The Spermine Synthase OsSPMS1 Regulates Seed ......of polyamines and their products, plants can effec-tively respond to environmental changes and thereby enhance their stress resistance,

1522 Plant Physiology®, December 2018, Vol. 178, pp. 1522–1536, www.plantphysiol.org © 2018 American Society of Plant Biologists. All Rights Reserved.

CM

Polyamines are a class of low-molecular-weight aliphatic cations with strong biological activities, and they are found in a wide variety of prokaryotic and eukaryotic organisms (Tabor and Tabor, 1984; Michael,

2016). Polyamine metabolism, one of the original met-abolic pathways present in the last universal com-mon ancestor, is essential for the cell expansion and growth of eukaryotes, archaea, and many bacteria (Miller-Fleming et al., 2015; Weiss et al., 2016; Michael, 2017). Therefore, polyamine biosynthesis is commonly used as an example in the analysis of basic evolution-ary mechanisms underlying biosynthetic diversity, including gene duplication, horizontal gene transfer, gene fusion, and gene loss.

Putrescine (Put) and spermidine (Spd) are by far the most common polyamines in the three domains of life. There are two main pathways of Put synthesis in plants. (1) Arg loses one molecule of urea to form Orn, which loses one molecule of carboxylic acid via the activity of Orn decarboxylase to yield Put. (2) One molecule of carboxylic acid is removed from Arg by arginine decarboxylase (ADC) to yield agmatine; then, one molecule of ammonia is removed by agmatine iminohydrolase to yield N-carbamoyl-putrescine and finally Put. Spermidine synthase (SPDS) catalyzes the conversion of Put to Spd using aminopropyl donated from decarboxylated S-adenosyl methionine (dcSAM), which then can be converted to Spm by spermine syn-thase (SPMS; Michael, 2016). These enzymes involved in polyamine anabolic synthesis were obtained via

The Spermine Synthase OsSPMS1 Regulates Seed Germination, Grain Size, and Yield1[OPEN]

Yajun Tao,a,2 Jun Wang,b,2 Jun Miao,a Jie Chen,a Shujun Wu,c Jinyan Zhu,a Dongping Zhang,a Houwen Gu,a Huan Cui,a Shuangyue Shi,a Mingyue Xu,a Youli Yao,a Zhiyun Gong,a Zefeng Yang,a Minghong Gu,a Yong Zhou,a,3 and Guohua Lianga,3,4

aJiangsu Key Laboratory of Crop Genetics and Physiology/Co-Innovation Center for Modern Production Technology of Grain Crops, Key Laboratory of Plant Functional Genomics of the Ministry of Education, Jiangsu Key Laboratory of Crop Genomics and Molecular Breeding, Yangzhou University, Yangzhou 225009, ChinabInstitute of Food Crops, Jiangsu Academy of Agricultural Sciences, Nanjing 210014, ChinacShanghai Academy of Agricultural Sciences, Shanghai 201403, ChinaORCID IDs: 0000‑0001‑6863‑1616 (J.W.); 0000‑0002‑0434‑1617 (J.M.); 0000‑0003‑3803‑011X (Y.Y.); 0000‑0002‑2071‑3150 (Z.G.); 0000‑0003‑4137‑3156 (G.L.)

Polyamines, including putrescine, spermidine, and spermine, play essential roles in a wide variety of prokaryotic and eukary-otic organisms. Rice (Oryza sativa) contains four putative spermidine/spermine synthase (SPMS)-encoding genes (OsSPMS1, OsSPMS2, OsSPMS3, and OsACAULIS5), but none have been functionally characterized. In this study, we used a reverse genetic strategy to investigate the biological function of OsSPMS1. We generated several homozygous RNA interference (RNAi) and overexpression (OE) lines of OsSPMS1. Phenotypic analysis indicated that OsSPMS1 negatively regulates seed germination, grain size, and grain yield per plant. The ratio of spermine to spermidine was significantly lower in the RNAi lines and consider-ably higher in the OE lines than in the wild type, suggesting that OsSPMS1 may function as a SPMS. S-Adenosyl-l-methionine is a common precursor of polyamines and ethylene biosynthesis. The 1-aminocyclopropane-1-carboxylic acid (ACC) and ethylene contents in seeds increased significantly in RNAi lines and decreased in OE lines, respectively, compared with the wild type. Additionally, the reduced germination rates and growth defects of OE lines could be rescued with ACC treatment. These data suggest that OsSPMS1 affects ethylene synthesis and may regulate seed germination and plant growth by affecting the ACC and ethylene pathways. Most importantly, an OsSPMS1 knockout mutant showed an increase in grain yield per plant in a high-yield variety, Suken118, suggesting that OsSPMS1 is an important target for yield enhancement in rice.

1This study was supported by grants from the National Key Research and Development Programme (2016YFD0100400), the National Natural Science Foundation of China (31100863), the China Postdoctoral Science Foundation (2016M601899), the Natural Science Foundation of Jiangsu Province (BK20161335), the Postgraduate Education Reform Project of Jiangsu Province (KYCX17_1885), the Innovation Project of Agricultural Science and Technology of Suzhou (SNG201506), and the Priority Academic Programme Development of Jiangsu Higher Education Institutions.

2These authors contributed equally to the article.3Senior authors.4Author for contact: [email protected] author responsible for distribution of materials integral to

the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Guohua Liang ([email protected]).

G.L. and Y.Z. designed the research; Y.T., J.W., J.M., J.C., S.W., J.Z., D.Z., H.G., H.C., S.S., and M.X. performed the research; Z.Y and Z.G. helped with the bioinformatics analysis; Y.Y. provided technical assistance; Y.T., M.G., G.L., and Y.Z. wrote and revised the article.

[OPEN]Articles can be viewed without a subscription.www.plantphysiol.org/cgi/doi/10.1104/pp.18.00877

www.plantphysiol.orgon April 12, 2020 - Published by Downloaded from Copyright © 2018 American Society of Plant Biologists. All rights reserved.

http://crossmark.crossref.org/dialog/?doi=10.1104/pp.18.00877&domain=pdf

https://orcid.org/0000-0001-6863-1616

https://orcid.org/0000-0002-0434-1617

https://orcid.org/0000-0003-3803-011X

https://orcid.org/0000-0002-2071-3150

https://orcid.org/0000-0003-4137-3156

http://dx.doi.org/10.13039/





http://www.plantphysiol.org

http://www.plantphysiol.org/cgi/doi/10.1104/pp.18.00877


Plant Physiol. Vol. 178, 2018 1523

diverse biosynthetic evolutionary pathways. It is likely that the ancestral enzyme was SPDS, which is found in all three domains of life. In flowering plants, SPMS evolved from SPDS via gene duplication (Hamasaki- Katagiri et al., 1998; Panicot et al., 2002). In certain sola-naceous plants, SPDS has been duplicated to produce the homologous protein putrescine N-methyltransferase, which catalyzes the first step in nicotine and tropane alkaloid biosynthesis (Hibi et al., 1994).

Through continuous evolutionary processes, organ-isms formed their own polyamine metabolic systems with different functions. Because polyamines play import-ant roles in molecular mechanisms, the relationships between polyamines and human diseases have been widely assessed (Moinard et al., 2005; Miller-Fleming et al., 2015). Cancer, aging, and neurodegeneration have been associated with polyamines (Vivó et al., 2001; Eisenberg et al., 2009). In plants, polyamines are involved in all stages of growth and development (Kumar et al., 1997; Michael, 2016). The stability of intracellu-lar polyamine levels in plants is maintained through feedback regulation (Seiler et al., 1996; Moschou et al., 2008; Hewezi et al., 2010). In the case of substrate de-ficiency or in the presence of exogenous polyamine synthesis inhibitors, the polyamine levels in plants decrease (Waduwara-Jayabahu et al., 2012), which trig-gers positive feedback to the transport system and promotes the uptake of extracellular polyamines. Exces-sive accumulation of polyamines triggers negative feedback, followed by a decrease in polyamine uptake by the transport system (Fujita et al., 2012; Mulangi et al., 2012a, 2012b). By regulating the concentrations of polyamines and their products, plants can effec-tively respond to environmental changes and thereby enhance their stress resistance, including their resis-tance to heat damage, salt stress, and osmotic stress. At present, research on polyamines has been focused primarily on abiotic and biotic stress tolerance, and the underlying mechanisms have been gradually illumi-nated (Liu et al., 2007, 2015; Yang et al., 2007; Hewezi et al., 2010; Bitrián et al., 2012; Li et al., 2015).

S-Adenosyl-l-methionine (SAM) not only supplies the aminopropyl group for polyamine synthesis but also is the precursor of ethylene, indicating an antago-nistic interaction between the polyamine and ethylene mechanisms (Bitrián et al., 2012). In tomato (Solanum lycopersicum), transgenic lines overexpressing yeast SPDS accumulated more Spd (Nambeesan et al., 2012), while the expression levels of certain genes related to ethylene biosynthesis and signaling were suppressed, leading to increased susceptibility to Botrytis cine-rea (Nambeesan et al., 2012). Additionally, the ADC2, SPDS1, and SPMS genes were strongly up-regulated under drought stress, which was not observed in abscisic acid (ABA)-insensitive and ABA-deficient mutants (Alcazar et al., 2006a). Several ABA-responsive ele-ments also have been found in the promoters of ADC2, SPDS1, and SPMS (Alcázar et al., 2006b). Taken to-gether, these results indicate that ABA is involved in regulating the expression of ADC2, SPDS1, and SPMS

during exposure to drought. Therefore, the polyamine metabolic pathway may interact with plant hormonal pathways to help plants respond to different environ-mental challenges.

