A Stowaway transposon disrupts the InWDR 1 gene controlling flower and seed coloration in a medicinal cultivar of the Japanese morning glory

Atsushi Hoshino*, Yoshiaki Yoneda and Tsutomu Kuboyama National Institute for Basic Biology, Nishigonaka 38, Myodaiji, Okazaki 444-8585, Japan Department of Basic Biology, SOKENDAI (The Graduate University for Advanced Studies), Nishigonaka 38, Myodaiji, Okazaki 444-8585, Japan Emeritus Professor, Shizuoka University, Ohya 836, Suruga-ku, Shizuoka 422-8529, Japan College of Agriculture, Ibaraki University, Chuou 3-21-1, Ami, Ibaraki 300-0393, Japan

The Japanese morning glory (Ipomoea nil) is a traditional floricultural plant in Japan.A number of its spontaneous mutants have been isolated since the early 19th century (Imai, 1927;Yoneda, 1990;Iida et al., 2004).Most of the mutants exhibit alterations of floricultural traits in flower and leaves.Recent studies revealed that transposons in the Tpn1 family cause a large proportion of the I. nil mutations.The a3 f , sp, r1, pr and efp mutations related to flower coloration, and the dp and fe mutations involving leaf and flower shapes, are caused by the insertion of Tpn1 family elements (Inagaki et al., 1994;Fukada-Tanaka et al., 2000;Hoshino et al., 2001Hoshino et al., , 2009;;Nitasaka, 2003;Iwasaki and Nitasaka, 2006;Morita et al., 2014).These elements are DNA transposons and belong to the En/Spm or CACTA superfamily (Inagaki et al., 1994;Kawasaki and Nitasaka, 2004).DNA transposons including Tpn1 family elements sometimes generate mutable alleles that spontaneously show recurrent somatic and germinal mutations.
DNA transposons generate target site duplication (TSD) upon their integration.They often leave small DNA rearrangements, called footprints, at their excision sites.Footprints sometimes include sequences of duplicated target sites.The length of the TSD of the Tpn1 family elements is 3 bp.Among the mutations in I. nil, several-bp insertions have been found repeatedly.The dy, ca and dk mutations are 4-bp (GGAT and CGAT in dy-1 and dy-2, respectively), 7-bp (GGAGTAC and TCCG-TAC in ca-1 and ca-2, respectively), and 4-bp (GGCC in dk-1) insertions (Morita et al., 2005(Morita et al., , 2006(Morita et al., , 2015)).The 3′most 3-bp sequences of the presumed insertion sequences are identical to the 3-bp sequences adjacent to their 5′ ends.These sequence features suggested that they were footprints of Tpn1 family elements.Thus far, the transposons generating these putative footprints have not been identified.
In the seed stock made by Nariyasu Shimazu of Ibaraki University, a line named Yakuyou-shirohana was collected.In Japanese, the name Yakuyou means medicinal.I. nil was introduced into Japan from China as a medicinal herb, and its seeds have been used as laxatives.In addition to wild-type black seeds, whitish seeds also have long been known as laxatives.Yakuyou-shirohana shows white green stems and whitish seeds (Fig. 1), while wild-type plants produce blue flowers, red-pigmented stems and black or dark brown seeds.From these phenotypes, we hypothesized that the line carries a ca mutation.The Ca gene, also called the InWDR1 gene, encodes a WD40 repeat protein (Morita et al., 2006).It activates anthocyanin pigmentation in flowers and stems, and is essential for normal seed pigmentation by proanthocyanidins and phytomelanins and for seed-trichome formation (Morita et al., 2006;Park and Hoshino, 2012).
To characterize the InWDR1 gene in Yakuyou-shirohana, we employed PCR analysis.Genomic DNA was isolated from flower petals using an automated DNA isolation system, NA-2000 (Kurabo, Osaka, Japan).InWDR1 fragments that included the whole InWDR1 coding sequence were amplified using the primers CAKPN-F1 (CCGGGG-TACCGGATCCCATGCATCCAGTGT) and CAKPN-R1 (GCGCGGTACCTACTGGGAGGTGGACTGTCT) (Park and Hoshino, 2012).The fragment from the mutant was approximately 4.5 kb long-about 800 bp longer than that of the wild-type plant-indicating that the mutant carries an insertion sequence (data not shown).To characterize the insertion further, the PCR products were purified using the QIAquick PCR Purification Kit (QIAGEN, Hilden, Germany) and were directly sequenced on an ABI Prism 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).The insertion sequence was found to be a Stowaway-like transposon, which we named InSto1 (Fig. 2A).Stowaway elements are miniature invertedrepeat transposable elements (MITEs) related to the Tc1/ Mariner superfamily (Bureau and Wessler, 1994;Feschotte et al., 2002Feschotte et al., , 2003)).MITEs are short non-autonomous class II elements in eukaryotic genomes (Feschotte et al., 2002).Although Stowaway elements are abundant in various plant genomes, very few Stowaway insertion mutations that affect the visible phenotype have been reported.In the potato, a Stowaway insertion within the gene that encodes flavonoid 3′,5′-hydroxylase resulted in alteration of the tuber anthocyanin coloration (Momose et al., 2010).These elements have a preference for insertion at the dinucleotide TA.InSto1 is 849 bp long and is inserted within the unique exon of the InWDR1 gene.The element is flanked by a TSD of the dinucleotide TA.It has an 11-bp terminal inverted