Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2012 Feb 2;109(5):897–910. doi: 10.1093/aob/mcs005

Spatiotemporal variation of leaf epidermal cell growth: a quantitative analysis of Arabidopsis thaliana wild-type and triple cyclinD3 mutant plants

Joanna Elsner 1, Marek Michalski 1,2, Dorota Kwiatkowska 1,*
PMCID: PMC3310487  PMID: 22307569

Abstract

Background and Aims

The epidermis of an expanding dicot leaf is a mosaic of cells differing in identity, size and differentiation stage. Here hypotheses are tested that in such a cell mosaic growth is heterogeneous and changes with time, and that this heterogeneity is not dependent on the cell cycle regulation per se.

Methods

Shape, size and growth of individual cells were followed with the aid of sequential replicas in expanding leaves of wild-type Arabidopsis thaliana and triple cyclinD3 mutant plants, and combined with ploidy estimation using epi-fluorescence microscopy.

Key Results

Relative growth rates in area of individual epidermal cells or small cell groups differ several fold from those of adjacent cells, and change in time. This spatial and temporal variation is not related to the size of either the cell or the nucleus. Shape changes and growth within an individual cell are also heterogeneous: anticlinal wall waviness appears at different times in different wall portions; portions of the cell periphery in contact with different neighbours grow with different rates. This variation is not related to cell growth anisotropy. The heterogeneity is typical for both the wild type and cycD3.

Conclusions

Growth of leaf epidermis exhibits spatiotemporal variability.

Keywords: Leaf epidermis, cell mosaic, cell expansion, growth heterogeneity, variation at cellular and sub-cellular levels, Arabidopsis thaliana, cyclinD3

INTRODUCTION

In symplastic growth, typical of plant tissues, adjacent cell walls do not alter position relative to each other, i.e. there is no cell sliding (Priestley, 1930; Erickson, 1986). Symplastic growth is therefore frequently called co-ordinated (Erickson, 1986; Evert, 2006), and it has been postulated by Nakielski (2008) that the growth of neighbouring cells is co-ordinated at the supracellular level. In meristems or organ primordia in which cells are dividing, regulation that couples cell growth and divisions is expected, leading to the maintenance of specific cell size. Also this regulation is postulated to operate at the supracellular level (John and Qi, 2008). A manifestation of the robust co-ordination of cell behaviour in developing plant organs is a compensation between the cell size and cell number that enables the maintenance of the organ size in a number of mutants where the cell cycling is affected (Doonan, 2000; Tsukaya, 2002, 2008).

The symplastic mode of plant cell growth does not imply that growth of neighbouring cells is the same (growth homogeneity). For example, in the shoot apex of Anagallis arvensis, the region where a boundary between the apical meristem and the leaf primordium is formed is characterized by growth rate and growth anisotropy (the variation of growth rate in various directions) that differ appreciably from growth of the adjacent primordium and the meristem periphery (Kwiatkowska and Dumais, 2003).

Patterns exist in dicot leaf growth (e.g. Pyke et al., 1991; Donnelly et al., 1999; Cookson et al., 2005; Fleming, 2007) although absolute values of parameters characterizing leaf growth are strongly dependent on environmental conditions (Massonnet et al., 2010). In the expansion phase of leaf development, during which the cell division rate is slowing down and the expansion rate remains high (Granier and Tardieu, 2009), leaves of numerous species exhibit a longitudinal (distal–proximal) gradient of growth rates, with higher rates in the proximal portion of the lamina (Richards and Kavanagh, 1943; Poethig and Susssex, 1985; Wolf et al., 1986; Granier and Tardieu, 1998; Wiese et al., 2007; reviewed by Walter et al., 2009). This growth rate gradient is accompanied by a gradient of cell division frequency (Pyke et al., 1991; Donnelly et al., 1999). The growth rate gradient may disappear if the leaf continues to grow but the growth distribution becomes uniform, as for example in tobacco Nicotiana tobacum (Poethig and Sussex, 1985). In the maturation phase cessation of cell divisions precedes cessation of cell expansion (Granier and Tardieu, 2009). The cell division cessation is related to two fronts of cell cycle arrest (White, 2006). The primary front is responsible for the arrest of general cell divisions. In the jaw-D mutant of Arabidopsis thaliana (Palatnik et al., 2003) or cincinnata of Antirrhinum majus (Nath et al., 2003) where the front of cell division termination is altered, the geometry of the leaf lamina is severely affected. The secondary front, in which regulation PEAPOD (PPD) is involved, imposes arrest of the cell cycle in dispersed meristematic cells, including the epidermal meristemoids or procambium (White, 2006). The fronts of cell cycle cessation are followed by the expansion (growth) cessation front, as shown for leaves of sunflower Helianthus annuus (Granier and Tardieu, 1998).

Recent reviews on plant growth and morphogenesis highlight the need for further quantification of spatial and temporal growth patterns at high resolution in leaves (Coen et al., 2004; Walter et al., 2009). Taking into account that the epidermis of an expanding leaf is often more non-uniform in a number of traits than other leaf tissues, especially mesophyll, one may expect local complexity of the leaf growth pattern in this tissue. Firstly, in the leaf epidermis, formative divisions and proliferation (Gunning et al., 1978; Fosket, 1994) take place side by side. Secondly, neighbouring cells are often at various stages of differentiation (Granier and Tardieu, 2009), e.g. differentiated cells frequently are next to cells that are still undergoing mitotic division. Thirdly, neighbouring cells often exhibit different identities, e.g. pavement cells, stomata guard cells, meristemoids, and trichome-associated support cells in A. thaliana leaf epidermis. In many dicot leaves, the differentiation of epidermal pavement cells, starting in cells when the leaf is still expanding, involves formation of wavy anticlinal walls leading to a characteristic jigsaw puzzle cell shape (Fu et al., 2005; Panteris and Galatis, 2005). The unique, and as yet incompletely elucidated, cell wall growth involved in this process (Hejnowicz, 2011; Zhang et al., 2011) adds to the complexity of epidermal growth. In A. thaliana leaf epidermis there is also variation in the cell ploidy level that increases due to endoreduplication (Joubès and Chevalier, 2000); as a result, the tissue becomes a mosaic of cells differing substantially in their DNA content (De Veylder et al., 2001; Boudolf et al., 2004; Vlieghe et al., 2007). The endocycles in shoot lateral organs of A. thaliana are restrained by the CYCLIND3 (CYCD3) proteins that promote mitotic cycles (Dewitte et al., 2003, 2007). A loss of function of the complete CYCD3 sub-group in the triple cycD3 mutant of A. thaliana leads to earlier cessation of mitotic cell cycling and reduction of cell number in leaves, but in adult leaves both the cell size and ploidy level are increased and consequently leaf size is not severely affected, i.e. the above-mentioned compensation mechanism operates (Dewitte et al., 2007). However, this is not the case in juvenile leaves where the cell size is increased but the cell number is not reduced, resulting in increased leaf size (Dewitte et al., 2007).

The present investigation focuses on the leaf epidermis growth in A. thaliana. Three hypotheses concerning the growth complexity of A. thaliana leaf epidermis are tested: (1) the local growth rate in area is heterogeneous, displaying patchiness, and variation with time; (2) even within a single cell, the cell wall expansion and development of a jigsaw puzzle shape are heterogeneous, the latter depending on neighbouring cells; and (3) these heterogeneities are not dependent on the cell cycle regulation per se.

