Ecient and Genotype Independent Maize pollen Transfection Mediated by Magnetic Nanoparticles

Background: Biotechnological engineering of maize to introduce favorable new traits relies on delivery of foreign DNA into its cells. Current gene delivery methods for maize is limited to specic genotypes, and depend on time-consuming and labor-intensive processes of tissue culture. Results: Here, we report a new method to transfect maize that is culture-free and genotype independent. Enhanced green uorescent protein gene (EGFP) or bialaphos resistance gene (Bar) bound with magnetic nanoparticles (MNPs) was delivered into maize pollens, and female orets of ve maize varieties were pollinated. Green uorescence was detected in 92% transfected pollens and 70% immature embryos. EGFP gene detected by PCR ranged from 29 to 68% in T1 generation of these ve transfected varieties, and 7-16% of the T1 seedlings showed immunologically active EGFP protein. Moreover, 1.41% of the Bar transfected T1 plants were glufosinate resistant, and heritable Bar gene was integrated into the maize genome effectively (veried by Southern blot), expressed normally and inherited stably in their progenies. Conclusion: These results demonstrate that exogenous DNA could be delivered into maize eciently and expressed normally through our genotype-independent pollen transfection system, providing a reliable, fast and large-scale gene delivery choice for most elite maize varieties recalcitrant to tissue culture.


Background
Maize (Zea mays L.) is a major crop cultivated globally to meet the increasing food, feed, and fuel demand. Although traditional breeding has signi cantly improved maize yield and quality during the past two hundred years, it still faces severe challenges, from various biotic and abiotic stresses [1]. To conquer these challenges, genetically-modi ed maize cultivars have been developed to introduce desired traits, such as insect and herbicide resistance, drought and cold tolerance, and increased nutritional quality [2]. E cient genetic transformation is crucial for the development of genetic modi ed maize. Since the rst report of successful maize transformation based on protoplast electroporation [3], several DNA delivery methods have been developed to generate genetic modi ed maize plants, including particle bombardment, Agrobacterium-mediation, PEG-mediation, liposome-meditation, silicon carbide-mediation, microinjection [4,5]. These methods rely mainly on tissue culture system, which is high-cost, laborintensive, time-consuming, and genotype-dependent especially. There are only few maize varieties such as Hi-II, B104 [6] and A188 [7] with relatively high gene transformation e ciencies. And most elite maize inbred lines have technical hurdle, thus limiting their commercial applications especially for the development of advanced breeding techniques like precise genome editing in elite crop germplasm directly. Although, transcription factors Baby boom (Bbm) and Wuschel2 (Wus2) can increase transformation e ciency in several maize varieties [8], the strenuous tissue culture process to introduce Bbm and Wus2 into desired recipient material is still inevitable. Therefore, it is very important to establish an e cient gene delivery system capable of direct and precise molecular improvement in all maize varieties.
With the development of nanobiotechnology, nanoparticles are being exploited as DNA carriers for gene delivery [2,9]. Exogenous DNA carried by nanoparticles were delivered into rice, Leucaena leucocephala [10], mustard [11], tobacco, maize [12], etc., based on the genotype dependent tissue culture procedure. Recently, a novel tissue culture independent gene delivery method using ferroferric oxide (Fe 3 O 4 ) magnetic nanoparticles (MNPs)-mediated pollen transfection was developed. Transgenic cotton, pumpkin and pepper plants were generated e ciently [13]. To our knowledge, pollen transfection application in various maize genotypes growing in the eld through MNPs has not been reported. In this study, we established a maize pollen transfection system applicable for large-scale, fast and e cient transfection in the eld, independent of maize genotype.

