The test material seeds of maize inbred lines, H8183, H8184, and H8185, were sown in the experimental field of the Key Laboratory of Crop Molecular Breeding in Jilin Province. After artificial pollination, immature maize embryos were stripped by selecting 10–13-day-old, pollinated, female maize ears. The outer husk was stripped off (leaving only a layer of young leaves) and then the maize were soaked in 75% ethanol for 15 min. Next, the young leaves and corn silk were removed. The remaining immature embryos were disinfected using 0.5% NaClO for 10 min and then rinsed three or four times with sterile water. Finally, a scalpel was used to remove the seed coat and endosperm at the top of the ear and to pull out immature embryos. The immature embryos of H8183, H8184, and H8185 that were pollinated on days 10–13 were then inoculated into the N6 induction medium. Every subsequent 2 weeks, stable callus were formed in approximately 40 days for subsequent infection and transformation.

The Agrobacterium strain EHA105 harboring the empty vector pCAMBIA3301 preserved in our laboratory was used in this study (Fig. 1). The binary vector contains the β-glucosidase (GUS) reporter gene and the herbicide resistance bar gene. Phosphinothricin acetyltransferase (PAT) encoded by the bar gene can inactivate phosphinothricin. Glufosinate-ammonium was used as a screening agent.

A single colony of Agrobacterium strain harboring the pCAMBIA3301 plasmid was selected and inoculated in YEP liquid medium containing appropriate antibiotics and incubated overnight on a shaker set at 28°C and 200 rpm. Bacteria were collected using centrifugation when OD600 reached 0.6. Following this, the bacterial pellet was resuspended in the infection solution (Agrobacterium-mediated transformation solution) containing 200 μmol/L acetosyringone (AS).

Maize callus, precultured on the induction medium for 3 days, were used for Agrobacterium-mediated transformation. The maize callus and bacterial solution were poured into conical flasks and shaken using a shaker at 28°C and 80 rpm for 20–30 min. The bacterial solution was discarded and the callus were transferred to a Petri dish with 2–3 layers of filter paper to absorb any excess bacterial solution. Subsequently, two coculture methods were adopted: (1) The callus were incubated directly on a solid coculture medium in the dark at 25°C for 2–3 days. (2) The callus were transferred to a solid coculture medium with a layer of sterile filter paper and incubated at 25°C for 2–3 days.

After coculture, the callus were transferred to the first screening medium, containing 0, 2, 4, or 6 mg/L glufosinate-ammonium (screening agent), for approximately 25 days. The callus from the first screening were then transferred into the corresponding screening medium for the second screening, which lasted for approximately 25 days. The survival and adventitious bud induction rates were calculated. The results were analyzed using GraphPad Prism v.7.0 (A scientific software officially launched by GraphPad).

The screened resistant callus were transferred to a differentiation medium (Tab. 1) containing different concentrations of additives and exposed to light at 25°C for 16 h/d. Each flask was inoculated with 10 callus from each group with five replicates, using three factors and four levels of orthogonal test. The data from nine groups were processed by the L₉ (3⁴) method, and the results were observed after the callus were differentiated into stems and leaves. The differentiation averages and differentiation rates of callus were also calculated.

Plants generated from callus differentiation were transplanted into a rooting medium post selection. When three to five long roots of approximately 5 cm were observed, the plants were cleaned for hydroponic culture.

GUS staining was performed on young roots of the transplanted resistant regenerated plants. Before staining, the young roots were thoroughly washed with a phosphate buffer, following which they were stained according to the previously described GUS staining protocol [13]. The samples were placed in a 1.5-mL centrifuge tube, followed by the addition of GUS staining solution (freshly prepared) until the samples were completely covered. The tubes were wrapped in aluminum foil and placed at 37°C overnight. Staining was observed under a microscope, and the instantaneous expression level of GUS gene was calculated (number of stained immature embryos/number of infected immature embryos).

Leaves from the transplanted putative transgenic plants were used for performing the Bar protein antibody test-strip analysis. Approximately 1 cm of the leaves was taken in a 1.5-mL centrifuge tube and snap-frozen using liquid nitrogen; 0.4 mL extraction buffer was added, and the test paper was removed. The end of the strip marked with an indicator arrow was immersed in the buffer for 1 min. The expression of Bar protein was observed. CATB method [14] was used to extract maize genomic DNA, which was later used as a template for polymerase chain reaction (PCR). Specific primers for Bar-R: ATGAGCCCAGAACGACGCC and Bar-L: TCAAATCTCGGTGACGGGC were used for PCR. The PCR amplification cycle was as follows: predenaturation: 94°C for 5 min; denaturation: 94°C for 30 s; annealing: 58°C for 30 s; extension: 72°C for 30 s, 35 cycles; and extension: 72°C for 8 min. After the reaction, the PCR products were detected using 1% agarose gel by electrophoresis. Accordingly, the number of transgenic plants was affirmed.

