Two winter wheat cultivars, Piko and Slawa, were used in our study. Both cultivars were registered in The Polish National List, Piko in 2002, Slawa in 2001. Piko was derived from the cross (Semidwarf × Kraka) × (Kranich × Caribo) (Nordsaat Saatzucht GmbH, Germany), whilst Slawa was obtained from the cross Lanca × Juwell 15 (Poznanska Hodowla Roslin, PHR Sp.z o.o, Poland). Both cultivars develop red grains which, when reacting with phenol, change color to brown. Both cultivars were used for baking.

These cultivars were chosen on the basis of our previous (2005-2007) field and laboratory experiments in which their pre-harvest sprouting capabilities were evaluated. According to our data, Piko grains exhibited higher susceptibility to PHS than Slawa grains. For our purposes both cultivars were cultivated at the Experimental Station of the Department of Plant Breeding and Seed Science, located in Prusy near Krakow (N 50°07’03” and E 20°05’13”), and the trials were prepared each year by sowing the seeds collected from isolated ears. For this study we used the plant material from the 2009 field experiment. The average temperatures in May, June, and July were 13.1, 17.6 and 20.8°C, respectively, the rainfall was 302.5, 135.1 and 105.2 mm, respectively (the average weather conditions in the years 2005-2015, recorded at the experimental station in May, June and July were as follows: temperatures - 13.9, 17.5 and 19.9°C, respectively; rainfalls - 92.5, 95.1 and 96.7 mm, respectively). Forty-five grams of seeds of each cultivar were sown manually on three plots. Each plot consisted of 10 rows 1.5 m long - the rows were 25 cm apart. The plants were evaluated in respect of the date when the pollen was released from the florets in the middle part of the spike. We harvested ten spikes from the middle part of each plot, 10, 14, 17, 24, 31, 38 and 45 days after pollination and the grains from middle parts of the spikes were used for preparing the biological samples for germination tests and expression analyses. The grains from each plot were used as one biological replication. The germination tests (see below) were prepared immediately after harvesting the grains, while for RNA isolation the grains were stored at -80°C. At the last developmental stage (45 DAP, grain humidity was 113 g/kg and 107 g/kg for Piko and Slawa, respectively) also spikes without any signs of external sprouting were collected and used for the induction of sprouting (see below). The grains harvested at that time were also stored (for 5 weeks) at room temperature before using them in an ABA responsiveness assay. The picked spikes and grains from each plot were used as individual biological replicates.

Grains harvested at different developmental stages (from 24 to 45 DAP), as well as the physiologically mature grains stored for 5 weeks at room temperature after harvest, were germinated. Grains were placed in 14-cm Petri dishes with three layers of filter paper and 10 mL of distilled water. Each dish had 25 grains of each cultivar. Incubation was performed at 20°C in the darkness. Grains after 5 weeks of storage were also incubated in 10, 50, 100 and 150 µmol/L of ABA (Sigma-Aldrich) solutions, at 20 and 25°C. Three replications for each cultivar were performed. Each replication consisted of the grains collected from one field plot. Germinating grains were counted daily and removed. Germination was considered when the grain coat over the embryo was at least broken or ruptured, resulting in partial exposure of the embryo. Germination capability was assessed after a 7-d induction using the germination index (GI) value ( Walker-Simmons, 1987 ), which is calculated as follows: GI = [(7×n1) + (6×n2) + (5×n3) + (4×n4) + (3×n5) + (2×n6) + (1×n7)]/(total days × total number of grains), where 7, 6,....1 are weights given to the grains germinated on the first (n1), second (n2), and subsequent days until the seventh day (n7), respectively. The maximum value of GI is 1.0.

Spikes collected at grain physiological maturity (45 DAP) without visible sprouting damage were placed in a moisture chamber at 20°C ± 1 for maximum 6 days. Three replications for each cultivar, with each sample consisting of 10 spikes, were performed. Each replication consisted of the spikes collected from one field plot. The spikes were placed in a vertical position on a special device (a frame with a stretched 0.5 cm mesh net). The spacing between the spikes was 2 cm in a row and 3 cm between rows. High humidity was maintained during the experiment: the spikes were sprayed using a fine nozzle (1 mm), with 2 L of water four times a day at: 7.00-8.00 a.m., 11.00 a.m.-12.00 noon, 3.00-4.00 p.m., and 6.00-7.00 p.m. After the last spraying, the spikes were covered with a plastic foil for overnight. Estimation of grain sprouting was conducted on the third day of incubation for five randomly chosen spikes from each replication when first germination signs were visible. The percentage of sprouting grains was determined. Among the sprouting grains, three stages of germination were distinguished: 1 - broken or ruptured grain coat over the embryo resulting in a partial embryo exposure, 2 - grains with the emerging radicle up to 2 mm in length, and 3 - grains with the coleoptile and radicle longer than 2 mm. The remaining spikes were left in the same conditions until the sixth day. For molecular analysis, sprouting grains (stage 1) collected on the third day and non-sprouting ones collected on the sixth day of incubation were used. For RNA isolation, each sample was performed in three biological replicates and stored at -80°C.

