Evidence of arrested silk growth in maize at high planting density using phenotypic and transcriptional analyses

ZHANG Min ,XING Li-juan ,REN Xiao-tian ,ZOU Jun-jie ,SONG Fu-peng ,WANG Lei ,XU Miao-yun

1 Biotechnology Research Institute,Chinese Academy of Agricultural Sciences,Beijing 100081,P.R.China

2 College of Bioscience and Resources of Environment,Beijing University of Agriculture,Beijing 102206,P.R.China

3 National Engineering Laboratory for Efficient Utilization of Soil and Fertilizer Resource,College of Resources and Environment,Shandong Agricultural University,Tai’an 271018,P.R.China

Abstract Increasing the planting density is an effective way to increase the yield of maize (Zea mays L.),although it can also aggravate ovary apical abortion-induced bald tips of the ears,which might,in turn,reduce the yield. While the mechanism underlying the regulation of drought-related abortion in maize is well established,high planting density-related abortion in maize remains poorly understood. Therefore,the present study was designed to investigate the mechanism underlying the ovary apical abortion response to high density. This was achieved by evaluating the effects of four different plant densities (60 000 plants ha-1 (60 k),90 k,120 k,and 150 k) on plant traits related to plant architecture,the plant ear,flowering time,and silk development in two inbred lines(Zheng58 and PH4CV) and two hybrid lines (Zhengdan958 and Xianyu335). The phenotypes of both inbred and hybrid plants were observed under different planting density treatments,and the high planting density was found to increase the phenotypic performance values of the evaluated traits. The anthesis-silking interval (ASI) was extended,and the amount of the silk extruded from husks was reduced upon increasing the planting density. Delayed silk emergence resulted in asynchronous flowering and ear bald tips. Observations of the silk cells revealed that the silk cells became smaller as planting density increased. The changes in transcript abundances in the silks involved the genes associated with expansive growth rather than carbon metabolism. These findings further our understanding of silk growth regulation under high planting density and provide a theoretical basis for further research on improving high planting density breeding in maize.

Keywords: maize (Zea mays L.),high planting density,bald tip,ASI,silk expansive growth

1.Introduction

Maize is one of the most economically important crops in the world,and maize yield plays a dominant role in global crop production. Improvements in maize genotypes and agronomic management practices have greatly promoted the maize yield. Planting density is one of the most important cultural practices that determine grain yield and the other important agronomic attributes of maize(Sangoi 2001). Increasing planting density has contributed immensely to the continually increasing yields in China and the United States over the past few decades (Mansfield 2012;Wanget al.2020). However,high planting density leads to competition for water,nutrients,and light among the crop plants,which frequently results in ovary abortion,and ultimately in low yield (Willey and Heath 1969). Serious ovary abortion resulting in a long bald tip is the greatest obstacle to high-density planting. Sugar deprivation has been proposed as a major cause of ovary abortion,based on a series of experiments that revealed that the addition of sucrose could partially reduce ovary abortion (Boyleet al.1991;Zinselmeieret al.1995a,b,c;McLaughlin and Boyer 2004). However,there is another argument that is independent of assimilation. The fate of a flower relies on its position in the racemes,i.e.,when buckwheat racemes open,the flowers at the base have a higher probability of developing normally and producing mature seeds compared to the terminal flowers (Cawoyet al.2007). The maize plant ear may be viewed as a coalesced inflorescence,which initiates spikelet pair meristems that are arranged sequentially in rings (Bonnett 1954;Kaplinsky and Freeling 2003). The basal kernels develop first,while the apical kernels develop later. Under stress,the female panicle can terminate the development of young grains at the top to preserve the basal grains and allow the completion of their growth process (Freieret al.1984;Cárcova and Otegui 2001). These cited studies on the mechanisms leading to abortion were based on water deficit conditions,but the mechanisms leading to abortion under high planting density conditions remain unknown at this point.

