JXB Advance Access originally published online on April 23, 2007
Journal of Experimental Botany 2007 58(8):1915-1925; doi:10.1093/jxb/erm046
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Published by Oxford University Press [2007] on behalf of the Society for Experimental Biology.
RESEARCH PAPER |
Physiological relationships among physical, sensory, and morphological attributes of texture in tomato fruits
1INRA, UMR408 Sécurité et Qualité des Produits d'Origine Végétale, Site Agroparc, Domaine Saint-Paul, F-84914 Avignon Cedex 9, France
2INRA, Unité de Génétique et Amélioration des Fruits et Légumes, Domaine Saint-Maurice, BP94, F-84143 Montfavet Cedex, France
3INRA, UR1266 Biopolymères, Interactions Assemblages URBIA, Rue de la Géraudière, BP71627, F-44316 Nantes Cedex 3, France
4Maison de l'Alimentation, CCI d'Avignon et de Vaucluse, Laboratoire d'Evaluation Sensorielle, Technopole Agroparc, BP1201, F-84911 Avignon Cedex 9
* To whom correspondence should be addressed. E-mail: marty{at}avignon.inra.fr
Received 9 November 2006; Revised 25 January 2007 Accepted 31 January 2007
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Tomato texture is one of the critical components for the consumer's perception of fruit quality. Texture is a complex character composed of several attributes that are difficult to evaluate and which change during fruit ripening. This study investigated the texture of tomato fruits at the rheological, sensory, morphological, and genetic levels, and attempted to correlate several parameters. Analyses were performed on tomato fruits from introgressed lines carrying quantitative trait loci (QTLs) associated with texture traits localized on different chromosomes, in two genetic backgrounds. Rheological measurements were used to determine resistance to deformation and fruit elasticity. Sensory analysis was used to assess flesh firmness, juiciness, mealiness, and skin toughness. Image analysis was used to study fruit morphology and to define the cellular structure and heterogeneity of the pericarp. A highly significant correlation was observed between instrumental and sensory firmness. Moreover, correlations were also established between some texture traits and parameters of the pericarp cellular structure. Compared with QTLs detected in a previous study, the phenotypic effects expected for mealiness were confirmed in all lines, whereas, for firmness, they were not confirmed. Significant interactions between QTL and genetic background were observed for several traits. In addition, kinetic analysis showed that differences in firmness occurred from the early stages of fruit development. These results provide both a broad description of texture components and preliminary information to understand their genetic control.
Key words: Firmness, fruit, genetic background, juiciness, mealiness, near isogenic lines (NILs), quantitative trait locus (QTL), ripening, Solanum lycopersicum, texture
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Texture is one of the critical components for the consumer's perception of tomato fruit quality (Causse et al., 2003; Serrano-Megias and Lopez-Nicolas, 2006). Many traits are involved in fruit texture, mainly sensory attributes such as flesh firmness, mealiness, meltiness, juiciness, and crispness (Harker et al., 1997; Redgwell and Fischer, 2002; Szczesniak, 2002). Major changes in texture occur during fruit ripening, mainly associated with softening which considerably influences post-harvest performance, i.e. transportation, storage, shelf life and pathogen resistance (Brummell and Harpster, 2001). Softening is considered to be associated with a decline in cell wall rigidity and cell wall adhesion (Crookes and Grierson, 1983; Hallett et al., 1992). Although a number of hydrolases affecting cell wall structure have already been identified (Koch and Nevins, 1989; Fry et al., 1992), the relationships between these enzymes and the rheological properties of the fruit are still not clear (Brummell and Harpster, 2001). This is partly due to the difficulties encountered in measuring and characterizing the textural properties of the fruit.
Instrumental measurements are generally used to evaluate firmness related to the mechanical properties of fruit tissues. The most widely used measurement of fruit texture is the Magness–Taylor firmness test, which assesses the maximum force needed to compress the sample in a specific way (Abbott, 2004). This system was mainly developed to evaluate commercial quality with respect to the tactile assessment of fruit firmness by consumers at the time of purchase. Many other types of destructive or non-destructive mechanical measurements exist, including compression, puncture, tension, and vibration. Rheological measurements inform about the mechanical properties of the fruit or of a localized part of the fruit (peel, pericarp, etc.). Sensory analysis provides complementary information on human perception of whole fruits. Mastication of the fruit enables characterization of several texture attributes that are difficult to measure mechanically, such as juiciness, i.e. release of juice in the mouth; mealiness, i.e. dry, powdery, non-sticky; or skin toughness, i.e. the skin remains in the mouth. In addition, the morphology of the fruit and the combination of the diverse tissues also play a crucial role in texture (Duprat et al., 1991). In the face of such complexity, very few studies have described fruit texture as a whole (Seymour et al., 2002). Most studies have focused on a single aspect of fruit texture (Liebhard et al., 2003), and more specifically on cell wall properties (Waldron et al., 2003 Devaux et al., 2005; Brummell, 2006). In order to improve our understanding of the biophysical and biochemical events involved in changes in texture in fleshy fruits, it is essential to characterize texture at different scales (rheological, sensory, and morphological) using complementary phenotypic descriptors. Improving fruit texture also requires an understanding of the genetic control of the various traits involved. Several approaches can be used to assess each component of tomato fruit texture (Barrett et al., 1998), but very few studies have combined physical, chemical, and physiological approaches (Causse et al., 2003; Serrano-Megias and Lopez-Nicolas, 2006). Finally, genetic control of such traits in tomato is poorly understood (Fulton et al., 2000; Doganlar et al., 2002; Frary et al., 2003). The aim of this study was thus to investigate the texture of tomato fruits at the rheological, sensory, morphological, and genetic levels.
