The Ubiquitin Receptor DA1 Regulates Seed and Organ Sizeby Modulating the Stability of the Ubiquitin-Specific ProteaseUBP15/SOD2 in ArabidopsisW

Liang Du,a,b,1 Na Li,a,1 Liangliang Chen,a,b Yingxiu Xu,a,c Yu Li,a,b Yueying Zhang,a,b Chuanyou Li,c and Yunhai Lia,2

a State Key Laboratory of Plant Cell and Chromosome Engineering, Institute of Genetics and Developmental Biology, ChineseAcademy of Sciences, Beijing 100101, ChinabUniversity of Chinese Academy of Sciences, Beijing 100039, Chinac State Key Laboratory of Plant Genomics, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing100101, China

ORCID ID: 0000-0002-0025-4444 (Y.H.L.)

Although the control of organ size is a fundamental question in developmental biology, little is known about the genetic andmolecular mechanisms that determine the final size of seeds in plants. We previously demonstrated that the ubiquitinreceptor DA1 acts synergistically with the E3 ubiquitin ligases DA2 and ENHANCER1 OF DA1 (EOD1)/BIG BROTHER torestrict seed growth in Arabidopsis thaliana. Here, we describe UBIQUITIN-SPECIFIC PROTEASE15 (UBP15), encoded bySUPPRESSOR2 OF DA1 (SOD2), which acts maternally to regulate seed size by promoting cell proliferation in the integumentsof ovules and developing seeds. The sod2/ubp15 mutants form small seeds, while overexpression of UBP15 increases seedsize of wild-type plants. Genetic analyses indicate that UBP15 functions antagonistically in a common pathway with DA1 toinfluence seed size, but does so independently of DA2 and EOD1. Further results reveal that DA1 physically associates withUBP15 in vitro and in vivo and modulates the stability of UBP15. Therefore, our findings establish a genetic and molecularframework for the regulation of seed size by four ubiquitin-related proteins DA1, DA2, EOD1, and UBP15 and suggest thatthey are promising targets for increasing seed size in crops.

INTRODUCTION

Seed size in flowering plants is crucial for evolutionary fitnessand is also an important agronomic trait (Alonso-Blanco et al.,1999; Moles et al., 2005; Fan et al., 2006; Song et al., 2007; Orsiand Tanksley, 2009; Gegas et al., 2010; Linkies et al., 2010). Inflowering plants, a seed consists of three major components:the embryo, the endosperm, and the seed coat, each withdifferent genetic compositions. Seed development involvesa double-fertilization event. One sperm nucleus fertilizes the eggto produce the diploid embryo, while the other sperm nucleusfertilizes the central cell to form the triploid endosperm (Lopesand Larkins, 1993; Reiser and Fischer, 1993). The outermostlayer of the seed is seed coat, which differentiates from maternalinteguments. Therefore, the size of a seed is regulated by thecoordinated growth of the embryo, endosperm, and maternaltissue.

Several factors that act maternally to regulate seed size havebeen described in Arabidopsis thaliana. For example, KLUH/CYTOCHROME P450 78A5 (CYP78A5) influences seed sizeby promoting cell proliferation in the maternal integuments

(Adamski et al., 2009), while AUXIN RESPONSE FACTOR2(ARF2)/MEGAINTEGUMENTA affects seed growth by restrictingcell proliferation in the integuments (Schruff et al., 2006). Bycontrast, TRANSPARENT TESTA GLABRA2 (TTG2) promotesseed growth by increasing cell elongation in the maternal in-teguments (Garcia et al., 2005). APETALA2 (AP2) may repressseed growth by restricting cell elongation in the integuments(Jofuku et al., 2005; Ohto et al., 2005, 2009). Overexpression ofEOD3/CYP78A6 promotes both cell proliferation and cell elon-gation in the maternal integuments, resulting in large seeds(Fang et al., 2012). Seed size is also influenced by zygotic tis-sues. Mutations in HAIKU (IKU) or MINISEED3 (MINI3) reduceseed size due to precocious cellularization of the endosperm(Garcia et al., 2003; Luo et al., 2005; Wang et al., 2010). SHORTHYPOCOTYL UNDER BLUE1 (SHB1) promotes seed growth byincreasing endosperm proliferation (Zhou et al., 2009). SHB1associates with the promoters of MINI3 and IKU2, and this as-sociation requires MINI3 (Kang et al., 2013). A recent study hasshown that cytokinin acts downstream of the IKU pathway toregulate seed size (Li et al., 2013). In addition, the endospermgrowth is influenced by epigenetic mechanisms associated withparent-of-origin effects (Scott et al., 1998; Xiao et al., 2006).The ubiquitin pathway has been known to play an important

part in seed size determination in plants. The ubiquitin receptorDA1 acts maternally to affect seed growth by restricting cellproliferation in the integuments (Li et al., 2008; Xia et al., 2013).In a screen for modifiers of da1-1, we identified enhancer of da1-1(eod1), which has a mutation in the E3 ubiquitin ligase BIGBROTHER (BB) (Disch et al., 2006; Li et al., 2008). DA1 functions

1 These authors contributed equally to this work.2 Address correspondence to yhli@genetics.ac.cn.The author responsible for distribution of materials integral to the findingspresented in this article in accordance with the policy described in theInstructions for Authors (www.plantcell.org) is: Yunhai Li (yhli@genetics.ac.cn).W Online version contains Web-only data.www.plantcell.org/cgi/doi/10.1105/tpc.114.122663

The Plant Cell, Vol. 26: 665–677, February 2014, www.plantcell.org ã 2014 American Society of Plant Biologists. All rights reserved.

synergistically with EOD1 to restrict seed growth (Li et al., 2008).Similarly, we recently reported that DA1 acts synergistically withanother E3 ubiquitin ligase DA2 to regulate seed growth by re-stricting cell division in the maternal integuments (Xia et al., 2013).Interestingly, DA2 shares high similarity with rice (Oryza sativa)GRAIN WIDTH AND WEIGHT2 (GW2), encoded by a quantitativetrait locus for grain size, which influences grain growth by re-stricting cell proliferation (Song et al., 2007; Xia et al., 2013). An-other quantitative trait locus, SEEDWIDTH ON CHROMOSOME 5(qSW5/GW5), affects grain width by limiting cell proliferation inrice (Shomura et al., 2008; Weng et al., 2008). The qSW5/GW5gene encodes an unknown protein that interacts physically withpolyubiquitin, suggesting that qSW5/GW5 might be related to theubiquitin-proteasome protein degradation pathway (Weng et al.,2008). However, the target substrates of these ubiquitin-relatedregulators remain to be discovered.

To identify the downstream targets of DA1, we isolated sup-pressors of the large-seed phenotype of da1-1. Here, we de-scribe sod2 mutants, which suppress the seed size phenotypeof da1-1. SOD2 encodes the ubiquitin-specific protease UBP15,which has known roles in leaf development (Liu et al., 2008) butno previously identified function in seed size determination.UBP15 influences seed size by promoting cell proliferation inthe maternal integuments. Genetic analyses show that SOD2/UBP15 functions in a common pathway with the ubiquitin re-ceptor DA1 to regulate seed growth. We further demonstratethat DA1 physically interacts with UBP15 and modulates thestability of UBP15. Thus, our findings define the genetic andmolecular mechanisms of four ubiquitin-related proteins, DA1,DA2, EOD1, and UBP15, in seed size determination.

RESULTS

The sod2 Mutations Suppress the Seed Size Phenotypeof da1-1

To identify other components in the DA1 pathway, we initiateda genetic screen for modifiers of da1-1 in seed size. From thisscreen, we identified several suppressors of da1-1 (sod) fromethyl methanesulfonate–treated M2 populations of da1-1. Wedesignated two of these suppressors sod2-1 and sod2-2, whichare allelic to each other. The sod2 da1-1 seeds were smaller andlighter than da1-1 seeds (Figures 1A to 1C and 1J). In addition,sod2 da1-1 plants produced small flowers and leaves comparedwith da1-1 plants (Figures 1D to 1I, 1K and 1L). The size of cellsin sod2 da1-1 petals and leaves was similar to that in wild-typepetals and leaves (Figures 1K and 1L), indicating that mutationsin SOD2 cause a decrease in cell number. Taken together, theseresults show that the sod2 mutations suppressed the seed andorgan size phenotypes of da1-1.

