Shuo Zhou, Cuimian Zhang,He Zhao,Mengyu Lyu, Fushuang Dong,Yongwei Liu,Fan Yang, Haibo Wang, Jianfang Chai

Hebei Province Plant Genetic Engineering Center, Institute of Genetics and Physiology, Hebei Academy of Agriculture and Forestry Sciences,Shijiazhuang 050051,Hebei,China

Keywords:

ABSTRACT Wheat-rye T1BL·1RS translocation lines are widely used, especially in China, but their processing quality is generally poor.An interfering expression vector targeting the ωsecalin genes was constructed with the 1Bx7 seed-specific promoter.Biolistic-mediated genetic transformation of the wheat cultivar KN199 carrying the T1BL·1RS translocation generated 10 transgenic lines.Two representative transgenic lines, 8-2 and 13-7, were selected for analysis.Compared with the control, the two transformants showed an up to 4.5-fold decrease in total ω-secalins and various levels of decrease in ω-gliadins,γ-gliadins,and low-molecular-weight glutenins.A decrease in high molecular weight(HMW)glutenin 1Bx7 was detected only in 8-2,owing possibly to promoter methylation.Increased levels of α-gliadins were observed in both transformants, but increased levels of HMW glutenins were observed only in 13-7.Line 13-7 showed increases in gluten index, Zeleny sedimentation value, stabilization time, and maximum resistance.Its bread volume was 849.6 mL, an 11.9% increase over that of the control.Line 8-2 showed decreases in these parameters, but its total cake-making quality score was 88, an 17.3% increase over that of the control.The study demonstrates that the same RNAi construct may produce different effects on wheat processing quality and highlights the influence of the vector promoter in RNA interference.

1.Introduction

Wheat(Triticum aestivum L.)cultivars with the T1BL·1RS translocation have been widely used owing to their disease resistance,yield potential, and wide adaptability [1-6].Rabinovich [7]listed around 300 wheat cultivars carrying chromosome arm 1RS of rye(Secale cereale L.).Approximately 70%of wheat cultivars released in southwestern China and 55%in northern China from the 1980s until the 2000s carry this chromosome arm[8,9].In recent years,more than 60% of wheat cultivars and breeding lines in Henan province, the largest wheat-growing region in China[10], have been reported to be T1BL·1RS translocation lines.However,most wheat T1BL·1RS translocations are associated with defects in end-use quality, including poor mixing tolerance, dough stickiness,low loaf volume,and poor noodle processing[11-13].The low quality of cultivars with T1BL·1RS translocations is considered to be caused partially by the presence of ω-secalins encoded by genes on chromosome arm 1RS.ω-secalins are a family of small monomeric proteins related to wheat ω-gliadin[11,14,15].

Several attempts have been made to overcome the negative effects of the ω-secalin genes on the processing quality of wheat with the 1B/1R translocation.Chromosome engineering by induced homologous recombination has been exploited to remove the secalin loci[3,16,17].Although the ωsecalin gene family comprises 15 members [18], only four,with high sequence similarity, have transcription and translation activities[19],offering the possibility of silencing all the ω-secalin genes by RNA interference (RNAi).In our previous study [20], several transgenic lines with partially silenced ωsecalin genes were obtained by RNAi of the ω-secalin genes.Although the transgenic lines showed higher gluten index,Zeleny precipitation value, and stabilization time than the non-transgenic control,these differences were limited,possibly owing to the low expression level of the interfering fragment driven by the maize ubiquitin promoter in the young seeds of wheat.When Blechl et al.[21]used a seedspecific promoter from the wheat high-molecular-weight(HMW) glutenin subunit 10 gene, the ω-secalin genes were highly silenced in the resulting transgenic lines, but the expression of HMW glutenin subunit 7 was also reduced.Although the flour from the transgenic lines showed a longer development time and greater tolerance to mixing than that from the control, no differences in SDS sedimentation were observed.The reason may have been non-specific interference caused by the long ω-secalin gene-interfering fragment(481 bp)in the vector.In our previous study[20],the length of the ω-secalin gene-interfering fragment was 381 bp.The purpose of the present study was to determine whether replacing the maize ubiquitin promoter with a wheat seedspecific promoter in our original ω-secalin gene-interfering expression vector would highly down-regulate the expression of ω-secalins while not interfering with the expression of HMW glutenin genes.