Quantitative trait locus (QTL) detection for several organoleptic quality traits (Saliba-Colombani et al., 2001; Causse et al., 2002) was performed in a previous study using the progeny [recombinant inbred lines (RILs)] from an intraspecific cross between a cherry tomato line, Cervil, and a common tomato line, Levovil. Cervil [Solanum lycopersicum var. cerasiforme (Dun.)] is characterized by small fruits with a high organoleptic potential, while Levovil (S. lycopersicum Mill.) has larger fruits with a classical taste. The main QTLs for fruit texture components have been shown to be located on chromosomes 4 and 9 as for flesh firmness and mealiness. In the present experiments, analyses were performed on fruits from isogenic lines that differed in the presence of one [QTL-near isogenic lines (NILs)] or several QTLs previously shown to be associated with texture traits (Fig. 1) and/or related to organoleptic quality. Chromosome fragments corresponding to the regions of interest were introgressed from Cervil into two texture-contrasted lines used as genetic background, one soft variety, Levovil, and one firm, VilB (Lecomte et al., 2004; Chaïb et al., 2006). Phenotypic characterization of texture in these lines was performed on red ripe tomatoes using physical and physiological measurements and sensory analysis. Some of the traits were also evaluated during the course of fruit development. Finally, as the anatomy of tomato fruit presents a large diversity of tissues, analysis of compartment composition provided information about the relative contribution of each compartment to the perception of texture. This was performed not only at the macroscopic scale by measuring pericarp thickness, volume, and number of locules, but also at the cellular scale by observing cell structure and the organization of the pericarp.
|
The objectives of this work were 3-fold: first, to determine multiple texture characteristics of tomatoes using different approaches and to correlate the different traits; secondly, to assess the genetic control of texture-related traits by analysing the effect of QTLs in different genetic backgrounds; and thirdly, to understand the physiological basis of texture, in particular fruit firmness, during fruit development.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Plant material
The experiments were performed on six introgressed lines developed in two different genetic backgrounds. The genetic backgrounds used were Levovil (hereafter referred to as Lgb) and VilB (hereafter Bgb) a modern line with large fruits, firmer than those of Levovil and with good post-harvest performance. The chromosomic fragments introgressed were provided from a cherry tomato line (Cervil) with a high organoleptic potential. These regions, located on four chromosomes, were shown to be involved in fruit quality using sensory, physical, and chemical analysis (Saliba-Colombani et al., 2001; Causse et al., 2002). A QTL for sourness was detected on chromosome 1, QTLs for sweetness, tomato aroma intensity, mealiness, and meltiness were detected on chromosome 2, a QTL for mealiness was detected on chromosome 4, QTLs for sourness, tomato aroma intensity, mealiness, meltiness, and flesh firmness were detected on the top of chromosome 9, and a QTL for pharmaceutical aroma was detected on the bottom of chromosome 9. Two introgressed lines cumulate the five chromosomic regions of interest in Lgb and Bgb (Lecomte et al., 2004) and were named Lx and Bx, respectively. Four isogenic lines carry QTLs mainly involved in fruit texture located on chromosome 4 or on chromosome 9 (Fig. 1) (Chaïb et al., 2006); they were named NIL-L4, NIL-L9, NIL-B4, and NIL-B9, with the letter corresponding to their genetic background, and a number corresponding to the QTL region. NIL-B9 is a line cumulating the top of chromosome 9 and a region from chromosome 1; no QTL for fruit texture was detected on chromosome 1.
Growth periods and design of trials
Two trials were performed during spring in a heated glasshouse. In 2004, the parental lines, the lines cumulating the five regions, and the QTL-NILs in the two genetic backgrounds were studied. Each line was represented by a single plot of six plants, grown in soil. Several types of analyses were performed on red ripe tomatoes: physical measurements, sensory profiling, fruit morphology, and cellular structure of the pericarp. In 2006, measurements were performed at four stages of fruit development following anthesis on the same lines, except NIL-B4. Five plants per line were grown in pots. The four stages of development were: 14 days post-anthesis (14 DPA), 35 DPA for mature green tomatoes, breaker (B), and red ripe tomatoes (RR). To harvest the two first stages, flowers were tagged at anthesis. For the two other stages, fruits were harvested based on fruit colour. Physical traits and ethylene production were measured.