SOD2 Encodes the Ubiquitin-Specific Protease UBP15

The sod2-1 mutation was identified by map-based cloning in anF2 population of a cross between sod2-1 da1-1 and da1-1Ler. Thesod2-1 mutation was mapped into an interval between markersF19K19-122 and F11A6-129 on chromosome I (SupplementalFigure 1A). This mapping region includes UBIQUITIN-SPECIFIC

PROTEASE15 (UBP15; At1g17110), which is known to influenceleaf growth in Arabidopsis (Liu et al., 2008), yet no function for it indetermination of seed size has been described. UBP15 containsa signature MYND-type zinc finger domain (Zf-MYND), which hasbeen suggested to be a protein–protein interaction domain inmammalian cells (Lutterbach et al., 1998), and a ubiquitin-specificprotease (UBP) domain that is required for deubiquitination ac-tivity (Figure 2B) (Liu et al., 2008). To define the molecular basis forthe sod2 mutations, the At1g17110 genomic DNA sequences oftwo sod2 mutant alleles were amplified by PCR and sequenced.We identified a single mutation in each of the mutant alleles. Insod2-1, the 59-intron-exon boundary of intron 4 is changed fromG to A (Figure 2A; Supplemental Figure 1B), replacing a G that ishighly conserved at plant gene splice sites. Similarly, sod2-2 hasa G-to-A transition in the 59-intron-exon boundary of the last intron(Figure 2A; Supplemental Figure 1B). Mutations in sod2 allelesaltered the splicing of At1g17110 mRNA (Supplemental Figures 1Cand 1D).To further investigate whether mutations in At1g17110/

UBP15 could suppress the seed and organ size phenotypes ofda1-1, we obtained the ubp15-1 mutant harboring a T-DNA in-sertion in the At1g17110 gene (Liu et al., 2008) and generated theubp15-1 da1-1 double mutant. Like sod2 mutations, the ubp15-1mutation suppressed the seed and organ size phenotypes ofda1-1 (Supplemental Figure 2), suggesting that UBP15 is theSOD2 gene. We further transformed sod2-2 da1-1 plants witha genomic fragment (gUBP15) that includes 2074-bp promoterand the At1g17110 gene. Most transgenic plants exhibited theseed and organ size phenotypes of da1-1 (Figures 2C to 2G),indicating that At1g17110/UBP15 is the SOD2 gene.To evaluate the expression of UBP15, we generated transgenic

Arabidopsis plants containing a UBP15 promoter:b-glucuronidasefusion (pUBP15:GUS). The tissue-specific expression patterns ofUBP15 were investigated using a histochemical assay for GUSactivity. In seedlings, GUS activity was detected in cotyledons,leaves, and roots (Figure 2H). In flowers, GUS activity was ob-served in sepals, petals, stamens, and carpels (Figures 2I to 2L).GUS activity was stronger in younger floral organs than older floralorgans (Figures 2J to 2L). Similarly, relatively higher GUS activitywas observed in younger ovules than older ones (Figures 2M to2O). Thus, these results indicate that UBP15 expression is tem-porally and spatially regulated.

UBP15 Acts Maternally to Regulate Seed Size

The ubp15-1 single mutant formed small leaves and flowerscompared with the wild type (Supplemental Figure 3) (Liu et al.,2008). Considering that the sod2/ubp15 mutations suppressedthe seed size phenotype of da1-1 (Figures 2E and 2F;Supplemental Figure 2), we asked whether the ubp15 singlemutant influences seed size. We therefore measured the sizeand weight of wild-type and ubp15-1 seeds. As shown in Figures3A and 3E, seeds from ubp15-1 plants were significantlysmaller and lighter than wild-type seeds. The embryo con-stitutes the major volume of a mature seed in Arabidopsis. Theubp15-1 embryos were obviously smaller than wild-type em-bryos (Figure 3B). The changes in seed size were also reflectedin the size of seedlings (Figure 3C). Cotyledons of ubp15-1

666 The Plant Cell

seedlings were clearly smaller than those of wild-type seed-lings (Figures 3C and 3F). Thus, these results indicate thatUBP15 is required for seed size control.

As the size of a seed is controlled by the zygotic and/or ma-ternal tissues, we investigated whether UBP15 acts zygoticallyor maternally. To test this, we conducted reciprocal cross ex-periments between the wild type and ubp15-1. We pollinatedubp15-1 plants with wild-type pollen. The size of seeds fromsuch crosses was comparable to that from self-pollinatedubp15-1 mutant (Figure 3G). The size of seeds from wild-typeplants pollinated by ubp15-1 mutant pollen was similar to thatfrom the self-pollinated wild-type plants (Figure 3G). These re-sults indicate that UBP15 functions maternally to regulate seedsize. We further investigated the size of ubp15-1/Columbia-0(Col-0) F2 and Col-0/ubp15-1 F2 seeds. Seeds from ubp15-1/Col-0 F1 and Col-0/ubp15-1 F1 plants have wild-type, ubp15-1,and ubp15/+ embryos within ubp15/+ seed coats, respectively.The size of ubp15-1/Col-0 F2 and Col-0/ubp15-1 F2 seeds wassimilar to that of wild-type seeds (Figure 3G). Thus, these resultsshow that the embryo and endosperm genotypes for UBP15 donot influence seed size, and UBP15 is required in sporophytictissue of the mother plant to determine seed size.

UBP15 Regulates Cell Proliferation in theMaternal Integuments

The reciprocal cross experiments demonstrated that UBP15acts maternally to influence seed growth (Figure 3G). The ovuleinteguments are maternal tissues and develop into the seedcoat after fertilization, which could physically limit seed growth(Adamski et al., 2009; Fang et al., 2012). Consistent with thisnotion, several previous studies indicated that the integument

size of ovules influences the final size of seeds in Arabidopsis(Schruff et al., 2006; Adamski et al., 2009; Fang et al., 2012). Wetherefore examined mature ovules from wild-type and ubp15-1plants 2 d after emasculation. The ubp15-1 ovules were obvi-ously smaller than wild-type ovules (Figure 3D). The outer in-tegument length of ubp15-1 ovules was significantly decreased,compared with that of wild-type ovules (Figure 3H).The size of the integument is determined by cell number and

cell size. We therefore investigated the number of outer in-tegument cells in wild-type and ubp15-1 ovules. As shown inFigure 3I, the number of outer integument cells in ubp15-1 ov-ules was significantly reduced compared with that in wild-typeovules, indicating that UBP15 is required for cell proliferation inthe maternal integuments of ovules. After fertilization, cells in theinteguments predominantly carry on expansion but still undergodivision (Garcia et al., 2005). We further counted the number ofouter integument cells in wild-type and ubp15-1 seeds at 6 and8 d after pollination (DAP). In wild-type seeds, the outer in-tegument cell number at 6 DAP was comparable with that at 8DAP (Figure 3I), indicating that cells in the outer integuments ofwild-type seeds completely cease division by 6 DAP. In ubp15seeds, the outer integument cell number at 6 DAP was similar tothat at 8 DAP, showing that cells in the outer integuments ofubp15-1 seeds also stop dividing by 6 DAP (Figure 3I). As shownin Figure 3I, the number of outer integument cells in ubp15seeds was clearly decreased, compared with that in wild-typeseeds. We further measured the length of outer integument cellsin wild-type and ubp15-1 seeds at 6 and 8 DAP. Cells in ubp15-1outer integuments were slightly longer than those in wild-typeouter integuments (Figure 3J), suggesting a possible compen-sation mechanism between cell number and cell size in thematernal integuments. Thus, these results indicate that UBP15

Figure 1. sod2 Suppresses the Seed Size Phenotype of da1-1.