Results
MNP-DNA complexes were formed and delivered into maize pollens e ciently Electric negative plasmid DNA was bound by positively charged MNPs to form MNP-DNA complexes.
According to the results of gel retardation assay (Fig. 1A), DNA was completely coated on MNPs at the 4:1 and 2:1 mass ratio (DNA/MNP), the amount of free DNA decreased with the decline of DNA/MNP ratios (DNA/MNP ≥ 10:1). For DNA protection analysis (Fig. 1B), MNPs thoroughly bounded plasmid DNA and protected DNA against endonuclease digestion at the 10:1, 4:1 and 2:1 mass ratio (DNA/MNP), uncoated DNA were digested at the DNA/MNP mass ratio more than 10:1. These results indicated that the optimum ratio between plasmid DNA and MNPs was 4:1 (DNA/MNP). And the complex size of DNA/MNP = 4:1 measured by dynamic light scattering (DLS) was 212.4 nm (Fig. 1C).
Meanwhile, we con rmed the aperture structure of maize pollen under scanning electron microscopy (SEM). Maize pollen possessed only one aperture with a diameter of about 6 µm on its relatively smooth surface ( Fig. 2A), from where the MNP-DNA complexes (diameter 0.2 µm) were able to permeate into the pollen. Then we tracked MNPs spatially via transmission electron microscopy (TEM) to investigate whether magnetofection can transfer MNP-DNA complexes into maize pollen. As shown in Fig. 2B, plenty of MNP-DNA complexes were shown on the internal side of the transfected pollen wall, but none MNP-DNA complex was appeared within the untransfected pollen (Fig. 2C). These results were similar to the cotton pollen magnetofection practice, proving that MNP-DNA complexes could be transferred into pollen through aperture by magnetofection [13].
EGFP reporter was successfully expressed in transfected Jing92 pollens and plants Pollens from maize variety, Jing92, were transfected with MNP, DNA and MNP-DNA (carrying the p35S::EGFP expression cassette, Figure S1), respectively. After cultured in vitro for 24 h at 25℃, green uorescence was detected in 92% (491/536) of MNP-DNA transfected pollens (Fig. 3A), but none of MNP ( Fig. 3B) or DNA (Fig. 3C) transfected pollens, indicating that MNP was necessary and e cient to deliver DNA into maize pollen for transient expression. Furthermore MNP-DNA transfected pollens were applied to female orets of Jing92, and 70% (21/30) immature embryos (2 day after pollination, 2 DAP) showed green uorescence (Fig. 3D). Green uorescence was also detected in both leaves (Fig. 3E) and roots ( Fig. 3F) of 18% (14/79) Jing92 T1 seedlings (at the three-leaf stage). These results demonstrate that, functional EGFP protein was successfully expressed in EGFP-transfected maize plants and the transgene was recovered in seeds and their T1 seedlings.
Then, the EGFP gene in Jing92 T1 seedlings at the three-leaf stage was analyzed. The gene encoding EGFP was detected in 46% (36/79) of T1 seedlings using PCR (Fig. 4A). RT-PCR and Western blot were used to verify whether the gene was transcribed and translated. RT-PCR results showed that the EGFP gene was transcribed normally in 32% (25/79) seedlings ( Fig. 4B, with ZmActin1 as the reference gene), and EGFP protein was detected in only 16% (13/79) seedlings ( Fig. 4C, with ZmACTIN1 as the reference protein). These results demonstrated that the foreign EGFP gene was delivered into maize e ciently through pollen transfection and expressed successfully.

EGFP reporter was e ciently delivered into different maize varieties
In order to further test the e ciency of pollen transfection in different maize varieties, EGFP ( Fig. S1) was delivered into ve maize varieties 178, B73, HZ178, Jing92 and Zheng58. At the three-leaf stage, T1 seedlings from different transfected varieties were examined to ensure timely detection of transient and stable products of target genes. As shown in Table 1, the e ciency of target gene delivery calculated by PCR positive rate was quite high, ranging from 29-68%. The transfection e ciency analyzed by RNA recovery dropped to 18-32% possibly due to exogenous DNA degradation by the host nuclease and the incomplete introduction of expression cassette. The transfection e ciency scored by recovery of protein based on immunostaining reached 7-16%. Detailed data were shown in Fig. S3-S6. These results proved that target genes were delivered e ciently into different maize varieties through pollen transfection and expressed functional products.

Bar selective marker was heritable in transfected maize progenies
For the convenience of selecting stable integrated progenies, the selective marker gene Bar (Fig. S7) was transferred into maize variety Zheng58. After glufosinate screening, 1.41% (5/355) of T1 seedlings were survived (Fig. 5A). These 5 glufosinate resistant plants showed the positive BAR band (test line) during the quick strip test (Fig. 5B), indicating that Bar gene was delivered into maize successfully through pollen transfection and expressed normally. Furthermore, Southern blot indicated that these T1 plants had at least 2-3 integrations in their genomes (Fig. 5C), and their T2 progenies showed genetic segregation (Fig. 5D). The above results demonstrated that, through our maize pollen transfection, exogenous gene was integrated into the maize genome effectively, expressed normally and inherited stably in their progenies.

