupted loci at 3.1-fold redundancy. Finally, analysis of 9-bp target site duplications indicates that single-pass sequencing of a set of 384 randomly selected MuTAIL clones is sufficient to identify >67% of the transposon insertion sites in a single genome. Together, these data suggest that MuTAIL is an effective method for identifying genes disrupted by Mu elements within a genome or pool of genomes.

However, our data indicate that MuTAIL amplifies transposon insertions in a stochastic manner. The hybridization data show that a given locus will not necessarily be amplified by a specific arbitrary primer at 100% frequency. Instead, most loci will be amplified with two to four of the optimized primers in a given experiment. These results illustrate both strengths and limitations to the MuTAIL method. MuTAIL cannot be used to identify specific insertion sites as cosegregating bands with gel electrophoresis because a specific insertion site will not necessarily amplify as the same size fragment or with the same arbitrary primer in multiple related genotypes. In addition, several of the optimized arbitrary primers amplify a smear of products, making it difficult to resolve specific bands with these primers. Thus the MuTAIL-PCR products need to be analyzed either after a cloning step or as hybridization substrates/probes. Moreover, further optimization of MuTAIL is possible because the CST1 primer did not prove to be a robust primer after multiple replicates of the 10-locus experiment. Based on the replicate experiments, exchanging CST1 with the Arabidopsis AD10 primer would enhance coverage of the a2 locus without causing a significant loss in amplification of the bz2 locus.

Furthermore, a subset of maize gene sequences appears resistant to MuTAIL analysis. For example, even though a large fraction of the lines contained the bz1-mum9 locus in this study, we did not identify the locus in the MuTAIL sequence libraries. While hybridization experiments confirm that bz1 is amplified by several arbitrary primers in the optimized set, the locus was not amplified at all by the Arabidopsis arbitrary primers and only amplified at 1.75-fold redundancy with the optimized primers. In addition, the bz1 hybridizing signals are typically weaker in comparison with the other genes tested in the 10-locus panel, which suggests that the relative concentrations of bz1 sequences in pools of MuTAIL products are lower and less likely to clone. These results indicate that some genes will not amplify and/or clone easily with MuTAIL, but it is important to note that all other flanking PCR methods have this disadvantage. Methods such as adapter-mediated PCR or inverse PCR rely on restriction sites that are sufficiently close in order to amplify a flanking sequence. Even applying a broad collection of enzymes with these methods will not promote amplification of all genes within a genome. In addition, short flanking sequences are typically recovered when restriction enzymes with 4-bp recognition sequences are used. This makes subsequent gene identification difficult unless the products derive from organisms with complete genome sequence. The Maize Gene Discovery Project has overcome these technical challenges with the RescueMu transgene. Using RescueMu, this genome project has submitted ~178 000 GSS sequences to DDBJ/EMBL/GenBank that derive from ~65 000 rescued plasmids. However, these sequences only represent a small fraction of the transposon insertions within the plants sampled, because the RescueMu transgene accounts for ~5% of the Mu elements when propagated in the native high-copy Mu-active genetic background (15).

In contrast, MuTAIL has advantages over other methods by its ease of application, by the long average length of amplified sequences and by sampling from the native complement of transposons within the genome. TAIL-PCR is a method that requires a minimum number of sample manipulations and is therefore readily automated. In our adaptation, we use two PCRs that simply require mixing PCR reagents and diluted DNA samples. The low number of manipulations makes the method attractive for high-throughput applications, as shown by the use of TAIL-PCR for amplifying T-DNA insertions in Arabidopsis (17). However, the TAIL protocol has not been used in high-copy insertion lines potentially due to concerns about aberrant priming when a large number of specific priming sites are present in the DNA template. Our data suggest that MuTAIL both samples and represents the relative abundance of transposon insertion sites within a DNA template. DNA gel blots with the MuTAIL genomic templates show a higher complexity of Mu insertions in lines that gave more unique insertion sites. Furthermore, we tested eight single-copy flanking sequences from Mu-active lines for inheritance and found a distribution of 4:2:2 for parental, new germinal and somatic insertions consistent with the relative over-representation of parental insertions within a pooled DNA template.

Because active Mutator lines contain a family of mobile elements that have polymorphic TIR sequences, applications that utilize native Mu elements benefit from a minimum bias in representation of different family members. While we did not attempt a quantitative comparison of sequences amplified with the TIR4 and TIR8 nested primers, analysis of the Mu-TIR sequences recovered with TIR4 suggests an apparent bias against Mu1 and Mu4 sequences. This bias could be caused by differences in the relative numbers of Mu family members between the DNA templates examined, by differences in amplification/cloning efficiency of the Arabidopsis versus optimized arbitrary primers or by differences in amplification/cloning efficiency of the TIR4 versus TIR8 nested primers. We did not observe a marked difference in the 10-locus hybridization results when the MuTAIL products were primed with the TIR4 or TIR8 primers, suggesting that the nested Mu-TIR primer is not the primary determinant of the amplified Mu-flanking sequences (data not shown). However, the optimized MuTAIL protocol, which uses TIR8, does recover a larger percentage of matching 9-bp target site duplications. Together, these observations suggest that the optimized protocol more accurately reflects the population of Mu-TIRs found in a pool of insertions.

