Effects of salinity on the growth, physiology and relevant gene expression of an annual halophyte grown from heteromorphic seeds

Seed heteromorphism provides plants with alternative strategies for survival in unfavourable environments, which may not only increase the chances of successful germination, but may also have an impact on characters of the descendants. However, Cao et al. found that the different properties of the dimorphic seeds of Suaeda aralocaspica (i.e. black and brown) had no effects on their descendants' growth and physiological responses to salinity: all descendants required salinity for optimal growth and adaptation to their natural habitat.


Introduction
Seed heteromorphism is a phenomenon in which an individual plant is able to produce different types of seeds with diverse morphologies or germination and dormancy behaviours (Sorensen 1978;Mandá k and Pyšek 2001;Brä ndel 2004Brä ndel , 2007. These adaptations evolved in response to adverse environments (Imbert 2002). According to seed type, seed heteromorphism can be classified into di-or polymorphism. In dimorphic seeds, a brown seed is usually larger, non-dormant and can germinate quickly, but the seedling mortality is much higher (a highrisk strategy; Venable 1985a) than a black seed. The latter is smaller, dormant and more sensitive to the environment than the brown seed and can enter the seed bank to compensate population size in an unpredictable environment (Venable 1985a). Seed heteromorphism is an important strategy for plants to adapt to environmental stress by extending the germination time, minimizing the risk of a single germination and ensuring the successful expansion of the population (Weber 2009).
Besides enhancing the chances of successful germination, seed heteromorphism may also have a significant impact on the growth, physiological characters and stress tolerance of the descendants ( Table 1). The seedlings descended from heteromorphic seeds may present significant differences in growth rate and size, which could persist throughout the life of the plant or may disappear in the subsequent stages of development (Wang et al. 2010). The non-synchronous germination among heteromorphic seeds usually results in different size of the seedlings. Within a species, although the seeds with relatively larger embryos can grow into larger seedlings than those grown from seeds with smaller embryos, the root-to-shoot ratio is similar between the two, which suggests that the development of heteromorphic plants is in synchrony (Imbert and Ronce 2001). Seed heteromorphism can also have an effect in descendants in response to stress (Table 1).
Previously, the majority of studies have focussed on how environmental factors such as salinity and temperature affect dormancy and germination of heteromorphic seeds (Khan and Ungar 1984;Khan et al. 2001a, b;Wang et al. 2008;Aguado et al. 2011;Li et al. 2011); only a few attempts have been made to investigate the physiological carry-over on descendant plants under stress (Redondo-Gó mez et al. 2008;Wang et al. 2008;Jiang et al. 2012). In Suaeda splendens, the seedlings derived from both seed morphs showed little difference in growth and photosynthesis in the presence of high salinity, which was the first report on the carry-over of salinity tolerance from different seed morphs to established seedlings and plants (Redondo-Gó mez et al. 2008). To our knowledge, the first report, which linked physiological responses with differential gene expression in seedlings derived from dimorphic seeds, was from Atriplex centralasiatica (Xu et al. 2011) and in which a molecular description of differential salt tolerance in dimorphic seeds was provided.
In the present study, we extended these findings to responses on salt tolerance of the plants from dimorphic seeds of S. aralocaspica, including morphological, physiological and biochemical changes, as well as relevant gene expression profiles.
Suaeda aralocaspica, the only species of the Borszczowia section of the genus Suaeda in the Amaranthaceae, is a monoecious annual halophyte with a single-cell C4 photosynthesis pathway (Voznesenskaya et al. 2001) and which is distributed in saline -alkaline sandy soils in the southern margin of Junggar Basin in China (Li 1979;Mao 1994). Plants grow up to 20-50 cm high with unisexual flowers (Commissione Redactorum Florae Xinjiangensis 1994) and can produce heteromorphic seeds with disparate forms and different germination characteristics . The brown seed, with oblate shape and soft seed coat, has a size 3.2 × 2.8 × 0.7 mm (length × width × height) and can germinate quickly, while the black seed, with elliptical shape and hard seed coat, has a size 2.5 × 2.3 × 1.4 mm and is dormant. To understand the salt tolerance of the descendants grown from the two seed morphs of S. aralocaspica and the impact of seed heteromorphism on the progeny, a series of experiments were designed to address the following questions: (i) What are the growth, physiological and biochemical responses of the descendants from dimorphic seeds to long-term NaCl treatment, as well as changes in gene expression pattern? (ii) Is there any different impact of seed heteromorphism on the progeny with or without salinity?