Discussion
Maize is an ideal plant for pollen transfection, as it has large panicles, concentrated blossom and abundant pollen. It is possible to collect su cient maize pollen and carry out large-scale transfection in the eld. We expanded the pollen treatment system capable of handling 5 g maize pollen in each transfection. As maize pollen grains were mainly 60-100 µm in diameter ( Fig. 2A, 3A-C), the 150 µm aperture sifter was used to purify and separate the pollen grains, and the 25 µm aperture nylon fabric was applied to collect maize pollen and lter liquid, thus transfected pollen grains were mainly reserved. After transfection, maize pollen was mixed with corn starch to accelerate the dry process. Through these optimizations, the MNP-mediated maize pollen transfection system became e cient. Maize varieties that are recalcitrant to tissue culture were successfully transfected with EGFP or Bar. Transient EGFP signal can be observed in 92% transfected pollens and 70% immature embryos. The EGFP gene was detected in part or whole in as high as 68% T1 seedlings germinated from transfected seeds, and the functional protein was detected in around 11% of T1 seedlings (Table 1). Moreover, 1.41% of the Bar transfected T1 plants were glufosinate resistant, and heritable Bar gene was integrated into the maize genome effectively (veri ed by Southern blot), expressed normally and inherited stably in their progenies. Taken together, the maize pollen transfection system is genotype independent and e cient, we envision this system will bene t gene delivery for all maize varieties growing in eld.
The key limiting factor of the maize pollen transfection system is the relative seed set, which is directly Using the same or other nanomaterials as DNA carriers, several inheritable transgenic plants were generated [11,13], but most transfections are transient [12,14,15]. The nanomaterial-based plant transient transformation methods are also bene cial for plant genome editing technology where gene expression without transgene integration is desired [16]. Similar technology has been developed and proven effective, termed as DNA-free genome editing, in which protein, RNA, or ribonucleoprotein are directly delivered, from a regulatory perspective, to eliminate all risk of transgene integration [17][18][19]. Our maize pollen transfection system, in which e cient and genotype independent introduction and transient expression could be achieved, providing a reliable, fast and large-scale gene delivery choice for most elite maize varieties recalcitrant to tissue culture, especially for the development of the advanced breeding techniques like precise genome editing in elite crop germplasm directly [20].

Conclusions
The results of this practical-based study clearly demonstrated that exogenous DNA could be delivered into maize e ciently and expressed normally through our genotype-independent pollen transfection system, providing a reliable, fast and large-scale gene delivery choice for most elite maize varieties recalcitrant to tissue culture, especially for the development of the advanced breeding techniques like precise genome editing in elite crop germplasm directly.

Materials
Maize varieties (major inbred lines, tissue culture recalcitrant): 178, B73, HZ178, Jing92 and Zheng58, were kindly provided by Corn Research Center, Beijing Academy of Agriculture and Forestry Sciences. Two plasmids, pYBA1132 (Fig. S1) and pYBA1132-Bar (Fig. S7), were constructed and preserved by our research center. The pYBA1132 plasmid, whose NCBI accession number was KF876796, harboring EGFP gene [21] under the control of cauli ower mosaic virus (CaMV) 35S promoter and CaMV 35S terminator. While pYBA1132-Bar plasmid carried the Bar gene [22]   178, B73, HZ178, Jing92 and Zheng58, Fig. 6A, 6B) to prepare the pollen and MNP-DNA suspension. Then, the suspension was placed on top of the MagnetoFACTOR-96 plate and kept still at room temperature for 0.5 h to transfect pollens (Fig. 6C). After transfection, the supernatant was carefully removed and the transfected pollens were spread on a 20 cm × 20 cm nylon fabric (25 µm aperture, which could lter liquid while reserve maize pollens) and the fabric was folded in half, extra liquid was absorbed completely by 3 layers of lter paper (18 cm in diameter, Fig. 6D), then the wet sticky pollens were dried with 3 g corn starch (Fig. 1E), and pollinated arti cially to 20 ears of the same maize variety (Fig. 6F) to generate transfected seeds (Fig. 6G).

Electron microscopy of maize pollen
In order to con rm their aperture structures, fresh maize pollen grains were collected and spread onto the silicon slice surface. After gold spray, pollen grains were observed under the scanning electron microscope (S3400N, Hitachi Co., Ltd., Japan) at 5.0 kV. On the other hand, we tracked MNP-DNA complexes spatially to investigate whether magnetofection can transfer MNPs into maize pollen. Transfected and untransfected pollen grains were xed and cut to prepare for ultrathin sections. The sections were mounted on copper grids and checked by transmission electron microscopy (JEM-1400, JEOL Ltd., Japan) at 80 kV.
Observation of green uorescence in transfected maize pollens and plants In order to observe the expression of EGFP reporter (in plasmid pYBA1132, Fig. S1) in the transfected maize pollens and plants, the following materials from maize variety, Jing92, with or without transfection were collected for green uorescence observation. Pollens transfected with MNP, DNA and MNP-DNA (cultured in dark for 24 h at 25℃, on pollen medium containing 15 g/L agar, brushed and disperse in the pollen medium before observation), immature embryos (2 day after pollination, 2 DAP), roots and leaves of T1 seedlings (at three-leaf stage) germinated from EGFP transfected seeds were observed for GFP signal under Confocal laser scanning microscopy (A1+, Nikon, Japan), using 488 nm wavelength light for excitation.

Molecular analysis of EGFP reporter in T1 transfected seedlings
Molecular analysis was carried out at the three-leaf stage to ensure timely detection of EGFP products. Genomic DNA and total RNA in leaves from maize seedlings were extracted via CTAB (Hexadecyl trimethyl ammonium Bromide) and TRIzol method, respectively. Gene speci c primer EGFP-F/R (

Availability of data and materials
All data generated or analyzed during this study are included in this published article and its supplementary information les.
Competing interests Z-PW, Z-BZ and Z-YW have applied for a patent in China (patent number 201910623296.5). The authors declare that they have no other con ict of interest.