Introduction

In her early studies on the Ac/Ds transposable element system in maize, Barbara McClintock described two states of transposable elements [21, 22]. State I caused frequent chromosomal breakage, whereas state II elements transposed from one locus to another, but rarely caused chromosomal breakage. A typical state II element has a pair of terminal inverted repeats (5′- and 3′-TIRs) flanked by conserved subterminal sequences. The transposase recognizes these regions, excises the transposable element from the donor site, and inserts it at a different locus by a ‘cut and paste′ mechanism [9, 13]. The donor site is then joined by DNA repair machinery. Unlike state II elements, the state I elements that cause frequent DNA breakage are characterized by the presence of multiple transposon ends in proximity. McClintock observed that when chromosome ends are broken in the maize gametophyte they undergo the breakage-fusion-bridge cycle, which can persist in the endosperm. However, in the sporophyte the broken ends are ‘healed’ [20]. The ‘healing’ of broken chromosome ends occurs by de novo synthesis of telomere sequences (reviewed by [23]).

Several studies have been carried out to understand the nature and mechanism of chromosomal breakage at the state I elements. Molecular characterization of chromosome-breaking state I elements has revealed the presence of multiple transposon ends in proximity. The chromosome-breaking shrunken alleles generated by McClintock carry a double Ds element – one Ds element inserted into another [6, 7, 9, 10, 28]. Presence of an intact Ac element in proximity to a fractured Ac (fAc) can also cause chromosomal breakage [26, 33]. The sesquiDs element at the bz1-m1 locus in maize is also a chromosome-breaker. It is derived from loss of a 5′ TIR from a double Ds element. In addition to a normal pair of 5′ and 3′ TIRs, it has an internal 3′ TIR in the same orientation as the 5′ TIR [19].

Weil and Wessler [30] showed that chromosomal breakage at state I Ds elements is caused by an aberrant transposition reaction involving transposon ends in non-standard orientation with reference to each other. English et al. [11] used an engineered Ds element in tobacco and established that the key feature of a chromosome breaking Ds element is the presence of one pair of 5′ and 3′ TIRs in direct orientation with reference to each other. Such aberrant transposition involved transposon ends on sister chromatids [12]. A pair of 5′ and 3′ TIRs in reversed orientation with reference to each other can also cause chromosome breakage. Non-standard transposition involving a pair of Ds and Ac elements at the bz1 locus with a configuration where the 3′ TIR of the Ds element and the 5′ TIR of the Ac element are in reverse orientation with reference to each other leads to a range of chromosomal rearrangements including the breakage-fusion-bridge cycle [14]. In a study on non-standard transpositions involving a pair of Ac and fAc (fractured Ac) at the maize p1 locus, Yu et al. [33] established that chromosomal breakage can be induced either when the 5′ and 3′ TIRs are in direct orientation or reversed orientation with reference to each other.

The wide range of chromosomal rearrangements (including large scale deletions, inversions, duplications, and translocations) caused by non-standard transposition both in direct orientation and reversed orientation of the 5′ and 3′ TIRs opens up the potential for such transposable elements in chromosome engineering. The ability of an engineered Ds element with 5′ and 3′ TIRs in reversed orientation to cause large scale chromosomal rearrangements has been studied in Arabidopsis [17], and rice [32]. However, the utility of engineered transposable elements to cause chromosome truncation has not yet been explored in maize.

Here we have developed a chromosome-breaking Ac/Ds platform for targeted chromosome truncation. We have introduced an engineered Ds element with the 3′ and 5′ TIRs in direct orientation. This construct also carries a site-specific recombination site for the potential targeted introduction of transgenes. We have tested the efficacy of this construct to induce chromosome breakage upon activation by Ac by using two of the transgenic lines with insertions on chromosome arms 4S or 4L. Chromosomal breakage events at the 4S transgenic site cause mosaics for sugary1 and brittle2, when crosses are made to stocks with mutations of these 4S genes. Similarly, the combination of Ac with a transgene insertion on chromosome 4L causes mosaics for colorless (c2), a gene on 4L, when crosses are made with this mutant stock. The recovery of an insertion event on the supernumerary B chromosome generates an exceptional tool for the study of this supernumerary chromosome in that a phenotypic marker is present on an otherwise intact B chromosome.