2.Materials and methods

2.1.Plant materials

Three wheat cultivars: Kenong 199 (KN199), Chinese Spring(CS),and Gaocheng 8901(GC8901),and the rice(Oryza sativa L.)cultivar Zhonghua 3 (ZH3) were used.GC8901, with an HMW glutenin subunit composition of 1/7 + 9/5 + 10, was used to clone the promoter of the HMW glutenin 1Bx7 subunit gene.CS, with a known Wx gene sequence, was used to isolate the first intron sequence of the wheat Wx gene.ZH3,grown in our lab, was used to isolate the terminator of the rice Wx gene.KN199, with the T1BL·1RS translocation, was used as the recipient cultivar in genetic transformation.

2.2.ω-Secalin RNAi vector construction and wheat transformation

In our previous study [20], we had constructed a ω-secalin RNAi vector pCAMBIA0390-Sec.In this vector, the promoter and terminator of the ω-secalin interfering gene were Ubi from maize and nos, respectively.To increase the expression efficiency of the ω-secalin interfering gene, the expression cassette of the ω-secalin interfering gene was excised from the plasmid pCAMBIA0390-Sec by digestion with HindIII and SmaI and inserted into the pAHC15 vector [22]by replacing the gus gene expression cassette to produce the plasmid pAHC15-Ubi::Sec-Nos.The ubiquitin promoter in this plasmid was then replaced by the wheat 1Bx7 promoter, which was amplified from the genomic DNA of GC8901 with the following primers:forward 5′ACGT AAGCTTTGGAGGCCAGGGAAAGACAATG-3′(the underlined sequence was added at the HindIII site) and reverse 5′-ACTC CCCGGGCTGTCAGTGAATTGATCTGTAG-3′(the underlined sequence was added at the SmaI site),resulting in the plasmid pAHC15-1Bx7::Sec-Nos.The nos terminator in this plasmid was replaced by the rice waxy gene terminator,which was cloned from the genomic DNA of the rice cultivar ZH3 with the following primers: forward 5′-TGAAGAGCTCGAGATCTACATATGGAGTGAT-3′ (the underlined sequence was added at the SacI site) and reverse 5′-TAGT GAATTCGTATCCACTCCCTCCGTCACAT-3′(the underlined sequence was added at the EcoRI site), resulting in the final plasmid pAHC15-1Bx7::Sec-Waxy.

The ω-secalin gene RNAi expression cassette was obtained from pAHC15-1Bx7::Sec-Waxy by digestion with HindIII and EcoRI and transforming into immature embryos of KN199 by gene gun cotransformation with plasmid pAHC20,containing the selectable marker gene bar under the control of the maize Ubi promoter,at a 1:1 molar ratio using the method described previously[23].

2.3.Polyacrylamide gel electrophoresis of seed gluten proteins

Mature wheat kernels were crushed into fine powder and used to extract the endosperm gluten proteins.Gliadins were extracted and separated by acid polyacrylamide gel electrophoresis (A-PAGE) following Zhang et al.[24].The glutenin proteins were extracted and separated by SDS-PAGE using the previously described method [25].Densitometric analysis of A-PAGE and SDS-PAGE gels was performed with the image analysis software Image J 1.8.0[20].

2.4.Molecular identification of transformants

To detect the transgene, genomic DNA was isolated from young leaves of T0transformants and their progeny using the Plant Genomic DNA kit (Tiangen, Beijing, China).The presence of the secalin RNAi construct in transformants was detected by PCR of the terminator region, using the following primers: forward WxT/F2: 5′-GTTGTACTCTTCTGGGTGTGC-3′and reverse WxT/R2: 5′-GTCAAACTACTGCTCCTTCAAAC-3′.The PCR conditions were as follows:94°C for 5 min,followed by 31 cycles at 94°C for 30 s,60°C for 30 s,72°C for 30 s,with a final extension at 72 °C for 10 min, and the products were separated on 1% agarose gel in Tris-acetic acid-EDTA buffer.The expected amplification fragment was 423 bp.For Southern blotting, 20 μg of genomic DNA was digested with EcoR I,Sac I, and Sma I.The digested DNA was separated on a 1%agarose gel and transferred to a nylon membrane,which was hybridized with probes from DIG-labeled PCR products amplified from the terminator region of the vector with primers WxT/F2 and WxT/R2.

