Posted on

m1-56 seeds

Seed Morphology in Silene Based on Geometric Models

2 Department of Biogeography, Paleoecology and Nature Conservation, Faculty of Biology and Environmental Protection, University of Lodz, 1/3 Banacha Str., 90-237 Lodz, Poland; [email protected]

José Luis Rodríguez-Lorenzo

3 Plant Developmental Genetics, Institute of Biophysics v.v.i, Academy of Sciences of the Czech Republic, Královopolská 135, 612 65 Brno, Czech Republic; [email protected] (J.L.R.-L.); [email protected] (B.J.)

Bohuslav Janoušek

3 Plant Developmental Genetics, Institute of Biophysics v.v.i, Academy of Sciences of the Czech Republic, Královopolská 135, 612 65 Brno, Czech Republic; [email protected] (J.L.R.-L.); [email protected] (B.J.)

Emilio Cervantes

1 IRNASA-CSIC, Cordel de Merinas 40, 37008 Salamanca, Spain; [email protected]

Abstract

Seed description in morphology is often based on adjectives such as “spherical”, “globular”, or “reniform”, but this does not provide a quantitative method. A new morphological approach based on the comparison of seed images with geometric models provides a seed description in Silene species on a quantitative basis. The novelty of the proposed method is based in the comparison of the seed images with geometric models according to a cardioid shape. The J index is a measurement that indicates the seed percentage of similarity with a cardioid or cardioid-derived figures used as models. The seeds of Silene species have high values of similarity with the cardioid and cardioid-derived models (J index superior to 90). The comparison with different figures allows species description and differentiation. The method is applied here to seeds of 21 species and models are proposed for some of them including S. diclinis, an endangered species. The method is discussed in the context of previous comparison with the measures used in traditional morphometric analysis. The similarity of seed images with geometric figures opens a new perspective for the automatized taxonomical evaluation of samples linking seed morphology to functional traits in endangered Silene species.

1. Introduction

The Caryophyllaceae Juss. (Caryophyllales Jussieu ex Bercht. and J. Presl) comprises about 90 genera and 2625 species, of a wide distribution with their greatest diversity in temperate climate [1,2,3,4]. Silene L. is the largest genus of the Caryophyllaceae [5]. The number of species in the genus varies between taxonomic treatments, including from 700 [6,7] to 800–900 species [4,8]. Most of them are diploid (2n = 2x = 24) but there are also tetraploid, hexaploid and octoploid species [9]. Species of Silene are annual, biennial, and perennial herbs distributed mainly across the Northern Hemisphere with two main centers of diversity: The South Balkan Peninsula and South-West of Asia [6,7,9]. At least 12 endangered species have been reported in Silene, including five in Spain: S. diclinis (Lag.) M. Laínz [10], S. fernandezii Jeanm., S. gazulensis A.Galán de Mera, J.E. Cortés, J.A. Vicente Orellana and R. Morales Alonso, S. hifacensis (Rouy ex Willk.) O. Bolòs and Vigo [11,12], and S. sennenii Pau [13]. According to comprehensive gene tree analyses based on the nrDNA ITS and cpDNA rps16 and phylogenetic tree analysis including 262 samples representing 243 species, the genus Silene has been split in three subgenera: S. subg. Lychnis, S. subg. Behenantha, and Silene sensu stricto [14]. The authors of this article already indicated the lack of a morphological diagnostic key in support of this distribution, due to a high degree of homoplasy.

Seed morphology provides important information in taxonomy, and it has been applied to genera in the Caryophyllaceae, such as Arenaria L. [15,16], Gypsophila L. [17,18], Moehringia L. [19], Paronychia Mill. [20], Sagina L. [21], Stellaria L. [22], and Velezia L. [23]. Silene is the best studied genus in the Caryophyllaceae in terms of seed morphology [5,24,25,26,27,28,29,30,31].

Seed morphology studies in Silene have a long tradition and present descriptions of seed shape and ornamentation. In 1869, Rohrbach applied to Silene seeds the expression reniformia (kidney-shaped), and introduced a classification based on the structure of the back of the seed as plane (dorso plana) or deepened (dorso canaliculata). His research sought the importance of seed morphology in the description of Silene seeds indicating: ”denn in der That bietet die Gestalt des Samens das sicherste Kennzeichen zur Unterscheidung sehr vieler species” (because in fact the shape of the seed offers the most reliable indicator for the differentiation of many species) [25]. Variation was observed in size and shape as well as in seed coat microstructure and the structure of lateral and dorsal faces [27,28,29,30,31]. Hoseini et al., 2017 described species and sections keeping common traits not only macromorphologically but also in the microstructure of the seed [31].

Since Rohrbach’s work, seed shape description in Silene species is based on adjectives such as “reniform”, “circular”, “globular”, or “semi-globular” [6,9,29,30]. However, these adjectives are not precise because they refer simultaneously to two different characters: two-dimensional images of the seed and the three-dimensional shape. For example, the term “reniform” means that the overall structure resembles a kidney, and “globular” means that it resembles a sphere. In the first case, it is not possible to have quantitative data because the kidney is not a geometrically defined figure, and in both cases, there are no described means to determine the degree of similarity to a kidney or to quantify sphericity of the seeds. To address this situation, we present a method based on the comparison of the seed images with bi-dimensional geometric figures described mathematically [32,33]. This allows the quantification of the two-dimensional shape of the seeds by the percentage of similarity between the seed image and a given geometric figure. This measure was termed J index. The method was first applied to Arabidopsis thaliana (L.) Heynh. [34] and later to the model legumes Lotus japonicus L. and Medicago truncatula Gaertn. [35,36] among other species.

The bidimensional images of the seeds of Arabidopsis thaliana resemble a cardioid elongated by a factor of Phi (The Golden Ratio = 1.618) in the horizontal axis [34], while the seeds of Medicago truncatula resemble a cardioid elongated by a factor of Phi in the vertical axis [35], and Lotus japonicus seeds adjust well to a cardioid [35,36]. Also, the seeds of Capparis spinosa L. adjust well to the cardioid [37], as well as species of the Papaveraceae and Malvaceae [38,39]. Ovals and ellipses were the models for seed shape quantification in many species of the Cucurbitaceae [40] and the Euphorbiaceae, such as Ricinus L. and Jatropha L. [41,42]. The description of shape in wheat kernels was based on three geometric figures: (1) an ellipse of aspect ratio (AR) = 1.8 for the “round varieties” (Triticum aestivum subsp. aestivum var. Zebra and Torka), (2) a lens of AR = 3.2 for the elongated kernels (T. monococcum L.), and (3) an ellipse of AR = 2.4 with the intermediate-shaped varieties such as T. turgidum subsp. durum cv. Floradur [43]. The seeds of species in the Vitaceae, as well as diverse cultivars of Vitis vinifera, were accurately described with a set of morphological models based on heart-shaped and piriform curves [44]. Studies of comparative morphology in Silene species based on geometric models can provide original information about taxonomic relationships pointing as well to associations between seed shape and ecological properties.

The main objective of this work is to define, for the first time, geometric models adjusting the bi-dimensional images of the seeds of species from the genus Silene, including the threatened species S. diclinis and assess their value in taxonomic classification. In addition, other morphological aspects are described and their application in classification is discussed.

2. Results

2.1. General Morphological Description: Size and Shape in the Seeds of Silene Species

Table 1 and Table 2 present a general morphological description of the seeds in groups corresponding to S. subg. Behenantha and S. subg. Silene. Table 1 contains data for eleven species belonging to S. subg. Behenantha and Table 2 presents the data corresponding to ten species of S. subg. Silene. Figure A1 contains the box plots corresponding to four selected characters (Area, Aspect Ratio, Circularity, and Roundness), for S. subg. Behenantha and S. subg. Silene respectively.

Table 1

Mean values of the area (A), perimeter (P), length of the major axis (L), length of the minor axis (W), aspect ratio (AR is the ratio L/W), circularity (C) and roundness (R) in the seeds of species of S. subg. Behenantha. Values of A are given in mm 2 ; P, L and W, in mm. Standard deviation values are given in parentheses. Superscript letters indicate the results of Scheffé test: The mean values marked with the same letter in each column do not differ significantly at p < 0.05. n is the number of seeds analyzed.

Species n A P L W AR C R
S. acutifolia 40 1.18 de 4.55 de 1.35 efg 1.15 d 1.18 cde 0.72 de 0.85 bcd
(0.14) (0.32) (0.08) (0.07) (0.06) (0.03) (0.04)
S. conica 40 0.52 i 2.80 i 0.89 j 0.75 g 1.18 cde 0.83 a 0.85 bcd
(0.03) (0.08) (0.03) (0.02) (0.04) (0.02) (0.03)
S. diclinis 40 1.73 b 5.28 b 1.65 b 1.36 b 1.21 bcd 0.78 bc 0.83 cde
(0.14) (0.24) (0.07) (0.06) (0.05) (0.03) (0.04)
S. dioica Chk 40 1.30 c 5.30 c 1.42 cd 1.21 c 1.16 e 0.59 f 0.86 b
(0.17) (0.59) (0.10) (0.09) (0.04) (0.08) (0.03)
S. dioica Pol 40 0.94 h 4.07 h 1.20 i 1.02 f 1.18 de 0.71 de 0.85 bc
(0.12) (0.37) (0.08) (0.06) (0.05) (0.05) (0.03)
S. latifolia Chk 40 1.15 gh 4.52 gh 1.33 hi 1.12 f 1.18 cde 0.71 ab 0.85 bcd
(0.14) (0.43) (0.09) (0.07) (0.05) (0.06) (0.03)
S. latifolia Pol 1 40 1.07 fg 4.29 fg 1.29 g h 1.06 ef 1.21 abc 0.74 d 0.82 def
(0.16) (0.42) (0.09) (0.09) (0.03) (0.04) (0.02)
S. latifolia Pol 2 40 1.26 cd 4.67 cd 1.41 cde 1.14 d 1.22 ab 0.73 d 0.82 ef
(0.19) (0.33) (0.10) (0.08) (0.03) (0.05) (0.02)
S. latifolia Pol 3 40 1.02 ef 3.97 ef 1.23 fg 1.04 d 1.18 de 0.80 de 0.85 bc
(0.18) (0.31) (0.11) (0.09) (0.03) (0.04) (0.03)
S. noctiflora 40 1.28 cd 4.55 cd 1.36 def 1.21 c 1.12 f 0.78 c 0.90 a
(0.11) (0.21) (0.06) (0.05) (0.04) (0.02) (0.03)
S. pendula 40 1.14 ef 4.48 ef 1.34 efg 1.11 de 1.20 bcde 0.71 de 0.84 bcde
(0.10) (0.23) (0.07) (0.07) (0.08) (0.03) (0.05)
S. uniflora 40 1.31 c 4.51 c 1.43 c 1.16 cd 1.25 a 0.80 a b 0.80 f
(0.13) (0.23) (0.09) (0.06) (0.07) (0.01) (0.04)
S. viscosa 40 0.51 i 3.19 i 0.89 j 0.75 g 1.17 de 0.63 f 0.85 bc
(0.05) (0.18) (0.05) (0.04) (0.04) (0.04) (0.03)
S. vulgaris 40 1.20 cde 4.67 cde 1.37 cdef 1.14 d 1.20 bcd 0.69 e 0.83 cde
(0.15) (0.36) (0.09) (0.08) (0.06) (0.04) (0.04)
S. zawadzkii 40 2.02 a 5.74 a 1.79 a 1.46 a 1.23 ab 0.77 c 0.81 ef
(0.24) (0.36) (0.10) (0.08) (0.06) (0.02) (0.04)

Table 2

Mean values of the area (A), perimeter (P), length of the major axis (L), length of the minor axis (W), aspect ratio (AR is the ratio L/W), circularity (C) and roundness (R) in the seeds of species of S. subg. Silene. Values of A are given in mm 2 ; P, L, and W, in mm. Standard deviation values are given in parentheses. Superscript letters indicate the results of Scheffé test: the mean values marked with the same letter in each column do not differ significantly at p < 0.05. n is the number of seeds analyzed.

Species n A P L W AR C R
S. colpophylla 40 0.68 c 3.44 de 1.08 c 0.82 d 1.35 a 0.72 e 0.74 c
(0.07) (0.18) (0.05) (0.06) (0.08) (0.02) (0.04)
S. gallica 40 0.53 d 2.97 e 0.93 e 0.74 f 1.26 b 0.76 cd 0.79 c
(0.05) (0.17) (0.05) (0.04) (0.05) (0.03) (0.03)
S. italica 40 0.65 c 3.21 e 1.01 d 0.83 d 1.21 cd 0.79 a 0.83 ab
(0.07) (0.16) (0.05) (0.05) (0.07) (0.01) (0.05)
S. mellifera 40 0.99 a 3.95 c 1.25 ab 1.01 b 1.25 bc 0.79 ab 0.80 bc
(0.21) (0.40) (0.13) (0.12) (0.07) (0.03) (0.04)
S. nutans Chk 40 1.06 a 4.25 b 1.3 a 1.06 a 1.19 cd 0.73 d 0.81 ab
(0.12) (0.24) (0.05) (0.08) (0.06) (0.04) (0.05)
S. nutans Pol 40 1.07 a 4.22 b 1.29 a 1.07 a 1.21 cd 0.75 d 0.83 ab
(0.13) (0.23) (0.07) (0.07) (0.05) (0.03) (0.04)
S. otites 40 0.41 d 2.60 f 0.81 e 0.66 f 1.24 bcd 0.76 cd 0.81 abc
(0.06) (0.20) (0.07) (0.05) (0.07) (0.02) (0.05)
S. saxifraga 40 0.78 b 3.52 d 1.11 c 0.92 c 1.20 d 0.79 a 0.84 a
(0.11) (0.27) (0.08) (0.07) (0.07) (0.02) (0.04)
S. schafta 40 0.82 b 4.79 a 1.22 b 0.93 c 1.34 a 0.45 h 0.75 c
(0.11) (0.59) (0.09) (0.07) (0.06) (0.06) (0.03)
S. tatarica 40 0.54 d 2.96 e 0.93 e 0.76 ef 1.25 bc 0.77 bc 0.80 bc
(0.05) (0.14) (0.04) (0.04) (0.06) (0.02) (0.04)
S. wolgensis 40 0.57 d 3.21 e 0.97 de 0.80 de 1.23 bcd 0.70 f 0.82 abc
(0.05) (0.18) (0.05) (0.04) (0.07) (0.03) (0.05)

Notable differences between species were found for all parameters analyzed. In the species of S. subg. Behenantha, seed image area was comprised between 0.51 mm 2 in S. viscosa and 2.02 mm 2 in S. zawadzkii. These values were extreme, with most species having mean area values comprised between 1.10 and 1.73 mm 2 . The two stocks of S. dioica, one in the collection of Poland and the other in Czech Republic, had different values in A, P, L, W, and C, but not in AR or R. Differences between stocks of S. latifolia were observed in all measurements.