To test these hypotheses, the growth pattern is examined here with the aid of a sequential replica method, both in the wild type and in the triple cycD3 mutant, so that the wild type can be compared with a system with disturbed cell cycle regulation but in which the compensation between cell size and cell number operates and organ size is not affected. First, fates of individual epidermal cells are followed in order to recognize geometrical alterations accompanying pavement cell differentiation. Secondly, spatial variation in epidermal cell and nucleus size is analysed. Then the pattern of epidermal growth is quantified at the cellular and sub-cellular levels, and the question about the relationships between the above cellular parameters and cell growth is addressed. Since leaves of both the wild type and the triple cycD3 mutant are used in this investigation, a wide range of the parameter values can be examined.

MATERIALS AND METHODS

Plant material and growth conditions

Plants of Arabidopsis thaliana ‘Columbia-0’ (Col), obtained from the Nottingham Arabidopsis Stock Centre (NASC), and the triple cyclinD3 mutant (cycD3) on the Col background (kindly provided by Walter Dewitte and James A. H. Murray, Cardiff University), were grown in pots in a growth room under short-day conditions (9 h day; 15 h night), temperature 19–21 °C and illumination 60 µmol m−2 s−1. The third or fourth leaves of aerial rosettes that formed in the axil of the oldest cauline leaf (Fig. 1A) of plants 16–20 weeks after germination were used in the investigation. The advantage of using the aerial rosette leaves was that they are easy to access in order to take replicas, particularly in comparison with regular rosette leaves.

Fig. 1.

Fig. 1.

Open in a new tab

A fragment of A. thaliana Col shoot with aerial rosettes growing in cauline leaf axils. Scale bar = 10 mm (A). Leaf portions used in cellular parameters and growth computation overlaid on an SEM micrograph of an exemplary Col leaf. Scale bar = 500 µm (B). Nucleus size and DNA content assessment (C–F). The surface area of cell nuclei measured in central optical sections is linearly dependent on the DAPI fluorescence intensity (in 106) of the same nuclei (y = 1·1 × 10−5 x + 7·74; R2 = 0·596) measured in image cytometry in arbitrary units (C). To obtain values of various cellular parameters for individual cells, epi-fluorescence micrographs are taken from a whole leaf preparation at a different focus so that the anticlinal cell wall outlines (D) and the nucleus outlines (E) are distinct. The same fragment of epidermis is recognized in the nail polish replica (F) in which other cellular parameters are assessed. Exemplary cells visible in all the images are labelled with asterisks; arrowheads point to their nuclei. Scale bars = 20 µm.

Eight Col and nine cycD3 leaves were examined. The leaf lamina length of all the leaves at the beginning of observation was 1·4–3·0 mm. The leaves were in the expansion phase of development, distinguished by Granier and Tardieu (2009).

Sequential replicas

The sequential replica method (Williams and Green, 1988) was used to obtain sequences of silicon moulds, made of Take 1 impression material (the hydrophilic vinyl wash material, regular set, Kerr Corp., Romulus, USA), from the abaxial epidermis of individual leaves. At the moment of application the silicon polymer is a fluid of sufficiently low density to allow it to penetrate the spaces between trichomes and reach the pavement cells (thus there are trichomes on the surface of the epoxy resin replica shown in Fig. 1B).

Sequences of three replicas were taken at 48 h intervals from the entire abaxial surface of the lamina of individual leaves. Arabidopsis thaliana leaf growth is known to exhibit diel patterns (Wiese et al., 2007). However, in the case of 48 h intervals between consecutive replicas, this temporal variation does not affect assessment of growth parameters since they are averaged for the time interval between taking the replicas.

Nail polish replicas (Kagan et al., 1992; Geisler and Sack, 2002) of the entire surface of leaf laminas were obtained from all the silicon moulds and observed under a light microscope. Sequential images from nail polish replicas were used for the assessment of the cellular parameters described below.

From sequences of silicon moulds of four Col and four cycD3 leaves, epoxy resin replicas (casts made from Devcon 2 ton epoxy) were also obtained (Williams and Green, 1988). They were sputter-coated and observed by scanning electron microscopy (SEM; Philips XL 30 TMP ESEN). The sequences of SEM micrographs (pairs of stereoimages taken from each region of interest) were used for the analysis of epidermis growth parameters and pavement cell morphogenesis.

Measurements and computation of cellular parameters in nail polish replicas

Each leaf lamina surface was divided by a line perpendicular to the midrib into proximal and distal portions of equal dimension along the midrib (Fig. 1B). In sequences of nail polish replicas from individual leaf epidermis, large groups of cells from these two portions of lamina were mapped. These groups included approx. 100 neighbouring cells (including the stomata guard and guard mother cells) and were located between, but did not include, trichomes and trichome-associated support cells. Two cellular parameters were assessed: (1) the cell surface area and (2) the relative growth rate in cell area. During assessment of these parameters the pairs of stomata guard cells were treated as single cells. The cell surface area was measured with the aid of ImageJ (National Institutes of Health; downloaded from http://rsbweb.nih.gov/ij/). It should be kept in mind that such measurements represent the surface area in the orthogonal projection of epidermal cells and not the actual area of the outer periclinal cell wall, which is not flat. The relative growth rate in cell area (Ra) was computed as Ra = ln(A′/A)/Δt, where A represents the cell surface area measured in the first nail polish replica; A′ equals the same cell (or the cell progeny) area in the second replica, and Δt is the time interval between taking the replicas (Dumais and Kwiatkowska, 2002). All the cells examined were at least ten cells away from the leaf margin, and at a sufficient distance from the midrib that no cells elongated along the midrib were included in the analysis.

Colour maps of the distribution of the cellular parameters were plotted on the cell wall patterns with the aid of CorelDRAW Graphics Suite X4 (Corel Corp.) and Adobe Photoshop and Illustrator CS4 Extended (Adobe Systems Inc.).

Epi-fluorescence microscopy and DNA content assessment

Entire leaves were fixed in aceto-alcohol (Johansen, 1940) immediately after the last replica was taken. Schiff reagent (Sigma-Aldrich, S-5133) staining was followed by mounting in Vectashield with 4,6′-diamidino-2-phenylindole (DAPI; Vector Laboratories Inc., H-1200) allowing visualization of both cell walls and nuclei using epi-fluorescence microscopy (Olympus BX41). Close inspection of the anticlinal cell wall pattern allowed us to recognize the same epidermal cells in replicas and whole-leaf preparations (Fig. 1D–F). One or two images were taken from each of these cells at various focal levels so that the anticlinal cell wall and/or the cell nucleus outline were distinct (Fig. 1D, E). The DNA content was estimated as the surface area of the nucleus measured in the optical section with the most distinct nucleus outline. The surface area of nuclei was measured with the aid of ImageJ and assigned to individual cells for which other cellular parameters were assessed in nail polish replicas (Fig. 1F). In order to check to what extent such assessed nucleus size represents the DNA content, the DNA content of nuclei was estimated with the aid of image cytometry (Olympus Scan^R) on the basis of total DAPI fluorescence intensity. This was done for a group of about 80 cells of the epidermis peeled from mature leaves. For the same cells, the nucleus section surface area was measured using epi-fluorescence microscopy with the aid of ImageJ. Linear regression analysis (Zar, 1999) shows that the nucleus section area is linearly dependent on the DNA content (Fig. 1C).