Materials and methods

Construct components

LB, T-DNA left border; RB, T-DNA right border; Tnos, nos terminator from Agrobacterium; Bar, bialaphos resistance gene as a selection marker gene; UbiP, maize ubiquitin promoter; loxP, site specific recombination site; B1-Peru, Maize pigment synthesis gene used as a visible marker; 5′ TE, 5′ Terminal Inverted Repeat plus subterminal sequence (265 bp) of Ds element; 3′ TE, 3′ Terminal inverted repeat sequence plus subterminal sequence (250 bp) of Ds element.

Description

DRDs (Fig. 1), the ‘Direct Repeat Ds’ T-DNA construct has the 3′ and 5′ TIRs (red and green triangles, respectively) of Ds cloned in direct orientation with reference to each other. The bialaphos resistance selection marker Bar (with Ubi promoter and Tnos terminator) is cloned between the two direct repeats. A visible selection marker gene B1-Peru (with Ubi promoter and Tnos terminator) is placed flanking the 5′TIR. The site-specific recombination site loxP is included between the Ubi promoter and the B-Peru gene.

Fig. 1
figure 1

Diagram of the DRDs construct; the elements of the transgene construct are noted. Between the Right and Left Border of the Agrobacterium construct are the Ubiquitin Promoter (UbiP) driving the bialaphos resistance gene (BAR) with the Nos terminator (Tnos). This gene is flanked by direct repeats of Dissociation (red and green arrows). The other gene has a ubiquitin promoter followed by a loxP site and then the B-Peru gene terminated with the Nos terminator

Full size image

Construction of DRDs

The multiple cloning site of T-DNA pPZP201 was removed by cutting with HinDIII and EcoRI and replaced with the following polylinker: AAGCTA-KpnI-XhoI-HinDIII-XbaI-BamHI-SacI-AvrII-XmaI-SpeI-CAATTC. This construct is pZP1718. The following fragments were sequentially introduced into the pZP1718 Multiple Cloning Site (MCS):

Source of DNA fragments

The 3′ TIR and 5′ TIR of Ds are derived from pWL117 [29]. UbiBARTnos and Ubi are derived from pAHC20 [8]. B1-peru [25] was cloned into pBKS.

Sequence of oligos for the the polylinker

PL17

5′AGCTAGGTACCAAAAAAACTCGAGAAGCTTAAAAAAATCTAGAGGATCCAAAAAAAGAGCTCCCTAGGAAAAAAACCCGGGAAAAAAAACTAGTC-3′

PL18

5′AATTGACTAGTTTTTTTTCCCGGGTTTTTTTCCTAGGGAGCTCTTTTTTTGGATCCTCTAGATTTTTTTAAGCTTCTCGAGTTTTTTTGGTACCT-3′

A detailed map of DRDs is shown in Fig. S1.

Transformation

Medium formulations

In our experiments, N6 salts mainly have been used in Hi-II callus induction during tissue culture in Agrobacterium-mediated transformation procedures, including infection, co-cultivation, and selection media. MS based media was used in the stage of regeneration. For a complete list of media reagents see Lee and Zhang [18].

Agrobacterium culture initiation and inoculation

Agrobacterium tumefaciens strain EHA101 glycerol stock was removed from a  − 80 °C freezer and streaked onto a YEP medium plate containing appropriate antibiotics. Then, the plate was incubated at 28 °C for 2 to 3 days in the dark until bacterial colonies developed fully.

One full loop of EHA101 from a YEP plate was suspended with 5 ml of sterile PHI-A infection medium in a 15 ml Falcon tube. 1 ml of such suspension was transferred to a spectrophotometer cuvette and the cell density was checked of the suspension culture at λ = 550 (OD550). The value of OD550 should be between 0.3 and 0.4.

Maize immature embryo dissection

F2 immature zygotic embryos were obtained from maize Hi-II hybrid (F1) crossed by Hi-II A or Hi-II B line. After 9 to 12 days post self-pollination, ears were collected when the size of immature embryos was approximately 1.5–2 mm in length and sterilized with a solution consisting of 50% commercial bleach containing 5.25% sodium hypochlorite in autoclaved water with 2 drops of TWEEN 20 to cover the ears for 20 min at room temperature. The bleach solution was poured off and the ears washed with autoclaved water three times for 5 min each. Then, the top half of the kernels of each ear were cut off. The 1.5 to 2 mm embryos were removed and placed into a 2 ml microcentrifuge tube filled with infection medium. The embryos were washed with the same medium three times.

