De novoactivated transcription of inserted foreign coding sequences is inheritable in the plant genome

Abstract

The manner in which newborn genes become transcriptionally activated and fixed in the plant genome is poorly understood. To examine such processes of gene evolution, we performed an artificial evolutionary experiment in Arabidopsis thaliana. As a model of gene-birth events, we introduced a promoterless coding sequence of the firefly luciferase (LUC) gene and established 386 T2-generation transgenic lines. Among them, we determined the individual LUC insertion loci in 76 lines and found that one-third of them were transcribed de novo even in the intergenic or inherently unexpressed regions. In the transcribed lines, transcription-related epigenetic marks were detected across the newly activated transcribed regions. These results agreed with our previous findings in A. thaliana cultured cells under a similar experimental scheme. The comparison of the results of the T2-plant and cultured cell experiments revealed that the de novo-activated transcription caused by local chromatin remodelling was inheritable. During one-generation inheritance, it seems likely that the transcription activities of the LUC inserts trapped by the endogenous genes/transcripts became stronger, while those of de novo transcription in the intergenic/untranscribed regions became weaker. These findings may offer a clue for the elucidation of the mechanism via which newborn genes become transcriptionally activated and fixed in the plant genome.

Kozak sequence acts as a negative regulator forde novotranscription initiation of newborn coding sequences in the plant genome

“…We found analogous cases in transgenic plants, in which promoterless LUC genes became transcriptionally activated concomitant with chromatin remodelling around the LUC insertion loci (Kudo, Matsuo, and Satoh et al ., 2020; Hata and Takada et al ., 2020). From the massive analysis of transgenic cultured cells, we also found that transcriptional activation occurred stochastically at 30% of the insertion events across the genome and was independent of chromosomal loci, suggesting that this transcriptional activation reflects the stochastic nature of chromatin remodelling (Satoh and Hata et al ., 2020). Taken together, these findings suggest that gene insertion events stochastically activate local chromatin remodelling to form a transcription-competent chromatin configuration.…”

“…To scrutinize the manner in which newborn promoters occur in the plant genome, we analysed the transcription start sites (TSSs) of previously established transgenic A. thaliana T87 cell lines that harboured exogenously inserted promoterless luciferase ( LUC ) genes (Satoh and Hata et al ., 2020). Because the cells experienced only 5–10 somatic cell divisions without luciferase-based screening, we assumed that they had retained the characteristic features of newborn genes.…”

“…Among the 550 LUC inserts, 74% were associated with a single TSS and the remainder were associated with two or more TSSs (Figure 1d). The LUC inserts were unbiasedly distributed over the A. thaliana genome (Satoh and Hata et al ., 2020), whereas the LUC loci identified in this TSS analysis were over-represented in the genic regions (Figure 1e). This bias might reflect the fact that the inserts in the genic regions have relatively higher transcription levels and that their cDNAs were more easily obtained than were those located in intergenic regions.…”

“…De novo transcription had the following characteristics: (1) its TSS was located at a Py-Pu dinucleotide located ~100 bp upstream of the LUC insert; (2) it tended to have an AT-rich region located ~30 bp upstream of the TSS; (3) inherent promoter-like epigenetic profiles were not needed; and (4) its TSS avoided overlap with pre-existing Kozak-containing ORFs. These analyses were performed using transgenic cells that experienced only 5–10 vegetative cell divisions, and were not screened for luciferase activity (Satoh and Hata et al ., 2020). Therefore, these characteristics were intrinsic properties of noticeably young promoters that were observed right after their birth, before their exposure to evolutionary selective pressures.…”

“…Here, we experimentally simulated gene origination processes in the plant genome to elucidate the manner in which newborn genes become transcriptionally active shortly after their birth. To overcome the low-throughput drawback of promoter/gene-trapping experiments, we previously applied a massively parallel reporter assay (Inoue and Ahituv, 2015) to the conventional promoter-trapping screenings, and established transgenic Arabidopsis thaliana T87 cell lines individually harbouring promoterless LUC open reading frames (ORFs) (Satoh and Hata et al ., 2020). Based on the precise mapping of LUC -TSSs, we identified de novo TSSs; they occurred de novo about 100 bp upstream of the inserted coding sequences with specific avoidance of pre-existing putative ORFs containing a Kozak motif.…”