The aspect ratio was comprised between 1.12 and 1.25 with an opposed trend to both circularity and roundness, and roundness values were higher than those of circularity. In cases where seed surface has prolongations increasing seed perimeter, circularity is reduced while roundness maintains high values. The values of roundness were higher in S. noctiflora (0.90) than in the other species (comprised between 0.80 and 0.86). Circularity values were comprised between 0.59 (S. dioica Chk) and 0.83 in S. conica.

In the species of S. subg. Silene, seed image area was comprised between 0.41 mm 2 in S. otites and 1.07 mm 2 in S. nutans ( Table 2 ). The values of aspect ratio were between 1.20 (S. saxifraga) and 1.35 (S. colpophylla), while the circularity was particularly low in S. schafta, due to its surface protuberances, but it was comprised between 0.70 and 0.79 in the other species. Roundness values were comprised between 0.74 (S. colpophylla) and 0.84 (S. saxifraga). In both, the species of S. subg. Behenantha and the species of S. subg. Silene, values of aspect ratio, circularity and roundness had lower variation rates than the measurements of size.

2.2. Structural Aspects

Structural aspects include seed surface and other properties of seeds, including asymmetry, the presence of a dorsal face and the existence of a ridge.

2.2.1. Seed Surface

Surface texture is better appreciated in confocal microscopy ( Figure A2 ). The seed surface can be smooth, such as in S. colpophylla and S. conica or covered by colliculae, that may be conical (acute) resulting in a radiate structure (S. dioica), conical-obtuse (S. gallica, S. latifolia) or rounded (S. acutifolia, S. diclinis, S. noctiflora, and S. nutans).

2.2.2. Other Properties of Seeds

In addition to surface structure, other aspects in the morphological description of Silene seeds include: (1) asymmetry, (2) a pronounced dorsal face that may be plane or concave, and (3) presence of a ridge. Table 3 presents a summary of the variability in the morphological characteristics of the species studied.

Table 3

Summary of the variability in the morphological characteristics of the species of Silene studied. A minus sign (-) indicates the absence (or predominance of absence) and a plus sign (+) the presence (predominance of presence) of the character indicated for a given species. An asterisk indicates the species in which only few seeds were observed with ridge. When two equal signs coincide in two opposite characters it means that none of them is predominant.

Subg. Species Symmetry Dorsal Face Ridge Hilium
Plane Canaliculate Almost None Almost All
Behenantha S. acutifolia Link ex Rohrb. + +
S. conica L. + + +
S. diclinis (Lag.) M.Laínz +
S. dioica (L.) Clairv. + +
S. latifolia Poir. +
S. noctiflora L. + +
S. pendula L. + +
S. uniflora Roth + + -*
S. viscosa (L.) Pers. +
S. vulgaris (Moench) Garcke + + +
S. zawadzkii Herbich + -* +
Silene S. colpophylla Wrigley + + +
S. gallica L. + + + +
S. italica (L.) Pers. + + + +
S. mellifera Boiss. and Reut. + +
S. nutans L. + + -*
S. otites (L.) Wibel + +
S. saxifraga L. + +
S. schafta S.G.Gmel. + +
S. tatarica (L.) Pers. + + +
S. wolgensis (Hornem.) Otth + + +

A marked asymmetry is observed in the seeds of S. acutifolia, S. colpophylla, S. tatarica, and S. wolgensis ( Figure 1 ). In all the images used, the seeds are oriented with the micropile to the right so that approximately half of the seed is above a hypothetical horizontal line passing through the micropile, and the other half is below this line. The asymmetry consists of differences above and below the micropyle. Three types can be described: (1) Asymmetric seeds with the two lobes more or less rounded and of different size and/or shape; (2) asymmetric seeds with one rounded lobe and the other flattened, and vertical or almost vertical; and (3) asymmetric seeds with one rounded lobe and the other flattened and inclined. Figure 2 contains examples of these types of asymmetry.

Seed asymmetry is frequently observed in S. acutifolia, S. colpophylla, S. tatarica, and S. wolgensis. Bar represents 0.5 mm.

Types of asymmetry observed in Silene species. Left: a symmetric seed (S. nutans). Next, left to right: Type 1: S. schafta (the two lobes of different size and/or shape); Type 2: S. mellifera (one lobe rounded and the other flattened); Type 3: S. tatarica (one lobe rounded and the other flattened and inclined). Bars represent 0.5 mm.

Asymmetric seeds are more frequent in species of S. subg. Silene than in S. subg. Behenantha ( Table 2 ).

Seeds of some species have a dorsal structure that can be plane or concave, forming a channel (S. colpophyla, S. conica) corresponding to the two classes that were termed by Rohrbach as dorso-plana and dorso-canaliculata [25] (p. 50) ( Figure 3 ). These types of dorsal structure are more frequent in the seeds of species of S. subg. Silene than in S. subg. Behenantha ( Table 2 ).

Representative examples of seeds with their dorsal face plane and canaliculate. Left: S. dioica as a control (dorsal face rounded, no particular structures). Above, three examples of dorsal face plane: S. acutifolia, S. italica, and S. pendula. Below, three examples of dorsal face canaliculate: S. colpophylla, S. conica, and S. tatarica. Bars represent 0.5 mm.

The seeds of S. gallica present a conspicuous ridge ( Figure 4 ). This structure is also observed in other species, such as S. nutans, S. uniflora, and S. zawadzkii, but in a lower number of seeds and with a lesser definition than in S. gallica.

Seeds of S. gallica present a conspicuous ridge. Bar represents 1 mm.

2.3. Morphological Description of Silene Species by Similarity with Geometric Models

The morphological description includes the analysis of similarity with the cardioid, a comparison between subgenera and the analysis of similarity with other figures related to the cardioid in some species.

2.3.1. Similarity with the Cardioid. J Index Values

The images of Silene seeds resemble cardioids or modified cardioids: Figure 5 presents the silhouettes corresponding to 40 photographs of each species and Table 4 and Table 5 contain respectively the values of J index (percentage of similarity with the cardioid, Model 1) in the species of S. subg. Behenanta and S. subg. Silene. Figure 6 and Figure 7 contain the box plot representations for J index in species of S. subg. Behenanta and S. subg. Silene, respectively.

The silhouettes of 40 seeds of each of the species of Silene grouped according to subgenus (S. subg. Behenantha in red; S. subg. Silene in blue). The seed silhouettes have been scaled to fit a common proportional size without affecting shape.

Box plot representing the values of J index with Model 1 for species of S. subg. Behenantha. Left to right: S. acutifolia, S. conica, S. diclinis, S. dioica Chk, S. dioica Pol, S. latifolia Chk, S. latifolia Pol 1, S. latifolia Pol 2, S. latifolia Pol 2, S. noctiflora, S. pendula, S. uniflora, S. viscosa, S. vulgaris, S. zawadskii. Upper and lower limits of the discontinuous lines represent the maximum and minimum values not atypical (atypical values are outside, below the discontinuous lines). Lower and upper limits of the boxes represent respectively the first and third quartile. The thickened bar in the box is the median.

Box plot representing the values of J index with Model 1 for species of S. subg. Silene. Left to right: S. colpophylla, S. gallica, S. italica, S. mellifera, S. nutans Chk, S. nutans Pol, S. otites, S. saxifraga, S. schafta, S. tatarica, and S. wolgensis. Upper and lower limits of the discontinuous lines represent the maximum and minimum values not atypical (atypical values are outside, below the discontinuous lines). Lower and upper limits of the boxes represent respectively the first and third quartile. The thickened bar in the box is the median.

Table 4

Values of J index with the cardioid as a model in S. subg. Behenantha. Superscript letters indicate the results of Scheffé test: The mean values marked with the same letter in each column do not differ significantly at p < 0.05. n is the number of seeds analyzed.

Species n J Index (Cardioid) Min Max Standard Dev.
S. acutifolia 40 91.2 bcdef 86 94.1 1.84
S. conica 40 92.1 bcd 89 94.6 1.28
S. diclinis 40 90.3 ef 83 93.6 2.63
S. dioica Chk 40 91.5 bcde 88 94.1 1.18
S. dioica Pol 40 91.7 bcde 85 95.8 1.34
S. latifolia Chk 40 92.5 a 83 93.5 1.97
S. latifolia Pol 1 40 92.2 bcd 89 94 0.88
S. latifolia Pol 2 40 92.4 abc 91 95.2 1.18
S. latifolia Pol 2 40 93.5 abc 87 95.8 1.14
S. noctiflora 40 93.6 a 82 93.4 1.09
S. pendula 40 90.8 def 89 94.5 2.1
S. uniflora 40 90.3 ef 83 94.1 2.66
S. viscosa 40 92.6 ab 83 94.5 1.33
S. vulgaris 40 91.18 cdef 83 94 2.15
S. zawadskii 40 90.02 f 83 94.5 2.57

Table 5

Values of J index with the cardioid as a model in Silene species (subg. Silene). Superscript letters indicate the results of Scheffé test: the mean values marked with the same letter in each column do not differ significantly at p < 0.05. n is the number of seeds analyzed.

Species n J Index (Cardioid) Min Max Standard Dev.
S. colpophylla 40 85.6 e 77.7 91.1 2.96
S. gallica 40 90.0 b 85.9 93.1 1.51
S. italica 40 90.1 ab 85 92.7 1.74
S. mellifera 40 89.8 bc 81.2 94 2.31
S. nutans Chk 40 91.6 a 83.7 95.2 2.43
S. nutans Pol 40 91.6 a 83.8 94.9 2.46
S. otites 40 90.7 ab 85.9 93.9 2.11
S. saxifraga 40 90.3 ab 85.2 94.2 1.73
S. schafta 40 82.3 f 74.9 90.3 2.88
S. tatarica 40 87.8 d 81.6 90.9 1.64
S. wolgensis 40 88.3 cd 81.8 92.2 2.28

All the species in S. subg. Behenantha had mean values of J index superior to 90 ( Table 4 ). The standard errors had the higher values in S. diclinis, S. uniflora and S. zawadzkii corresponding to notable variation in seed shape in these species (see Figure 6 and, for example, the 40 seeds of S. diclinis in Figure A3 ). The analysis includes two seed stocks of S. dioica collected in Poland and Czech Republic, and four of S. latifolia, three in Poland and one in Czech Republic. The values of J index obtained with the cardioid (Model 1) were similar in each pair of seed stocks of the same species with the only exception of S. latifolia Pol 1 and S. latifolia Chk.

Five species of S. subg. Silene gave mean values of J index below 90 (S. colpophylla, S. mellifera, S. schafta, S. tatarica, and S. wolgensis; Table 5 ). The seeds in S. colpophylla, S. tatarica, and S. wolgensis are markedly asymmetric ( Figure 1 ), and thus different to a symmetric figure like the cardioid. The seeds of S. schafta have notable surface protuberances and many of them are also asymmetric ( Figure 2 ), while S. mellifera seeds are variable in size and shape. The remaining species resemble the cardioid with values of J index comprised between 90.0 in S. gallica and 91.6 in S. nutans. The values of J index in the two seed stocks of S. nutans are similar.

2.3.2. Comparison between Subgenera

The seeds of S. subg. Behenantha were larger than seeds of S. subg. Silene ( Table 6 ). S. subg. Silene had displayed higher aspect ratio values. Mean circularity was similar in both subgenera while roundness and J index were higher in S. subg. Behenantha. Variation rate (standard deviation compared to mean values) was lower for aspect ratio, circularity and roundness when compared to the measurements related with length and area (area, perimeter, length, width), and still lower in J index.

Table 6

Comparison between subgenera of the mean values representative of size and shape: Area (A), perimeter (P), length of the major axis (L), length of the minor axis (W), aspect ratio (AR is the ratio L/W), circularity (C), roundness (R) and percent similarity with the cardioid (J index M1). Values of A are given in mm 2 ; P, L, and W, in mm. Standard deviation values are given in parentheses. Superscript letters indicate the results of Scheffé test: The mean values marked with the same letter in each column do not differ significantly at p < 0.05. n is the number of seeds analyzed.

n A P L W AR C R JI M1
S. subg Behenantha 600 1.18 a (0.4) 4.44 a (0.8) 1.33 a (0.24) 1.11 a (0.19) 1.19 a (0.06) 0.73 a (0.07) 0.84 a (0.05) 91.8 a (2.09)
S. subg. Silene 400 0.7 b (0.22) 3.5 b (0.69) 1.07 b (0.17) 0.86 b (0.13) 1.25 b (0.08) 0.73 a (0.1) 0.8 b (0.05) 88.7 b (3.4)
2.3.3. Similarity with Other Figures Related the Cardioid

The cardioid (Model 1) has already been tested in model plants and other species [34,35,36,37,38,39], and it was applied in general to all species tested in this work. The comparisons of the seed images with the cardioid gave good results (J index superior to 90) with many species in both subgenera, but our interest is to obtain models specific for particular species, i.e., that give high values with one species and low in the others. The observation of the composed images of the seed silhouettes ( Figure 5 ) suggested that other cardioid-derived models could fit better the shape of some species. Models 2, 3 and 4 were specifically designed to increase values of J index in particular species. They were obtained by simple algebraic modifications of the cardioid equation (see Materials and Methods). Models 2 and 4 where designed considering those seeds whose ventral side is flatter than a cardioid, such as S. diclinis, S. latifolia, and S. noctiflora. Model 4 is slightly thinner than Model 2, and this can favor the resemblance to particular species, such as S. diclinis, while decreasing similarity to others, more rounded, such as S. noctiflora. Model 3 was designed based on the peculiar structure of the seeds of S. gallica, with a pronounced entry around the micropillar region. J index was calculated for the new models with the species resembling more to each of them. The results are shown in Figure 8 and Table 7 and Table 8 .