Agrobacterium infection and resistant culture selection

1 ml of Agrobacterium suspension was added into each 2 ml tube. After 5 to 10 min with the embryos submerged, the Agrobacterium suspension was poured off and the embryos placed onto co-cultivation medium plates. The embryos were placed flat face down on the medium using a spatula. Then, the plate was incubated at 20 °C in the dark for 3 days. Subsequently, embryos were transferred to the callus induction medium plate and placed in a 28 °C incubator in the dark for 10–12 days. Embryos were then sub-cultured on callus selection medium I for 2 weeks for initial herbicide selection. Next, these calli were moved to callus selection medium II plates and were sub-cultured to fresh selection medium II biweekly until all remaining herbicide resistant calli proliferated. Each callus from the same embryo was treated as a single transformation event.

Regeneration of transgenic plants

The herbicide-resistant calli were cultured on regeneration medium in a 28 °C incubator to initiate shoots and sub-cultured to the same medium biweekly until shoots are visible. Then, the shoots were moved to rooting medium and cultured at 25 °C under 16:8 h photoperiod with light intensity of 100–150 µmol m−2 s−1. After 2–4 weeks, each regenerated small plantlet was transferred to small pots containing Pro-mix soil and allowed to acclimatize for 2–3 weeks in growth chamber conditions at 25 °C under 16:8 h photoperiod with light intensity of 300 µmol m−2 s−1. Finally, these transgenic plants were transplanted to the greenhouse.

Fluorescence in situ hybridization

Fluorescence In Situ Hybridization (FISH) was conducted as described [2, 16] with the following modifications. To facilitate probe production, the HindIII fragment of DRDs was cloned into pBlueScript (Agilent, Santa Clara, CA, USA). A Texas-Red labeled DRDs probe (NEL426001EA, PerkinElmer-Revvity, Waltham, MA, USA) was made via nick translation [15] using either plasmid DNA or PCR product [Primers, M-13 Forward (5′-GTAAAACGACGGCCAGT-3′) and M-13 Reverse (5′-CAGGAAACAGCTATGAC-3′); enzyme, JumpStart REDTaq Ready Mix (P0982, Sigma-Aldrich, St. Louis, MO, USA)]. To identify each chromosome pair, two fluorescent oligonucleotides (Integrated DNA Technologies, Coralville, IA, USA) were used in conjunction with DAPI (4′,6-diamidino-2-phenylindole) incorporation into heterochromatic regions. Sequence of the Centromere C (Cent C) oligo is 5′- 6-FAM / CCTAAAGTAGTGGATTGGGCATGTTCG-3′. The TAG microsatellite repeat oligo is a 56-mer, sequence 5′- 6-FAM / AGT- (AGT)17—AG-3′. The hybridization mixture (for 10 µL/slide) contained 200—400 ng DRDs probe, 45 ng Cent C oligo, and 10 ng TAG oligo in 2X SSC – 1 X TE buffer [15]. Images were captured with a Cool-1300QS CCD camera (VDS Vosskühler GmbH, Osnabrück, Germany) on an Olympus BX61 fluorescence microscope (Olympus Corporation, Tokyo, Japan) using GenASIsis software (Applied Spectral Imaging, Carlsbad, CA, USA). Images were processed using the Curves, Levels, and/or Brightness-Contrast functions of Photoshop 2023 (Adobe, Inc., San Jose, CA, USA).

Results

An Agrobacterium construct was assembled that contained a Ubiquitin promoter driving the bialaphos resistance gene (BAR) and terminated with the nos terminator. This gene was flanked by direct repeats of the sequence of the Dissociation terminal inverted repeat. The other gene in the construct has a maize Ubiquitin promoter driving the B1-Peru allele of the b1 locus of maize that is a transcriptional regulator of the anthocyanin pigment pathway [25]. A loxP site [1] is intervening between the promoter and the B1-Peru coding sequence. The construct is illustrated in Fig. 1 (also Fig. S1) and described in more detail in the Methods. This construct was transformed into the Hi-II transformation line of maize [27].

