Differential transformation efficiency of Japonica rice varieties developed in northern China

The production of japonica rice in northern China plays an important role in food security for the world. Agrobacteriummediated gene transfer is becoming a powerful approach to generate germplasm and develop varieties. However, the transgenic efficiency of the japonica rice varieties in northern China has been completely unknown, which obstructs the development of transgenic breeding and the exploration of gene functions. In this study, the transgenic efficiencies of six japonica rice varieties developed in northern China are evaluated. The rates of primary and secondary callus induction of the varieties are similar. However, transgenic efficiency and the regeneration ability of the varieties are greatly different. The results have established a platform for transformation of the rice varieties and proposed a suitable variety, SN9816, for gene transfer. SN9816 can be applied as an elite germplasm for transgenic breeding and basic research of molecular biology in northern China or an area in the same latitude.


INTRODUCTION
Rice (Oryza sativa L.), which is cultivated in more than 115 countries and feeds over one third of the population worldwide, is one of the world's most important crops (Khush 1997, Raimondi et al. 2014).More than 90% of the world's rice is grown and consumed in Asia, where 60% of the earth's people live.Rice is also one of the major staple foods in China, a center of origin of Asian cultivated rice (Khush 1997, Londo et al. 2006).The production of japonica rice in northern China is famous for its higher yield and good quality and accounts for 50% of the total production.Enhancing the yield and improving the quality of rice through breeding are effective ways of responding to increasing food demands.Molecular breeding, including marker assistant selection (MAS), transgenic breeding, and genome-wide association study (GWAS) provides breeders with new challenges and opportunities to increase the efficiency of selection, shorten the breeding period, and generate new germplasm with improved traits, which are difficult to improve by conventional methods (Jung et al. 2008).
Improvement of rice varieties through transferring genes or cis-acting elements from rice, plants, or other organisms to rice to confer new agronomic traits in rice, termed transgenic breeding, is an effective way to create new germplasm and breed new varieties (McLaren 2005).Agrobacterium-mediated transformation is a general method for genetic modification in many plant species because it allows efficient insertion of stable, unrearranged, single-copy sequences into plant genomes (Nishimura et al. 2006).Efficient methods for Agrobacterium-mediated transformation of japonica rice have been developed (Hiei et al. 1994, Hiei et al. 1997, Hiei and Komari 2008).Two japonica varieties, Nipponbare and Kitaake (developed in Japan), are good genotypes for Agrobacterium-mediated transformation.Based on this method, Agrobacteriummediated transformation methods have been improved, with minor modification for many rice varieties, not only japonica but also indica and tropical japonica (Nishimura et al. 2006, Hiei and Komari 2008, Saika and Toki 2010).
It has been demonstrated that transformation efficiencies are genotype dependent (Hiei et al. 1997, Hiei andKomari 2008).Although efficient procedures for Agrobacteriummediated transformation of either japonica or indica rice have been developed, the transgenic efficiency of the rice varieties developed in northern China have been completely unknown (Hiei andKomari 2008, Saika andToki 2010).Additionally, because of the higher latitude in northeast China and the sensitivity of rice heading to day length, Nipponpare could not successfully develop seed in the field under normal conditions.Even though Kitaake could be used to replace Nipponbare in northern Asia, it was hard for breeders to use Kitaake as breeding germplasm because the plant size of Kitaake is smaller than desired and its agronomic traits are not as good as expected.These factors restrict the development of transgenic breeding and the performance of molecular biological study in northeast China.In this study, five varieties which are popularly cultivated in northern China are used as explants to compare their response to Argobacterium-mediated transformation by using Nipponbare as control.Our studies have established a platform for Agrobacterium-mediated transformation of the japonica rice varieties developed in northern China and proposed a suitable variety, Shen Nong 9816, for Agrobacterium-mediated gene transfer with a transgenic efficiency and regeneration rate similar to Nipponbare.

Plant materials
Six varieties of japonica-type rice were used in the study, including Shen Nong 9816 (SN9816), Shen Nong 265 (SN265), M31, Liao Xing 1 (LX1), Toyonishiki (Toyo, from Japan), and Nipponbare (Nipp, from Japan, sequenced).Among them, SN9816, SN265, M31, LX1, and Toyonishiki are widely cultivated in northern China, which is the major region for japonica rice production in China.Nipponbare was used as control, the seeds of which were produced in the Experimental Station of the Chinese Academy of Agricultural Science, Beijing (lat 39 o 92' N, long 116 o 46' E).The mature seeds of the other five varieties used for callus induction were produced in the Experimental Station of the Rice Research Institute at Shenyang Agricultural University (lat 41 o 80' N, long 123 o 38' E) in northeast China.