2.5.Methylation analysis of endogenous HMW glutenin subunit 1Bx7 gene promoter

Methylation analysis of the 1Bx7 promoter region in transgenic lines and the recipient cultivar KN199 was conducted using the sodium bisulfite conversion method [26].The genomic DNA isolated from seeds collected 15 days after flowering was modified using the EZ DNA Methylation-Gold Kit (cat#5005,ZYMO Research Co.,Alameda,CA,USA)following the manufacturer’s instructions.The 1Bx7 promoter sequence was amplified by PCR with the primers 5′-TAGTTTTTTTTGTGTTGGTAAATTG-3′ and 5′-TAATTATCCAAAAACCTCACCTTC-3′.The size of the expected amplification fragment was 300 bp, which includes a partial promoter sequence and partial coding region sequence of the wheat HMW 1Bx7 gene (?213 to +87 bp with respect to the translational start site ATG; GenBank accession DQ119142).The downstream PCR primer was selected from the coding region of the 1Bx7 gene in order to amplify the promoter sequence of the endogenous 1Bx7 gene rather than an exogenous 1Bx7 promoter sequence introduced by the transgene.The resulting PCR products were inserted into the pUC18-T vector(Beijing Tianmo Sci & Tech Development Co., Ltd., Beijing, China) and then sequenced (Sangon Biotech (Shanghai) Co., Ltd., Shanghai,China).Ten clones from each bisulfite treatment were sequenced.Sequences were analyzed using the online cytosine methylation analysis software CyMATE(http://www.cymate.org/)to quantify and analyze the CGN(N=any base),CHG(H=A or T),and CHH(H=A,T,or C)methylation patterns.

2.6.Protein fractionation and quantification

Fractionation and quantification of seed storage proteins in transgenic lines and the recipient cultivar KN199 followed Suchy et al.[27]with some modification.Briefly,duplicates of flour samples (10 mg at 14% moisture basis) were weighed into 2-mL centrifuge tubes.The first set of samples were extracted with 1.8 mL of 50% isopropanol in a temperaturecontrolled mixer for 30 min at 150 r min?1and 25 °C.After centrifugation at 13,500 ×g for 2 min, the absorbance of the supernatant was determined at 280 nm with a NANO DROP 2000 spectrophotometer (Thermo Scientific, USA).This fraction was called 50% isopropanol soluble protein (50PS).The second set of samples was extracted exactly as the first set but using a different solvent system of 50% isopropanol and 2%dithiothreitol(DTT)solution.This fraction was called the total soluble protein (TSP).The amount of insoluble glutenins in the flour (50PI) was calculated as the difference between TSP and 50PS.To determine the amount of albumin and globulin,another set of flour samples with the same specifications was prepared.This set of samples was first extracted exactly as the first set but with 0.5 mol L?1sodium chloride solution,and the precipitate after centrifugation was extracted exactly as the second set.This fraction was called salt-insoluble protein(SIP)and could also be called gluten protein(GP).The amount of salt-soluble protein (SSP) was expressed as the difference between TSP and SIP,which was composed mainly of albumin and globulin.

The proportion of 50PS (%50PS) protein in flour was expressed as the ratio of the absorbance of the 50PS fraction to the absorbance of TSP at 280 nm.The same was true for %50PI,%SIP,and%SSP.

2.7.Grain yield comparison and processing quality determination

T4-generation transgenic lines and their untransformed control KN199 were planted in a randomized complete block design with three replicates.The size of the plot was 1.2 m×4.0 m,containing four rows with row spacing of 30 cm and 240 seeds per row.After maturity, grain yield of each plot was measured.The seed samples from each plot were tempered and milled according to AACC Approved Method 26-21.02 [28].Flour samples were maintained at 23°C for one week.Moisture content and protein content were determined according to AACC Approved Methods 44-15.02 and 46-13.01, respectively.Flour wet gluten content,farinograph, extensograph, and Zeleny sedimentation were determined according to AACC Approved Methods 38-12.02, 54-21.02, 54-10.01, and 56-61.02, respectively.Bread-making and cake-making qualities were tested with 100 g of flour following GB/T 14611-2008[29]and GB/T 24303-2009[30],respectively.

2.8.Data analysis

One-way analysis of variance was performed using DPS software version 3.01.The significance was determined by Duncan’s test.The results with P-values less than 0.05 were considered significant.