In rice (Oryza sativa), polyamines have been reported to be involved in modulating the postanthesis devel-opment of spikelets (Yang et al., 2008). Spd and Spm exhibit much higher levels in superior spikelets com-pared with inferior spikelets based on the polyamine contents determined in six rice cultivars with different grain-filling rates (Yang et al., 2008). The application of exogenous Spd and Spm to rice panicles promotes grain filling and grain weight in inferior spikelets (Chen et al., 2013a). Additionally, the concentrations of Spd and Spm are positively correlated with rice grain quality. Milling and appearance qualities have been improved by the application of exogenous Spd and Spm at the early grain-filling stage (Wang et al., 2007). Research examining the physiological role of polyamines has made progress in rice. However, the mechanisms of genes involved in polyamine synthesis and the regulation of rice growth and yield production remain elusive.

In this study, we functionally characterized the OsSPMS1 gene, which may encode a SPMS in rice. The OsSPMS1 gene plays an essential role in con-verting Spd to spermine (Spm), and it also affects 1-aminocyclopropane-1-carboxylic acid (ACC) and eth-ylene synthesis. Manipulation of the OsSPMS1 gene has substantial effects on multiple traits, including plant height, grain size, seed germination, and yield production. More importantly, knockout of OsSPMS1 can further increase grain yield in a high-yield variety, indicating that OsSPMS1 is a key target gene for rice yield improvement.

RESULTS

Phylogenetic Analysis of SPDS/SPMS Members in Plants

Generally, SPDS/SPMS possesses a Spermine_synth domain. To investigate the evolution and protein se-quence similarity of the plant SPDS/SPMS, we char-acterized them from species representing the main lineages of green plants, including chlorophyte algae, charophyte algae, bryophytes, lycophytes, gymno-sperms, and basal angiosperms, as well as from three monocot and four dicot angiosperms. A total of 14 plant species were selected (Supplemental Table S1). Each of the selected genomes contained at least two members with Spermine_synth domains in its pro-tein product. The phylogenetic tree divided all these genes into two subfamilies with 100% bootstrap sup-port (Supplemental Fig. S1). According to the homo-logs in Arabidopsis (Arabidopsis thaliana), we defined them as SPDS/SPMS and ACAULIS5 (ACL5) subfam-ily members. Although all the proteins in these two subfamilies contained the conserved Spermine_synth

OsSPMS1 Regulates Seed Germination and Yield


http://www.plantphysiol.org/cgi/content/full/pp.18.00877/DC1



1524 Plant Physiol. Vol. 178, 2018

domain, the differentiation characteristics clearly dif-fered between the subfamilies. We further compared the sequence similarities between and within these two subfamilies using Arabidopsis and rice proteins (Supplemental Fig. S2). The protein products had 58% and 64% identities within the SPDS/SPMS and ACL5 subfamilies, respectively, while the average identity was 21% between these two subfamilies (Supplemental Table S2).

We further investigated the evolutionary patterns of plant SPDS/SPMSs. In plants, the SPDS/SPMS- encoding genes first emerged in green algae. All the SPDS/SPMSs in angiosperms were clustered into two groups with high bootstrap support, and both of these groups contained gymnosperm and angiosperm genes (Supplemental Fig. S1). This result suggests that a gene duplication event before the split of seed plants contributed to the evolution of the SPDS/SPMS- encoding genes. In addition, we found that group A of angiosperms contained two monocot-specific branches, while there was only one in group B. In rice, there were four putative SPDS/SPMS-encoding genes: LOC_ Os06g33710 (OsSPMS1), LOC_Os02g15550 (OsSPMS2), LOC_Os07g22600 (OsSPMS3), and LOC_Os02g14190 (OsACL5). The rice genes OsSPMS1 and OsSPMS2 were located in group A and arranged on the two monocot-specific branches. This result revealed that there was a duplication event before the split of grass species.

Expression Pattern and Subcellular Localization of the OsSPMS1 Protein

The functions of the proteins encoded by the four predicted rice SPDS/SPMS-encoding genes have yet to be characterized. In this study, we investigated the roles of OsSPMS1 in rice. Reverse transcription quanti-tative PCR (RT-qPCR) analysis suggested that OsSPMS1 was constitutively expressed in the leaf, node, sheath, root, stem, and panicle, with the highest expression levels in the leaf (Fig. 1A). During the earlier stage of panicle development, the expression level of OsSPMS1 increased with panicle elongation, but it declined significantly after reaching the summit (Fig. 1B). To gain further insight into the expression profiles of OsSPMS1, a 1.9-kb promoter region of OsSPMS1 was cloned to drive the GUS gene, and the promoter-GUS construct was transformed into rice plants. GUS stain-ing showed that OsSPMS1 was expressed abundantly in the leaf, node, stem, and sheath (Fig. 1, C–F); it also was expressed in the root, with particularly strong staining in root tips (Fig. 1G). We also compared the GUS activity among panicles at different stages of development and observed a similar result to the RT- qPCR data (Fig. 1H).

To investigate the subcellular localization of the OsSPMS1 protein, the coding region of OsSPMS1 was fused to GFP to generate a fusion construct. The re-sulting plasmid was transiently expressed in tobacco

Figure 1. Tissue-specific expression and subcellular localization of OsSPMS1. A, Relative expression level of OsSPMS1 in different tissues. B, Transcript levels of OsSPMS1 in developing panicles. Numbers on the horizontal axis represent young pani-cles with an average length of 1 to 3, 4 to 6, 7 to 9, 11 to 13, 15 to 17, and 18 to 20 cm. The data in A and B are presented as means ± sd (n = 3). C to G, GUS activity in the node (C), stem (D), sheath (E), leaf (F), and root tip (G). H, GUS activity in the elongation of young panicles. I, Subcellular localization of OsSPMS1. Localizations of GFP alone and OsSPMS1-GFP in tobacco epidermal cells are shown. The photographs were ob-tained under an optic field to examine cell morphology (Bright field), under a dark field to localize green fluorescence (GFP), and in combination (Merged). C, Cytoplasm; M, membrane; N, nucleus. The boxed images show the magnified cytoplasm-localized signal. Bars = 50 μm.

Tao et al.








(Nicotiana benthamiana) epidermal cells. As shown in Figure 1I, the OsSPMS1-GFP signal was similar to that of GFP alone, which indicated that the OsSPMS1 pro-tein was located in the nucleus, plasma membrane, and cytoplasm.

Manipulation of OsSPMS1 Affects Plant Architecture

To investigate the function of OsSPMS1 in rice, we generated RNA interference (RNAi) and overexpres-sion (OE) constructs of OsSPMS1 under the control of the maize (Zea mays) ubiquitin promoter, which were then transformed into rice variety Nipponbare. A total of eight positive OE lines and 11 RNAi lines were obtained. We selected four homozygous transgenic lines (RNAi1, RNAi2, OE1, and OE2) for further anal-ysis. RT-qPCR indicated that the expression levels of OsSPMS1 were suppressed significantly and elevated in the RNAi and OE lines, respectively, compared with those in the wild type (Fig. 2A). Due to the high se-quence similarity among the four homologs in rice, we also compared the transcripts of the other three genes in the two RNAi lines. As shown in Figure 2B, the tran-scripts of OsSPMS2, OsSPMS3, and OsACL5 were not significantly different between wild-type and RNAi plants. OsSPMS1 expression was suppressed specifi-cally, suggesting that the two RNAi lines could be used for further analysis.

From the seeding stage, the RNAi plants grew faster while the OE plants grew more slowly than the wild type (Supplemental Fig. S3). At the tillering stage, the difference became more pronounced (Supplemental Fig. S3). At maturity, the two OE lines, OE1 and OE2, were significantly shorter than the wild type in both years (Fig. 2, C and D). Conversely, the two RNAi lines were significantly taller than the wild type. Compared with the wild type, the plant height of the RNAi1 line increased by 4.9% and 4.91% in 2016 and 2017, respec-tively. The RNAi2 line was much taller (13.87% and 25.75% in 2016 and 2017, respectively; Fig. 2, C and D). Additionally, the expression level of OsSPMS1 in RNAi2 was lower than that in RNAi1 (Fig. 2A). These data suggest that plant height was negatively related to OsSPMS1 transcript accumulation in rice. To inves-tigate the elongation pattern of internodes, we com-pared the lengths of the three uppermost internodes between wild-type and transgenic plants. The first and second internodes from the top of RNAi1 were lon-ger than in the wild type in both 2016 and 2017. The lengths of all the internodes of RNAi2 lines were sig-nificantly greater than in the wild type in both years (Fig. 2E). In contrast, the first internode of OE1 and all three internodes of OE2 were shorter than those in the wild type in both years.