In the present study,two common maize inbred lines(Zheng58 and PH4CV) and two maize hybrid lines(Zhengdan958 and Xianyu335) were planted at different planting densities. Physiological phenotypical traits,such as height,stem diameter,chlorophyll content,and biomass,were found to be greatly affected in a negative manner. Increasing the planting density could significantly enhance the length of baldness and ovary abortion. As planting density increased,the anthesissilking interval (ASI) became extended,the amount of the silk extruded from the husk was reduced,and the silk cell length and size were smaller. Furthermore,the transcript abundances of the genes involved in tissue expansion and sugar metabolism were evaluated. The significance and number of differentially expressed genes were found to be greater for tissue expansion than carbon metabolism. Taken together,these findings revealed that bald tip under high planting density conditions is a result of silk that grows extremely slowly and,consequently,misses the pollination period.

2.Materials and methods

2.1.Plant materials,growth conditions,and pollination

Maize (ZeamaysL.) plants (both inbred lines Zheng58 and PH4CV,and hybrid lines Zhengdan958 and Xianyu335)were grown at Wanzhuang Agricultural Research Station of the Chinese Academy of Agricultural Sciences (CAAS),Langfang,Hebei,China,during the growing seasons of 2016 and 2017. On account of their small plant architecture,the inbred lines Zheng58 and PH4CV were planted at three planting densities of 90 000 plants ha-1(90 k),120 k,and 150 k,while the hybrid lines Zhengdan958 (Zheng58×Chang7-2) and Xianyu335(PH6WC×PH4CV) were planted at four planting densities of 60 k,90 k,120 k,and 150 k. The plant groups with the respective planting densities of 90 k and 60 k served as the control groups for the inbred and hybrid lines,respectively. Three plots,each with an area of 5×4.2 m2,were established for each planting density. Row spacing was 0.6 m for each planting density,while plant spacings were 27.8,18.5,13.9,and 11 cm for the planting densities of 60 k,90 k,120 k,and 150 k,respectively. All plants were open-pollinated,except for the emerging silks to be used for microscopic observations and RNA-seq,which were placed into white pollination bags to prevent pollination.

2.2.Measurement of physiological parameters in the field

Plant height (from ground to tassel-top) was measured using a horizontal ruler made of an aluminum alloy(NewMap,China). Stem diameter (of the second stem node from the bottom of the plant) was measured with tape. The relative chlorophyll content was determined using the Chlorophyll Meter SPAD-502plus (Konica Minolta,Japan) according to the manufacturer’s instructions. The parameters of the second leaf under the ear were measured by dividing the leaf equally into 10 portions,measuring the parameters of each portion,and calculating the average of all 10 measurement results as the final result for the whole leaf. The biomass of the above-ground parts of the plant was determined using an ordinary platform balance. All physiological parameters were measured every three days starting at the time of pollination.

Meanwhile,the ASI values of the maize plants at different planting densities were calculated. Pollination time was defined as the time when over one-third of the tassel had completely lost power. Time to silking was defined as the time when silk emerged from the husks to a length of 2-3 cm.

2.3.Ear sampling and measurements

Mature ears were harvested 40 days after pollination for bald tip length measurement. At least 60 ears were measured for each density. The ears harvested at 20 days after pollination were used for calculating abortion frequency. Bald tip length and ear total length were determined using the ImageJ Software.

2.4.Silk sampling and measurements

Silk numbers of the two hybrid lines were counted seven days after silk emergence (SE). The silk cell was measured at two days after SE in unfertilized silks,and all samples for this measurement were collected from the root of the silk near the cob and immediately placed into pre-cooled glutaraldehyde fixative. Cells within the silk were observed under an optical microscope Axio Imager(Carl Zeiss,Germany) at 20× magnification. The silk cell length and area were determined using ImageJ Software.

2.5.Pollen viability evaluation

Fresh pollen was spread evenly on a glass slide,1-2 drops of TTC staining solution were dropped onto it,and the slide was observed under a microscope at low magnification. If the pollen was stained red,it had vitality;if it was not stained,it was considered dead. If the pollen was stained partially red or with a low-intensity red color,then it was considered partially viable or as having weakened vitality,respectively.

