|Home | About | Journals | Submit | Contact Us | Français|
Although cooked potato tuber texture is an important trait that influences consumer preference, a detailed understanding of tuber textural properties at the molecular level is lacking. Previous work has identified tuber pectin methyl esterase activity (PME) as a potential factor impacting on textural properties. In this study, tuber PME isoform and gene expression profiles have been determined in potato germplasm with differing textural properties as assessed using an amended wedge fracture method and a sloughing assay, revealing major differences between the potato types. Differences in pectin structure between potato types with different textural properties were revealed using monoclonal antibodies specific for different pectic epitopes. Chemical analysis of tuber pectin clearly demonstrated that, in tubers containing a higher level of total PME activity, there was a reduced degree of methylation of cell wall pectin and consistently higher peak force and work done values during the fracture of cooked tuber samples, demonstrating the link between PME activity, the degree of methylation of cell wall pectin, and cooked tuber textural properties.
Potato tuber texture is a key quality determinant of cooked potato and a major trait that influences consumer preference (McGregor, 2007). Tuber texture is also a key issue in potato processing and is known to be affected by pre-processing procedures such as blanching and peeling, and by storage (Shomer and Kaaber, 2006; Thybo et al., 2006). Factors that may impact on cooked potato texture include starch content and distribution within the tuber, starch swelling pressure, cell size, cell-wall structure and composition, and the breakdown of the cell wall middle lamella during cooking (reviewed in Taylor et al., 2007). The relative importance of the different factors involved is not yet clear. Several studies have described potato germplasm that produces tubers with markedly different textural properties (Van Marle et al., 1994; Ducreux et al., 2008; Ross et al., 2010). In particular, members of the Solanum tuberosum group Phureja (Phureja) have been identified which exhibit a boiled tuber texture described as extremely floury or crumbly (De Maine et al., 1993, 1998). The cooking time of Phureja tubers is generally in the order of half that taken for typical Solanum tuberosum group Tuberosum (Tuberosum) tubers at the same developmental stage (Ducreux et al., 2008; Ross et al., 2010).
Mechanical methods have been developed for the quantitative assessment of textural properties. A wedge fracture test was first developed by Vincent et al. (1991) to provide a method that reflects the sensory experience of the consumer during consumption. Good correlations between the wedge fracture and panel sensory tests for a range of different food types were demonstrated. More recently, the method has been developed to assess cooked potato tuber texture (Ross et al., 2010) and it was demonstrated that tubers from the Phureja group, when cooked (by steaming or boiling) gave significantly lower peak force and work done values during fracture than typically observed for those from Tuberosum. Apart from the stem end of the tuber, which is enriched in vascular tissues, there was little variation in tuber textural properties throughout the rest of the tuber. Interestingly, the dry matter content of the pith region of the tuber was 25% lower than in the perimedullary region and yet the measured textural properties were very similar, suggesting that dry matter content did not have a great influence on tuber texture, at least in the potato lines assessed by this method. Other textural properties such as the degree of sloughing, defined as the flaking and disintegration of the outer layers of potato tubers cooked in water, also show considerable variation between potato cultivars (Hejlova and Blahovec, 2008).
A recently developed 44 000-element potato microarray was used to identify tuber gene expression profiles that correspond to differences between Phureja and Tuberosum germplasm differentiated in a range of traits including tuber textural properties (Ducreux et al., 2008). Based on differential expression patterns between Phureja and Tuberosum tubers, several genes related to cell wall metabolism are candidates that may account for tuber textural differences. The most clear-cut example is a gene encoding pectin methyl esterase (PME, MICRO.4403.C2_728). Orthologous PME genes have been shown to impact on the texture of fruit from many species (reviewed in Fischer and Bennett, 1991). As pectin is a major component of the cell wall and the middle lamella, its structure is likely to be an important factor in texture in potato tubers as well as other plant tissues (Fischer and Bennett, 1991). The pectin molecule consists of a backbone of α-(1-4)-D-galacturonic acid and contains some regions with alternating L-rhamnose and D-galacturonic acid (Voragen et al., 2001). Twenty to eighty per cent of the rhamnose units in the backbone can be branched, with side chains composed of galactose, arabinose, and small amounts of fucose and mannose. Some of the galacturonic acids in pectin are methyl esterified or acetylated. Non-methylated D-galacturonic acid sequences are sites for cross-linking polymeric chains through the site-specific interaction with Ca2+ (Pilnik and Voragen, 1991). The content of unesterified uronic acid residues can be increased by the action of PME. The binding of pectin chains has a co-operative nature and an increase in Ca2+ binding efficiency contributes to higher tissue firmness (Andersson et al., 1994). The much higher (c. 10-fold) expression level of the PEST1 gene and total PME activity in Tuberosum tubers may underpin changes in pectin structure and contribute to the differences in textural properties compared with Phureja (Ross et al., 2010). In all plants studied to date, several PME genes are expressed (reviewed in Pelloux et al., 2007). For example, tomato fruit contains at least three isoforms of PME, which have been termed PE1, PE2, and PE3 (Tucker et al., 1982), while at least five isoforms have been detected in leaf tissue (Gaffe et al., 1994). Detailed analysis has identified differences in the temporal and spatial patterns of expression of the tomato PME genes. Different functions for the isoforms are implied by the results of silencing experiments. For example, PE1 down-regulation leads to an increase in the rate of fruit softening (Phan et al., 2007) whereas PE2 down-regulation has no appreciable effect in this regard (Hall et al., 1994).