The nucleus size measurements coupled with the computation of other cellular parameters was performed for the proximal portion of leaves from each genotype (for 80–90 non-guard cells). Similar numbers of stomata guard cells were analysed for the two genotypes. For each pair of guard cells a nucleus size was measured in only one cell.

Stereoscopic reconstruction and computation of growth parameters from SEM micrographs of epoxy resin replicas

Sequences of epoxy resin replicas (made for four Col and four cycD3 leaves) were used for the stereoscopic reconstruction of abaxial leaf surface, based on stereopairs of SEM images (Routier-Kierzkowska and Kwiatkowska, 2008). Cell growth analysis using this method is more detailed and can be carried out for larger portions of young leaves than if nail polish replicas are used, for a number of reasons. First, the cellular pattern of the leaf epidermis is very distinct on the whole surface of epoxy resin casts, and secondly the stereoscopic reconstruction of the replica surface ensures that the measurements are unbiased even if locally the leaf surface is not ideally planar (e.g. folded or bent).

In the sequences of epoxy resin replicas, the fates of ten (five in Col and five in cycD3) groups of 50–90 adjacent cells were followed in order to compute four parameters characterizing cell growth: (1) relative growth rate in the approximated cell area; (2) relative growth rate in length for straight line segments between adjacent vertices of individual cells (i.e. two vertices joined by the common anticlinal wall); (3) directions and values of maximal and minimal growth rates in length averaged for each cell; and (4) average cell growth anisotropy. Relative growth rate in the approximated cell area (Rap) was computed as Rap = ln(A′ap/Aap)/Δt, where Aap and A′ap are the approximated surface area of the cell (or the cell progeny) before and after growth, respectively, and Δt is the time interval between taking the replicas (Dumais and Kwiatkowska, 2002). The area of a cell was approximated as the sum of surface areas of all the triangles defined by two adjacent cell vertices and the cell centroid. Such a computed relative growth rate in the approximated cell area produces values similar to the relative growth rate in cell area computed from nail polish replicas (Supplementary Data Fig. S1). Relative growth rates in length (Rs) were computed as Rs = ln(s′/s)/Δt, where s and s′ are the lengths of the straight line segment between adjacent cell vertices before and after growth, respectively. In order to assess anisotropy of growth rates in length for each cell, maximal and minimal growth rates and directions [principal directions of growth (strain) rates] were computed on the basis of relative changes in vertex positions (Dumais and Kwiatkowska, 2002). In this computation protocol, directions of maximal and minimal growth rates are first computed for each cell vertex on the basis of deformation of a triangle defined by the vertices that are joined with the vertex by a common cell wall (the natural vertex neighbours). Then, the values and directions for all the cell vertices are averaged. Consequently the cell growth anisotropy is assessed on the basis of vertices belonging to the cell plus their natural neighbours, and therefore referred to as the average cell growth anisotropy. The growth anisotropy (ani) was computed as ani = (rmax – rmin)/(rmax + rmin), where rmax represents the maximal growth rate and rmin is the minimal growth rate (Dumais and Kwiatkowska, 2002).

All codes used for this analysis were written in Matlab (The Mathworks, Natick, MA, USA).

Statistical analysis

Data distribution of the cellular parameters was shown not to be normal using the Kolmogorov–Smirnov test. Accordingly, the non-parametric Kruskal–Wallis test was used for multiple comparisons of mean ranks, with the significance level of 0·05. The interquartile range (the distance between the first and the third quantiles) was used as a measure of data dispersion, while the coefficient of variation (the ratio between the standard deviation and the sample mean) was used as a measure of relative dispersion. Relationships between various cellular parameters of individual cells were tested with a linear regression model (Zar, 1999). Descriptive statistics, tests and plots were performed with the aid of the Statistics Toolbox of Matlab (The Mathworks).

RESULTS

Geometry changes of pavement cells

In both Col and cycD3 leaf epidermis, changes in geometry of individual pavement cells visible in developmental sequences of SEM micrographs (Fig. 2A–D) comprise cell enlargement and development of anticlinal wall waviness leading to the jigsaw puzzle shape of cells. The shape of an anticlinal wall of an individual pavement cell is different at the portions where it contacts different neighbouring cells, the anticlinal wall segments. In numerous relatively large pavement cells, the anticlinal wall is very wavy in some segments and not wavy in others (e.g. cell 1 in Fig. 2A, B). Relatively long anticlinal wall segments where two large cells come into contact are most often wavy (the wall segment between cells 1 and 2 in Fig. 2A, B). Relatively short segments are not wavy. These are segments at contacts between large cells and small cells that are still dividing or have recently divided (segment between cells 1 and 3 in Fig. 2A, B), or short segments between two large pavement cells (segment between cells 1 and 4 in Fig. 2A, B). The anticlinal wall segments where small and large cells come into contact have a special shape: the small cell appears to bulge into the large cell (like cell 3 and later its progeny shown in Fig. 2A). Later, such curved wall segments develop invaginations, i.e. become wavy (like the wall segment indicated by the white arrow in Fig. 2C). Initially, the segment waviness develops as shallow invaginations. Later, the ‘wave amplitude’ of the segment increases (segment indicated by the white arrowhead in Fig. 2D). In some segments the number of invaginations increases due to the appearance of a new invagination (black arrows in Fig. 2C, D).

Fig. 2.

Fig. 2.

Open in a new tab

Pavement cell morphogenesis shown in sequences of SEM micrographs of the abaxial epidermis of expanding leaves. The time at which the replicas were taken is given on the top of each column of micrographs. Formation of wavy anticlinal cell walls is shown for proximal portions of Col (A, D) and cycD3 (B, C) leaves. Anticlinal wall segments of exemplary large cells, labelled by 1, exhibit different shapes at contacts with cells 2, 3 and 4 (A, B). Black arrows (C, D) point to cell wall segments on which secondary invaginations are formed; white arrows (C) label the wall that first bulges into the larger cell and afterwards attains a wavy shape; white arrowheads (D) point to the wall in which the wave amplitude increases in time without formation of secondary invaginations. Occurrence of divisions in jigsaw puzzle-shaped pavement cells (pointed by arrows) is exemplified by a proximal portion of Col leaf (E). Scale bars = 20 µm.

As a rule, once the waviness of a wall segment has started to develop, no new cell wall that would join this particular wavy segment is formed due to a division in contacting cells. Exceptions are divisions that occur infrequently in already jigsaw puzzle-shaped cells both in Col (as new walls indicated by arrows in Fig. 2E) and in the mutant. These are symmetric divisions that lead to formation of two pavement cells, i.e. divisions of the proliferation type.

Variation of cellular parameters in nail polish replicas of Col and cycD3 leaf epidermis

All the cellular parameters examined in nail polish replicas of expanding Col and cycD3 leaf epidermis exhibit high variation, i.e. large dispersion of the parameter values (high interquartile ranges shown in Fig. 3A–C) and high coefficient of variation (Table 1). Some tendencies are, nevertheless, clear. The mean cell surface area is bigger in distal than in proximal leaf portions, and the mean cell area in cycD3 leaf portions is significantly higher than in corresponding portions of Col leaves (Table 1; all differences between mean ranks are statistically significant; significance level of 0·05 of the Kruskal–Wallis test). Similarly, the cell surface area dispersion (Fig. 3A) is higher in distal leaf portions than in the proximal portions, and in cycD3 leaves than in Col leaves. The coefficient of variation of the cell surface area, a measure of the relative dispersion, is similar in the two genotypes (Table 1).