2.6.RNA preparation,RNA sequencing,and RNAseq data analyses

Newly emerging silks (2-3 cm) of Zheng58,which were placed in a bag to prevent pollination,were sampled randomly from each planting density plot. The collected silk samples were frozen immediately in liquid nitrogen(flash freezing at less than 30 s after sampling) and stored at -80°C until their use for RNA extraction.

Total RNA was extracted from the silk of Zheng58 using CTAB reagent (OE Biotech Co.,Shanghai,China)according to the manufacturer’s instructions. The extracted RNA sample was quantified using Nanodrop Spectrophotometry (Nanodrop Technologies,Wilmington,DE,USA),and the RNA integrity was evaluated using an Agilent 2100 Bioanalyzer (Agilent Technologies,Santa Clara,CA,USA). Libraries were constructed by using the TruSeq Stranded mRNA LT Sample Prep Kit (Illumina,San Diego,CA,USA) according to the manufacturer’s instructions and subsequently sequenced on the Illumina sequencing platform by OE Biotech Co.,Ltd.(Shanghai,China). A total of three RNA-seqs were surveyed in the silk of Zheng58,and their data have been submitted to the NCBI SRA database (https://trace.ncbi.nlm.nih.gov/Traces/sra/sra.cgi) with accession numbers SAMN11294228,SAMN11294230,and SAMN11294229.All clean reads were mapped to the maize B73 RefGen_V3 genome (http://archive.maizesequence.org/index.html).

The clean reads were obtained by removing the raw reads containing ploy-N and the low-quality reads,and were subsequently mapped to the reference genome using hisat2 (Kimet al.2015). Fragments per kilobase of transcript per million mapped reads (FPKM) (Trapnellet al.2010) values were calculated using the cufflinks Software (Trapnellet al.2012). The planting density of 90 k served as the control group. The fold changes between 90 k and 120 k and between 90 k and 150 k were determined by calculating the respective ratios between their FPKM values.

3.Results

3.1.Physiological parameters of inbred and hybrid lines

In order to determine the effects of high planting density on maize growth during the flowering phase,four physiological phenotypes (namely,plant height,stem diameter,chlorophyll content,and biomass) of the inbred lines Zheng58 and PH4CV and the hybrid lines Zhengdan958 and Xianyu335 planted at different planting densities,were evaluated during the flowering phase. The results indicated that as the planting density increased,plant height gradually decreased in the two inbred lines (Fig.1-A). At the 90 k,120 k,and 150 k planting densities,the plant heights in Zheng58 were 170,167,and 164 cm,respectively,while those in PH4CV were 184,183 and 180 cm,respectively. With the exceptions of Zhengdan958 planted at 60 k density and Xianyu335 planted at 120 k and 150 k densities,the plant height in the hybrid lines increased slightly with increasing density. At the 60 k,90 k,120 k,and 150 k planting densities,the plant heights in the hybrid line Zhengdan958 were 268,264,271,and 275 cm,respectively,while in Xianyu335 they were 331,333,334,and 330 cm,respectively. Moreover,the stem diameter,chlorophyll content,and biomass decreased with increasing planting density (Fig.1-B-D). In particular,the stem diameter and biomass decreased significantly with increasing planting density. These results suggested that excessive planting density could seriously impact various physiological parameters of the plants,which is consistent with previously reported findings (Pacala and Weiner 1991;Cox 1996;Sangoi 2001;Renet al.2017).

3.2.High planting density aggravated baldness and abortion

In order to investigate the effect of planting density on bald tip length and abortion in the inbred and hybrid lines,plants from all planting densities were evaluated 30 days after pollination. The results indicated that bald tip length increased with increasing planting density in all inbred and hybrid lines. At the 90 k,120 k and 150 k planting densities,the bald tip lengths of Zheng58 were 1.04,1.08 and 2.15 cm,respectively,while those of PH4CV were 1.74,1.83 and 2.11 cm,respectively. Evidently,the bald tip length was higher in the hybrid lines compared to the inbred lines.Meanwhile,the ear bald tip length at the highest planting density was more than twice the value at the lowest planting density for all plants,except for PH4CV (Fig.2-A).All these results indicated that increasing the planting density heavily impacted the length of baldness.