repeat (TIR) sequence (Fig. 2B) that shows high similarity with the conserved TIR sequence of Stowaway elements (Fig. 2C; Feschotte et al., 2002).A BlastX search revealed that InSto1  carries no coding sequences for transposases or known proteins.This strongly suggests that InSto1 is a nonautonomous element.We designated the InWDR1 allele with the InSto1 insertion as ca-3.Interestingly, the insertion site of InSto1 is 1 bp upstream to that identified for the 7-bp insertions in the ca-1 and ca-2 alleles (Fig. 2B).The original interpretations of the 7-bp insertion sites were therefore presumably wrong, and the 7-bp sequences (CGGAGTA and CTCCGTA in ca-1 and ca-2, respectively) that share an identical insertion site with InSto1 represent the correct insertions.The newly interpreted 7-bp insertions have 2 bp of TA at their 3′ end.This strongly suggests that InSto1 generated the 7-bp insertions as footprints, the wild-type sequence TA being changed into TA(CGGAG/ CTCCG)TA upon InSto1 excision.In other words, ca-3 was a precursor of ca-1 and ca-2.Consistent with this notion, Yakuyou-shirohana and a ca-1 line (NS/W1ca1) had identical InWDR1 sequences except for the InSto1 insertion site, while several polymorphisms were found between the 5′ upstream regions of Yakuyou-shirohana and a wild-type line, Tokyo Kokei Standard (Fig. 2A).The 5′ upstream regions in a second ca-1 (NS/W2ca3) and six ca-2 lines have not been sequenced.In contrast, the 4-bp insertions in the dy-1, dy-2 and dk-1 alleles cannot be footprints of InSto1 or related Stowaway elements because the 4-bp insertions have no TA dinucleotide at their 3′ ends.
According to our survey of Japanese historical literature, the first reliable description of whitish seeds of I. nil appeared in an agricultural book, Nogyo Zensho, published in 1697 (Miyazaki and Kaibara, 1697).The oldest picture of whitish seeds was printed in an encyclopedia, Wakan Sansai Zue (Terashima, 1712).Therefore, a mutation that confers whitish seeds was isolated at the end of the 17th century, at the latest.This is more than a hundred years before the appearance of most of the other I. nil mutants.All nine lines showing whitish seeds characterized thus far have been found to carry one of the recessive ca mutations, ca-1, ca-2 or ca-3 (Morita et al., 2006, this study).Therefore, the mutants described in Nogyo Zensho and Wakan Sansai Zue probably had ca-3 or its derivatives, and ca-3 would thus be one of the oldest mutations in I. nil.In other words, the origin of the ca-3 mutation is likely to have been different from that of most mutations related to floricultural traits.Floricultural mutations were assumed to have been isolated from lines carrying active Tpn1 family elements in Japan (Iida et al., 2004).Whitish seeds of I. nil have also been used as a medicinal herb in China for a long time.It can be hypothesized that mutants carrying ca-3 were first isolated in China and were introduced into Japan.Consistent with this hypothesis, the ca-3 allele was found in a Chinese traditional herbal line, Baichou.Wang et al. (2013) found a mutant allele of the InWDR1 gene from Baichou.While no details of the allele were reported, the sequence of the allele was deposited in GenBank (accession number KF384189) as the WDR1sino allele.Sequence comparison revealed that WDR1sino and ca-3 are the same allele, although InSto1 is annotated as a Tpn-like transposon rather than a Stowaway transposon in the GenBank record.
No evidence indicating the transposition activity of InSto1 could be detected.No fragments derived from footprints were amplified, even by nested PCR (data not shown).Interestingly, approximately 300 years ago, it was reported that blue sectors were occasionally generated in white flower petals in a whitish seed line, Ginsen (Mimura, 1723).Ginsen possibly carried the ca-3 allele.If this is the case, InSto1 occasionally transposed in that period and caused somatic reversions of the InWDR1 gene, resulting in the appearance of chimeric blue-colored sectors on white petals.We are currently screening ca-3 lines for active InSto1 and searching for InSto1-related autonomous elements.

Fig. 2 .
Fig. 2. Characterization of the ca-3 allele.(A) Structure of the 4.5-kb PCR fragment containing InWDR1 in the ca-3 mutant.The larger box indicates the unique exon of InWDR1; the coding region is shaded gray.The box with filled arrowheads shows InSto1.Vertical bars indicate the positions of polymorphisms between the ca-3 mutant and a wild-type line, Tokyo Kokei Standard: substitutions and 9-and 1-bp insertions are indicated as sub, 9-bp ins and 1-bp ins, respectively.(B) Sequences of the insertion site of InSto1 and the putative footprints in ca-1 and ca-2 alleles.The TA dinucleotide target site is indicated in boldface.The 5-bp insertion sequences putatively created upon InSto1 excision are shown in uppercase.The underlined 7-bp sequences were previously thought to be footprints of the Tpn1 family element (Morita et al., 2006) because their 3′-most 3-bp sequences (TAC) are identical to the 3-bp sequences immediately upstream of the putative footprint sequences.(C) Comparison of the conserved TIR sequence of Stowaway elements and the InSto1 TIRs.