**Xiaoru Fan 1, Jingguang Chen 1,2, Yufeng Wu 3, CheeHow Teo 4, Guohua Xu <sup>1</sup> and Xiaorong Fan 1,\*,**†


Received: 12 January 2020; Accepted: 1 March 2020; Published: 6 March 2020

**Abstract:** Transgenic technologies have been applied to a wide range of biological research. However, information on the potential epigenetic effects of transgenic technology is still lacking. Here, we show that the transgenic process can simultaneously induce both genetic and epigenetic changes in rice. We analyzed genetic, epigenetic, and phenotypic changes in plants subjected to tissue culture regeneration, using transgenic lines expressing the same coding sequence from two different promoters in transgenic lines of two rice cultivars: Wuyunjing7 (WYJ7) and Nipponbare (NP). We determined the expression of *OsNAR2.1* in two overexpression lines generated from the two cultivars, and in the RNA interference (RNAi) *OsNAR2.1* line in NP. DNA methylation analyses were performed on wild-type cultivars (WYJ7 and NP), regenerated lines (CK, T0 plants), segregation-derived wild-type from *pOsNAR2.1-OsNAR2.1* (SDWT), *pOsNAR2.1-OsNAR2.1*, *pUbi-OsNAR2.1*, and RNAi lines. Interestingly, we observed global methylation decreased in the T0 regenerated line of WYJ7 (CK-WJY7) and *pOsNAR2.1-OsNAR2.1* lines but increased in *pUbi-OsNAR2.1* and RNAi lines of NP. Furthermore, the methylation pattern in SDWT returned to the WYJ7 level after four generations. Phenotypic changes were detected in all the generated lines except for SDWT. Global methylation was found to decrease by 13% in *pOsNAR2.1-OsNAR2.1* with an increase in plant height of 4.69% compared with WYJ7, and increased by 18% in *pUbi*-*OsNAR2.1* with an increase of 17.36% in plant height compared with NP. This suggests an absence of a necessary link between global methylation and the phenotype of transgenic plants with *OsNAR2.1* gene over-expression. However, epigenetic changes can influence phenotype during tissue culture, as seen in the massive methylation in CK-WYJ7, T0 regenerated lines, resulting in decreased plant height compared with the wild-type, in the absence of a transformed gene. We conclude that in the transgenic lines the phenotype is mainly determined by the nature and function of the transgene after four generations of transformation, while the global epigenetic modification is dependent on the genetic background. Our research suggests an innovative insight in explaining the reason behind the occurrence of transgenic plants with random and undesirable phenotypes.

**Keywords:** genetic; epigenetic; global methylation; transgenic; phenotype; OsNAR2.1

#### **1. Introduction**

Transgenic technologies allow gene transfer to completely unrelated organisms and their application in agriculture has increased the global transgenic crop cultivation to 181 million hectares [1]. In addition, as the basis of transgenic technology, tissue culture is also used for the clonal propagation and regeneration of many plants [2,3]. Although cultured material is not expected to show many genetic changes compared with the original material, there are clear examples of tissue culture material showing heritable phenotypic differences [4–6]. Phenotypic changes in tissue culture-derived material have been found to be caused by epigenetic changes [6–8]. For instance, Rhee et al. [7] demonstrated the silencing of epialleles by epigenetic modifications and showed that the pericarp color1 (*p1)* epialleles were capable of functioning in the presence of the correct trans-acting factors in maize. Furthermore, there are many reports on epigenetic changes caused by plant regeneration in rice [9], garlic [10], triticale [11], pineapple [12], torenia, and rye [13,14].

Variation in plant phenotype is determined by both genetic and epigenetic factors [4]. Epigenetics refers to the study of heritable phenotype changes without genetic alteration [15]. DNA methylation is an epigenetic mechanism that involves the transfer of a methyl group to the C5 position of cytosine and contributes to the epigenetic regulation of nuclear gene expression and to genome stability [16]. In plants, DNA methylation occurs in three sequence contexts: CG, CHG, and CHH (H=A, C or T) [16,17]. The modulation of DNA methylation in culture is crucial to regeneration outcomes: successful regenerants of *Oryza sativa* ssp. *japonica* had lower CG methylation levels than failed regenerants [9]. The regeneration process, with or without genetic transformation, affects gene regulation at the transcriptional and post-transcriptional levels and correlates with changes in DNA methylation patterns [18]. Transgenic approaches have been successfully used to produce herbicide and pest-resistant varieties in several crop species. However, there is only sporadic research concerning global DNA methylation changes in transgenic plants. Potential global DNA methylation modification occurring in transgenic plants is still largely unexplored [18–23]. There have been reports of transgenic plants from different genetic backgrounds having random and undesirable phenotypes [24–26]. Transgenic maize with overexpressed *DREB3,* a dehydration-responsive element-binding transcription factor, showed higher yields in some genetic backgrounds but not others [24]. Researchers have suggested that insert transgene position, metabolic imbalances and environmental constraints [24,27] could be the reasons behind undesirable phenotypes of transgenic plants. However, we suspect that epigenetic changes in transgenic lines may be another reason for the occurrence.

Rice (*Oryza sativa* L.) is a major staple food for a large part of the global population. To determine whether genetic and global epigenetic modification influences the phenotype of transgenic rice, we used whole-genome bisulfite sequencing (WGBS) and methylated DNA immunoprecipitation sequencing (MeDIP-seq) methods to determine DNA methylation at the genomic level in various transgenic plants and their controls. The transgenic process includes tissue induction, selection pressure and insertion of the transgene [28]. We used wild-type Wuyunjing7 (WYJ7), its regenerated line (CK), and segregation-derived wild-type from *pOsNAR2.1-OsNAR2.1* (SDWT) to represent different transgenic processing, and wild-type of Nipponbare (NP), its *OsNAR2.1* overexpression line under the ubiquitin promoter (*pUbi-OsNAR2.1*), and the *OsNAR2.1* RNA interference (RNAi) NP line to represent different backgrounds and different transgenic expression. Previous studies have shown that OsNAR2.1 is a partner protein of rice high-affinity nitrate transporters (OsNRT2s) [29–31] and plays a key role in enabling the plant to cope with variable environmental nitrate supplies [29–31]; overexpression of *OsNAR2.1* can lead to an increase in grain yield and higher nitrogen-use efficiency (NUE) in rice cultivation systems [32,33]. We used *OsNAR2.1* transgenic lines as representative lines for the transgenic process.
