Plant Growth Regulation 41: 185–195, 2003. # 2003 Kluwer Academic Publishers. Printed in the Netherlands.

185

Hormones in the grains in relation to sink strength and postanthesis development of spikelets in rice Jianchang Yang1,*, Jianhua Zhang2, Zhiqin Wang1 and Qingsen Zhu1 1

College of Agriculture, Yangzhou University, Yangzhou, Jiangsu 225009, China; 2Department of Biology, Hong Kong Baptist University, Kowloon Tong, Hong Kong, China; *Author for correspondence (e-mail: [email protected]; phone: +86 514 7979087; fax: +86 514 7349817)

Received 26 March 2003; accepted in revised form 1 July 2003

Key words: Abscisic acid, Cytokinins, Gibberellins, Indole-3-acetic acid, Rice (Oryza sativa L.), Sink strength

Abstract Inferior spikelets usually exhibit a slower grain filling rate and lower grain weight than superior spikelets in a rice (Oryza sativa L.) panicle. This study investigated whether the variations in grain filling between the two kinds of spikelets were attributed to their sink strength and whether the sink strength was regulated by the hormonal levels in the grains. Using two field-grown rice genotypes, the division rate of endosperm cells, hormonal levels in the grains, and grain weight of both superior and inferior spikelets were determined during the grain filling period. The results showed that superior spikelets had dominance over inferior spikelets in endosperm cell division rate and cell number, grain filling and grain weight. Changes in zeatin (Z) and zeatin riboside (ZR) contents paralleled and were very significantly correlated with the cell division rate and cell number. Cell division rate and the content of indole-3-acetic acid (IAA) in the grains were also significantly correlated. Gibberellin (GAs; GA1 + GA4) content of the grains was high but ABA levels were low at the early grain filling stage. ABA increased substantially during the linear phase of grain growth and was very significantly correlated with grain dry weight during this period. Application of kinetin at 2 through 6 days post anthesis (DPA) significantly increased cell number, while spraying ABA at 11 through 15 DPA significantly increased the grain filling rate. The results suggest that differences in sink strength are responsible for variations in grain filling between superior and inferior spikelets. Both cytokinins and IAA in the grains may mediate cell division in rice endosperm at early grain filling stages, and therefore regulate the sink size of the grain, whereas ABA content correlates with sink activity during the linear period of grain growth. Abbreviations: ABA – abscisic acid; DPA – days post anthesis; ELISA – enzyme-linked immunosorbent assay; GAs – gibberellins; IAA – indole-3-acetic acid; Z – zeatin; ZR – zeatin riboside

Introduction A rice panicle is composed of a number of spikelets or caryopses. Based on their flowering date and locations within a panicle, the spikelets can be classified as superior or inferior (Zhu et al. 1988;

Iwasaki et al. 1992; Umemoto et al. 1994). In general, superior spikelets flower earlier and are located at the top of primary branches, whereas inferior spikelets flower later and are located at the base of secondary branches. Superior spikelets usually exhibit a faster rate of increase in dry

186 weight during development and higher final grain weight than inferior spikelets (Sikder and Gupta 1976; Kato 1989). Slow grain filling and low grain weight of inferior spikelets have often been attributed to a limitation in carbohydrate supply (Sikder and Gupta 1976; Wang 1981; Murty and Murty 1982). However, more recent work has shown that there is no clear causative relationship between assimilate concentration and spikelet development in rice (Mohapatra and Sahu 1991; Mohapatra et al. 1993). Our earlier work (Wang et al. 1998; Yang et al. 1999) showed that inferior spikelets contained a higher level of soluble sugars than superior spikelets at the early grain filling stage. However, the intrinsic factors responsible for variations in grain filling between superior and inferior spikelets remain elusive. It is generally believed that grain-filling rate in cereals is closely associated with sink strength (Venkateswarlu and Visperas 1987; Liang et al. 2001). During the grain filling period, rice grains are strong carbohydrate sinks (Cao et al. 1992). The sink strength can be described as the product of sink size and sink activity (Warren 1972; Venkateswarlu and Visperas 1987). Sink size is a physical restraint that includes cell number and cell size and sink activity as the physiological constraint upon a sink organ’s assimilate import (Ho 1988). It is assumed that hormones in a sink organ are prominent factors in determining sink strength (Bangerth 1989; Brenner and Cheikh 1995). In cereals, high levels of cytokinins are generally found in the endosperm of developing seeds, and may be required for cell division during the early phase of seed filling (Michael and Seiler-Kelbitsch 1972; Saha et al. 1986; Morris et al. 1993; Dietrich et al. 1995; Banowetz et al. 1999). Although cytokinins are generally considered to play a major role in increasing sink size by promoting cell division, little information is available with regard to their effect on endosperm cell number (Brenner and Cheikh 1995). Auxins, gibberellins (GAs), and abscisic acid (ABA) are also thought to be involved in regulating sink strength either by mediating the division and enlargement of endosperm cells or controlling import of assimilates to the sink (Karssen 1982; Davies 1987; Kende and Zeevaart 1997; Hansen and Grossmann 2000). There are many reports on the correlation between the level of ABA and the growth rate of fruits or seeds (e.g., Eeuwens and