Fig. 3.

Fig. 3.

Open in a new tab

Variation of cellular parameters measured in nail polish replicas of expanding leaf epidermis (A–C) and test for their dependence (D–F). Three cellular parameters are examined: cell surface area (A); relative growth rate in cell area (B); and nucleus section area (C). In box-type plots (A–C) all measurements for large groups of approx. 100 cells, made in the first replicas of all the leaves in a given genotype, are taken together. The relative growth rates between the first and the second replicas only are considered (B). In (C) measurements are from all the leaves in a given genotype, separate for pavement cells and stomata guard cells (GC). A solid line within each box represents the median. The box delimits the first and third quantiles. Whiskers extend from each end of the box to the adjacent values in the data as long as the most extreme values are within 1·5 times the interquartile range from the ends of the box. Crosses represent outliers, i.e. data with values beyond the ends of the whiskers. (D) Cell surface area is linearly dependent on the nucleus surface area both in Col (y = 28·98 x – 142·69; R2 = 0·651) and in cycD3 (y = 34·94 x – 462·76; R2 = 0·628). (E) The relative growth rate in area during the time interval directly preceding material fixation for the nucleus size measurement is not linearly dependent on the nucleus surface area, both in Col (R2 = 0·108) and in cycD3 (R2 = 0·028). (F) The relative growth rate in area (in the time interval between taking the first and the second replica), plotted with respect to the initial cell surface area, also shows no linear dependence either in Col (R2 = 0·09) or in cycD3 (R2 = 0·122). Stomatal guard cells or guard mother cells are excluded from this analysis.

Table 1.

Descriptive statistics of cellular parameters computed for proximal and distal portions of expanding Col and cycD3 leaves

Col leaf
 cycD3 leaf
Cellular parameter Proximal Distal Proximal Distal
Cell surface area (μm2) 152 ± 6 (98/618) 388 ± 24 (125/406) 197 ± 9 (105/491) 684 ± 38 (113/406)
Relative growth rate in area (10−3 μm2 μm−2 h−1) 12·2 ± 0·3 (38/200) 7·4 ± 0·3 (49/171) 14·2 ± 0·3 (31/240) 9·6 ±.3 (45/200)
Nucleus area: non-guard cells (μm2) 22·8 ± 1·8 (75/90) 64·5 ± 3·7 (54/88)
Nucleus area: guard cells (μm2) 9·3 ± 0·2 (25/88) 12·3 ± 0·3 (24/83)
Open in a new tab

Values are means ± s.e., with the coefficient of variation (%) and sample number given in parentheses.

Samples are the same as used for the box-plots shown in Fig. 3A–C. Data were obtained from nail polish replicas. Differences between mean ranks in each row are statistically significant (P < 0·05; Kruskal–Wallis test).

Strikingly, variation of relative growth rate in cell area is also very high. The mean rate is significantly higher in proximal than in distal leaf portions, and mean rates in cycD3 leaf portions are higher than in corresponding portions of Col leaves (Table 1; Kruskal–Wallis test). In both Col and cycD3 leaves the dispersion of the growth rate values is lower in distal than in proximal portions (Fig. 3B), and the coefficient of variation is high and similar in the two genotypes (Table 1).

In order to check whether correlations existed between the observed variation in cell size, relative growth rates and cell ploidy, the mean size of the nucleus (the surface area of a central optical nucleus section) was assessed for groups of cells in proximal leaf portions, and for the same cells other parameters were computed (Table 1). As reported earlier (Melaragno et al., 1993), the smallest mean nucleus size is that of stomatal guard cells. Also the data dispersion and coefficient of variation are the lowest for guard cell nuclei (Fig. 3C; Table 1). In non-guard cells, similarly to the cell surface area, the mean nucleus size is significantly bigger in cycD3 than in Col epidermis (Table 1; Kruskal–Wallis test). Accordingly, there is a linear dependence between the surface area of non-guard cells and the cell nucleus size in both genotypes (Fig. 3D). This is not a strong dependence (R2 between 0·6 and 0·7). However, the shapes of large nuclei are variable and often non-spherical (compare the nucleus shape in cells labelled with asterisks in Fig. 1D, E), which implies that the relationship between the section area and nucleus volume (thus also the DNA content) is different in different nuclei. No linear dependence was found between the relative growth rate in cell area and the nucleus size (Fig. 3E), or between the growth rate and the initial cell surface area (Fig. 3F).

Local spatial variation of the relative growth rate in cell area measured in nail polish replicas

The relative growth rate in cell area exhibits a high local spatial variation in distal and proximal leaf lamina portions (Figs 4 and 5). Both the Col (Fig. 4) and cycD3 (Fig. 5) epidermis is a mosaic of single cells or groups of a few cells characterized by growth rate values which differ from that of their neighbours (arrows in Figs 4 and 5). Differences between cell growth rates within the same region can be up to 6-fold, with a 3-fold difference between adjacent cells commonly occurring (e.g. Figs 4A and 5E). These differences are not related to cell size or differentiation (compare, for example, cells in Fig. 5D). This is in agreement with the lack of linear dependence of the relative growth rate in area and initial cell size described above (Fig. 3F). Moreover, often a group of cells or a cell may exhibit fluctuation in expansion rate (e.g. cells labelled with arrows in Figs 4A–C, 5A–C and Fig. 4D–F). In consecutive intervals, however, members of such groups of cells often change (e.g. cells labelled with arrows in Fig. 5D–F).

Fig. 4.

Fig. 4.

Open in a new tab

Local spatial variation in relative growth rate in cell area computed from nail polish replicas, for large groups of cells from the distal (A–C) and proximal (D–F) portion of an expanding Col leaf (initial leaf lamina length was 1·4 mm). The time at which the replica was taken is given in the lower right corner of each map. The orientation of all maps is such that the leaf midrib is vertical and the apical direction points upward. Maps representing the relative growth rates are plotted on the cell pattern as it appeared at the beginning of the given time interval (A, B; D, E). In the last maps of each row the cell outlines only are plotted. In these maps, pairs of stomata guard cells are labelled with ‘x’. Exemplary groups of cells that differ in growth rate from surrounding cells are labelled with arrows. Scale bar = 100 µm.

Fig. 5.

Fig. 5.

Open in a new tab

Local spatial variation in relative growth rate in cell area computed from nail polish replicas, for large groups of cells from the distal (A–C) and proximal (D–F) portion of a expanding cycD3 leaf (1·7 mm long). Labelling as in Fig. 4. Scale bar = 100 µm.

Local spatial variation of relative growth rates at cellular and sub-cellular levels measured in epoxy resin replicas

Computation of the growth rates in cell area from nail polish replicas is based on real cell shapes and thus accounts for formation of wavy anticlinal walls. The observed local spatial variation in growth rate might thus be related to this process. Therefore, in order to investigate further the local growth variation, epidermal cell growth was assessed as the relative growth rate in area of polygons approximating the cell shape, which are defined by vertex positions in the reconstructed epidermis surface (based on epoxy resin replicas observed in SEM). Vertices are identified as contact points between the anticlinal walls of three adjacent cells and therefore their precise positions can be tracked in consecutive replicas. The relative growth rates in approximated cell area were computed for groups of cells in all the sequences of epoxy resin replicas examined, and are exemplified by three groups. One group is from the proximal portion of Col leaf (Fig. 6). Two groups, one located closer to the lamina base than the other, are from the proximal portion of a cycD3 leaf (Figs 7 and 8), the length of which was much smaller than the length of the Col leaf.