(A) to (C) Seeds of Col-0 (A), da1-1 (B), and sod2-1 da1-1 (C).(D) to (F) Ten-day-old seedlings of Col-0 (D), da1-1 (E), and sod2-1 da1-1 (F).(G) to (I) Flowers of Col-0 (G), da1-1 (H), and sod2-1 da1-1 (I).(J) Seed area (SA) and seed weight (SW) of Col-0, da1-1, sod2-1 da1-1, and sod2-2 da1-1 plants.(K) Fifth leaf area (LA) and leaf palisade cell area (LCA) of Col-0, da1-1, sod2-1 da1-1, and sod2-2 da1-1 plants.(L) Petal area (PA), petal length (PL), petal width (PW), and petal cell area (PCA) of Col-0, da1-1, sod2-1 da1-1, and sod2-2 da1-1 plants.Values in (J) to (L) are given as mean6 SE relative to the respective wild-type values, set at 100%. **P < 0.01 compared with the da1-1 (Student’s t test).Bars = 0.5 mm in (A) to (C) and 1 mm in (D) to (I).

Ubiquitin-Mediated Control of Seed Size 667

is required for cell proliferation in the maternal integuments ofovules and developing seeds.

Plants Overexpressing UBP15 Show Similar Phenotypesto da1-1

We transformed Col-0 plants with a genomic fragment(gUBP15). Interestingly, some transgenic plants had dramaticincreases in UBP15 mRNA compared with wild-type plants

(Figure 4D). It is possible that the integration regions of thetransgenes in the genome affect the promoter activity ofgUBP15. gUBP15 transgenic plants formed larger and heavierseeds than wild-type plants (Figures 4A and 4E). gUBP15transgenic plants also produced larger flowers and leaves thanwild-type plants (Figures 4B, 4C, 4F. and 4G). The increased sizeof petals and leaves in gUBP15 transgenic plants was notcaused by larger cells (Figures 4F and 4G), indicating that it isthe number of petal and leaf cells that is higher. We previously

Figure 2. Identification and Molecular Characterization of the SOD2/UBP15 Gene.

(A) The SOD2/UBP15 gene structure. The start codon (ATG) and the stop codon (TAG) are indicated. Closed boxes indicate the coding sequence, openboxes indicate the 59 and 39 untranslated regions, and lines between boxes indicate introns. The mutation sites of sod2-1 and sod2-2 and the T-DNAinsertion site in ubp15-1 are shown.(B) The SOD2 protein contains a Zf-MYND and a UBP domain.(C) Plants (21-d-old) of da1-1, sod2-2 da1-1, COM#1, and COM#2 (from left to right). COM is sod2-2 da1-1 transformed with the genomic sequence ofthe UBP15 gene (gUBP15).(D) Flowers of da1-1, sod2-2 da1-1, COM#1, and COM#2 (from left to right).(E) Seeds of da1-1, sod2-2 da1-1, COM#1, and COM#2 (from left to right).(F) Seed area (SA) and seed weight (SW) of da1-1, sod2-2 da1-1, COM#1, and COM#2.(G) Petal area (PA), petal length (PL) and petal width (PW) of da1-1, sod2-2 da1-1, COM#1, and COM#2.(H) to (O) UBP15 expression activity was monitored by pUBP15:GUS transgene expression. Six GUS-expressing lines were observed, and all showeda similar pattern, although they differed slightly in the intensity of the staining. Histochemical analysis of GUS activity in an 8-d-old seedling (H), thedeveloping sepals (I), the developing petals (J), the developing stamens (K), the developing carpels (L), and the developing ovules ([M] to [O]).Values in (F) and (G) are given as mean 6 SE relative to the respective da1-1 values, set at 100%. **P < 0.01 compared with the sod2-2 da1-1 (Student’st test). Bars = 5 cm in (C), 1 mm in (D) and (H), 0.5 mm in (E), and 0.1 mm in (I) to (O).

668 The Plant Cell

reported that the da1-1 mutant formed large seeds, flowers, andleaves by promoting cell proliferation (Li et al., 2008). Thus, theseed and organ size phenotypes of gUBP15 transgenic plantswere similar to those observed in da1-1 plants (Li et al., 2008).

UBP15 Acts in a Common Pathway with DA1 to Affect SeedSize, but Does so Independently of DA2 and EOD1

As sod2/ubp15 mutations suppressed the seed size phenotypeof da1-1, we sought to determine genetic relationships between

UBP15 and DA1 in seed size. We therefore measured the sizeof seeds from wild-type, ubp15-1, da1-1, and ubp15-1 da1-1plants. As shown in Figures 5A and 5C, the average size andweight of ubp15-1 da1-1 seeds were indistinguishable fromthose of ubp15-1 single mutant seeds, indicating that ubp15-1is epistatic to da1-1 with respect to seed size and weight. Thechanges in seed size were reflected in the size of the embryos(Figure 5B). The size of ubp15-1 da1-1 embryos was compa-rable with that of ubp15-1 embryos (Figure 5B). We further in-vestigated mature ovules from wild-type, ubp15-1, da1-1, andubp15-1 da1-1 plants. The outer integument length of ubp15-1da1-1 ovules was similar to that of ubp15-1 ovules (Figure 5D).Similarly, the outer integument length of ubp15-1 da1-1 seedswas indistinguishable from that of ubp15-1 seeds at 6 and 8DAP. In addition, the size of ubp15-1da1-1 flowers and petalswas comparable with that of ubp15-1 flowers and petals(Supplemental Figure 4). Together, these genetic analyses showthat ubp15-1 is epistatic to da1-1 with respect to seed and or-gan size, indicating DA1 and UBP15 function antagonistically ina common pathway to regulate seed and organ growth.To further investigate the cellular basis of epistatic inter-

actions between DA1 and UBP15 in seed size, we counted theouter integument cell number of ovules and developing seeds inthe wild type, ubp15-1, da1-1, and ubp15-1 da1-1. The numberof outer integument cells in ubp15-1 da1-1 ovules was com-parable to that in ubp15-1 ovules (Figure 5E). Similarly, thenumber of outer integument cells in ubp15-1 da1-1 developingseeds was indistinguishable from that in ubp15-1 developingseeds (Figure 5E). These results show that ubp15 is epistatic toda1-1 with respect to the number of outer integument cells.Thus, DA1 and UBP15 act antagonistically in a common path-way to regulate cell proliferation in the maternal integuments.We previously reported that DA1 functions synergistically with

EOD1 and DA2 in the control of seed and organ size in Arabi-dopsis (Li et al., 2008; Xia et al., 2013). We further asked whetherUBP15 acts antagonistically with EOD1 or DA2 in a commonpathway to regulate seed and organ growth. To test this, wegenerated ubp15-1 eod1-2 and ubp15-1 da2-1 double mutantsand investigated their seed and organ size phenotypes. Thegenetic interaction between ubp15-1 and eod1-2 was essen-tially additive for both seed and petal size compared with theirparental lines (Supplemental Figures 5A and 5B). Similarly, theseed and petal size phenotypes of ubp15-1 da2-1 were additivecompared with those of ubp15-1 and da2-1 single mutants(Supplemental Figures 5C and 5D). Therefore, these geneticanalyses suggest that UBP15 acts independently of EOD1 andDA2 to influence seed and organ growth.

DA1 Modulates UBP15 Stability

DA1 encodes a ubiquitin receptor that has been proposed to beinvolved in ubiquitin-mediated protein degradation (Li et al.,2008; Xia et al., 2013). The ubiquitin proteasome system is in-volved in the degradation of many proteins. To test whetherUBP15 could be degraded in a proteasome-dependent manner,we treated the Arabidopsis gUBP15-GFP (green fluorescentprotein) transgenic line with the proteasome inhibitor MG132.After MG132 treatment, the level of UBP15 protein was higher

Figure 3. UBP15 Acts Maternally to Regulate Seed Size.