Furthermore,the ratio of bald tip (bald tip length to ear total length) was calculated for all lines. At the 90 k,120 k and 150 k planting densities,the ratios of bald tip for Zheng58 were 7.03,7.83 and 17.28%,respectively,and those for PH4CV were 12.69,12.83 and 15.67%,respectively. The ratio of ear bald tip increased sharply as the planting density increased. At the 60 k,90 k,120 k,and 150 k planting densities,the ratios of bald tip for Zhengdan958 were 13.36,21.12,23.17,and 29.30%,respectively and those for the hybrid line Xianyu335 were 11.67,12.68,16.97,and 23.76% (Fig.2-B). Thus,the ratio of baldness was higher in the hybrids compared to their parent lines.

Furthermore,the range of abortion rates caused by various planting densities was explored in the two hybrid lines Zhengdan958 and Xianyu335. The results indicated that 100% abortion occurred at the 40th,35th,33rd,and 15th ovary positions in Zhengdan958 (Fig.2-C) and at the 46th,42nd,42nd,and 34th ovary positions in Xianyu335,at the 60 k,90 k,120 k,and 150 k planting densities,respectively (Fig.2-D). These results demonstrated that the abortion frequency increased with the ovary position on the ear at all planting densities in both inbred and hybrid lines. Moreover,there were wide gaps of 25 kernels in Zhengdan958 and 12 kernels in Xianyu335 between the 60 k and 150 k planting densities,indicating a severe impact of high planting density on the bald tip.The bald tip length (or ratio) results and ovary abortion results indicated that an excessive increase in the planting density had a huge negative impact on maize yield.

3.3.Increasing planting density lengthened ASI

Flowering is a key developmental switch and also a critical determinant of the adaptation of the plant to different environments (Ribautet al.1996). According to previous reports,ASI is highly correlated with grain yield and demonstrates high heritability under stress (B?nzigeret al.1999;Chapman and Edmeades 1999). Therefore,in the present study,the effects of planting density on maize pollen shedding days and ASI were determined in both inbred and hybrid lines. At the planting densities of 90 k,120 k and 150 k,the pollen shedding days for Zheng58 occurred at 65.1,64.1 and 64.3 days,respectively,while those for PH4CV occurred at 48.7,49 and 48.3 days,respectively (Fig.3-A). As the planting density increased,flowering in Zheng58 advanced slightly,while there was no significant change in the flowering time of PH4CV(Fig.3-A). At the planting densities of 90 k,120 k and 150 k,the ASI values of Zheng58 were 2.3,4.2,and 4.3 days,respectively,and the ASI values of PH4CV were 1.5,1.7 and 2.2 days,respectively (Fig.3-B). The pollen shedding day and ASI results indicated that in both Zheng58 and PH4CV,silking was delayed as the planting density increased. However,there were no evident changes in the pollen shedding day or the ASI of PH4CV with different planting densities (Fig.3-B),which indicated that its flowering period was less impacted by the planting density.

High planting density altered the pollen shedding time and the ASI in both Zhengdan958 and Xianyu335. At the 60 k,90 k,120 k,and 150 k planting densities,the pollen shedding times for Zhengdan958 were 55.8,58,57.1,and 58.5 days,respectively,while those for the other hybrid Xianyu335 were 58.8,57.6,56.2,and 55.1 days,respectively (Fig.3-C). At the same planting densities,the ASI values of Zhengdan958 were 2.3,2.5,3.1,and 3 days,respectively,while the ASI values of Xianyu335 were 2.1,2.6,3,and 3.5 days,respectively (Fig.3-D). Thus,flowering in Zhengdan958 did not accelerate as the planting density increased,while it accelerated in Xianyu335.However,the ASI values of both the hybrid lines increased with the increasing density. These results suggested that the acceleration of flowering at higher densities varies among the different varieties of the maize plant,although the ASI increases consistently with increasing density.These findings revealed that silk emergence is delayed upon increasing the planting density.