Comparison of seed shape in Silene species with geometric models. The first row shows the cardioid (Model 1; 1A), the composed image of forty silhouettes (1B) and representative seeds of S. viscosa (J index with Model 1 = 92.6; 1C). The second row shows the flattened cardioid (Model 2; 2A), the composed image of forty silhouettes (2B) and representative seeds of S. noctiflora (J index with Model 2 = 94.4; 2C). The third row contains the open cardioid (Model 3; 3A), the silhouettes (3B) and seed images of S. gallica (J index with Model 3 = 90.4; 3C), and the fourth row contains the flattened and elongated cardioid (Model 4; 4A), silhouettes (4B) and images of S. latifolia (J index with Model 2 = 93.7; 4C). Bars represent 1 mm.

Table 7

Comparison between values of J index with Model 1 (M1; cardioid), Models 2 and 4 in seed images of S. diclinis, S. latifolia and S.noctiflora. Standard deviation values are given in parentheses. Superscript letters indicate the results of Scheffé test: The mean values marked with the same letter in each column do not differ significantly at p < 0.1 (S. diclinis), or at p < 0.05 (S. latifolia and S. noctiflora). n is the number of seeds analyzed.

J Index Values S. diclinis
(n = 40)
S. latifolia
(n = 160)
S.noctiflora
(n = 40)
M1 90.3 c
(2.66)
92.6 b
(1.46)
93.6 a
(1.11)
M2 91.2 c
(2.18)
93.0 b
(1.68)
94.4 a
(1.14)
M4 91.5 b
(2.42)
92.5 a
(1.97)
89.6 c
(2.53)

Table 8

Values of J index with Model 1 and Model 3 in S. conica, S. gallica, and S. otites. Superscript letters indicate the results of Scheffé test: The mean values marked with the same letter in each column do not differ significantly at p < 0.05 (S. conica and S. otites) or p < 0.16 (S. gallica). n is the number of seeds analyzed.

J Index Values S. conica
(n = 40)
S. gallica
(n = 40)
S. otites
(n = 40)
M1 92.1 a (1.29) 90.0 b (1.52) 90.7 b (2.13)
M3 86.2 c (2.22) 90.4 a (1.01) 88.8 b (2.48)

Model 4 represents an improvement respect to Model 1 for the description and quantification of seed shape in S. diclinis. The value obtained with Model 4 in S. diclinis is remarkable considering that the seeds of this species present variable shapes, with some of them resembling the model ( Figure 9 ) and others different from it (see the image composition in Figure A3 ). In addition, the values of J index with Model 4 also decreased for other species (see for example the values obtained with Model 4 with S. noctiflora in Table 7 ), this result adding value in favor of the specificity of this model for S. diclinis.

From left to right: The silhouettes of 40 images of S. diclinis seeds merged together in one image, Model 4 and three representative seeds of this species. Bar represents 1 mm.

With Model 1 higher values of J index were obtained in S. conica than in S. gallica or S. otites, while the J index obtained with Model 3 was higher in S. gallica. From the value of 90 obtained with Model 1, the J index value increased in S. gallica to 90.4 with Model 3 ( Table 8 ). Thus, Model 3 represents an improvement for the description and quantification of seed shape in S. gallica. Not only because the values of J index increased for this species, also because they decreased with other species.

2.4. Multivariate Analysis

A PCA with the distribution of all the variables clustered them into two main groups in dimension 2. Those from direct measurements are found in the positive values of the ordinate axis and those fitting geometrical models are found in the negative values of the ordinate axis ( Figure A4 ). Geometrical models from complex calculations are more informative, and less influenced by the high levels of homoplasy. For this reason, a new PCA using only the variables belonging to geometrical models was performed. In Figure 10 we see the PCA graph with an explained variance of 96.7%. Red circles (with names in red color) represent species from S. subg. Behenantha, blue triangles (with names in blue color) represent species from S. subg. Silene. Variables corresponding to the geometrical models are represented in black.

Principal Component Analysis showing the distribution of representative species belonging to S. subg. Behenantha (red) and S. subg Silene (blue) regarding geometrical models based on seed morphological data.

In the dimension 2 of the PCA, J index and roundness are present together in the negative values of the ordinate axis. Regarding the different species, their mean distribution for S. subg. Behenantha and for S. subg Silene is opposite in both dimensions. In dimension 2, the same distribution is observed for S. subg. Behenantha and the variables Roundness and J index, while S. subg Silene and variable Circularity are found in the positive values of the ordinate axis. It is important to highlight that in the PC, those stocks from the same species share the same axis, including S. latifolia, which practically overlapped this distribution. According to the MANOVA ( Table A1 ), we found significant differences when the analysis was applied according to the subgenera.

3. Discussion

Seed morphology in the genus Silene has been studied for decades as it provides important keys for the taxonomy and understanding the evolution of this genus [5,6,25,27,30,31]. In the genus Silene, seed morphological features like shape, structure of lateral face including macro and micromorphological traits have been useful for species identification and classification with consistent data for species in sections Conoimorphae, Melandriformes, and Sclerocalycinae [31].

We present here an original approach to seed morphology in Silene species based on the comparison of the seed images with geometric figures taken as models. This work includes data from twenty six seed stocks belonging to twenty one species of two subgenera, S. subg. Behenantha and S. subg. Silene, including the seeds of S. diclinis, an endangered species [10]. First, a general description based on the measurements of seed area, perimeter, length, width, aspect ratio, circularity, and roundness, is presented. Distinctive characters relevant for seed shape description include the texture of the seed surface, seed symmetry, presence of a conspicuous ridge and presence and characteristics of the dorsal face. These are useful for the identification of some species and may be associated with more general morphological patterns. Our main objective was to describe the shape of seeds based on the comparison of seed images with geometric figures as it was done in other species [32,33,34,35,36,37,38,39,40,41,42,43,44].

Geometrical models for the description of seed shape in Silene seeds have been used for the first time in this work and include the cardioid and three figures related to it. The fact that in the distribution of the different parameters, those belonging to geometrical models clustered together, supports the idea of using geometric figures as models for seed morphological analysis. The percentage of similarity of forty seed images with the cardioid (J index) has been calculated. J index values were more stable than the other measures considered (area, perimeter, length, width, aspect ratio, circularity and roundness) based on two arguments: First, the values of their standard deviations were low in relation to the means; and second, when J index was compared with the values of seed stocks from different origins belonging to the same species (S. dioica, S. latifolia) this values were generally held with a single exception. Our results indicate that the association of seed morphology with geometrical figures is very robust and could be used for classification purposes [43,44,45].

The mean values of J index were higher in S. subg. Behenantha than in S. subg. Silene. This being related with higher variation and more frequent presence of asymmetric seeds in the latter. In general, asymmetric seeds are expected to give lower values of J index with symmetric models. J index showed the same distribution with subgenus in both, PC1 and PC2. This confirms the observation of Rohrbach [25] (p.49) that seed shape offers the most reliable indicator for the differentiation of many species. Not only the distribution in the PCA showed differences, but also this result was supported by the MANOVA analysis, which showed statistical differences as well. This result indicates that values of a geometrical index (cardioid-shape in this case) may be useful to find differences in subgenera of Silene. Morphological traits have been usually applied for phylogenetic classification in plants [46,47]. On the other hand, Silene possess a high level of homoplasy in morphological characters, which leads to special difficulties in the phylogenetic interpretation [14]. As a matter of fact, the diagnostic characters of the recognized taxonomic groups of Silene significantly overlap, and they may be characterized by a set of combined morphological features especially related to the inflorescences, indumentum of the calyx and their morphology in flower and in fruit [14]. In our study, homoplasy is observed in S. uniflora and S. zawadzkii which do not follow the trend of subg. Behenantha. Several taxa from S. subg. Behenantha also show classification problems in phylogenetic analysis and specific morphological studies have been recommended [14]. For this reason, we suggest geometric indexes applied to seed morphology as a tool. Combined with molecular phylogenetic data, geometric indexes could be used for phylogenetic classification, as it has been proposed in other plant species of difficult classification [48].

Silene latifolia seeds showed a good adjustment to Models 2 and 4. This is of particular interest because sexual determination in this species is based on a system of remarkable similarity to the human XY pair, but where the Y chromosome is more susceptible to analysis [49]. This species is considered a model for the study of evolution and chromosome modification [50], and it may be of interest to evaluate diverse populations in this species for their morphology, as well as to correlate seed shape with chromosomic variations and genetic alterations. Model 4 describes seed shape in S. diclinis, one of the five endangered species of Silene in the Flora of the Iberian Peninsula. High similarity to Model 3 is associated to S. gallica with the presence of a ridge. It could be relevant to check if these two characteristics may coincide in other species of Silene, and check for the presence of seeds resembling Model 3 in other species.

The presented method allows the description of seed morphology based on geometric models. It is reproducible and useful to differentiate between seeds of different species providing a new approach for taxonomic studies of the genus Silene. With this basis the application of the adjectives reniform, circular and globular should be made with reserve. Reniform means kidney-shaped, but the kidney is not a geometric figure defined by an equation that can be objectively reproduced for comparison with other objects. In addition, a kidney, a circle or a globule refer to three dimensional figures and, in so far, there is no way to measure a three-dimensional shape in seeds. In comparison with elliptic Fourier analysis this method compares bi-dimensional seed shape with a figure of reference, thus the results are not purely numerical or statistical, but also testable directly by visual observation. A high value in J index means that the seeds of a species have a given shape, defined by the similarity to a model; and the difference detected between two species or varieties as the result of a statistical test means that there are morphological differences related with a different degree of similarity to a given model between the species or varieties tested. Thus, the results of a statistical test are compared directly with visual information given by the models.

Quantification of J index in a seed lot requires a certain degree of homogeneity in the sample under analysis, which may be difficult in seeds under stress conditions, stored for long time and for those belonging to endangered species, due to reduced seed availability, or absence in extreme cases [51]. In general, good adjustments to models were found in species with lower variation in shape. The case of S. diclinis is exceptional with a good adjustment to Model 4 associated with high diversity in shape. In the sample analyzed of this species there is a clear difference between seeds conforming to the model and others diverging from it. In the attempt to improve reproduction of this species it may be interesting to check for physiological differences associated with the morphological types, as well as to detect the relative frequency of morphological types in population studies. This opens up the possibility that the method here reported may be of help for the analysis and identification of different natural populations of this and other protected species. Perhaps this first indication of a certain morphological separation in S. diclinis could be indicating some separation of the genetic information of these populations that could be of interest for conservation purposes in the future.

Seed shape is closely related to those gene families playing a role in the proper morphological development, and changes in these genes greatly affect seed morphology [52]. Morphological seed parameters can be used to delimit genera and are also used together with sequencing data in phylogeny analyses [53,54,55]. Our results indicate a common morphological pattern conserved by seeds from subgenus Behenantha which shows some similarities to the phylogenetic gene tree analysis in Silene [48].

The method presented links seed morphology to functional traits through a geometrical approach. Not all the criteria used in Phylogeny have the same informative value. Individual morphological traits may be homoplastic which means very low informative value. On the other hand, a geometric model includes several traits which may overcome this problem. In addition, area, perimeter, length or width don’t provide information on shape. Circularity or roundness are more informative in this regard, but their usefulness depends on the similarity of the figure with a circle, being of scarce utility in low values. In the cases of a good adjustment to a model, J index provides valuable information in a single measurement and it may be, in consequence, useful in Phylogeny studies. Other analyses similar to the reported here for Silene species have revealed differences in species belonging to other families [38,39,40,41,42,43,44,45] and describe a morphological character that could support the taxonomic treatments based on molecular phylogenies of the groups considered.

The combination of different seed traits could be used as a tool for seed management in genebanks, avoiding in some cases more expensive and time-consuming analysis. Likewise, variations in seed morphology due to stress because of climatic change could be identified on those individual seeds that do not fall into the range of the geometrical index assigned to a concrete plant species. A correct evaluation of the different morphological indexes could be a suitable tool in phylogenetic analyses.

4. Materials and Methods

4.1. Species of Silene

We examined seeds of 26 seed stocks belonging to 21 species of Silene conserved in the plant collections of the Laboratory of Plant Ecology and Adaptation, University of Lodz (Poland) and the Botanic Garden of the University of Warsaw and laboratories of the Academy of Sciences of Brno (Czech Republic). Species are listed in Table 9 with an indication of their locations of origin (when known). Of them, three were annuals (S. conica, S. noctiflora, S. pendula), one annual or biennial (S. gallica), four biennials or perennials (S. colpophylla, S. otites, S. viscosa, S. wolgensis) and thirteen perennials [29,56,57,58,59,60,61,62,63,64,65] (see Table 9 ). S. latifolia is mentioned as being perennial and never annual by Bojňanský and Fargašová [29], but as being annual or short-lived perennial by efloras.org [56]. Species nomenclature was adopted after Oxelman et al. [8] and Plants of the world online [66]. For the classification of species as S. subg. Behenanta or S. subg. Silene, we followed Sileneae classification [67].

Table 9

List of species described, indicating the origin of the seeds.

Species Lab. Origin Annual, Biannual or Perennial
S. acutifolia Link ex Rohrb. CzR u (unknown) P [57]
S. colpophylla Wrigley CzR France B,P [58]
S. conica L. CzR Germany A [59]
S. diclinis (Lag.) M.Laínz CzR Pla de Mora (Spain) P [60]
S. dioica (L.) Clairv. CzR/Pol Tišnov (CzR)/u P [29,58]
S. gallica L. CzR u AB [29,61]
S. italica (L.) Pers. CzR u BP [29,61]
S. latifolia Poir. CzR/Pol
(3 seed stocks)
Panenská Rozsíčka (CzR)/Dubidze(Pol) P [29,56,61]
S. mellifera Boiss. and Reut. Pol u BP [62]
S. noctiflora L. CzR Kuřim (CzR) A [29,61]
S. nutans L. CzR/Pol u P [29,61]
S. otites (L.) Wibel CzR Rohatec (CzR) P [58,63]
S. pendula L. CzR u A [29]
S. saxifraga L. CzR u P [29]
S. schafta S.G.Gmel. CzR u P [29]
S. tatarica (L.) Pers. CzR u P [29,61]
S. uniflora Roth Pol u P [64]
S. viscosa (L.) Pers. CzR Rohatec (CzR) BP [61]
S. vulgaris (Moench) Garcke CzR Lomnička (CzR) P [29,61,65]
S. wolgensis (Hornem.) Otth CzR Bashkortostan (RU) B [29]
S. zawadzkii Herbich CzR u P

4.2. Seed Images

Photographs were taken with a Nikon Stereomicroscope Model SMZ1500 equipped with a camera Nikon DS-Fi1 of 5.24 megapixels. The seeds were oriented with the micropyle to the right ( Figure 11 ). Composed images containing 40 seeds per accession were prepared with Corel Photo Paint and are stored in: https://zenodo.org/record/4057708#.X3LRpRRxeM8 and https://zenodo.org/record/4035649#.X3LRShRxeM8.