Seed collection
Mature fruits of S. aralocaspica were harvested from plants growing in their natural habitat in a desert saline soil at the Wujiaqu 103 regiment (44829 ′ 821 ′′ N, 87831 ′ 181 ′′ E) in October 2010, in Xinjiang, China. Seeds were air-dried at room temperature, then cleaned and sieved to remove the impurities and stored at 4 8C, 10-12 % relative humidity, in brown paper bags for experiments performed between 2011 and 2012. half-strength Hoagland solutions containing 100, 300 or 500 mM NaCl for 2 months. To avoid osmotic shock, 300 and 500 mM NaCl treatments were imposed as follows: seedlings were first treated with 100 mM for 12 h, then the NaCl concentration gradually increased to 200, 300 mM or 400, 500 mM every 12 h until the final concentrations were achieved. Thereafter, the final concentration was applied every 2 weeks. For all assays, four samples were collected from young fresh leaves in the upper part of the plant and immediately frozen in liquid nitrogen until use.

Measurement of growth relevant parameters
The plant height and leaf length were measured with a rectilinear scale, and the stem diameter and leaf width were determined using a Vernier caliper. Ten plants of similar size from the population were used to determine the growth parameters.

Determination of chlorophyll concentration
Chlorophyll (Chl) concentration was measured in extracts of fresh leaves (0.15 g) with 96 % ethanol as described by Porra et al. (1989). The resulting homogenate was centrifuged at 1200g for 10 min, and the absorbance of the supernatant recorded (UV-3010, Shimadzu, Japan) at 649 and 665 nm. The Chl a and Chl b concentrations were calculated by the following equations: Chl a ¼ 13.95D 665 2 6.88D 649 and Chl b ¼ 24.96D 649 2 7.32D 665 , where D 649 and D 665 are the absorbance values at 649 and 665 nm, respectively. The total Chl concentration in leaves was calculated by the addition of Chl a and Chl b .
Determination of Na 1 and K 1 concentrations Na + and K + measurements were conducted according to Sun et al. (2007). Young fresh leaves (1.0 g) from the upper part of the plant were washed three times with distilled water and quickly placed at 105 8C for 10 min to deactivate enzymes, then transferred to an aerated oven at 70 8C for 6 -12 h and weighed every 3 h until constant dry weight (DW). Cations were extracted from dry plant material (0.3 g) with HNO 3 (25 mL), and the Na + and K + concentrations determined by inductively coupled plasma atomic emission spectrometry (Agilent, USA). Pure HNO 3 was used as control.
Determination of water content and osmolyte concentration: proline, glycinebetaine, total soluble sugar and total protein Leaf water content (WC) was measured according to Pujol et al. (2001). Young fresh leaves were detached from each treatment and weighed immediately to record the fresh weight (FW). Then the leaves were transferred into an oven at 70 8C drying for 24 h to achieve the DW. The WC was calculated as follows: WC (% FW) ¼ (FW 2 DW) × 100/FW. Proline concentration was determined following Zhao et al. (2009). Young fresh leaves (0.15 g) were homogenized in aqueous sulfosalicylic acid (5 mL, 3 % (w/v)) and boiled for 10 min (shaking from time to time). A mixture containing leaf extract (200 mL), glacial acetic acid (200 mL) and acidic ninhydrin (300 mL, 2.5 % (w/v)) was incubated in a boiling water bath for 40 min and then cooled on ice. The product was extracted with toluene (500 mL) by vigorous shaking and the absorbance measured at 520 nm; toluene was used as the blank. The concentration of proline was determined by a calibration standard curve (0-20 mg mL 21 ).