In view of the reported differences in texture between Phureja and Tuberosum, an objective of this study was to investigate the activity profile of PME isoforms in tuber extracts from the Phureja and Tuberosum types. A further objective was to associate the changes in PME expression and activity with differences in pectin structure, revealed both biochemically and by the use of monoclonal antibodies specific for different pectic epitopes.
Commercially-available fresh market potato cultivars examined in this study (Phureja cultivars Mayan Gold and Inca Sun and Tuberosum cultivars Desiree, Montrose, and Pentland Dell) were grown in field trials during 2007 at Gourdie Farm, Dundee, UK (56°29' N 3°3' W) using standard agronomic practices. The trial design was a randomized complete block (three blocks) with plots consisting of 20 plants. Tubers for each cultivar were harvested at maturity for each plot within the three blocks. Representative-sized tubers were selected and formed the basis for all the analysis described in this paper.
All tests were carried out on a QTS 25 Texture Analyser (Brookfield Engineering, Harlow, UK). The potato samples were boiled for 10 min and the wedge fracture test (as outlined by Vincent et al., 1991) was used as described by Ross et al. (2010). Briefly, using an acrylic wedge descending through the cooked tuber cube at 5 mm min−1, two measurements were recorded for each sample; work done is the energy required to penetrate to 10 mm while peak force is the maximum force required for the wedge initially to cut and then force the tissue apart and propagate a crack in the cube ahead of the wedge.
Tubers were extracted in 2 vols ice-cold 20 mM potassium phosphate buffer (pH 7.0) containing 1 mM PMSF, 1 M NaCl, and 10 mM Na2SO3 in a cooled Waring Blender. The extracts were centrifuged at 4 °C at 20 000 g for 30 min and the volume of the removed supernatant was measured. Protein from the supernatant fractions was precipitated by the addition of (NH4)2SO4 to 80% saturation at 4 °C and resuspended in 10 mM TRIS-HCl (pH 7.5). The protein extracts were then dialysed overnight against 10 mM TRIS-HCl (pH 7.5) (2× 4.0 l) prior to column separation of the isoforms. The separation of the PME isoforms was carried out using Affi-Gel Heparin Gel (Bio-Rad, Hemel Hempstead, UK) affinity chromatography (20 ml column volume) similar to that described by Phan et al. (2007). A HPLC 432 UV/VIS detector (Kontron, Chichester, UK) was used along with a Kontron HPLC pump 420 and HPLC gradient former 425. The column was maintained at 4 °C using a cooled water bath pumped to a sleeve surrounding the column. The cooled column was equilibrated with buffer A (10 mM TRIS-HCl, 10 mM NaCl, pH 7.5). Prior to loading, the dialysed protein extracts were centrifuged for 5 min at 16 000 g, and 2 ml samples were applied in buffer A. The column was eluted at 1 ml min−1 with buffer A for 20 min and then a linear gradient was applied to 25% buffer B (10 mM TRIS-HCl, 1 M NaCl, pH 7.5) over 100 min. Finally, step gradients were applied to 50% and 100% buffer B. Following each step, elution was continued until all protein had been washed off the column as determined by eluent absorbance at 280 nm. Fractions (96×2 ml) were collected on a Frac-100 fraction collector (Pharmacia, Stockholm, Sweden) and kept on ice prior to assay. The assay procedure was as described by Phan et al. (2007) using 20 μl of each fraction in a 96-well microtitre plate assay. Assay buffer (200 μl) consisting of 0.5% citrus pectin, 2 mM TRIS-HCl, 150 mM NaCl, and 0.002% phenol red (pH 8.2) was added into each well. Absorbance at 405 nm was read on a Spectra-Max M5 plate reader (Molecular Devices, Sunnyvale, CA, USA).