An image of Silene (S. latifolia). The perimeter (P) is indicated by the blue color. Length of the major axis (L) and length of the minor axis (W) are indicated. The hilium is included as a part of the seed in all measurements.

Confocal images were obtained with a Leica DM IRB TCS SP2 confocal microscope and are limited to a total of 13 species. Each figure contains the mean projection of a series of 20 images.

4.3. Surface Characteristics and other Structural Properties of Seeds

The presence and types of colliculae were evaluated in the stereomicroscopic images and by confocal microscopy. Other structural properties analyzed in the sets of stereomicroscopic images are: (1) seed asymmetry, (2) presence of ridges, and (3) pronounced dorsal surface (plane or concave).

4.4. General Morphological Description by Image Analysis

Photographs were used to obtain the area (A), perimeter (P), length of the major axis (L), length of the minor axis (W), aspect ratio (AR is the ratio L/W), circularity (C) and roundness (R). All the measurements are obtained with ImageJ program [68] by the conversion of pixel units to length or surface units (mm or mm 2 ) using a ruler as a reference. The circularity index and roundness were calculated as described [69]. Circularity is the ratio (4π × A)/P 2 , while roundness is (4 × A)/πL 2 .

4.5. Comparison with Geometric Models: Calculation of the J Index

A new approach to the morphological description of seed shape in Silene species is based in the comparison with geometric figures used as models. A set of four models is described and applied for the first time in this work. These are as follows ( Figure 12 ):

Models used in the geometric description of seeds from Silene species. Model 1, the cardioid curve, corresponding to Equation (1), was obtained from Mathematica [70]. Models 2 to 4 were obtained by the modification of Equation (1) searching for similarity with the outlines of particular seed images. Model 1 is superimposed with thin-line overlay in Models 2 to 4 to appreciate the differences.

Model 1: The cardioid curve is described by the equation:

Model 2: A flattened cardioid, with a reduced discontinuity in the region of the hilium in relation to Model 1, is given by:

Model 3: An open cardioid is given by:

Model 4: A flattened and elongated cardioid, also with a reduced discontinuity in the region of the hilium but thinner, less rounded, than Model 2 (see Figure 12 ) is given by:

In all four equations a and b are real positive parameters. Models 2 and 4 were selected visually by their similarity with the silhouettes of S. latifolia and S. noctiflora. Model 3 was selected for its similarity to S. gallica. Mathematica code for the models is available at: https://zenodo.org/record/4120172#.X5nooUeg-M9.

Graphic compositions were done departing from one image containing forty seeds for each seed stock and subsequently elaborated with Corel PHOTO-PAINT X7. For the comparison of a group of seeds with the models and quantification of J index, the geometric figures used as models were superimposed to each seed image in the group of forty, searching a maximum adjustment between both shapes, the seeds and the model. An image scaled of the cardioid was adapted to the seed images and the percent of similarity between the image scaled of the cardioid and the seed image was estimated. Three graphic documents were kept for each composition: (1) A file in PSD format with the forty seeds and the geometric figure adapted to each of them, in which it is possible to make changes and corrections; (2) A file in JPG format with the geometric models in black, that served (later on) to obtain total area (T) with ImageJ, and (3) Another file in JPG format with the geometric models in white, that was used to obtain the values of area shared between the geometric figure and the seed image (S) in ImageJ. All the process of image composition with seeds and models was done in Corel PHOTO-PAINT X7, while area quantification was calculated in ImageJ. Figure 13 presents examples of the adjustment between seed images and the geometric models with indication of the areas measured for the calculation of the J index.

Representative samples of the composition of seed images and the geometric models used in the calculation of the J index. Four seed images are presented with the model superimposed in white (left, above) and the same images with the model in black (left, below). In the center (red colored), the corresponding images after modification in ImageJ (Image Type: 8 bit; adjust threshold). This way, the shared areas are observed and can be quantified as the larger areas in red (above), while total area is obtained as the total surface limited by the red limit (below). In the right, the four silhouettes above correspond to shared (S) and the four silhouettes below represent total area (T). The surfaces are observed and quantified with ImageJ. J index is the ratio S/Tx100.

The images used are provided as Supplementary Files in Zenodo (https://zenodo.org/record/4057744; 4057740; 4057809; 4057810; 4020369; 4020382; 4057838; 4057831).

Seeds from all species were compared to Model 1; in addition, S. latifolia and S. noctiflora were compared to Models 2 and 4, because their seeds resemble more these models and in consequence, J index values obtained are higher. For the same reason, S. conica, S. gallica, and S. otites were compared to Model 3. To have good definitions of seed shape it is important to obtain high values of J index (high similarity with a given model means a good definition of a given shape). The areas corresponding to the two regions needed for the calculation of J index were obtained with ImageJ: the region that is common to the model and the seed image (“Shared” area, S) and the total region occupied by both areas, the seed and the model (“Total” area, T). The J index is defined by:

S is represented in Figure 13 as the area in the four silhouettes above (right), while T is the total area in the four silhouettes below (right). Note that J index is a measure of seed shape, not of its area. It ranges between 0 and 100 decreasing when the size of the not-shared region grows and equals 100 when the geometric model and the seed image areas coincide. High value of J index (high similarity with a given model) means a precise definition of seed shape for a particular species.

4.6. Statistical Analysis

The distribution of the raw data was skewed, and therefore had to be transformed to achieve homoscedasticity. One-way ANOVA was used to show significant differences between species for the measured variables, followed by Scheffé post-hoc tests to provide specific information on which means were significantly different from one another. This analysis was done with software IBM SPSS statistics v25 (SPSS 2017).

Multivariate Analysis

Multivariate analysis was done in R Studio, V.1.2.1335 [71]. Raw data was analyzed to check if the dataset was homoscedastic, what it means homogeneity of variance and normal distribution. A normalization of raw data to lower the weight of outliers prior to start with multivariate analysis was required and achieved using powerTransform function. A lambda value to transform the data was used according to mathematical procedure. This function uses the maximum likelihood-like approach of Box and Cox [72] to select a transformation of a univariate or multivariate response for normality. Principal Component Analysis is a procedure for dimension reduction used to see the total variation of variables and individuals in multidimensional data. The distribution of the species is based on their value for each parameter combined. This offers the possibility to gather species among them and to specific traits or, like in this analysis, to specific geometrical model indexes. In order to identify the statistical differences, a multivariate analysis of variance (MANOVA), a procedure for comparing multivariate sample means using the covariance, was applied. From the morphological description data, only those considered shape indexes (circularity index, roundness and J index) were used.

5. Conclusions

Images taken from well oriented seeds of Silene are described by comparison with a cardioid. J index is the percentage of similarity between the cardioid and the seed image. J index values are higher in species of S. subg. Behenantha than in S. subg. Silene.

Geometric figures derived from the cardioid by slight modifications in the corresponding algebraic equation adjust well to particular species of Silene. Thus, specific models are described for S. gallica as well as the endangered species S. diclinis.

The quantification of seed shape based in the comparison of Silene seeds with cardioid or cardioid-derived figures opens the way to new semi-automated methods of phenotyping. Variations in seed morphology due to stress in general, or climatic change, could be identified on those individual seeds that do not fall into the range of the geometrical index assigned to a concrete plant species. The asymmetry of the data does not affect the main conclusions and supports the hypothesis that in Silene the seeds of a population of the same species may follow different morphological patterns.

Acknowledgments

The authors would like to thank Izabela Kirpluk (Botanic Garden of the University of Warsaw) and Adam Kapler (Polish Academy of Sciences Botanical Garden in Warsaw Powsin) for their help in collecting material. We acknowledge support of the publication fee by the CSIC Open Access Publication Support Initiative through its Unit of Information Resources for Research (URICI). We thank Ana Juan, from the Department of Environmental Sciences and Natural Resources (University of Alicante) for critical reading of the manuscript, Ángel Tocino, from the Department of Mathematics (University of Salamanca) for help in design and parametrization of the models and Francisco Peinador for English corrections.

A giant, 165-inch TV for Optimus Prime and his fans

Ever seen a 60-inch TV and said, “Naah, too small for me?” Worry not. A TV straight out of the Transformers is here to help. Provided you have deep pockets. Like real, deep pockets.

Austrian company C Seed has just brought a monstrosity to life with their latest 7ft-high 165-inch foldable 4K MicroLED TV called the C-Seed M1.

CONTINUE READING BELOW

The TV is aimed at the elite class (because who else can afford a $400,000 TV?) to fit their decor as it emerges out of the floor and unfolds its four panels to make up a humongous 4K HDR display.

The company has implemented Adaptive Gap Calibration Technology which aims to create a seamless effect, making the multi-sectional design resemble one giant 165-inch TV.

People familiar with C-Seed know that the Austrian company specialises in luxurious indoor and outdoor TVs, many of which are the size of a small bus. The M1 is the firm’s latest masterpiece and will be available in black, gold, silver and ‘Titan’ grey from July 2021 for about $400,000 or Rs 2,94,38,560.

Even if we eliminate the price of the TV, buying one might be easy, but burying a 1,500kg TV under the floor of your house might be the hard part.

S-nitrosylation triggers ABI5 degradation to promote seed germination and seedling growth

Plant survival depends on seed germination and progression through post-germinative developmental checkpoints. These processes are controlled by the stress phytohormone abscisic acid (ABA). ABA regulates the basic leucine zipper transcriptional factor ABI5, a central hub of growth repression, while the reactive nitrogen molecule nitric oxide (NO) counteracts ABA during seed germination. However, the molecular mechanisms by which seeds sense more favourable conditions and start germinating have remained elusive. Here we show that ABI5 promotes growth via NO, and that ABI5 accumulation is altered in genetic backgrounds with impaired NO homeostasis. S-nitrosylation of ABI5 at cysteine-153 facilitates its degradation through CULLIN4-based and KEEP ON GOING E3 ligases, and promotes seed germination. Conversely, mutation of ABI5 at cysteine-153 deregulates protein stability and inhibition of seed germination by NO depletion. These findings suggest an inverse molecular link between NO and ABA hormone signalling through distinct posttranslational modifications of ABI5 during early seedling development.

Introduction

In the plant life cycle, the development of a new seedling depends on both appropriate timing of seed germination and the perception of environmental conditions. Consequently, germination and early seedling development must be tightly regulated by exogenous and endogenous signal molecules. Among them, abscisic acid (ABA) plays an important role in the inhibition of seed germination 1 and in post-germinative seedling arrest under unfavourable environmental conditions 2 . ABA-dependent growth arrest after germination relies on the basic leucine zipper-type transcription factor ABI5 (refs 3, 4, 5). In many plant developmental processes and stress responses, the precise changes in cellular status are regulated not solely by ABA signalling but by a complex network of ABA and other signalling pathways. Several examples of cross-talk between ABA and the gaseous signalling molecule nitric oxide (NO) have been recently reported. These interactions include the link of SnRK2.6/OST1 S-nitrosylation to the negative regulation of ABA signalling in the stomata 6 and the NO regulation of ABI5 transcription through control of group VII ethylene response factors (ERFs) stability 7 . NO is a signalling molecule involved in a variety of physiological processes during plant growth and development. Extensive research has shown that NO affects seed dormancy, seed germination and ABA sensitivity 8,9 , as evidenced by exogenous application of NO donors 10 or by genetic analysis of mutants with altered endogenous NO levels in Arabidopsis 11 . However, despite the abundant involvement of NO in different plant cell signalling pathways, the actual knowledge about its direct targets is poorly understood. A key feature of NO biology is the posttranslational modification of cysteine thiol to form nitrosothiols (S-nitrosylation) in target proteins 12 . In animals, this posttranslational modification has also been related to protein degradation via the ubiquitin-dependent proteasome pathway 13 . Here we establish a molecular mechanism for NO and ABA antagonism in the regulation of seed germination and post-germinative development through ABI5 protein stability. Genetic analysis identified that abi5 mutants are insensitive to NO scavenging during seed germination. ABI5 protein levels are high in NO-deficient mutant backgrounds and low in NO-overaccumulating plants. S-nitrosylation of ABI5 at Cys 153 facilitates its degradation and promotes seed germination. Conversely, mutation of ABI5 Cys 153 reduces protein degradation through CULLIN4 (CUL4)-based and KEEP ON GOING E3 (KEG) ligases, and deregulates the inhibition of seed germination by NO depletion. Thus, ABI5 is regulated through the antagonistic action of ABA and NO, as evidenced by the synergistic effect of S-nitrosoglutathione (GSNO) on ABI5 destabilization. These findings suggest an inverse molecular link between NO and ABA hormone signalling through distinct posttranslational modifications of ABI5 during gene regulation of early seedling development.

Results

Identification and genetic characterization of gap mutants

NO affects seed dormancy, seed germination and ABA sensitivity 8,9 , as shown by exogenous application of NO donors 10 or by genetic analysis of mutants with altered endogenous NO levels in Arabidopsis 11 . By exploiting the cross-talk between ABA and NO in the transition from dormancy to germination, we performed a genetic screen using (+)-S-ABA coupled to the effect of the NO scavenger 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (cPTIO). We isolated 7 (+)-S-ABA-insensitive mutants that were able to germinate on 3 μM (+)-S-ABA and also display a cPTIO-insensitive phenotype (Supplementary Fig. 1a). These mutants were named gap (germination in ABA and cPTIO). Allelism tests indicated that these fell into two different loci, five of these mutants corresponding to new abi5 alleles verified by candidate gene sequencing (Supplementary Fig. 1b). Germination of Col-0 wild-type seeds was delayed by 100 μM cPTIO (Fig. 1a,b), in agreement with previous reports 9 ; however, germination of the abi5 mutant alleles was less affected by NO depletion than the wild type. Hence, these results showed that abi5 mutants were insensitive not only to ABA but also to NO scavenging by cPTIO during seed germination.