Glycinebetaine (GB) was determined according to Huang et al. (2009). Young fresh leaves (0.15 g) were ground with distilled water (1.5 mL). After centrifugation at 9000g for 15 min, saturated Reinecke's salt (500 mL, 15 mg mL 21 ) was added to the supernatant (300 mL)  (2005) and the mixture incubated on ice for 1 h, followed by centrifugation at 9000g for 15 min. The supernatant was discarded and the precipitate was washed two to three times with 99 % (v/v) ether and then dissolved in acetone (70 %). The absorbance was immediately recorded at 525 nm; a standard curve of GB (0 -200 mg mL 21 ) was used to determine the concentration. The concentration of total soluble sugars (SSs) was determined using the method described by Palma et al. (2009). Young fresh leaves (0.12 g) were sheared into pieces and boiled in distilled water (10 mL) for 30 min. Extraction mixture (50 mL) was added to distilled water (150 mL) and reacted with anthrone ethyl acetate (50 mL, 2 % (w/v)). Concentrated sulfuric acid (500 mL) was added to the mixture to develop colour, then immediately boiled for 1 min and cooled in ice water. The absorbance was determined at 630 nm and a standard curve (0-100 mg mL 21 of glucose) used to estimate SS.
Total protein concentration was determined by measuring the absorbance at 595 nm according to Bradford (1976). Young fresh leaves (0.1 g) were homogenized in liquid nitrogen and then transferred into a 1.5-mL microtube with extraction buffer (500 mL, 12.5 mM Tris-Cl, pH 6.8). The homogenate was centrifuged at 10 000g for 15 min at 4 8C. All of the antioxidant enzyme activities were expressed as unit per mg total protein.

Measurement of O 2
− production, H 2 O 2 concentration and lipid peroxidation O 2 − production was estimated according to the method described by Li and Gong (2005). Young fresh leaves (0.2 g) were homogenized with sodium phosphate buffer (phosphate-buffered saline (PBS), 50 mM, pH 7.8) on ice and the mixture centrifuged at 10 000g for 20 min at 4 8C. The supernatant (1 mL) was then incubated with hydroxylamine (1 mL, 1 mM) at 25 8C for 1 h, then 4-aminobenzenesulfonic acid (1 mL, 17 mM) and a-naphthalenamine (1 mL, 7 mM) were added and the test tubes incubated at 25 8C for 20 min. Reaction mixture (500 mL) was combined with an equal volume of ether to eliminate the pigments. The mixture was centrifuged at 9000g for 3 min and the absorbance of the supernatant recorded at 530 nm. A standard curve using NaNO 2 was plotted within the range of 0 -50 mM.
The H 2 O 2 concentration was determined by a method described by Sun et al. (2010) with some minor changes. Young fresh leaves (0.15 g) were homogenized with cooled acetone and the homogenate was centrifuged at 1000g for 10 min at 4 8C. Then supernatant (1 mL) was transferred into a new 1.5-mL micro-tube followed by the addition of titanium sulfate (100 mL, 5 % (w/v)) and ammonia (200 mL) to precipitate the titanium-hydroperoxide complex, followed by centrifugation as above. The supernatant was discarded and the precipitate suspended and washed three to five times with acetone to remove the pigments and then dissolved in sulfuric acid (5 mL, 2 M). The absorbance was measured at 415 nm and compared with a standard hydrogen peroxide curve (0-100 mM of H 2 O 2 ).
Lipid peroxidation was determined by measuring the concentration of malondialdehyde (MDA) produced by the thiobarbituric acid (TBA) reaction as described by Heath and Packer (1968). Young fresh leaves (0.12 g) were ground in trichloroacetic acid (TCA, 500 mL, 612 mM). Trichloroacetic acid (700 mL, 612 mM) was added to the homogenate, which was centrifuged at 1200g for 10 min. The supernatant (300 mL) was mixed with an equal volume of TBA (34.7 mM), boiled for 10 min, cooled in an ice-bath and then centrifuged at 12 000g for 10 min. The absorbance of the supernatant was monitored at 532, 450 and 600 nm. The concentration of MDA (C ) was calculated by using the following equation: C ¼ 6.45 × (D 532 2 D 600 ) 2 0.56 × D 450 , where D 450 , D 532 and D 600 represent absorbance values at 450, 532 and 600 nm, respectively, and the results expressed as mmol g 21 DW.
For the glutathione (GSH) assay, tissue extract (as that in the AsA assay) was mixed with NaH 2 PO 4 (130 mL, 150 mM) and 5,5-dithio-bis(2-nitrobenzoic acid) (90 mL). The mixture was incubated at 30 8C for 5 min, and the absorbance at 412 nm was recorded (Otto and Moon 1996). The total GSH was calculated from a standard curve prepared with 0 -0.12 mM GSH solution.