The PME activity of total protein samples prepared as described above, were assayed over a range of NaCl concentrations from 10 mM to 180 mM using the method described by Ross et al. (2010).
Tubers from each of the potato lines harvested at maturity from the field were used for testing the degree of sloughing during boiling. The tubers were cut into seven 20 mm cubes to ensure all areas of the tuber were tested and the group of seven cubes weighed prior to boiling. The cubes were placed in a wire basket in 1.0 l of vigorously boiling water in a pan. A pretest of cooking times for each of the lines was carried out by subjective assessment and this showed that the Phureja lines were adequately cooked after 5–6 min, whereas the Tuberosum lines required 14 min. Consequently, Inca Sun and Mayan Gold were tested over 4, 6, 8, and 10 min and Desirée and Pentland Dell over 10, 14, 16, and 18 min. At each of the time points, the cubes in the wire baskets were removed and the boiling water containing any sloughed material was poured through double, preweighed coffee filters. Four tubers of each line for each time period were used. The filters were placed in polythene bags and frozen at –20 °C along with blank control filters through which 1.0, l of boiled water had been passed. The opened polythene bags containing the filters were then lyophilized (Christ Gamma 1-16 LSC, Shrewsbury, UK) and reweighed to obtain weight gain due to freeze-dried sloughed material.
To show the location of the sloughing across the tuber, cubes of tuber were cooked and photographed. Pentland Dell, Montrose, Inca Sun, and Mayan Gold tubers were cut to produce a 20 cm wide ring of tissue across the centre of the tuber. This was further dissected to produce 20 cm ‘cubes’ of tuber that incorporated the epidermis. Tissue samples were cooked for 10 min in boiling water, cooled rapidly in chilled water, and photographed using a digital SLR camera (EOS 350D; Canon, Woodhatch, UK).
Reverse transcription of 10 μg of RNA was performed using Invitrogen Superscript™ II reverse transcriptase (www.invitrogen.com) using random hexamers as primer. cDNA (25 ng) was used as template for real-time PCR using the Universal Probe Library System (https://www.roche-applied-science.com/sis/rtpcr/upl/index.jsp). Reactions were performed in 25 μl containing 1× FastStart TaqMan® Probe Master (supplemented with ROX reference dye). Gene-specific primers and probe were used at a concentration of 0.2 μM and 0.1 μM, respectively. Thermal cycling conditions were: 10 min denaturation at 95 °C followed by 40 cycles (15 s at 94 °C, 60 s at 60 °C). The reactions were repeated in triplicate with independent cDNAs. Relative expression levels were calculated and the primers validated using the ΔΔCt method (Livak and Schmittgen, 2001) using data obtained with the elongation factor-1 alpha specific primers as an internal reference control. In the case where relative efficiencies of the target and reference amplicons were not within recommended limits, an alternative method for calculating relative quantification was used (Pfaffl, 2001). Universal probe library (UPL) primer and probe sequences were as follows: EF1 alpha fwd, 5′-CTTGACGCTCTTGACCAGATT-3′, EF1 alpha rev, 5′-GAAGACGGAGGGGTTTGTCT-3′, UPL probe number 117 (5′-AGCCCAAG-3′); Micro.2764.c1 fwd, 5′-GGACTGAAACCAGATGTGACAG-3′, Micro.2764.c1 rev, 5′-CGTCTCATAATCGCAACCAC-3′, UPL probe number 66 (5′-GGCTGCTG-3′); Micro.2676.c2 fwd, 5′-GGGAACTTACAAAGAAAATGTTGAG-3′, Micro.2676.c2 rev, 5′-GCCATCACCCACAAGCATA-3′, UPL probe number 143 (5′-AGAAGAAGA-3′); Pest1 fwd, 5′-TGGGTGAGTTCGAGGGATAG-3′, Pest1 rev, 5′-TGCATTTGCTTTGATGTCCT-3′, UPL probe number 156 (5′-GCTGATGG-3′); Pest2 fwd, 5′-TGAGGCCCAAGCTTATACTCC-3′, Pest2 rev, 5′-CCAGTGGAGCTCAACCAACT-3′, UPL probe number 67 (5′-TGCTGGAG-3′).