We also compared the culm diameter of the three uppermost internodes between the wild type and transgenic lines. With the decreased expression level of OsSPMS1, the culms of RNAi lines became thicker than those of the wild type, whereas the elevated tran-script level of OsSPMS1 led to thinner culms (Fig. 3,

A and B). Cross sections of the second internode from the top further showed that the RNAi plants had much better developed vascular systems and larger paren-chyma cells, while the OE plants exhibited smaller vascular bundles and parenchyma cells (Fig. 3C). Moreover, there was a marked difference in cell size between wild-type and transgenic plants in the second uppermost internodes (Fig. 3D; Supplemental Fig. S4). Taken together, these findings showed that OsSPMS1 had marked effects on stem growth and development by negatively regulating cell size.

OsSPMS1 Negatively Regulates Grain Size and Yield

The grain yield components and yield production were examined in 2016 and 2017 (Table 1). No signif-icant changes were observed in panicle number be-tween the wild type and transgenic lines. The panicle length and grain number per panicle from the main

Figure 2. Phenotypic measurements of wild-type (WT) and transgenic plants. A, Expression levels of OsSPMS1 in wild-type and transgenic plants detected by RT-qPCR. B, Comparison of expression levels of OsSPMS1 homologous genes (OsSPMS1, OsSPMS2, and OsACL5) be-tween wild-type and RNAi plants. The data in A and B are presented as means ± sd of triplicate experiments. C, Plant architecture of wild-type and transgenic plants at the mature stage. Bar = 10 cm. D, Plant height of the wild type and transgenic lines at the mature stage in 2016 and 2017. E, Length of the uppermost internodes in wild-type and trans-genic plants. I, II, and III represent the first, second, and third internode from the top, respectively. The data in D and E are presented as means ± sd (n = 15). *, P < 0.05 and **, P < 0.01, Student’s t test.









culms of the two RNAi lines were increased signifi-cantly compared with those in the wild type. By con-trast, the OE lines produced a reduced panicle size. In addition, OsSPMS1 negatively regulated grain length but had no effect on grain width and thickness (Fig. 4). As a result, the average 1,000-grain weight of the RNAi1 and RNAi2 lines increased by 4.48% to 5.05% in 2016 and by 4.46% to 6.67% in 2017 (Table 1). Con-versely, the grains of the OE1 and OE2 lines decreased

in weight by 3.59% to 4.32% in 2016 and by 6.15% to 8.08% in 2017 (Table 1). We also compared the grain yield per plant between the wild type and transgen-ic lines and found that down-regulation of OsSPMS1 resulted in a 16.47% to 21.16% (mean, 18.5%) enhance-ment of yield in 2016 and 2017, while overexpression of OsSPMS1 led to obvious yield losses (Table 1). These data suggest that OsSPMS1 negatively regulates grain size and yield.

Figure 3. Effects of OsSPMS1 on the stem. A, Cross sections of the main culms (I–III show the first, sec-ond, and third internodes from the top). WT, Wild type. Bar = 5 mm. B, Diameter of the axes from the first internode to the third internode from the top. The data are presented as means ± sd (n = 15). *, P < 0.05 and **, P < 0.01, Student’s t test. C, Transverse sections of the middle part of inter-node II from the main culms. LVB, Large vascular bundles; PC, parenchyma cells. Bars = 200 μm. D, Longitudinal sections of internode II from the main culms. Bars = 50 μm.

Table 1. Comparison of yield-related traits between the wild type and transgenic linesAsterisks indicate significant differences compared with the wild type: *, P < 0.05 and **, P < 0.01.

Trait Year Wild Type RNAi1 RNAi2 OE1 OE2

Panicle number 2016 9.07 ± 1.32 9.58 ± 1.31 9.15 ± 1.34 9.36 ± 1.12 9.50 ± 1.272017 9.77 ± 1.65 9.40 ± 1.55 9.57 ± 1.45 9.46 ± 1.20 9.01 ± 1.44

1,000-grain weight (g)

2016 24.76 ± 1.03 25.87 ± 0.61** 26.01 ± 0.36** 23.87 ± 0.43** 23.69 ± 0.73**2017 24.88 ± 0.52 25.99 ± 0.52** 26.54 ± 0.58** 23.35 ± 0.45** 22.87 ± 0.53**

Panicle length (cm)

2016 22.59 ± 1.77 23.68 ± 1.31* 23.86 ± 1.37** 20.58 ± 0.70** 20.77 ± 0.79**2017 22.39 ± 1.40 23.83 ± 1.62** 24.76 ± 1.80** 20.46 ± 0.75** 21.21 ± 1.72**

Grain number per panicle

2016 133.71 ± 10.41 149.21 ± 19.08** 152.93 ± 14.53** 122.07 ± 10.67** 118.86 ± 13.88**2017 138.13 ± 11.08 157.33 ± 12.54** 168.38 ± 28.14** 130.33 ± 8.37* 125.13 ± 10.25**

Grain yield per plant (g)

2016 18.94 ± 1.87 22.06 ± 1.81** 22.64 ± 2.67** 17.39 ± 1.04** 16.24 ± 2.06**

2017 20.13 ± 3.43 23.52 ± 2.13** 24.39 ± 2.23** 17.05 ± 3.65* 16.57 ± 2.79**

Tao et al.




To elucidate how OsSPMS1 regulates grain length, we investigated cell size and cell number in the outer epidermis of spikelets using a scanning electron mi-croscope (Fig. 5). We divided the spikelet hull into up-per, middle, and lower parts to measure the cell size in these three locations (Fig. 5, A–C). Both the lemma and palea in the RNAi plants contained obviously larger cells in the upper, middle, and lower parts than the corresponding sites in the wild type (Fig. 5, D and E). By contrast, the cells in the OE spikelets were signifi-cantly smaller than those in the wild type (Fig. 5, D and E). Additionally, there were no differences in total cell numbers in the lemma along the longitudinal axis be-tween wild-type and transgenic plants (Fig. 5F). These results suggest that OsSPMS1 regulates spikelet size by affecting cell expansion.

OsSPMS1 Affects Polyamine Synthesis

In plants, SPDS catalyzes the conversion from Put to Spd, which then can be converted to Spm by SPMS (Fig. 6A; Michael, 2016). To verify whether OsSPMS1 affects polyamine synthesis, we detected the polyamine content in flag leaves at the heading stage. Down-regulation of OsSPMS1 resulted in decreases in the contents of endogenous Put and Spm but an increase in Spd con-tent (Fig. 6B). Overexpression of OsSPMS1 led to a sharp consumption of Spd and accumulation of Put (Fig. 6B). However, there was no significant difference in Spm content between wild-type and OE plants. We also calculated the ratios of Spm to Spd, which were re-duced significantly in the two RNAi lines and extreme-ly elevated in the two OE lines compared with those in the wild type (Fig. 6C). Additionally, the expression of OsSPMS1 could be induced rapidly by Spd and in-hibited by Spm but was unaltered after treatment with Put (Fig. 6D). Based on these findings, we propose that OsSPMS1 can consume Spd and functions as an SPMS.

OsSPMS1 Influences Ethylene Homeostasis to Regulate Seed Germination and Shoot and Root Development

When sowing seeds, we found that the seed germi-nation rates of RNAi lines were higher than those of the wild type, while OE lines showed lower germina-tion rates. This result was confirmed further by a seed germination assay. Four days after soaking, freshly harvested seeds of the wild type germinated at a rate of 43.33%, whereas OE1 and OE2 seeds germinated at rates of 6.46% and 6.67%, respectively (Fig. 7, A and B). By contrast, the germination rates of RNAi1 and RNAi2 were 75.01% and 82.23%, respectively (Fig. 7, A and B). GUS staining indicated that many transcripts of OsSPMS1 accumulated at the early stage of seed germination (Fig. 7C). We conclude that OsSPMS1 is involved in regulating seed germination in rice.

Considering that the Spd contents were up- and down-regulated in RNAi and OE plants, respectively, we were interested in the role of Spd in seed germina-tion. A low concentration of Spd (10 and 100 μm) had no obvious effect on germination (Fig. 7E). However, seed germination was suppressed significantly after treatment with 1,000 μm Spd for 5 d (Fig. 7, D and E). Interestingly, we observed that the RNAi plants nearly stopped growing 9 d after treatment, while the shoot lengths of OE plants were significantly longer than those of the wild-type plants (Fig. 7, D and F). A possi-ble explanation is that the overexpression of OsSPMS1 resulted in the degradation of exogenous Spd. By com-parison, down-regulation of OsSPMS1 promoted the accumulation of Spd, which suppressed postgermina-tion seedling growth.

SAM is the common precursor of polyamines and ethylene biosynthesis, which raises the question whether seed germination was affected by the ethylene path-way. We measured the ACC and ethylene contents in

Figure 4. Effects of OsSPMS1 on grain size. A and B, Grain shape in the wild type (WT) and transgenic lines. Bars = 1 cm. C to E, Mean values of grain length (C), grain width (D), and grain thickness (E) in wild-type and transgenic plants. The data are presented as means ± sd (n ≥ 70). **, P < 0.01, Student’s t test.





mature seeds and found that both were increased in RNAi lines and decreased in OE lines compared with those in the wild type (Fig. 8, A and B). The germination rates of OE seeds could be rescued with 100 μm SAM and ACC treatments (Fig. 8, C and D). Furthermore, the plant growth suppression of OE lines was rescued under 10 and 50 μm ACC conditions (Supplemental Fig. S5). We also observed that the root length of RNAi plants cultivated in a growth chamber was significantly shorter than that of wild-type plants (Supplemental Fig. S6, A, B, and E), a finding that resembled the behavior

of plants with enhanced ethylene responses, such as the rice constitutive triple response2 mutant (Wang et al., 2013) and OE lines of MAOHUZI7 (Ma et al., 2013). When treated with 10 µL L−1 1-methylcyclopropene (1-MCP; an inhibitor of ethylene perception), the root length of RNAi plants increased significantly (Supplemental Fig. S6, C–E). By contrast, Put, Spd, and Spm had negative effects on seedling growth at all detected concentrations (Supplemental Fig. S7). We propose that OsSPMS1 may regulate seed germination and plant growth by affect-ing the ACC and ethylene pathways.