Schwabe 1975; Browning 1980; Ber€ uter 1983; Wang et al. 1987; Ross and McWha 1990; Schussler et al. 1991; Kato et al. 1993; Yang et al. 2001). However, the proposal that ABA is involved in the regulation of assimilate partitioning towards developing seeds has remained disputable (Jones and Brenner 1987; Barratt et al. 1989; Ober and Setter 1990; Schussler et al. 1991; de Bruijn and Vreugdenhil 1992; Sharp and LeNoble 2002). The purposes of this study were to investigate the processes of division of endosperm cells and grain filling and changes in cytokinins, indole-3-acetic acid (IAA), GAs and ABA in both the superior and inferior spikelets of rice during grain filling, and determine whether the variations in grain filling between superior and inferior spikelets were attributed to their sink strength and whether the sink strength was regulated by the hormonal levels in the grains.

Materials and methods Plant materials The experiment was conducted at a farm of Yangzhou University, Jiangsu Province, China (32
300 N, 119
250 E) during rice growing season (May–October) of 2000, and repeated in 2001. Two rice genotypes, Wuyujing 3 (a japonica inbred cultivar) and LYP-9 (Pei-ai 64S/Yangdao 7, an indica/ indica F1 hybrid), were grown in the paddy field. Seedlings were raised in the field with a sowing date 10 May and transplanted on 11 June at a hill spacing of 0.2  0.2 m with two seedlings per hill. Plot dimension were 4  5 m. Each of the genotypes had three plots as repetitions in a complete randomized block design. The soil of the field was sandy loam with 24.5 g kg1 organic matter and available N–phosphorus–potassium at 106, 33.8 and 66.4 mg kg1, respectively. N (60 kg ha1 as urea), phosphorus (30 kg ha1 as single superphosphate) and potassium (40 kg ha1 as KCl) were applied and incorporated before transplanting. N as urea was also applied at mid-tillering (40 kg ha1) and at panicle initiation (25 kg ha1). Both genotypes (50% of plants) headed on 21 through 23 August, and were harvested on 10 October. Except for drainage at end-tillering (11–15 July), the field was kept at 1–2-cm water level during the

187 whole growth period. The temperatures, averaged per 10 days from anthesis (21–23 August) to harvest, were 27.4, 26.5, 25.3, 24.1, and 23.5
C, respectively. Sampling Three hundred panicles that headed on the same day were chosen and tagged for each plot. The flowering date and the position of each spikelet on the tagged panicles were recorded. Fifteen tagged panicles from each plot were sampled at 2-day intervals from anthesis to 22 days post anthesis (DPA) and at 4-day intervals from 24 DPA to maturity. Superior spikelets which flowered on the first 2 days within a panicle and inferior spikelets which flowered on the last 2 days within a panicle were separated from the sampled panicles. The difference in flowering date between superior and inferior spikelets within a panicle was 3 days for Wuyujing 3 and 5 days for LYP-9. Two hundred to 240 sampled grains from superior or inferior spikelets formed each sample. Half-sampled grains were frozen in liquid nitrogen for 1 min and then stored at 80
C for hormonal assay. Eighty to 100 sampled grains were used for measurements of grain dry weight and soluble carbohydrate. These were dried at 70
C to constant weight for 72 h, and weighed. Soluble carbohydrate in the grains was determined as described by Yoshida et al. (1976). After cutting a small hole on the edge of a hull, 10–12 grains either from the superior or the inferior spikelets were fixed in Carnoy’s solution (absolute ethanol : glacial acetic acid : chloroform ¼ 9 : 3 : 1, v/v) for 48 h, then kept in 70% (v/v) ethanol pending examination of endosperm cell number. Nuclear/cell counting The method for isolation and counting of endosperm cells was modified from Singh and Jenner (1982). Briefly, fixed grains were dehulled and transferred into 50 and 25% (v/v) ethanol, respectively, and finally into distilled water for 5–7 h prior to dissection of the endosperm. The endosperm was isolated under a dissecting microscope and dyed with a Delafied’s haematoxylin solution for 24–30 h, washed several times with distilled water and then hydrolysed in 0.1% (w/v) cellulase