Binary vector and Agrobacterium-mediated transformation
The pCambia 1301 binary vector was applied, harboring the expression cassettes of the 35S:intron-GUSA gene (tagged by hexaHis), the 35S:aadA gene (conferring kanamycin resistant to bacterium), and the 35S:hptII gene (conferring hygromycin resistance to callus or plantlets) (Figure 1).The binary vector was transformed into the Agrobacterium strain EHA 105 through the Freeze-Thaw method.

Methods used in statistical analysis
Three independent biological replicates were performed.The data were analyzed using the GraphPad Prism 5 software.The Tukey multiple comparison test of one-way analysis of variance (ANOVA) was used to detect the significance of differences.

Genomic DNA extraction and PCR analysis
For extraction of genomic DNA, rice seedlings were harvested, immediately frozen in liquid N 2 , and stored at -80℃.Genomic DNA was extracted through the 2×CTAB method.PCR analysis to detect the hpt gene was performed by using rTaq DNA polymerase (Takara Biotechnology, Cat# R500Z) with the primer pair of forward primer, HygF CTTCTGCGGGCGATTTGT, and reverse primer, HygR CAGCGTCTCCGACCTGAT.

Callus formation abilities are similar in the varieties
It is known that some rice varieties are tissue-culture sensitive, especially the indica rice varieties.To evaluate the ability for callus formation, sterilized mature seeds of the varieties were inoculated in N6D medium and grown for 7 days or 3 weeks.The status of primary callus growth, the rate of primary or secondary callus induction, and the rate of secondary callus growth were observed (Figure 3).
Yellow, compacted primary callus can be induced from all six varieties, but the emergence of primary calli from Liao Xing 1 (LX1), M31, and Shen Nong 265 (SN265) was later and slower than that of Nipponbare (Nipp), Shen Nong 9816 (SN9816), and Toyonishiki (Toyo) (Figure 3A).The rates of primary or secondary callus induction of the varieties were similar (Figures 3B, C).No significant difference in the rate of primary or secondary callus induction of the varieties was analyzed.However, a highly significant difference in the fresh weight of 100 callus cluster of the secondary calli (FWSC) after 15 days of subculture was observed among the varieties (Figure 3D).The FWSC of LX1 (13.94g) was the highest, and that of Toyo (7.97g) was the lowest.The FWSC of LX1 was significantly higher than that of Nipp (10.16g).The FWSC of LX1 (13.94g),SN265 (12.33g),M31 (11.36g), or SN9816 (11.09g) was significantly higher than that of Toyo.However, no significant difference of the FWSC was detected between Nipp and SN9816, between Nipp and Toyo, and among SN9816, LX1, M31, and SN265.These results indicated that the secondary calli induced from the five varieties developed in northern China are of good status for Agrobacterium-mediated transformation compared to Nipp.Thus, we conclude that it is easier to induce calli from the mature seeds of the varieties, including SN9816, LX1, Toyo, M31, and SN256.

Transformation efficiencies of the six varieties are vary
To investigate the transformation rates of the japonica rice varieties, 3-week-old secondary calli were infected with the Agrobacterium strain EHA105 harboring the pCambia 1301 vector.After 3 days of co-cultivation with Agrobacterium, transient expression efficiency of GUS was determined (Figure 4A).Intron-GUS is a convenient reporter gene since this gene is strongly expressed in rice cells but not in Agrobacterium cells that attach to the tissues (Chan et al. 1993).The rates of the GUS stained calli were analyzed (Figure 4B).The results showed that the GUS staining rate of SN9816 (59.17%) was the highest, which was higher than that of Nipp (48.78%).However, the rates of LX1 (20%), Toyo (18.89%),M31 (17.50%), and SN265 (25.54%) were lower than that of Nipp (48.78%).
After 3-week selection in the N6D medium with 50 mg L -1 Hyg (N6D-S), the proliferation of Hyg-resistant calli was observed from all the varieties (Figure 4C).However, the rates of Hyg-tolerant calli or transgenic efficiencies of the varieties were different (Figure 4D), which is consistent with the previous results that transgenic efficiency is genotype dependent (Hiei et al. 1997, Hiei and Komari 2008, Saika and Toki 2010).The highest rate of Hyg-resistant calli was obtained from Toyo (92.69%) and the lowest rate from SN265 (36.61%).Toyo (92.69%),SN9816 (91.60%), and Nipp (82.96%) had much more significant rates than LX1 (33.62%),M31 (34.52%), and SN265 (36.61%) (Figure 4D).There was no significant difference among LX1, M31, and SN265.These results indicated that, in addition to Nipp, both SN9816 and Toyo were genotypes tolerant to Agrobacterium-mediated gene transfer, which gave rise to higher transformation efficiency after secondary selection on Hyg containing N6D-S medium (Figure 4D).