Figure 5. Scanning electron microscopy analysis of spikelet hulls in wild-type (WT) and transgenic plants. A, Spikelet hulls in wild-type and transgenic plants. Bar = 1 mm. B, Scanning electron microscopy observation of lemma. Bars = 50 μm. C, Scanning electron microscopy observation of palea. Bars = 50 μm. D, Cell size of lemma in wild-type and transgenic plants. E, Cell size of palea in wild-type and transgenic plants. Cell size was estimated using ImageJ software. F, Total cell number of full spikelets along the longitudinal axis. Total cell number was calculated according to the length of the longitudinal axis and the average cell number in the upper, middle, and lower parts. The data in D to F are presented as means ± sd (n ≥ 35). *, P < 0.05 and **, P < 0.01, Student’s t test.

Tao et al.











Transcriptomic Analysis of Young Panicles from the Wild Type and OsSPMS1 Transgenic Lines

Because of the significant influences of OsSPMS1 on panicle elongation and spikelet size, we performed a transcriptome analysis to explore the possible molec-ular pathway of OsSPMS1 using young wild-type, RNAi2, and OE2 panicles. Compared with the wild type, a total of 2,002 differentially expressed genes

(DEGs; ratio ≥ 2) were detected in RNAi2 (Supple-mental Fig. S8A), of which 34.1% (682 genes) were up- regulated and 65.9% (1,320 genes) were down-regulated. A total of 1,699 DEGs were detected in OE2 (Supple-mental Fig. S8A), of which 37.3% (634 genes) were up-regulated and 62.7% (1,065 genes) were down- regulated compared with those in the wild type. Enrich-ment analysis of DEGs showed that these genes were

Figure 6. Comparison of endogenous polyamine content between wild-type (WT) and transgenic plants. A, Schematic diagram of Put, Spd, and Spm biosynthesis in plants. SPDS and SPMS are respon-sible for the transfer of aminopropyl groups to Put and Spd, respectively. B, Contents of Put, Spd, and Spm in wild-type and transgenic lines. FW, Fresh weight. C, Ratio of endogenous Spm to Spd in wild-type and transgenic lines. The data in B and C are presented as means ± sd (n = 3). **, P < 0.01, Student’s t test. D, Effects of Put, Spd, and Spm on transcript levels of OsSPMS1. Ten-day-old seedlings of Nipponbare grown in a plastic box were treat-ed with 100 μm Put, Spd, and Spm; then, the RNA from leaves was extracted for expression analysis. The data in D are presented as means ± sd of three biological triplicate experiments.

Figure 7. Germination of wild-type (WT) and transgenic seeds. A, Gemi-nated seeds of wild-type and transgen-ic plants after soaking in water for 4 d. Bars = 1 cm. B, Time course of germina-tion. Germination rates at the indicated time points are shown. Values represent means ± sd of triplicate experiments with 25 seeds per sample. C, GUS staining of dehulled OsSPMS1 promoter-GUS seeds. Transgenic seeds that were soaked in water for 24 h were stained for the photographs. Bars = 1 mm. D, Wild-type and transgenic seeds were soaked in water and Spd (1,000 μm) for 5 and 9 d. Bars = 1 cm. E, Germination of wild-type and transgenic seeds after treatment for 5 d. F, Shoot length of wild-type and transgenic seeds after treatment for 9 d. All seeds were soaked in water and Spd. After 5 and 9 d, the seeds were observed and photographed. The data in E and F are presented as means ± sd of triplicate experiments with 25 seeds per sample. **, P < 0.01, Student’s t test.









enriched significantly in biological process, cellular component, and molecular function Gene Ontology terms. For biological process, metabolic process was the top annotated term. For molecular function, most of the DEGs were enriched in binding and catalytic activity terms (Supplemental Fig. S8B). Additionally, Cys and Met metabolism (ko00270) was the top annotated path-way, and Met is the precursor of ethylene and polyam-ine synthesis. These results indicate that OsSPMS1 may perform a catalytic function.

Knockout of OsSPMS1 Increases Yield in the Background of a High‑Yielding Variety

The OsSPMS1 RNAi lines showed better vigor and higher grain yield per plant. We explored whether knockout of this gene still has the potential to increase

grain yield in the background of a high-yielding variety. The OsSPMS1 gene was knocked out in the high-yielding variety Suken118 using a clustered regu-larly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein9 (Cas9) system, and two mutants (osspms1#1 and osspms1#2) were obtained. Sequencing analysis indicated that one- and two- nucleotide deletions were found in the target regions of osspms1#1 and osspms1#2, respectively (Fig. 9A), each of which caused a premature termination codon to yield shortened polypeptide products (Supplemen-tal Fig. S9). T2 homozygous lines were planted for the assessment of agronomic traits. The two mutation lines showed higher growth vigor from the seedling stage, similar to the RNAi lines. At the mature stage, both osspms1#1 and osspms1#2 exhibited significantly increased plant height and enlarged panicles and grain size (Fig. 9, B–D and F). Compared with the wild type, osspms1#1 and osspms1#2 also showed an accelerated germination rate (Supplemental Fig. S10). Most impor-tantly, the grain yield per plant of osspms1#1 and osspms1#2 increased by 14.39% and 13.08%, respective-ly, compared with that of the wild type (Fig. 9, E and F). In conclusion, we propose that knockout of OsSPMS1 enhances yield production in a high-yield variety and that the OsSPMS1 gene is an important target site for rice yield improvement.

DISCUSSION

SPDS, which catalyzes the transfer of aminopropyl groups to Put, is the universal pathway for the forma-tion of Spd in eukaryotes (Hoyt et al., 2000a, 2000b; Clark et al., 2010; Colotti and Ilari, 2011; Willert and Phillips, 2012). SPMS evolved from the duplication of SPDS (Wu et al., 2008; Pegg and Michael, 2010). In this study, we investigated the evolutionary patterns of plant SPDS/SPMS members (Supplemental Fig. S1). Phylogenetic analyses suggested that gene duplica-tion before the split of seed plants contributed to the evolution of the SPDS/SPMS-encoding genes. Addi-tionally, the SPDS/SPMS subfamily was classified into two groups, with one containing two monocot-specific branches and the other containing one branch. The rice genes OsSPMS1 and OsSPMS2 are located in group A and arranged on these two branches. These results revealed that a duplication event occurred before the split of grass species.

The crystal structure of SPDS has been revealed, and the substrate-binding sites have been identified (Korolev et al., 2002; Dufe et al., 2005). Spm and ther-moSpm synthases exhibit a similar spatial arrange-ment of these binding sites (Knott et al., 2007). AtACL5 from Arabidopsis acts as a thermoSpm synthase (Knott et al., 2007). The conserved amino acid sequence GG-GDG constitutes the binding site for dcSAM in SPDSs, which also is found in SPMSs (Knott et al., 2007). The Asp residue in this sequence is responsible for an ionic interaction with the amino group of the aminopropyl

Figure 8. Effects of SAM and ACC on seed germination of the wild type (WT) and transgenic lines. A, ACC content in wild-type and trans-genic seeds. Mature seeds were immediately ground into powder with liquid nitrogen for ACC quantification. B, Ethylene content in wild-type and transgenic seeds. Mature seeds were placed in a sealed peni-cillin bottle for ethylene detection. For A and B, the data are presented as means ± sd of triplicate experiments with 20 seeds per sample. *, P < 0.05 and **, P < 0.01, Student’s t test. FW, Fresh weight. C, Wild-type and transgenic seeds treated with water, SAM (100 μm), and ACC (100 μm). Seed germination was observed after treatment for 5 d. Bars = 1 cm. D, Germination of wild-type and transgenic seeds after treatment for 5 d. The data are presented as means ± sd of triplicate experiments with 25 seeds per sample. **, P < 0.01, Student’s t test.

Tao et al.









moiety of dcSAM. All ACL5 family members replace Asp (D) with Glu (E), resulting in the domain se-quence GGGEG/L. Sequence analysis indicated that the OsSPMS1 protein possesses the amino acid sequence GGGDG (Supplemental Fig. S2) and may act as an SPMS or SPDS. We compared the ratio of Spm to Spd between the wild type and transgenic lines and found that the Spm/Spd ratio was reduced significantly in the two RNAi lines and extremely elevated in the two OE lines (Fig. 6C). We propose that OsSPMS1 can con-sume Spd and functions as an SPMS.