(No. c-2415, Sigma, St Louis, MO, USA) solution (pH 5.0) at 40
C for 4–6 h and oscillated. The isolated endosperm cells were diluted to 2–10 ml according to the development stage of endosperm, from which eight to 10 sub-samples (20
l for each sub-sample) were taken to a counting chamber (1 cm2 area). Using a light microscope, the endosperm cell number of 10 fields of view for each counting chamber was noted. Within 2 DPA for superior spikelets and 4–6 DPA for inferior ones, the number of nuclei was counted as endosperm cell number. The total cell number per endosperm was calculated according to Liang et al. (2001). Six grains (endosperms) were examined for each genotype at each measurement. The division process of endosperm cells was fitted by Richards’ (1959) growth equation as described by Zhu et al. (1988): M ¼ A=ð1 þ B ekt Þ1=N

ð1Þ

where M is the cell number, A is the maximum cell number, t is the time after anthesis (days), and B, k, and N are coefficients determined by regression. The rate of endosperm cell division (R) was calculated the derivative of the Equation (1): R ¼ AKB ekt =Nð1 þ B ekt ÞðNþ1Þ=N

ð2Þ

The active period of endosperm cell division was defined as that when M was from 5% (t1) to 95 (t2) of A. The average rate of endosperm cell division during this period was calculated from t1 to t2. Hormonal extraction, purification and quantification The methods for extraction and purification of zeatin (Z), zeatin riboside (ZR), IAA, GAs (GA1 + GA4), and (±) ABA were modified from those described by Bollmark et al. (1988) and He (1993). Samples consisting of 50–80 dehulled and frozen grains were ground in an ice-cold mortar in 5–10 ml 80% (v/v) methanol extraction medium containing 1 mM butylated hydroxytoluene as an antioxidant. The extract was incubated at 4
C for 4 h and centrifuged at 4000 rpm for 15 min at the same temperature. The supernatants were passed through Chromosep C18 columns (C18 Sep-Park Cartridge, Waters, Millford, MA, USA), prewashed with 10 ml 100% (v/v) and 5 ml 80% (v/v)

188 methanol, respectively. The hormone fractions were dried under N2, and dissolved in 2 ml phosphate-buffered saline (PBS) containing 0.1% (v/v) Tween 20 and 0.1% (w/v) gelatin (pH 7.5) for analysis by an enzyme-linked immunosorbent assay (ELISA). The mouse monoclonal antigens and antibodies against Z, ZR, IAA, GAs (GA1 + GA4), and ABA, and immunoglobulin G-horseradish peroxidase (Ig G-HRP) used in ELISA were produced at the Phytohormones Research Institute (China Agricultural University; see He 1993). The method for quantification of Z, ZR, IAA, GAs, and ABA by ELISA was described previously (Yang et al. 2001).

et al. 2001). Twenty plants from each treatment were harvested at maturity for examination of final grain weight. Statistical analysis The results were analyzed for variance using the SAS statistical analysis package (version 6.12, SAS Institute, Cary, NC, USA). Data from each sampling date were analyzed separately. Means were tested by least significant difference at P0.05 level (LSD0.05). Linear regression was used to evaluate the relationship between hormonal contents in the grains with the cell division rate of division of endosperms and grain dry weight during grain filling.

Application of plant growth regulators Results Plants of the hybrid LYP-9 were grown in porcelain pots that were placed in a field. Each porcelain pot (30 cm height, 25 cm diameter, 14.72 l volume) was filled with 18 kg sandy loam soil with the same nutrient contents as the field soil. Thirty-day-old seedlings raised in the field were transplanted on 11 June into the pots with three hills per pot and one seedling per hill. Two grams of N as urea, 0.2 g of phosphorus as single superphosphate, and 0.3 g of potassium as KCl were mixed into the soil in each pot before transplanting. At mid-tillering and panicle initiation, 0.6 and 1.2 g of N as urea were top dressed into each pot, respectively. The plants were watered daily by hand and the pot was kept at 1–2-cm water level during the whole growth period. Gibberellic acid (GA3) (Shanghai Chem & Biol Co., China), kinetin (6-furfuryl amino purine), and ABA (both from Sigma) were applied at two stages either at initial grain filling starting 2 DPA (T1) or at early grain filling starting 11 DPA (T2). At each stage, either 25  106 M ABA, or 50  106 M kinetin, or 100  106 M GA3 were sprayed at the rate of 50 ml per pot on the leaves and panicles daily for 5 days with 0.5% (v/v) Teepol (Fluka, Riedel-de-Haen) as surfactant. The plants sprayed with the same volume of 0.5% Teepol solution were taken as a control. Each treatment had 50 pots. Endosperm cell number and grain dry weight were measured with the methods described above, with four replications for each measurement. The grain filling process was fitted by Richards’ (1959) growth equation as described previously (Yang