The regeneration ability of the varieties differ from each other
To explore the regeneration capability of the varieties, the secondary calli were transferred directly to the MS-NK regeneration medium.The regeneration rates without transformation of the varieties on an inoculated callus basis were counted.The results revealed that the calli derived from all the varieties had the ability to regenerate (Figure 5A), but the rates of regeneration of the varieties differed.The regeneration rate of Toyo (73.99%) was the highest, and that of M31 (22.93%) was the lowest (Figure 5C).The regeneration rates of Toyo (73.99%),Nipp (66.12%),SN9816 (64.66%), and SN265 (52.50%) were significantly higher than LX1 (35.55%) and M31 (22.93%) (Figure 5C).The regeneration rate of SN265 was significantly higher than that of M31 (Figure 5C).No significant difference was detected among Toyo, Nipp, SN9816, and SN265, and between LX1 and SN265 (Figure 5C).These results suggest that there is a greater possibility of growing plants from the callus for Toyo, Nipp, SN9816, and SN265.However, the question remains as to whether the regeneration ability of

Regeneration of the varieties is transgenic sensitive
To investigate the response of regeneration to Agrobacterium-mediated transformation, the Hyg-tolerant calli of the varieties in N6D-S medium were transferred to the MS-NK regeneration medium.Regeneration ability was evaluated.The results showed that the regeneration rates after transformation of all the varieties were lower than the rate of calli subcultured in the regeneration medium without transformation (Figures 5C, D).The highest regeneration efficiency after Agrobacterium-mediated transformation was observed from the Nipp variety (60.98%), and the lowest was from LX1 (23.33%) (Figure 5D).The regeneration rates after Agrobacterium-mediated transformation of Nipp (60.98%) and SN9816 (52.75%) were significantly higher than those of LX1 (23.33%),SN265 (27.14%), and M31 (29.74%) (Figure 5D).The rates of Nipp (60.98%) and Toyo (47.43%) were significantly higher than the rates of Toyo (47.43%) and LX1 (23.33%), respectively (Figure 5D).There was no significant difference between Nipp and SN9816 (Figure 5D).These results revealed that the regeneration of all six varieties was sensitive to Agrobacterium-mediated transformation, especially the Toyo variety (Figures 5C, D).Agrobacterium-mediated transformation reduced regeneration ability.However, the regeneration rates of Nipp and SN9816 were similar, at the level of 50% to 60%, based on the number of Hyg-tolerant calli (Figure 5D).

The rates of root growth of the regenerated shoots were similar among the varieties
After regeneration, the regenerated shoots were transplanted to the root growth medium (MS-HF) to allow vigorous root growth (Figure 5B).The root-generating ability among the varieties was similar.No significant difference of root-emergence rates was observed among the varieties without or after Agrobacterium-mediated transformation (Figures 5E, F).We conclude that the regenerated shoots from the calli transformed or non-transformed varieties had  similar ability to generate roots (Figures 5E, F).
To determine the rate of positive transgenic plants, Polymerase Chain Reaction (PCR) was performed on the genomic DNA of the candidate transgenic plants obtained with the primer pair (HygF/HygR) of the Hyg gene in the pCambia 1301 binary vector (Figure 6A).The transgenic rates of the plants from the varieties were measured with the statistical method (Figure 6B).The transgenic rates of the candidates were from 93.75% to 98.58%.No significant difference existed among the varieties.The transgenic plants were then transplanted to the soil to harvest seeds for the subsequent studies (Figure 6C).
To summarize, the transgenic efficiency of five varieties commonly grown in northern China were evaluated in this study and compared to Nipp.Though the rates of the secondary callus growth vary, the proliferation rates of primary and secondary calli of the varieties are similar, indicating that it is easier to induce calli from the mature seeds of the varieties (Figure 3).The transgenic efficiency of the varieties is different after secondary selection in the Hyg-containing N6D-S medium.The transgenic efficiencies of Toyo, Nipp, and SN9816 are similar, but much more highly significant than those of LX1, M31, and SN265 (Figure 4).Agrobacterium-mediated transformation suppresses regenera-tion of the varieties.The regeneration rates of Toyo, Nipp, and SN265 without transformation are higher than those of LX1 and M31 (Figure 5C).The regeneration rates after transformation of Nipp and SN9816 are higher than those of LX1, M31, and SN265 (Figure 5D).The regeneration of Toyo is hypersensitive to Agrobacterium-mediated transformation (Figures 5C, D).We may easily conclude that the performance of SN9816 during Agrobacterium-mediated gene delivery, including callus formation, transgenic efficiency, regeneration rate, and root generation rate, is similar to Nipp.Our data suggest that SN9816, which is the major cultivated super japonica variety with erect panicle, ideal plant architecture, and higher yield, is an elite germplasm for Agrobacterium-mediated transformation in northern China and the area at the same latitude.These results lay the sound foundation for us to generate the large number of T-DNA insertion plants needed for functional analysis of the rice genome and for breeding super japonica rice varieties with additional agronomic value.