Measurement of the activities of antioxidant enzymes: superoxide dismutase, peroxidase, catalase, ascorbate peroxidase and glutathione reductase Young fresh leaves (0.15 g) were homogenized in ice-cold buffer (1.5 mL) containing PBS (pH 7.8, 50 mM), EDTA (0.5 mM), polyvinylpyrrolidone (1 %), AsA (1 mM) and glycerol (10 %). The homogenate was filtered and AoB PLANTS www.aobplants.oxfordjournals.org & The Authors 2015 5 centrifuged at 10 000g for 15 min at 4 8C. The supernatant was transferred onto ice as crude enzyme and immediately used for analysis of the activity. Superoxide dismutase (SOD) activity was determined by measuring its ability to inhibit the photochemical reduction of nitroblue tetrazolium (NBT) chloride, as described by Basu et al. (2010). Crude enzyme extract (50 mL) was added into test tubes with reaction mixture (3 mL) consisting of PBS (pH 7.8, 50 mM), EDTA-Na 2 (100 mM), methionine (130 mM), NBT (750 mM) and riboflavin (2 mM); the test tubes were incubated at 25 8C under a light source of 72 mmol m 22 s 21 for 25 min. As blank treatment, an aluminium foil wrapped blank tube was employed. The absorbance of the reaction mixture was read at 560 nm. One unit (U) of SOD activity was defined as the amount of enzyme required to cause 50 % inhibition of the NBT photoreduction rate.
Ascorbate peroxidase (APX) activity was assayed according to Nakano and Asada (1981) with a minor modification. The reaction mixture (3 mL) contained PBS (pH 7.0, 50 mM), ascorbate (15 mM), H 2 O 2 (0.3 mM) and enzyme extract (100 mL), and the activity was indicated by the change in absorbance at 290 nm. The values were calculated in terms of micromoles of ascorbate oxidized per minute.
Glutathione reductase (GR) activity was determined by the method of Foyer and Halliwell (1976), with minor modifications, based on the oxidation of NADPH at 340 nm for 1 min in a reaction mixture (1 mL) containing NADPH (1 mM), Tricine -NaOH (pH 7.8, 100 mM), glutathione disulfide (5 mM) and enzyme extract (200 mL). Corrections were made by subtraction of the background absorbance at 340 nm in the absence of NADPH.
The absorbance was recorded for 3 min at 340 nm. Enzyme activities of PEPC and RUBPC were defined as 0.01 optical density value decrease per minute is 1 U.

Quantitative real-time polymerase chain reaction analysis of gene expression
Total RNA was extracted from young fresh leaves (0.1 g) using the RNAprep Pure Plant Kit (Tiangen, Beijing, China). Each reverse transcription reaction was performed with total RNA (1.5 mg) using the Quant First Strand cDNA Synthesis Kit (Tiangen) according to the manufacturer's instruction. Based on the published sequences, conserved sequences of 10 target genes and the internal control of b-actin were employed to design degenerate primers of each gene. The subsequent polymerase chain reaction (PCR) products were sequenced with the gene-specific primers for quantitative real-time (qRT)-PCR that were generated (Table 2).
For PCR amplification, the reactions were first incubated for 2 min at 95 8C, followed by 40 cycles of 30 s at 95 8C, 30 s at 56-62 8C (annealing temperature was calculated according to the melting temperature of each pair of genespecific primers), 50 s at 72 8C, then for 10 min extension at 72 8C. Three samples (biological replicates) per gene were duplicated (technical replicates) in qRT-PCR experiment.

Statistical analysis
All data were analysed using the software of GraphPad Prism Version 4.02 for Windows (GraphPad Software, San Diego, CA, USA). Data were subjected to unpaired t-test for independent samples. Two-way analyses of variance were used to test the significance of main effects between seed type and salinity and their interactions. If significant main effects existed, differences were tested by a multiple comparison Tukey test at the 0.05, 0.01 and 0.001 significance level.

Primer names
Primer   concentration than that of the plants from black seed. Only GB concentration was significantly affected by the interaction of seed type and salinity concentration (F 3,24 ¼ 6.265, P ¼ 0.0027).