Fig. 6.

Fig. 6.

Open in a new tab

Scanning electron micrographs (A–C) and growth parameter plots (D–K) computed from the sequence of epoxy resin replicas taken from the proximal portion of an expanding Col leaf (initial lamina length was 2·8 mm). The time at which the replica was taken is given in the lower right corner of each micrograph. Only cells for which the parameters were computed are shown in the micrographs. Colour maps representing growth parameters are plotted on the cell pattern as it appeared at the beginning of the given time interval. Growth parameters visualized in maps are: relative growth rate in approximated cell area (D, E); relative growth rate in length of straight line segments between two vertices joined by the same anticlinal cell wall (F, G); coefficient of variation (CV) of the growth rates in length of segments belonging to individual cells (H, I); directions of maximal growth rate and average cell growth anisotropy (J, K). Growth rates in length are plotted as maps in which the colour code is assigned to the considered line segment. In anisotropy plots, short line segments representing the direction of maximal growth are plotted on colour maps representing average cell growth anisotropy. An exemplary small group of cells and a single cell outlined in black change their areal growth rates in consecutive time intervals. Dots mark cells with similar growth anisotropy but a different CV. Scale bar = 50 µm.

Fig. 7.

Fig. 7.

Open in a new tab

Scanning electron micrographs (A–C) and growth parameter plots (D–K) for the sequence of replicas taken from the proximal portion of an expanding cycD3 leaf (1·7 mm initial length). Labelling as in Fig. 6. Exemplary adjacent cells of similar size that differ a lot in growth rates in area are labelled with asterisks; arrows point to an exemplary pair of cells exhibiting different values of the coefficient of variation despite the similar size. Scale bar = 50 µm.

Fig. 8.

Fig. 8.

Open in a new tab

Scanning electron micrographs (A–C) and growth parameter plots (D–K) for the sequence of replicas taken from the proximal portion of an expanding cycD3 leaf. This epidermis fragment originates from the same sequence of replicas as that shown in Fig. 7 but was located farther away from the leaf lamina base. Labelling as in Figs 6 and 7. Scale bar = 50 µm.

Local variation in growth rate in approximated cell area is as high as the variation in growth rate in cell area computed from nail polish replicas, both in Col and in cycD3 (compare Figs 6D, E with 4D, E; 7D, E and 8D, E with 5D, E; see also Supplementary Data Fig. S1). No relationship between the growth rate in approximated area and the approximated initial cell area is apparent (compare pairs of cells labelled with asterisks in Fig. 7A–E). Differences in the growth rates of cells from the same epidermis fragment are sometimes as much as 4-fold even if stomatal guard cells or guard mother cells are not taken into consideration (e.g. Figs 6D, 7E and 8D). In the case of stomata, since a pair of guard or guard mother cells was treated as a single polygon, most often a triangle, computed growth rates are usually lower than the actual rates. This is because the ontogenetic shape change in these cells comprises rounding, while no new vertices are formed. Similar to relative growth rate in cell area computed from nail polish replicas, relative growth rate in approximated cell area often changes in consecutive time intervals (e.g. cells outlined in Figs 6D, E; 8D, E grow faster than surrounding cells in the first time interval and slower in the second).

Because of this striking local variation of the areal growth rates, growth rates were also quantified at higher spatial resolution. For this purpose, relative growth rate in length was computed for straight line segments joining adjacent vertices of each individual cell. Such line segments, representing different regions of an individual cell, often differ several fold in the growth rate (Figs 6F, G; 7F, G; 8F, G). The differentiation stages of the two cells that have the two vertices in common do not seem to influence the growth rate of the line segment. The coefficient of variation computed for all the segments assigned to the individual cell (Figs 6H, I; 7H, I; 8H, I) is within the range of about 10 % to >100 % with no apparent relationship to the cell size (compare the cells indicated by arrows in Fig. 7H, I, that have similar size but differ substantially in the coefficient of variation).

If growth is anisotropic the growth rate in length of a line segment depends on the segment direction, and the difference between growth rate in different directions increases with the anisotropy. Thus in order to check whether the observed variation of growth rate in length is not the manifestation of strongly anisotropic growth, the average cell growth rate anisotropy was assessed (Figs 6J, K; 7J, K; 8J, K). Short line segments plotted within cell outlines point to the directions of maximal growth rate, and colour maps show anisotropy of growth rates (the higher the anisotropy value, the bigger the difference between maximal and minimal growth rates). If the anisotropy value is zero, growth is isotropic, while if this value equals, for example, 0·5, the maximal growth rate is three times bigger that the minimum. Growth anisotropy of the epidermis portions examined is generally low though not uniform (Figs 6J, K; 7J, K; 8J, K). No relationship between growth anisotropy and coefficient of variation of growth rate in length of segments assigned to the cell is apparent. Cells exhibiting similar anisotropy, i.e. both the anisotropy value and the direction of maximal growth, have different coefficients of variation (compare adjacent cells labelled with dots in Figs 6I and K; 7H and J; 8H and J).

DISCUSSION

Heterogeneity of leaf epidermis growth at cellular and sub-cellular levels

Examination of leaf epidermis in A. thaliana at a single time point shows that it is a mosaic of cells differing in identity, size and ploidy level (Boudolf et al., 2004). In expanding dicot leaves, adjacent epidermal cells differ in differentiation stages (Granier and Tardieu, 2009). Time-lapse imaging of individual cells allowed us to investigate local spatial variation of growth of such a cell mosaic in order to test the hypothesis that its growth is heterogeneous and alters with time. Indeed, the epidermis of all the expanding A. thaliana leaves examined is heterogeneous not only in cell differentiation stage and size, but also in growth. This is true for both wild-type and triple cycD3 mutant plants, with disturbed cell cycle regulation and the compensation between cell size and number operating at the organ level. Differences in relative growth rate in cell area between individual cells in the same epidermis fragments are several fold. Furthermore, the growth rate alters over time (unsteady): often either a cell or a small group of cells growing relatively quickly in one time interval grow slowly in another, or the opposite change takes place, as if the epidermis comprised small pulsating domains of different growth. The boundaries of these domains, however, are not stable with respect to cells. Such spatial growth variation is related neither to the variation in cell size nor to variation in the size of the nucleus. The measurement of variable growth rate in cell area in nail polish replicas may reflect the formation of wavy anticlinal walls if, for example, invaginations ‘grew deeper’ into one cell than the other. Examining such a phenomenon, however, would require a different computation method from that used in the present investigations. Nevertheless, computation of the relative growth rate in cell area approximated by a polygon, the growth parameter that omits wall waviness formation, confirmed the existence of several fold differences between cells.