Our sequencing results raise the question of whether MuTAIL will be effective for identifying the genes responsible for phenotypic mutants. The sequencing studies presented here do not focus on identifying causative transposon disruptions in novel phenotypic mutants. Instead, we have validated that the MuTAIL protocol can be used for insertion site identification in high-copy lines. More recently, we have focused on the maize vp13 mutation and identified a MuTAIL flanking sequence that is associated with a Mu insertion in the Vp13 locus (T. Porch, D.R. McCarty and A.M. Settles, in preparation).

As more species are studied at a genomics level, the need for high-copy insertional mutagenesis will increase due to the following factors: the practical constraints of achievable population sizes for large organisms, the challenges of transforming non-model species and the widespread use of transposons as mutagens. Because the MuTAIL protocol amplifies large genomic sequences ranging to sizes above 2 kbp, the flanking sequences can be used either as complex probes or as substrates for hybridization-based identification of genomic disruptions. In organisms with incomplete genome sequence information, efficient recovery of cloned insertion sites with sufficient flanking DNA is essential to characterize insertions. A larger flanking sequence can be used as a hybridization probe to allow for the identification of cDNA or genomic clones in addition to giving a greater chance of finding homologous sequences with database searches. Indeed, the Maize Mapping Project has screened a subset of 196 of the MuTAIL clones we describe here for RFLP probes. Most of the clones screened (155/196) had four or fewer copies on DNA gel blots, and 56 primarily single-copy clones were selected for genetic mapping (MaizeGDB) (http://www.maizegdb.org/probe.php). Our database searches and the Maize Mapping Project data suggest that further sequencing of transposon flanks would be a valuable source of low-copy genomic sequence, particularly from a genome such as maize that has a high proportion of repetitive sequences. Importantly, these sequences are derived from insertion events that can cause mutations and potentially give insight into gene function. Based on our data presented here, we and the Maize Endosperm Development Project have begun a larger-scale sequencing project to sequence 384 MuTAIL clones from each of 90 seed mutants. Currently, ~30 000 MuTAIL sequences from these mutants have been deposited to the GSS database in DDBJ/EMBL/GenBank, but these sequences have yet to be analyzed in detail.


We thank Bao Cai Tan and Masaharu Suzuki for help in constructing flanking sequence libraries and Alex Raimos, Savita Shankar, Christian von Kleist, Regina Shaw and William Farmerie for assistance in sequencing and informatics. This work was supported by USDA grant 2000–35300–10309 (to A.M.S.), NSF Plant Genome award 0077676 (to D.R.M.) to fund the Maize Endosperm Functional Genomics Project and University of Florida Agricultural Experiment Station funding.


DDBJ/EMBL/GenBank accession nos+ BH417702–BH418377, CC050461–CC050722 and CG785063–CG786032