Cell division rate and grain filling Figure 1 illustrates the division progression and rate of endosperm cells in both superior and inferior spikelets. Both genotypes exhibited a faster division rate and higher cell number in superior spikelets than in inferior ones, indicating that the former had dominance over the latter in both the speed of sink enlargement and sink size. There were no obvious differences in the division rate and cell number in superior spikelets in the inbred cultivar and the hybrid. However, inferior spikelets of the hybrid had a slower division rate and smaller number of endosperm cells, when compared with the same kind of spikelets of the inbred cultivar. Very similar to the changing pattern of endosperm cell number, grain dry weight increased much faster in the superior than in inferior spikelets (Figure 2). The more endosperm cells in spikelets, the higher the grain weight, suggesting that sink size played dominant role in determination of grain weight in rice. As opposed to their lower grain dry weight, inferior spikelets contained a much higher concentration of soluble carbohydrate than superior spikelets at the early grain filling stage (Figure 3A,B), consistent with previous observations (Mohapatra et al. 1993; Wang et al. 1998; Yang et al. 1999). The high concentration of soluble carbohydrate was mainly attributable to a high concentration of sucrose in these spikelets (Figure 3C,D).

189

Figure 1. Cell numbers (A,B) and cell division rates (C,D) in the endosperms of superior (closed circles) and inferior (open circles) spikelets of the japonica cultivar Wuyujing 3 (A,C) and indica hybrid LYP-9 (B,D). The division rate of endosperm cells was calculated according to Richards’ (1959) equation. Vertical bars in (A,B) represent ± S.E. of the mean (n ¼ 6) where these exceed the size of the symbol.

Hormonal changes in the grains The methods used in this study for hormone extraction and purification and for quantification hormones in the grains by ELISA recovered 73.5% of Z, 75.7% of ZR, 68.9% of IAA, 74.8% of GAs (GA1 + GA4), and 82.1% of ABA (Table 1). Crossactivities for antibodies used were very small (<0.1). The specificity of the monoclonal antibodies and the other possible nonspecific immunoreactive interference were checked previously and proved reliable (Wu et al. 1988; Zhang et al. 1991; Yang et al. 2001). The grains contained about 40% more ZR than Z, and both showed a similar pattern of change in superior and inferior spikelets during grain filling (data not shown). Z + ZR contents in grains increased transiently in both kinds of spikelets at the early grain filling stage, and reached a

maximum at 6 DPA for the superior and 8 DPA for the inferior spikelets, decreasing thereafter (Figure 4A,B), in good agreement with the division rate of endosperm cells as shown in Figure 1. At the early grain filling stage, superior spikelets contained more Z + ZR than inferior spikelets. Similar to the changes in Z + ZR, IAA content in the grains increased sharply during the early grain filling stage, reaching a maximum at 4 DPA for the superior and 6 DPA for the inferior spikelets, and falling quickly afterwards (Figure 4C,D). In comparison with that in superior spikelets, IAA content in inferior spikelets, especially for the hybrid, was much lower at 2–6 DPA. GAs (GA1 + GA4) contents in both superior and inferior spikelets showed a similar changing pattern, increasing sharply at the early grain filling stage, reaching a maximum at 8–10 DPA, and decreasing very quickly thereafter (Figure 3E,F).

190 with r ¼ 0.96** to 0.97** (P < 0.01). IAA contents in the grains were also positively and significantly correlated with the cell division rate (r ¼ 0.71* to 0.74*, P < 0.05), whereas neither GAs nor ABA was significantly correlated with cell division (r ¼ 0.41 to 0.23, P > 0.05). ABA contents in the grains, however, were significantly and positively correlated with grain dry weight (r ¼ 0.98**, P < 0.01) during the linear increase period of grain dry weight (4–28 DPA for superior and 8–40 DPA for inferior spikelets). Effects of plant growth regulators Application of kinetin at the initial grain filling stage (2–6 DPA) significantly increased the total cell number per endosperm and grain weight, whereas ABA had the opposite effect (Table 2). Spraying ABA at the linear grain filling stage (11–15 DPA) significantly increased the grain filling rate and grain weight. Spraying kinetin at this stage and application of GA3 at either the initial or linear grain filling stage had no significant effect (Table 2). Figure 2. Grain dry weight of superior (closed circles) and inferior (open circles) spikelets of the japonica cultivar Wuyujing 3 (A) and indica hybrid LYP-9 (B). Vertical bars represent ± S.E. of the mean (n ¼ 3) where these exceed the size of the symbol.