(A) Seeds of Col-0 (left) and ubp15-1 (right).(B) Mature embryos of Col-0 (left) and ubp15-1 (right).(C) Seven-day-old seedlings of Col-0 (left) and ubp15-1 (right).(D) Mature ovules of Col-0 (left) and ubp15-1 (right).(E) Seed area (SA) and seed weight (SW) of Col-0 and ubp15-1.(F) Cotyledon area (CoA) of Col-0 and ubp15-1.(G) Seed area of Col-0 3 Col-0 F1, ubp15-1 3 ubp15-1 F1, Col-0 3

ubp15-1 F1, ubp15-1 3 Col-0 F1, Col-0 3 ubp15-1 F2, and ubp15-1 3

Col-0 F2.(H) The outer integument length of Col-0 and ubp15-1 at 0, 6, and 8 DAP.Ovules at 0 DAP are mature ovules from wild-type and ubp15-1 plants2 d after emasculation.(I) The number of cells in the outer integuments of Col-0 and ubp15-1 at0, 6, and 8 DAP.(J) The length of cells in the outer integuments of Col-0 and ubp15-1 at 0,6, and 8 DAP.Values in (E) to (J) are given as mean 6 SE relative to the respective wild-type values, set at 100%. **P < 0.01 and *P < 0.05 compared with thewild type (Student’s t test). Bars = 0.5 mm in (A), 0.25 mm in (B), 1 mm in(C), and 0.1 mm in (D).

Ubiquitin-Mediated Control of Seed Size 669

than that in untreated plants (Figure 6A), indicating that theubiquitin proteasome influences the stability of UBP15. Con-sidering that DA1 and UBP15 act antagonistically in a commongenetic pathway to regulate seed and organ growth, we askedwhether DA1 could affect the stability of UBP15. To test this, wecrossed da1-1 with two independent gUBP15-GFP transgeniclines. After the selection of plants containing da1-1 andgUBP15-GFP transgene, we measured and compared the levelsof UBP15 proteins in different genetic backgrounds. As shownin Figure 6B, the UBP15-GFP protein levels in the da1-1 mutantbackground were obviously higher than those in wild-typeplants, while the transcript levels of UBP15 in gUBP15-GFP andgUBP15-GFP;da1-1 plants were similar (Figure 6E). Thus, theseresults indicate that DA1 influences the stability of UBP15 inArabidopsis.As the accumulation of UBP15 was observed in the da1-1

mutant background, one would expect that the da1-1 alleleshould increase the seed and organ size phenotypes ofgUBP15-GFP transgenic plants. We therefore measured theseed and organ size of gUBP15-GFP and gUBP15-GFP/da1-1plants. As shown in Figures 6C, 6D, 6F, and 6G, the seed andorgan size phenotypes of gUBP15-GFP transgenic plants werestrongly enhanced by da1-1. Thus, our genetic analyses furthersupport the role of DA1 in modulating UBP15 stability.

DA1 Interacts with UBP15 in Vitro and in Vivo

Considering that DA1 modulates UBP15 stability, we askedwhether DA1 could physically interact with UBP15 and influenceits degradation. To test this, we performed in vitro pull-downassays. DA1 was expressed as a glutathione S-transferase(GST) fusion protein, while UBP15 was expressed as a maltosebinding protein (MBP) fusion. As shown in Figure 7A, GST-DA1was pulled down with MBP-UBP15, but not with MBP alone,indicating that DA1 physically interacts with UBP15 in vitro.Considering that the da1-1 mutation (DA1R358K) affects UBP15stability, we further asked whether the DA1R358K mutation in-fluences the interaction between DA1 and UBP15. Usinga GST-DA1R358K fusion protein in pull-down experiments withMBP-UBP15, we showed that the DA1R358K mutation does notaffect the interaction between DA1 and UBP15 (Figure 7A).DA1 contains two ubiquitin interaction motifs (UIMs) in the

N terminus, a single LIM (Lin-11, Isl-1, and Mec3) domain, andthe C-terminal region (Figure 7E) (Li et al., 2008). We further in-vestigated which domain of DA1 is responsible for the interactionbetween DA1 and UBP15. The UIM domains (DA1-UIM), the LIMdomain (DA1-LIM), the C-terminal region (DA1-C), and the LIMdomain and C terminus (DA1-LIM+C) were expressed as GST

Figure 4. Plants Overexpressing UBP15 Show Similar Phenotypes to da1-1.

(A) Seeds of Col-0, gUBP15#1, gUBP15#2, and gUBP15#3 plants (fromleft to right). gUBP15 is Col-0 transformed with the genomic sequence ofthe UBP15 gene.(B) Plants (28 d old) of Col-0, gUBP15#1, gUBP15#2, and gUBP15#3(from left to right).(C) Flowers of Col-0, gUBP15#1, gUBP15#2, and gUBP15#3 (from left toright).(D) Quantitative real-time RT-PCR analysis of the UBP15 gene expres-sion in Col-0, gUBP15#1, gUBP15#2, and gUBP15#3 seedlings.(E) Seed area (SA) and seed weight (SW) of Col-0, gUBP15#1,gUBP15#2, and gUBP15#3.

(F) Petal area (PA), petal length (PL), petal width (PW), and petal cell area(PCA) of Col-0, gUBP15#1, gUBP15#2, and gUBP15#3.(G) The average area of the fifth leaves (LA) and palisade cells (LCA) ofCol-0, gUBP15#1, gUBP15#2, and gUBP15#3.Values in (D) to (G) are given as mean6 SE relative to the respective wild-type values, set at 100%. **P < 0.01 compared with the wild type (Stu-dent’s t test). Bars = 0.5 mm in (A), 5 cm in (B), and 1 mm in (C).

670 The Plant Cell

fusion proteins in Escherichia coli (Figure 7E). UBP15 wasexpressed as an MBP fusion protein and used in pull-downexperiments. As shown in Figure 7B, GST-DA1-LIM+C andGST-DA1-C were pulled down with MBP-UBP15, but not withMBP alone. However, GST-DA1-UIM and GST-DA1-LIM didnot bind to MBP-UBP15. Thus, these results show that theC-terminal region of DA1 is required for the interaction withUBP15 in vitro.

UBP15 contains a signature Zf-MYND near the N terminus,which has been proposed to be a protein–protein interactiondomain in mammalian cells (Lutterbach et al., 1998), and a UBPdomain in the C terminus (Figure 7F). We further asked whichregion of UBP15 is responsible for the interaction between DA1and UBP15. The N-terminal region containing Zf-MYND domain(UBP15-N) and the C-terminal region (UBP15-C) were ex-pressed as MBP fusion proteins in E. coli. DA1 was expressedas a GST fusion protein and used in pull-down experiments.As shown in Figure 7C, GST-DA1 was pulled down with MBP-UBP15-C, but not with MBP-UBP15-N or MBP alone. Thus,these results show that the C-terminal region of UBP15 isrequired for the interaction between DA1 and UBP15 in vitro.

To investigate possible association between DA1 andUBP15 in planta, we used coimmunoprecipitation assays todetect their interactions in vivo. We transiently coexpressed35S:GFP-UBP15 and 35S:Myc-DA1 in Nicotiana benthamianaleaves. 35S:GFP coexpressed with 35S:Myc-DA1 was usedas a negative control. Although the majority of GFP-UBP15fusion proteins were degraded, a very weak band corre-sponding to GFP-UBP15 could be detected (Figure 7D). The

coimmunoprecipitation analysis indicated that Myc-DA1 wasdetected in the immunoprecipitated GFP-UBP15 complexbut not in the negative control (GFP) (Figure 7D), revealinga physical association between DA1 and UBP15 in vivo. Wefurther transiently coexpressed 35S:GFP-UBP15-C and 35S:Myc-DA1 in N. benthamiana leaves. The coimmunoprecipita-tion analysis showed that the C-terminal region of UBP15 isrequired for the interaction with DA1 in vivo (SupplementalFigure 6).