3.Results

3.1.Molecular identification of ω-secalin RNAi transgenic lines

Transformation of wheat cultivar KN199 carrying theT1BL·1RS translocationwith the ω-secalin RNAi expression cassette and pAHC20 plasmid yielded 22 bialaphos-resistant T0plants.Thirteen were confirmed by PCR to be positive for the ω-secalin RNAi expression cassette.Fig.1 shows the PCR detection results for some T0generation plants.PCR-positive transgenic plants with normal growth characteristics were self-pollinated for two to three generations to yield 10 homozygous transgenic lines.Southern blotting of two homozygous transgenic lines,13-7 and 8-2,showed that they were two independent transformants,each with a transgenic copy number of one(Fig.2).

3.2.Functional analysis of transgenic lines

A-PAGE was used to determine the effect of the ω-secalin RNAi.Among the 10 homozygous transgenic lines, six showed lower levels of ω-secalins than the recipient KN199,indicating that the RNAi was functional.Based on the expression differences among HMW glutenin subunits as detected by SDS-PAGE, the six transgenic lines with lower levels of ω-secalins could be divided into two groups of three lines each.Transgenic lines 8-2 and 13-7 as representative lines of the two groups were selected for further analysis.A-PAGE and SDS-PAGE of 8-2,13-7,and the control are shown in Fig.3.These two lines showed similar gliadin expression profiles.The total amount of ω-secalins in the transgenic lines was reduced to 21.2%-22.4%of that in the control(Fig.3A).In addition to ω-secalins, the expression of two ωgliadins and one γ-gliadin,indicated by arrows in Fig.3A,was also reduced,while that of all the α-gliadins was increased.The lowmolecular-weight(LMW)glutenin expression profiles of the two transgenic lines were also similar.Compared with the control,two LMW glutenins in the transgenic lines were clearly downregulated (arrows in Fig.3B).The HMW glutenin expression profiles of the two transgenic lines were markedly different.In line 8-2,1Bx7 was almost undetectable and the other four HMWglutenin subunits did not show decreases,whereas in line 13-7,1Bx7 and the other four HMW glutenin subunits showed clear increases(Fig.3B).

Fig.2–Southern blotting analysis of transgenic lines 8-2 and 13-7.M, DNA marker;1, EcoR V;2,Sma I;3, Sac I.

Fig.1– Electropherogram of the PCR products of T0 bialaphos-resistant plants.M, DNA marker;A,without DNA;B,plasmid pAHC15-1Bx7,Sec HP;C,recipient cultivar KN199;1–12, T0 transgenic plants;Arrow,target band.

3.3.Changes in methylation of the endogenous 1Bx7 gene promoter in transgenic lines

The changes in methylation of the endogenous 1Bx7 gene promoter region (?1 to ?188 bp upstream of the translation initiation codon)are shown in Fig.4 and Table 1.In the recipient cultivar KN199 and the transgenic line 13-7,the endogenous 1Bx7 gene was expressed normally, and almost no methylation occurred in the promoter region, with a total methylation rate of less than 1%.In contrast, in transgenic line 8-2 in which the expression of the endogenous 1Bx7 gene was inhibited,methylation was detected at a rate of ＞10%.Among the three methylation types,the proportion of class 1 CGN(N=any base)(66.7%)was higher than that of class 2 CHG(H=A or T)(14.3%)or class 3 CHH(H=A,T,or C)(4.3%).The highly methylated bases were not evenly distributed in the sequenced region, and they were all located upstream of the TATA box (?85 to ?91 bp upstream of the translation start site),suggesting that methylation directly inhibited the transcription of 1Bx7 in line 8-2.

Fig.3–A-PAGE of gliadins(A)and SDS-PAGE of glutenins(B).CK,KN199;1,transgenic line 8-2;2,transgenic line 13-7;Arrows in B, reduced low-molecular-weight glutenins in the two transgenic lines compared to CK.