1. Coelho P.S., Kumar,A. and Snyder,M. (2000) Genome-wide mutant collections: toolboxes for functional genomics. Curr. Opin. Microbiol., 3, 309–315. [PubMed]
2. Parinov S. and Sundaresan,V. (2000) Functional genomics in Arabidopsis: large-scale insertional mutagenesis complements the genome sequencing project. Curr. Opin. Biotechnol., 11, 157–161. [PubMed]
3. May B.P., Liu,H., Vollbrecht,E., Senior,L., Rabinowicz,P.D., Roh,D., Pan,X., Stein,L., Freeling,M., Alexander,D. et al. (2003) Maize-targeted mutagenesis: a knockout resource for maize. Proc. Natl Acad. Sci. USA, 100, 11541–11546. [PMC free article] [PubMed]
4. Wendel J.F. (2000) Genome evolution in polyploids. Plant Mol. Biol., 42, 225–249. [PubMed]
5. Shoemaker D.D., Lashkari,D.A., Morris,D., Mittman,M. and Davis,R.W. (1996) Quantitative phenotypic analysis of yeast deletion mutants using a highly parallel molecular bar-coding strategy. Nature Genet., 14, 450–456. [PubMed]
6. Winkler R.G., Frank,M.R., Galbraith,D.W., Feyereisen,R. and Feldmann,K.A. (1998) Systematic reverse genetics of transfer-DNA-tagged lines of Arabidopsis: isolation of mutations in the cytochrome p450 gene superfamily. Plant Physiol., 118, 743–750. [PMC free article] [PubMed]
7. Walbot V. (2000) Saturation mutagenesis using maize transposons. Curr. Opin. Plant Biol., 3, 103–107. [PubMed]
8. Frey M., Stettner,C. and Gierl,A. (1998) A general method for gene isolation in tagging approaches: amplification of insertion mutagenised sites (AIMS). Plant J., 13, 717–722.
9. Hanley S., Edwards,D., Stevenson,D., Haines,S., Hegarty,M., Schuch,W. and Edwards,K.J. (2000) Identification of transposon-tagged genes by the random sequencing of Mutator-tagged DNA fragments from Zea mays. Plant J., 23, 557–566. [PubMed]
10. Lund A.H., Turner,G., Trubetskoy,A., Verhoeven,E., Wientjens,E., Hulsman,D., Russell,R., DePinho,R.A., Lenz,J. and van Lohuizen,M. (2002) Genome-wide retroviral insertional tagging of genes involved in cancer in Cdkn2a-deficient mice. Nature Genet., 32, 160–165. [PubMed]
11. Mikkers H., Allen,J., Knipscheer,P., Romeyn,L., Hart,A., Vink,E. and Berns,A. (2002) High-throughput retroviral tagging to identify components of specific signaling pathways in cancer. Nature Genet., 32, 153–159. [PubMed]
12. Suzuki T., Shen,H., Akagi,K., Morse,H.C., Malley,J.D., Naiman,D.Q., Jenkins,N.A. and Copeland,N.G. (2002) New genes involved in cancer identified by retroviral tagging. Nature Genet., 32, 166–174. [PubMed]
13. Singh M., Lewis,P.E., Hardeman,K., Bai,L., Rose,J.K., Mazourek,M., Chomet,P. and Brutnell,T.P. (2003) Activator mutagenesis of the pink scutellum1/viviparous7 locus of maize. Plant Cell, 15, 874–884. [PMC free article] [PubMed]
14. Samson F., Brunaud,V., Balzergue,S., Dubreucq,B., Lepiniec,L., Pelletier,G., Caboche,M. and Lecharny,A. (2002) FLAGdb/FST: a database of mapped flanking insertion sites (FSTs) of Arabidopsis thaliana T-DNA transformants. Nucleic Acids Res., 30, 94–97. [PMC free article] [PubMed]
15. Lunde C.F., Morrow,D.J., Roy,L.M. and Walbot,V. (2003) Progress in maize gene discovery: a project update. Funct. Integr. Genomics, 3, 25–32. [PubMed]
16. Walbot V. and Warren,C. (1988) Regulation of Mu element copy number in maize lines with an active or inactive Mutator transposable element system. Mol. Gen. Genet., 211, 27–34. [PubMed]
17. Alonso J.M., Stepanova,A.N., Leisse,T.J., Kim,C.J., Chen,H., Shinn,P., Stevenson,D.K., Zimmerman,J., Barajas,P., Cheuk,R. et al. (2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science, 301, 653–657. [PubMed]
18. Liu Y.G., Mitsukawa,N., Oosumi,T. and Whittier,R.F. (1995) Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal asymmetric interlaced PCR. Plant J., 8, 457–463. [PubMed]
19. Arumuganathan K. and Earle,E.D. (1991) Nuclear DNA content of some important plant species. Plant Mol. Biol. Rep., 9, 208–218.
20. Timmermans M.C., Hudson,A., Becraft,P.W. and Nelson,T. (1999) ROUGH SHEATH2: a Myb protein that represses knox homeobox genes in maize lateral organ primordia. Science, 284, 151–153. [PubMed]
21. Das L. and Martienssen,R. (1995) Site-selected transposon mutagenesis at the hcf106 locus in maize. Plant Cell, 7, 287–294. [PMC free article] [PubMed]
22. Becraft P.W., Stinard,P.S. and McCarty,D.R. (1996) CRINKLY4: A TNFR-like receptor kinase involved in maize epidermal differentiation. Science, 273, 1406–1409. [PubMed]
23. McCarty D.R., Hattori,T., Carson,C.B., Vasil,V., Lazar,M. and Vasil,I.K. (1991) The Viviparous-1 developmental gene of maize encodes a novel transcriptional activator. Cell, 66, 895–905. [PubMed]