3.4.High planting density arrested silk growth

If the male tissue of the maize plant is not considered,there could be two main reasons for the occurrence of bald-tip. One possibility is that the silk of the ear tip did not emerge from the husk,resulting in no pollination.The other reason is that although the silk emerged,the kernel was aborted during development. The appearance of a fertilized kernel is very different from that of a nonfertilized kernel. Therefore,it is clear from Fig.2 that the depicted bald tip occurred due to the failure of pollination.

Additional data indicated that pollen viability was not greatly impacted by planting density (Appendix A),and the number of pollen grains was sufficient for pollination to occur. Therefore,it was inferred that pollen was not the main reason for abortion,and so the subsequent step in this study was focused on silk growth. We counted silk number per ear of the two hybrids Zhengdan958 and Xianyu335,and the results revealed that silk number per ear decreased with increasing density (Appendix B). At the 60 k,90 k,120 k,and 150 k planting densities,the silk numbers per ear for Zhengdan958 were 694,660,597,and 310,respectively,and those of Xianyu335 were 790,772,741,and 614,respectively. The silk number decreased sharply with increasing density,showing a huge gap between the 60 k and 150 k planting densities of 384 in the case of Zhengdan958 and 176 in the case of Xianyu335. Silk growth is determined by the number of silk threads that emerge from the husk. In order to explore the reasons for the huge difference observed in the silk number at different planting densities,silk cells in both inbred and hybrid lines at different planting densities were observed under a microscope at 20×magnification when the silk had emerged to ＞2-3 cm in length from the bract (Fig.4-A and B). The average silk cell length and area were calculated. At the planting densities of 90 k,120 k and 150 k,the silk cell lengths in Zheng58 were 231,214 and 128 μm,respectively,and those in PH4CV were 247,193 and 128 μm,respectively. At the planting densities of 60 k,90 k,120 k,and 150 k,the silk cell lengths in Zhengdan958 were 428,181,116,and 111 μm,respectively,and in Xianyu335 they were 624,196,158,and 88 μm,respectively. The silk areas in Zheng58 at the 90 k,120 k and 150 k planting densities were 4 979,5 052,and 2 460 μm2,respectively,and those in PH4CV were 4 932,2 708,2 274,and 1 881 μm2,respectively (Fig.4-C). The Zhengdan958 at the 60 k,90 k,120 k,and 150 k planting densities were 7 797,2 834,2 160,and 1 610 μm2,respectively,and those in Xianyu335 were 10 355,7 255,3 326,and 1 881 μm2,respectively (Fig.4-D). Therefore,the evidence indicates that both silk cell length and silk area declined as the planting density increased. This result was consistent with the previously reported results of silk cell elongation under drought stress (Fuad-Hassanet al.2008;Ouryet al.2016).

3.5.Changes in the silk transcript abundances of genes related to expansive growth

In order to explore the molecular-level changes in the silk,transcript abundances of the genes involved in tissue expansion and carbon metabolism were studied.The genes involved in tissue expansion exhibited a higher differential expression between the planting densities of 90 k and 120 k and the planting densities of 90 k and 150 k (Fig.5-A). The genes related to categories such as cellulose synthase-like,cellulose synthase,pectin synthesis,cellulose,and WAK were significantly downregulated,while a few of the genes related to pectinase,pectinesterase,and expansive growth were upregulated. Conversely,transcripts of the genes involved in carbon metabolism exhibited only small differences in expression between the 90 k and 120 k and between the 90 k and 150 k planting densities(Fig.5-B). Only two genes,namely,GRMZM2G105791andGRMZM2G366659,which are related to starch synthase and trehalose-6-phosphate synthase (TPS),exhibited small differences in their transcript levels. The significance and number of differentially expressed genes(DEGs) in tissue expansion were greater than those for the DEGs in carbon metabolism. These results indicated that the slow silk emergence may not be due to a lack of carbon,but rather due to the blockage of silk elongation.