3.4.Changes in contents of seed storage protein fractions in transgenic lines

Compared with the control KN199,transgenic line 8-2 showed a significant increase of the gliadin proportion and a significant reduction of the glutenin proportion in total extractable protein, whereas for transgenic line 13-7 the proportion of gliadin was significantly reduced and the proportion of glutenin was significantly increased (Table 2), in agreement with the decreased expression of HMW glutenin subunit 7 in 8-2 and the increased expression of HMW glutenin subunits in 13-7.With respect to the proportion of salt-soluble and saltinsoluble protein in total soluble protein, the two transgenic lines performed similarly.The proportion of salt-soluble protein was significantly increased and that of salt-insoluble protein significantly decreased, indicating that RNA interference with the ω-secalin genes reduced the proportion of saltinsoluble gluten protein and correspondingly increased the proportion of salt-soluble albumin and globulin in seed storage protein.

3.5.Quality analysis of transgenic lines

There was no significant difference in seed protein content between the two transgenic lines 8-2 and 13-7 and the recipient cultivar KN199.Compared to the control,the wet gluten content in the two transgenic lines decreased significantly(P＜0.05).For other quality parameters, the two lines presented contrasting results (Table 3).The gluten index,Zeleny sedimentation value,stabilization time, and maximum resistance of 13-7 were significantly higher than those of the control, but the stretched area and extensibility were not significantly different from those of the control(Table 3).The bread volume of 13-7 was 849.6 mL,significantly higher than that of the control.But the gluten index,Zeleny sedimentation value, stabilization time, stretched area,extensibility, and maximum resistance of 8-2 were significantly decreased compared to those of the control, whereas the cakemaking quality was significantly increased,with a score of 88,13 higher than that of the control.

Fig.4–Comparison of methylation of endogenous 1Bx7 gene promoters between the two transgenic lines 8-2 and 13-7 and the control KN199.The numbers ?5 to ?184 at the top indicate the base location upstream of the translation start codon ATG;the numbers ?1 to ?10 at the left of the figure indicate the 10 clones.

3.6.Grain yield of transgenic lines

The grain yields of the two representative transgenic lines and the control were evaluated at the plot level.Although the yields of the two lines were lower than that of the control,the difference was not significant(Fig.5).

4.Discussion

4.1.Choice of promoter and interference fragment has an important influence on the interference effect

Table 1–Number of methylation sites within the endogenous 1Bx7 gene promoter region in the transgenic lines 8-2 and 13-7 and the control KN199 for class 1(CGN,where N=any base),class 2(CHG,H=A or T),and class 3(CHH,H=A,T,or C)sites.

In our previous RNAi experiment [20], the maize ubiquitin promoter was used to drive the expression of ω-secalin interference gene, and the expression of ω-secalins was decreased by only approximately 50%with negligible changes in the other gluten proteins.In the present study, the ωsecalins were reduced by approximately 78%, furthermore,the levels of some ω-gliadins, γ-gliadin, and LMW glutenin subunits were also decreased to different extents in the two selected transgenic lines, indicating that the wheat 1Bx7 promoter has a greater ability than the maize ubiquitin promoter to drive gene expression in developing wheat seeds.The ω-gliadins have been reported [31]to have an adverse effect on gluten strength.Blechl et al.[21]used the wheat seed-specific 1Dy10 promoter to drive the expression of a ω-secalin interference fragment and obtained interference effects similar to those in the present study.The difference was that the expression of 1Bx7 was consistently decreased in the two transgenic lines of Blechl et al.[21]; whereas in our study, the expression of 1Bx7 was decreased in line 8-2 but increased in line 13-7,and the decreased expression of 1Bx7 in 8-2 could be caused by methylation of its promoter (Fig.4),rather than RNAi.The 1Bx7 promoter used in our RNAi expression vector may be the cause of the methylation of the endogenous 1Bx7 promoter in the transgenic line 8-2[32].The difference in the interfering effect of RNAi between the two studies may have resulted from the difference in the ωsecalin-interfering fragments used.The RNAi construct used by Blechl et al.[21]contained a 481-bp coding region of the ωsecalin gene,whereas the RNAi vector that we used had a 381-bp trigger from the coding region of the ω-secalin gene.We speculate that the longer trigger sequence used in the previous study [21]contains a short sequence similar to the 1Bx7 gene and thus interfering with its expression.

Table 2–Amounts and proportions of protein fractions in transgenic lines 8-2 and 13-7 and their control KN199 determined by UV spectrophotometry at 280 nm.