DISCUSSION

Seed size in flowering plants is an important trait for plant fit-ness and agricultural purposes (Alonso-Blanco et al., 1999;Orsi and Tanksley, 2009). Although several factors that func-tion maternally to influence seed size were described inArabidopsis, including TTG2, AP2, ARF2, CYP78A5/KLUH,CYP78A6/EOD3, DA1, and DA2, little is known about theirgenetic and molecular mechanisms in seed size determination.We previously demonstrated that the ubiquitin receptor DA1acts synergistically with the E3 ubiquitin ligases DA2 and EOD1to restrict seed growth in Arabidopsis (Li et al., 2008; Xia et al.,2013), but their target substrates remain to be discovered.Here, we show that DA1 physically associates with the ubiquitin-specific protease UBP15/SOD2 and modulates the stability ofUBP15. Further results reveal that UBP15 and DA1 act antago-nistically in a common genetic pathway to regulate seed size.Thus, our findings define the genetic and molecular mechanisms

Figure 5. ubp15-1 Is Epistatic to da1-1.

(A) Seeds of Col-0, ubp15-1, da1-1, and ubp15-1 da1-1.(B) Mature embryos of Col-0, ubp15-1, da1-1, and ubp15-1 da1-1.(C) Seed area (SA) and seed weight (SW) of Col-0, ubp15-1, da1-1, and ubp15-1 da1-1.(D) The outer integument length of Col-0, ubp15-1, da1-1, and ubp15-1 da1-1 at 0, 6, and 8 DAP.(E) The number of cells in the outer integuments of Col-0, ubp15-1, da1-1, and ubp15-1 da1-1 at 0, 6, and 8 DAP.Values in (C) to (E) are given as mean 6 SE relative to the respective wild-type values, set at 100%. **P < 0.01 compared with the wild type (Student’s ttest). Bars = 0.5 mm in (A) and 0.25 mm in (B).

Ubiquitin-Mediated Control of Seed Size 671

of four ubiquitin-related proteins: DA1, DA2, EOD1, and UBP15in seed size control.

UBP15 Functions Maternally to Regulate Seed Size

The sod2/ubp15 mutants were identified as suppressors of thelarge seed and organ phenotypes of da1-1. The ubp15 singlemutants produced small seeds and organs, whereas over-expression of UBP15 resulted in large seeds and organs (Figures3 and 4; Supplemental Figure 3), indicating that UBP15 isa positive regulator of seed and organ size. Several studiessuggested that there is a possible correlation between organsize and seed growth. For example, arf2, da1-1, and da2-1mutants with large organs produced large seeds (Schruff et al.,2006; Li et al., 2008; Xia et al., 2013), while klu mutants withsmall organs formed small seeds (Adamski et al., 2009).However, organ and seed size is not always positively related.For example, several other mutants with altered organ size hadnormal-sized seeds (Horiguchi et al., 2005; White, 2006; Leeet al., 2009; Xu and Li, 2011). Thus, these studies suggest that

organs and seeds may possess both common and distinctmechanisms to determine their respective size.The size of a seed is determined by maternal sporophytic

and/or zygotic tissues. Our reciprocal cross experiments in-dicated that UBP15 functions maternally to promote seedgrowth, and the embryo and endosperm genotypes for UBP15do not determine seed size (Figure 3G). Several previousstudies demonstrated that the maternal integuments are cru-cial for determining seed size in Arabidopsis. For example,the arf2, da1-1, da2-1, and eod3-1D mutants produced largeseeds due to the enlargement of the maternal integuments(Schruff et al., 2006; Li et al., 2008; Fang et al., 2012; Xia et al.,2013), whereas klu mutants formed small seeds as a result ofreduced integument size (Adamski et al., 2009). The ubp15-1integuments were smaller than wild-type integuments (Figures3D and 3H). The ubp15-1 mutation also suppressed the largeintegument size phenotype of da1-1 (Figure 5D). Thus, thesestudies suggest that the regulation of maternal integumentsize is one of the important mechanisms for control of seedsize.

Figure 6. DA1 Modulates UBP15 Stability.

(A) The proteasome inhibitor stabilizes UBP15. Ten-day-old gUBP15-GFP seedlings were treated with or without 20 mM MG132. Total protein extractswere subjected to immunoblot assays using anti-GFP and anti-RPN6 (as loading control) antibodies.(B) The UBP15-GFP protein accumulates at higher levels in the da1-1 mutant. Total protein extracts were subjected to immunoblot assays using anti-GFP and anti-RPN6 (as loading control) antibodies.(C) Seeds of gUBP15-GFP#7 and gUBP15-GFP#7;da1-1.(D) Flowers of gUBP15-GFP#7 and gUBP15-GFP#7;da1-1.(E) The relative expression of UBP15 in gUBP15-GFP#7 and gUBP15-GFP#7;da1-1. Data shown are mean 6 SD of three replicates.(F) Seed area (SA) and seed weight (SW) of gUBP15-GFP#7 and gUBP15-GFP#7;da1-1.(G) Petal area (PA), petal length (PL), and petal width (PW) of gUBP15-GFP#7 and gUBP15-GFP#7;da1-1.Values in (F) and (G) are given as mean 6 SE relative to the respective gUBP15-GFP#7 values, set at 100%. **P < 0.01 compared with gUBP15-GFP#7(Student’s t test). Bars = 0.5 mm in (C) and 1 mm in (D).

672 The Plant Cell

The size of the integument or seed coat is coordinatelydetermined by cell proliferation and cell expansion. The numberof cells in ubp15-1 integuments was less than that in wild-typeinteguments (Figure 3I), indicating that UBP15 is required forcell proliferation in the maternal integuments. By contrast, the

size of cells in ubp15-1 integuments was slightly increasedcompared with that in wild-type integuments (Figure 3J),suggesting a possible compensation mechanism between cellproliferation and cell expansion in maternal integuments. Theinteguments of da1-1 had more cells than those of the wild type

Figure 7. DA1 Physically Interacts with UBP15 in Vitro and in Vivo.

(A) DA1 directly interacts with UBP15 in vitro. GST-DA1 and GST-DA1R358K were pulled down (PD) by MBP-UBP15 immobilized on amylose resin andanalyzed by immunoblotting (IB) using an anti-GST antibody.(B) The C terminus of DA1 interacts with UBP15 in vitro. Both GST-DA1-LIM+C and GST-DA1-C were pulled down by MBP-UBP15 immobilized onamylose resin and analyzed by immunoblotting using an anti-GST antibody.(C) The C-terminal region of UBP15 interacts with DA1 in vitro. GST-DA1 was pulled down by MBP-UBP15-C immobilized on amylose resin andanalyzed by immunoblotting using an anti-GST antibody.(D) UBP15 interacts with DA1 in vivo. N. benthamiana leaves were transformed by injection of Agrobacterium tumefaciens EHA105 cells harboring 35S:GFP-UBP15 and 35S:Myc-DA1 plasmids. Myc-DA1 was detected in the immunoprecipitated GFP-UBP15 complex, indicating that there is a physicalassociation between DA1 and UBP15 in vivo.(E) Schematic diagram of DA1 and its derivatives containing specific domains. The DA1 contains two UIM motifs in the N terminus, a single LIM domain,and the C-terminal region. aa, amino acid.(F) Schematic diagram of UBP15, UBP15-N, and UBP15-C. The UBP15 contains one predicted Zf-MYND domain in the N terminus and the UBPdomain in the C-terminal region.(G) A genetic and molecular framework for DA1, DA2, EOD1/BB, and UBP15/SOD2-mediated regulation of seed and organ size. DA1 and DA2 actsynergistically to restrict seed andorgan size, suggesting that DA1 andDA2may have a commondownstreamsubstrate. Similarly, DA1 and EOD1may sharea common target. However, DA2 acts independently of EOD1 to influence seed and organ size, suggesting that DA2 and EOD1 may target distinct cellproliferation stimulators (substrate 1 and substrate 2) for degradation, with common regulation via DA1. DA1 acts the upstream of UBP15 and modulatesUBP15 stability. However, UBP15 acts independently of DA2 and EOD1, suggesting that UBP15 is not the target of DA2 or EOD1 for degradation.