In this study, we used a reverse genetic strategy to elucidate the function of OsSPMS1. Two RNAi and OE lines of OsSPMS1 were generated and used for further analysis. To investigate the biochemical function of OsSPMS1, the endogenous polyamine content in the transgenic lines was measured (Fig. 6B). In the RNAi plants, the contents of Spd and Spm increased and

decreased, respectively. In contrast, Spd was overcon-sumed in the OE plans. We postulated that OsSPMS1 could consume Spd and may act as an SPMS. We also attempted to detect the enzymatic activity in vitro. The OsSPMS1 coding sequence was PCR amplified and cloned into pCzn1, and the pCzn1-OsSPMS1 con-struct was introduced into Escherichia coli for protein expression and purification. The obtained recombinant proteins were incubated with the substrates, including Put or Spd, according to a previously reported method (Yoon et al., 2000). However, no expected product was detected (data not shown). The first possible explana-tion for this finding is that the recombinant protein did not possess activity in vitro. In fact, there is evidence that plant SPDS/SPMS may occur in the form of a mul-tienzyme complex (Panicot et al., 2002). In Arabidop-sis, interactions of SPDS1-SPDS2, SPDS1-SPMS, and SPDS2-SPMS were confirmed to regulate polyamine

Figure 9. Molecular and phenotypic identification of the CRISPR/Cas9 mutants of OsSPMS1. A, Targeting sequence results for osspms1#1 and osspms1#2. The targeted sites are labeled with the black line; red and green triangles show the termination codon of OsSPMS1 in osspms1#1 and osspms1#2, respectively. B, Plant morphology of the wild type (WT), osspms1#1, and osspms1#2. Bar = 20 cm. C, Panicle morphology of the wild type, osspms1#1, and osspms1#2. Bar = 2 cm. D, Grains of the wild type, osspms1#1, and osspms1#2. Bar = 1 cm. E, Grains from whole plants of the wild type, osspms1#1, and osspms1#2. F, Comparison of plant height, panicle number, grain number per panicle, grain length, 1,000-grain weight, and grain yield per plant between Suken118 (the wild type) and the two osspms1 mutants. The data for plant height and grain number per panicle were obtained from the main culms of each plant. All data are presented as means ± sd (n = 15). **, P < 0.01, Student’s t test.






biosynthesis (Panicot et al., 2002). Multienzyme com-plexes broadly exist in plants, and they can decrease the transit time required for an intermediate to reach the active site of the next enzyme (Howard et al., 2011; Schoberer et al., 2013). Using the yeast two-hybrid system, we found that OsSPMS1 could interact with OsSPMS2 (Supplemental Fig. S11A). The interactions between OsSPMS1 and OsSPMS2 were validated fur-ther in a bimolecular fluorescence complementation (BiFC) assay. OsSPMS1 was fused to the C-terminal end of yellow fluorescent protein (cYFP), while Os-SPMS2 was fused to the N-terminal end of yellow flu-orescent protein (nYFP). Both were cotransferred into tobacco leaf cells via Agrobacterium tumefaciens infiltra-tion. The resulting YFP fluorescence was observed in the nucleus, cytoplasm, and membrane of tobacco leaf epidermal cells (Supplemental Fig. S11B). Therefore, another possible reason why the expressed OsSPMS1 protein showed no activity is that OsSPMS1 and Os-SPMS2 may work as a complex.

Additionally, we found that overexpression of Os-SPMS1 did not promote the accumulation of Spm (Fig. 6B). In Arabidopsis, overexpression of SPDS slightly increased the Spd content (1.3- to 2-fold), which is rela-tively low in plants (Kasukabe et al., 2004). In tobacco, increased activity of SPDS only elevated the content of Spd, while it had no effects on Spm and total poly-amine contents (Franceschetti et al., 2004). Therefore, polyamine levels are not only determined by the activ-ities of SPDS/SPMS but also affected by metabolism or unknown mechanisms.

The debate on whether polyamines belong to the hormone family is ongoing. Regardless, cross talk oc-curs between polyamines and ethylene due to their common precursor SAM. A competitive relationship has been reported between polyamines and ethylene (Apelbaum et al., 1981). In sorghum (Sorghum bicolor) under salt stress conditions, silicon promotes polyam-ine accumulation but suppresses ethylene accumula-tion (Yin et al., 2016). In this study, manipulation of OsSPMS1 not only modulated polyamine contents but also altered ACC and ethylene accumulation. For the synthesis of Spd and Spm, one and two SAM molecules are needed, respectively. When OsSPMS1 was over-expressed, Spd and SAM were consumed simultane-ously. As a result, the residual SAM for ACC synthesis declined, which negatively affected the content of ACC and ethylene (Fig. 8, A and B). ACC and ethylene have various effects on rice plant growth and development, including seed germination, root development, flower-ing, grain filling, and leaf senescence (Yang et al., 2006; Chen et al., 2013b; Ma et al., 2013; Wang et al., 2013). In this study, seed germination was promoted signifi-cantly in RNAi lines and suppressed in OE lines (Fig. 7A). This phenotype difference raised the question of whether the modified germination rate was caused by alterations of endogenous ACC and ethylene. In bar-ley (Hordeum vulgare), polyamines can be substituted for ethylene during germination (Locke et al., 2000). We did not observe an increase in germination when

seeds were treated with 10 or 100 μm Spd. A signif-icant decrease in germination was found under con-ditions of 1,000 μm Spd (Fig. 7, D and E). However, the suppressed germination of OE seeds could be recovered by the addition of exogenous SAM and ACC (Fig. 8, C and D). The root lengths in the RNAi lines were suppressed significantly, and this suppression could be rescued by 1-MCP (Supplemental Fig. S6). Ethylene-mediated inhibition of root growth is widely observed in plants (Ma et al., 2013). By contrast, mod-erate exogenous ACC was shown to promote shoot growth (Jun et al., 2004). Under deepwater conditions, ethylene accumulates in the plant and induces the ex-pression of SNORKEL1 and SNORKEL2, which trigger remarkable internode elongation via GA (Hattori et al., 2009). Recently, another ethylene-responsive transcrip-tion factor, OsEIN3-LIKE1a, also was found to tran-scriptionally activate SEMIDWARF1 to increase the synthesis of GAs, largely GA4, which promotes inter-node elongation in deepwater conditions (Kuroha et al., 2018). In this study, we also observed that exogenous ACC, the precursor of ethylene, obviously promoted shoot elongation and rescued the growth defects of OE plants (Supplemental Fig. S5). Taken together, OsSPMS1 may regulate seed germination and plant growth main-ly via the ACC and ethylene pathways.

Reports on the influence of polyamine biosynthesis on plant architecture and yield in plants are scarce, possibly due to the complex nature of metabolic net-works, which often are spatially and temporally reg-ulated at different levels. In our study, we found that OsSPMS1 had multiple effects on agronomic traits and negatively regulated plant height, grain number, grain size, and yield production (Table 1). OsSPMS1 also had a strong effect on stem growth. The cross sections of culm internodes showed that the RNAi plants had larger vascular bundles and parenchyma cells than wild-type plants. Vascular tissue plays an important role in mediating whole-plant carbohy-drate partitioning from source leaves to sink organs and is essential for the yield potential of rice (Eom et al., 2012). Moreover, knockout of OsSPMS1 enhanced yield production in a high-yield variety, suggesting that this gene is an important target site for rice molecular breeding.

The effect of ACC and ethylene on rice yield-related traits is largely obscure and must be elucidated. In rice, evidence suggests that an enhanced ethylene response may lead to larger grains, and plants with reduced sensitivity to ethylene produce small grains (Wuriyanghan et al., 2009; Ma et al., 2013; Yang et al., 2015). However, these ethylene-sensitive mutants or transgenic plants with increased grain size do not always have a higher 1,000-grain weight and grain yield due to ethylene-induced leaf senescence (Ma et al., 2013). Leaf senescence has an important effect on grain filling, which is a key determinant of grain weight and yield production in rice. However, there was no obvious leaf senescence in OsSPMS1 RNAi plants. The physiological and molecular mechanisms

Tao et al.








by which OsSPMS1 improves rice yield components remain unknown.

Transcriptome analysis showed that most DEGs were enriched in the binding and catalytic activity terms of molecular function. Decreased expression of OsSPMS1 caused a lower synthesis efficiency of Spm and a greater flux of aminopropyl groups flowing into the biosynthesis of Spd. Therefore, the endogenous contents of Put and Spm were decreased in RNAi lines. When OsSPMS1 was overexpressed, Spd was overcon-sumed, with a reduction in the aminopropyl groups used in the biosynthesis of Spd, which led to the accu-mulation of Put. SAM is a common precursor of poly-amines and ACC synthesis, which plays an essential role in the biosynthesis of both polyamines and ACC. In OsSPMS1 RNAi plants, more SAM flowed into the ACC and ethylene pathways. By contrast, elevated expression of OsSPMS1 led to an overconsumption of SAM, leading to a decreased accumulation of ACC and ethylene. Based on all of our results, we have summarized a working model in which OsSPMS1 par-ticipates in polyamine and ethylene homeostasis and negatively regulates seed germination, grain size, and yield in rice (Fig. 10).