Decreases in the expression of certain genes in the RNAi assay are often accompanied by a compensatory increase in the expression of other genes.Gil-Humanes et al.[33]reported that the reduction in γ-gliadins in their transgenic lines appeared not to have a direct effect on mixing and breadmaking properties; the compensatory increase in total glutenins increased the dough strength.In the transgenic line 13-7, the levels of all five HMW glutenin subunits were increased relative to those in the control, increasing dough strength and bread-making quality[34,35].The SDS sedimentation values of the two transformants analyzed in the previous study [21]did not increase, possibly because of a decrease in 1Bx7 expression.In the transgenic line 8-2, the effects of silencing of 1Bx7 outweighed the beneficial effects of RNAi targeting the ω-secalin genes and thus reduced gluten strength.

4.2.High-protein, weak-gluten wheat can also yield highquality cake

The aim of this study was to improve the bread processing quality of wheat 1B/1R translocation lines, and this goal was reached by production of a line, 13-7, whose bread volume was greater than that of a control.It was not expected that the 1Bx7 gene would be silenced in the transgenic line 8-2,reducing its dough strength as a result.Transgenic line 8-2 could make high-quality cake.The use of high-protein(＞14%)wheat to make high-quality cake has not been previously reported.

Generally, weak-gluten wheat has low protein content,gluten content, and gluten strength.In China, weak-gluten wheat cultivars are required to have ＜12%protein content(dry basis), ＜24% wet gluten content (14% moisture basis), ＜55%water absorption rate, and ＜3 min stabilization time.At present, there are few weak-gluten wheat cultivars in China,and they are grown mainly in the upper, middle, and lower reaches of the Yangtze River as low-protein cultivars.The present study provides a new method to develop weak-gluten wheat cultivars for high-protein wheat growing areas.

Table 3–Quality-related parameters of the transgenic lines 8-2 and 13-7 and their non-transformed parent KN199.

The characteristics of 8-2 suitable for making cakes may be accounted for by the low gluten content caused by RNAi and low gluten strength due to silencing of 1Bx7.Although the seed protein contents of line 8-2 and the control remained at the same level(both ＞14.2%),the wet gluten content(20.9%)of 8-2 was significantly lower than that (32.2%) of the control.Becker et al.[36]used RNAi to decrease the level of α-gliadins,and the gluten content in the transgenic lines was also decreased, whereas the total content of globulins and albumins was increased compared to those of their nontransgenic control.In the present study, the proportions of globulins and albumins in total soluble protein in the two transgenic lines were also increased in comparison to those in the control.Silencing of the 1Bx7 gene in line 8-2 reduced the proportion of HMW glutenin in gluten, in turn reducing the glutenin proportion, gluten strength, and stabilization time.Although the protein content of the transgenic line 8-2(14.2%)was higher than 12%, its wet gluten content (20.9%) and stabilization time(2.4 min)were in line with the requirements of weak-gluten wheat.The results of line 8-2 suitable for making cakes indicate that cake-making quality is affected mainly by gluten content and gluten strength rather than protein content, given that the globulin and albumin in the seed are not involved in the formation of gluten.

Fig.5–Grain yield of two transgenic lines 8-2 and 13-7 and their non-transformed parent.No significant difference between the two transgenic lines and their non-transformed parent at P＜0.05.

In summary, using our new ω-secalin gene RNAi vector,the dough strength and bread-making quality of wheat can be improved by selection of the appropriate transformant.In evaluating the interference effect of RNAi, it is necessary to consider the size of the interference fragment and the choice of promoter in the expression vector.Silencing of the 1Bx7 gene in transgenic line 8-2 changed the RNAi effect and reduced dough strength.This high-protein transgenic line 8-2 could make high-quality cake.How the same expression vector differentially influences the methylation of an endogenous gene promoter awaits further study.

CRediT authorship contribution statement

Shuo Zhou and Cuimian Zhang performed most of the experiments; He Zhao participated in field experiments;Mengyu Lyu generated transgenic lines; Fushuang Dong and Yongwei Liu performed Southern blot and promoter methylation assay; Fan Yang measured the content of storage protein components; Haibo Wang participated in the formulation of the experimental plan; Jianfang Chai designed the experiments and assisted in the writing.

Declaration of competing interest

Authors declare that there are no conflicts of interest.

Acknowledgments

The financial support by Key Research and Development Plan of Hebei Province(20326348D) and Hebei Modern Agricultural Science and Technology Innovation Project (2019-4-08-1) are gratefully acknowledged.