Physical and physiological measurements
In 2004, red ripe fruits were harvested in bulk on the six plants twice a week for 6 weeks. At the first harvest each week, seven fruits from each batch were taken randomly in each plot in order to obtain six sets per line (six independent repetitions). This strategy was preferred to harvesting individual plants as there is more variation between weeks rather than among plants (unpublished results). A total of 42 fruits per plot were evaluated for physical traits. Fruit-by-fruit evaluations were performed for fruit weight (FW), colour, firmness (FIRd, FIRp) and ‘total elasticity’ (ELA). Fruit colour was measured using the CIE L*a*b (lightness/green-to-red scale/blue-to-yellow scale) colour space with a Minolta chromameter CM-1000R (Minolta, Ramsey, NJ, USA). Fruit firmness was measured with a Durofel® (FIRd) and with a Penelaup® penetrometer (FIRp). For FIRd, a probe was applied at two points on the fruit equator, the displacement of the probe when compressing the fruit was recorded, and the average of the two measures was used. The Penelaup compressed the fruit between two steel plates. Force was applied by the upper plate to deform the fruit by 3% of its initial diameter, and the resulting deformation and force were measured. FIRp is the mechanical stress needed to deform the fruit (force per surface unit). Total elasticity, ELA, was the slope of the force/displacement curve. As fruits from the Cervil parental line are small, FIRd, FIRp, and ELA were not measured.
In 2006, measurements of FW, fruit colour, FIRp, ELA, and ethylene production were performed during fruit development on an average of 30 fruits per line and stage, from several trusses of the five plants. Ethylene production was measured by gas chromatography (IGC 121 FL) after 1.5 h of confinement of one fruit in a jar (Chambroy et al., 1995).
Sensory profiling
Red ripe tomatoes were harvested in the morning of the day they were to be tasted. Homogeneous fruit samples were selected and stored at 20 °C in an air-conditioned room. The sensory panel was composed of 15 judges, who had previously been trained in the quantitative description of tomato attributes according to selection trials based on French norms (ISO8586-1, AFNOR V09-003). For each line, fruits were tasted twice by each judge, giving 30 scores per genotype. Fifteen sessions were held in a sensory analysis laboratory (AFNOR norm V09-105), on 2 d per week, and eight fruits were tasted by each judge on each occasion. The attributes chosen to describe fruit texture were firmness, juiciness, mealiness, and skin toughness. Each was scored on a 10-point scale.
Macroscopic study of fruit morphology
The internal structure of the fruit was investigated on red ripe tomatoes. Five fruits were analysed per line. An approximately 0.5 cm thick transversal section was cut in the tomatoes, and locules were drained by removing gel and seeds with a vacuum pump. The section was placed on a dark support and an image with a resolution of 768x574 pixels was acquired using a CCD camera. Three types of calibration were used to observe the minimum area in relation to the size of the fruit: calibration 0 for a field of view 51.48 mm in width, calibration 1 for 86.16 mm, and calibration 2 for 130.20 mm. A computer program developed by F Duprat at INRA-Montfavet was used to measure automatically the locule number (LONB), the percentage of locule area (%LO), and the relative thickness of the pericarp (PERIr) in relation to the total area.
Analysis of the cellular structure of the pericarp
Analysis of the cellular structure of the pericarp was performed on red ripe tomatoes. As the largest parenchyma tomato cells can be >500 µm long, macroscopic vision was used to observe a representative number of cells and their arrangement within the pericarp. For each genotype, eight fruits were analysed and three samples were collected in the equatorial region of the tomato using a circular punch 1 cm in diameter. Sections were cut in the middle of the sample using a vibrating microtome (Microm HM 650 V). The thickness of the section was set at 200 µm. Images were acquired as described in Devaux et al. (2005). For each image, the field of view was 10.7x14.4 mm with a pixel size of 18.6x18.6 µm.
Image analysis was performed as follows. The pericarp was segmented in the image. Cells were closed by a reconstruction operator using a squared structuring element of size 50 and a connectivity of 4 (Soille, 2003). The pericarps were extracted by thresholding the resulting image using a fixed grey level value of 50. A measure of the pericarp thickness was achieved as the average length of the 100 central columns of the images. The final region of interest within pericarp was obtained after removing the cuticle by an erosion using a squared structuring element of size 20. Grey levels within the pericarp region were standardized by histogram equalization. Information concerning cell size distribution was extracted within the pericarp by mathematical morphology using a global technique based on grey level granulometry (Soille, 2003). Such techniques can be applied when objects cannot easily be segmented, which was the case for the scale and resolution used in the present work. The method consists of applying a basic procedure called ‘closing’. Closing is based on the comparison of images with structuring elements of a given size and shape. Only particles larger than the structuring element remain after closing. A size distribution can be obtained by applying a sequence of closings with structuring elements of increasing size. After each closing step, the sum of grey levels is computed. This sum depends both on the objects remaining and on their grey levels. Variations in grey levels between each step can be plotted as a function of the size of the structuring elements. The resulting curves are called granulometric curves. In the present work, 40 successive openings were applied that used linear structuring elements of lengths ranging from 56 µm to 1505 µm by increments of 37.2 µm. Linear elements were chosen to assess cell elongation in the direction perpendicular to the skin. The orientation of the element was set at 90°. Image analysis procedures were developed using Aphelion 3.2g software (ADCIS France).