Such growth heterogeneity has not been reported for expanding A. thaliana cotyledon epidermis (Zhang et al., 2011), which may be a manifestation of the relative simplicity of cotyledon development. Patches of different growth rates have, however, been reported for leaves of tobacco exhibiting a growth pattern of leaf lamina generally similar to A. thaliana (Walter and Schurr, 2005), and also for Populus deltoides leaves (Matsubara et al., 2006) where the general growth rate distribution is uniform. Similar to domains reported herein, shapes and positions of these patches altered with time, i.e. were temporary not uniform. However, they were apparent at much lower spatial resolution and at much higher temporal resolution than those reported here (at time intervals of minutes as compared with 48 h in the present investigations).

The existence of patches of varying growth rates is interesting in that it requires some compensation by adequate spatial and temporal growth processes, otherwise the leaf surface would buckle (Walter and Schurr, 2005). This concurs with the present observation that local growth rates change in time: cells that grow relatively quickly in one time interval grow more slowly in the other. The growth heterogeneity (in space) is also interesting from the perspective of symplastic growth, which imposes the requirement of identical growth pattern at the two contacting cell walls. Therefore, in such a mosaic-type distribution of cells with various growth rates as exists in leaf epidermis, non-uniform growth of various portions of an individual cell may be expected. Indeed the present analysis shows that even within a single cell the expansion is heterogeneous: there are sub-cellular differences in growth rates around the cell periphery, whose portions are approximated by straight line segments for which the growth rate in length was computed. This variation is not simply a manifestation of cell growth anisotropy. First, the average growth anisotropy computed for cells is rather low. Secondly, no relationship was found between the cell growth anisotropy and the variation of growth rates of the cell periphery. The sub-cellular differences in growth rates of the cell periphery reported here are in agreement with the postulate of Jarvis and collaborators (2003) to consider growth as quantized not into individual cells but into pairs of adherent cell walls. Such individuality of wall portions is further supported by the heterogeneity in waviness formation described below.

Complex morphogenesis of pavement cells of A. thaliana epidermis

A specific feature of epidermal pavement cells in many leaves are wavy anticlinal cell walls known only from a few other tissues (Panteris et al., 1994). Several aspects of this cell shape formation, especially the role of the cytoskeleton in the process, have been thoroughly investigated (Panteris and Galatis, 2005; Kotzer and Wasteneys, 2006; Zhang et al., 2011). However, to our knowledge, only a single publication exists reporting the individual pavement cell growth during and after wavy wall formation, namely in the case of expanding A. thaliana cotyledons (Zhang et al., 2011). The replica method and computation protocols used in the present investigation enable the assessment of growth parameters for the outer periclinal walls of epidermal cells. In the case of anticlinal walls their shape changes can be followed but local growth of anticlinal walls cannot be assessed because no marker points apart from the vertices are available that could be recognized at consecutive replicas. Nevertheless, examination of sequential replicas provides some important information on waviness formation. First, the process starts in cells whose surface area will increase many fold later on. Secondly, at least in some cells, the waviness formation is a step-wise process where development of initial invaginations is followed by formation of secondary invaginations. Such additional invagination (lobe) formation events have also been indirectly recognized for cotyledon pavement cells (Zhang et al., 2011).

The examination of sequential replicas confirms the hypothesis that the formation of anticlinal wall waviness is heterogeneous within a single cell. Individual segments of the anticlinal cell wall behave independently in terms of waviness formation, and this formation is apparently related to the length of the segment and differentiation stages of cells contacting the wall segment. At contacts between small and large cells, walls are not wavy, but small cells appear to bulge into the large ones, which could possibly be explained by differences in turgor in adjacent cells (as postulated by Corson et al., 2009, on the basis of theoretical consideration). Anticlinal wall segments common to the two large cells are usually wavy, unless the segment is relatively short. Interestingly, even though in general cells with wavy walls stop dividing, there are examples where divisions take place in large cells with well-developed wall waviness that are most probably endoreduplicating. Rare divisions in already endoreduplicating cells were also reported for A. thaliana sepal epidermis (Roeder et al., 2010).

Size and growth of epidermal cells differing in DNA content

Both ploidy-dependent and independent factors are postulated to control the cell size (Cookson et al., 2006; Vlieghe et al., 2007). Accordingly, many reports show the relationship between the cell size and the ploidy level in endoreduplicating cells (e.g. Melaragno et al., 1993; Roeder et al., 2010), though such a relationship has not always been confirmed (e.g. Beemster et al., 2002). Indeed, the only significant linear dependence that is reported in the present paper is between the cell surface area and the nucleus section area, which is related to the DNA content. This dependence is not strong but the DNA content assessment by nucleus section area is probably biased by the variability of nucleus shapes occurring, especially in the case of large nuclei. However, also in the case of epidermal cells of the A. thaliana sepal where a significant linear correlation between the surface area and nucleus size was shown, the DNA content in giant epidermal cells varies (Roeder et al., 2010).

Though accommodating large cell size or rapid cell expansion might be a function of endoreduplication (Galbraith et al., 1991), no correlation between the nucleus size and cell growth was found here. This may be because the endoreduplication takes place prior to cell enlargement in order to make it possible in future. The exact amount of enlargement may, however, not be set by the ploidy, especially when cells are embedded in a tissue, like epidermal pavement cells (Traas et al., 1998). Another explanation for the lack of a close link between the cell growth rate and ploidy is that ploidy may control an increase in the mass of cytoplasmic macromolecules (referred to as ‘cell growth’ by Sugimoto-Shirasu and Roberts, 2003), but the control of cell volume increase through vacuolation (‘cell expansion’) may be different (Sugimoto-Shirasu and Roberts, 2003).

Comparison of leaf epidermis in triple cycD3 mutant and wild-type plants

The present results confirm the hypothesis that the growth heterogeneity of leaf epidermis is not dependent on the cell cycle regulation per se. At both the cellular and the sub-cellular levels, the general tendencies in spatial variation of cell growth, as well as cell and nucleus size, are similar in the wild-type and triple cycD3 mutant plants, with the loss of function of the complete CYCD3 sub-group. Only the means and distributions in the mutant are shifted toward higher values. This may be yet another manifestation of the compensation between cell size and cell number in leaves of the mutant. These results are in agreement with previous reports (Dewitte et al., 2007). The new observation is the significant though small difference in the relative growth rate in cell area in expanding aerial rosette leaves, which has higher values in the mutant. One reason for this may be in that in the mutant where there are more large cells, the growth due to vacuole enlargement, which is faster than growth of proliferating cells (Fleming, 2007), may contribute more to the leaf growth than in the wild type. The other difference apparent in sequential replicas is in the duration of epidermal meristemoid activity and the frequency of satellitary meristemoid formation. This is apparent in maps of cell surface area variation where in expanding leaves of the mutant, patches of small cells with few anticlinal wall invaginations (i.e. mainly meristemoid cells or stomata guard mother cells) are smaller than in the wild type (compare Figs 4C and 5C). This problem, however, requires further investigation.

SUPPLEMENTARY DATA

Supplementary data are available online at www.aob.oxfordjournals.org/ and consist of Fig. S1: variation of areal growth rates, presented as a box-plot and as colour-coded maps of relative growth rates.

Supplementary Data
supp_109_5_897__index.html (955B, html)

ACKNOWLEDGEMENTS

We thank Professor Zygmunt Hejnowicz, Professor Jerzy Nakielski and Dr Agata Burian for discussions during preparation of this manuscript; two anonymous reviewers for helpful comments on the initial version of the manuscript; Dr Ewa Teper from the Laboratory of Scanning Electron Microscopy, Faculty of Earth Sciences, University of Silesia, for help in the preparation of SEM micrographs; and Professor James Murray and Dr Walter Dewitte for the seeds of triple cycD3 mutant. This work was supported by the Polish Ministry of Science and Higher Education [grant number N N303 3917 33] in 2007–2010.