Figure 2 .
Figure 2. Procedures of Agrobacterium-mediated transformation for japonica rice varieties in northern China.(A) Sterilization of Seeds.(B) Inoculation of seeds in N6D medium.(C) Formation of primary calli in N6D medium.(D) Induction of secondary calli in N6D medium.(E) Preculture of the secondary calli in N6D medium three days before transformation.(F) Infection of the preculture secondary calli with the Agrobacterium harboring pCambia 1301 vector.(G) Drying of the infected calli before co-cultivation.(H) Co-cultivation with the Agrobacterium for three days in the dark.(I) Proliferation of Hyg-tolerant calli on the secondary selection medium, N6D-S medium.(J) Regeneration of transgenic shoot in root generation medium, MS-HF.The white or black line is the scale bar, representing 2cm or 1cm as indicated.

Figure 3 .
Figure 3. Callus formation of japonica rice varieties in northern China.(A) The 7-day-old primary calli of the varieties.(B) The rate of primary callus induction of the varieties.(C) The rate of secondary callus induction of the varieties.(D) The fresh weight of 100 callus cluster of secondary calli after 15 days of subculture.The white line is the scale bar, representing 1cm.The values are the mean and standard deviation (SD) from three independent biological experiments.SN9816: Shen Nong 9816; LX1: Liao Xing 1; SN265: Shen Nong 265.One asterisk (*) represents significant difference (P<0.05).Two asterisks (**) represent highly significant difference (P<0.01).

Figure 4 .
Figure 4. Agrobacterium-mediated transformation efficiency of the varieties in northern China.(A) Transient expression of 35S:intron-GUS after 3 days of co-cultivation with Agrobacterium.(B) The rate of GUS stained calli of the varieties.(C) The proliferation of the Hyg-tolerant calli in secondary selection medium, N6D-S medium.(D) Transgenic efficiencies of the varieties in northern China after secondary selection.Three biological replicates were performed.Similar results were obtained.The results from two or three replicates are shown.The values are the mean and standard deviations from three independent biological experiments.One asterisk (*) represents significant difference (P<0.05).Two asterisks (**) represent highly significant difference (P<0.01).SN9816: Shen Nong 9816; LX1: Liao Xing 1; SN265: Shen Nong 265.White lines are the scale bar, representing 1cm or 2cm as indicated.

Figure 5 .
Figure 5. Regeneration ability of the varieties in northern China.(A) Regenerated shoots from the Hyg-tolerant tissues.(B) Plantlets with roots derived from Agrobacterium-mediated transformation.(C) Regeneration rates without Agrobacterium-mediated transformation of the varieties in northern China.(D) Regeneration rates with Agrobacterium-mediated transformation of the varieties in northern China.(E) The rate of root generation from the regenerated shoots without transformation.(F) The rate of root generation from the regenerated shoots with Agrobacteriummediated transformation.Three biological replicates were performed.Similar results were obtained.The results from two or three replicates are shown.The values are the mean and standard deviationsof two or three independent biological experiments.One asterisk (*) represents significant difference (P<0.05).Two asterisks (**) represent highly significant difference (P<0.01).SN9816: Shen Nong 9816; LX1: Liao Xing 1; SN265: Shen Nong 265.White lines are the scale bar, representing 2cm or 4cm as indicated.

Figure 6 .
Figure 6.Detection of transgenic plantlets with PCR.(A) A gel image of PCR products with the HygF/HygR primer pair on Hyg-tolerant transgenic plantlets.The number represents the individual transgenic plantlets.The asterisk (*) labeled 5 and 12 represents negative plantlets.(B) Transgenic efficiency detected by PCR based on the number of the rooted regeneration plantlets.(C) Transgenic plants in heading stages.The white line is the scale bar, representing 10cm.