MATERIALS AND METHODS

Homologous Detection and Phylogenetic Analysis

To detect the homologs of SPDS/SPMS in green plants, the rice (Oryza sativa) OsSPMS1 protein sequence was used as a query to search the Phytozome, National Center for Biotechnology Information nonredundant, and ref_seq protein and spruce (Picea spp.) genome project databases. Then, the Pfam tool

was used to predict the Spermine_synth domain (PF01564). Multiple alignments were performed using ClustalX. Maximum-likelihood (ML) and neighbor-join-ing (NJ) methods were adopted to analyze the phylogenesis using IQ-TREE and MEGA version 7.0, respectively (Nguyen et al., 2015). The most optimal models of protein substitution and rate heterogeneity were selected automatically when the ML phylogenetic analysis was performed using IQ-TREE. The NJ phyloge-netic analyses were conducted with the following parameters: Dayhoff model, four rate categories, and estimated gamma distribution parameter. A total of 1,000 nonparametric bootstrap samplings were carried out to estimate the support level for each internal branch for both the ML and NJ trees. The branch lengths and topologies of all phylogenies were calculated with the distance method. The phy-logenetic tree was displayed using the tree explorer tool in MEGA.

Vector Construction and Rice Transformation

To construct the OsSPMS1-RNAi construct, the end of the last exon and partial 3′ untranslated region of OsSPMS1 were amplified due to the high sequence sim-ilarity among four putative SPDS/SPMS-encoding genes in rice. The target frag-ment then was cloned into the p1022 vector and transferred into p1301UbiNOS expressed under the control of the maize (Zea mays) ubiquitin promoter (Zhou et al., 2009). To generate the OsSPMS1-OE construct, the coding region of OsSPMS1 was amplified from Nipponbare cDNA and then inserted into the p1301UbiNOS vector. The 1.9-kb promoter region of OsSPMS1 was cloned into pCAMBIA1301 to drive the expression of the GUS gene. At the heading stage, each tissue was col-lected for GUS staining according to a previous method (Weingartner et al., 2011). All the constructs were transformed into the wild-type japonica rice Nipponbare via Agrobacterium tumefaciens-mediated transformation.

To generate the CRISPR/Cas9 mutant, we selected single-guide RNAs targeting the first exon of the OsSPMS1 gene. The final fragment was insert-ed into a CRISPR/Cas9 system (Baige) in which the Cas9 destination vector was driven by the maize ubiquitin promoter for expression in rice, and sin-gle-guide RNA expression was driven by the U6 promoter. The CRISPR/Cas9 construct was transformed into Suken118 via A. tumefaciens-mediated trans-formation. Suken118 is a high-yielding japonica variety that was released in Jiangsu Province, China, in 2016.

Plant Materials and Growth Conditions

Nipponbare transgenic lines (RNAi and OE) of the T2 and T3 generations were used for phenotypic analyses in 2016 and 2017, respectively. Suken118

Figure 10. Proposed working model of OsSPMS1 in regulating seed germina-tion, grain size, and yield production in rice. Rectangular and elliptical boxes represent catalytic enzymes. The blue arrows indicate changes in substance content in the RNAi lines; the red arrows indicate changes in substance content in the OE lines; arrows pointing up or down indicate that the content has increased or decreased, respectively. KMTB, 2-Ke-to-4-methylthiobutyrate; MET, Met; MTA, 5′-methylthioadenosine; MTR, meth-ylthioribose; MTR-P, MTR-1-phosphate; OsACO, ACC oxidase; OsACS, ACC syn-thase; OsSAMDC, SAM decarboxylase; OsSAMS, SAM synthetase; PAO, polyam-ine oxidase.





and the two CRISPR/Cas9 mutants were grown in 2017. These materials were planted at an experimental farm at Yangzhou University according to normal agricultural practices. For seed germination, seeds of the wild type and RNAi and OE transgenic lines harvested after 2 months were dehulled and soaked in water in a growth chamber under dark conditions for the first 48 h, followed by a 12/12-h light/dark cycle at 30°C during the light period and 28°C during the dark period. The germination rate was defined according to a previous report (Ye et al., 2015), and every experiment was repeated at least three times. For 1-MCP treatments, rice seedlings were grown on plastic plates placed in an air-tight plastic box and incubated in a growth chamber for 7 d (Ma et al., 2013). For hydroponic growth experiments, germinated seeds were placed onto plastic plates and grown in an illumination incubator (12 h of light, 30°C and 12 h of dark, 28°C). One week later, seedlings were treated with different concentrations of Put, Spd, Spm, and ACC for 7 d.

RNA Extraction and RT‑qPCR

Total RNA was extracted using an RNA extraction kit (with gDNase; Tiangen). cDNAs were synthesized using a FastQuant RT Kit (with gDNase) following the manufacturer’s instructions (Tiangen). Gene expression levels were analyzed using qPCR. The rice β-Actin gene (LOC_Os03g50890) was used as an internal control. qPCR was carried out in a total volume of 20 µL containing 2 µL of the cDNA, 10 μm of each primer, 10 µL of 2×SYBR Green PCR master mix, and 0.4 µL of 50×ROX Reference Dye 2 (Vazyme). qPCR was performed on an ABI ViiA7 real-time PCR system (Applied Biosystems) using the following program: 95°C for 5 min, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s.

Morphological and Cellular Analyses

Fresh spikelets from the young panicles of wild-type and transgenic plants at the heading stage were collected and observed with a scanning electron microscope. For histology, spikelets and culms were fixed in 2.5% (v/v) glu-taraldehyde for more than 24 h and then dehydrated in a graded ethanol series and embedded in Spurr resin. Transverse sections of each sample were pro-duced using an ultramicrotome (EM UC7; Leica) and then stained with 0.5% Toluidine Blue and viewed using a microscope (DM1000; Leica). The cell size of each sample was measured using ImageJ software.

Subcellular Localization of OsSPMS1 Proteins

For subcellular localization analysis, the coding sequence of OsSPMS1 was fused in frame to the GFP coding sequence in the pCAMBIA1300-221GFP vec-tor to generate OsSPMS1-GFP. GFP alone was used as a control. Both were driven by the CaMV 35S promoter. These two fusion constructs were intro-duced into A. tumefaciens strain EHA105. Then, a tobacco (Nicotiana benthami-ana) leaf was infiltrated by A. tumefaciens to express the GFP fusion protein. Approximately 48 h later, the tobacco epidermal cells were observed and pho-tographed using a confocal microscope (LSM710; Zeiss).

Yeast Two‑Hybrid and BiFC Assays

To assay the interactions between OsSPMS1 and OsSPMS2, the coding re-gions of these two genes were amplified from Nipponbare cDNA and then inserted into the pGBKT7-BD and pGADT7-AD vectors (Clontech). Yeast two-hybrid assays were performed in accordance with the manufacturer’s instructions.

For BiFC assays, the coding region of OsSPMS1 was cloned into the cYFP vector and the coding region of OsSPMS2 was cloned into the nYFP vector. The plasmids were introduced into A. tumefaciens strain EHA105 and coin-filtrated into tobacco leaves. YFP fluorescence was visualized with a confocal scanning microscope (LSM710; Zeiss) after infiltration for 72 h.

Evaluation of Agronomic Traits

At the mature stage, several yield-related agronomic traits were mea-sured in 2016 and 2017. Plant height was measured from the tallest panicle to the ground surface. The panicle of the main stem was selected for mea-suring panicle length and counting the grain number per panicle. The flag leaf length and width of wild-type and transgenic plants also were mea-sured. Grain traits, including grain length, grain width, and 1,000-grain

weight, were measured after harvesting and storage at 37°C for at least 2 weeks. The total filled grains of one plant were weighed to measure the grain yield per plant.

Transcriptome Analysis

Young panicles of wild-type and transgenic plants were sampled for RNA sequencing analyses. Total RNA was extracted from each sample using TRIzol Reagent (Invitrogen), an RNeasy Mini Kit (Qiagen), and other kits. Total RNA from each sample was quantified and qualified using an Agilent 2100 Bioan-alyzer (Agilent Technologies), NanoDrop (Thermo Fisher Scientific), and 1% agarose gels. Approximately 1 μg of total RNA with an RNA integrity number value greater than 7 was used for subsequent library preparation. Next- generation sequencing library preparations were constructed according to the manufacturer’s protocol (Next Ultra RNA Library Prep Kit for Illumina; New England Biolabs). DEGs were defined by a 2-fold expression difference with a false discovery rate value less than 0.05. The raw RNA sequencing data were successfully submitted to the SRA database of the National Center for Bio-technology Information (https://www.ncbi.nlm.nih.gov/), and the accession number is SRP153908.

Measurement of Polyamines, ACC, and Ethylene

For total polyamine extraction, the appropriate amount of sample was ac-curately weighed in a 100-mL Erlenmeyer flask, followed by the addition of 10 mL of 0.6 mol L−1 perchloric acid, shaking, and extraction for 1 h. The superna-tant was filtered, combined, centrifuged at 4,000 rpm for 8 min, and brought to 50 mL. One milliliter of the sample solution and standard amine solution was collected, and the pH value was neutralized with 2 mol L−1 NaOH solution prior to the addition of 100 μL of sodium bicarbonate solution and 1 mL of derivatizing agent (5 mg mL−1 dansyl chloride acetone solution), followed by shaking and mixing. After standing for 30 min at room temperature, 100 μL of ammonia was added, mixed, and allowed to stand for an additional 20 min. The acetone was removed under reduced pressure in a 60°C water bath. Next, 3 mL of diethyl ether was added with mixing and vortexing on a vortex shaker for 1 min and then allowed to stand. After delamination, the upper organic phase (ether layer) was added to 3 mL of ether and again extracted. The ether extracts were combined and evaporated to dryness under reduced pressure in a 40°C water bath. Then, 3 mL of methanol was added to dissolve the residue with shaking and mixing, followed by passing the sample through a 0.45-μm microporous membrane and quantifying it via HPLC as described previously (Chen et al., 2013a).