Transformants were selected by resistance to bialaphos. The Hi-II line does not support anthocyanin pigment. Therefore, initially some events were confirmed by Fluorescence In Situ Hybridization (FISH). To take advantage of the B1-peru marker for phenotypic studies, each candidate transformation event was crossed to a line recessive for the anthocyanin gene r1. The r1-r allele converged to the inbred line W22 was used. The B1-Peru gene will complement the recessive r1-r allele. A selfing regime for multiple generations followed until a line was produced that was full color. The Ubi-B1-Peru transgene expresses in most tissues and thus, these lines have purple pigment to varying degrees in most tissues. The lines that fulfilled this criterion were examined for purple pigment in the embryo. Few lines of maize express anthocyanin in the embryo of the kernel but the Ubi-B1-Peru transgene activates the pathway genes in this tissue. Nevertheless, the various lines showed different intensities and distributions of pigment. Unlike the others, the insertion in 4S described below shows no color in most tissues. The lines that were developed from the crossing and selfing regime were subjected to FISH with the DRDs clone to localize each to a chromosome arm. The locations of this collection are noted in Table 1.

Table 1 Chromosome arm locations of DRDs insertion events
Full size table

Tests of chromosomal breakage

Two of the early identified events were localized to near the centromere on the short arm of chromosome 4 (4S) and on the long arm of chromosome 4 (4L). Each of these was crossed by a stable version of Activator (stAc5145) [31] to test whether the transgene would cause chromosomal breakage. This version of Activator has a lesion in one of its terminal inverted repeats at the p1 locus on the short arm of chromosome 1 such that it is incapable of transposition itself. It does, however, supply the transposase that can act on Ds elements in the genome.

When the F1 between the transgene on 4S and with stAc5145 on 1S was crossed as a male parent to a line that was homozygous for sugary1 (su1), which is located in 4S, mosaic kernels were found that had a pitted endosperm that in most cases could be reliably identified as sectors that uncover the sugary1 mutation, indicating chromosomal breakage in proximal 4S during endosperm development. Figure 2 shows kernels with a mosaic phenotype together with sugary kernels for comparison.

Fig. 2
figure 2

Uncovering of the sugary1 mutation; a Kernels with the wild-type Sugary1 gene, b Kernels showing the sugary1(su1) mutant phenotype, c Mosaicism for the sugary phenotype in a selfed su1/+ ; DRDs4Sa/+ ; stAc5145/+ plant, d Mosaicism for the sugary phenotype in kernels from a cross of a sugary1 line as a female by the homozygous line for DRDs4Sa and stAc5145. Arrows indicate three regions uncovering su1

Full size image

By following the FISH signals for the 4S transgene and the stAc5145 on the short arm of chromosome 1, which can be visualized among the multiple Ac elements present in the genome [34], a line homozygous for the transgene and the stAc5145 element was established (Fig. 3). When this line was crossed to sugary1 females, all the resulting kernels showed a finely pitted phenotype. The homozygous line itself shows no pitting but over several generations of self-pollination, it regularly has small or defective kernels present (Fig. S2). Apparently, homozygous deficiencies are sufficiently rare as not to hinder the survival of most individuals in this stock.

Fig. 3
figure 3

FISH localization of a transgene on proximal 4S and stAc5154 on 1S in a line homozygous for both; a FISH image illustrating the homozygous DRDs4Sa insertion in this line. The arrows designate the transgene signal in red, b FISH image of the same line probed with Ac, illustrating stAc5145 on 1S. The red signals represent the multiple Ac elements in the genome, including stAc5154 (arrows). Green signals are chromosomal landmarks used to identify the chromosomes (Scale bars 5 µm)

Full size image

In crosses of DRDs4Sa/+ ; stAc5154/+ to heterozygotes of brittle2 (bt2) with normal (heterozygotes were used to facilitate germination), kernels segregated for pitted kernels and normal (Fig. 4). On the female side bt2 and + will segregate 1:1. On the male side, the transgene and stAc5154 independently segregate for a combination of the two being present in the male gametes at a frequency of 25%. If the transgene causes chromosomal fracture in the presence of stAc5154, a predicted frequency of mosaic kernels in this cross would be 12.5%, assuming there is no impact of the combination on transmission. For one ear, there were 18 pitted kernels in 175 total for a frequency of 10.3%. For a second ear, there were 19 pitted kernels among 198 for a frequency of 10%. Combined, there were 37 pitted in a total of 373 kernels. These values are highly significantly different from the 12.5% prediction with no detrimental impact (Chi2 = 8.08, d.f. = 1, p < 0.005) but nevertheless are a close approximation. However, these results indicate that DRDs4Sa or stAc5154 alone are incapable of producing this phenotype and that no other active Ac elements are present in these stocks. In subsequent analyses, bt2/+ females were crossed by the line homozygous for DRDs4Sa and stAc5154, which produced ears segregating 1:1 for pitted and normal kernels. Together, these results indicate that the combination of DRDs4Sa and stAc5154 are responsible for the pitted phenotype indicative of chromosomal breakage on 4S during endosperm development.

