Introduction
Since the first approval of commercial genetically modified (GM) maize in 1996, the extent of global GM plant cultivation has increased rapidly over the past two decades, reaching 191.7 million hectares in 2018. The principal GM crops in 2018 were soybean, occupying 50% or 95.9 million hectares of the global GM area (ISAAA 2019). Most GM soybeans have been modified with agronomically valuable traits, such as herbicide tolerance, insect resistance, and disease resistance (Kamthan et al. 2016), with herbicide tolerance consistently being the dominant trait. Cultivation of GM crops not only provides enormous economic and social benefits but also has positive effects on the environment in terms of decreasing pollution.
Although GM crops have great benefits, some ecological risks are emerging, such as the evolution of resistance in target insects, the parallel movement of foreign genes, and a decrease in the biodiversity of nontarget organisms (Romeis et al. 2008; Lazebnik et al. 2017). One of the main issues in the safety assessment of GM crops is pollen-mediated gene flow between the GM crop and its wild relatives (Song et al. 2010; Devos et al. 2018).
Annual wild soybean (Glycine soja ) is widely distributed in China, Japan, Korea and northeastern Russia, and China is one of the main distribution areas of wild soybean. It has been reported that wild soybean has excellent characteristics, such as high protein, high yield, and tolerance to salt (Peng et al. 2013), which are valuable genetic resources for cultivated soybean breeding. Since wild soybean is the wild ancestor of cultivated soybean and its chromosome number is the same as that of soybean (2n = 40), outcrossing between cultivated soybean and wild soybean can frequently occur under natural field conditions, although some studies have shown that gene flow between cultivated and wild soybean occurred at very low frequencies (Mizuguti et al. 2009; Nakayama and Yamaguchi 2002; Liu et al. 2020). However, wild soybean commonly grows throughout almost all of China, and their distributions largely overlap with the distributions of cultivated soybean fields, especially in northeastern and southeastern China (Wang and Li 2012). While more favorable conditions, such as flowering synchrony and certain climatic conditions, are available, greater gene flow may be observed (Yook et al. 2020).
Transgene flow from GM crops to their wild relatives may have the potential to exacerbate weed problems by providing novel traits that allow these plants to compete better, produce more seeds, and become more abundant (Snow, 2002), which could result in changing variations in wild populations (Lu and Xia 2011), and will probably cause unwanted ecological consequences.
Models to predict consequences of gene flow from GM crops to wild relatives require a measure of hybrid fitness (Hails et al. 2005; Weis et al. 2005; Allainguillaume et al. 2006), which is mainly reflected by the ability of individuals to survive and reproduce in their new environment (Burke and Rieseberg 2003; Snow et al. 2003; Halfhill et al. 2005; Yang et al. 2011). Individuals with a higher fitness level might be associated with a higher ability to adapt to the environment and may therefore be beneficial for the establishment of a new plant generation.
At present, a few studies were conducted to evaluate the growth performance of hybrids between GM soybean and wild soybean under greenhouse or field conditions (Guan et al. 2015; Kan et al. 2015; Yook et al. 2020). Kan et al. (2015) measured the F1 and F2 hybrids of four wild soybeans and glyphosate-resistant soybean in a greenhouse and found that hybrids had similar pod and seed numbers per plant as their wild relatives. Field experimental studies conducted by Yook et al. (2020) and Guan et al. (2015) also found that F1 and F2 hybrids (especially F2 hybrids) exhibited intermediate characteristics of their parental soybeans in their vegetative and reproductive stages, and all parameters were close to those of wild soybean. The results of previous studies (Guan et al. 2015; Kan et al. 2015; Yook et al. 2020) indicated that hybrids acquired by gene flow from GM soybean to wild soybean were not associated with a fitness cost when there was no glyphosate selection pressure, hybrids resulting from gene flow from GM soybean to wild soybean possess a higher potential of persistence in ecosystem.
The expression of endogenous genes such as Bt andCP4-EPSPS could improve resistance to insects or herbicides; if the transgene is normally expressed in crop-wild hybrids and progenies and inherited between different generations, the transgene may change a certain trait of wild plants, possibly leading to further undesired environmental consequences (Lu 2008). Therefore, to evaluate the risk of GM soybean and its hybrids resulting from gene flow, it is necessary to investigate protein expression data for assessing and monitoring the biosafety of GM crops and hybrids; however, previous studies mainly focused on vegetative and reproductive hybrids (Guan et al. 2015; Kan et al. 2015; Yook et al. 2020), and the protein levels in hybrid plants were not well investigated. In addition, wild soybean seeds have strong dormancy, while cultivated soybean seeds do not (Kubo et al. 2013). The seed dormancy of the progeny of the hybrid obtained from a cross of wild soybean and GM soybean is still not clear, especially in higher hybrid generations, such as the F2 and F3 generations, the hybrid populations will segregate as homozygous resistant plants (RR), heterozygous resistant plants (RS) and homozygous susceptible plants (SS) based on endogenous genes, and the seed dormancy, vegetative growth and fecundity of these three groups are unknown.
As a developing country, China has always attached great importance to the application of GM technology to improve agricultural productivity. After over 20 years of development, GM soybean in China is now closer to the commercialization stage. It is now necessary to monitor the possible gene transfer from GM crops to wild soybean and investigate the characterization of such hybrids before large-scale commercial production of GM soybean. In the present study, we generated and characterized F1, F2 and F3 hybrids between wild soybean and GM soybean lines 40-3-2, and seed dormancy, vegetation and fecundity between herbicide-resistant (homozygous and heterozygous resistant) and susceptible plants in hybrids were measured under greenhouse conditions. The results of this study will provide valuable scientific information for assessing the environmental risk of GM soybeans.