Cell walls were prepared from tuber samples collected within 1 week of harvest at full maturity. Cell walls were prepared from frozen tuber material largely following the method of Jardine et al. (2002). Briefly, 100 g of tubers were chopped with knives, whilst still frozen, and placed into 500 ml ice-cold mixed cation buffer (MCB) consisting of 10 mM sodium acetate, 3 mM KCl, 2 mM MgCl2, and 1 mM CaCl2, pH 6.5 containing 0.5 ml Triton X-100 then homogenized using an Ultra-Turrax (IKA, Staufen, Germany; 10×1 min bursts; 3/4 power). All extractions were carried out in triplicate using tubers from the three separate field plots for each tuber type. Insoluble material was collected by sieving on 53 μM nylon mesh and washed with 500 ml MCB to remove Triton X-100. All of the wash was collected as Wash A. The crude cell wall material (CWM) was washed with 500 ml of 50% acetone/ultra pure water (UPW) then recovered by filtration on a grade 2 glass sinter and weighed.
The CWM was deproteinized by extraction in 4 vols of 80% (w/v) saturated phenol for 30 min at room temperature with stirring at 80 rpm. After extraction, the CWM was recovered by filtration on a glass sinter then washed with 500 ml ice-cold MCB.
After frozen storage, the CWM was de-starched. The material was suspended in MCB preheated to 80 °C and incubated at 80 °C in a water bath for 45 min to gelatinize the starch. The material was then cooled rapidly to 20 °C by placing in cold water. Starch digestion was initiated by adding 1 g of pancreatic amylase powder (Sigma type VI-B) and 25 μl of pullulanse suspension (Sigma P-5420) and incubating overnight at 25 °C with rotary shaking at 80 rpm.
Starch digestion was monitored by removing small aliquots of the insoluble material and staining with iodine. Samples were satisfactorily de-starched following the addition of another 0.5 g of amylase and further incubation for 8 h. The de-starched CWM was recovered on a glass sinter and the filtrate collected as Wash B. The CWM was washed with excess MCB then dried by washing in 50% acetone then from water. Samples were frozen then lyophilized.
Cell wall recovery ranged from 2–5 g dry weight per 100 g tuber material and the residual starch content was below 5% (Karkalas, 1985). Samples of washes A and B were centrifuged at 15 000 g for 10 min at 4 °C to remove insoluble starch. Soluble pectins in the wash A and wash B supernatants were recovered by precipitation with 3 vols of ice-cold ethanol and centrifugation at 15 000 g for 15 min at 4 °C. The pelleted polysaccharides were re-suspended in UPW and the galacturonic content assessed (Filisetti-Cozzi and Carpita, 1991). The wash A samples contained very little pectin whereas substantial amounts were present in Wash B.
The degree of pectin methylation was determined for the cell wall samples and the solubilized pectins. After saponification, released methanol was measured using the alcohol oxidase/Purpald method as described by Anthon and Barrett (2004). Fourier transform infra-red (FT-IR) spectroscopy was performed using the DRIFT method on a Bruker IFS-66 spectrometer (Bruker UK Ltd, Coventry, UK) as described previously (Stewart et al., 1997). Triplicate spectra were obtained for each sample and averaged, then normalized using the OPUS 5.5 software. Spectra were also obtained from commercially-available pectin standards with different degrees of methylation to direct assignment of spectral peaks.
Wedges of potato tuber tissue 2–3 mm thick were cut from mature tubers of Phureja cultivar Inca Sun and Tuberosum cultivar Pentland Dell. Using a TP1020 tissue processor (Leica, Milton Keynes, UK), the samples were fixed in 4% paraformaldehyde in PEM buffer (0.1 M PIPES pH 6.95, 1 mM EGTA, 1 mM MgSO4) for 6 h, washed in 0.85% NaCl (30 min), dehydrated in an ethanol series (70% for 1 h, 80% for 1.5 h, 90% for 2 h, 3× 100% for 1 h, 1.5 h, and 2 h) and xylene (100% for 1 h and 1.5 h) before infiltration with Paramat extra wax (VWR-Gurr, Lutterworth, UK) (2 h and 4 h) at 65 °C. Following embedding, 10 μm thick sections were cut using a RM2265 rotary microtome (Leica, Milton Keynes, UK) and mounted on Polysine slides (VWR, Lutterworth, UK). The sections were dewaxed in Histo-Clear II (National Diagnostics, Hessle, UK) prior to rehydration in an ethanol series (100%, 95%, 85%, 70%, 40%, 0%) made in the appropriate buffer (see below).