ACC was extracted from 20 seeds of wild-type and transgenic plants. The samples were immediately ground into powder with liquid nitrogen in 80% (v/v) ethanol and centrifuged. The supernatant was evaporated to dryness using nitrogen. The residue was resuspended in 2 mL of water. The ACC content was determined as described previously (Lizada and Yang, 1979).

Twenty seeds were transferred to a known volume in a penicillin bottle, the rubber stopper was sealed, and the seeds were placed in the dark at room temperature for 8 h. Then, 1 mL of gas was sampled with a syringe and measured with a gas chromatograph (HP5890 Series II; Hewlett Packard). The column was a Propark Q (0.3 cm × 250 cm, 50–80 mesh), and the detector was a hydrogen flame detector with an inlet temperature of 140°C, column temperature of 80°C, and detector temperature of 200°C. N2 was used as the carrier gas and was fed at a rate of 30 mL min−1. H2 and air were used as the flame ionization detection gas, and the intake airspeeds were 30 and 300 mL min−1, respectively.

Statistical Analysis

All numerical data are presented as means ± sd (error bars indicate the sd). Statistical analyses were carried out by comparing the raw data for all indi-viduals using Excel (2016) and SigmaPlot software (version 11.0). Significance levels were determined according to Student’s t test: *, P < 0.05; **, P < 0.01; and ns, not significant.

Primers

The nucleotide sequences of all primers used for vector construction, as well as PCR and RT-qPCR analyses, are provided in Supplemental Table S3.

Tao et al.


https://www.ncbi.nlm.nih.gov/




Accession Numbers

Accession numbers are as follows: OsSPMS1 (LOC_Os06g33710), Os-SPMS2 (LOC_Os02g15550), OsSPMS3 (LOC_Os07g22600), and OsACL5 (LOC_Os02g14190).

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. Phylogenetic tree of green plant genes encoding proteins with the Spermine_synth domain.

Supplemental Figure S2. Sequence alignment of Arabidopsis and rice pro-teins with the Spermine_synth domain.

Supplemental Figure S3. Plant morphology of wild-type and transgenic plants at the seedling and tillering stages.

Supplemental Figure S4. Comparison of the cell length and cell width of the second internodes from the top in wild-type and transgenic plants.

Supplemental Figure S5. Effects of different concentrations of ACC treat-ment on rice plant growth.

Supplemental Figure S6. Comparison of root length in wild-type and transgenic plants treated with 1-MCP.

Supplemental Figure S7. Effects of different concentrations of polyamines on rice plant growth.

Supplemental Figure S8. Transcriptome analysis using young wild-type, RNAi, and OE2 panicles.

Supplemental Figure S9. Protein sequences of OsSPMS1 in wild-type, osspms1#1, and osspms1#2 plants.

Supplemental Figure S10. Time course of germination of wild-type, osspms1#1, and osspms1#2 seeds.

Supplemental Figure S11. Protein-protein interactions between OsSPMS1 and OsSPMS2.

Supplemental Table S1. Summary of 62 genes encoding proteins with the Spermine_synth domain in 14 representative green plant genomes.

Supplemental Table S2. Sequence identities of proteins with the Sper-mine_synth domain in Arabidopsis and rice.

Supplemental Table S3. List of primers used in this study.

ACKNOWLEDGMENTS

We thank Chen Chen for suggestions for the improvement of the article.

Received August 9, 2018; accepted August 16, 2018; published September 6, 2018.

LITERATURE CITED

Alcazar R, Cuevas JC, Patron M, Altabella T, Tiburcio AF (2006a) Abscisic acid modulates polyamine metabolism under water stress in Arabidopsis thaliana. Physiol Plant 128: 448–455

Alcázar R, Marco F, Cuevas JC, Patron M, Ferrando A, Carrasco P, Tiburcio AF, Altabella T (2006b) Involvement of polyamines in plant response to abiotic stress. Biotechnol Lett 28: 1867–1876

Apelbaum A, Burgoon AC, Anderson JD, Lieberman M (1981) Polyamines inhibit biosynthesis of ethylene in higher plant tissue and fruit protoplasts. Plant Physiol 68: 453–456

Bitrián M, Zarza X, Altabella T, Tiburcio AF, Alcázar R (2012) Polyamines un-der abiotic stress: metabolic crossroads and hormonal crosstalks in plants. Metabolites 2: 516–528

Chen T, Xu Y, Wang J, Wang Z, Yang J, Zhang J (2013a) Polyamines and eth-ylene interact in rice grains in response to soil drying during grain filling. J Exp Bot 64: 2523–2538

Chen Y, Xu Y, Luo W, Li W, Chen N, Zhang D, Chong K (2013b) The F-box protein OsFBK12 targets OsSAMS1 for degradation and affects pleiotropic phenotypes, including leaf senescence, in rice. Plant Physiol 163: 1673–1685

Clark K, Niemand J, Reeksting S, Smit S, van Brummelen AC, Williams M, Louw AI, Birkholtz L (2010) Functional consequences of perturbing poly-amine metabolism in the malaria parasite, Plasmodium falciparum. Amino Acids 38: 633–644

Colotti G, Ilari A (2011) Polyamine metabolism in Leishmania: from arginine to trypanothione. Amino Acids 40: 269–285

Dufe VT, Lüersen K, Eschbach ML, Haider N, Karlberg T, Walter RD, Al‑Karadaghi S (2005) Cloning, expression, characterisation and three- dimensional structure determination of Caenorhabditis elegans spermidine synthase. FEBS Lett 579: 6037–6043

Eisenberg T, Knauer H, Schauer A, Büttner S, Ruckenstuhl C, Carmona‑ Gutierrez D, Ring J, Schroeder S, Magnes C, Antonacci L, (2009) Induction of autophagy by spermidine promotes longevity. Nat Cell Biol 11: 1305–1314

Eom JS, Choi SB, Ward JM, Jeon JS (2012) The mechanism of phloem loading in rice (Oryza sativa). Mol Cells 33: 431–438

Franceschetti M, Fornalé S, Tassonia A, Zuccherelli K, Mayer MJ, Bagni N (2004) Effects of spermidine synthase overexpression on polyamine biosyn-thetic pathway in tobacco plants. J Plant Physiol 161: 989–1001

Fujita M, Fujita Y, Iuchi S, Yamada K, Kobayashi Y, Urano K, Kobayashi M, Yamaguchi‑Shinozaki K, Shinozaki K (2012) Natural variation in a poly-amine transporter determines paraquat tolerance in Arabidopsis. Proc Natl Acad Sci USA 109: 6343–6347

Hamasaki‑Katagiri N, Katagiri Y, Tabor CW, Tabor H (1998) Spermine is not essential for growth of Saccharomyces cerevisiae: identification of the SPE4 gene (spermine synthase) and characterization of a spe4 deletion mutant. Gene 210: 195–201

Hattori Y, Nagai K, Furukawa S, Song XJ, Kawano R, Sakakibara H, Wu J, Matsumoto T, Yoshimura A, Kitano H, (2009) The ethylene response fac-tors SNORKEL1 and SNORKEL2 allow rice to adapt to deep water. Nature 460: 1026–1030

Hewezi T, Howe PJ, Maier TR, Hussey RS, Mitchum MG, Davis EL, Baum TJ (2010) Arabidopsis spermidine synthase is targeted by an effector pro-tein of the cyst nematode Heterodera schachtii. Plant Physiol 152: 968–984

Hibi N, Higashiguchi S, Hashimoto T, Yamada Y (1994) Gene expression in tobacco low-nicotine mutants. Plant Cell 6: 723–735

Howard TP, Fryer MJ, Singh P, Metodiev M, Lytovchenko A, Obata T, Fernie AR, Kruger NJ, Quick WP, Lloyd JC, (2011) Antisense suppression of the small chloroplast protein CP12 in tobacco alters carbon partitioning and severely restricts growth. Plant Physiol 157: 620–631

Hoyt MA, Broun M, Davis RH (2000a) Polyamine regulation of ornithine de-carboxylase synthesis in Neurospora crassa. Mol Cell Biol 20: 2760–2773

Hoyt MA, Williams‑Abbott LJ, Pitkin JW, Davis RH (2000b) Cloning and expression of the S-adenosylmethionine decarboxylase gene of Neurospora crassa and processing of its product. Mol Gen Genet 263: 664–673

Jun SH, Han MJ, Lee S, Seo YS, Kim WT, An G (2004) OsEIN2 is a positive component in ethylene signaling in rice. Plant Cell Physiol 45: 281–289

Kasukabe Y, He L, Nada K, Misawa S, Ihara I, Tachibana S (2004) Overex-pression of spermidine synthase enhances tolerance to multiple environ-mental stresses and up-regulates the expression of various stress-regulated genes in transgenic Arabidopsis thaliana. Plant Cell Physiol 45: 712–722