Fig. 4
figure 4

Mosaic kernels resulting from chromosomal breakage; bt2/+ plants were crossed as a female by the heterozygote DRDs4Sa/+ ; stAc5154/+ . a Normal kernels, b Phenotype of kernels from a bt2 line, c Segregants with fine grain mosaicism for bt2

Full size image

The other event that was crossed to stAc5145 involved a transgene on the long arm of chromosome 4, DRDs4La. The transgene construct itself exhibits solid purple color in the aleurone of kernels and it also shows strong purple pigment in plant tissues. After crossing this construct to the stAc5154 stock and subsequent self-pollinations, individual plants from the F2 were crossed as male to a line that was homozygous for the colorless (c2) anthocyanin mutation located in 4L in an otherwise heterozygous background for W22 and Mo17 maize lines. All other genes for anthocyanin production in the kernel are present in this c2 tester line including the R-scm2 allele of the r1 gene. Resulting ears showed mosaicism for C2/c2 indicative of chromosome fracture in the long arm of chromosome 4 (Fig. 5).

Fig. 5
figure 5

Chromosome breakage in the long arm of chromosome 4; c2 is one of the genes needed in the anthocyanin pigment pathway. a Kernels from a cross between a female homozygous for c2 and a male parent containing DRDs4La. The solid purple kernels indicate no chromosome breakage, b Kernels in which the female is c2 and the male parent contained DRDs4La and stAc5145. The endosperm is triploid; therefore, one functional C2 allele is present in the primary endosperm nucleus. With increasing dosage of the C2 allele, there is an increase in the level of anthocyanin in the aleurone layer of the endosperm. The mosaic pattern on the selected kernels (b) and the ear (d) is attributed to chromosomal breakage during endosperm development. The yellow areas indicate the absence of a C2 allele. In some kernels one can observe darker sectors (c, arrow) adjacent to the absence of purple pigment sectors, which is indicative of a breakage-fusion-bridge cycle in the endosperm in which chromosome breakage eliminates the C2 allele in a cell lineage and the sister lineage has two copies

Full size image

The mosaic kernels from the cross to the c2 tester were used as males in subsequent crosses to an r1-r tester that is recessive for the r1 gene but contains the dominant alleles for all other genes required for anthocyanin in the kernel. As noted above, the B1-Peru gene in the transgene can complement the r1-r allele. The genotype of the mosaic kernels used in this cross is R-scm2/r1-r; c2/DRDs4La; +/stAc5145. In the resulting progeny, the expectation is that half of the kernels would be full color from the paternally derived R-scm2 allele. In the remaining half of the kernels, the inheritance of the normal chromosome 4 alone without R-scm2 or the complementing B1-Peru transgene would be in a quarter of the total and have a colorless phenotype. The inheritance of the DRDs4La chromosome alone without R-scm2 would be present in one quarter of the total progeny kernels. In the absence of stAc5154, they would condition full color due to the B1-Peru transgene. In the presence of the independently assorting Ac element, breakage is expected at the DRDs position. If this position is proximal to the centromere relative to the B1-Peru gene, the B1-Peru gene will be eliminated from the construct creating a mosaic phenotype. If the B1-Peru gene is centromere proximal, mosaicism for B1-Peru can still occur due to a breakage-fusion-bridge (BFB) cycle in the gametophytes and endosperm that is initiated at the sight of breakage. In either scenario, mosaic kernels would be predicted in one eighth of the total kernels. From a pair of crosses, 472 total kernels were found with 72 showing evidence of mosaicism for B-Peru. A Chi Square test with the expected frequency of 12.5% found that the observed frequency was not significantly different from the expectation (Chi2 = 2.86, d.f. = 1; p > 0.05). This result indicates that the various insertions of DRDs that express pigment could use a uniform system involving r1-r in the presence of an Ac element to generate mosaic kernels and would not need chromosome arm specific tester stocks.