For total RNA isolation, 1 g of grain samples (corresponding to 12-20 grains depending to the developmental stage of the grains) were homogenized in liquid nitrogen using a mortar and pestle, and 100 mg of the resulting powder was transferred to 2 mL tubes. The RNA pellet, obtained according to the protocol described by Singh et al. (2003) was resuspended in 100 µL of sterile Milli-Q water and incubated for 4 h on ice. After incubation, the mixture was centrifuged (15000 rpm, 10 min, 4°C) (Biofuge Stratos, Kendro Laboratory Products, Hanau, Germany). The supernatant was transferred into new Eppendorf tubes, and RNA precipitation using LiCl was performed as described by Simlat et al. (2013) . The obtained RNA pellet was resuspended in 20 µL of sterile Milli-Q water and stored at -80°C until further use. The quantity of RNA was determined by measuring absorbance at 260 nm and by standard agarose electrophoresis.

To synthesize the first strand of cDNA, we used 1 µg of RNA. Prior to reverse transcription, RNA was treated with DNase I without RNase activation (Fermentas Int Inc) (2 U DNase I / 1 µg RNA) and incubated for 30 min at 37°C. Reverse transcription was performed using the First Strand cDNA Synthesis Kit (Fermentas Int Inc) with random hexamer primers in 20 µL reactions according to the manufacturer’s instructions. Quantitative PCR was performed using the 7500 Fast Real Time PCR System (Applied Biosystems, USA) with SYBR Green to monitor double-stranded DNA synthesis. Three technical replicates of each cDNA sample were performed for qPCR analysis. For amplification, cDNA was diluted 10-fold, and 1 µL was used in 20 µL qPCR reaction mixture with 2X Fast SYBR Master Mix (Applied Biosystems, USA) and 0.5 µmol/L of each primer. The following thermal profile was used for all reactions: 95°C for 20 s, 40 cycles of 95°C for 3 s, and 60°C for 1 min. Melting curve analysis was performed to evaluate the presence of non-specific PCR products and primer dimers. Relative gene expression was calculated in different samples using the Ct method. Actin gene was used as an internal control. The relative mRNA level was normalized for each sample to that of 10 DAP for the respective cultivar. For all genes, standard curves were generated using a dilution series of cDNA. The obtained data were analyzed using 7500 Software v2.0 (Applied Biosystems, USA). The primers used for detection of transcript accumulation were designed on the basis of the sequence records: AK332105 (AAO3), AB439590 (CPS3), AB047554 (VP1), and AY663392 (actin). The VP1 cDNA was amplified with primers anchored in the region coding for the B3-domain and therefore both the correctly and alternatively spliced mRNAs were targeted in the assay. The primers were designed and tested for possibility of homo- and hetero-dimers formation using PrimerQuest and Oligo Analyzer programs available online ( http://eu.idtdna.com/Home/Home.aspx ). The gene specific primers were as follows: AAO3-fwd: GGC GTT GTG ATT GCT GAA AC, AAO3-rev: TAG CTG TCA TGT TGG ATG GC, CPS3-fwd: GCG GTC TCC ATG AAG GAA AT, CPS3-rev: TCC ACC AAC AGG TAG TGT CT, VP1-fwd: CAC TGG TGA CTT TGT TCG GT, VP1-rev: GTG CTT GGC TAG ATC CTG TT, actin-fwd: AGG TGC CCT GAG GTC CTC TT, and actin-rev: GTA GGT CGT CTC GTG GAT TCC A.

Differences among mean values of germination index, percentage of sprouted grains, and relative quantity of gene expression were analyzed using a factorial-design analysis of variance (ANOVA). In these analyses, the least significant difference (LSD) was calculated using the Student’s t-test at 0.05 and 0.01 probability. The statistical analyses of data were performed using Statistica, v 10 ( www.statsoft.com ).

The grains of both cultivars harvested at 24, 31, 38, and 45 DAP were induced to germinate in Petri dishes. Each day, the grains with visible signs of external germination were removed. For both cultivars, germination capabilities increased progressively during grain development. For Piko, the values of the germination index ranged from 0.05 at 24 DAP to 0.82 at 45 DAP. For Slawa, the GI values ranged from 0.04 to 0.6, but they did not differ significantly at 24 and 31 DAP. In all analyzed developmental stages, except 24 DAP, the germination index of Slawa was significantly lower than that of Piko ( Fig. 1 ).