Three inbred lines, H8183, H8184, and H8185, were inoculated into the N6 medium to generate immature embryos after 10–13 days of pollination. The immature embryos from all three lines produced type II embryogenic callus with yellow or milky white, dry, rice-like grain structure. The number of callus obtained from each line was different. Fig. 2 shows the callus of the three inbred lines after one successive generation. According to the results of repeated experiments, the induction rate of the three genotypes was H8185 > H8183 > H8184. The recovery rate of H8185 was the highest at approximately 93.2%. The recovery rates of H8183 and H8184 were very similar, 82.7% and 80.1%, respectively (Tab. 2).

After infection, the callus were cocultured either directly on a solid medium or on a solid medium with a layer of sterile filter paper. As observed in Fig. 3, the callus that were cocultured directly on a solid medium showed no plaque after 1 day, obvious plaque after 2 days, and a large area of plaque around the callus after 3 days. However, in the coculture method using a layer of sterile filter paper on the solid medium, no plaque appeared within 1–3 days. This indicates that coculture with filter paper-containing media can effectively reduce the growth of bacteria in the callus. The lower the Agrobacterium concentration, the more effectively the growth of bacteria carried by the explants can be reduced. Therefore, the method of using a solid medium with a layer of sterile filter paper for coculture has better antibacterial effects than that using a solid medium directly.

The cocultured callus were transferred to the first screening medium containing 0, 2, 4, or 6 mg/L glufosinate-ammonium for preliminary culture screening. The selection medium without glufosinate-ammonium was used as a control, and the growth status was evaluated after 25 days. The selection medium containing 2 mg/L glufosinate-ammonium resulted in the highest survival rates of callus, reaching 82.4%, and adventitious bud differentiation (Fig. 4). The selection medium containing 4 mg/L glufosinate-ammonium resulted in significantly decreased survival rates of callus, with less-differentiated adventitious buds. On the selection medium containing 6 mg/L glufosinate-ammonium, most of the callus were albino and eventually died. The growth was extremely weak, with only few differentiated adventitious buds.

The callus selected on 2, 4, and 6 mg/L glufosinate-ammonium–supplemented medium were transferred to the corresponding secondary selection medium, and the survival and adventitious bud induction rates were evaluated. When the concentration of glufosinate-ammonium was 0 and 2 mg/L, the survival rate of the callus was >70%, and the explants grew well. However, with the increase in glufosinate-ammonium concentration, the survival rate of callus and the induction rate of adventitious buds gradually decreased (Fig. 5). When the concentration of glufosinate-ammonium was 4 mg/L, some explants could not differentiate into callus, but the adventitious bud induction rate could still reach 41.8%. When the concentration of glufosinate-ammonium was 6 mg/L, most of the callus appeared yellow, and the adventitious bud induction rate was <10%. These results indicated that 6 mg/L glufosinate-ammonium had an inhibitory effect on maize callus and inhibited adventitious bud induction. Therefore, in the screening process, the first screening medium containing 2 mg/L glufosinate-ammonium was used, followed by the second screening medium containing 4 mg/L glufosinate-ammonium, which could ensure the normal germination of transformed callus and reduce the generation of escaped buds.

To further study the relationship between the factors affecting the differentiation rate in maize callus and the relative intensity of the influence of each factor on the differentiation rate of maize callus, the L₉(3⁴) orthogonal test was used to design different differentiation combinations. Results showed that the differentiation rate of callus changed depending on the changes in the three factors (Tab. 3). The value of range “R” intuitively reflects the degree of influence of each factor on the differentiation rate of maize callus. The greater the R value, the greater the degree of influence; accordingly, the smaller the R value, the smaller the effect. According to the R value, the influence of various factors on the differentiation rate of maize callus in this study was as follows: IBA concentration (A) > 6-BA concentration (C) > Vc concentration (D).

According to the average callus differentiation rate at each level of each factor, the optimal level of each factor was determined as A3 (1 mg mL⁻¹ IBA), B3C1 (1.6 mg mL⁻¹ 6-BA), and D3 (1.5 mg mL⁻¹ Vc). The results of variance analysis (Tab. 4) revealed that there was no significant difference between the third group, while the seventh, eighth, and other remaining groups showed a significant difference. The seventh group had the most optimized combination, which was A3B3C1D3, and the corresponding concentrations were 1 mg mL⁻¹ IBA, 1.6 mg mL⁻¹ 6-BA, and 1.5 mg mL⁻¹ Vc.

GUS histochemical staining was performed on the roots of the putative transgenic plants. Partial staining results are presented in Fig. 6. The roots of the resistant plants were stained blue to varying degrees, while the roots of nontransgenic plants were not stained. A total of 23 plant roots were used for staining, of which 13 were successfully stained, and the transient expression rate of GUS reached 55.9%. This indicates that the GUS gene was integrated into maize and can be expressed stably.

Genomic DNA of putative transgenic plants was extracted, and the bar gene was detected using PCR, with nontransgenic wild-type as a control. The target band was amplified only in the resistant transgenic plants, indicating that the obtained resistant regenerated materials were transgenic plants (Fig. 7). To further test whether the target gene introduced into the maize genome was expressed at the protein level, the resistant plants were tested using Bar test-strip analysis (Fig. 8), and finally five transgenic plants were obtained.