LITERATURE CITED

  1. Beemster GTS, Vusser KD, Tavernier ED, Bock KD, Inzé D. Variation in growth rate between Arabidopsis ecotypes is correlated with cell division and A-type cyclin-dependent kinase activity. Plant Physiology. 2002;129:854–864. doi: 10.1104/pp.002923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Boudolf V, Vlieghe K, Beemster GTS, et al. The plant-specific cyclin-dependent kinase CDKB1;1 and transcription factor E2Fa-DPa control the balance of mitotically dividing and endoreduplicationg cells in Arabidopsis. The Plant Cell. 2004;16:2683–2692. doi: 10.1105/tpc.104.024398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Coen E, Rolland-Lagan A-L, Matthews M, Bangham JA, Prusinkiewicz P. The genetics of geometry. Proceedings of the National Academy of Sciences, USA. 2004;101:4728–4735. doi: 10.1073/pnas.0306308101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cookson SJ, van Lijsebettens M, Granier C. Correlation between leaf growth variables suggest intrinsic and early controls of leaf size in Arabidopsis thaliana. Plant, Cell and Environment. 2005;28:1355–1366. [Google Scholar]
  5. Cookson SJ, Radziejwoski A, Granier C. Cell and leaf size plasticity in Arabidopsis: what is the role of endoreduplication? Plant, Cell and Environment. 2006;29:1273–1283. doi: 10.1111/j.1365-3040.2006.01506.x. [DOI] [PubMed] [Google Scholar]
  6. Corson F, Hamant O, Bohn S, Traas J, Boudaoud A, Couder Y. Turning a plant tissue into a living cell froth: a case of isotropic growth in plants. Proceedings of the National Academy of Sciences, USA. 2009;106:8453–8458. doi: 10.1073/pnas.0812493106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. De Veylder L, Beeckman T, Beemster GTS, et al. Functional analysis of cyclin-dependent kinase inhibitors of Arabidopsis. The Plant Cell. 2001;13:1653–1667. doi: 10.1105/TPC.010087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dewitte W, Riou-Kamlichi C, Scofield S, et al. Altered cell cycle distribution, hyperplasia, and inhibited differentiation in Arabidopsis caused by the D-type cyclin CYCD3. The Plant Cell. 2003;15:79–92. doi: 10.1105/tpc.004838. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dewitte W, Scofield S, Alcasabas AA, et al. Arabidopsis CYCD3 D-type cyclins link cell proliferation and endocycles and are rate-limiting for cytokinin responses. Proceedings of the National Academy of Sciences, USA. 2007;104:14537–14542. doi: 10.1073/pnas.0704166104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Donnelly PM, Bonetta D, Tsukaya H, Dengler RE, Dengler NG. Cell cycling and cell enlargement in developing leaves of Arabidopsis. Developmental Biology. 1999;215:407–419. doi: 10.1006/dbio.1999.9443. [DOI] [PubMed] [Google Scholar]
  11. Doonan J. Social controls on cell proliferation in plants. Current Opinion in Plant Biology. 2000;3:482–487. doi: 10.1016/s1369-5266(00)00117-5. [DOI] [PubMed] [Google Scholar]
  12. Dumais J, Kwiatkowska D. Analysis of surface growth in shoot apices. The Plant Journal. 2002;31:229–241. doi: 10.1046/j.1365-313x.2001.01350.x. [DOI] [PubMed] [Google Scholar]
  13. Erickson RO. Symplastic growth and symplasmic transport. Plant Physiology. 1986;82:1153. doi: 10.1104/pp.82.4.1153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Evert RF. Esau's plant anatomy. Meristems, cells, and tissues of the plant body: their structure, function, and development. Hoboken, NJ: John Willey & Sons; 2005. [Google Scholar]
  15. Fleming A. Cell cycle control during leaf development. In: Inzé D, editor. Cell cycle control and plant development. Oxford: Blackwell Publishing; 2007. pp. 203–226. [Google Scholar]
  16. Fosket DE. Plant growth and development. A molecular approach. San Diego: Academic Press; 1994. [Google Scholar]
  17. Fu Y, Gu Y, Zheng Z, Wasteneys G, Yang Z. Arabidopsis interdigitating cell growth requires two antagonistic pathways with opposing action on cell morphogenesis. Cell. 2005;120:687–700. doi: 10.1016/j.cell.2004.12.026. [DOI] [PubMed] [Google Scholar]
  18. Galbraith DW, Harkins KR, Knapp S. Systemic endopolyploidy in Arabidopsis thaliana. Plant Physiology. 1991;96:985–989. doi: 10.1104/pp.96.3.985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Geisler MJ, Sack FD. Variable timing of developmental progression in the stomatal pathway in Arabidopsis cotyledons. New Phytologist. 2002;153:469–476. doi: 10.1046/j.0028-646X.2001.00332.x. [DOI] [PubMed] [Google Scholar]
  20. Granier C, Tardieu F. Spatial and temporal analyses of expansion and cell cycle in sunflower leaves. Plant Physiology. 1998;116:991–1001. doi: 10.1104/pp.116.3.991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Granier C, Tardieu F. Multi-scale phenotyping of leaf expansion in response to environmental changes: the whole is more than the sum of parts. Plant, Cell and Environment. 2009;32:1175–1184. doi: 10.1111/j.1365-3040.2009.01955.x. [DOI] [PubMed] [Google Scholar]
  22. Gunning BES, Hughes JE, Hardham AR. Formative and proliferative cell divisions, cell differentiation, and developmental changes in the meristem of Azolla roots. Planta. 1978;143:121–144. doi: 10.1007/BF00387786. [DOI] [PubMed] [Google Scholar]
  23. Hejnowicz Z. Plants as mechano-osmotic transducers. In: Wojtaszek P, editor. Mechanical integration of plant cells and plants. Berlin: Springer-Verlag; 2011. pp. 241–268. [Google Scholar]
  24. Jarvis MC, Briggs SPH, Knox JP. Intercellular adhesion and cell separation in plants. Plant, Cell and Environment. 2003;26:977–989. [Google Scholar]
  25. Johansen DA. Plant microtechnique. New York: McGraw-Hill; 1940. [Google Scholar]
  26. John PCL, Qi R. Cell division and endoreduplication: doubtful engines of vegetative growth. Trends in Plant Science. 2008;13:121–127. doi: 10.1016/j.tplants.2008.01.004. [DOI] [PubMed] [Google Scholar]
  27. Joubès J, Chevalier Ch. Endoreduplication in higher plants. Plant Molecular Biology. 2000;4:735–745. doi: 10.1023/a:1006446417196. [DOI] [PubMed] [Google Scholar]
  28. Kagan ML, Novoplansky N, Sachs T. Variable cell lineages form the functional pea epidermis. Annals of Botany. 1992;69:303–312. [Google Scholar]
  29. Kotzer AM, Wasteneys GO. Mechanisms behind the puzzle: microtubule–microfilament cross-talk in pavement cell formation. Canadian Journal of Botany. 2006;84:594–603. [Google Scholar]
  30. Kwiatkowska D, Dumais J. Growth and morphogenesis at the vegetative shoot apex of Anagallis arvensis L. Journal of Experimental Botany. 2003;54:1585–1595. doi: 10.1093/jxb/erg166. [DOI] [PubMed] [Google Scholar]
