*3.3. TEs with Methylcytosine Affected the Expression of Anthocyanin-Related Genes*

Various evidence supports the role of TE insertions in gene expression changes and phenotypic variation in higher plants. In anthocyanin biosynthetic genes, for example, TE-induced insertions cause null mutations that result in variations in seed, peel, and flower coloration. The insertion and excision of the *Ds* transposon has given rise to variation in the size and intensity of colored spots in maize kernels, and a *Candystripe 1* insertion in the second intron domain of the *y-candystripe* allele has altered the pigmentation of the sorghum grain pericarp from solid red to variegated [8,73,74]. Similarly, the *TRANSPARENT TESTA8* (*BrTT8*) locus, encoding a bHLH protein, lost its function with the insertion of a *Helitron* transposon, resulting in yellow seeds in *Brassicarapa* [75]. The golden

pigmentation in hulls and internodes of *Oryza sativa* mutants is due to the complete suppression of the *OsCHI* gene following insertion of a *Dasheng* retrotransposon into its 50 untranslated region (UTR), while a retrotransposon insertion in the upstream sequence of the pigmentation-related gene *VvmybA1* is regarded as the molecular basis for white-skinned coloration in grape cultivars [15,76]. TE-mediated insertional mutations have also caused alterations in seed coat and flower color in both *Ipomoea purpurea* and *Glycine max* [77–80]. Transformed tobacco plants carrying an inserted *Tag1* element between the *CaMV* 35S promoter and the maize *R* gene have variegated flowers, and the insertion of either *Ty1dic1* or *Retdic1* transposons can disrupt the *AA5GT* (*acyl-glucose-dependent anthocyanin 5-O-glucosyltransferase*) gene to prevent glycosylation of the 50 position of anthocyanins in *Dianthus caryophyllus* [14,17].

One unanswered question concerns how TEs are activated or repressed to ensure a stable phenotype. Emerging evidence is demonstrating that TEs are silenced or reactivated by epigenetic mechanisms such as DNA methylation modification [39,41,42,68,81,82]. In the model plant *A. thaliana*, the imprinted gene *FWA* is a flowering-time modifier, with its silencing dependent on the cytosine methylation of a SINE retro element in the promoter region [83]. Transposons in *Arabidopsis* are heavily methylated at both CG and non-CG sites, whereas non-CG methylation is rarely found in active genes [70]. In rice, the essential *chloroplast protease 5* (*OsClpP5*) gene with the insertion of an epigenetically silenced *autonomous DNA-based active rice transposon 1* (*aDart1*) may induce leaves to show variegation [84]. Similarly, in our study, we detected 197 DMRs associated with 196 TEs inserted into 106 DEGs. We annotated these DEGs with GO and KEGG functional terms and pathways. The genes *Pm008680* (*UGFGT*), *Pm031359* (*UGT79B6*), *Pm031359* (*UGT79B6*), *Pm013782* (*DFRA*), and *Pm011195* (*GSTXC*), all participating directly in the color regulation pathway and anthocyanin transport, were enriched. This result suggests that the insertion of methylcytosine-modified transposons affects the expression of anthocyanin genes, resulting in chimerism.