Superior spikelets also contained more GAs than inferior spikelets at the early grain filling stage but such differences were rather small. In contrast to rapid increase of Z + ZR, IAA, and GAs, ABA content in both types of spikelets increased very slowly at the initial grain filling stage (Figure 4G,H). ABA increased markedly from 6 DPA for the superior and 10–12 DPA for inferior spikelets, reaching a maximum at 24–28 DPA for the superior and 36–40 DPA for inferior spikelets, in good agreement with the increase of grain dry weight (refer to Figure 2). ABA content was greater in the superior than in the inferior spikelets at early and mid grain filling stages, but this was reversed at the late grain filling stage. During the active cell division period (2–8 DPA for superior and 4–12 DPA for inferior spikelets), Z + ZR contents in grains were positively and very significantly correlated with the cell division rate

Discussion The endosperm of rice contributes more than 90% of the final grain weight of a caryopsis (Murata and Matsushima 1975; Cao et al. 1992). Thus, sink size is determined by the number of cells and cell size in the endosperm and the number of amyloplasts per endosperm cell. Our results show that increase in grain dry weight was closely associated with the rate of division of endosperm cells (Figures 1 and 2). The higher the number of endosperm cells, the higher was the grain weight. This suggests that cell number of rice endosperm plays a dominant role in determining sink size and grain weight. Slow grain filling and low grain weight for inferior spikelets would be mainly attributed to their slow sink enlargement and small sink size. It is generally presumed that constraints in assimilate availability for inferior spikelets, especially in hybrid rice, results in their poor sink strength and leads to producing poor quality grains (Venkateswarlu and Visperas 1987; Cao et al. 1992; Yuan 1997; Liang et al. 2001). We observed here that the active cell division period in inferior

191

Figure 3. Concentrations of soluble carbohydrate (A,B) and sucrose (C,D) in superior (closed circles) and inferior (open circles) spikelets of the japonica cultivar Wuyujing 3 (A,C) and indica hybrid LYP-9 (B,D). Vertical bars represent ± S.E. of the mean (n ¼ 3) where these exceed the size of the symbol.

Table 1. Recovery test of enzyme-linked immunosorbent assay for zeatin (Z), zeatin riboside (ZR), indole-3-acetic acid (IAA), gibberellins (GAs), and abscisic acid (ABA).

Hormone

Graina (pmol g1 FW)

Grain + standardb (pmol g1 FW)

Recovery (%)

Z ZR IAA GAs ABA

468 ± 44 657 ± 53 964 ± 78 751 ± 86 723 ± 69

932 ± 82 1102 ± 97 1215 ± 113 1160 ± 96 1250 ± 112

73.5 ± 5.1 75.7 ± 6.2 68.9 ± 7.5 74.8 ± 6.9 82.1 ± 6.3

a

Grains only. Grains + synthetic compounds of Z, ZR, IAA, GAs (GA1 + GA4), and ABA, respectively. Eight hundred picomole of each compound was added to 1 g of fresh grains before purification. Z, ZR, IAA, and ABA were from Sigma Chem Co. and GAs were provided by China Agricultural University. The test cultivar was Wuyujung 3 (Japonica). Data are expressed as means ± S.E. of four replications. b

spikelets was 4–12 DPA (Figure 1). In this period, the concentration of soluble carbohydrate in inferior spikelets doubled that in superior spikelets (Figure 3). The hybrid had the highest concentration of soluble carbohydrate in inferior spikelets at the early grain filling stage, but showed the smallest

cell division rate. Thus, it can be concluded that the intrinsic concentrations of assimilates should not limit the division rate of endosperm cells in inferior spikelets. Our data clearly showed that the division rate of endosperm cells paralleled and was very

192

Figure 4. Contents of zeatin (Z) + zeatin riboside (ZR) (A,B), indole-3-acetic acid (IAA) (C,D), gibberellins (GAs) (E,F), and abscisic acid (ABA) (G,H) in superior (closed circles) and inferior (open circles) spikelets of the japonica cultivar Wuyujing 3 (A,C,E,G) and indica hybrid LYP-9 (B,D,F,H). Vertical bars represent ± S.E. of the mean (n ¼ 3) where these exceed the size of the symbol.

193 Table 2. Effects of applying abscisic acid (ABA), kinetin, and gibberellic acid (GA3) on the total cell number per endosperm, grain filling rate, and grain weight of pot-grown rice.

Type of spikelet

Treatment

Total cell number (103 cells endosperm1)

Grain filling rate (mg grain1 day1)

Grain weight (mg grain1 day1)

Superior

Control 25  106 M ABA (T1) 50  106 M kinetin (T1) 100  106 M GA3 (T1) 25  106 M ABA (T2) 50  106 M kinetin (T2) 100  106 M GA3 (T2) Control 25  106 M ABA (T1) 50  106 M kinetin (T1) 100  106 M GA3 (T1) 25  106 M ABA (T2) 50  106 M kinetin (T2) 100  106 M GA3 (T2)

265 247* 278* 264 264 266 263 192 146** 218* 196 187 194 191

1.01 0.96 1.09 0.94 1.25* 0.91 0.92 0.58 0.57 0.58 0.55 0.90** 0.53 0.52

29.8 27.9* 31.5* 28.5 31.4* 29.4 28.4 22.7 16.6** 25.5* 23.1 26.7** 23.6 21.9

Inferior

An indica hybrid LYP-9 was used for the test. Plant growth regulators were applied either at the initial grain filling starting 2 days post anthesis (T1) or at linear grain filling starting 11 days post anthesis (T2). The leaves and panicles were sprayed either with 25  106 M ABA, 50  106 M kinetin, or 100  106 M GA3 daily for 5 days at each stage. The grain-filling rate was calculated according to Richards’ (1959) equation. Values of grain weight were the means of 20 plants harvested at maturity. Statistical comparison is within the same column and the same type of spikelets. *,**Values significantly different from the control at P ¼ 0.05 and P ¼ 0.01 levels, respectively.