Expression pattern of ion-transport-and osmoticregulation-related genes. The expression level of vacuolar pyrophosphatase gene (V-PPase) was significantly increased under 300 and 500 mM NaCl treatment in plants from brown and black seeds (F 3,40 ¼ 22.32, P , 0.0001) (Fig. 4A), while that of the key subunit B of vacuolar ATP synthase gene (V-ATPase) remained unchanged except for the plants from black seeds under 300 mM NaCl treatment (F 3,40 ¼ 130.0, P , 0.0001) (Fig. 4B). In comparison between the plants derived from the two types of seeds, an apparent difference was observed in the transcript accumulation of V-PPase (F 1,40 ¼ 43.95, P , 0.0001) and V-ATPase (F 1,40 ¼ 101.0, P , 0.0001) at 300 mM NaCl treatment: expression was significantly higher in plants from black seed compared with that of plants from brown seed (t 10 ¼ 11.40, P , 0.0001 for V-PPase; t 10 ¼ 29.65, P , 0.0001 for V-ATPase).
The expression level of pyrroline-5-carboxylate synthase gene (P5CS) was significantly increased under 100 mM NaCl treatment in plants from heteromorphic seeds (F 3,40 ¼ 26.60, P , 0.0001) (Fig. 4C). There was no significant difference between two types of plants (F 1,40 ¼ 0.6608, P ¼ 0.4211). The transcript accumulation of betaine aldehyde dehydrogenase (BADH) gene was significantly increased under various NaCl concentrations in plants from both types of seeds compared with plants grown in the absence of salt (F 3,40 ¼ 60.92, P , 0.0001), which was obviously affected by seed type (F 1,40 ¼ 19.50, P , 0.0001) and significantly higher in plants from black seed at 300 mM NaCl than that of brown seed (t 10 ¼ 4.118, P ¼ 0.0021) (Fig. 4D).
With the increasing NaCl concentration, the AsA (F 3,16 ¼ 175.6, P , 0.0001) and GSH (F 3,16 ¼ 41.40, P , 0.0001) increased significantly in both plant types and reached to their highest values at 300 and/or 500 mM NaCl concentrations (Fig. 7A and B), although a significant decrease in AsA concentration was observed at 500 mM compared with 300 mM NaCl treatment (t 10 ¼ 10.89, P , 0.0001), it was still significantly higher than that of plants grown in the absence of salt (Fig. 7A). There was no significant difference in antioxidant concentration between plants from brown and black seeds (F 1,16 ¼ 0.8069, P ¼ 0.3824 for AsA; F 1,16 ¼ 0.5116, P ¼ 0.4848 for GSH).

Gene expression pattern of PEPC and RUBPC
With the increase in salt concentration, PEPC was significantly up-regulated (F 3,40 ¼ 153.8, P , 0.0001) (Fig. 10A), while RUBPC was significantly decreased (F 3,40 ¼ 334.6, P , 0.0001) (Fig. 10B) in both types of plants compared with plants grown without salt. Significant differences between plants from the two types of seeds were observed (F 1,40 ¼ 37.98, P , 0.0001 for PEPC; F 1,40 ¼ 148.6, P , 0.0001 for RUBPC). In plants from brown seed, the transcript accumulation of PEPC at 100 mM (t 10 ¼ 4.448, P ¼ 0.0012) and 300 mM (t 10 ¼ 4.304, P ¼ 0.0016) was significantly higher than that of plants from black seed; in plants from black seed, the RUBPC expression level was significantly higher than that of plant from brown seed under various salt concentrations (t 10 ¼ 11.27, P , 0.0001 for 100 mM; t 10 ¼ 17.25, P , 0.0001 for 300 mM; t 10 ¼ 6.685, P , 0.0001 for 500 mM). Both of above indexes was significantly affected by the interaction of seed type and salt concentration (F 3,40 ¼ 6.413, P ¼ 0.0012 for PEPC; F 3,40 ¼ 17.24, P , 0.0001 for RUBPC). The expression patterns of PEPC and RUBPC agreed with the enzyme activities of PEPC and RUBPC, respectively.

Discussion
Suaeda aralocaspica is a C4 type halophyte with dimorphic seeds, whose difference in germination and dormancy has been well documented Wang et al. 2008;He et al. 2013). However, the effects of the dimorphic seeds on the descendants have not been well understood. Wang et al. (2014) recently compared the biomass and mineral profiles of plants from dimorphic seeds of S. aralocaspica under different levels of nutrient and salinity. The present work systematically investigated responses of growth, physiology and relevant gene expression of plants from different seed morphs to long-term salt treatment for the first time. Results indicate that plants grown from brown and black seeds showed no significant difference for most indexes, which suggests that the different properties between the dimorphic seeds of S. aralocaspica are not transferred to the descendants on growth and physiological responses with or without salinity and both types of plants require salinity in growth to cope with the saline habitat.
