Variable Transposition of Eight Maize Activator (Ac) Elements Located on the Short Arm of Chromosome 1

Eight Activator (Ac) transposable elements mapped to the maize chromosome arm 1S were assessed for Ac transposition rates. For each of the Ac stocks, plants homozygous for the single Ac element and the Ds reporter r1-sc:m3 on chromosome 10 were crossed as females by a homozygous r1-sc:m3 tester color-converted W22 line. The resulting ears produced mostly coarsely spotted kernels and a low frequency of either near-colorless fine-spotted kernels or nonspotted kernels. The relative frequency of these two types of near-colorless kernels differed among the eight Ac stocks. The extent to which increased Ac dosage results in nonspotted kernels may be Ac-specific. Although all of the Ac elements are in near-isogenic inbred W22 lines, they varied to a large extent in their transposition frequency. These differences might possibly result from structural differences among the Ac elements. Because one pair of Ac elements derived from Ac33 on chromosome arm 5S differed about 13-fold in transposition frequency and a second pair of Ac elements derived from Ac12 on chromosome arm 1S differed about 3-fold in transposition frequency, this is not a likely explanation for all eight Ac elements. The data presented here support the notion that the differences in transposition frequency of the eight Ac elements may be a reflection of variability in Ac transcription or accessibility of the transposase to the Ac element, resulting from differences in the chromatin environments wherein the Ac elements are located. This is the first report of variability in transposition rates among different Ac donor lines.

Transposon tagging with the maize Ac element is a useful tool for regional mutagenesis (Federoff et al. 1984;Brutnell and Conrad 2003;Singh et al. 2003). Two features of the Ac element make it a tractable system for gene tagging. There are several genes controlling anthocyanin synthesis in the aluerone and embryo that contain Ds element insertions and can serve as reporter loci for the presence of an Ac element (McClintock 1955;Dooner et al. 1994). In addition, the delayed timing of Ac transposition in tissues containing increased Ac copy number provides a means of assessing the occurrence of an Ac transposition event by observing the size of pigmented sectors in the aleurone and embryo tissues of kernels.

MATERIALS AND METHODS
A collection of Ac-containing, near-isogenic, color-converted W22 inbred lines was produced by Kolkman et al. (2005); these included 41 precisely mapped Ac elements. The present report concerns 8 of these lines, all containing Ac elements mapped to the short arm of chromosome 1. The results presented here concern the variability in transposition frequency of these Ac elements observed while pursuing a regional mutagenesis program on this chromosome arm. For each of the Ac stocks, plants homozygous for the single Ac element and the Ds reporter r1-sc:m3 on chromosome 10 were crossed as females by a homozygous r1-sc:m3 tester color-converted W22 line (Figure 1).

Scoring of ears for Ac transposition events
The ears borne on plants homozygous for the Ac element produce mostly coarsely spotted kernels and a low frequency of near-colorless kernels. The coarsely spotted kernels display the pattern of colored sectors expected when the nuclei of the female gametophyte contain a single copy of the Ac element, whereas the near-colorless kernels manifest either a transposition event resulting in an increased dosage of the Ac element in these nuclei or the absence of an Ac element. Increased dosage may result from the Ac element replicating prior to its transposition and the subsequent presence of both the donor Ac element and the tr-Ac element in the functional megaspore that develops into the female gametophyte (embryo sac) (Greenblatt 1968). The increased dosage of the Ac element results in the delay in the development of the fine spotting of colored sectors in these kernels (see Figure 2). A total of 394 scoreable ears were screened for nearcolorless kernels. Each such kernel was scored as a transposition event and was removed from the ear and examined for fine spots under magnification. The kernels were sorted into two groups: spotted kernels containing at least one fine spot on either the aleurone or the embryo, and nonspotted kernels lacking any colored spots on both aleurone and the embryo ( Figure 2). Initially, the transposition frequencies were calculated on the basis of transposition events per ear. The number of ears scored for each of the Ac stocks ranged from 24 to 77. Among the eight Ac stocks, the mean values ranged from a high frequency of 9.67 fine-spotted kernels and 6.34 nonspotted kernels per ear (bti00252::Ac) to a low frequency of 0.58 fine-spotted kernels and 0.03 nonspotted kernels per ear (mon00106::Ac) (supporting information, Table S1, Figure S1). Subsequently, all of the kernels on the scored ears were weighed to provide an estimate of the total number of kernels per ear. The total estimated number of kernels examined was approximately 108,000. These data were used to calculate an estimated frequency of finespotted and nonspotted kernels per 1000 kernels for each of the eight Ac elements. The mean values ranged from a high frequency of 38.80 fine-spotted kernels and 26.28 nonspotted kernels per 1000 kernels (bti00252::Ac) to a low frequency of 2.15 fine-spotted kernels and 0.15 nonspotted kernels per 1000 kernels (mon00106::Ac) ( Table 1). The frequency of transposition of Ac elements per 1000 kernels for the individual families of the eight Ac elements is shown in Table S2.
Variation in Ac dosage effects An average transposition frequency of 2 to 4% (20 to 40 per 1000 kernels) was reported by Kolkman et al. (2005) based on their examination of approximately 12,400 kernels generated from 10 different Ac lines. No data were provided for any of the individual lines. In this report, the Ac transposition frequency as evidenced by the frequency of near-colorless, fine-spotted kernels ranged from 0.215% to 3.880%. When the near-colorless nonspotted kernels are included in the Figure 2 Results of crossing an Ac stock as shown in Figure 1 and generating transposed Acs. An F1 ear produced by crossing a stock homozygous for mon00068::Ac by the r1-sc:m3 reporter tester line. Most of the kernels have an aleurone layer containing two copies of the Ac element and are coarsely spotted. A small number of kernels have an aleurone layer containing four copies of the Ac element (two copies of the original Ac element and two copies of the newly transposed Ac element) and are finely spotted (near-colorless). Figure 1 Crossing scheme to produce finely spotted kernels. The chromosome constitutions of the embryos are shown. In most cases, the Ac element does not transpose and the embryo contains a single copy of the Ac element, whereas the aleurone cells of the endosperm contain two copies (one Ac from each of the two polar nuclei of the embryo sac). This dosage results in coarse spotting of the kernel aleurones. In a few cases, the Ac element duplicates, and subsequently, one copy transposes to a new site. When both copies are transmitted to the embryo sac (whether they are linked as shown in the figure or are on different chromosomes), the embryo will have two copies of the Ac element and the aleurone will have four copies, resulting in a finely spotted aleurone.
n calculations, then the frequency of Ac transposition ranges from 0.230% to 6.508%. Kolkman et al. (2005) noted differences in the degree of variation in Ac-mediated Ds variegation patterns (the size of sectors or spots) in kernels homozygous for independent Ac insertions. These researchers further noted that, inasmuch as all of the Ac elements they studied were in near-isogenic lines using the same Ds reporter, it was not likely that the variations they observed resulted from differences in the reporter gene or segregating modifier loci.