(a) Insensitivity of abi5 mutants to NO scavenging by cPTIO during seed germination. Photographs of 2-day-old germinated seeds after imbibition of wild type (Col-0) and the ABA-insensitive abi5-1 and abi5-7 mutants, in the absence of (Control) or the presence of 100 μM cPTIO. Scale bar, 1 mm. (b) Germination of wild-type (Col-0), abi5-1 and abi5-7 seeds in media containing 0 and 100 μM cPTIO after 2 days. Error bars represent±s.e. (n=3). Asterisk indicates significant differences compared with Col-0 (Control) (t-test, P<0.05). (c) Co-localization of ABI5 expression, protein localization and NO production. pABI5:ABI5-GUS seeds were stratified for 3 days at 4 °C and grown for 1 to 2 days at 21 °C on MS agar plates and then subjected to DAF-2DA incubation or GUS staining after treatment with NO scavenger (cPTIO) and donor (SNAP). Arrows indicate high NO accumulation (left), and ABI5 expression and protein localization (middle). Scale bars, 100 μm. (d) qRT–PCR analysis of ABI5 relative transcript abundance in Col-0 seeds untreated (Control) and after treatments with ABA, cPTIO, SNAP and GSNO after 24, 48 and 72 h and in the abi5-1 background. Error bars represent±s.e. (n=3).

NO induces degradation of the ABI5 protein

To extend these findings, we determined the localization of the ABI5 protein (Fig. 1c) and ABI5 transcript (Fig. 1d) in seeds after treatment with ABA, the NO donor S-nitroso-N-acetyl- DL -penicillamine (SNAP) and the NO scavenger cPTIO. ABI5 accumulated to high levels after treatment with ABA for 48 or 72 h and even earlier after NO was depleted (Fig. 1d). In contrast, application of the NO donor SNAP quickly reduced ABI5-GUS levels (Fig. 1c and Supplementary Fig. 2).

To further investigate the role of NO in the regulation of seed germination in Arabidopsis, we examined endogenous NO levels in 2-day-old wild-type seeds by using the fluorescence indicator 4,5-diaminofluorescein diacetate (DAF-2DA). We observed NO-dependent fluorescence in the seed endosperm and rapid elongation zone after radicle protrusion (Fig. 1c), as previously described 14,15 . Application of the NO scavenger cPTIO reduced the DAF-2DA fluorescence pattern in treated seeds, signal being limited to the autoflorescence of the testa tissues (Supplementary Fig. 1c–e). Co-localization of NO burst and ABI5 tissue-specific reduction of ABI5-GUS protein during radicle protrusion (seed imbibition/germination) was confirmed after GUS histochemical analysis using pABI5:ABI5-GUS lines as highlighted in Fig. 1c.

Recently, the N-end rule pathway of targeted proteolysis has been shown to regulate ABI5 expression through group VII ERF transcription factors 7 . To this end, application of MG132 did not transcriptionally induce ABI5 gene expression, as a putative consequence of enhanced group VII ERF stabilization (Fig. 2a).

(a) qRT–PCR analysis of ABI5 relative transcript abundance in Col-0 seeds untreated (C) and after treatments with GSNO, SNAP, cPTIO, cycloheximide (CHX), MG132 proteasome inhibitor and the combinations indicated after 3 h. Error bars represent±s.e. (n=3). (b) SNAP and GSNO treatments promote ABI5 degradation in dormant seeds (DS). Immunoblot analysis of ABI5 protein levels in seed extracts of Col-0 and abi5-1, treated with or without (C) NO scavenger (cPTIO) and donors (GSNO and SNAP), and the MG132 proteasome inhibitor. Actin protein levels are shown as a loading control. (c) Immunoblot analysis of ABI5 protein levels in seed extracts of Col-0 dormant seeds treated with or without (C) NO donor (GSNO), the proteasome inhibitor cocktail (PIC, composed of MG115, MG132 and epoxomicin), cycloheximide (CHX) and the combinations indicated. Actin protein levels are shown as a loading control. (d) GSNO treatment promotes ABI5 degradation in 2 days ABA (5 μM)-treated after-ripened seeds. Immunoblot analysis of ABI5 protein levels in seed extracts of Col-0, treated with or without (C) NO scavenger (cPTIO) and donor (GSNO). Actin protein levels are shown as a loading control. (e) ABI5 protein levels in wild type (WT; Col-0), abi5-1 and NO-deficient (atnoa1-1, nia1;nia2, atnoa1-2;nia1;nia2) mutant backgrounds. Stratified seeds were sown on control MS (Control, left) and 0.1 μM ABA after 48 h (right). Immunoblot analysis of ABI5 protein levels in seed extracts of WT and NO-deficient mutants. Actin protein levels are shown as a loading control. (f) ABI5 protein levels in WT, 35S:AHb1 (H3, H7) and 35S:antiAHb1 (L1, L3) lines. Stratified seeds were sown on MS (Control) and 0.1 μM ABA for 96 h. Immunoblot analysis of ABI5 protein levels in seed extracts of AHb1-overexpressing and -silencing lines. Actin protein levels are shown as a loading control.

The effect of SNAP and the physiological NO donor GSNO in the promotion of seed germination was correlated by the disappearance of ABI5 during germination and post-germinative growth (Fig. 2b–d and Supplementary Fig. 3a–c). Conversely, endogenous NO depletion by the cPTIO scavenger inhibited seed germination and maintained high ABI5 protein levels, similar to ABA (Fig. 2b,d). It is noteworthy that the proteasome inhibitor MG132 or the proteasome inhibitor cocktail (including MG115, MG132 and epoxomicin) restored ABI5 accumulation, even in the presence of NO donors, and prevented seed germination (Fig. 2b,c and Supplementary Fig. 3c). In addition, cycloheximide alongside the MG132 and NO donor/scavenger treatments proved that ABI5 protein stability was being affected (Fig. 2a,c). Collectively, these findings implied that NO function during seed germination was through ABI5 degradation. After this developmental checkpoint and at high concentrations, NO dramatically affected post-germinative seedling growth, inhibiting root growth and development 15 .

ABI5 level is altered when NO homeostasis is impaired

We corroborated the above pharmacological findings by the use of mutants and transgenic lines impaired in NO homeostasis. Thus, NO-deficient atnoa1-2;nia1;nia2 triple mutant impaired in NIA/NR- and AtNOA1-dependent NO biosynthesis in Arabidopsis was hypersensitive to ABA, underscoring its effect on germination inhibition 11 (Supplementary Fig. 3d). Non-symbiotic haemoglobin 1 (AHb1) is an endogenous scavenger of NO, and thus AHb1-overexpressing and -silenced lines contain lower and higher NO levels, respectively 16 . As expected, the ABA response phenotype of these AHb1 lines during seed germination differed from the hypersensitivity to ABA of the AHb1-overexpressing lines to the wild-type germination of the AHb1-silenced lines (Supplementary Fig. 3e,f).

We investigated ABI5 protein accumulation in the Arabidopsis atnoa1, nia1;nia2 and atnoa1;nia1;nia2 mutant seeds, which exhibit decreased levels of cellular NO 11 (Fig. 2e). NO-deficient mutants, defective in either the oxidative or reductive NO synthesis pathways, respectively, or in both pathways, accumulated higher ABI5 protein levels. In agreement with the enhanced ABA sensitivity of NO-deficient mutant backgrounds, ABI5 protein levels were increased with respect to those observed in the wild type (Col-0). To corroborate these findings, AHb1-overexpressing and -silenced lines were also analysed for their ABI5 accumulation pattern (Fig. 2f). Collectively, these findings implied that genetic backgrounds where the endogenous NO levels were enhanced (AHb1-silencing lines) or diminished (atnoa1, nia1;nia2, atnoa1;nia1;nia2 and AHb1-overexpressing lines) displayed altered ABI5 levels, which were mirrored by changes in their ABA response. These pharmacological and genetic approaches suggest that NO promoted ABI5 protein degradation via a proteasome-dependent pathway.

ABI5 is S-nitrosylated in vivo and in vitro

One possible mechanism of NO action in plant tissues is the redox-based posttranslational modification of target proteins through S-nitrosylation. NO is able to reversibly modify thiol groups of specific cysteine residues in target proteins, hence altering protein function 12 . To determine whether ABI5 was S-nitrosylated by NO, the recombinant protein was exposed to either GSNO or SNAP, which are typically used to evaluate S-nitrosylation (that is, SNO formation) in vitro 17 . The formation of SNO-ABI5 was monitored by the biotin switch method 18 . As shown in Fig. 3a–c, ABI5 was S-nitrosylated by either GSNO or SNAP. Furthermore, the addition of dithiothreitol (DTT) strongly reduced the formation of SNO-ABI5, consistent with the presence of a reversible thiol modification. The biotin switch was assayed without ascorbate, to assure that the immune reactivity of the protein was ascorbate dependent (Supplementary Fig. 4a) and to prevent switching SNO for biotin, as the method describes 18 . ABI5 contains four cysteine residues (Cys 56, Cys 153, Cys 293 and Cys 440) that might serve as sites for this redox-based modification (Supplementary Fig. 4b). In-silico prediction suggested that Cys 153 was potentially S-nitrosylated (Supplementary Fig. 4d,e). Mass spectrometry analysis confirmed S-nitrosothiol formation at only Cys 153 of ABI5 (Fig. 3a,b and Supplementary Figs 5 and 6). This residue was therefore mutated individually and the resulting protein was expressed and treated with GSNO before analysis with the biotin switch method. The Cys153 to Ser mutation abolished S-nitrosylation of ABI5 (Fig. 3c). Collectively, these findings indicated that Cys 153 of ABI5 was specifically S-nitrosylated in vitro. This Cys 153 is only present in the closest ABI5 homologue AtbZIP67 (Supplementary Fig. 4c), sharing a seed expression pattern with ABI5 (Supplementary Fig. 7). Together, these data suggest that redox modification by S-nitrosylation may regulate the activity of these b-ZIP transcription factors in plants to govern seed germination and seedling establishment.

(a) Mass spectrometric analyses identify C153 as the S-nitrosylation site. MS/MS spectra of C153 from the tryptic fragment QGSLTLPAPLCR (peptide MS/MS spectra shown with Cys modified by biotin-HPDP). (b) The LC–MS spectra of the corresponding peaks (*562 m/z (3+) and 842,49 m/z (+2)) of this peptide fragment is shown in the inset. (c) The C153S mutation blocks S-nitrosylation of ABI5. In vitro S-nitrosylation of wild-type ABI5 and mutant ABI5C153S recombinant proteins by the NO donors GSNO (200 μM) and SNAP (200 μM). This modification is reversed by treatment with DTT (20 mM). No signal was observed with glutathione (200 μM) treatment showing specificity of the biotin-switch assay. ABI5 protein loading was detected by anti-His antibody. (d,e) S-nitrosylation of ABI5 induced by GSNO in after-ripened seed extracts. Samples were initially immunopurified with anti-biotin before immunoblot analysis of ABI5 protein levels in seed extracts of Col-0 (d), 35S:ABI5 (e) and abi5-1 untreated (C) or treated with the indicated compounds. No signal was observed in the absence of biotin (−Biotin) or after DTT (20 mM) treatment. Actin protein levels are shown as a loading control. (f) In-vivo S-nitrosylation of ABI5 in abi5-1;35S:cMyc-ABI5 and abi5-1;35S:cMyc-ABI5C153S after-ripened seed extracts 24 h after proteasome inhibitor MG132 (100 μM) incubation. Immunoblot analysis of in vivo ABI5 protein levels after immunopurification of S-nitrosylated proteins. No signal was observed in the absence of sodium ascorbate (−Asc) or after cPTIO (1 mM) treatment. Input protein levels were also determined using anti-ABI5 anti-serum.

To understand whether S-nitrosylation of Cys 153 could modulate the function of ABI5, we assessed the effect of this NO-mediated posttranslational modification on the previously reported homodimerization of the ABI5 protein 19 by using a yeast two-hybrid assay. Interestingly, neither the ability of ABI5 to interact (Supplementary Fig. 8) nor the DNA-binding capacity to the ABRE cis-consensus motif (Supplementary Fig. 9) were disturbed by the Cys–Ser mutation or in the presence of NO-related compounds.

To determine whether ABI5 was S-nitrosylated in vitro during seed germination, protein extracts from wild-type and 35S:ABI5 transgenic plants were treated with GSNO and MG132, and assayed by the biotin switch method, then S-nitrosylated proteins were immunopurified (Fig. 3d,e). Protein gel blot analysis of purified proteins probed with an antibody specifically recognizing the ABI5 protein showed the corresponding S-nitrosylation in vitro. In addition, in vivo studies with transgenic ABI5 lines expressing either Myc-tagged wild-type ABI5 or mutant derivatives were analysed during seed germination and treatment with MG132. To this end, endogenous proteins were subjected to biotin-switch analysis, immunoprecipitated and detected with an anti-ABI5 antibody. Wild-type ABI5 was S-nitrosylated during seed germination, but the Cys153Ser mutant was not (Fig. 3f). Therefore, these results suggest that NO may regulate ABI5 redox state by S-nitrosylation at Cys 153 during seed imbibition.

ABI5 is a NO sensor during seed germination

To assess a physiological role for S-nitrosylation of the ABI5 protein during seed germination, ABI5- and ABI5Cys153Ser-overexpressing lines were generated in Arabidopsis by expressing this protein under control of the 35S promoter, in the wild-type and abi5-1 mutant background (Supplementary Fig. 10). The phenotype of the transgenic plants was evaluated in response to NO and ABA, and was compared with that of abi5 mutant and 35S:ABI5 transgenic plants. Intact ABI5– and mutated ABI5Cys153Ser-overexpressing lines exhibited ABA-hypersensitive phenotypes during seed germination and seedling establishment (Supplementary Fig. 10d), demonstrating that ABI5Cys153Ser was able to largely restore the abi5-1 mutation (Supplementary Fig. 10e). To further explore the possible biological consequence of ABI5-SNO formation at Cys 153, ABI5 protein accumulation was monitored in the presence of the protein synthesis inhibitor cycloheximide with or without GSNO (Fig. 4a). The data obtained implied that the mutation of ABI5 at Cys 153, considering similar seed germination stages, impaired NO-promoted ABI5 degradation, resulting in elevated levels of ABI5 protein. CUL4 and KEG are well-known factors that directly bind and regulate ABI5 ubiquitin-mediated proteolysis 20,21,22,23,24 . Accordingly, mutations in either CUL4 or KEG abolished NO-promoted ABI5 protein degradation (Fig. 4b and Supplementary Fig. 11). Intact ABI5 protein interacted with both CUL4 and KEG in the presence of GSNO, whereas ABI5Cys153Ser mutation failed to interact (Fig. 4c,d), supporting that S-nitrosylation triggered ABI5 destabilization through CUL4-based and KEG E3 ligases.