In the present study, plant growth of S. aralocaspica was stimulated by the increasing salt concentration, and no significant difference was presented between both types of plants. Similar results were observed from the seed dimorphic halophyte S. splendens: seedlings grown from AoB PLANTS www.aobplants.oxfordjournals.org dimorphic seeds all grew well at high salinity (400 mM) (Redondo-Gómez et al. 2008). Although the effect of salinity on growth varies among halophytes, many halophytes show optimal growth in concentrations 200 mM NaCl (Flowers et al. 1986), and organic dry mass is stimulated by growth in saline conditions in at least some dicotyledonous halophytes (Yeo and Flowers 1980;Glenn and O'Leary 1984). Many natural halophytes can grow at similar rates to glycophytes even though they use energy from the ion accumulation and compartmentalization for  osmotic adjustment (Yeo 1983). Our results suggest that S. aralocaspica is a euhalophyte and can benefit from salt for vegetative growth.
In the present study, only the higher salt concentration (500 mM) caused the Na + to accumulate to a significant level in plants of both seed morphs, whereas K + concentration and the consequent K + /Na + ratio were reduced, which indicates that competitive inhibition between Na + and K + absorption results in a change in the balance of intracellular K + -Na + . Under salinity, the up-regulation of the vacuolar H + -ATPase genes (VHA), SsVHA-H and SsVHA-B and the increase in the activity of V-H + -ATPase provide the proton-driving force for sequestering Na + in leaf vacuoles of Suaeda salsa (Li et al. 2004(Li et al. , 2006. In the present study, the expression levels of B subunit of V-H + -ATPase gene were significantly increased under salt treatment in plants from black seed (V-H + -ATPase) and the expression of V-H + -PPase increased in plants from both seed types, but why the expression pattern of SaV-ATPase showed no obvious change in plants from brown seed was not clear. These results are in agreement with those found in another halophyte Salicornia europaea, in which the transcript accumulation of SeVHA-A and SeVP1 (encoding for V-H + -PPase) in shoots increased significantly with rising NaCl concentration (Lv et al. 2012). This sequestration requires Na + being actively transported into the vacuole against the electrochemical gradient and Na + in the vacuole being prevented from leaking back into the cytosol (Bonales-Alatorre et al. 2013), the high expression of these genes and corresponding proteins play important roles in this process.
Proline, GB and SSs serve as the important organic osmolytes that contribute to alleviate the low cellular water potential (Sakamoto and Murata 2002;Flowers and Colmer 2008;Trovato et al. 2008). In the present study, with the increasing salt concentration, proline, GB and all SSs measured were significantly increased in plants from both seed morphs. Consistent with the above results, the expression levels of P5CS (for proline synthesis) and BADH (for GB synthesis) genes increased with the rising NaCl concentration in plants from both seed types. SsP5CS in S. salsa was also up-regulated under salt stress (Wang et al. 2002). The BADH level increased under salt treatment in S. aralocaspica (Park et al. 2009). The above results suggest that these osmolytes must have important function in osmo-regulation of the cells in S. aralocaspica.