Knott JM, Römer P, Sumper M (2007) Putative spermine synthases from Thalassiosira pseudonana and Arabidopsis thaliana synthesize thermosper-mine rather than spermine. FEBS Lett 581: 3081–3086

Korolev S, Ikeguchi Y, Skarina T, Beasley S, Arrowsmith C, Edwards A, Joachimiak A, Pegg AE, Savchenko A (2002) The crystal structure of spermi-dine synthase with a multisubstrate adduct inhibitor. Nat Struct Biol 9: 27–31

Kumar A, Altabella T, Taylor M, Tiburcio A (1997) Recent advances in poly-amine research. Trends Plant Sci 2: 124–130

Kuroha T, Nagai K, Gamuyao R, Wang DR, Furuta T, Nakamori M, Kitaoka T, Adachi K, Minami A, Mori Y, (2018) Ethylene-gibberellin signaling un-derlies adaptation of rice to periodic flooding. Science 361: 181–186

Li J, Hu L, Zhang L, Pan X, Hu X (2015) Exogenous spermidine is enhancing tomato tolerance to salinity-alkalinity stress by regulating chloroplast anti-oxidant system and chlorophyll metabolism. BMC Plant Biol 15: 303

Liu JH, Kitashiba H, Wang J, Ban Y, Moriguchi T (2007) Polyamines and their ability to provide environmental stress tolerance to plants. Plant Biotechnol 24: 117–126

Liu M, Chu M, Ding Y, Wang S, Liu Z, Tang S, Ding C, Li G (2015) Exogenous spermidine alleviates oxidative damage and reduce yield loss in rice sub-merged at tillering stage. Front Plant Sci 6: 919



















Lizada MC, Yang SF (1979) A simple and sensitive assay for 1-aminocyclopro-pane-1-carboxylic acid. Anal Biochem 100: 140–145

Locke JM, Bryce JH, Morris PC (2000) Contrasting effects of ethylene percep-tion and biosynthesis inhibitors on germination and seedling growth of barley (Hordeum vulgare L.). J Exp Bot 51: 1843–1849

Ma B, He SJ, Duan KX, Yin CC, Chen H, Yang C, Xiong Q, Song QX, Lu X, Chen HW, (2013) Identification of rice ethylene-response mutants and characterization of MHZ7/OsEIN2 in distinct ethylene response and yield trait regulation. Mol Plant 6: 1830–1848

Michael AJ (2016) Biosynthesis of polyamines and polyamine-containing mol-ecules. Biochem J 473: 2315–2329

Michael AJ (2017) Evolution of biosynthetic diversity. Biochem J 474: 2277–2299

Miller‑Fleming L, Olin‑Sandoval V, Campbell K, Ralser M (2015) Remaining mysteries of molecular biology: the role of polyamines in the cell. J Mol Biol 427: 3389–3406

Moinard C, Cynober L, de Bandt JP (2005) Polyamines: metabolism and im-plications in human diseases. Clin Nutr 24: 184–197

Moschou PN, Delis ID, Paschalidis KA, Roubelakis‑Angelakis KA (2008) Transgenic tobacco plants overexpressing polyamine oxidase are not able to cope with oxidative burst generated by abiotic factors. Physiol Plant 133: 140–156

Mulangi V, Chibucos MC, Phuntumart V, Morris PF (2012a) Kinetic and phylogenetic analysis of plant polyamine uptake transporters. Planta 236: 1261–1273

Mulangi V, Phuntumart V, Aouida M, Ramotar D, Morris P (2012b) Func-tional analysis of OsPUT1, a rice polyamine uptake transporter. Planta 235: 1–11

Nambeesan S, AbuQamar S, Laluk K, Mattoo AK, Mickelbart MV, Ferruzzi MG, Mengiste T, Handa AK (2012) Polyamines attenuate ethylene-medi-ated defense responses to abrogate resistance to Botrytis cinerea in tomato. Plant Physiol 158: 1034–1045

Nguyen LT, Schmidt HA, von Haeseler A, Minh BQ (2015) IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol 32: 268–274

Panicot M, Minguet EG, Ferrando A, Alcázar R, Blázquez MA, Carbonell J, Altabella T, Koncz C, Tiburcio AF (2002) A polyamine metabolon in-volving aminopropyl transferase complexes in Arabidopsis. Plant Cell 14: 2539–2551

Pegg AE, Michael AJ (2010) Spermine synthase. Cell Mol Life Sci 67: 113–121Schoberer J, Liebminger E, Botchway SW, Strasser R, Hawes C (2013)

Time-resolved fluorescence imaging reveals differential interactions of N-glycan processing enzymes across the Golgi stack in planta. Plant Physi-ol 161: 1737–1754

Seiler N, Delcros JG, Moulinoux JP (1996) Polyamine transport in mammali-an cells: an update. Int J Biochem Cell Biol 28: 843–861

Tabor CW, Tabor H (1984) Polyamines. Annu Rev Biochem 53: 749–790Vivó M, de Vera N, Cortés R, Mengod G, Camón L, Martínez E (2001) Poly-

amines in the basal ganglia of human brain: influence of aging and degen-erative movement disorders. Neurosci Lett 304: 107–111

Waduwara‑Jayabahu I, Oppermann Y, Wirtz M, Hull ZT, Schoor S, Plotnikov AN, Hell R, Sauter M, Moffatt BA (2012) Recycling of methylthioadenos-ine is essential for normal vascular development and reproduction in Ara-bidopsis. Plant Physiol 158: 1728–1744

Wang Q, Zhang W, Yin Z, Wen CK (2013) Rice CONSTITUTIVE TRIPLE- RESPONSE2 is involved in the ethylene-receptor signalling and regulation of various aspects of rice growth and development. J Exp Bot 64: 4863–4875

Wang ZQ, Zhang H, Wang XM, Zhang ZC, Yang JC (2007) Relationship be-tween concentrations of polyamines in filling grains and rice quality. Zuo Wu Xue Bao 33: 1922–1927

Weingartner M, Subert C, Sauer N (2011) LATE, a C2H2 zinc-finger protein that acts as floral repressor. Plant J 68: 681–692

Weiss MC, Sousa FL, Mrnjavac N, Neukirchen S, Roettger M, Nelson‑Sathi S, Martin WF (2016) The physiology and habitat of the last universal com-mon ancestor. Nat Microbiol 1: 16116

Willert E, Phillips MA (2012) Regulation and function of polyamines in Afri-can trypanosomes. Trends Parasitol 28: 66–72

Wu H, Min J, Zeng H, McCloskey DE, Ikeguchi Y, Loppnau P, Michael AJ, Pegg AE, Plotnikov AN (2008) Crystal structure of human spermine syn-thase: implications of substrate binding and catalytic mechanism. J Biol Chem 283: 16135–16146

Wuriyanghan H, Zhang B, Cao WH, Ma B, Lei G, Liu YF, Wei W, Wu HJ, Chen LJ, Chen HW, (2009) The ethylene receptor ETR2 delays floral transi-tion and affects starch accumulation in rice. Plant Cell 21: 1473–1494

Yang C, Ma B, He SJ, Xiong Q, Duan KX, Yin CC, Chen H, Lu X, Chen SY, Zhang JS (2015) MAOHUZI6/ETHYLENE INSENSITIVE3-LIKE1 and ETH-YLENE INSENSITIVE3-LIKE2 regulate ethylene response of roots and cole-optiles and negatively affect salt tolerance in rice. Plant Physiol 169: 148–165

Yang J, Zhang J, Wang Z, Liu K, Wang P (2006) Post-anthesis development of inferior and superior spikelets in rice in relation to abscisic acid and eth-ylene. J Exp Bot 57: 149–160

Yang J, Zhang J, Liu K, Wang Z, Liu L (2007) Involvement of polyamines in the drought resistance of rice. J Exp Bot 58: 1545–1555

Yang J, Yunying C, Zhang H, Liu L, Zhang J (2008) Involvement of polyam-ines in the post-anthesis development of inferior and superior spikelets in rice. Planta 228: 137–149

Ye H, Feng J, Zhang L, Zhang J, Mispan MS, Cao Z, Beighley DH, Yang J, Gu XY (2015) Map-based cloning of Seed Dormancy1-2 identified a gibberellin synthesis gene regulating the development of endosperm-imposed dor-mancy in rice. Plant Physiol 169: 2152–2165

Yin L, Wang S, Tanaka K, Fujihara S, Itai A, Den X, Zhang S (2016) Sili-con-mediated changes in polyamines participate in silicon-induced salt tolerance in Sorghum bicolor L. Plant Cell Environ 39: 245–258

Yoon SO, Lee YS, Lee SH, Cho YD (2000) Polyamine synthesis in plants: iso-lation and characterization of spermidine synthase from soybean (Glycine max) axes. Biochim Biophys Acta 1475: 17–26

Zhou Y, Zhu J, Li Z, Yi C, Liu J, Zhang H, Tang S, Gu M, Liang G (2009) Deletion in a quantitative trait gene qPE9-1 associated with panicle erect-ness improves plant architecture during rice domestication. Genetics 183: 315–324

Tao et al.



Documents

The Spermine Synthase OsSPMS1 Regulates Seed ......of polyamines and their products, plants can effec-tively respond to environmental changes and thereby enhance their stress resistance,