Sections to be incubated with monoclonoal antibodies (mAbs) JIM5, JIM6, LM5, and LM6 were rehydrated in PBS buffer (pH 7.4) before blocking with 1% BSA in PBS. Sections were incubated overnight at 4 °C in this blocking buffer including mAbs diluted 1:20, rinsed in blocking buffer and incubated in Alexa 488 nm goat anti-rat (Invitrogen, Paisley, UK) diluted 1:100 in blocking buffer for 3 h at 20 °C. Sections to be incubated with mAb 2F4 were rehydrated in TCaS buffer (20 mM TRIS-HCl, pH 8.2, 1 mM CaCl2, 150 mM NaCl) and blocked in 1% BSA in TCaS. The 2F4 antibody was diluted 1:75 in this blocking buffer before incubating as described above and the sections were then treated with Alexa 488 nm goat anti-mouse (Invitrogen) diluted 1:100 in blocking buffer. The sections were rinsed in blocking buffer and the appropriate individual buffer prior to staining with 0.25% Alcian blue in 3% acetic acid for 10 min and mounting in Fluoromount G (Southern Biotech, Birmingham, Alabama, USA).
Confocal sections were collected using a Leica SP2 confocal laser scanning microscope (Leica Microsystems, Heidelberg GmbH, Germany) fitted with a HC PL Fluotar 10.0× 0.30 dry objective, with 488 nm excitation and emission collected between 500 nm and 530 nm. The laser intensity and gain settings were maintained at the same levels for all samples. Multiple z-stacked images were collected across the whole of the tissue using the tile-scan function and motorized stage of the microscope, maximum intensity projections were then produced from the z-stacks and stitched together using Leica LCS software to create maps of the whole tissue with microscopic resolution.
Pieces of tuber tissue were fixed, dehydrated, and embedded in LR White resin as described by Bush and McCann (1999). Ultrathin sections were collected on pyroxylin-coated nickel grids. Detection with antibody LM6 was performed as described by Bush and McCann (1999), except that grids were blocked on Goat Gold Conjugate Blocking Solution (Aurion, Wageningen, Netherlands) and LM6 was diluted in 0.1% BSA-c (Aurion) in phosphate-buffered saline (PBS). Probing with antibody 2F4 was performed as described by Bush and McCann (1999), except a dilution of 1:50, instead of 1:300, in 1% (w/v) skimmed milk in TCaS buffer was used. Following incubations with primary antibodies, grids were washed five times with the same buffer used for dilution of the primary antibody. Grids were then incubated with 10 nm gold conjugates (British BioCell International, Cardiff, UK) diluted 1:50 in the same buffer as used for the primary antibody. Subsequently, grids were washed five times on the buffers used for antibody dilution, three times on PBS or TCaS buffers, and twice on distilled water. Ultrathin sections were counterstained with uranyl acetate and lead citrate and examined in a JEM1400 (Jeol UK Ltd, Welwyn Garden City, UK) transmission electron microscope.
Recently, the textural characteristics of potato tubers during the boiling process have been described, comparing tuber textural properties in three Phureja cultivars (Inca Sun, Mayan Gold, and breeding line DB333/16) with those from four table Tuberosum cultivars (Pentland Dell, Montrose, Maris Piper, and Desiree) (Ross et al., 2010). Using a wedge fracture test, for an equivalent cooking time (10 min), it was demonstrated that the Phureja types required a substantially (3–5-fold) lower peak force to initiate fracture and required a lower total amount of work done to propagate a fracture of 10 mm. Similar differences in textural properties were measured in the Phureja and Tuberosum samples used in this study, demonstrating that this trait was stable year-to-year (data not shown). As a further measurement of the differences in textural properties between the Phureja and Tuberosum types, the degree of sloughing (Hejlova and Blahovec, 2008) from cubes of potato during boiling was measured (Table 1). Comparing values after 10 min of boiling, the Phureja samples had sloughed appreciably whereas the amount of material sloughed from the Tuberosum samples was at least 8-fold lower. Tuberosum tubers (Fig. 1A, B) were largely intact after 10 min cooking whereas Phureja tubers (Fig. 1C, D) showed marked separation of the tissue in the region of the vascular ring and cortex resulting in sloughing of material from the cortical/epidermal layers.
The total pectin methyl esterase activity was significantly (3–5-fold) lower in Phureja than Tuberosum tubers (Fig. 2A). PME isoform profiles were investigated in tubers from the Phureja (Inca Sun and Mayan Gold) and Tuberosum (Pentland Dell and Desiree) cultivars using heparin column chromatography. Profiles from Pentland Dell and Inca Sun are shown (Fig. 2B (i) and (ii), respectively), but similar profiles were obtained for extracts from Mayan Gold and Desiree. Although three peaks of activity could be resolved in samples from all four cultivars, much greater activity was found in the Tuberosum samples. Therefore, Tuberosum profiles were obtained after 100 min but 600 min for the Phureja profiles. The Phureja and Tuberosum profiles were consistently qualitatively similar and further analysis, perhaps using transgenic approaches to manipulating individual isoforms, would be required to confirm whether the peaks represent the same isoforms. The salt sensitivity of the three resolved isoforms was determined as both salt-sensitive and insensitive isoforms have been identified in tomato (Phan et al., 2007). In contrast to tomato fruit, no salt-insensitive PME activity could be detected. Activity associated with peak 1 required the lowest salt concentration for maximal activity (40 mM), whilst that in peaks 2 and 3 required 80 mM for maximal activity (data not shown).
Previous microarray analysis of Phureja and Tuberosum tuber gene expression has identified that five genes, annotated as encoding PME, are expressed in potato tubers (Ducreux et al., 2008). Quantitative RT-PCR analysis was undertaken to investigate the expression levels of these genes in the samples used for this study (Fig. 2C). Genes encoding two PME isoforms were expressed at significantly higher levels in Tuberosum than Phureja tubers (MICRO.2676 and PEST1) whereas two genes MICRO.2764 and PEST2 were expressed at higher levels in Phureja tubers than in Tuberosum. It was not possible to associate particular peaks in the isoform profiles with the expression of specific PME genes although in view of the greatly elevated level of PME activity in Tuberosum compared with Phureja tubers, it would appear that the isoforms encoding MICRO 2764 and PEST 2 contribute little to total PME activity.
The isolated tuber cell wall preparations gave largely similar FT-IR spectra with subtle differences. The main difference between the Phureja and Tuberosum spectra was seen in the region around 1630 cm−1 (Fig. 3A, arrow). Absorbance in this region has been associated with differences in the abundance of acid groups within pectins (McCann et al., 1997) and has been assigned to C–O stretching frequencies in acid groups. Indeed, FT-IR spectra of standard pectins with varying degrees of methylation differ in the ratio of this peak to the peak at ~1740 cm−1, which is associated with pectate acid groups acylated with methyl groups (results not shown). The monosaccharide composition of the cell walls of Phureja and Tuberosum types was similar with no significant differences in monosaccharide composition (data not shown). The degree of cell wall methylation differed between the Phureja and Tuberosum types (Fig. 3B) with the Phureja cell walls consistently releasing 1.5-fold higher amounts of methanol upon saponification. This confirms the suggestion from the FT-IR spectral data that the Tuberosum walls were less methylated.
The release of soluble pectins during de-starching of the tuber cell walls (wash B), where the material is maintained at >70 °C to ensure gelatinization, is analogous to cooking. There was a marked difference in the amount of solubilized pectin between the Phureja and the Tuberosum lines (Table 2). Note that the solubilized pectin fraction from Desiree was not retained but another Tuberosum type (variety Montrose) gave similarly low levels of solubilized pectin. The methanol content of the solubilized pectin fractions was also higher in the Phureja than Tuberosum varieties (Table 2). This suggests that the Phureja types release a higher proportion of easily solubilizable pectin than the Tuberosum types, which could be related to their textural differences.
To study the spatial distribution of pectic polysaccharides in situ, indirect immunofluorescence labelling was performed on longitudinal sections of tuber tissue using a range of well-characterized monoclonal antibodies (Bush et al., 2001). JIM5 and JIM7 recognize homogalacturonan (HG), LM5 recognizes (1-4)-β-D-galactan, LM6 recognizes (1-5)-α-L-arabinan, and 2F4 binds to homogalacturonans with degrees of methyl-esterification (DM) up to 40%. All cell walls of the cortex, perimedulla, and vascular bundles were labelled equally strongly along their length with mAb LM6. However, there was stronger labelling of the cortex for the Phureja cultivars (Fig. 4A, and insert) than for the Tuberosum cultivars (Fig. 4B, and insert). At the ultrastructural level, probing of cortical tissue sections from Phureja and Tuberosum cultivars with mAb LM6 gave gold labelling across the entire width of parenchymal cell walls when imaged using an electron microscope, but with lower abundance in the region of the middle lamella (Fig. 5A). In accordance with the light microscopic analysis, denser gold labelling seen by electron microscopy was obtained for the Phureja tissue than for the Tuberosum tissue. The amount of labelling of the vascular tissue was similar to the surrounding cells in both cultivars.
There was no apparent difference between the Phureja and Tuberosum cultivars in the pattern of labelling with any of the other antibodies tested. Although all the cells were labelled with mAb 2F4, the walls of the cortex and, in particular, the corners were more densely labelled than the cells of the perimedulla (Fig. 4C, D, and inserts). The distribution of mAb 2F4 epitopes was investigated at the ultrastructural level through probing of Phureja and Tuberosum sections from pith and perimedulla tissue. The densest gold labelling was associated with electron opaque material occluding apoplastic spaces at the cell wall junctions with sparse gold labelling along the length of parenchymal cell walls. As reported previously (Bush and McCann, 1999), gold label was concentrated in cell walls at junctions and plasmodesmata (Fig. 5B). The abundance of 2F4 epitopes at cell wall junctions was assessed by counting gold particles in randomly imaged junctions, not including particles in occluded apoplastic spaces. Although labelling was found to be higher in pith tissue than perimedulla tissue, no statistically significant differences were found between Phureja and Tuberosum cultivars (data not shown).
The labelling with mAb JIM5 was higher at the corners of the cell walls compared with the intervening walls and was more prominent in the cells of the cortex than in tissues of the perimedulla (Fig. 4E, F, and inserts). The amount of labelling of the vascular tissue was much higher than the surrounding cells resulting in the appearance of prominent vascular bundles.
The mAbs LM5 (Fig. 4G, H, and inserts) and JIM7 (data not shown) produced a uniform distribution of label, with all areas of the cell walls and all tissues showing a similar intensity of labelling. Light microscopical analysis of antibody distribution was also carried out on cooked tuber sections and showed similar results to the uncooked tissue. Again, there were no differences detected in the localization of each antibody between Phureja and Tuberosum apart from the increased labelling of the cortical tissue of Phureja tubers with LM6 antibody (data not shown).
PME activity has been associated with the texture of cooked potato in several previous studies (Ng and Waldron, 1997; Jarvis, 1999; Ross et al., 2010). However, direct evidence that links PME activity, cooked tuber textural properties, and pectin structure has, until now, been lacking. In this study, we have taken advantage of the diverse textural properties exhibited by Phureja potatoes in comparison with the Tuberosum types. Our previous comparison of the textural properties of the Phureja and Tuberosum types revealed a more rapid softening of the Phureja tuber type on boiling and that the total PME activity in Phureja tubers was 2–4-fold lower than could be extracted from the Tuberosum types (Ross et al., 2010).
The mode of action of PME has been described in some detail. Some PME isoforms can remove methyl esters from pectin in a block-wise fashion (Limberg et al., 2000). The demethylated pectin chains can then link via calcium inter-chelation into egg box structures described by Grant et al. (1973), which may lead to strengthening of the cell wall. Alternatively, demethylated pectin may be more susceptible to polygalacturonase-catalysed degradation, particularly pectin that is not demethylated in a blockwise fashion, but in a random pattern as catalysed by some plant PME isoforms (Willats et al., 2001). Thus PME activity can either strengthen or weaken cell wall structure dependent on the pattern of demethylation and the environment in which it acts. In this study, the expression pattern of four PME genes was described, with only two exhibiting clear-cut elevated expression levels in the Tuberosum types. Although two PME genes were expressed at significantly higher levels in Phureja tubers than Tuberosum, only isoforms with greater activity in Tuberosum could be detected. Thus it would appear that the isoforms encoded by the Phureja up-regulated PME genes contribute little to overall PME activity. It is conceivable that these isoforms may have a cell wall weakening effect and an activity below that detectable in the assay system employed in this study. Further transgenic studies will be required to associate particular PME genes with peaks in PME activity resolved in the chromatogram of total PME activity (Fig. 2B).
Two independent approaches were used to investigate the level of pectin methylation in cell walls from the Tuberosum and Phureja types. Firstly, FT-IR analysis revealed differences consistent with a higher degree of methylation in cell walls from the Phureja type. Secondly, direct chemical analysis of pectin methylation clearly confirmed its elevation in the Phureja type cell walls. Thus the higher levels of pectin methylation consistently measured in the Phureja samples compared with those from Tuberosum are consistent with the hypothesis that pectin methylation depends on the total PME activity extracted from the sample. The question then arose as to whether the changes in pectin methylation were reflected in any alteration in either the distribution or levels of different pectin epitopes. Previously, the low methoxyl epitopes labelled by mAbs JIM5 and 2F4 have been shown to be generated in response to PME de-esterification (Knox et al., 1990; Morvan et al., 1998). In contrast to previous observations in which JIM5 labelled all cell walls of the cortex and perimedulla equally strongly across their entire width and length (Bush and McCann, 1999), in the present study this antibody showed enhanced labelling at the cell corners along with prominent labelling of the vascular bundles. This was similar to the 2F4 (‘egg-box’ dimer) epitope, localized to the middle lamellar particularly at cell corners and the linings of intercellular spaces as has previously been observed (Bush et al., 2001). However, at the light microscopy level, there was no obvious difference between the labelling of the Phureja compared to the Tuberosum tubers with either JIM5 or 2F4, or in the labelling by mAb JIM7 for HGA pectin with a high degree of methylation. Quantification of relatively small changes in localized pectin methylation may be confounded by sample to sample variability and therefore antibody labelling may be difficult to quantify. In addition, at the electron microscopic level, although the pattern of the distribution of the low esterified HG epitope (recognized by the 2F4 antibody) was similar to previous work (Bush et al., 2001), consistent quantitative differences in the level of labelling between Phureja and Tuberosum sections were not evident.
The main difference in pectin epitope distribution observed between the Phureja and Tuberosum types was for the LM6 epitope, (1-5)-α-L-arabinan. The LM6 epitope was present at much higher levels in the cortex of Phureja tubers than in Tuberosum. The outer cell layers (including the periderm and cortex) of Phureja but not Tuberosum tubers were frequently detached during cooking (see Fig. 1) and the difference in arabinan distribution could be involved in this phenomenon. Arabinans are very flexible molecules in aqueous solution (Cros et al., 2004) and they are also very mobile in muro (Renard and Jarvis, 1999). Indeed, it has been suggested that arabinans are not structural cell wall components but they act as plasticizers and water-binding agents in the wall (Renard and Jarvis, 1999). Whether the arabinan distribution has an effect on tuber texture or is involved in the sloughing of the outer tissue layers remains to be determined.
Although the focus of this study was on the changes in pectin and pectin methyl esterase in relation to cooked tuber texture, it is likely that the texture trait is also affected by other factors. Our previous study suggested that there was little correlation between starch content and textural properties as the textural properties of the low starch pith were similar to other regions of the tuber containing higher levels of starch (Ross et al., 2010). In addition, there were no obvious correlations between cell size and shape and textural properties in the germplasm used in this study, although this has also been put forward as a factor in cooked tuber texture (Taylor et al., 2007). However, other differences in cell wall structure may well have an influence on cooked tuber texture. A genetic analysis of a potato population that segregated for tuber textural properties identified a gene encoding a cell wall protein as underlying a QTL for tuber texture (Kloosterman et al., 2010). The tyrosine and lysine-rich cell wall protein (StTLRP) was present at higher levels in individuals within the population which exhibited a firmer cooked tuber texture. Interestingly, microarray analysis of gene expression in Phureja and Tuberosum tubers revealed a significantly higher level of expression of this gene in Tuberosum tubers compared with Phureja (Ducreux et al., 2008). Further genetic and transgenic studies will be required to determine more precisely the combination of gene expression required to achieve different degrees of cooked potato texture, although the evidence presented in this study clearly identifies that modulation of pectin methylation through PME is one of the key mechanisms that impacts on the texture trait.
The financial support of the Scottish Government Rural and Environment Research and Analysis Directorate is gratefully acknowledged. We thank Susan Verrall for assistance in preparation of figures.