Salt stress can generate secondary oxidative stress by production of excess reactive oxygen species (ROS), e.g. AoB PLANTS www.aobplants.oxfordjournals.org significantly increased only at higher rather than lower salt concentration in both types of plants. These data indicated that only higher salinity may lead to damage of the cellular membrane (corresponding to significant accumulation of MDA), and the results suggest that the antioxidant agents must function actively to scavenge excess ROS under higher salinity. The metabolism of ROS depends on the synergetic function of multiple antioxidant enzymes, in which SOD has been regarded as the first defence line by catalysing the dismutation of O 2 − to molecular oxygen and H 2 O 2 (Badawi et al. 2004;Ashraf 2009), and the H 2 O 2 can be further scavenged predominantly by CAT and POD (Willekens et al. 1997;Azevedo Neto et al. 2006). In the present study, the activity of SOD significantly increased with the rising O 2 − , while that of CAT and POD decreased significantly in both types of plants under higher salinity, suggesting that the failure of an increase of the activity of CAT and POD might mean that the excess H 2 O 2 was not effectively scavenged, thereby causing more serious oxidative stress. Several other enzymes also play important roles in detoxification of H 2 O 2 via the AsA-GSH cycle (Bowler et al. 1992). Ascorbate peroxidase and GR, which are the first and last enzymes in this cycle, respectively, are responsible for H 2 O 2 scavenging in green leaves (Foyer and Harbinson 1994). Ascorbate peroxidase employs AsA as a specific electron donor to reduce H 2 O 2 to H 2 O (Asada 1992), and GR has a central role in maintaining the reduced GSH pool during stress (Pastori et al. 2000;Alhdad et al. 2013). In the present study, the activities of APX and GR increased significantly with rising salt concentration in both types of plants, which was consistent with the increasing trend of two corresponding non-enzymatic antioxidants AsA and GSH. These two agents were reported as the most powerful water-soluble antioxidants and could minimize the oxidative damage either by directly scavenging dangerous ROS or via the AsA-GSH cycle (Foyer et al. 1997;Ashraf 2009). The fully oxidized AsA has a short half-life and would be lost unless it is reduced back, in which GSH plays a key role in regenerating AsA via the AsA -GSH cycle (Foyer and Halliwell 1976;Gill and Tuteja 2010). GSH concentration usually declines with the increasing of stress intensity (Tausz et al. 2004). However, in the present study, the levels of AsA and GSH accumulated significantly when both types of plants were exposed to higher salt concentration, suggesting that these two non-enzymatic antioxidants play a predominant role in endowing this halophyte with resistance to potential oxidative damage compared with antioxidant enzymes. Plant cells contain three SOD types (i.e. Fe-SODs, Mn-SODs and Cu/Zn-SODs) that differ in their metal cofactors (Attia et al. 2008). The transcriptional expression of the most abundant isoform, Cu/Zn-SOD (gene as cSOD), located in cytosol and chloroplast (Bowler et al. 1994), was analysed in the present study. Results showed that cSOD was up-regulated at low salt concentration, whereas down-regulated at higher salt concentration, which was incompletely correlated with the activity of SOD at higher NaCl concentration. This difference may correspond to post-translational regulation, inactivation of the enzymes by their product H 2 O 2 or direct inhibition by Na + at higher salinity (Scandalios 1993;Madamanchi et al. 1994). The transcriptional level of SasAPX in the present study increased significantly in both types of plants with the rising salt concentration, which was consistent with the corresponding enzyme activity: it suggests an important role of this gene in the protection against higher salinity-induced oxidative stress in S. aralocaspica. Analysis of the expression of SaGR in the present study showed that gene expression with the rising salt concentration correlated with the changes of GR activity and GSH concentration in S. aralocaspica plants from dimorphic seeds, which implies that the increases in gene expression and GR activity are not only required for the scavenging of ROS. Glutathione reductase activity needed sustained high levels of GSH for adjusting cellular redox homeostasis (Pastori et al. 2000;Contour-Ansel et al. 2006).