An insertion on the B chromosome

The transformation recipient line (Hi II) contained supernumerary B chromosomes that had been introgressed into the line. One of the lines identified by FISH showed an insertion into this supernumerary chromosome (Fig. 6). When crossed into the W22 line with recessive, r1-r, the plants and the kernels exhibit a strong purple pigmentation. Because of the ubiquitin promoter, it is expressed in both the embryo and endosperm. The B chromosome has the property of nondisjunction at the second pollen mitosis that produces the two sperm and therefore the sperm differ in their B chromosome content [3, 4]. Thus, with double fertilization, the embryo and endosperm will be pigmented differentially when nondisjunction occurs, and the pigmentation in either the embryo or endosperm will depend on whether the B containing sperm fertilizes the egg or the central cell, respectively. When crossed to the r1-r tester, some kernels are produced that have anthocyanin in the embryo but not the endosperm. These contain two B chromosomes that are inherited to the next generation. In the W22 background, the rate of nondisjunction is 99%. This event provides a phenotypic marker for an intact B chromosome that is very useful for studies of B chromosome behavior. A companion study illustrates its properties and utility for studying the preferential fertilization property of the B chromosome [5].

Fig. 6
figure 6

FISH localization of DRDsB, an insertion on the supernumerary B chromosome; The red hybridization signal designated by the arrow represents the transgene insertion. The other red signals depict hybridization of the DRDs construct components to endogenous genes (Ubiquitin and B-Peru). Green signals represent landmarks for chromosome identification [2, 15, 16] (Scale bar 5 µm)

Full size image

Discussion

A test was conducted to determine whether the transgenic introduction of directly oriented terminal inverted repeats of the transposable element system Ac/Ds would cause chromosomal breakage at the site of insertion. An event localized to the short arm of chromosome 4 near the centromere was combined with a stable version of Ac in the p1 locus, namely stAc5145. When either a segregating male parent for the transgene and Ac or homozygous for both was crossed to a sugary tester, sectors in the resulting endosperm showed evidence of uncovering of the recessive mutation inherited from the female parent. The frequency from the segregating parent was consistent with the need for both the transgene and the stable Ac for the sectoring to occur. The cross with the homozygous line as a male resulted in all the progeny with sugary sectors. A heterozygote of the brittle2 mutation (bt2/+) was crossed by males of the homozygous DRDs4Sa line, which produced a segregating ear for uncovering of the recessive bt2, which resides on 4S. This cross illustrates that the male parent alone does not condition the pitted phenotype. The results are consistent with chromosomal cleavage at the site of the transgene insertion in the presence of Activator.

A second event localized to the long arm of chromosome 4 was also combined with the stAc5145 and selfed. Individual plants from the F2 were crossed as a male parent to a colorless (c2) anthocyanin gene tester. The c2 gene resides in 4L. The resulting ears showed evidence of mosaicism for the C2 allele introduced from the male parent indicating chromosomal breakage in 4L. These results indicate a second example in which the DRDs construct produced chromosomal breakage in the presence of transposase supplied by Activator.

Other events were progressed through the crossing scheme with r1-r W22 to produce homozygous lines of the transgene. These lines provide the potential for use with other chromosome arms to generate mosaic plants/endosperms for the respective regions when Activator is introduced.

The recovery of DRDs on the B chromosome provides a means to follow the behavior of this supernumerary chromosome without an involvement of a translocation with an A chromosome as has been the case previously [3]. This marked B chromosome allows the study of the B chromosome without the involvement of any aberrations and has many potential uses.

With the recently developed ability to transform more lines of maize than previously, the DRDs plasmid could be transformed to lines that allow the development of anthocyanin pigment. Although BAR selection could be used, it is potentially the case that transformed cells will be purple due to the expression of the B1-Peru allele. Additional insertions of the transgene within the genome would provide a visible marker (B1-Peru) for each of them and permit the ability to create mosaics for the respective chromosome arms. Neuffer [24] described the use of transposed chromosome breaking Ds elements to 11 chromosomes and their use for genetic analysis. The transgene described here could be used in a similar manner for chimeric studies and has the advantage of showing the same phenotypic marker that is expressed in many maize tissues.

Here we describe the development of a transgenic chromosome-breaking system based on the use of directly oriented terminal repeats of the transposable element Dissociation. Ten transgenic events were recovered and partially introgressed into an r1-r W22 background. The B1-Peru anthocyanin regulator present in the transgene complements the r1-r allele. Two examples were crossed to a stable Activator autonomous element and then crosses to testers were performed to assay for chromosomal breakage. In both examples, phenotypic mosaicism expected from chromosomal breakage at frequencies expected from the combination of the transgene and the Activator element was observed. One event located on the supernumerary B chromosome provides an excellent phenotypic marker to follow this chromosome in studies of its behavior.