(a) Immunoblot analysis of ABI5 protein levels in 8-day-old seedling extracts of similar germination stages abi5-1;35S:cMyc-ABI5 and abi5-1;35S:cMyc-ABI5C153S, in the presence of cycloheximide (1 mM) and cycloheximide (1 mM) plus GSNO (500 μM) from 0 to 9 h. Actin protein levels are shown as a loading control. (b) ABI5 protein levels in wild type (Col-0), cul4cs, keg4 and abi5-1 mutant backgrounds. Stratified seeds were incubated with 5 μM ABA for 48 h and treated with GSNO (1 mM) for 6 h after ABA removal. Immunoblot analysis of ABI5 protein levels in seed extracts of wild type and mutants. Actin protein levels are shown as a loading control. (c) Co-immunoprecipitation assays between CUL4 and transgenic ABI5/ABI5C153S proteins in the presence of GSNO. Input protein levels were also determined using anti-FLAG and anti-MYC antisera, respectively. (d) Co-immunoprecipitation assays between KEG and ABI5/ABI5C153S proteins in the presence of GSNO. Input protein levels were also determined using anti-HA and anti-MYC antisera, respectively. (e) NO-insensitive inhibition of seed germination to NO scavenging in 35S:ABI5C153S lines as compared with 35S:ABI5 plants. Total seed germination of wild type (Col-0), abi5-1, abi5-7 and two (1, 2) 35S:ABI5– and 35S:ABI5C153S-independent lines grown for 2 days on MS agar plates untreated (Control) or supplemented with 50 and 100 μM of the NO-scavenger cPTIO. Values represent the mean ±s.e. (n=3). Asterisks indicate significant differences compared with 0 μM cPTIO (t-test, *P<0.05, **P<0.01). (f) ABI5 levels in 35S:ABI5 and 35S:ABI5C153S transgenic lines used for the germination assay. Immunoblot analysis of ABI5 protein levels in seed extracts. Actin protein levels are shown as a loading control. (g) NaCl- and mannitol-hypersensitive inhibition of post-germinative growth in two 35S:ABI5 and 35S:ABI5C153S lines as compared with wild-type plants. Seedling growth of wild type (Col-0), abi5-1, 35S:ABI5 and 35S:ABI5C153S lines grown for 9 days on MS agar plates untreated (Control) or supplemented with 100 mM of NaCl and 250 mM of mannitol. Values represent the mean±s.e. (n=3). Letters indicate significant differences compared with wild-type (Col-0) (a), 35S:ABI5-1 (b), 35S:ABI5-2 (c), abi5-1;35S:ABI5-1 (e), abi5-1;35S:ABI5-2 (d), (t-test, P<0.05).

To determine the possible impact of the ABI5Cys153Ser mutation on NO sensing during seed germination and seedling establishment, we treated intact ABI5– and mutated ABI5Cys153Ser-overexpressing lines with NO scavengers (cPTIO) and donors (SNAP) (Fig. 4e,f and Supplementary Fig. 10f). As depicted in Fig. 4e,f, inhibition of seed germination by NO depletion during radicle protrusion was enhanced in ABI5 but not in ABI5Cys153Ser overexpressors and also in the establishment of a new plant (Supplementary Fig. 10f), resulting in a prominent decrease in the rate of germination. In addition, there was a significant increase in seedling establishment after NO treatment in the ABI5-overexpressing lines relative to the overexpression of the ABI5Cys153Ser mutant version that instead was able to maintain growth arrest (Supplementary Fig. 10f). Germination of the transgenic plants under unfavourable conditions such as high salinity and hyperosmotic stress showed NaCl- and mannitol-hypersensitive inhibition of post-germinative growth in 35S:ABI5 and 35S:ABI5C153S lines as compared with wild-type plants (Fig. 4g). Indeed, 35S:ABI5C153S lines were more deeply hypersensitive than the ABI5-overexpressing lines. Together, this information implies that one way in which seeds sense NO was by the S-nitrosylation of ABI5 at Cys 153, and that disruption of this mechanism was able to arrest seedling growth under adverse environmental conditions.

The involvement of group VII ERF transcription factors in NO sensing and NO downregulation of ABI5 transcription has been previously reported 7 . Consistently, ABI5 protein levels accumulated to a greater extent after ABA treatment in prt6-1 mutant seeds and seedlings, which are impaired in the degradation of group VII ERFs by the N-end rule pathway 25,26 (Fig. 5a–c). Thus, Rubisco large subunit was detected in those conditions/genotypes where the development of green and expanded cotyledons had taken place, producing seedlings with the ability to fix carbon during photosynthesis (Fig. 5c). The key role of ABA in the inhibition of seed germination and in post-germinative seedling arrest (that is, the absence of green and expanded cotyledons) was also highlighted, as no detection of Rubisco large subunit could be found in those ABA-hypersensitive genotypes (prt6) or in the Col-0 wild type under high ABA concentrations. Conversely, ABI5 was only detected in those conditions/genotypes where the seed germination was inhibited and or seedling growth arrested at post-germination developmental checkpoints. Here we also demonstrated that NO-induced ABI5 degradation by the proteasome could occur in a prt6-1 mutant background (Fig. 5d). Thus, NO could promote ABI5 protein degradation independently of the NO-dependent transcriptional control of ABI5 by the N-end rule pathway. Taken together, our results demonstrated a pivotal role of ABI5 in antagonistic ABA and NO signalling in the regulation of seed germination and highlighted ABI5 as a critical NO sensor in seeds.

(a,b) ABA treatment promotes ABI5 accumulation in germinating seeds. Immunoblot analysis of ABI5 protein levels in seed extracts of Col-0 and prt6-1 before (a) and after 48 h treatment with (+) or without (−) 0.25 μM ABA (b). Actin protein levels are shown as a loading control. (c) Post-germinative ABI5 accumulation in seedlings. Immunoblot analysis of ABI5 protein levels in 10-day-old extracts of Col-0, prt6-1 and abi5-1, treated with or without (−) 0.25, 0.5 and 1 μM ABA. Actin protein levels are shown as a loading control and Rubisco large subunit (RbcL) detection is indicated. (d) ABI5 protein levels in wild-type (Col-0), prt6-1 and abi5-1 mutant backgrounds. Stratified seeds were incubated with 5 μM ABA for 48 h (T0) and treated after ABA removal with H2O, ABA, SNAP (1 mM) and GSNO (1 mM) for 12 h. Immunoblot analysis of ABI5 protein levels in seed extracts of wild type and mutants. Actin protein levels are shown as a loading control.

Discussion

In conclusion, our data establish a molecular mechanism for NO and ABA antagonism in the regulation of seed germination and post-germinative growth (Fig. 6). We identify two new loci involved in ABA and NO signalling (GERMINATION IN ABA AND cPTIO, GAP) and characterize one of these loci that correspond to the basic leucine zipper transcription activator ABI5. S-nitrosylation of ABI5 targets proteasomal degradation through CUL4-based and KEG ligases and acts as a regulatory switch for seed germination in Arabidopsis. Thus, S-nitrosylation offers new insights into the regulation of protein stability through CULLIN-related degradation pathways in plants.

Transcriptional control of ABI5 expression via the group VII ERFs 7 and posttranslational S-nitrosylation of ABI5 protein are included. Dormant and dry seeds accumulate high levels of the ABA-induced ABI5 growth reppressor. On seed imbibition, a burst of NO is early produced to degrade group VII ERFs via the N-end rule pathway of targeted proteolysis (PRT6) 25,26 and induces ABI5 S-nitrosylation promoting the interaction with CUL4-based and KEG E3 ligases. Consequently, ABI5 is rapidly degraded by the proteasome during seed germination.

The findings by Gibbs et al. 7 and those reported here demonstrate that NO targets ABI5 at both the transcriptional level (through NO-mediated degradation of ERFVIIs) and the posttranslational level (through NO-mediated degradation of ABI5), respectively, highlighting the fact that ABI5 is depleted via a dual NO-responsive mechanism. These two independent mechanisms converging on ABI5 may have evolved to ensure that NO irrevocably removes ABI5 from the seed to promote germination. Several previous reports emphasize that ABI5 protein stability is regulated tightly by multiple different mechanisms 21,22,23,24 . Thus, it is reasonable to speculate that S-nitrosylation-mediated ABI5 protein degradation plays an important role in the NO function to regulate seed germination, rather than only through transcriptional regulation 7 . In addition, it is likely to be that other mechanisms of NO-induced ABI5 degradation independently of C153 S-nitrosylation may be involved, probably nitrosylating different regulators for ABI5 degradation.

The identification of ABI5 as a direct NO target involved in NO-mediated effects in plant growth and development contributes to our understanding of NO role in plant signal transduction networks and establishes a molecular framework for the NO function during seed germination.

Methods

Plant materials and treatments

Arabidopsis thaliana ecotype Columbia-0 (Col-0) was the genetic background for all wild-type plants used in this work. Seed stocks of abi5, atnoa1, nia1;nia2 and keg4 mutants were obtained from Arabidopsis Biological Resource Center. The atnoa1;nia1;nia2 mutants 11 and 35S:AHb1 (H3, H7) and 35S:antiAHb1 (L1, L3) lines 27 were kind gifts from Dr José León (IBMCP-CSIC, Valencia, Spain) and Dr Massimo Delledonne (University of Verona, Verona, Italy), respectively. Dr Xing Wang Deng (Yale University, Connecticut, USA) kindly provided cul4cs and 35S:FLAG-CUL4 lines 20 . DEX::KEG-HA lines were previously described 28 .

Arabidopsis plants were grown in a growth chamber or greenhouse under 50–60% humidity, a temperature of 22 °C and with a 16-h light/8-h dark photoperiod at 80–100 μE m −2 s −1 in pots containing a 1:3 vermiculite/soil mixture.

For in vitro culture, Arabidopsis seeds were surface sterilized in 75% (v/v) sodium hypochlorite and 0.01% (v/v) Triton X-100 for 5 min, and washed three times in sterile water before sowing. Seeds were stratified for 3 days at 4 °C and then sowed on Murashige and Skoog (MS) 29 solid medium with 2% (w/v) Suc and 0.6% (w/v) agar, and the pH was adjusted to 5.7 with KOH before autoclaving. Seeds were sown in different treatments and plates were sealed and incubated in a controlled environment growth chamber. For NO treatments, Arabidopsis seeds were sown in MS medium supplemented with 300 μM NO donors GSNO and SNAP or 300 μM 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (cPTIO) NO scavenger. MG132 proteasome inhibitor (100 μM) and 0.1–5 μM of ABA were used. Whenever required, seeds were transferred every 24 h to a new medium containing fresh compounds. At the end of the treatment, tissues were collected and frozen (−80 °C).

NO donors (SNAP and GSNO), NO scavenger (cPTIO), proteasome inhibitor (MG132), cycloheximide and ABA were purchased from Sigma.

Mutant screening

M2 seeds of ethyl methane sulfonate-mutagenized Columbia, ethyl methane sulfonate-mutagenized Landsberg erecta and fast neutron-mutagenized Columbia were purchased from Lehle Seeds (Round Rock, TX) and each of ∼ 20,000 M2 seeds (20 batches of ∼ 1,000 M2 seeds harvested from ∼ 1,000 M1 seeds) was used for screening to isolate ABA- and cPTIO-insensitive mutants. Screening conditions were described previously 30 , except for (+)-S-ABA and cPTIO were used. Briefly, M2 seeds were surface sterilized, sown on 0.8% agar plates supplemented with half strength of MS and 3 μM (+)-S-ABA or 100 μM cPTIO. Seeds were stratified for 4 days and incubated in the presence of ABA for 4 days at room temperature under continuous light condition. Germination in cPTIO was tested under continuous light conditions at room temperature without stratification. Seedlings were transferred to hormone-free media, incubated for several days and transfered to pots for seed harvest.

Germination assays

To measure ABA, cPTIO, NaCl and mannitol sensitivity, seeds were plated on solid medium composed of MS basal salts, 2% (w/v) Suc or 0.8% w/v agar plates (adjusted to pH 5.8 by 2-(N-Morpholino)ethanesulfonic acid buffer) and different concentrations of ABA (0.1–1 μM), cPTIO (100 μM), SNAP (300 μM), NaCl (100 mM) and mannitol (250 mM). Seed lots to be compared were harvested on the same day from individual plants grown under identical environmental conditions. Each value represents the average germination percentage of 50–100 seeds with the s.e. of three replicates. Experiments were repeated at least three times. For the dormancy assay, seed lots to be compared were freshly harvested on the same day from individual plants grown under identical environmental conditions and sown immediately without stratification. The after-ripening status of dry seed lots was determined following storage for 1–4 weeks. The percentage of seeds with an emerged radicle (germination) or seedlings that germinated and developed green, fully expanded cotyledons was determined every day during 10 days after sowing.

Detection of endogenous NO

One- to 3-day-old seeds were incubated in a 500-μl solution containing 5 μM of DAF-2DA (Sigma) in buffer Tris–HCl 10 mM, pH 7.4, during 1 h at 25 °C in the dark. Seedlings were then washed three times for 15 min with fresh buffer Tris–HCl 10 mM, pH 7.4. Finally, fluorescence emitted by DAF-2DA was detected on a Leica magnifying glass by excitation at 495 nm and emission at 515 nm. NO depletion was also achieved by adding the scavenger cPTIO (1 mM) to the solution.

GUS histochemical staining

GUS staining of ABI5pro:ABI5-GUS seeds and seedlings was performed as previously described 31 using 50 mM potassium phosphate buffer (pH 7.0) containing 0.05% (v/v) Triton X-100, 1 mM K3Fe(CN)6, 1 mM K4Fe(CN)6, 0.05 M EDTA and 1 mg ml −1 X-Gluc (Duchefa). Twenty-four-hour Arabidopsis germinated seeds in MS medium as well as endosperms from imbibed seeds with embryo removed after GSNO (1 mM), cPTIO (1 mM) and control (MetOH) treatment were used. Staining was examined after 2 h incubation at 37 °C for germinated seeds and 48 h for endosperms. Samples were mounted in 50% glycerol for imaging.