High salinity also considerably affects the process of photosynthesis in most plants by altering the ultrastructure of the organelles, concentration of various pigments and metabolites and enzyme activities involved in this process (Ashraf and Harris 2013). In the present study, the total Chl concentration significantly decreased after exposing plants to higher salinity, which may result in the destruction of the chloroplast structure. The impaired biosynthesis or accelerated degradation of photosynthetic pigment was caused by cholorophyllase (Rao and Rao 1981;Singh and Dubey 1995). In the present study, the activities of two key photosynthetic enzymes were measured and the results showed that activity of RUBPC was significantly decreased with the rising salt concentration, while that of PEPC increased significantly. It is known that S. aralocaspica is a kind of single-cell C4 plant species (Voznesenskaya et al. 2001), although RUBPC as major carbon fixation enzyme is the limiting factor of the rate of photosynthesis (Mitra and Baldwin 2008). In C4 species, CO 2 is initially fixed by PEPC in the mesophyll cells (Aldous et al. 2014) and refixed by RUBPC (Sage 2004). Increased salt concentration is suggested to repress RUBPC activity of both glycophytes and halophytes (Osmond and Greenway 1972;Kaiser and Heber 1981). The decreased activity of RUBPC may be caused by a salt-induced reduction of total water content (TWC) and accumulation of proline (Sivakumar et al. 1998;Lawlor and Cornie 2002), which is in agreement with the results of TWC and proline analysis in the present study. The suppression of RUBPC activity by accumulated proline might weaken hydrophobic interactions between subunits of the enzyme and finally dissociating the small subunits from the large subunits (Sivakumar et al. 2001). However, the activity of PEPC can be used as biochemical indicator of salt tolerance (Guerrier 1988). The increased PEPC activity upon salt stress could potentially improve carbon metabolism during periods of reduced stomatal conductance by reassimilating respired CO 2 and/or increasing rates of CO 2 fixation at night when stomata are open (Cushman and Borland 2002;García-Mauriñ o et al. 2003;Carmo-Silva et al. 2008). In a C4 shrub Atriplex lentiformis (grown in saline habitats), the net CO 2 assimilation rate and the PEPC activity in leaf increase linearly with salinity rising (Zhu and Meinzer 1999). Phosphoenolpyruvate carboxylase may also support the biosynthesis of biocompatible osmolytes such as proline (Chen et al. 2010). Being consistent with the corresponding enzyme activity, our data showed that the transcriptional level of SaPEPC was increased significantly, while that of SaRUBPC was decreased significantly in both types of plants with the rising salt concentration. An up-regulated PEPC in response to salinity or drought stress has been well documented in C3, C4 and crassulacean acid metabolism plants (Li and Chollet 1994;Gonzá lez et al. 2003;Carmo-Silva et al. 2008). Specifically, a major up-regulation of PEPC and concomitant down-regulation of RUBPC are typical of this response (Fontaine et al. 2003;Dizengremel et al. 2009).
There is evidence that brown seeds are able to germinate faster and at higher percentage than black seeds in species with dimorphic seeds Redondo-Gó mez et al. 2008;Song et al. 2008;Wang et al. 2008). In the halophyte Suaeda maritima, large seeds have Values are means + SE of six replicates (including three biological replicates and two technical replicates). Bl, black seed; Br, brown seed.
AoB PLANTS www.aobplants.oxfordjournals.org larger cotyledons than small seeds, which imply that more nutrient reserves can be used for better germination of the large S. maritima seeds (Wetson et al. 2008). In the halophyte Atriplex triangularis, large seeds produce larger plants than do in small seeds and this dichotomy can be maintained throughout the whole life of the species (Ellison 1987). Similarly, in Hedypnois cretica and Crepis aspera, larger peripheral achenes produced bigger seedlings than did in smaller central achenes and the effect of different seed size on plant growth could sustain up to 40 days post emergence in the progeny of both species (El-Keblawy 2003). We found in our previous observation that the seedlings from brown seeds of S. aralocaspica emerged earlier and with higher uniformity than those of black seeds; however, the difference between two types of seedlings became indistinguishable in 2-3 weeks. After stratification of black seeds, the seedlings from brown and black seeds showed similar performance in emergence time, uniformity and the size. Further comparison between plants derived from heteromorphic seeds in the present study, we found that there was no significant difference on the performance with most tested physiological parameters and relevant gene expressions with or without salinity (Table 3), most of the indexes were induced under salt treatment and most of the performance of physiological parameters was matched with corresponding gene expression patterns. It suggests that the difference between two types of seedlings in the early stage in S. aralocaspica arose because of different rates of emergence time rather than for other reasons.

Conclusions
Our data suggest that different concentrations of NaCl stimulated similar active responses of all descendants from dimorphic seeds of S. aralocaspica on osmotic, antioxidative and photosynthetic systems, which were consistent with plant growth and morphological changes, as well as gene expression patterns (Table 3). Plants grown from both seed morphs all performed well at higher salinity, which may be attributed to synergetic actions among osmotic homeostasis, antioxidative defence and photosynthetic functions. Osmolytes (proline and betaine) were significantly increased and the excess ROS produced by higher salinity were scavenged by increased levels of antioxidant enzymes (SOD, APX and GR) and corresponding antioxidants (AsA and GSH). Enhancement of PEPC activity at high salt intensity should be a positive effect on maintaining normal photosynthesis. Most tested parameters showed no difference between plants from the two types of seeds, but a few were different: the significance of these few differences is not clear. Our results suggest that there