Introduction

Pears (Pyrus L.) are one of the leading cultivated fruit trees of temperate regions in China following apples and grapes in planting area and fruit yield. As the top country in pear industry, China produces more than 60% of the world pear production (Wu et al. 2013), but the low sugar content in fresh pear directly influences the flavor of fruit, leading to great economic losses in postharvest storage periods (Hudina and Śtampar 2000). In pears and other Rosaceae plants, sorbitol is the main sugar for transportation from leaves to fruit, while sucrose is the minor one. They are basic material for fruit quality components and flavor substances such as vitamins, pigments, volatiles, and have important physiological functions, such as the driving force in providing penetration for fruit cell enlargement (Kanayama 2009). In pears, sorbitol is synthesized in the leaves by sorbitol 6-phosphate dehydrogenase (S6PDH, EC 1.1.1.200), and unloads from sieve element-companion cell complexes into the cell wall space in fruit (Yamaki 1986). It is taken up into the cytosol of parenchyma cells by sorbitol transporter (SOT), and then converted to fructose by sorbitol dehydrogenase (SDH, EC 1.1.1.14) (Zhang et al. 2004). Sucrose is directly transported into parenchyma cells by sucrose transporters (SUT) located on fruit, and is converted to glucose and fructose by acid invertase (AIV, EC 3.2.1.26), or converted to fructose and UDP-glucose (UDPG) by sucrose synthase (SUS, EC 2.4.1.13). Glucose and fructose are then phosphorylated to glucose 6-phosphate (G6P) and fructose 6-phosphate (F6P). Both F6P and UDPG can be combined to re-synthesize sucrose-by-sucrose phosphate synthase (SPS, EC 2.4.1.14) (Li et al. 2012). Among sucrose metabolizing enzymes, invertase, including AIV and NI, as well as SUS-cleavage, are responsible for the accumulation of reducing sugars (Sturm and Tang 1999). Additionally, SUS-synthesis and SPS are involved in catalyzing sucrose biosynthesis (Guo et al. 2002). AIV plays an important role in functions related to sucrose metabolism and presumably hydrolyzes sucrose to supply reducing sugars necessary for cell growth and development (Tang et al. 1999). AIV activity is high in young fruit and decreases in Japanese pear during fruit maturation (Tanase and Yamaki 2000). Sucrose deposited in the vacuole is cleaved by vacuolar invertase. Yamada et al. (2007) isolated cDNA clones for AIV from Japanese pear fruit and found that during fruit maturation, AIV seems to have important roles in supplying hexoses at the young stage and regulating the sucrose-to-hexoses ratio in the vacuole at mature stage. The concentration and distribution of sugars in the leaves and fruit are modulated by sucrose metabolism and sorbitol metabolism cycle system. This system not only ensures the carbon portioning between the leaves (source) and fruit (sink), but also adjusts the balance of sugar metabolism, and maintains cell osmotic potential and turgor pressure between the leaves and fruit. A number of studies have proved that the sorbitol in apple phloem accounts for about 60–70% of total sugars, up to 80% in some varieties; the sucrose accounts for only 10–24% (Loescher 1987; Yamaki 1986). While the proportion of sugar types in pear fruit vary with pear cultivars. Some pear fruit mainly accumulate fructose, while the others may mainly accumulate sucrose or sorbitol. Sorbitol of peach and pear from phloem transport accounts for about 70–80% (Nadwodnik and Lohaus 2008), and sucrose is accumulated up to 10–53% of total sugar in mature fruit flesh (Moriguchi et al. 1992; Zhang et al. 2012).

Potassium (K), as one of the most important and abundant cations in living plant cells, plays an important role in many fundamental processes including membrane potential, anion neutralization, osmoregulation, enzyme activation, and signal transmission (Wang and Wu 2013). It also plays a key role in pear leaf photosynthesis, flower bud quality, fruit size, color, and flavor (Basile et al. 2003). K fertilizer application is required to sustain an adequate supply of soluble K to crops (Fernández et al. 2009). Balanced fertilization can improve the yield and quality of the fruit as well as the phloem translocation of photoassimilates from leaves to fruit. Philippar et al. (2003) provide evidence that a balance between the sugar and K concentration can maintain osmotic pressure in the phloem sap and drive photosynthate transportation in a sieve tube. K is crucial for most of the steps of the protein synthesis process beginning with enzyme activation and continuing through ribosome synthesis, and mRNA turnover (Pettigrew 2008). Although the changes of enzyme activity or gene expression in the process of carbohydrate accumulation in fruit have been reported in some previous research (Moriguchi et al. 1992; Itai and Tanahashi 2008; Li et al. 2015; Dai et al. 2015), the expression levels of key genes involved in sugar metabolism in the leaves and fruit in response to different K levels remain almost unknown.

K is well known as the most important quality element. Many studies have showed that K can increase yield and promote fruit quality in pear (Brunetto et al. 2015), apple (Nava and Dechen 2009) and citrus (Quaggio et al. 2006). The flow of photosynthetic products (soluble sugars) from the leaves into the fruit can be promoted to provide a certain osmotic potential and accumulate more soluble sugars in fruit by K supply (Teo et al. 2006). However, little research has been reported on the mechanism of how to improve fruit sugar accumulation by K. Farmers tend to apply excessive or insufficient K fertilizers in pear orchards in China, which usually resulted in the deterioration of fruit quality and the aggravation of physiological disease (Brunetto et al. 2015). In this study, a main pear cultivar ‘Huangguan’ (Pyrus bretschneideri Rehd.) cultivated in Hebei, Shandong, Jiangsu and Gansu provinces in China was used to evaluate the effect of different K levels on leaf photosynthetic characteristics and the accumulation process of primary metabolites in leaves and fruit; to study the expression patterns of multiple members of gene families encoding key enzymes, or transporters involved in sugar metabolism in different organs, and to provide a scientific basis for a better understanding of the effect of K fertilizer on yield and fruit quality of pear trees.

Materials and methods

Pear tree management

The field study was carried out with 16-year-old Asian pear (P. bretschneideri Rehd. cv. ‘Huangguan’) trees planted at 4 × 5 m2 (the root stock was Pyrus betulaefolia) during the 2014 and 2015 growing seasons at the town of Fangcun in the city of Xuzhou, Jiangsu Province, China (34°05′N and 117°29′E). Fruit thinning every year was performed manually to keep a constant yield of 30,000 kg ha−1. The soil was classified as a sandy soil with the following chemical characteristics in 2012: pH 8.38, 44.91 mg kg−1 available nitrogen (N), 19.11 mg kg−1 available phosphorus (P), and 65 mg kg−1 available K.

K treatments

Orchard management followed normal commercial practices. Based on the same N (525 kg ha−1) and P2O5 (225 kg ha−1) levels, the experiment was set up with four K2O levels: 0 (control), 150 (K150), 300 (K300), 450 (K450) kg ha−1. Each K treatment was replicated six times in a completely randomized design from 2014 to 2015. The trees received fertilizer twice each year in the mid-March and mid-June. Each tree around the drip line in the orchard was supplied with urea (46% N) (bud sprouting period accounted for 60%, enlargement stage I accounted for 40%), calcium superphosphate (12% P2O5) at the bud sprouting period, and potassium sulfate (51% K2O) (bud sprouting period accounted for 40%, enlargement stage I accounted for 60%).

Samples collection

A total of five developmental stages were collected from young fruit stage, which were 30 days after full blooming (30DAB, 10/5). The leaves and fruit samples were selected from the enlargement phase I (73DAB, 22/6), enlargement phase II (101DAB, 20/7), maturity (123DAB, 11/8), and 1 month after harvest (154DAB, 11/9) in 2014. This was repeated at young fruit stage (31DAB, 11/5), enlargement phase I (67DAB, 16/6), enlargement phase II (97DAB, 16/7), and maturity (128DAB, 16/8) in 2015. The sampled fruit and leaves were weighed and immediately frozen in liquid N2, then stored at −80 °C for the analysis of soluble sugars and gene expression.

Measurement of leaf photosynthetic parameters

The SPAD value was measured using a SPAD-502 chlorophyll meter (SPAD-502, Minolta Camera Co., Tokyo, Japan) at the enlargement phase II in the summer of 2014–2015. The net photosynthetic rate (Pn) was measured using the portable photosynthesis system Li-6400 (LI-COR Biosciences, Lincoln, NE, USA) under a photon flux density of 1000 μmol m−2 s−1. Ten leaves were selected as a replicate for a total of three replicates per treatment.

Chemical analysis of soil and measurements of total K concentration in leaves and fruit

The pH was determined using 1:2.5 (w/v) soils:water extracts. Alkalihydrolysable N was determined using the method used by Stehman et al. (1999). Available phosphorus was measured using extracts of hydrochloric acid–ammonium fluoride (HCl–NH4F). Available potassium was determined by extracting the soil with 1 mol L−1 ammonium acetate (CH3COONH4), and then measured by flame photometry. The fruit skin was removed and the fruit flesh was homogenized. The collected leaves and fruit flesh were denatured at 105 °C, and dried at 70 °C to a uniform moisture level in an oven before grinding in a stainless steel mill. Approximately 0.1 g of dried ground leaves and fruit samples were digested with 5 mL 98% H2SO4 and 1 mL 30% hydrogen peroxide in a digestion furnace (at 280 °C, 2 kW, 1 h) before being diluted to 100 mL with distilled water. The K concentration of the dried leaves and fruit samples was determined using an AP1200 Flame Photometer (Aopu, Shanghai, China). We corrected the ‘Huangguan’ pear leaf to fruit ratio as 37:1, by which we can calculate the ‘Huangguan’ pear fruit biomass at the base of leaf biomass. The total K accumulation in leaves and fruit were calculated according to the following equation:

\({\text{Leaf}}\;\text{K}\;\text{accumulation}\;(\text{kg}\;{\text{plant}}{^{-1}})=\text{leaf}\;\text{biomass}\;(\text{kg}\;{\text{plant}}{^{ -1}}) \times \text{leaf}\;\text{K}\;\text{concentration}\,\left( \% \right) \times {10^{ - \,2}}.\)

\({\text{Fruit}}\;\text{K}\;\text{accumulation}\;(\text{kg}\;\text{plant}{^{-1}})=\text{fruit}\;\text{biomass}\;(\text{kg}\;\text{plant}{^{-1}}) \times \text{fruit}\;\text{K}\;\text{concentration}\;\left( \% \right) \times {10^{ - \,2}}.\)

Measurements of fruit quality and sugar content in leaves and fruit

In 2014–2015 at the maturity stage three replicates of fresh fruit samples with at least 15 fruit in each tree were harvested and immediately weighed. Fruit firmness was assessed with a penetrometer (FT327; Effegi, Alfonsine, Italy) with a probe size of 11.3 mm. The soluble solids content (SSC) was determined using a refractometer (PAL-1; ATAGO, Tokyo, Japan). Total soluble sugar was measured using the anthrone–sulphuric acid colorimetric method (Alexander and Edwards 2003). Titratable acidity (TA) was determined using standard acid-base titration (Sánchez et al. 2015).

High performance liquid chromatography (HPLC) was used to analyze the carbohydrates present in fresh leaves and fruit (Melgarejo et al. 2000). Pear fresh leaves and fruit samples (2 g) were homogenized to a purée and diluted to 20 mL with 80% (v/v) ethanol, and then centrifuged at 20000×g for 15 min at 4 °C. The supernatant was recovered and immediately filtered through a 0.45 mm SepPak filter (Waters, Milford, MA) to eliminate large particles. The extraction was stored at −80 °C in a sealed tube for high performance liquid chromatography (HPLC) (Agilent 1200; Agilent, Santa Clara, CA). The individual sugar concentration was determined using the following specific conditions as described by Colaric et al. (2007) with some modifications: CAPCELL PAK NH2 column (4.6 × 250 mm, 5 μm) (Shiseido, Tokyo, Japan), constant temperature of 50 °C, and mobile phase of acetonitrile–water (80/20; v/v). An evaporative light scattering detector (ELSD) detector (Alltech 3300 ELSD; Grace, Deerfield, IL) was used with a flow rate of 1.0 mL min−1, drift tube temperature of 80 °C, and nitrogen flow rate of 2.0 mL min−1. Standard curves for sorbitol, fructose, glucose, and sucrose (Sigma, St Louis, MO) were generated as references to quantify the sugar content in the samples. Total sugar = fructose + sorbitol + glucose + sucrose.

RNA isolation and cDNA synthesis

Total RNA was extracted from fresh leaves and fruit tissues according to the CTAB (cetyltrimethyl ammonium bromide) method (Gasic et al. 2004), and DNase I (Invitrogen) was used to remove genomic DNA contamination. An aliquot of RNA was quantified by light absorbance at 260 nm, and then electrophoretically separated on a 1.0% agarose gel to check the integrity.

Single-stranded cDNA was synthesized from 2.0 mg of total RNA using a Revert AidTM First Strand cDNA Synthesis Kit (Fermentas, Ontario, Canada). All cDNA samples were diluted 1:10 with RNase-free water, and stored at −20 °C before being used as templates in cloning and qRT-PCR.

Real-time PCR analysis of sugar metabolism related genes in leaves and fruit

AIV1 (AB334115.1), AIV2 (AB334116.1), SPS1 (AB334114.1), SUS1 (AB045710.1), S6PDH (KC506733.1), NAD-SDH1 (KC506730.1), NAD-SDH2 (KC506731.1), NAD-SDH3 (KC506732.1) (Liu et al. 2013), SUT1 (KC801340), SOT2 (AB719045), and SOT3 (AB719046) (Ito et al. 2012) were searched for expression analysis in the National Center for Biotechnology Information (NCBI). The cDNA samples were amplified with gene-specific primers using Primer Premier 5.0 software (Supplemental Table 1). Each treatment was made in the six biological samples. Triplicate quantitative assays for each biological sample were performed using the Step One Plus Real-Time PCR Systems (Applied Biosystems, Bio-Rad, CA). The melting-curve analysis following the final cycle of the qRT-PCR and 2% agarose gel electrophoresis were used to check the specificity of the PCR amplification. The total reaction volume for each qRT-PCR was 20 µL, which comprised 10 µL SYBR Green PCR SuperMix (Takara, Tokyo, Japan), 0.4 µL of each primer, 2 µL of 1:10 diluted cDNA, 0.4 µL Passive reference, and 6.8 µL double-distilled water. The PCR reaction conditions were as follows: 94 °C for 4 min, and 40 cycles of 94 °C for 5 s. Followed by 60 °C for 30 s. Actin (JN684184), which was chosen as a reference gene (Yu et al. 2012). The relative gene expression levels were calculated using the 2−ΔΔCT method (Livak and Schmittgen 2001).

Statistical analysis

Data was analyzed by analysis of variance (ANOVA) using SPSS 18.0 (SPSS Inc., Chicago, IL, USA). The drawing was completed in Origin8.5 (Origin Inc., Chicago, USA), and Canoco 4.5 was used to the redundancy analysis (RDA) (Biometris, Wageningen, The Netherlands). All measured values are presented as the mean ± standard error. Means were compared for treatment effects using a Tukey’s protected least significant difference (LSD) at P < 0.05.

Results

Effect of K application rates on K concentration and accumulation in leaves and fruit

Leaf K concentration was declined then increased, and was lowest at the mature stage during all the stages in 2014–2015 (Supplemental Table 2). Leaf and fruit K concentration and accumulation increased with K application rates at each stage. Compared to the control, leaf K concentration significantly increased with the K450 treatment at different stages. Leaf K concentration again raised with the increase of K application rates at the 1 month after harvest stage, which under the K450 treatment was 23.46 and 12.19% higher compared to the control in 2014–2015. The change of leaf K accumulation in the 2 years was basically the same, and K supply increased the K accumulation of the leaves especially in 2015. Leaf K accumulation between the 2014 and 2015 years were both the highest at enlargement phase II, and leaf K accumulation of K450 treatment was 16.19 and 50.50% significantly higher than that of the control. Leaf K accumulation at maturity stage was 4.52 and 14.05% less than enlargement phase II, respectively (Supplemental Table 3).

From young fruit stage to maturity fruit K concentration significantly declined. With the increase of K application rates, there was no change significantly in 2015 for the fruit K concentration at the enlargement phases, but over the 2 years the fruit K accumulation increased gradually at the mature stage. At maturity fruit K concentration of all K treatments significantly increased by 11.05 and 8.13% compared to the control in 2014–2015 (Supplemental Table 2). The fruit K accumulation increased gradually with the fruit development stages, and rapidly increased from the enlargement phase II, and then reached a peak at maturity. In 2014–2015, K absorption at the enlargement phase II accounted for 55.73 and 52.23% of the total uptake at maturity, respectively. Fruit K accumulation with K450 treatment significantly increased by 29.95 and 44.06%, respectively compared to the control (Supplemental Table 3). During the fruit enlargement stage in 2014–2015, fruit K accumulation accounted for 83.73 and 52.23%, respectively in the total K accumulation at mature stage, and leaf K accumulation at the enlargement phase II decreased by 4.52 and 14.05% compared to that at the enlargement phase I.

Effect of K application rates on SPAD and photosynthetic characteristics in leaves

In the enlargement phase II, it was shown (Supplemental Fig. 1) that determination of pear leaf net photosynthetic rate (Pn), and in the relative content of chlorophyll (SPAD). We found that the net photosynthetic rates of the leaves between the two years were basically the same, but the SPAD value in 2015 was higher than that in 2014. Compared to the control, the Pn and SPAD of the K450 treatment in 2 years had an average of 8.44 and 2.71% higher, respectively.

Effect of K application rates on the single fruit weight

As shown in Supplemental Fig. 2 the single fruit weight in 2014–2015 from young fruit stage to the enlargement phase I grew slowly, but increased sharply from the enlargement phase II to maturity. The daily growth rate of fruit weight at the expansion phase II reached the peak of 5.64 and 2.92 g day−1 among the all stages in 2014–2015, respectively. Fruit weight increased with the increase of K application rates especially in 2015. In 2015, the single fruit weight at different stages was lower than that in 2014 except for the young fruit stage in 2014. The single fruit weight with the K450 treatment at each stage of 2 years significantly increased compared to the control, and the single fruit weight of K450 treatment at the mature stage were 12.03 and 11.53% higher than that of the control, respectively.

Effect of K application rates on fruit yield and quality index

In 2014 the fruit yield was lower than that in 2015. In the two seasons an increase was observed in pear fruit yield subjected to the application of increasing rates of K fertilizer. The yield of K300 treatment had the highest yield; K450 treatment slightly decreased, but compared to the control, the average fruit yield under the K450 and K300 treatments in 2014–2015 increased by 19.21 and 20.88%, respectively (Fig. 1a).

Fig. 1
figure 1

The effect of K application rates on fruit quality index: a yield, b firmness, c soluble solids, d soluble sugars, e titratable acidity of Asian pear in 2014–2015. The experiment was set up with four K2O levels: 0 (control), 150 (K150), 300 (K300) and 450 (K450) kg ha−1. Different letters indicate significant differences between means (Turkey’s test, P < 0.05)

The effect of K application rates on fruit quality was not the same in two seasons (Fig. 1b–e). We found that there was no significance on the fruit hardness with the K application rates, but the soluble solids content (SSC), and soluble sugar remarkably increased. TA under the K300 treatment was the highest, and under the K450 treatment was an average of 8.11 and 11.39% lower than under the K300 treatment in 2014–2015, respectively.

Effect of K application rates on the concentration of soluble sugars in leaves and fruit

Soluble sugars concentrations of the leaves at the young fruit stage were the highest (Fig. 2). They decreased firstly, and then remained unchanged while the concentration of soluble sugars in fruit increased with the fruit development. The sorbitol concentration was a major component of the sugars in leaves where the fructose concentration in fruit was dominant. The concentration of sorbitol, glucose, and sucrose in leaves significantly increased by K treatments in the enlargement phase II, but the concentration of the K300 treatment in the leaves was significantly higher than those of the K450 treatment at maturity. The fruit sugar component concentration; except for fructose concentration, during the fruit development stages significantly improved under the K treatments compared to the control. At maturity, when compared to the control, the concentration of sorbitol, sucrose, and total sugar significantly increased by 11.52, 38.63 and 12.24%, respectively in the K450 treatment. Except for mature stage the concentration of total sugar of each stage in the leaves also significantly increased by high K treatment.

Fig. 2
figure 2

The effect of K application rates on concentration of soluble sugars: a, f fructose, b, g sorbitol, c, h glucose, d, i sucrose and e, j total sugar in leaves (left row) and fruit (right row) in 2015. The experiment was set up with four K2O levels: 0 (control), 150 (K150), 300 (K300) and 450 (K450) kg ha−1. Total sugar = fructose + sorbitol + glucose + sucrose. FW fresh weight. Different letters indicate significant differences between means (Tukey’s test, P < 0.05)

Effect of K application rates on expression levels of key genes involved in sugar metabolism

Expression levels of key genes involved to sugar metabolism in leaves

Metabolite levels in leaves and relative expression of key genes involved in sugar metabolism were presented as color-coded heat maps in Fig. 3. Results showed that sorbitol, glucose, sucrose, and total sugar accumulation increased with the increase of K application rates, while there was very little difference in fructose accumulation among K treatments at young fruit stage (11/5). AIV was responsible for sucrose hydrolysis to fructose and glucose in melon fruits. SPS specifically utilized K as a cofactor to synthesize sucrose from glucose and fructose (Lester et al. 2001). The relative expression of genes (AIV1, AIV2 and SPS1) involved in sucrose metabolism and gene (S6PDH) involved in sorbitol synthesis increased with the increase of K application rates (Fig. 3c), while sucrose transporter (SUT) involved in sucrose metabolism displayed the opposite pattern. The relative expression of the other with sorbitol degradation and transport related genes (SDH1, SDH2, SDH3, SOT2 and SOT3) decreased with the increase of K application rates.

Fig. 3
figure 3

The effect of K application rates on the sugar accumulation and the expression levels of sugar metabolism key genes in leaves in 2015. The experiment was set up with four K2O levels: 0 (control), 150 (K150), 300 (K300) and 450 (K450) kg ha−1. a, b Red color indicates an increase and green color indicates a decrease in the soluble sugars accumulation and the relative expression volume in K treatments compared to the control. T1: log2(K150/control), T2: log2(K300/control), T3: log2(K450/control). Different shades of green and red express the extent of the change according to the color bar provided (log2 ratio of K treatments to control). Black color indicates no change. c The gene relative expression under the control at young fruit stage (11/5) multiplied by 10 is regarded the relative expression of key genes in different stages as the standard. AIV acid invertase, SPS sucrose phosphate synthase, SUS sucrose synthase, SUT sucrose transporters, S6PDH sorbitol 6-phosphate dehydrogenase, SDH sorbitol dehydrogenase, SOT sorbitol transporter. Different letters indicate significant differences between means (Tukey’s test, P < 0.05). (Color figure online)

The accumulation of soluble sugars in the leaves increased with the increase of K application rates in the enlargement phase I (11/6) at the T3, which was especially evident in sorbitol accumulation. The relative expression of S6PDH gene was up-regulated by the K treatment. The sugars accumulation in leaves at the enlargement phase II increased significantly with the increase of K application rates. The relative expression of sucrose metabolism and sorbitol metabolism involved in genes also increased with K fertilizer application (Fig. 3c).

The accumulation of fructose, sorbitol, and glucose in leaves increased with the increase of K fertilizer application at maturity (16/8) (Fig. 3a), however the sucrose accumulation decreased. The relative expression of sucrose decomposition and sucrose transport involved in genes (AIV1, AIV2 and SUT) were up-regulated by the K treatments. We found that at the T3 the expression levels of sugar metabolism involved in genes was the highest, and then gradually reduced with the increase of K application rates.

Expression levels of key genes involved to sugar metabolism in fruit

The accumulation of soluble sugars in fruit; as a sink organ, was from the synthesis itself. The other part was mainly the transport of leaf photosynthetic products. As shown in Fig. 4, at the young fruit stage (11/5) there was little accumulation for fruit sucrose (Fig. 4a). The relative expression of genes (AIV and SUS) involved in sucrose decomposition was up-regulated with the increase of K fertilizer application, which was more conducive to the decomposition from sucrose to fructose and glucose for fruit growth and development. With the increase of the amount of K, the accumulation of sorbitol in fruit also increased, while the relative expression of S6PDH decreased (Fig. 4c).

Fig. 4
figure 4

The effect of K application rates on the sugar accumulation and the expression levels of sugar metabolism key genes in fruit in 2015. The experiment was set up with four K2O levels: 0 (control), 150 (K150), 300 (K300) and 450 (K450) kg ha−1. a, b Red color indicates an increase and green color indicates a decrease in the soluble sugars accumulation and the relative expression volume in K treatments compared to the control. T1: log2(K150/control), T2: log2(K300/control), T3: log2(K450/control). Different shades of green and red express the extent of the change according to the color bar provided (log2 ratio of K treatments to control). Black color indicates no change. c The gene relative expression under the control at young fruit stage (11/5) multiplied by 10 is regarded the relative expression of key genes in different stages as the standard. AIV acid invertase, SPS sucrose phosphate synthase, SUS sucrose synthase, SUT sucrose transporters, S6PDH sorbitol 6-phosphate dehydrogenase, SDH sorbitol dehydrogenase, SOT sorbitol transporter. Different letters indicate significant differences between means (Tukey’s test, P < 0.05). (Color figure online)

The accumulation of soluble sugars in fruit increased as a result of the K treatments at the enlargement phase (11/6 and 16/7). The relative expression of S6PDH gene was up-regulated by the K treatments (Fig. 4b, c), while the relative expression of SDH1, SDH2, SDH3, SOT2, and SOT3 were down-regulated by the K treatments. The express levels of sucrose metabolism involved in genes (AIV, SPS, SUS and SUT) were down-regulated with the increase of K application rates.

At maturity (16/8), the accumulation of fructose, sorbitol, and glucose in fruit increased with the increase of K fertilizer application. This was especially true for the sucrose accumulation at the T3, which was significantly higher than that of the T1 (Fig. 4a). The relative expression of SPS1, SUS, and SUT genes involved in sucrose metabolism were significantly promoted by K fertilization in T2 and T3 compared to the T1. The relative expression of SDH1 and SDH3 genes was also up-regulated (Fig. 4b, c).

Redundancy analysis (RDA) of K concentration, sugar metabolites, and key genes involved in sugar metabolism

Through the RDA analysis of the relationship among the K concentration, sugar concentration, and sugar metabolism key genes we found sugar concentration and total sugar concentration in leaves were positively correlated with SOT3, SDH1, SDH3, and S6PDH involved in sorbitol metabolism in leaves and AIV2 and SPS involved in sucrose metabolism (Fig. 5a). The K concentration in fruit has a significantly positive correlation with the expression of key genes SDH1 (r = 0.573) and SUS (r = 0.766), which were involved in sorbitol metabolism and sucrose metabolism, respectively. The sorbitol concentration in fruit had a significantly negative correlation with the expression of key genes SUS (r = −0.732), SUT (r = −0.577), and SDH1 (r = −0.735). The sucrose concentration in fruit was significantly negative when correlated with the expression of key genes SUS (r = −0.526), and SDH1 (r = −0.501). The total sugar concentration in fruit was also significantly negative when correlated with the expression of key genes SUS (r = −0.515), SUT (r = −0.514), and SDH1 (r = −0.504) (Fig. 5b).

Fig. 5
figure 5

Redundancy analysis (RDA) of K concentration, sugar metabolites and the expression of key genes involved in sugar metabolism. Samples are represented as circles; sugar metabolites and key genes involved in sugar metabolism indicate red and blue lines with filled arrows. a Leaf-K: leaf K concentration; Fru-L, Sor-L, Glu-L, Suc-L and TS-L indicate the concentration of fructose, sorbitol, glucose, sucrose and total sugar in leaves, respectively; LAIV2, LSOT3, LSDH1, LSDH3, LS6PDH and LSPS1 indicate the expression of key genes in leaves, respectively. b Fruit-K: fruit K concentration; Fru-F, Sor-F, Glu-F, Suc-F and TS-F indicate the concentration of fructose, sorbitol, glucose, sucrose and total sugar in fruit, respectively; FSUS, FSDH1, FSDH2, LSUT and FSPS1 indicate the expression of key genes in fruit, respectively. AIV acid invertase, SPS sucrose phosphate synthase, SUS sucrose synthase, SUT sucrose transporters, S6PDH sorbitol 6-phosphate dehydrogenase, SDH sorbitol dehydrogenase, SOT sorbitol transporter. (Color figure online)