Ubiquitin-Mediated Control of Seed Size 673

(Xia et al., 2013). Genetic analyses show that ubp15-1 is epi-static to da1-1 with respect to cell number in the integuments(Figure 5E), further supporting the role of UBP15 in cell pro-liferation. Consistent with this, expression of UBP15 in youngerovules was higher than that in older ones (Figures 2H to 2O).Thus, it is possible that the integument cell number maydetermine seed size by setting the growth potential of theseed coat.

A Genetic Framework for DA1, DA2, EOD1, andUBP15-Mediated Control of Seed Size

Several factors involved in ubiquitin-related activities havebeen reported to affect seed and organ growth in plants. Wepreviously identified the ubiquitin receptor DA1 as a negativefactor of seed and organ size (Li et al., 2008). A modifier screenidentified an enhancer of da1-1 (EOD1) (Li et al., 2008), whichencodes the E3 ubiquitin ligase BB (Disch et al., 2006). DA1acts synergistically with EOD1 to restrict seed and organgrowth, suggesting that DA1 and EOD1 might share a com-mon downstream target. DA1 functions synergistically withanother E3 ubiquitin ligase DA2 to influence seed and organgrowth (Li et al., 2008; Xia et al., 2013). However, geneticanalyses show that DA2 and EOD1 act in different pathways toregulate seed and organ size (Li et al., 2008; Xia et al., 2013),suggesting that DA2 and EOD1 may target distinct growthstimulators for degradation, with common regulation via DA1.In this study, we identified two allelic suppressors of da1-1(sod2) mutated in the gene encoding UBP15, which possessesdeubiquitination activity (Liu et al., 2008). With respect to seedand organ size, ubp15-1 is epistatic to da1-1 (Figure 5), andoverexpression of UBP15 caused similar large seed and organphenotypes to da1-1 (Figure 4). In addition, UBP15 and DA1showed similar expression patterns. For example, UBP15 andDA1 were highly expressed during early stages of organ for-mation, but levels were largely reduced at the later stages oforgan development (Figures 2H to 2O) (Li et al., 2008). Thus,these results support that DA1 and UBP15 act antagonisticallyin a common genetic pathway to regulate seed and organgrowth. However, our genetic analyses show that UBP15 actsindependently of EOD1 and DA2 to influence seed and organsize, suggesting that UBP15 is not the substrate of theE3 ubiquitin ligases DA2 or EOD1 for degradation. Thus, ourfindings define a framework for the control of seed and organsize by four ubiquitin-related proteins DA1, DA2, EOD1, andUBP15 in Arabidopsis (Figure 7G).

DA1 Modulates UBP15 Stability

Our genetic analyses showed that DA1 and UBP15 act antag-onistically in a common pathway to control seed and organgrowth (Figure 5). DA1 encodes a ubiquitin receptor with twoUIMs that can bind ubiquitin in vitro (Li et al., 2008). Ubiquitinreceptors have been proposed to interact with polyubiquitinatedsubstrates via their UIM domains and facilitate substratedegradation by the proteasome (Verma et al., 2004). In thisstudy, we showed that UBP15 stability is affected by theproteasome (Figure 6A). A recent proteomic study has shownthat UBP15 is a ubiquitination target (Kim et al., 2013). Thus, it is

possible that DA1 could be involved in mediating the degrada-tion of UBP15 by the proteasome. Supporting this, the level ofUBP15 protein was obviously increased in the da1-1 mutantcompared with that in the wild type (Figure 6B). We further re-vealed that DA1 physically associates with UBP15 in vitro andin vivo (Figures 7A to 7D). The interaction between DA1 andUBP15 may help DA1 specifically recognize UBP15 for protea-somal degradation.Currently, plant seeds and organs are not only used as foods,

but also used as sustainable fuel and energy sources. Ourfindings establish a genetic and molecular framework for theregulation of seed and organ size by DA1, DA2, EOD1, andUBP15 in Arabidopsis. Interestingly, overexpression of themaize da1-1 mutant allele (Zea mays da1-1) has been known toincrease seed mass of maize (Wang et al., 2012), and the ho-molog of DA2 in rice is GW2, a negative regulator of rice grainsize (Song et al., 2007). Thus, our current understanding of DA1,DA2, EOD1, and UBP15 functions suggests that these fourubiquitin-related proteins and their orthologs in other plantspecies could be used to engineer large seed size and increasedbiomass.

METHODS

Plant Materials and Growth Conditions

Arabidopsis thaliana ecotype Col-0 was used as the wild-type line.All mutants used in this study were in the Col-0 background. sod2-1and sod2-2 were identified as suppressors of da1-1 from an ethylmethanesulfonate–treated M2 population of da1-1. The ubp15-1(SALK_018601) was obtained from the Nottingham Arabidopsis StockCentre. Arabidopsis plants were grown under long-day conditions (16 hlight/8 h dark) at 22°C.

Map-Based Cloning, Constructs, and Plant Transformation

The sod2-1mutation wasmapped in the F2 population of a cross betweensod2-1 da1-1 and da1-1Ler. Molecular markers were developed usingpublic database (Supplemental Table 1).

The 6424-bp genomic sequence containing 2074-bp promoter, theUBP15 gene, and 276-bp 39 untranslated region was amplified using theprimers gUBP15 (99)-F and gUBP15 (99)-R. PCR products were thencloned to pDONR207 using BP enzyme (Invitrogen). The UBP15 genomicDNA was then subcloned to the pMDC99 binary vector by LR reaction togenerate the gUBP15 construct. The plasmid gUBP15 was introducedinto the sod2-2 da1-1 mutant plants using Agrobacterium tumefaciensGV3101, and transformants were selected on medium containing hy-gromycin (30 mg/mL).

The gUBP15-GFP construct was made using a PCR-based Gatewaysystem. The 6145-bp genomic sequence containing 2074-bp promoterand the UBP15 gene without the stop code was amplified using theprimers gUBP15 (99)-F and gUBP15 (107)-R. PCR products were clonedto pDONR207 using BP enzyme (Invitrogen). The UBP15 genomic DNAwas then subcloned to the pMDC107 binary vector with the GFP geneto generate the gUBP15-GFP construct. The plasmid gUBP15-GFPwas introduced into Col-0 plants using Agrobacterium GV3101,and transformants were selected on medium containing hygromycin(30 mg/mL).

The 2111-bp promoter sequence of UBP15 was amplified using theprimers pUBP15-F and pUBP15-R (Supplemental Table 1). PCR productswere cloned into pDONR207 using BP enzyme. TheUBP15 promoter was

674 The Plant Cell

then subcloned to the binary vector pGWB3 with the GUS reporter geneto generate the transformation plasmid pUBP15:GUS. The plasmidpUBP15:GUS was introduced into Col-0 plants using AgrobacteriumGV3101, and transformants were selected on medium containing hy-gromycin (30 mg/mL).

Morphological and Cellular Analysis

To determine seed size, we photographedmature dry seeds under a Leicamicroscope (Leica S8APO) using a Leica charge-coupled device camera(DFC420). The projective area of seeds was measured using ImageJ software. Average seed weight was determined by weighing mature dryseeds in batches of 500 using an electronic analytical balance (MettlerMoledo AL104). The weights of five sample batches were measured foreach seed lot.

Petals (stage 14), leaves, and cotyledons were scanned to producedigital images, respectively. The average area of petals, leaves, andcotyledons was calculated using Image J software. Leaf and petal cellsizes were measured from differential interference contrast images.

GUS Staining

Seedlings, sepals, petals, stamens, carpels, and ovules of pUBP15:GUStransgenic plants were stained in a GUS staining solution (1 mM 5-bromo-4-chloro-3-indolyl-b-D-glucuronic acid, 100 mM Na3PO4 buffer, 3 mMeach K3Fe(CN)6/K4Fe(CN)6, 10 mM EDTA, and 0.1% Nonidet P-40) andincubated at room temperature for 5 h. After GUS staining, chlorophyllwas removed using 70% ethanol.