With respect to sprouting in spikes harvested at 45 DAP, both cultivars showed different tendencies. The grains of Piko sprouted faster and better than those of Slawa. The mean percentage of sprouted grains observed for Piko reached 50.4%, whereas in Slawa it was 14.9% ( Table 1 ). The cultivars also differed in the progress of grain sprouting. In Piko, more than 50% of the germinated grains were those with well-developed radicles and coleoptiles (stage 3), whilst in Slawa the most numerous were grains with the radicles no longer than 2 mm (stage 2).

Table 2 shows that the cultivars differed in the GI, also when germination was tested in the presence of ABA (the experiment was performed for the physiologically mature grains stored at room temperature for 5 weeks after harvest). Exogenous ABA inhibited germination of grains in both cultivars. However, the inhibitory effect of ABA was more pronounced for PHS-tolerant Slawa than for PHS-susceptible Piko. A greater inhibitory effect was observed when grains were germinated at 25°C, and only for Slawa the concentration of exogenous ABA had significant effect on grain germination.

Figure 2 shows that relative quantitation (RQ) of the AAO3 mRNA in Piko and Slawa grains harvested at successive stages of development clearly indicates greater accumulation of this transcript during last three stages, i.e. from 31 to 45 DAP. In both cultivars, the greatest accumulation of this transcript was observed at 31 DAP. Afterward, the AAO3 mRNA level decreased, but in mature grains (45 DAP) of Slawa, this transcript accumulation was significantly higher than that of Piko. An increase in accumulation was also detected in Piko at 17 DAP and in Slawa at 20 DAP, but observed differences were not significant. In both cultivars, the lowest accumulation of mRNA was observed at the first analyzed developmental stage (10 DAP) ( Fig. 2a ).

The CPS3 transcript accumulation showed two distinct peaks during grain development in both cultivars. In Piko, the highest levels of CPS3 mRNA were observed at 14 and 20 DAP. However, in Slawa the first peak was observed at 17 and the second at 31 DAP. Relatively low expression values (relative to 10 DAP) were found for Piko grains at 17, 31, 38, and 45 DAP and for Slawa grains at 20, 24, 38, and 45 DAP. However, for both cultivars at each developmental stage the level of expression was higher than that at 10 DAP. It is also worth mentioning that in Piko, the expression level was higher at 20 DAP than at 17 DAP, while in Slawa at those stages an opposite relationship was observed. Moreover, at 24 and 31 days after pollination, the tested cultivars showed differences in the accumulation of the examined transcript. In Piko, at 24 and 31 DAP, the relative accumulation decreased from 7.8 to 2.8, while in Slawa it increased from 3.0 to 6.5. At the last two analyzed stages of grain development, transcript accumulation showed similar values in both cultivars ( Fig. 2b ).

The expression analysis of the VP1 gene showed two peaks in both examined cultivars. In Piko (the PHS-susceptible cultivar), the relative VP1 gene expression was low at the beginning and during the middle stages of grain development, and the first peak was observed at 20 DAP. After 24 DAP, the expression level significantly increased and reached the second peak at 45 DAP. In Slawa (the PHS-tolerant cultivar), the relative VP1 gene expression was also low at the beginning of grain development, but it significantly increased during the middle stages. The first peak was observed at 20 days after pollination. Transcript accumulation then dropped down at 24 DAP and peaked again at 38 DAP. The relative accumulation of the VP1 mRNA remained high in mature grains of both cultivars ( Fig. 2c ).

For this analysis, sprouting grains collected on the third day and non-sprouting ones collected on the sixth day of sprouting induction were used ( Table 3 ). The relative amount of the AAO3 transcript in sprouting grains was similar in both examined cultivars. In non-sprouting grains, a higher expression level was observed in Piko than in Slawa but observed difference was not significant. It is also worth mentioning that in non-sprouting grains of both cultivars, the expression level was higher than that in sprouting grains. This difference for Piko was significant.

The relative CPS3 expression exhibited differences between the two cultivars. In both sprouting and non-sprouting Piko grains, the CPS3 expression was significantly higher than for both samples of Slawa. On the other hand, in both samples of Slawa, a similar relative quantity value of the examined gene was observed.

The relative VP1 expression showed an accumulation profile similar to that for the AAO3 gene for both cultivars: in grains without visible signs of external sprouting, the amount of transcript was greater than that in sprouting grains. The cultivars did not differ significantly in the levels of the VP1 transcript in sprouting grains. However, in non-sprouting grains of Piko, the level of this mRNA was about 2.5 times higher than that observed in Slawa.