RESULTS
The size of spots on self-pollinated kernels on ears of plants homozygous for the eight different Ac stocks differed only slightly in size, and there was no obvious relationship with their Ac transposition frequency. The same degree of similarity of spotting size was observed on kernels on ears of plants homozygous for Ac elements that were crossed as females by the reporter stock. When plants homozygous for Ac elements were crossed by pollen of the reporter stock, the relative frequency of fine-spotted to nonspotted kernels differed among the eight Ac elements (Table 1, Figure S1). In the case of the two most distally located Ac elements (mon03080::Ac and bti95004::Ac), the number of nonspotted kernels per 1000 kernels was greater than the number of spotted kernels per 1000 kernels. However, for the six most proximally located Ac elements, the frequency of spotted kernels was greater than the frequency of nonspotted kernels. The relative frequency of nonspotted kernels may depend on the level of transposase transcription by the Ac elements, and this level may vary among the Ac elements at both their original sites on chromosome arm 1S and at the target sites of the tr-Ac elements. Where the sum of the resulting transposase levels is high enough, then the negative dosage effect (Heinlein 1996;Kunze and Weil 2002) exerted upon transposition of Ds from the reporter locus may suppress all spotting, even though Ac elements are present in the kernel. However, inasmuch as the differences in spotting frequency are a reflection of the frequency of the transposition of the Ds element from the r1-sc:m3 reporter locus, these differences may not be directly linked to the frequency of transposition of the Ac element. Whereas the accessibility of the transposase to the Ds element is likely to be similar in these isogenic lines, the accessibility of the Ac element to transposase may differ among the Ac elements and may be influenced by its position in the chromosome. Consequently, the extent to which an increased Ac dosage results in nonspotted kernels may be Ac element-specific but not directly related to Ac transposition frequency.
The same considerations regarding the use of isogenic lines containing the same Ds reporter apply to the present study wherein significant differences in Ac transposition frequencies are documented (Table 1, Table S3, Table S4). It was earlier noted by Kolkman et al. (2005) that previous studies have indicated that methylation plays an important role in altering transcription patterns of Ac and that, therefore, the variation in Ac-mediated excision patterns of the Ds element is likely a result of differences in the transcriptional activity of the different Ac elements and that their mapped Ac lines may be useful for analyzing "cis-acting elements that control gene expression throughout the genome." The differences in Ac transposition frequency reported here might possibly result from structural differences among the Ac elements. However, this is not a likely explanation for all eight cases. An examination of Table 1 reveals that the source of both mon00106::Ac and mon00068::Ac was Ac33 on chromosome arm 5S, yet their transposition frequencies differed about 13-fold. Likewise, the source of both mon00192::Ac and bti00252::Ac was Ac12 on chromosome arm 1S, yet their transposition frequencies differed about 3-fold. The data presented here support the notion that the differences in transposition frequency of these eight mapped Ac elements may be a reflection of variability in Ac transcription or accessibility of transposase to the Ac element, resulting from differences in the chromatin environments wherein the Ac elements are located.

ACKNOWLEDGMENTS
The author thanks Tom Brutnell for providing the Ac lines, Elasa Ludvigsen for assistance in weighing kernels, Dale Brunelle for assistance in statistical analysis, and Don Auger, Jan Clark, and two anonymous reviewers for helpful comments on the manuscript.