RNA Isolation, RT-PCR, and Quantitative Real-TimeRT-PCR Analysis

Total RNA was extracted from Arabidopsis seedlings using an RNApreppure plant kit (Tiangen). Total mRNA was reverse transcribed into cDNAusing SuperScript III reverse transcriptase (Invitrogen). Quantitative real-time RT-PCR analysis was performed with a Lightcycler 480 engine(Roche) using the Lightcycler 480 SYBR Green Master (Roche). Theprimers used for RT-PCR and quantitative real-time RT-PCR are de-scribed in Supplemental Table 1.

In Vitro Protein–Protein Interaction

The coding sequences of UBP15 and its N terminus (UBP15-N ) weresubcloned into XbaI and SbfI sites of the pMAL-c2 vector to generateMBP-UBP15 andMBP-UBP15-N constructs. The coding sequence of theC terminus (UBP15-C ) was subcloned into XbaI and HindIII sites of thepMAL-c2 vector to generate the MBP-UBP15-C construct. The specificprimers forMBP-UBP15,MBP-UBP15-N, andMBP-UBP15-C constructswere MBP-UBP15-F, MBP-UBP15-R, MBP-UBP15N-R, MBP-UBP15C-F,and MBP-UBP15C-R, respectively (Supplemental Table 1). GST-DA1,GST-DA1R358K, GST-DA1-UIM, GST-DA1-LIM+C, GST-DA1-LIM, andGST-DA1-C constructs were made as previously described (Xia et al.,2013).

For in vitro pull-down assay, bacterial lysates containing ;15 mgMBP-UBP15, MBP-UBP15-N, or MBP-UBP15-C fusion proteins weremixed with lysates containing ;30 mg GST-DA1, GST-DA1R358K, GST-DA1-UIM, GST-DA1-LIM, GST-DA1-LIM+C, or GST-DA1-C fusionproteins. Amylose resins (25 mL; New England Biolabs) were added intoeach combined solution with rocking at 4°C for 30 min. Beads werewashed five times with the buffer (50 mM HEPES, pH 7.5, 150 mMNaCl, 1.5 mM MgCl2, 1 mM EGTA, pH 8.0, 1% Triton X-100, 10%glycerol, 1 mM PMSF, and 13 Complete protease inhibitor cocktail[Roche]), and the isolated proteins were further separated on a 10%SDS-polyacrylamide gel and detected by immunoblot analysis with

anti-GST (Abmart) and anti-MBP antibodies (New England Biolabs),respectively.

Coimmunoprecipitation

The coding sequence of UBP15 was amplified using the primers GFP-UBP15-F and GFP-UBP15-R. PCR products were subcloned into thepCR8/GW/TOPO TA vector (Invitrogen) using TOPO enzyme. UBP15wasthen subcloned into the Gateway binary vector pMDC43 containing the35S promoter and the GFP gene to generate the 35S:GFP-UBP15construct. The 35S:Myc-DA1 construct was made as previously de-scribed (Xia et al., 2013).

Nicotiana benthamiana leaves were transformed by injection ofAgrobacterium EHA105 cells harboring 35S:GFP-UBP15, 35S:Myc-DA1,and 35S:GFP plasmids. Total protein was extracted with extraction buffer(50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 20% glycerol, 2% Triton X-100,1 mM EDTA, 13Complete protease inhibitor cocktail, and 20 mMMG132)and incubated with GFP-Trap_A (Chromotek) for 30 min at 4°C. Beadswere washed three times with wash buffer (50 mM Tris/HCl, pH 7.5,150 mM NaCl, 20% glycerol, 0.1% Triton X-100, 1 mM EDTA, pH 8.0, 13Complete protease inhibitor cocktail, and 20 mM MG132). The im-munoprecipitates were separated in 10% SDS-polyacrylamide gel anddetected by immunoblot analysis with anti-Myc (Abmart) and anti-GFP(Beyotime) antibodies, respectively.

Accession Numbers

Arabidopsis Genome Initiative locus identifiers for genesmentioned in thisarticle are as follows: At1g19270 (DA1), At1g17110 (SOD2/UBP15),At3g63530 (EOD1/BB), and At1g78420 (DA2).

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure 1. Identification and Molecular Characterizationof SOD2/UBP15.

Supplemental Figure 2. ubp15-1 Suppresses the Seed- and Organ-Size Phenotypes of da1-1.

Supplemental Figure 3. ubp15-1 Forms Small Organs.

Supplemental Figure 4. ubp15-1 Is Epistatic to da1-1.

Supplemental Figure 5. UBP15 Acts Independently of EOD1 and DA2to Regulate Seed and Organ Size.

Supplemental Figure 6. The C-Terminal Region of UBP15 Interactswith DA1 in Vivo.

Supplemental Table 1. Primers Used in This Study.

ACKNOWLEDGMENTS

We thank the Nottingham Arabidopsis Stock Centre for the ubp15-1mutant. This work was supported by the grants from National BasicResearch Program of China (2009CB941503), the National NaturalScience Foundation of China (91017014, 31221063, 30870215, and31300242), and the Hundred Talent Program of the Chinese Academy ofsciences.

AUTHOR CONTRIBUTIONS

L.D., N.L., and Y.H.L. designed the research. Y.H.L. conceived andsupervised the project. L.D. performed most of the experiments. N.L. and

Ubiquitin-Mediated Control of Seed Size 675

L.D. conducted pull-down assays. L.C. performed coimmunoprecipita-tion experiments. Y.X. identified the ubp15-1 da1-1 double mutant. Y.L.and Y.Z. helped do cellular analysis. L.D., N.L., C.L., and Y.H.L. analyzeddata. Y.H.L. wrote the article.

Received January 5, 2014; revised February 4, 2014; accepted February6, 2014; published February 28, 2014.

REFERENCES

Adamski, N.M., Anastasiou, E., Eriksson, S., O’Neill, C.M., andLenhard, M. (2009). Local maternal control of seed size by KLUH/CYP78A5-dependent growth signaling. Proc. Natl. Acad. Sci. USA106: 20115–20120.

Alonso-Blanco, C., Blankestijn-de Vries, H., Hanhart, C.J., andKoornneef, M. (1999). Natural allelic variation at seed size loci inrelation to other life history traits of Arabidopsis thaliana. Proc. Natl.Acad. Sci. USA 96: 4710–4717.

Disch, S., Anastasiou, E., Sharma, V.K., Laux, T., Fletcher, J.C.,and Lenhard, M. (2006). The E3 ubiquitin ligase BIG BROTHERcontrols arabidopsis organ size in a dosage-dependent manner.Curr. Biol. 16: 272–279.

Fan, C., Xing, Y., Mao, H., Lu, T., Han, B., Xu, C., Li, X., and Zhang,Q. (2006). GS3, a major QTL for grain length and weight and minorQTL for grain width and thickness in rice, encodes a putativetransmembrane protein. Theor. Appl. Genet. 112: 1164–1171.

Fang, W., Wang, Z., Cui, R., Li, J., and Li, Y. (2012). Maternal controlof seed size by EOD3/CYP78A6 in Arabidopsis thaliana. Plant J. 70:929–939.

Garcia, D., Fitz Gerald, J.N., and Berger, F. (2005). Maternal controlof integument cell elongation and zygotic control of endospermgrowth are coordinated to determine seed size in Arabidopsis. PlantCell 17: 52–60.

Garcia, D., Saingery, V., Chambrier, P., Mayer, U., Jürgens, G., andBerger, F. (2003). Arabidopsis haiku mutants reveal new controls ofseed size by endosperm. Plant Physiol. 131: 1661–1670.