significantly correlated with the changes in cytokinin (Z + ZR) levels in the grains. Superior spikelets that had higher Z + ZR contents also exhibited a higher cell division rate. In contrast, inferior spikelets of the hybrid that contained the lowest levels of Z + ZR had the slowest cell division rate. When kinetin was applied to the plant at the initial grain filling stage (2–6 DPA), both the total cell number per endosperm and grain weight were significantly increased (Table 2). These results suggest that cytokinin levels in the grains mediate the cell division in rice endosperm, and therefore regulate the sink size of the grain. Low cytokinin levels in inferior spikelets result in reduced sink size by limiting cell number and lead to low grain weight. We observed that changes in IAA content in the grains were similar to the changing pattern of Z + ZR (Figure 4). IAA content in the grains was significantly correlated with the cell division rate during the active cell division period. The results support the hypothesis that auxin may stimulate cell division in combination with cytokinins (Davies 1987). High auxin levels in the sink could create an ‘attracting power’, leading to increased cytokinin levels in the grain (Seth and Waering 1967; Singh and Gerung 1982).

Two sharply contrasting patterns were found between GAs (GA1 + GA4) and ABA contents in the grains during the grain filling period (Figure 4). GAs in both superior and inferior spikelets were high at the early grain filling stage and had maximal levels at 8–10 DPA, which were associated with the rapid enlargement of the embryo (Qin and Tang 1984). However, GAs levels in the grains were not correlated with the cell division rate. Spraying GA3 on the plants at either the initial or linear grain filling stage had no significant effect on the cell number and grain weight (Table 2). A similar observation was made by Pharis (1985) who reported that the rice cultivar Tan-ginbozu had only one-fifth as much bioactive GAs as a normal cultivar or the heavy grain, semidwarf Tong-il, but the final dry weight per grain was essentially the same for all cultivars. These results suggest that GAs may not regulate the cell division of endosperm during grain filling, although they may play an important role in embryogenesis in rice. In contrast to GAs, ABA content in rice grains was rather low at the initial grain filling stage, and increased substantially at the linear phase of grain growth (Figure 4). There have been controversial

194 debates about the role of ABA in sink growth (Jones and Brenner 1987; Barratt et al. 1989; Schussler et al. 1991; de Bruijn and Vreugdenhil 1992; Sharp and LeNoble 2002). Our results showed that ABA content in rice grains was negatively correlated with the cell division rate during the active division period of endosperm cells, though the coefficient was not significant. Spraying ABA at the initial grain filling stage significantly reduced the cell number and grain weight (Table 2). However, increase in ABA content in the grains paralleled the increase in grain dry weight during the linear stage of grain growth. Application of ABA at 11 through 15 DPA significantly increased grain filling rate and grain weight (Table 2). Similar results were also reported for wheat (Dewdney and McWha 1979; Bai et al. 1989), barley (Tietz et al. 1981) and soybean (Ackerson 1985). The results imply that functions of ABA in the grains are determined developmentally. At the initial grain filling stage, ABA in rice grains may not correlate with grain growth or even negatively affect cell division. During the linear growth period, however, ABA levels in the grains do correlate with grain growth. Acknowledgements The work was financed by the National Natural Science Foundation of China (Project No. 30270778). References Ackerson R.C. 1985. Invertase activity and abscisic acid in relation to carbohydrate status in developing soybean reproductive structures. Crop Sci. 25: 615–618. Bai X.F., Cai Y.P. and Nie F. 1989. Relationship between abscisic acid and grain filling of rice and wheat. Plant Physiol. Commun. (China) 3: 40–41. Bangerth F. 1989. Dominance among fruits/sinks and the search for a correlative signal. Physiol. Plant. 76: 608–614. Banowetz G.M., Ammar K. and Chen D.D. 1999. Postanthesis temperatures influence cytokinin accumulation and wheat kernel weight. Plant Cell Environ. 22: 309–316. Barratt D.H.P., Whitford P.N., Cook S.K., Butcher G. and Wang T.L. 1989. An analysis of seed development in Pisum sativum. VIII. Does abscisic acid prevent precocious germination and control storage protein synthesis? J. Exp. Bot. 40: 1009–1014. Ber€ uter J. 1983. Effect of abscisic acid on sorbitol uptake in growing apple fruits. J. Exp. Bot. 34: 737–743.