  31. Massonnet C, Vile D, Fabre J, et al. Probing the reproducibility of leaf growth and molecular phenotypes: a comparison of three Arabidopsis accessions cultivated in ten laboratories. Plant Physiology. 2010;152:2142–2157. doi: 10.1104/pp.109.148338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Matsubara S, Hurry V, Druart N, et al. Nocturnal changes in leaf growth of Populus deltoides are controlled by cytoplasmic growth. Planta. 2006;223:1315–1328. doi: 10.1007/s00425-005-0181-0. [DOI] [PubMed] [Google Scholar]
  33. Melaragno JE, Mehrotra B, Coleman AW. Relationship between endopolyploidy and cell size in epidermal tissue of Arabidopsis. The Plant Cell. 1993;5:1661–1668. doi: 10.1105/tpc.5.11.1661. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Nakielski J. The tensor-based model for growth and cell divisions of the root apex. I. The significance of principal directions. Planta. 2008;228:179–189. doi: 10.1007/s00425-008-0728-y. [DOI] [PubMed] [Google Scholar]
  35. Nath U, Crawford BCW, Carpenter R, Coen E. Genetic control of surface curvature. Science. 2003;299:1404–1407. doi: 10.1126/science.1079354. [DOI] [PubMed] [Google Scholar]
  36. Palatnik JF, Allen E, Wu X, et al. Control of leaf morphogenesis by microRNAs. Nature. 2003;425:257–263. doi: 10.1038/nature01958. [DOI] [PubMed] [Google Scholar]
  37. Panteris E, Galatis B. The morphogenesis of lobed plant cells in the mesophyll and epidermis: organization and distinct roles of cortical microtubules and actin filaments. New Phytologist. 2005;167:721–732. doi: 10.1111/j.1469-8137.2005.01464.x. [DOI] [PubMed] [Google Scholar]
  38. Panteris E, Apostolakos P, Galatis B. Sinuous ordinary epidermal cells: behind several patterns of waviness, a common morphogenetic mechanism. New Phytologist. 1994;127:771–780. doi: 10.1111/j.1469-8137.1994.tb02981.x. [DOI] [PubMed] [Google Scholar]
  39. Poethig RS, Sussex IM. The developmental morphology and growth dynamics of the tobacco leaf. Planta. 1985;165:158–169. doi: 10.1007/BF00395038. [DOI] [PubMed] [Google Scholar]
  40. Priestley JH. Studies in the physiology of cambial activity. II. The concept of sliding growth. New Phytologist. 1930;29:96–140. [Google Scholar]
  41. Pyke KA, Marrison JL, Leech RM. Temporal and spatial development of the cells of the expanding first leaf of Arabidopsis thaliana (L.) Heynh. Journal of Experimental Botany. 1991;42:1407–1416. [Google Scholar]
  42. Richards OW, Kavanagh AJ. The analysis of the relative growth gradients and changing form of growing organisms Illustrated by tobacco leaf. American Naturalist. 1943;77:385–399. [Google Scholar]
  43. Roeder AHK, Chickarmane V, Cunha A, Obara B, Manjunath BS, Meyerowitz EM. Variability in the control of cell division underlies sepal epidermal patterning in Arabidopsis thaliana. PLoS Biology. 2010;8:e1000367. doi: 10.1371/journal.pbio.1000367. http://dx.doi.org/10.1371/journal.pbio.1000367 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Routier-Kierzkowska A-L, Kwiatkowska D. New stereoscopic reconstruction protocol for scanning electron microscope images and its application to in vivo replicas of the shoot apical meristem. Functional Plant Biology. 2008;35:1034–1046. doi: 10.1071/FP08047. [DOI] [PubMed] [Google Scholar]
  45. Sugimoto-Shirasu K, Roberts K. ‘Big it up’: endoreduplication and cell-size control in plants. Current Opinion in Plant Biology. 2003;6:544–553. doi: 10.1016/j.pbi.2003.09.009. [DOI] [PubMed] [Google Scholar]
  46. Traas J, Hülskamp M, Gendreau E, Höfte H. Endoreduplication and development: rule without dividing? Current Opinion in Plant Biology. 1998;1:498–503. doi: 10.1016/s1369-5266(98)80042-3. [DOI] [PubMed] [Google Scholar]
  47. Tsukaya H. Interpretation of mutants in leaf morphology: genetic evidence for a compensatory system in leaf morphogenesis that provides a new link between cell and organismal theories. International Review of Cytology. 2002;217:1–39. doi: 10.1016/s0074-7696(02)17011-2. [DOI] [PubMed] [Google Scholar]
  48. Tsukaya H. Controlling size in multicellular organs: focus on the leaf. PloS Biology. 2008;6:e174. doi: 10.1371/journal.pbio.0060174. http://dx.doi.org/10.1371/journal.pbio.0060174 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Vlieghe K, Inzé D, De Veylder L. Physiological relevance and molecular control of the endocycle in plants. In: Inzé D, editor. Cell cycle control and plant development. Oxford: Blackwell Publishing; 2007. pp. 227–248. [Google Scholar]
  50. Walter A, Schurr U. Dynamics of leaf and root growth: endogenous control versus environmental impact. Annals of Botany. 2005;95:891–900. doi: 10.1093/aob/mci103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Walter A, Silk WK, Schurr U. Environmental effects on spatial and temporal patterns of leaf and root growth. Annual Review of Plant Biology. 2009;60:279–304. doi: 10.1146/annurev.arplant.59.032607.092819. [DOI] [PubMed] [Google Scholar]
  52. White DWR. PEAPOD regulates lamina size and curvature in Arabidopsis. Proceedings of the National Academy of Sciences, USA. 2006;103:13238–13243. doi: 10.1073/pnas.0604349103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wiese A, Christ MM, Virnich O, Schurr U, Walter A. Spatio-temporal leaf growth patterns of Arabidopsis thaliana and evidence for sugar control of the diel leaf growth cycle. New Phytologist. 2007;174:752–761. doi: 10.1111/j.1469-8137.2007.02053.x. [DOI] [PubMed] [Google Scholar]
  54. Williams MH, Green PB. Sequential scanning electron microscopy of a growing plant meristem. Protoplasma. 1988;147:77–79. [Google Scholar]
  55. Wolf SU, Silk WK, Plant RE. Quantitative patterns of leaf expansion: comparison of normal and malformed leaf growth in Vitis vinifera cv. Ruby Red. American Journal of Botany. 1986;73:832–846. [Google Scholar]
  56. Zar JH. Biostatistical analysis. New Jersey: Prentice-Hall; 1999. [Google Scholar]
  57. Zhang C, Hasley LE, Szymanski DB. The development and geometry of shape change in Arabidopsis thaliana cotyledon pavement cells. BMC Plant Biology. 2011;11:27. doi: 10.1186/1471-2229-11-27. http://dx.doi.org/10.1186/1471-2229-11-27 . [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_109_5_897__index.html (955B, html)
supp_mcs005_mcs005supp.pdf (86.5KB, pdf)

Articles from Annals of Botany are provided here courtesy of Oxford University Press

RESOURCES