Quantitative reverse transcriptase–PCR analysis

Total RNA for quantitative reverse transcriptase–PCR (qRT–PCR) was extracted from 1- to 3-day-old Arabidopsis seeds treated with GSNO (300 μM), SNAP (300 μM), cPTIO (300 μM), ABA (5 μM), cycloheximide (1 mM), MG132 (100 μM) and control condition as previously described 32 . Briefly, frozen tissue was powdered with mortar and pestle, and incubated on ice with 550 μl of extraction buffer (0.4 M LiCl, 0.2 M Tris, pH 8, 25 mM EDTA, 1% SDS) and 550 μl chloroform followed by centrifugation for 3 min at 15,800g at 4 °C. Next, 500 μl of water-saturated acidic phenol and 200 μl of chloroform were added to the supernatant. After centrifugation for 3 min, 1/3 volume of 8 M LiCl was used for precipitation at −20 °C for 1 h before centrifugation again for 30 min at 4 °C. Pellet was dissolved in 26 μl Diethyl pyrocarbonate treated water and incubated with 10 units of RNase-free DNase I (Roche) 30 min at 37 °C. RNA was precipitated in 3 M NaAc, pH 5.2, and ethanol, washed with 70% ethanol and resuspended in 20 μl Diethyl pyrocarbonate treated water. For 7-day-old abi5-1;35S:cMyc:ABI5, abi5-1;35S:cMyc:ABI5C153S, Col-0;35S:cMyc:ABI5 and Col-0;35S:cMyc:ABI5C153S Arabidopsis seedlings, total RNA was extracted using TRIzol Reagent as directed by the manufacturer (Invitrogen) and complementary DNA was synthesized using SuperScript Kit (Roche). qRT–PCRs were performed as described previously 15 in ABI PRISM 7000 Sequence Detection System (Applied Biosystems, Framingham, MA, USA). Amplification was carried out with ‘Brilliant SYBR Green QPCR MasterMix’ (Stratagene) according to the manufacturer’s instructions. The thermal profile for SYBR Green real-time PCR was 50 °C for 2 min, 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. To generate the standard curves, cDNA isolated from Arabidopsis seedlings was serially diluted 10 × and aliquots of the dilutions were used in standard real-time PCRs. Each value determination was repeated three times, to ensure the slope of the standard curves and to determine the s.d. The concentration of unknown samples was calculated with the ABI-Prism 7000 SDS software, which created threshold cycle values (Ct) and extrapolated relative levels of PCR product from the standard curve. Primers used were ABI5 (forward 5′-AACATGCATTGGCGGCGGAGT-3′, reverse 5′-TTGTGCCCTTGACTTCAAACT-3′) and ACTIN8 (forward 5′-AGTGGTCGTACAACCGGTATTGT-3′, reverse 5′-GAGGATAGCATGTGGAAGTGAGAA-3′) as a control.

Production of recombinant proteins and polyclonal antibodies

Wild-type ABI5 (kindly provided by Dr Luis Lopez-Molina, University of Geneva, Geneva, Switzerland) and mutated ABI5C153S cDNAs were recombined into pET-28a + vector, to obtain fusion proteins with amino-terminal His-tag. Recombinant proteins were expressed in Escherichia coli and purified using His-Select Nickel Affinity Gel (IMAC) (Sigma) according to the manufacturer’s protocols (Biomedal).

A primary dose of 750 μg purified recombinant protein 6 × His-ABI5 was emulsified in Freunds complete adjuvant (Sigma) and administered subcutaneously in two rabbits. Three doses of the protein (375 μg) emulsified in Freunds incomplete adjuvant (Sigma) were administered at intervals (21 days). After the third booster injection (10 days), blood was collected from the rabbits and the serums were separated. Antibodies (anti-ABI5) were isolated by column chromatography with a protein-G column (GE Healthcare Life Sciences).

Western blotting

Total protein for western blot analysis was extracted from dormant seeds of Col-0, abi5-1 and after-ripened seeds of Col-0, abi5-1 and transgenic lines untreated or treated during 72 h with ABA, cPTIO, GSNO, SNAP, GSNO and SNAP plus MG132 or the proteasome inhibitor cocktail including MG115 (1 μM), MG132 (100 μM) and epoxomicin (0.75 μM). For cycloheximide treatments, transgenic plants were germinated and grown on MS liquid medium during 7 days and cycloheximide (1 mM) alone or in combination with GSNO (500 μM) was added to the medium, and samples were taken at indicated intervals. Tissue was powdered using mortar and pestle, and incubated for 10 min on ice with extraction buffer (50 mM Tris–HCl, pH 7.5, 75 mM NaCl, 15 mM EGTA, 15 mM MgCl2, 1 mM DTT, 0.1% Tween 20, 1 mM NaF, 0.2 M NaV, 2 mM Na-pyrophosphate, 60 mM β-glycerolphosphate and 1 × proteases inhibitor mix, Roche) followed by centrifugation for 10 min at 15,800g at 4 °C. Protein concentration was determined by the Bio-Rad Protein Assay (Bio-Rad) based on the Bradford method 33 . Sixty micrograms of total protein was loaded per well in SDS-acrylamide/bisacrylamide gel electrophoresis using Tris–glycine–SDS buffer. Proteins were electrophoretically transferred to an Inmobilon-P polyvinylidene difluoride membrane (Millipore) using the Trans-Blot Turbo (Bio-Rad). Membranes were blocked in Tris buffered saline-0,1% Tween 20 containing 5% Blocking Agent and probed with antibodies diluted in blocking buffer. Anti-ABI5 Purified Rabbit Immunoglobulin (Biomedal, 1:5,000), anti-Actin clone 10-B3 Purified Mouse Immunoglobulin (Sigma A0480, 1:20,000), monoclonal anti-polyHistidine−Peroxidase produced in mouse (Sigma A7058, 1:2,000) and ECL-Peroxidase-labelled anti-rabbit (Amersham NA934, 1:20,000) and anti-mouse (Amersham NA931, 1:20,000) antibodies were used in the western blot analyses. Detection was performed using ECL Advance Western Blotting Detection Kit (Amersham) and the chemiluminescence was detected using an Intelligent Dark-Box II, LAS-1000 scanning system (Fujifilm). Quantification of band intensity was performed with ImageJ software. Full-sized uncropped immunoblots of cropped blottings used in figures are included in Supplementary Fig. 12.

Site-directed mutagenesis

Site-directed mutagenesis of ABI5 was performed using the QuickChange II Site-Directed Mutagenesis Kits (Stratagene Corporate). Plasmid pET28a-ABI5 was used as template and primers were designed using the tools from Stratagene and synthesized by Isogen. The primers were as follows: forward primer: 5′-CACTTCCAGCTCCGCTTAGTAGGAAGACTGTTGAT-3′and reverse primer 5′-ATCAACAGTCTTCCTACTAAGCGGAGCTGGAAGTG-3′. Mutations were confirmed by sequencing.

In vivo and in vitro S-nitrosylation assays

We used the biotin switch method 18 that converts –SNO into biotinylated groups, to detect S-nitrosylated proteins in Arabidopsis seed extracts and recombinant purified ABI5 and ABI5C153S, with slight modification.

For in vitro S-nitrosylation, purified ABI5 recombinant protein was pre-treated with NO donors SNAP and GSNO (200 μM, Calbiochem), and the glutathionylating agent glutathione (200 μM, Sigma) in the dark at room temperature for 30 min with regular vortexing. Treatment with the reducing agent (DTT, 20 mM; Sigma) after GSNO incubation was carried out for 1 h under the same conditions, to check reversibility of the modification. Reagents were removed by precipitation with two volumes of cold acetone and proteins were assayed by the biotin switch method.

In vivo S-nitrosylation of ABI5 was carried out with Arabidopsis dry seeds homogenized in extraction buffer (50 mM Tris–HCl, pH 7.5, 75 mM NaCl, 15 mM EGTA, 15 mM MgCl2, 0.1% Tween 20, 1 mM NaF, 0.2 M NaV, 2 mM Na-pyrophosphate, 60 mM β-glycerolphosphate) containing Complete Protease Inhibitor Cocktail (Roche). Seeds extracts (1 mg) were incubated with GSNO (1 mM) and proteasome inhibitor MG132 (100 μM) in the dark for 30 min with repeated vortexing. Samples treated with DTT (20 mM) were also kept for 1 h under the same conditions and were used as a negative control. Proteins were recovered by precipitation with two volumes of acetone for 20 min at −20 °C, to remove excess of GSNO/DTT and assayed by the biotin switch method.

For the biotin switch, extracts or recombinant proteins were incubated with 20 mM S-methyl-methanethiosulfonate and 2.5% SDS at 50 °C for 30 min with frequent vortexing, to block free Cys. S-methyl-methanethiosulfonate was removed by protein precipitation with two volumes of cold acetone and proteins were disolved in 0.1 ml of RB buffer (25 mM Hepes, 1 mM EDTA and 1% SDS, pH 7.7) per mg of protein. After addition of 1 mM HPDP biotin (Pierce, Rockford, IL) and 1 mM ascorbic acid, the mixture was incubated for 1 h at room temperature in the dark with intermittent vortexing. In vivo biotinylated proteins were purified by immunoprecipitation with neutravidin or an IPA (protein A/G Ultralink Resin, Pierce) anti-biotin antibody, overnight at 4 °C with 15 μl IPA per mg of protein, preincubated with 2 μl of anti-biotin antibody (Sigma B7653, 1:2,000) as described elsewhere 34 . Briefly, beads were washed three times with HEN buffer (100 mM Hepes, 1 mM EDTA and 0.1 mM neocuproine, pH 7.8) and bound proteins were eluted with 10 mM DTT in SDS–PAGE solubilization buffer, loaded in 10% SDS–PAGE and transferred to a polyvinylidene difluoride membrane, to detect ABI5 with anti-ABI5 Purified Rabbit Immunoglobulin (Biomedal, 1:5,000). In-vitro biotinylated proteins with SDS–PAGE solubilization buffer were loaded in 12% SDS–PAGE, visualized by Brilliant Blue-G Colloidal Stain (Sigma) and protein bands were analysed by matrix-assisted laser desorption/ionization–tandem time of flight (MALDI–TOF/TOF). Alternatively, purified proteins were analysed directly by nano-liquid chromatography (LC) and ion-trap tandem mass spectrometry (MS/MS).

Mass spectrometry

For tryptic digestion in reducing conditions, Coomassie-stained bands were excised manually, deposited in 96-well plates and processed automatically in a Proteineer DP (Bruker Daltonics, Bremen, Germany). The digestion protocol 35 was used with minor modifications. Gel plugs were washed first with 50 mM ammonium bicarbonate and second with acetonitrile (ACN) before reduction with 10 mM DTT in 25 mM ammonium bicarbonate solution, and alkylation was carried out with 55 mM iodoacetamide in 50 mM ammonium bicarbonate solution. Gel pieces were then rinsed with 50 mM ammonium bicarbonate and with ACN, and then dried under a stream of nitrogen. Modified porcine trypsin (sequencing grade; Promega, Madison, WI) was added at a final concentration of 16 ng μl −1 in 25% ACN/50 mM ammonium bicarbonate solution and the digestion was led to proceed at 37 °C for 6 h. The reaction was stopped by adding 0.5% trifluoroacetic acid for peptide extraction. Digestion in non-reducing conditions was done essentially with the same protocol, but no DTT and iodoacetamide were used. In both cases, tryptic peptides were dried by speed-vacuum centrifugation and resuspended in 4 μl of MALDI solution (30% ACN+15% isopropanol+0.1% trifluoroacetic acid). Twenty per cent of each peptide mixture was deposited onto a 386-well OptiTOF Plate (Applied Biosystems) and allowed to dry at room temperature. A 0.8-μl aliquot of matrix solution (3 mg ml −1 CHCA in MALDI solution) was then added onto the dried digests and allowed to dry at room temperature.

For MALDI peptide mass fingerprinting (PMF) and MS/MS analysis, samples were automatically processed in an ABi 4800 MALDI–TOF/TOF mass spectrometer (Applied Biosystems) in positive ion reflector mode (ion acceleration voltage was 25 kV for MS acquisition and 1 kV for MS/MS) and the spectra were stored into the ABi 4000 Series Explorer Spot Set Manager. PMF and MS/MS fragment ion spectra were smoothed and corrected to zero baseline using routines embedded in ABi 4000 Series Explorer Software v3.6. Each PMF spectrum was internally calibrated with the mass signals of trypsin autolysis ions to reach a typical mass measurement accuracy of <25 p.p.m. Known trypsin and keratin mass signals, as well as potential sodium and potassium adducts (+21 Da and +39 Da) were removed from the peak list.

To submit the combined PMF and MS/MS data to MASCOT software v.2.1 (Matrix Science, London, UK), GPS Explorer v4.9 was used, searching in the non-redundant NCBI protein database (Viridiplantae) with peptide mass tolerance of 75 p.p.m. for MALDI–TOF/TOF analysis.

The remaining 80% of the samples were analysed by LC coupled to electrospray ion-trap mass spectrometry MS/MS using Ultimate 3000 nano LC (Dionex, Amsterdam, The Netherland) and a 75-mm I.D., 100 mm reverse-phase column, at 300 nl min −1 flow, coupled to a Bruker HCT Ultra ion-trap mass spectrometer (Bruker Daltonics, Bremen, Germany) working in dynamic exclusion mode. For protein identification, LC coupled to electrospray ion-trap mass spectrometry MS/MS spectra were transferred to BioTools 2.0 interface (Bruker Daltonics), to search in the NCBInr database using a licensed version of Mascot v.2.2.04 search engine (www.matrixscience.com; Matrix Science). Search parameters were set as follows: in reduced samples, carbamidomethyl cystein was set as fixed modification by the treatment with iodoacetamide, oxidized methionines as variable modification, peptide mass tolerance of 0.5 Da for the parental mass and fragment masses, and one missed cleavage site. In the case of non-reduced samples, biotin-HPDP cysteine modification was set as variable modification.