Statistical tests
For sensory analysis, an analysis of variance (ANOVA; genotype, judge, and their interaction) test was performed for each attribute to evaluate panellist performance and to eliminate judges and attributes showing extreme variance. For physical measurements and sensory profiling, data from the QTL-NILs were first compared by a two-factor ANOVA to test differences between the introgressed chromosomic fragment, the genetic background, and their interaction (data not shown). All the genotypes were then ranked using a Duncan test. The Cervil line was not included in these analyses because it was too different from the other lines. These statistical analyses were performed using Splus software (Splus, 1993). Pearson coefficients of correlations were calculated between physical measurements, sensory analysis, fruit morphology, and cellular structure of pericarp data, measured on red ripe tomatoes. Analysis of data concerning the cellular structure of pericarp was performed using principal component analysis as described in Devaux et al. (2005).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Physical and sensory analysis of texture related traits
To assess fruit texture, physical and sensory analysis were performed in various genotypes, in two genetic backgrounds, that differed in one or five regions of the tomato genome involved in fruit quality and particularly in texture: Levovil, NIL-L4, NIL-L9, Lx, VilB, NIL-B4, NIL-B9, and Bx. Three components were measured to describe physical texture (Fig. 2): fruit elasticity (ELA) and firmness using two instrumental methods (FIRp and FIRd). Firmness, mealiness, skin toughness, and juiciness described sensory attributes of texture (Fig. 3).
|
|
The difference in firmness between VilB and Levovil was confirmed by both physical measurements and sensory analysis. VilB showed higher fruit elasticity than Levovil, while Levovil was juicier than VilB. No difference was observed for mealiness or skin toughness. The individual effects of the QTLs of chromosome 4 on texture parameters were generally not significant compared with the parental lines. Significant differences were observed only for NIL-L4, which was juicier and less mealy than Levovil, and for fruit elasticity for NIL-B4, which was slightly less elastic than VilB. The QTLs of chromosome 9 showed more effects on texture traits. Differences were observed between NIL-B9 and VilB in firmness, mealiness, and skin toughness, the QTL-NIL being less mealy and firmer than the parental line (at physical and sensory levels). Similarly, NIL-L9 was firmer and less mealy than Levovil, but less significantly. There were notable differences in the effects of the introgression of several QTLs in Lx and Bx. These lines presented less elasticity and mealiness compared with the corresponding parental lines. The firmness attribute evolved differently depending on the genetic background: Lx was firmer than Levovil, while Bx was less firm than VilB. Only NIL-Bx was juicier than the parental line. For Lx and Bx, no significant difference was noted for skin toughness.
Morphological analysis at macroscopic and cellular levels
The macroscopic study of fruit morphology was conducted by analysing transverse sections of tomatoes at the red ripe stage (Fig. 4). Locule number (LONB), the percentage of locule area in relation to total area (%LO), and the relative thickness of the pericarp (PERIr) were measured and means were compared. For LONB, two groups were detected, one corresponding to lines with many locules including Levovil, NIL-L4, and NIL-L9, and one corresponding to a mean of two locules for Lx and all the lines in Bgb (VilB, NIL-B4, NIL-B9, and Bx). For %LO, Levovil had the highest value, and NIL-L4, Lx, and VilB had the lowest means. The other lines were situated between these two groups and very close, except for NIL-B4 which was closer to the group containing VilB. The relative thickness of the pericarp, PERIr, showed less variation and was distributed in three groups. The lines with the thickest pericarp were VilB and NIL-B4, while NIL-L4 and NIL-L9 have the thinnest. Means for Levovil, Lx, NIL-B9, and Bx were situated between the two groups.
|
The distribution of different sized and shaped cells in the pericarp of red ripe tomatoes was assessed by macroscopic image analysis (Fig. 5). In all images, white spots were observed in the middle of the pericarp that corresponded to vascular bundles. In all lines, smaller cells were observed under the cuticle, which is the region where cell divisions occur (Cheniclet et al., 2005). Cell morphology differed depending on the parental line: Cervil displayed small cells (data no shown), while cells were larger in Lgb and more elongated in Bgb lines. In addition, elongation occurred perpendicularly to the cuticle. Image analysis was therefore performed to measure cell length observed at an angle of 90° to the fruit skin (Fig. 6A). Granulometric curves exhibited one large peak corresponding to the length at which the majority of cells were removed by image transformation. The peaks were located between 100 µm and 500 µm: Lx, Bx, NIL-B9, and NIL-L9 lines showed smaller cells than NIL-L4, NIL-B4, and the parental lines. Principal component analysis was applied to the set of granulometric curves computed for all macroscopic images (Fig. 6B). The first two principal components took into account 67% and 19% of the total variance, respectively. The first principal component (pc1) described the average length with the small cell size on the left and the larger cell size on the right. NIL-L4 and NIL-B4 were not distinct from their parental lines, and NIL-B9, NIL-L9, Lx, and Bx cells were shorter. The second principal component (pc2) highlighted the heterogeneity of cell distribution with an average cell length ranging between 200 µm and 400 µm located at the top of the map (positive scores) and the occurrence of simultaneously small and elongated cells located at the bottom of the map (negative scores). The two genetic backgrounds were separated by this component. VilB, NIL-B4, NIL-B, and Bx were mostly located on the negative side, while Levovil, NIL-L4, NIL-L9, and Lx were mostly located on the positive side.
|
|
Evolution of texture traits during fruit development
In order to define when differences in texture behaviour occur, fruits were characterized during the course of fruit development. Fruits were collected at four stages of development: 14 DPA corresponding to the growth stage, 35 DPA corresponding to mature green fruits, at the breaker stage (B) when fruits began ripening, and at the red ripe stage (RR). Fruits were characterized by colour, weight, ethylene production, and firmness (Fig. 7).
|
No difference in fruit weight (Fig. 7A) was found between lines at the 14 DPA stage, whereas there was a significant difference in fruit weight between the two groups from 35 DPA until the ripe stage. The first group included the parental lines and NIL-L9, and showed higher weight (
100 g at 35 DPA and 140 g at the ripe stage) than the second group (50 g) containing NIL-B9 and lines carrying the five QTLs (Lx and Bx). The colour of fruits (Fig. 7B) in all lines was similar during the developmental phases as long as fruits remained green, up to 35 DPA (a around –13). A significant difference was observed at the breaker stage (a from 9 to 15) when fruits began to ripen and to change from green to orange. At the red ripe stage, a very slight but non-significant difference was observed between lines. Ethylene production was only measured on fruits with Lgb. As expected, ethylene (Fig. 7C) was produced at the breaker stage when fruits began to ripen, and there was a significant difference between lines. The lines with the five introgressed QTLs or only with the region on chromosome 9 produced more ethylene than the parental line. Concerning fruit firmness, in Lgb, only Lx displayed higher FIRp values throughout fruit development, whereas Levovil and NIL-L9 were almost identical. Very few differences between lines were observed in Bgb. Bx displayed the highest values in the green stages, whereas NIL-B9 displayed the highest values in the breaker and red ripe stages. Two types of kinetic pattern were distinguished. Lx and NIL-B9 curves decreased regularly, while for the other lines loss of firmness was greatest between 35 DPA and the breaker stage.
Correlations between texture-related traits
Correlations between traits measured at physical and sensory levels were calculated (Table 1). The parameters pc1 and pc2 of the principal component analysis of the granulometric curves described above were used to take the cellular structure of pericarp into account.
|
For physical traits, positive correlations were highly significant between the texture parameters FIRd and FIRp, FIRd and ELA, and ELA and PERIa. Regarding cellular structure parameters, the cell size, pc1, was positively correlated with PERIa, with FW, and with the colour component a. For the heterogeneity of cell distribution, negative relationships were observed between pc2 and both FIRp and FIRd, and a positive correlation with LONB. The colour component a was positively correlated with ELA and PERIa, as were FW and LONB. For sensory traits, correlations were positively significant between firmness and skin toughness, and negatively significant between firmness and juiciness. Concerning the links between physical and sensory traits, a highly significant positive correlation was observed between firmness and FIRp, and this was slightly lower with FIRd. The same relationship was detected between skin toughness and FIRp/FIRd. Firmness and LONB showed a negative correlation, as did firmness and pc2. Juiciness was negatively correlated with FIRp and to PERIr. Mealiness was correlated positively with FW, with a, with pc1, and with PERIa. A negative correlation was observed between skin toughness and pc2.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Many studies describe tomato fruit texture (Barrett et al., 1998), but very few combine such different approaches as rheological, sensory, chemical, and physiological measurements, and even fewer analyse genetic determinism (Fulton et al., 2000; Causse et al., 2002; Doganlar et al., 2002; Frary et al., 2003). In this study, complementary methods were used to describe several components of tomato fruit texture and to identify instrumental measurements associated with sensory texture attributes. The physical measurements provided information about the deformation resistance of whole fruit and fruit elasticity, while sensory analysis provided detailed information about internal perception of texture through attributes such as flesh firmness, juiciness, mealiness, or skin toughness. Image analysis provided information about the morphology of the fruit and the structure and heterogeneity of the pericarp cells. In addition, the genetic control of these traits was assessed by studying the effects of texture-related QTLs and their interaction with genetic background. Indeed, different chromosome regions were previously identified as being involved in fruit firmness and mealiness (Saliba-Colombani et al., 2001; Causse et al., 2002) and were introgressed in two texture-contrasted parental lines, VilB and Levovil. In this study, physical and sensory evaluations showed that ViLB and Levovil differed for several texture-related traits and confirmed the interest in their use to study the impact of genetic background on the effect of the previously described QTLs.
QTLs effects on texture traits
Although QTLs for mealiness, firmness, and fruit elasticity were detected on chromosomes 4 and 9 in the RIL population used for QTL detection, derived from the cross between Cervil and Levovil (Saliba-Colombani et al., 2001; Causse et al., 2002), the individual effects of these QTLs on texture parameters were generally not significant in QTL-NILs compared with the parental lines. For sensory firmness, the major QTL was detected in RILs on chromosome 9 (Fig. 1) and another QTL was detected on chromosome 2 (in the same region as for mealiness), the Cervil alleles providing the highest values for these traits. For instrumental firmness, the Cervil and the Levovil alleles provided the higher values of the trait for the QTL on chromosome 4 and on chromosome 9, respectively. The inconsistency of QTL effects in QTL-NILs compared with RILs was particularly observed in Lgb. Nevertheless, the expected QTLs based on RIL data were detected in the lines cumulating the five regions of interest, Lx and Bx, but with an opposite effect for firmness in both lines. New QTLs for juiciness and for skin toughness were detected on chromosomes 4 and 9 in QTL-NILs, respectively, but these effects were not observed in Lx and Bx. The variations in the QTL effects observed in Lx and Bx could be explained by the fact that these lines cumulate several QTLs for the same trait, sometimes with opposite effects depending on the chromosomic region. Such inconsistency in the behaviour of the introgressed lines could be due to interactions between QTLs but also between different parts of the same region since they represent large chromosome fragments (from 20 cM to 50 cM). Concerning fruit morphology, no significant QTL effect was observed among introgressed lines with respect to the locule area or the relative thickness of the pericarp. A difference was observed for the locule number between Levovil and Lx in relation to the presence of the region of chromosome 2 where a QTL with large effect for this trait was detected in RILs. In addition, the classification of genotypes by variance analysis for the cellular structure and organization of pericarp (pc1 associated with pc2) was similar to that obtained for firmness (data not shown), suggesting that the pericarp of firm tomatoes is composed of small cells under the cuticle. Tomatoes with the thickest pericarp showed more elongated cells.
Genetic background and environment effects on texture traits
This study demonstrated the impact of the genetic background and/or of the environment on the expression of texture-related QTLs (at the physical and sensory levels) and possible interactions between them. Differences in the effects of QTLs as a function of genetic background were observed for all the characters studied: firmness and skin toughness for NIL-B9, juiciness for NIL-L4 and Bx, and mealiness for all the QTL-NILs. In a previous study (Chaïb et al., 2006), the individual effects of QTLs for firmness in both Levovil and VilB genetic backgrounds were not observed. This lack of consistency of QTLs on firmness was in part explained by the different instrumental methods that were used in the determination of the firmness: it was measured with a Penelaup® (FIRp) in the RIL population while a Durofel® (FIRd) was used in the evaluation of the additive effect in the QTL-NILs. In this study, Durofel® measurements were highly correlated to Penelaup® data, confirming the inconsistency of the individual effect of QTLs for firmness, suggesting allelic differences at the loci of interest. A significant individual effect on firmness was observed for the QTL located on chromosome 9 in the VilB genetic background, which was not detected by Chaïb et al. (2006). Such variations confirm that firmness interacts strongly with environmental conditions such as year of cultivation or location (Bernacchi et al., 1998; Chaïb et al., 2006).
Correlations between texture traits
The statistical correlations between physical, sensory, and morphological results revealed that several parameters are linked. As a limited number of lines in two contrasted genetic backgrounds were used to calculate the Pearson coefficient of correlation, the impact of the genetic background was tested. Differences in the significant correlations detected were studied using Lgb lines or Bgb lines separately (data not shown). Consequently, only correlations that were confirmed graphically within each genetic background are detailed hereafter. A highly positive correlation was revealed between sensory and physical firmness (measured by penetrometry); this confirmed previous results observed during the detection of QTLs in the RIL population (Causse et al., 2002). However, the negative correlation between juiciness and pericarp elasticity observed in the first study was not reproduced here. The present work showed a negative relationship between firmness and juiciness, locule number, and the heterogeneity of cell distribution in the pericarp, and a positive correlation between firmness and skin toughness. These results suggested that the firmest fruits were less juicy with a tough skin, had a low locule number, and a pericarp with a heterogeneous cell distribution. Many of these traits associated with firmness were also correlated between themselves, as were skin toughness and the presence of small cells under the cuticle. However, it has been demonstrated that locule number is not associated with firmness since firm tomatoes with a high locule number exist (data not shown). Finally, a positive correlation was shown between mealiness and absolute pericarp thickness and between mealiness and cell size, suggesting that mealy tomatoes had a thicker pericarp with elongated cells.
Evolution of texture during fruit development
In order to understand when changes in texture behaviour occur, measurements of firmness were analysed throughout fruit development. It appeared that differences in firmness between lines occurred from the early stages of development. Parameters such as fruit weight, colour, and ethylene production enabled four physiological stages to be defined. As expected, differences in fruit weight appeared at 35 DPA, corresponding to the cell elongation stage. Fruit weight was lower in relation to the number of Cervil alleles unfavourable for this trait. The breaker stage, corresponding to the beginning of the ripening stage, was defined by the change from green to orange (lycopene synthesis) and the beginning of ethylene production. It is interesting to note that ethylene emission was high in Lx which had the firmest fruit, but it is difficult to relate to a particular physiological mechanism. Ethylene is known to regulate the expression of several genes connected with ripening, including fruit texture. These genes are principally involved in cell wall degradation, such as polygalacturonase, expansins, or pectin methylesterase (Alexander and Grierson, 2002). Finally, as expected, firmness decreased during ripening associated with softening (Brummell, 2006). Nevertheless, the differences in firmness observed between lines appeared as early as the cell expansion stage (14 DPA) and subsequently remained constant. Thus, differences in firmness observed between red ripe tomatoes could result not only from cell wall degradation during ripening but also from the cellular organization of the pericarp that occurs early during fruit development.
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
This study enabled (i) the characterization of several texture traits of tomato fruit and correlations between some of them; (ii) the identification of genetic and environmental factors that influence the expression of some of these traits; and finally (iii) the identification of the stage at which differences in firmness occur during fruit development. Sensory and instrumental evaluations of firmness were shown to be highly correlated, whereas only sensory analysis was available for other attributes such as mealiness, juiciness, and skin toughness (expensive experiments for routine measurements). For the parameters studied, correlations were established between (i) firmness and a heterogeneous cell distribution in the pericarp; (ii) skin toughness and the presence of small cells under the cuticle; and (iii) mealiness and a thick pericarp composed of elongated cells. Moreover, analysis of the cellular structure of the pericarp and measurements of firmness statistically distinguished lines in the same way. Genetic control of firmness is complex, as are interactions between QTLs and genetic background, and interactions with the environment were observed. QTL effects for mealiness appeared to be more consistent, even though the influence of the genetic background was not negligible. In addition, while most studies on fruit texture showed that changes in firmness result from softening and cell wall degradation (Crookes and Grierson, 1983; Hallett et al., 1992), it was shown that these differences could also occur at early stages in fruit development. This could be linked to a previous relationship between firmness and the cellular structure of pericarp, so it will be interesting to complete these observations by macroscopic image analysis of early stages of development. Preliminary results suggest very few differences in pericarp pattern at anthesis, whereas marked variations are observed at the breaker stage as described in Cheniclet et al. (2005). To complete the study of cellular organization, another approach would be to take into account the number of cells as a function of their shape and of their location in the tissue.
|
Acknowledgements |
|---|
Many thanks to Emmanuel Botton for taking care of the plants, to Marielle Boge, Philippe Duffé, Esther Pelpoir, and Cecile Garchery for their technical support, to Barbara Gouble and Sylvie Bureau for managing the physical evaluations, to Guy Jacquemin and Emilie Terrel for managing sensory profiling, and to François Duprat and Abdelkrim Sadoudi, respectively, for developing the program for image analysis of transverse tomato sections and for computing the macroscopic data. This project is partially funded by a trilateral co-operation initiative between France, Spain, and Germany on plant genome research. The experiments comply with the current French laws.
|
Abbreviations |
|---|
Bgb, VilB genetic background; DPA, days post-anthesis; ELA, total elasticity; FIRd, firmness measured with Durofel; FIRp, firmness measured with Penelaup; FW, fruit weight; Lgb, Levovil genetic background; LONB, locule number; %LO, percentage of locule area in relation to the total area of a transverse section of tomato; NIL, near-isogenic line; pc1, principal component 1 (cell size); pc2, principal component 2 (heterogeneity of cell distribution); PERIa, absolute thickness of the pericarp; PERIr, relative thickness of the pericarp; RIL, recombinant inbred line.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Abbott JA. Textural quality assessment for fresh fruits and vegetables. Quality of Fresh and Processed Foods (2004) 542:265–279.[Web of Science]
Alexander L, Grierson D. Ethylene biosynthesis and action in tomato: a model for climacteric fruit ripening. Journal of Experimental Botany (2002) 53:2039–2055.
Barrett DM, Garcia E, Wayne JE. Textural modification of processing tomatoes. CRC Critical Reviews in Food Science and Nutrition (1998) 38:173–258.[CrossRef]
Bernacchi D, Beck-Bunn T, Emmatty D, et al. Advanced backcross QTL analysis of tomato. II. Evaluation of near-isogenic lines carrying single-donor introgressions for desirable wild QTL-alleles derived from Lycopersicon hirsutum and L. pimpinellifolium. Theoretical and Applied Genetics (1998) 97:170–180.[CrossRef][Web of Science]
Brummell DA. Cell wall disassembly in ripening fruit. Functional Plant Biology (2006) 33:103–119.[CrossRef][Web of Science]
Brummell DA, Harpster MH. Cell wall metabolism in fruit softening and quality and its manipulation in transgenic plants. Plant Molecular Biology (2001) 47:311–340.[CrossRef][Web of Science][Medline]
Causse M, Buret M, Robini K, Verschave P. Inheritance of nutritional and sensory quality traits in fresh market tomato and relation to consumer preferences. Journal of Food Science (2003) 68:2342–2350.[CrossRef][Web of Science]
Causse M, Saliba-Colombani V, Lecomte L, Duffe P, Rousselle P, Buret M. QTL analysis of fruit quality in fresh market tomato: a few chromosome regions control the variation of sensory and instrumental traits. Journal of Experimental Botany (2002) 53:2089–2098.
Chaïb J, Lecomte L, Buret M, Causse M. Stability over genetic backgrounds, generations and years of quantitative trait locus (QTLs) for organoleptic quality in tomato. Theoretical and Applied Genetics (2006) 112:934–944.[CrossRef][Web of Science][Medline]
Chambroy Y, Souty M, Jacquemin G, Gomez RM, Audergon JM. Research on the suitability of modified atmosphere packaging for shelf-life and quality improvement of apricot fruit. Acta Horticulturae (1995) 384:633–638.
Cheniclet C, Rong WY, Causse M, Frangne N, Bolling L, Carde JP, Renaudin JP. Cell expansion and endoreduplication show a large genetic variability in pericarp and contribute strongly to tomato fruit growth. Plant Physiology (2005) 139:1984–1994.
Crookes PR, Grierson D. Ultrastructure of tomato fruit ripening and the role of polygalacturonase isoenzymes in cell wall degradation. Plant Physiology (1983) 72:1088–1093.
Devaux MF, Barakat A, Robert P, Bouchet B, Guillon F, Navez B, Lahaye M. Mechanical breakdown and cell wall structure of mealy tomato pericarp tissue. Postharvest Biology and Technology (2005) 37:209–221.[CrossRef][Web of Science]
Doganlar S, Frary A, Ku HM, Tanksley SD. Mapping quantitative trait loci in inbred backcross lines of Lycopersicon pimpinellifolium (LA1589). Genome (2002) 45:1189–1202.[Medline]
Duprat F, Roudot F, Grotte-Nicolas M, Roudot AC. Fruit inhomogeneity. Sciences des Aliments (1991) 11:613–626.[Web of Science]
Frary A, Doganlar S, Frampton A, Fulton T, Uhlig J, Yates H, Tanksley S. Fine mapping of quantitative trait loci for improved fruit characteristics from Lycopersicon chmielewskii chromosome 1. Genome (2003) 46:235–243.[Medline]
Fry SC, Smith RC, Renwick KF, Martin DJ, Hodge SK, Matthews KJ. Xyloglucan endotransglycosylase, a new wall-loosening enzyme-activity from plants. Biochemical Journal (1992) 282:821–828.[Web of Science][Medline]
Fulton TM, Grandillo S, Beck-Bunn T, Fridman E, Frampton A, Lopez J, Petiard V, Uhlig J, Zamir D, Tanksley SD. Advanced backcross QTL analysis of a Lycopersicon esculentumxLycopersicon parviflorum cross. Theoretical and Applied Genetics (2000) 100:1025–1042.[CrossRef][Web of Science]
Hallett IC, MacRae EA, Wegrzyn TF. Changes in kiwifruit cell wall ultrastructure and cell packing during postharvest ripening. International Journal of Plant Sciences (1992) 153:49–60.[CrossRef][Web of Science]
Harker FR, Redgwell RJ, Hallett IC, Murray SH, Carter G. Texture of fresh fruit. Horticultural Reviews (1997) 20:121–224.
Koch JL, Nevins DJ. Tomato fruit cell wall. I. Use of purified tomato polygalacturonase and pectinmethylesterase to identify developmental changes in pectins. Plant Physiology (1989) 91:816–822.
Lecomte L, Duffé P, Buret M, Servin B, Hospital F, Causse M. Marker-assisted introgression of five QTLs controlling fruit quality traits into three tomato lines revealed interactions between QTLs and genetic backgrounds. Theoretical and Applied Genetics (2004) 109:658–668.[Web of Science][Medline]
Liebhard R, Kellerhals M, Pfammatter W, Jertmini M, Gessler C. Mapping quantitative physiological traits in apple (Malusxdomestica Borkh.). Plant Molecular Biology (2003) 52:511–526.[CrossRef][Web of Science][Medline]
Redgwell RJ, Fischer M. Fruit texture, cell wall metabolism and consumer perceptions. In: Fruit quality and its biological basis—Knee M, ed. (2002) Oxford: Blackwell. 46–88.
Saliba-Colombani V, Causse M, Langlois D, Philouze J, Buret M. Genetic analysis of organoleptic quality in fresh market tomato. 1. Mapping QTLs for physical and chemical traits. Theoretical and Applied Genetics (2001) 102:259–272.[CrossRef][Web of Science]
Serrano-Megias M, Lopez-Nicolas JM. Application of agglomerative hierarchical clustering to identify consumer tomato preferences: influence of physicochemical and sensory characteristics on consumer response. Journal of the Science of Food and Agriculture (2006) 86:493–499.[CrossRef][Web of Science]
Seymour GB, Manning K, Eriksson EM, Popovich AH, King GJ. Genetic identification and genomic organization of factors affecting fruit texture. Journal of Experimental Botany (2002) 53:2065–2071.
Soille P. Morphological image analysis. In: Principles and applications (2003) Berlin: Springer-Verlag.
Splus. Splus guide to statistical and mathematical analyses (1993) Seattle, Washington.
Szczesniak AS. Texture is a sensory property. Food Quality and Preference (2002) 13:215–225.[CrossRef][Web of Science]
Waldron KW, Parker ML, Smith AC. Plant cell walls and food quality. Comprehensive Reviews in Food Science and Food Safety (2003) 2:101–119.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. I. Zanor, J.-L. Rambla, J. Chaib, A. Steppa, A. Medina, A. Granell, A. R. Fernie, and M. Causse Metabolic characterization of loci affecting sensory attributes in tomato allows an assessment of the influence of the levels of primary metabolites and volatile organic contents J. Exp. Bot., May 1, 2009; 60(7): 2139 - 2154. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Mounet, A. Moing, V. Garcia, J. Petit, M. Maucourt, C. Deborde, S. Bernillon, G. Le Gall, I. Colquhoun, M. Defernez, et al. Gene and Metabolite Regulatory Network Analysis of Early Developing Fruit Tissues Highlights New Candidate Genes for the Control of Tomato Fruit Composition and Development Plant Physiology, March 1, 2009; 149(3): 1505 - 1528. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Bertin, M. Causse, B. Brunel, D. Tricon, and M. Genard Identification of growth processes involved in QTLs for tomato fruit size and composition J. Exp. Bot., January 1, 2009; 60(1): 237 - 248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Paran and E. van der Knaap Genetic and molecular regulation of fruit and plant domestication traits in tomato and pepper J. Exp. Bot., November 23, 2007; (2007) erm257v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||