Gegas, V.C., Nazari, A., Griffiths, S., Simmonds, J., Fish, L., Orford,S., Sayers, L., Doonan, J.H., and Snape, J.W. (2010). A geneticframework for grain size and shape variation in wheat. Plant Cell 22:1046–1056.

Horiguchi, G., Kim, G.T., and Tsukaya, H. (2005). The transcriptionfactor AtGRF5 and the transcription coactivator AN3 regulate cellproliferation in leaf primordia of Arabidopsis thaliana. Plant J. 43:68–78.

Jofuku, K.D., Omidyar, P.K., Gee, Z., and Okamuro, J.K. (2005).Control of seed mass and seed yield by the floral homeotic geneAPETALA2. Proc. Natl. Acad. Sci. USA 102: 3117–3122.

Kang, X., Li, W., Zhou, Y., and Ni, M. (2013). A WRKY transcriptionfactor recruits the SYG1-like protein SHB1 to activate geneexpression and seed cavity enlargement. PLoS Genet. 9: e1003347.

Kim, D.Y., Scalf, M., Smith, L.M., and Vierstra, R.D. (2013).Advanced proteomic analyses yield a deep catalog of ubiquitylationtargets in Arabidopsis. Plant Cell 25: 1523–1540.

Lee, B.H., Ko, J.H., Lee, S., Lee, Y., Pak, J.H., and Kim, J.H. (2009).The Arabidopsis GRF-INTERACTING FACTOR gene family performsan overlapping function in determining organ size as well as multipledevelopmental properties. Plant Physiol. 151: 655–668.

Li, J., Nie, X., Tan, J.L., and Berger, F. (2013). Integration ofepigenetic and genetic controls of seed size by cytokinin inArabidopsis. Proc. Natl. Acad. Sci. USA 110: 15479–15484.

Li, Y., Zheng, L., Corke, F., Smith, C., and Bevan, M.W. (2008).Control of final seed and organ size by the DA1 gene family inArabidopsis thaliana. Genes Dev. 22: 1331–1336.

Linkies, A., Graeber, K., Knight, C., and Leubner-Metzger, G.(2010). The evolution of seeds. New Phytol. 186: 817–831.

Liu, Y., Wang, F., Zhang, H., He, H., Ma, L., and Deng, X.W. (2008).Functional characterization of the Arabidopsis ubiquitin-specificprotease gene family reveals specific role and redundancy ofindividual members in development. Plant J. 55: 844–856.

Lopes, M.A., and Larkins, B.A. (1993). Endosperm origin, development,and function. Plant Cell 5: 1383–1399.

Luo, M., Dennis, E.S., Berger, F., Peacock, W.J., and Chaudhury,A. (2005). MINISEED3 (MINI3), a WRKY family gene, and HAIKU2(IKU2), a leucine-rich repeat (LRR) KINASE gene, are regulators ofseed size in Arabidopsis. Proc. Natl. Acad. Sci. USA 102: 17531–17536.

Lutterbach, B., Sun, D., Schuetz, J., and Hiebert, S.W. (1998). TheMYND motif is required for repression of basal transcription fromthe multidrug resistance 1 promoter by the t(8;21) fusion protein.Mol. Cell. Biol. 18: 3604–3611.

Moles, A.T., Ackerly, D.D., Webb, C.O., Tweddle, J.C., Dickie, J.B.,and Westoby, M. (2005). A brief history of seed size. Science 307:576–580.

Ohto, M.A., Fischer, R.L., Goldberg, R.B., Nakamura, K., andHarada, J.J. (2005). Control of seed mass by APETALA2. Proc.Natl. Acad. Sci. USA 102: 3123–3128.

Ohto, M.A., Floyd, S.K., Fischer, R.L., Goldberg, R.B., and Harada,J.J. (2009). Effects of APETALA2 on embryo, endosperm, and seedcoat development determine seed size in Arabidopsis. Sex. PlantReprod. 22: 277–289.

Orsi, C.H., and Tanksley, S.D. (2009). Natural variation in an ABCtransporter gene associated with seed size evolution in tomatospecies. PLoS Genet. 5: e1000347.

Reiser, L., and Fischer, R.L. (1993). The ovule and the embryo sac.Plant Cell 5: 1291–1301.

Schruff, M.C., Spielman, M., Tiwari, S., Adams, S., Fenby, N., andScott, R.J. (2006). The AUXIN RESPONSE FACTOR 2 gene ofArabidopsis links auxin signalling, cell division, and the size ofseeds and other organs. Development 133: 251–261.

Scott, R.J., Spielman, M., Bailey, J., and Dickinson, H.G. (1998).Parent-of-origin effects on seed development in Arabidopsisthaliana. Development 125: 3329–3341.

Shomura, A., Izawa, T., Ebana, K., Ebitani, T., Kanegae, H.,Konishi, S., and Yano, M. (2008). Deletion in a gene associatedwith grain size increased yields during rice domestication. Nat.Genet. 40: 1023–1028.

Song, X.J., Huang, W., Shi, M., Zhu, M.Z., and Lin, H.X. (2007). AQTL for rice grain width and weight encodes a previously unknownRING-type E3 ubiquitin ligase. Nat. Genet. 39: 623–630.

Verma, R., Oania, R., Graumann, J., and Deshaies, R.J. (2004).Multiubiquitin chain receptors define a layer of substrate selectivityin the ubiquitin-proteasome system. Cell 118: 99–110.

Wang, A., Garcia, D., Zhang, H., Feng, K., Chaudhury, A., Berger,F., Peacock, W.J., Dennis, E.S., and Luo, M. (2010). The VQ motifprotein IKU1 regulates endosperm growth and seed size inArabidopsis. Plant J. 63: 670–679.

Wang, X., Liu, B., Huang, C., Zhang, X., Luo, C., Cheng, X., Yu, R.,and Wu, Z. (2012). Over expression of Zmda1-1 gene increasesseed mass of corn. Afr. J. Biotechnol. 11: 13387–13395.

Weng, J., et al. (2008). Isolation and initial characterization of GW5,a major QTL associated with rice grain width and weight. Cell Res.18: 1199–1209.

676 The Plant Cell

White, D.W. (2006). PEAPOD regulates lamina size and curvature inArabidopsis. Proc. Natl. Acad. Sci. USA 103: 13238–13243.

Xia, T., Li, N., Dumenil, J., Li, J., Kamenski, A., Bevan, M.W., Gao,F., and Li, Y. (2013). The ubiquitin receptor DA1 interacts with theE3 ubiquitin ligase DA2 to regulate seed and organ size inArabidopsis. Plant Cell 25: 3347–3359.

Xiao, W., Brown, R.C., Lemmon, B.E., Harada, J.J., Goldberg,R.B., and Fischer, R.L. (2006). Regulation of seed size by

hypomethylation of maternal and paternal genomes. Plant Physiol. 142:1160–1168.

Xu, R., and Li, Y. (2011). Control of final organ size by Mediator complexsubunit 25 in Arabidopsis thaliana. Development 138: 4545–4554.

Zhou, Y., Zhang, X., Kang, X., Zhao, X., Zhang, X., and Ni, M. (2009).SHORT HYPOCOTYL UNDER BLUE1 associates with MINISEED3and HAIKU2 promoters in vivo to regulate Arabidopsis seed development.Plant Cell 21: 106–117.

Ubiquitin-Mediated Control of Seed Size 677

DOI 10.1105/tpc.114.122663; originally published online February 28, 2014; 2014;26;665-677Plant Cell

Liang Du, Na Li, Liangliang Chen, Yingxiu Xu, Yu Li, Yueying Zhang, Chuanyou Li and Yunhai LiArabidopsisUbiquitin-Specific Protease UBP15/SOD2 in

The Ubiquitin Receptor DA1 Regulates Seed and Organ Size by Modulating the Stability of the

 This information is current as of August 6, 2020

 

Supplemental Data /content/suppl/2014/02/14/tpc.114.122663.DC1.html

References /content/26/2/665.full.html#ref-list-1

This article cites 38 articles, 23 of which can be accessed free at:

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists