Though fruit flavor is related to both taste (e.g., sugar and acid content) and aroma components, the main component dictating pineapple fruit eating quality is influenced by total sugar content as well as composition [12]. In this study, three major sugar (fructose, sucrose and glucose) contents of pineapple fruits were determined throughout development stages.

In the early stages of fruit growth, hexose (fructose and glucose) was the predominant sugar in nearly equal amounts except for medial section, accounting for ca. 80% of total soluble sugars (Figure 1a,b). A parallel gradual accumulation of glucose and fructose contents was detected in the apical and basal sections from 50 DAA to 70 DAA with the exception of core section, and then decreased in mature fruits. In medial sections, the fructose and glucose of medial sections also increased dramatically from 50 to 70 DAA and reached the highest levels at 70 DAA (23.86247 mg·g⁻¹ FW) and 60 DAA (36.37503 mg·g⁻¹ FW), respectively. Diametrically, the dynamic activities of two sugars are slightly decreased during fruit growth and maturation in core sections, which may be related to metabolizing-enzyme activities [13–15].

As reported by many researchers, sucrose is the chief contributor to sweetness, and is transported through sieve elements and can enter into sink organs directly by the plasmodesmata or the apoplastic space [15–16]. In our study, gradual increases of sucrose contents were detected starting from 20 DAA up to the end of fruit growth regardless of fruit sections, and reached 7 fold of initial content in basal section (Figure 1c,d). This may explain why sweet has a certain difference in the various portions and periods.

However, the variation tendency of sucrose was moderate in different parts of fruit between 50 and 60 DAA, accordingly, hexose (fructose and glucose) contents were sharply increased (Figure 1a,b). It indicated that the conversion of sucrose to hexoses was indispensable and may have other benefits. For example, hexose accumulation may function to increase the osmotic potential of the tissues and allow improved nutrient uptake [17]. Hexoses also play a role in regulating gene expression and therefore, accumulation in the vacuole would permit hexose storage without adverse effects on gene expression since the hexoses would be effectively excluded from the cytosol [17].

In order to evaluate the importance of carbohydrate-metabolizing enzymes throughout pineapple fruit development, the spatial and temporal expression patterns revealed that three genes were differentially expressed in immature and mature fruits. 18S rRNA expression levels were chosen as standards for constitutive expression and gave closely similar results. Expression levels of Ac-sps mRNA revealed some differences in all developmental stages and different parts of pineapple fruits.

Recent evidence suggested that SPS may play a more important role in sucrose synthesis and accumulation [7,15,17,19], as supported by the transcript characteristics of Ac-sps in sucrose accumulation (Figure 5a). In contrast to mRNA levels during the initial phase of anthesis, Ac-sps transcript levels of fruit different sections were nearly devoid in 20 DAA except for medial parts, which was associated with low levels of sucrose in immature fruit sections, as also observed by Botha & Black (2000) [24]. They gradually reached two peaks of accumulation with maxima at 30 and 70 DAA in basal and medial sections, respectively, and then were down-regulated up to 80 DAA. However, a transcript peak of 30 DAA was lost in apical and core sections. This reason may be caused by the differences of organic acid and secondary metabolites in various fruit sections [17]. Experiments with banana had consistently demonstrated SPS activity increase with accumulation of sucrose in agreement with our results [25]. Strand et al. also provided the evidence that decreased expression of sps can inhibit sucrose synthesis in transgenic A. thaliana [26]. Though a remarkable rise is showed along with fruit maturation, there is a significant difference between the core and apical sections and the basal and medial sections (p < 0.05, n = 3) throughout the ripening process of pineapple fruits. Zhang et al. (2010) also found that different harvest seasons had serious influence on the activities of carbohydrate-metabolizing enzymes and sugar component in the different stages of fruit development [15].

The other most significant contributions of our study are that not only SPS but also SuSy and NI activities are positively correlated with sugar accumulation. Though SPS activity had also been implicated in sugar metabolism, the roles of other two enzymes had been less clear [11]. In our experiments, Ac-susy transcription levels were gradually up-regulated, and exhibited obvious profiles in the different sections with fruit development (Figure 5b). Thereafter, Ac-susy transcripts in four different sections showed a marked decrease to basal levels after 70 DAA until harvest. In medial and basal sections, there are more obvious than the other parts in Ac-susy transcript levels. However, some studies had attributed sucrose breakdown roles to these two latter enzymes [16], but our study clearly indicated that activities of SuSy were increased with sucrose accumulation in different sections of pineapple fruits, rather than the opposite, similar with results detected in watermelon [3] as well as other fruit [27]. These results indicated Ac-susy transcript was bound up with sucrose increase. Ciereszko et al. also demonstrated that the sucrose/glucose-dependent SuSy expression was strongly induced in transgenic Arabidopsis hexokinase-overexpressing plants [28]. Although details of this regulation are unclear, SuSy activity and expression of the corresponding gene(s) were more or less related with hexokinase.

Due to low enzyme activity and gene transcription levels, NI was scarcely studied [11]. Nevertheless, higher levels of the Ac-ni transcripts were detected between 30 DAA and 50 DAA with a transient increase of hexose contents in our study (Figure 5c). On the other hand, though the abundance of Ac-ni mRNA also increased and had a single peak around 40 DAA regardless of different sections in the early development of the fruits, Ac-ni expression levels decrease come up more early than that of Ac-susy and Ac-sps. From 40 to 80 DAA, Ac-ni transcription was down-regulated to a negligible level, especially in the core and basal sections. These data together suggest the putative role of hexose in regulating Ac-ni transcription through a negative feedback control had been evaluated. This hypothesis may be further supported by the Ac-ni transcript levels of different fruit sections, because hexose accumulation of the basal section analogous to Ac-ni transcript levels precedes the other parts. Nonis et al. (2007) also reported Ac-ni transcription levels can be improved using the hexokinase inhibitor NAG [29]. While previous studies on NI revealed small changes in gene expression among different tissues and during organ development [22], the pineapple NI studied in this article displayed a highly modulated pattern of expression in response to developmental clues. The major function of the high and constant transcription activity of Ac-ni in immature pineapple fruit is to maintain the cellular hexose concentrations. A decreased invertase activity would lead to increases in sucrose/hexose ratios [30,31]. This emphasizes not only the continuous hydrolysis of sucrose catalysed by invertase, but also the possibility of continuous sugar exchange between cytosol and vacuole (sucrose influx and hexose efflux).

As expected, analogous results had been obtained in a comparison of enzyme activities and gene transcript levels. Those clearly indicated that sucrose accumulation occurs only in the presence of threshold activities of the three enzymes directly involved in sucrose metabolism. However, the dynamic patterns of three enzyme activities and transcripts during fruit development could not compare directly with each other, because multiple gene families could be differentially expressed in various parts and developmental stages of plants [18]. SPS, SuSy and NI had been reported in different flowering plants and database searches supported the existence of many isoforms of these enzymes [10,19,30]. The enzyme activity is synthetically performed by translations of many isozymes, but the mRNA transcript is one of the isozymes. Moreover, the protein contents of enzyme increased, and the increase of transcripts may push up the activity of enzymes.

In some fruits, sucrose is translocated from the source to the fruit and is either apoplastically hydrolysed to fructose and glucose by invertase, or is cleaved symplastically to fructose and UDP-glucose by sucrose synthase [31]. In the present work, the Ac-NI activities were very low in basal and core sections at 20 DAA, but rapidly increased and reached maxima (14-fold and 10-fold, respectively) at 40 DAA. A pronounced increase was also detected in apical and medial activities from 30 to 40 DAA (Figure 6a). Thereafter, activities of various sections showed similar patterns with a slight peak at 70 DAA, followed by a gradual decrease towards becoming undetectable at 80 DAA. A positive correlation between sugar accumulation (Figure 1) and levels of SuSy and SPS activities (Figure 6b,c) were indicated in fruit different sections during development, which is paralleled with the transcript characteristics of two genes. These results suggest that sucrose synthase can only cleave the sucrose translocated to fruit tissue at a mature stage, and that NI activity is closely related to the accumulation of fructose and glucose in immature pineapple fruit. This is consistent with the suggestion that the two steps of sucrose degradation by SuSy and NI and the re-synthesis by SPS in fruit could generate a locally increased sucrose concentration gradient in the zone of phloem unloading, thus favoring sucrose import [17]. The potential capacity for re-synthesis of sucrose by the action of SPS in fruit is also supported with the presence of SPS activity and transcript of Ac-sps, which increase until the mature stage in fruit different sections (Figure 6c). A series of actions by SuSy and SPS during maturation could be important to sink strength in citrus fruit at this stage [5]. Thus, NI may have a role in supplying materials for growth and cell wall construction, whereas SuSy may supply SPS with substrates (UDP-glucose and fructose) for the re-synthesis of sucrose [9]. Substantial progress has been also made in studies on SuSy synthesizing UDPG which is essential for sucrose and cell wall biosynthesis [32]. Since pineapple fruits have a high capacity to cleave sucrose to glucose and fructose in immature stages (Figure 1), it seems likely that it is glucose rather than sucrose that serves as a signal in upregulating Ac-susy expression. Though the physiological role of NI and SuSy in sucrose metabolism is not completely clear in fruit development, the positive correlation with sucrose accumulation in our study suggests the need to rethink this step of sucrose metabolism in the context of sink sugar accumulation.

Field-grown pineapples (A. comosus cv. Comte de paris) were cultivated in the pineapple resource bank of South Subtropical Crops Research Institute. The experimental design used in this study was under the same management conditions such as irrigation, fertilization, soil management, disease control and pruning. Ninety fruit samples had been selected after the full florescence period from April to June in 2007. The fruits were randomly sampled every 10 days from the 20th day after anthesis (DAA) and cut transversely into apical, medial, basal and core sections, after the size and weight of crowns and fruits were measured. And then, the tissues were immediately frozen in liquid nitrogen and stored at −80 °C before being analyzed. These sliced fleshes of 30 fruits were pooled together as one of three replications at each harvesting time.

For glucose, fructose and sucrose content, the sugars were extracted by grinding flesh tissues (10 g) in 80% ethanol, adjusted to pH 7.0 with 0.1 M NaOH, and heated for 5 min at 80 °C. The extract corresponding to 0.5 g fresh weight (FW) was dried in vacuum, redissolved in water and the solution was passed through an ion-exchange column (Dowex 50W-X8 and Dowex 1-X8). The eluate (10 μL) was then analyzed by high-performance liquid chromatography (HPLC) [13] (Shimadzu LC-6A; Kyoto, Japan) equipped with a RI detector and a SP1010 column (Showa Denko K. K., Tokyo, Japan) at a flow rate of 0.5 mL·min⁻¹.

Sugar-metabolizing enzymes were prepared essentially in frozen flesh tissue as described by Zhang et al. [15]. Fruit material was homogenised in 10 mL of ice-cold buffer containing 50 mmol·L⁻¹ Hepes-NaOH (pH 7.5), 0.5 mmol·L⁻¹ Na-ethylenediaminetetraacetic acid (EDTA), 2.5 mmol·L⁻¹ DTT, 3 mmol·L⁻¹ diethyldithiocarbamic acid, 0.5% (w/v) bovine serum albumin (BSA) and 1% (w/v) insoluble polyvinylpyrrolidone (PVP). After centrifugation at 12,000 g for 20 min at 4 °C, supernatants were dialyzed for approximately 20 h against 25 mmol·L⁻¹ Hepes-NaOH (pH 7.5) and 0.25 mmol·L⁻¹ Na-EDTA and used as the crude soluble enzyme extract. The insoluble pellet was washed twice in homogenising medium and then incubated with shaking for 4 h in ice-cold homogenising medium containing 1 mol·L⁻¹ NaCl.

NI activities were assayed in a final volume of 25 mL, containing 0.2 mL of dialyzed enzymatic extraction, 0.8 mL of reaction solution (pH 4.8 or 7.2, 0.1 mol·L⁻¹ Na₂HPO₄, 0.1 mol·L⁻¹ sodium citrate, 0.1 mol·L⁻¹ sucrose for neutral invertase). The activities were measured by the quantity of reducing sugars released in the assay media with dinitrosalicylic acid [14].

Activities of SPS were assayed of using 0.15 mL of reaction medium and 0.2 mL of enzyme sample. The reaction medium is composed of 50 mmol·L⁻¹ Mops-NaOH (pH 7.5), 10 mM MgCl₂, 5 mmol·L⁻¹ glucose-6-phosphate, 10 mmol·L⁻¹ fructose-6-phosphate and 5 mmol·L⁻¹ UDPG. After the mixture was incubated for 30 min at 37 °C, the reaction was stopped by adding 0.1 mL 30% (w/v) NaOH and kept in boiling water for 5 min. When cooled to room temperature, the resorcinol solution (12%, v/v) of 0.5 mL and HCl (12 mol·L⁻¹) of 0.5 mL were added into the mixture and held in an 80 °C water bath for 10 min. Blank controls were obtained by adding the sterile water to the reaction medium containing resorcinol. The procedure for the SuSy assay was identical to that of SPS except for the reaction mixture, where fructose 6-phosphate or glucose 6-phosphate was replaced by contained 10 mmol·L⁻¹ fructose.

DPSv3.01 for the variance analysis and the correlation analysis by SAS 9.0 according to different requirements was done. Quantitative analyses were presented as mean values and the reproducibility of the results was expressed as standard error (S.E.).