Discussion

K supply increased K nutrition, fruit weight and yield of pear

K availability is often limited in natural and agricultural ecosystems. K deficiency directly affects plant growth leading to decreased crop yield and production (Marschner 1995). This is the reason why supplemental K fertilizers are required for sustainable agricultural practices. Our results showed that the nutritional levels of pear leaves and fruit K were significantly improved by K fertilization. This had also been reported in many plants such as tomatoes (Schwarz et al. 2013), oranges (Quaggio et al. 2011), and grapes (Niu et al. 2008). Niederholzer et al. (1991) and Southwick et al. (1996) reported there was no effect on the size and yield of plums with the use of potassium fertilizer, but Restrepo-Diaz et al. (2008) and Nava et al. (2008) found that the size and yield of plums and apples significantly increased by K, which was consistent with the results of this study. K was reported to stimulate large number of flowers and early fruit setting in tomatoes, thus increasing fruit numbers per plant (Idowu and Aduayi 2007). In grape, the initial rapid increase in grape size was followed by a lag period of slow or no growth, while the second phase of rapid growth was due entirely to cell enlargement, which was characterized by a rapid accumulation of sugars (Mpelasoka et al. 2003). At the fruit ripening stage, the fruit K acquisition was dramatically increased leading to a quick redistribution of K from the leaves into the fruit. Understanding the dynamic changes of K concentration in the leaves and fruit would help us grasp K demand and best K application stage. In our study, we found the daily growth rate of fruit reached the peak at the enlargement phase II (Supplemental Fig. 2), at the same time, K flow was fast accumulated in fruit (Supplemental Table 3). This suggested that the key stage of K supply, which that promoted fruit size and increased fruit yield by improving the synthesis and transport of carbon in the leaves, was at the expansion phase II.

K improved fruit quality by enhancing the synthesis and transportation of photosynthetic products among the leaves and fruit

K is generally considered to be the quality element of the crop, and involved in some biochemical enzymatic functions, and promoted the synthesis and transportation of photosynthetic products (sucrose) in storage organs (Zhao et al. 2001; Beckles 2012). The transportation and distribution of photosynthetic products (sugars) are the material basis of fruit development, and an important factor in determining the yield and quality of crops. It is reported that photosynthetic products from the leaves were mainly transported through the phloem to the sink (fruit), and the most of K+ from the old leaves was transferred to young leaves or fruits through the phloem (Marschner 1995). The K+ accumulation in phloem is the primary condition to establish and maintain high osmotic potential in the sieve tube, and made sure higher transport efficiency of photosynthate (Van Volkenburgh 1999). Marschner (2012) pointed out that in the K supply the photosynthetic product produced by C14 marker in 90 min was transferred from leaves to other organs, and about 20% of the product was stored in the main storage organs, which accumulated a large amount of carbohydrates during fruit development, and were a strong metabolic pool in plant cells. The effects of K on fruit size, appearance, color, soluble solids, titratable acid, ascorbic acid content, and shelf life were important (Kanayama 2009). It was reported by Mpelasoka et al. (2003) that the increase of SSC in grapes with K fertilization rates resulted from the transport of soluble substances in fruit by K, which was consistent with the results of this study. Chapagain et al. (2003) found that the K content and acid content in tomato fruit was positively correlated, but in our study we found acid content under the K450 treatment in 2014 and 2015 was relatively lower than that under the K300 treatment. This was mostly due to the increase of SSC in fruit, especially sucrose (Fig. 2i). Hutchings (1978) reported that in Ricinus cotyledons increasing K+ influx with increasing pH suggested a link between K+ influx and H+ efflux by an H+ pump, which provides the driving force for an H+-sucrose cotransport and the movement of K+.

K-promoted sugar accumulation of the leaves and fruit might result from up-regulated expression levels of key genes involved in sugar metabolism in leaves and fruit

Fructose, sorbitol, glucose, and sucrose in pear fruit are the major components of soluble sugars (Zhang et al. 2012). The accumulation of the fructose at mature stage plays a very important role in sugar sweetness, while the changes of components of soluble sugars in leaves and fruit are regulated by the genes involved in sugar metabolism. Sugar metabolism was studied in apples (Li et al. 2012), loquats (Cao et al. 2013), pears (Liu et al. 2013), kiwifruits (Nardozza et al. 2013), and peaches (Lombardo et al. 2011). Therefore, the expression of genes putatively related to pear sugar synthesis and accumulation between leaves and fruit were analyzed (Fig. 6). In this study the relative expression of AIV, SUS, S6PDH, SPS, and SDH genes were higher in the early stage of fruit development (11/5), and were regulated by the fruit development. These observations were in high concordance with that of the study of apples (Li et al. 2012). We found that the sucrose content was very low or little at young fruit stage, which might explain up-regulation of the expression of AIV genes in fruit, mainly being decomposed to into monosaccharides (fructose and glucose). Itai and Tanahashi (2008) reported the expression of SUS (PpSUS1), AIV (PpAIV1 and PpAIV2) and SPS (PpSPS1) during storage and found that PpAIV1 was highest at a very early fruit stage (34 DAFB: 34 days after full bloom) and decreased rapidly during fruit development, while the transcript of PpAIV2, increased gradually from a very young stage (34 DAFB) until the middle stage (79 DAFB) in Japanese pear (Pyrus pyrifolia Nakai cv. Gold Nijisseiki and Hosui). The expression of PpSPS1 and PpAIV1 might be responsible for sucrose degradation during storage in Japanese pear. The fruit enlargement stage was a critical period of cell division and enlargement especially in the fruit vacuolar enlargement; the most fructose and glucose were stored in the large central vacuole of parenchyma cells (Yamaki 1984). At the maturity stage the sucrose accumulation in fruit (sink) increased with K application rates, while the sucrose accumulation in leaves (source) was reduced. The increase of the expression of SUS genes by K promoted the decomposition of sucrose in leaf, which was conductive to the synthesis of sucrose in fruit. Furthermore, the relative expression of SUT gene in leaves and fruit was up-regulated by the K treatments at maturity stage (Fig. 6). Sucrose was loaded as a carbon photoassimilate into the phloem by sucrose transporters (SUTs) and unloaded in sink tissues (Williams et al. 2000). This illustrated that K enhanced the synthesis and transport of sucrose from leaves to fruit.

