Cell expansion and endoreduplication show a large genetic variability in pericarp and contribute strongly to tomato fruit growth

Abstract

Postanthesis growth of tomato (Solanum lycopersicon) as of many types of fruit relies on cell division and cell expansion, so that some of the largest cells to be found in plants occur in fleshy fruit. Endoreduplication is known to occur in such materials, which suggests its involvement in cell expansion, although no data have demonstrated this hypothesis as yet. We have analyzed pattern formation, cell size, and ploidy in tomato fruit pericarp. A first set of data was collected in one cherry tomato line throughout fruit development. A second set of data was obtained from 20 tomato lines displaying a large weight range in fruit, which were compared as ovaries at anthesis and as fully grown fruit at breaker stage. A remarkable conservation of pericarp pattern, including cell layer number and cell size, is observed in all of the 20 tomato lines at anthesis, whereas large variations of growth occur afterward. A strong, positive correlation, combining development and genetic diversity, is demonstrated between mean cell size and ploidy, which holds for mean cell diameters from 10 to 350 microm (i.e. a 32,000-times volume variation) and for mean ploidy levels from 3 to 80 C. Fruit weight appears also significantly correlated with cell size and ploidy. These data provide a framework of pericarp patterning and growth. They strongly suggest the quantitative importance of polyploidy-associated cell expansion as a determinant of fruit weight in tomato.

Figure 1.

Structure and development of pericarp after anthesis. A to C, Cross sections of ovary wall at anthesis for Wva106 (A), Ferum 26 (B), and Grosse de Gros (C) lines. D, Pericarp cross section of 4-DPA Wva106 fruit, showing anticlinal (vertical arrows), periclinal (horizontal arrows), and randomly oriented (arrowheads) cell divisions in inner epidermal and subepidermal layers and in central cells. E, Enlarged portion of the outer pericarp of the same fruit as in D showing anticlinal divisions (vertical arrows) in outer epidermal layer and periclinal divisions (horizontal arrows) in outer subepidermal layer. oe, Outer epidermis; ie, inner epidermis; vb, vascular bundle. Scale bars: 50 μm (A–D) or 20 μm (E).

Figure 2.

Cellular parameters of pericarp at anthesis and breaker stages. Tomato plants from 20 lines were grown in a greenhouse (Avignon). For each line, three to seven fruits were sampled for structural and cytological analyses at the time of anthesis (white columns) and at the beginning of breaker stage (gray columns). A, Pericarp thickness at anthesis (mm, left) and breaker (mm, right) stages. B, Mean cross-sectional area of one pericarp cell at anthesis (μm², left) and breaker (mm², right) stages. C, Number of cell layers across pericarp. D, MCV of pericarp cells at breaker stage.

Figure 3.

Kinetics of pericarp growth in the Wva106 line. Six Wva106 plants were grown in a greenhouse (Bordeaux). At each developmental stage, three to five fruits (total no. 102) were selected for measurement (mean ± sd). A, Fruit diameter. B, Pericarp thickness. C, Mean cross-sectional area of one pericarp cell. D, Number of cell layers across pericarp. E, MCV of pericarp cells, as described in Figure 5. Insets in C and E show enlarged kinetics for the 0 to 8 DPA period. Arrows show the breaker stage.

Figure 4.

Pericarp structure at breaker stage. A to C, Cross sections of fruit (left) and pericarp (right) at breaker stage are shown for Grosse de Gros (A), Ferum 26 (B), and Wva106 (C) lines. D, Enlargement of Wva106 outer pericarp at the same magnification as Figure 1E; these show that epidermal and subepidermal cells have mainly enlarged tangentially since anthesis. The rectangle in A, right, is an example of an area within which mean cell size has been estimated, as detailed in “Materials and Methods.” Bars: 1 cm (A–C, left sections), 1 mm (A–C, right sections), and 20 μm (D).

Figure 5.

Endoreduplication in tomato fruit. Ploidy has been analyzed by flow cytometry in pericarp (A and B), locular gel (C and D), columella (E and F), and sepals (G and H) of Bubjekosoko fruit grown in Bordeaux and harvested at breaker stage. Left sections show histograms of one representative fruit. Right sections (gray bars) show the frequency of each C value class as calculated from measurements performed on 29, 20, 19, and 12 fruits for B, D, F, and H, respectively. In B, the results from another set of three fruits from plants of the same line grown in Avignon are also shown (white bars, MCV').

Figure 6.

Developmental kinetics of C-value classes in Wva106 pericarp. Data are from the same experiment as in Figure 3. A, Evolution of 2-C (white triangles, dashed line), 4-C (white squares, dashed line), 8-C (black triangles, solid line), and 16-C (black squares, solid line) classes in fruit pericarp. B, Evolution of 32-C (white circles, dashed line), 64-C (white diamonds, dashed line), 128-C (black circles, solid line), and 256 C (black diamonds, solid line) classes in fruit pericarp.

Figure 7.

Ploidy distribution in pericarp of four tomato lines. Data are mean ± sd of the frequency of each C-value class of three to seven fruits at breaker stage for four selected tomato lines, two lines with a low MCV value, Wva106 (white bars) and Grosse de Gros (light-gray bars), and two lines with a high MCV, Montfavet 136-11 (dark-gray bars) and Ferum 26 (black bars). Growth conditions and line characteristics are detailed in Figure 2.

Figure 8.

Relationship between cell size and ploidy. Each point shows the MCV and cell diameter in pericarp of a single fruit. The mean cell diameter was calculated from mean cross-sectional cell areas estimated as described in “Materials and Methods,” by approximating cells as spheres. White symbols in A and C report data from the same experiment as in Figure 3, where fruit development was analyzed in Wva106 line from anthesis to ripening (n = 102 fruits). Black symbols in B and C report data from the same experiment as in Figure 2, where 20 tomato lines were compared at breaker stage (all lines are represented by the same symbol; n = 81 fruits). Dashed lines show the polynomial regression curves for each set of data. R² = 0.93 and 0.69 for A and B, respectively. The equation in C is y = 0.05x² + 8.9x − 7.0, R² = 0.96, n = 183.

Figure 9.

Relationship between cell size and fruit size. Each point shows the weight of one fruit as a function of its mean pericarp cell diameter. Mean cell diameters were calculated from mean cross-sectional cell areas estimated as described in “Materials and Methods,” by approximating cells as spheres. Note the log scale of fruit weight in all panels. A, Same experiment as in Figure 3, where fruit development was analyzed in Wva106 line from anthesis to ripening. The dashed line is the regression curve (R² = 0.96, n = 107). B, Same experiment as in Figure 2, where fruit from 20 tomato lines was compared at the breaker stage. Black symbols in B represent fruits with only two to three carpellar locules and the dashed line is the regression curve (R² = 0.85, n = 41). White symbols in B show 40 fruits with four or more carpellar locules. C, Combination of all data from A (white symbols) and of data from fruits with only two to three carpellar locules in B (black symbols). The dashed line is the regression curve (y = 3.10⁻⁶x^2.79, R² = 0.96, n = 148).

Figure 10.

Relationship between fruit size and ploidy. Fruit weight is shown as a function of MCV in pericarp. Each point represents one tomato line (mean of three to seven fruits per line, 20 tomato lines). Data are the same as in Figure 2. The number beside each symbol is the line number shown in Table I. Black symbols represent lines with only two to three carpellar locules, and the associated dashed line shows the polynomial regression curve (y = 0.0264x² + 0.2622x − 21.758, R² = 0.87, n = 12). White symbols show eight lines with four or more carpellar locules.