Yeast two-hybrid assay

ABI5 and ABI5C153S were cloned in the pDEST22 and pDEST32 vectors using the GATEWAY technology. Prey and bait clones were grown for 3 days on DOB-W and DOB-L (MP Biomedicals) plates from their corresponding frozen stocks. YPAD medium was inoculated in parallel with bait and prey cells, and incubated overnight at 28 °C with shaking (200 r.p.m.). After overnight incubation, the bait culture was added to prey and mating was allowed 48 h by incubating at 28 °C without shaking. Settled cells were resuspended and used to inoculate plates containing diploid selection media (DOB-L-W). After 1 day of growth at 28 °C and vigorous shaking, diploid cells were resuspended and spotted onto diploid selection and screening (DOB-L-W-Histidine±3-Amino-1,2,4-triazole; 3-AT, Sigma) plates. Positive colonies were visible after 2–5 days of growth at 28 °C.

Electrophoretic mobility shift assay

For the electrophoretic mobility shift assay, 10 ng of a double-stranded biotinylated probe 5′-GATCCTCTCGCGTACAATAAAGTCAGACACGTGGCATGTCACCAACGTAGCGTATGCGTA-3′(ABRE-binding site motif underlined) 36 was incubated with 1.5 μg wild-type ABI5 or mutant ABI5C153S recombinant protein in the presence and absence of GSNO (1 mM) and DTT (1 mM) as described previously 37 . Basically, DNA-binding reactions were performed in a buffer containing 10 mM Tris, pH 8, 1 mM EDTA, 100 mM NaCl, 2 mM DTT and 10% glycerol. Protein extracts were mixed with labelled DNA, 0.5 μg poly(dI-dC) and BSA (250 μg μl −1 ) and incubated for 30 min on ice. Electrophoretic mobility shift assay to separate free and bound DNA was in 6% polyacrylamide gel (40:1 bisacrylamide cast in 0.5 × TBE; TBE is 89 mM Tris, 89 mM boric acid and 2 mM EDTA).

Co-immunoprecipitation and pull-down assays

For haemagglutinin (HA) and FLAG pull-down assays, the proteins were extracted with lysis buffer (100 mM Tris–HCl, pH 7.5, 150 mM NaCl, 0.1% Tween-20) supplemented with 1 mM phenylmethanesulfonyl fluoride and protease inhibitors (Roche). Extracts were cleared by centrifugation and protein concentration was determined by Bradford assay. One milligram of soluble proteins was treated with GSNO (1 mM) in darkness at room temperature during 1 h. After treatment, extracts were immunoprecipitated using anti-HA Affinity Matrix (Roche) and anti-FLAG M2 Matrix (Sigma) for ABI5/KEG and ABI5/CUL4 interactions, respectively. Extracts and beads were incubated during 2 h at 4 °C and beads were washed two times in lysis buffer. Next, the beads were incubated at 4 °C during 2 h with 1 mg of ABI5 and ABI5C153S soluble proteins previously treated with GSNO (1 mM) in darkness at room temperature during 1 h. After incubation, beads were washed three times in lysis buffer and proteins were eluted from beads in SDS sample buffer. The proteins were visualized using anti-MYC antibodies (Abcam ab62928, 1:10,000), anti-HA antibodies (Roche 12013819001, 1:2,000) and anti-FLAG antibodies (Sigma F1804, 1:1,000).

Generation of transgenic Arabidopsis plants

ABI5 and ABI5C153S were cloned in the pEarleyGate 203 vector 38 using the GATEWAY technology and the following primers (ABI5-F 5′-ATGGTAACTAGAGAAACGAAGTTGACG-3′; ABI5-R 5′-TTAGAGTGGACAACTCGGG-3′). Similarly, ABI5pro:ABI5 was cloned in the binary pGWB3 vector 39 fused to GUS. The constructs generated (Supplementary Fig. 11a) were used to transform the C58C1 (pGV2260) Agrobacterium strain 40 . For plant transformation, Arabidopsis plants (Col-0, abi5-1 or abi5-7) were transformed by the floral dip method 41 as described previuosly 42 . Seeds were harvested and plated on selection medium to identify T1 transgenic plants. Approximately 100 of the T2 seeds were plated on selection medium MS agar plates and transgenic lines with a 3:1 (resistant/sensitive) segregation ratio were selected. T3 progenies homozygous for selection medium resistance were used for further studies.

Additional information

How to cite this article: Albertos, P. et al. S-nitrosylation triggers ABI5 degradation to promote seed germination and seedling growth. Nat. Commun. 6:8669 doi: 10.1038/ncomms9669 (2015).

References

Nambara, E. et al. Abscisic acid and the control of seed dormancy and germination. Seed Sci. Res. 20, 55–67 (2010).

Lopez-Molina, L., Mongrand, S. & Chua, N. H. A post- germination developmental arrest checkpoint is mediated by abscisic acid and requires the ABI5 transcription factor in Arabidopsis. Proc. Natl Acad. Sci. USA 98, 4782–4787 (2001).

Finkelstein, R. R. & Lynch, T. J. The Arabidopsis abscisic acid response gene ABI5 encodes a basic leucine zipper transcription factor. Plant Cell 12, 599–609 (2000).

Bensmihen, S. et al. The homologous ABI5 and EEL transcription factors function antagonistically to fine-tune gene expression during late embryogenesis. Plant Cell 14, 1391–1403 (2002).

Lopez-Molina, L., Mongrand, S., McLachlin, D., Chait, B. & Chua, N. H. ABI5 acts downstream of ABI3 to execute an ABA-dependent growth arrest during germination. Plant J. 32, 1–12 (2002).

Wang, P. et al. Nitric oxide negatively regulates abscisic acid signaling in guard cells by S-nitrosylation of OST1. Proc. Natl Acad. Sci. USA 112, 613–618 (2014).

Gibbs, D. J. et al. Nitric oxide sensing in plants is mediated by proteolytic control of group VII ERF transcription factors. Mol. Cell 53, 369–379 (2014).

Bethke, P. C., Badger, M. R. & Jones, R. L. Apoplastic synthesis of nitric oxide by plant tissues. Plant Cell 16, 332–341 (2004).

Bethke, P. C., Gubler, F., Jacobsen, J. V. & Jones, R. L. Dormancy of Arabidopsis seeds and barley can be broken by nitric oxide. Planta 219, 847–855 (2004).

Beligni, M. V. & Lamattina, L. Nitric oxide stimulates seed germination and de-etiolation, and inhibits hypocotyl elongation, three light-inducible responses in plants. Planta 210, 215–221 (2000).

Lozano-Juste, J. & Leon, J. Enhanced abscisic acid-mediated responses in nia1nia2noa1-2 triple mutant impaired in NIA/NR- and AtNOA1-dependent nitric oxide biosynthesis in Arabidopsis. Plant Physiol. 152, 891–903 (2010).

Stamler, J. S., Lamas, S. & Fang, F. C. Nitrosylation: the prototypic redox-based signaling mechanism. Cell 106, 675–683 (2001).

Kwak, Y. D. et al. NO signaling and S-nitrosylation regulate PTEN inhibition in neurodegeneration. Mol. Neurodegener. 5, 49 (2010).

Liu, Y. G. et al. Nitric oxide-induced rapid decrease of abscisic acid concentration is required in breaking seed dormancy in Arabidopsis. New Phytol. 183, 1030–1042 (2009).

Fernández-Marcos, M., Sanz, L., Lewis, D. R., Muday, G. K. & Lorenzo, O. Nitric oxide causes root apical meristem defects and growth inhibition while reducing PIN-FORMED 1 (PIN1)-dependent acropetal auxin transport. Proc. Natl Acad. Sci. USA 108, 18506–18511 (2011).

Hill, R. D. Non-symbiotic haemoglobins-what’s happening beyond nitric oxide scavenging? AoB Plants pls004, 1–13 (2012).

Romero-Puertas, M. C. et al. S-nitrosylation of peroxiredoxin II E promotes peroxynitrite-mediated tyrosine nitration. Plant Cell 19, 4120–4130 (2007).

Jaffrey, S. R. & Snyder, S. H. The biotin switch method for the detection of S-nitrosylated proteins. Sci. STKE 86, pl1 (2001).

Nakamura, S., Lynch, T. J. & Finkelstein, R. R. Physical interactions between ABA response loci of Arabidopsis. Plant J. 26, 627–635 (2001).

Chen, H. et al. Arabidopsis CULLIN4 forms an E3 ubiquitin ligase with RBX1 and the CDD complex in mediating light control of development. Plant Cell 18, 1991–2004 (2006).

Lee, J. H. et al. DWA1 and DWA2, two Arabidopsis DWD protein components of CUL4-based E3 ligases, act together as negative regulators in ABA signal transduction. Plant Cell 22, 1716–1732 (2010).

Stone, S. L., Williams, L. A., Farmer, L. M., Vierstra, R. D. & Callis, J. KEEP ON GOING, a RING E3 ligase essential for Arabidopsis growth and development, is involved in abscisic acid signaling. Plant Cell 18, 3415–3428 (2006).

Liu, H. & Stone, S. L. Abscisic acid increases Arabidopsis ABI5 transcription factor levels by promoting KEG E3 ligase self-ubiquitination and proteasomal degradation. Plant Cell 22, 2630–2641 (2010).

Liu, H. & Stone, S. L. Cytoplasmic degradation of the Arabidopsis transcription factor ABSCISIC ACID INSENSITIVE 5 is mediated by the RING-type E3 Ligase KEEP ON GOING. J. Biol. Chem. 288, 20267–20279 (2013).

Holman, T. J. et al. The N-end rule pathway promotes seed germination and establishment through removal of ABA sensitivity in Arabidopsis. Proc. Natl Acad. Sci. USA 106, 4549–4554 (2009).

Gibbs, D. J. et al. Homeostatic response to hypoxia is regulated by the N-end rule pathway in plants. Nature 479, 415–418 (2011).

Perazzolli, M. et al. Arabidopsis nonsymbiotic hemoglobin AHb1 modulates nitric oxide bioactivity. Plant Cell 16, 2785–2794 (2004).

Wawrzynska, A., Christiansen, K. M., Lan, Y., Rodibaugh, N. L. & Innes, R. W. Powdery mildew resistance conferred by loss of the ENHANCED DISEASE RESISTANCE1 protein kinase is suppressed by a missense mutation in KEEP ON GOING, a regulator of abscisic acid signaling. Plant Physiol. 148, 1510–1522 (2008).

Murashige, T. & Skoog, F. A revised medium for rapid growth and bioassays with tobacco tissue culture. Physiol. Plant. 15, 473–497 (1962).

Nambara, E. et al. A screen for genes that function in abscisic acid signaling in Arabidopsis thaliana. Genetics 161, 1247–1255 (2002).

Jefferson, R. A., Kavanagh, T. A. & Bevan, M. W. GUS fusions: β-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6, 3901–3907 (1987).

Oñate-Sánchez, L. & Vicente-Carbajosa, J. DNA-free RNA isolation protocols for Arabidopsis thaliana, including seeds and siliques. BMC Res. Notes 1, 93 (2008).

Bradford, M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254 (1976).

Ortega-Galisteo, A. P. et al. S-Nitrosylated proteins in pea (Pisum sativum L.) leaf peroxisomes: changes under abiotic stress. J. Exp. Bot. 63, 2089–2103 (2012).

Schevchenko, A., Wilm, M., Vorm, O. & Mann, M. Mass spectrometric sequencing of proteins from silver stained polyacrylamide gels. Anal. Chem. 68, 850–858 (1996).

Carles, C. et al. Regulation of Arabidopsis thaliana Em genes: role of ABI5. Plant J. 30, 373–383 (2002).

Godoy, M. et al. Improved protein-binding microarrays for the identification of DNA-binding specificities of transcription factors. Plant J. 66, 700–711 (2011).

Earley, K. W. et al. Gateway-compatible vectors for plant functional genomics and proteomics. Plant J. 45, 616–629 (2006).

Nakagawa, T. et al. Development of series of gateway binary vectors, pGWBs, for realizing efficient construction of fusion genes for plant transformation. J. Biosci. Bioeng. 104, 34–41 (2007).

Deblaere, R. et al. Efficient octopine Ti plasmid-derived vectors for Agrobacterium-mediated gene transfer to plants. Nucleic Acids Res. 13, 4777–4788 (1985).

Clough, S. J. & Bent, A. F. Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16, 735–743 (1998).

Saavedra, X. et al. The nuclear interactor PYL8/RCAR3 of Fagus sylvatica FsPP2C1 is a positive regulator of abscisic acid signaling in seeds and stress. Plant Physiol. 152, 133–150 (2010).

Acknowledgements

We thank Roberto Solano, Salomé Prat, Javier Paz-Ares and Dolores Rodriguez for critical reading of the manuscript and stimulating discussions. We also thank proteomics facility CNB-CSIC for technical proteomic assistance. We are grateful to Luis Lopez-Molina, Xing Wang Deng, Massimo Delledonne and José León for providing seeds and full-length cDNAs. This work was financed by grants BIO2014-57107-R, BIO2011-26940 and CSD2007-00057 (TRANSPLANTA) from the Ministerio de Economia y Competitividad (Spain), EcoSeed Impacts of Environmental Conditions on Seed Quality ‘EcoSeed-311840’ ERC.KBBE.2012.1.1-01 and SA239U13 from Junta de Castilla y León (to O.L.). P.A. was supported by a FPU fellowship from the Ministerio de Educación y Ciencia (Spain).

Author information

Affiliations

Dpto. de Microbiología y Genética, Instituto Hispano-Luso de Investigaciones Agrarias (CIALE), Facultad de Biología, Universidad de Salamanca, C/ Río Duero 12, Salamanca, 37185, Spain

Pablo Albertos, Isabel Mateos, Inmaculada Sánchez-Vicente & Oscar Lorenzo

Dpto. de Bioquímica, Biología Celular y Molecular de Plantas, Estación Experimental del Zaidín, Consejo Superior de Investigaciones Científicas, Profesor Albareda 1, Granada, E-18008, Spain

María C. Romero-Puertas

Laboratory of Plant Organ Development, National Institute for Basic Biology, Nishigonaka 38, Myodaiji, 444-8585, Okazaki, Japan

Department of Cell and Systems Biology, University of Toronto, 25 Willcocks Street, Toronto, M5S3B2, Ontario, Canada