In order to validate the transcriptome data,five genes that were expressed differentially in silk,including three genes involved in expansive growth (GRMZM2G108600(cellulose synthase-like),GRMZM2G107854(pectin synthesis),andGRMZM2G154678(cellulase)) and two genes involved in carbon metabolism (GRMZM2G055489(SPP) andGRMZM2G055331(SPS)),were selected for further validation. In comparison to the planting density of 90 k,the planting densities of 120 k and 150 k presented a downregulation of all five genes (Fig.6). These results were consistent with the transcriptome expression results.

4.Discussion

Hybrid lines are widely recognized to exhibit heterosis(Reifet al.2005;Birchleret al.2010;Chen 2010),although the reason why hybrids do not exhibit any advantages regarding bald tip traits under high planting density remains unknown. The results of the present study revealed that silk expansive growth might be the reason for longer bald tips at higher planting densities.As the planting density increases,the bald tip length also increases. The ASI also increased compared to that at a moderate planting density. The most direct evidence for this conclusion is that the amount of silk that emerged from the husks declined sharply with increasing planting density. This mechanism was further confirmed by cellular observations,which revealed that silk cell length and area both decreased significantly. The changes in the transcription abundance of silks involved the genes related to expansive growth rather than carbon metabolism.

ASI is defined as the time that elapses from male flowering to the emergence of silks from the husks enclosing the ear;and a shorter ASI is reported to be associated with yield maintenance (Bola?os and Edmeades 1996;Varshneyet al.2014). On the other hand,a longer ASI indicates that the male and female ears of the corn cannot flower simultaneously,which results in poor pollination and a more severe abortion (Edmeadeset al.2000;Welckeret al.2007). The ASI values of Zheng58 in the present study were 2.3,4.2 and 4.3 days at the planting densities of 90 k,120 k and 150 k,respectively,while those of PH4CV were 1.5,1.7 and 2.2 days,respectively. The ASI values of Zhengdan958 were 2.3,2.5,3.1,and 3 days at the planting densities of 60 k,90 k,120 k,and 150 k.Moreover,the ASI values of Xianyu335 were 2.1,2.6,3,and 3.5 days at planting densities of 60 k,90 k,120 k,and 150 k. There was no significant difference between the ASI values of plants grown at the 150 k and 120 k planting densities,although the bald tip of plants grown at a planting density of 150 k was much longer compared to that of the plants grown at a planting density of 120 k,which was probably due to the differences in the silk growth rates and silk expansion genes.

Silk emergence occurs in the same order as the development of ovary cohorts,from the basal to apical regions (Cárcova and Otegui 2007). The bottom leaves sprout the silk first,and then the silk is gradually extruded from the husks in a basal-to-apical order(Fuad-Hassanet al.2008). Plants provide sufficient nutrients to prioritize the growth of basal grains,and abortion occurs in the youngest ovaries that do not reach a critical stage by a particular time (Newet al.1994).Owing to the position effect of the silk at the base,the emergence time of basal silks over the husks,i.e.,the silking time,is less strongly impacted by the difference in the planting density. This is the reason why the ASI values of the plants grown at 120 k and 150 k planting densities were almost the same. However,the higher the planting density,the slower the elongation of silk;so the silk at the apical region would not emerge from the husk as scheduled,resulting in poor pollination and a more severe abortion. Therefore,the bald tip was much longer in the plants grown at 150 k compared to those grown at a 120 k planting density,particularly in the hybrid line Xianyu335. Moreover,the sharp decrease in the final amount of extruded silk with an increase in the planting density also confirmed this result.

5.Conclusion

The results of the present study revealed that the arrest of silk expansive growth was the main reason for kernel abortion at a high planting density,which consequently caused severe bald tip. Therefore,it is suggested that increasing the planting density to the extent possible,in order to reduce the bald tip length,is an effective approach for increasing maize yield. Future studies should explore the molecular mechanisms underlying the phenomenon of silk growth arrest at high planting density.The present study provides reliable physiological data for the subsequent research to decipher these molecular mechanisms.

Acknowledgements

This study was supported by the National Key R&D Program of China (2016YFD0101002),the National Natural Science Foundation of China (32072068) and the Central Public-interest Scientific Institution Basal Research Fund,China (1610392021001).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm