Sepal shape variability is robust to cell size heterogeneity in Arabidopsis

Abstract

How organisms produce organs with robust shapes and sizes is still an open question. In recent years, the Arabidopsis sepal has been used as a model system to study this question because of its highly reproducible shape and size. One interesting aspect of the sepal is that its epidermis contains cells of very different sizes. Previous reports have qualitatively shown that sepals with more or less giant cells exhibit comparable final size and shape. Here, we investigate this question using quantitative approaches. We find that a mixed population of cell size modestly contribute to the normal width of the sepal but is not essential for its shape robustness. Furthermore, in a mutant with increased cell and organ growth variability, the change in final sepal shape caused by giant cells is exaggerated but the shape robustness is not affected. This formally demonstrates that sepal shape variability is robust to cell size heterogeneity.

Keywords: growth; morphogenesis; paf1c; reproducibility; variability.

Figure 1.

The effects of giant cells on sepal shape (a) Scanning electron microscopy pictures of WT, pATML1::LGO (LGO), vip3-1 and vip3-1 pATML1::LGO (vip3-1 LGO) sepals. Giant cells make up most of the outer epidermal cell population in LGO and vip3-1 LGO sepals. n ≥ 5 samples for each genotype. Scale bar = 200 µm. (b) Representative images of mature WT, pATML1::LGO (LGO), vip3-1 and vip3-1 pATML1::LGO (vip3-1 LGO) sepals. Note the similarity between sepals in (b) and the corresponding average sepal contours in (c). Scale bar = 0.5 mm. (c) Plots showing the contours of sepals of the four genotypes. The contours are normalized to the area. The red outlines are the average shapes. n = 40, 48, 50, 53 sepals for WT, pATML1::LGO (LGO), vip3-1 and vip3-1 LGO, respectively. (d) Overlapping average shapes of WT and pATML1::LGO (upper half) and vip3-1 and vip3-1 LGO (lower half). (e) Aspect ratios (width/length) of sepals of the four genotypes. A higher aspect ratio means a wider shape and vice versa. n = 40, 48, 50, 53 sepals for WT, pATML1::LGO (LGO), vip3-1 and vip3-1 LGO, respectively. Welch’s t‐test. Different letters indicate statistically significant differences (p ≤ 0.05 for vip3-1 and vip3-1 pATML1::LGO comparison, p ≤ 0.0001 for other pairwise comparisons).

Figure 2.

The MWP index to quantify the effects of giant cells on sepal shape. (a) Two shapes of the same aspect ratio can be vastly different. To distinguish them, we identify the MWP where the width is the widest along the sepal length. The ratio L1/L is the MWP index, with L being the sepal length and L1 being the distance from the tip to the MWP. (b) MWP index of sepals of the four genotypes. A lower MWP index means the sepal is widest near the tip. n = 40, 48, 50, 53 sepals for WT, pATML1::LGO (LGO), vip3-1 and vip3-1 LGO, respectively. Two-sided Welch’s t‐test. Different letters indicate statistically significant differences (p ≤ 0.05 for pATML1::LGO and vip3-1 comparison, p ≤ 0.0001 for other pairwise comparisons).

Figure 3.

Sepal shape variability quantification in the WT and lines with different mix of cell sizes. Sepal shape variability is expressed as an S ₂ score (squared deviation of sepal contours from the average contour) in the log10 scale to aid with visualization. A higher score means higher shape variability. n = 40, 48, 50, 53 sepals for WT, pATML1::LGO (LGO), vip3-1 and vip3-1 LGO, respectively. Two-sided Welch’s t‐test. Different letters indicate statistically significant differences (p ≤ 0.001).