24. Tan B.C., Schwartz,S.H., Zeevaart,J.A. and McCarty,D.R. (1997) Genetic control of abscisic acid biosynthesis in maize. Proc. Natl Acad. Sci. USA, 94, 12235–12240. [PMC free article] [PubMed]
25. Walbot V. (1992) Developmental regulation of excision timing of mutator transposons of maize—comparison of standard lines and an early excision bz1-Mu1 line. Dev. Genet., 13, 376–386.
26. McLaughlin M. and Walbot,V. (1987) Cloning of a mutable bz2 allele of maize by transposon tagging and differential hybridization. Genetics, 117, 771–776. [PMC free article] [PubMed]
27. Chomet P., Lisch,D., Hardeman,K.J., Chandler,V.L. and Freeling,M. (1991) Identification of a regulatory transposon that controls the Mutator transposable element system in maize. Genetics, 129, 261–270. [PMC free article] [PubMed]
28. Ewing B., Hillier,L., Wendl,M.C. and Green,P. (1998) Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res., 8, 175–185. [PubMed]
29. Ewing B. and Green,P. (1998) Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res., 8, 186–194. [PubMed]
30. Gordon D., Abajian,C. and Green,P. (1998) Consed: a graphical tool for sequence finishing. Genome Res., 8, 195–202. [PubMed]
31. Altschul S.F., Madden,T.L., Schaffer,A.A., Zhang,J., Zhang,Z., Miller,W. and Lipman,D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res., 25, 3389–3402. [PMC free article] [PubMed]
32. Chandler V., Rivin,C. and Walbot,V. (1986) Stable non-mutator stocks of maize have sequences homologous to the Mu1 transposable element. Genetics, 114, 1007–1021. [PMC free article] [PubMed]
33. Oishi K.K. and Freeling,M. (1988) A new Mu element from a Robertson’s Mutator line. In Nelson,O. (ed.) Plant Transposable Elements. Plenum Press, New York, pp. 289–291.
34. Schnable P.S., Peterson,P.A. and Saedler,H. (1989) The bz-rcy allele of the Cy transposable element system of Zea mays contains a Mu-like element insertion. Mol. Gen. Genet., 217, 459–463. [PubMed]
35. Fleenor D., Spell,M., Robertson,D. and Wessler,S. (1990) Nucleotide sequence of the maize Mutator element, Mu8. Nucleic Acids Res., 18, 6725. [PMC free article] [PubMed]
36. Felsenstein J. (1981) Evolutionary trees from DNA sequences: a maximum likelihood approach. J. Mol. Evol., 17, 368–376. [PubMed]
37. Smith D., Yanai,Y., Liu,Y.G., Ishiguro,S., Okada,K., Shibata,D., Whittier,R.F. and Fedoroff,N.V. (1996) Characterization and mapping of Ds-GUS-T-DNA lines for targeted insertional mutagenesis. Plant J., 10, 721–732. [PubMed]
38. Palmer L.E., Rabinowicz,P.D., O’Shaughnessy,A.L., Balija,V.S., Nascimento,L.U., Dike,S., de la Bastide,M., Martienssen R.A. and McCombie,W.R. (2003) Maize genome sequencing by methylation filtration. Science, 302, 2115–2117. [PubMed]
39. Whitelaw C.A., Barbazuk,W.B., Pertea,G., Chan,A.P., Cheung,F., Lee,Y., Zheng,L., van Heeringen,S., Karamycheva,S., Bennetzen,J.L. et al. (2003) Enrichment of gene-coding sequences in maize by genome filtration. Science, 302, 2118–2120. [PubMed]
40. SanMiguel P., Tikhonov,A.P., Jin,Y.-K., Motchoulskaia,N., Zakharov,D., Melake-Berhan,A., Springer,P., Edwards,K., Lee,M. and Avramova,Z. (1996) Nested retrotransposons in the intergenic regions of the maize genome. Science, 274, 765–768. [PubMed]
41. Cresse A.D., Hulbert,S.H., Brown,W.E., Lucas,J.R. and Bennetzen,J.L. (1995) Mu1-related transposable elements of maize preferentially insert into low copy number DNA. Genetics, 140, 315–324. [PMC free article] [PubMed]
42. Robertson D.S. (1978) Characterization of a Mutator system in maize. Mutat. Res., 51, 21–28.
43. Barker R.F., Thompson,D.V., Talbot,D.R., Swanson,J. and Bennetzen,J.L. (1984) Nucleotide sequence of the maize transposable element Mu1. Nucleic Acids Res., 12, 5955–5967. [PMC free article] [PubMed]
44. Singer T., Yordan,C. and Martienssen,R.A. (2001) Robertson’s Mutator transposons in A.thaliana are regulated by the chromatin-remodeling gene Decrease in DNA Methylation (DDM1). Genes Dev., 15, 591–602. [PMC free article] [PubMed]
45. Lisch D.R., Freeling,M., Langham,R.J. and Choy,M.Y. (2001) Mutator transposase is widespread in the grasses. Plant Physiol., 125, 1293–1303. [PMC free article] [PubMed]
46. Rudenko G.N. and Walbot,V. (2001) Expression and post-transcriptional regulation of maize transposable element MuDR and its derivatives. Plant Cell, 13, 553–570. [PMC free article] [PubMed]

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • GSS
    Published GSS sequences
  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • SRA
    Links to SRA
  • Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...