Bollmark M., Kubat B. and Eliasson L. 1988. Variations in endogenous cytokinin content during adventitious root formation in pea cuttings. J. Plant Physiol. 132: 262–265. Brenner M.L. and Cheikh N. 1995. The role of hormones in photosynthate partitioning and seed filling. In: Davies P.J. (ed.), Plant Hormones, Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 649–670. Browning G. 1980. Endogenous -abscisic acid and pea seed development: evidence for a role in seed growth from changes induced by temperature. J. Exp. Bot. 31: 185–197. Cao X., Zhu Q. and Yang J. 1992. Classification of source-sink types in rice varieties with corresponding cultivated ways. In: Xiong Z. and Min S. (eds), Prospects of Rice Farming for 2000, Zhejiang Sci. & Tech. Press, Hangzhou, pp. 361–372. Davies P.J. 1987. The plant hormones: their nature, occurrence, and functions. In: Davies P.J. (ed.), Plant Hormones and Their Role in Plant Growth and Development, Martinus Nijhoff Publishers, The Netherlands, pp. 1–11. de Bruijn S.M. and Vreugdenhil D. 1992. Abscisic acid and assimilate partitioning to develop seeds. I. Dose abscisic acid influence the growth rate of pea seeds? J. Plant Physiol. 140: 201–206. Dewdney S.J. and McWha J.A. 1979. Abscisic acid and the movement of photosynthetic assimilates towards developing wheat (Triticum aestivum L.) grains. Z. Pflanzenphysiol. 92: 186–193. Dietrich J.T., Kaminek M., Blevins D.G., Reinbott T.M. and Morris R.O. 1995. Changes in cytokinins and cytokinin oxidase activity in developing maize kernels and effects of exogenous cytokinin on kernel development. Plant Physiol. Biochem. 33: 327–336. Eeuwens C.J. and Schwabe W.W. 1975. Seed and pod wall development in Pisum sativum L. in relation to exacted and applied hormones. J. Exp. Bot. 26: 1–14. Hansen H. and Grossmann K. 2000. Auxin-induced enthylene triggers abscisic acid biosynthesis and growth inhibition. Plant Physiol. 124: 1437–1448. He Z. 1993. Method for an indirect enzyme-linked immunosorbent asssay. In: He Z.P. (ed.), Guidance to Experiment on Chemical Control in Crop Plants, Beijing Agricultural University Publishers, Beijing, pp. 60–68. Ho L.C. 1988. Metabolism and compartmentation of imported sugars in sink organs in relation to sink strength. Annu. Rev. Plant Physiol. Plant Mol. Biol. 39: 355–378. Iwasaki Y., Mae T., Makino A., Ohira K. and Ojima K. 1992. Nitrogen accumulation in the inferior spikelet of rice ear during ripening. Soil Sci. Plant Nutr. 38: 517–525. Jones R.J. and Brenner M.L. 1987. Distribution of abscisic acid in maize kernel during grain filling. Plant Physiol. 83: 905–909. Karssen C.M. 1982. The role of endogenous hormones during seed development and the onset of primary dormancy. In: Wareing P.F. (ed.), Plant Growth Substances, Academic Press, London, pp. 623–632. Kato T. 1989. Relationship between grain filling process and sink capacity in rice. Jpn. J. Plant Breed. 39: 431–438. Kato T., Sakurai N. and Kuraishi S. 1993. The changes of endogenous abscisic acid in developing grains of two rice cultivars with different grain size. Jpn. J. Crop Sci. 62: 456–461. Kende H. and Zeevaart J.A.D. 1997. The five ‘classical’ plant hormones. Plant Cell 9: 1197–1210.