Fig. 6
figure 6

A comprehensive model of the effects of K on the expression of key genes involved in sucrose and sorbitol metabolism in leaf and fruit in Asian pear from young fruit stage to maturity. Based on previous studies of the sugar biosynthesis pathway in plant (Li et al. 2012; Shangguan et al. 2014), the modified model was as follows in this study: the express levels of gene INV and S6PH were up-regulated by increasing the K application rates at the fruit young stage, leading to the increase of all soluble sugars in leaves. At enlargement phase II, the expression of AIV, SPS and SUS, S6PDH and SDH1/3 involved in sugar metabolism in leaves were up-regulated by increasing the K application rates, resulting in higher accumulation of soluble sugars in leaves. At the fruit maturity stage, the expression of SUT in leaves, and SPS, SUS and SUT in fruit was significantly up-regulated, leading to higher sucrose accumulation in fruit. The red and blue indicate the up-regulation and down-regulation of express levels of key genes with increase of K application rates. The sucrose and sorbitol metabolism steps were marked by solid lines, and the trans-membrane transporters are marked by orange, deep orange and deep blue boxes, respectively. The dashed boxes indicates a hypothesis that K+ transporters and channels play a role on the transportation of K+ and sugars (sucrose and sorbitol) from pear leaves to fruit. Young fruit young stage, Enlargement II fruit enlargement phase II, Maturity fruit maturity stage, TP phosphotriose, G-6-P glucose-6-phosphate, SPS sucrose phosphate synthase, SUS sucrose synthase, INV acid invertase, S6PDH sorbitol-6-phosphatedehydrogenase, SDH sorbitol dehydrogenase, SUT sucrose transporters, SOT sorbitol transporter, KTs K+ transporters, KChs K+ channels. (Color figure online)

K accumulation drove tissue expansion and contributed to the large accumulation of sugars after veraison in grape (Davies et al. 2006). In this study, K promoted the accumulation of soluble sugars in leaves and fruit. This was especially true during the second application of K fertilizer. The expression of key genes involved in sucrose metabolism and sorbitol metabolism in leaves were up-regulated with the increase of K application rates (Figs. 3b, c, 6), and these gene expressions had a significantly positive correlation with the components of sugar concentrations in the leaves (Fig. 5a), where it is negatively correlated in fruit (Fig. 5b). This further explained that the up-regulation of the key gene expressions involved in sugar metabolism in leaves by K was beneficial to K-promoted sugar accumulation in leaves and fruit. Philippar et al. (2003) demonstrated that specific K channels were linked to sugar loading and unloading in Vicia faba and maize. The timing and location of expression of two transporters (VvKUP1 and VvKUP2) were consistent with an involvement in K accumulation in grape berries. But, there was a shortcoming that we did not concern the express profiles of K+ transporters or channels in pear leaves and fruit in this experiment, and future experiments need to be further studied. The increase of leaves and fruit K accumulation (Figs. 4a, 5a, 6) promoted the synthesis of photosynthetic products; this may be involved in the activation of K+ to the enzymes (Leigh and Jones 1984; Amtmann et al. 2006; Anschütz et al. 2014). Our data indicated that the accumulation of fruit soluble sugars were not only affected by the source-sink strength but also regulated by the inner regulating mechanism between sorbitol and sucrose metabolism among the leaves and fruit.

Conclusion

Results obtained in this 2-year study confirmed that the application of K increased K nutrition of the pear leaves and fruit. Compared to the control, the fruit single weight and yield between the K300 and K450 treatments significantly increased. The accumulations of soluble sugars (fructose, sorbitol, glucose, and sucrose) in leaves were correlated with the up-regulated expression of genes AIV and S6PDH during the early stages. Furthermore, with fruit development the expression of AIV1, SPS1, SUS, S6PDH, and SDH3 involved in sugar metabolism in the leaves were up-regulated by increasing the K application rates. This resulted in higher accumulations of soluble sugars (fructose, sorbitol, glucose, and sucrose) in leaves. At the fruit maturity stage the up-regulation of the expressions of SUT in leaves and SPS1, SUS, and SUT in fruit; which were involved in sucrose metabolism, lead to higher sucrose accumulation in fruit. Our data clearly illustrates that K plays a key role on promoting the assimilation of photosynthetic products in leaves and fruit, and up-regulating the expression of the key genes involved in sugar metabolism in leaves and fruit.