195 Liang J., Zhang J. and Cao X. 2001. Grain sink strength may be related to the poor grain filling of indica-japonica rice (Oryza sativa) hybrids. Physiol. Plant. 112: 470–477. Michael G. and Seiler-Kelbitsch H. 1972. Cytokinin content and kernel size of barley grains as affected by environmental and genetic factors. Crop Sci. 12: 162–165. Mohapatra P.K. and Sahu S.K. 1991. Heterogeneity of primary branch development and spikelet survival in rice in relation to assimilates of primary branches. J. Exp. Bot. 42: 871–879. Mohapatra P.K., Patel R. and Sahu S.K. 1993. Time of flowering affects grain quality and spikelet partitioning within the rice panicle. Aust. J. Plant Physiol. 20: 231–242. Morris R.D., Blevins D.G., Dietrich J.T., Durly R.C., Gelvin S.B., Gray J., Hommes N.G., Kaminek M., Mathews L.J., Meilan R., Reinbott T.M. and SagavendraSoto L. 1993. Cytokinins in plant pathogenic bacteria and developing cereal grains. Aust. J. Plant Physiol. 20: 621–637. Murata Y. and Matsushima S. 1975. Rice. In: Evans L.T. (ed.), Crop Physiology, Cambridge University Press, London, pp. 75–99. Murty P.S.S. and Murty K.S. 1982. Spikelet sterility in relation to nitrogen and carbohydrate contents in rice. Indian J. Plant Physiol. 25: 40–48. Ober E.S. and Setter T.L. 1990. Timing of kernel development in water stressed maize: water potential and abscisic acid accumulation. Ann. Bot. 66: 665–672. Pharis R.P. 1985. Gibberellins and reproductive development in seed plants. Annu. Rev. Plant Physiol. 36: 517–568. Qin Z. and Tang X. 1984. Dynamics of some large bio-molecules during the formation of rice endosperm. China Sci. 12: 1103–1110. Richards’ F.J. 1959. A flexible growth function for empirical use. J. Exp. Bot. 10: 290–300. Ross G.S. and McWha J.A. 1990. The distribution of abscisic acid in Pisum sativum plants during seed development. J. Plant Physiol. 136: 137–142. Saha S., Nagar P.K. and Sircar P.K. 1986. Cytokinin concentration gradient in the developing grains and upper leaves of rice (Oryza sativa) during grain filling. Can. J. Bot. 64: 2068–2072. Schussler J.R., Brenner M.L. and Brun W.A. 1991. Relationship of endogenous abscisic acid to sucrose level and seed growth rate of soybeans. Plant Physiol. 96: 1308–1313. Seth A.K. and Waering P.E. 1967. Hormone-directed transport of metabolites and its possible role in plant senescence. J. Exp. Bot. 18: 65–77. Sharp R.E. and LeNoble M.E. 2002. ABA, ethylene and the control of shoot and root growth under water stress. J. Exp. Bot. 53: 33–37. Sikder H.P. and Gupta D.K.D. 1976. Physiology of grain in rice. Indian Agric. 20: 133–141.

Singh B.M. and Jenner C.F. 1982. A modified method for the determination of cell number in wheat endosperm. Plant Sci. Lett. 26: 273–278. Singh G. and Gerung S.B. 1982. Hormonal role in the problem of sterility in Oryza sativa. Plant Physiol. Biochem. 9: 22–23. Tietz A., Ludwig M., Dingkuhn M. and Dorffling K. 1981. Effect of abscisic acid on the transport of assimilates in barley. Planta 152: 557–561. Umemoto T., Nakamura Y. and Ishikura N. 1994. Effect of grain location on the panicle on activities involved in starch synthesis in rice endosperm. Phytochemistry 36: 843–847. Venkateswarlu B. and Visperas R.M. 1987. Source-sink relationships in crop plants. Int. Rice Res. Paper Ser. 125: 1–19. Wang T.L., Cook S.K., Francis R.J., Ambrose M.J. and Hedley C.L. 1987. An analysis of seed development in Pisum sativum. VI. Abscisic acid accumulation. J. Exp. Bot. 38: 1921–1932. Wang Y. 1981. Effectiveness of supplied nitrogen at the primordial panicle stage on rice characteristics and yields. Int. Rice Res. News Lett. 6: 23–24. Wang Z., Yang J., Zhu Q., Zhang Z., Lang Y. and Wang X. 1998. Reasons for poor grain filling in intersubspecific hybrid rice. Acta Agron. Sin. 24: 782–787. Warren W.J. 1972. Control of crop processes. In: Rees A.R., Cockshull K.E., Hand D.W. and Hurd R.G. (eds), Crop Processes in Controlled Environments, Academic, London/ New York, pp. 7–30. Wu S., Chen W. and Zhou X. 1988. Enzyme linked immunosorbent assay for endogenous plant hormones. Plant Physiol. Commun. (China) 5: 53–57. Yang J., Su B., Wang Z. and Zhu Q. 1999. Characteristics and physiology of grain filling in intersubspecific hybrid rice. Chin. Agric. Sci. 1: 61–70. Yang J., Zhang J., Wang Z., Zhu Q. and Wang W. 2001. Hormonal Changes in the grains of rice subjected to water stress during grain filling. Plant Physiol. 127: 315–323. Yoshida S., Forno D., Cock J. and Gomez K. 1976. Determination of sugar and starch in plant tissue. In: Yoshida S. (ed.), Laboratory Manual for Physiological Studies of Rice, International Rice Research Institute, The Philippines, pp. 46–49. Yuan L.P. 1997. Hybrid rice breeding for super high yield. Hybrid Rice 12(6): 1–6. Zhang J., He Z. and Wu Y. 1991. Establishment of an indirect enzyme-linked immunosorbent asssay for zeatin and zeatin riboside. J. Beijing Agric. Univ. (China) 17: 145–151. Zhu Q., Cao X. and Luo Y. 1988. Growth analysis in the process of grain filling in rice (in Chinese with English abstract). Acta Agron. Sin. 14: 182–192.