Body size varies widely among species, populations and individuals, depending on the environment. Transitioning between proliferation and differentiation is a crucial determinant of final organ size, but how the timing of this transition is established and maintained remains unknown. Using cell proliferation markers and genetic analysis, we show that CHIQUITA1 (CHIQ1) is required to maintain the timing of the transition from proliferation to differentiation in Arabidopsis thaliana. Combining kinematic and cell lineage-tracking studies, we found that the number of actively dividing cells in chiquita1-1 plants decreases prematurely compared with wild-type plants, suggesting CHIQ1 maintains the proliferative capacity in dividing cells and ensures that cells divide a specific number of times. CHIQ1 belongs to a plant-specific gene family of unknown molecular function and genetically interacts with three close members of its family to control the timing of proliferation exit. Our work reveals the interdependency between cellular and organ-level processes underlying final organ size determination.

Body size control is fundamental for growth and development, and can impact reproductive fitness and ecosystem structure (Aarssen, 2015; Luan et al., 2020). Understanding how plants control their body size has direct implications on agricultural productivity and resource use. Although body size control is well established for insects and mammals (Penzo-Méndez and Stanger, 2015; Texada et al., 2020), much less is known for plants (Karamat et al., 2021).

Body size is driven by the rate and duration of organ growth (Texada et al., 2020), which determines the number of cell divisions and how large cells become. The final number of cells in an organ plays an important role in determining final organ size (Conlon and Raff, 1999; Gázquez and Beemster, 2017), and derives from the rate and duration of cell proliferation (Gonzalez et al., 2012). In animals, cell death also contributes to the final cell number in an organ (Conlon and Raff, 1999). The rate of cell proliferation within an organ depends on the number of dividing cells at any given time and the length of the cell cycle. Cell proliferation ends when all cells have exited the cell cycle. The genetic basis that controls proliferation exit during development is still unknown (Vollmer et al., 2017).

Three cell-intrinsic mechanisms of proliferation exit (i.e. intracellular mechanisms that stops proliferation; Conlon and Raff, 1999) have been proposed to control the timing of the transition from proliferation to differentiation: (1) a timer mechanism (cells divide for a fixed period of time); (2) a counter mechanism (cells undergo a fixed number of divisions); and (3) a sizer mechanism (an optimal threshold cell size is required for cell cycle progression). Exit from proliferation is also controlled by systemic mechanisms at the organ or organism level, including long-distance signaling and cell-to-cell contact interactions (Texada et al., 2020). In animals, organs and organisms have evolved different strategies to control when cells exit proliferation (Hayflick and Moorhead, 1961; Durand et al., 1998; Burton et al., 1999; Stanger et al., 2007; Lui and Baron, 2011; Penzo-Méndez and Stanger, 2015; Pickering et al., 2018; Chen et al., 2019; Neurohr et al., 2019), indicating that no universal mechanism controls the timing of proliferation exit. For example, a timer composed of at least two cyclin-dependent kinase inhibitors controls how many times an oligodendrocyte precursor (Durand et al., 1998) and a myocyte divide (Burton et al., 1999). A cell sizer mechanism controls the number of cell divisions in Xenopus laevis embryos before the onset of the zygotic genome activation (ZGA) (Chen et al., 2019). Before ZGA, cells divide without growing, which reduces cytoplasmic volume and increases DNA concentration and the DNA:cytoplasm ratio. When individual cells in X. laevis embryos reach a threshold size and DNA:cytoplasm ratio, ZGA is triggered, leading to the onset of germ-layer specification and cell differentiation (Chen et al., 2019). In plants, although a cell cycle counter mechanism has been proposed for developing energy sink organs such as flowers and roots (Sun et al., 2009; Kumpf et al., 2014), it is not known which mechanisms operate to control proliferation exit of the major energy source organ: the leaf.

We previously used a bioinformatic pipeline to find uncharacterized transcriptional regulators and identified CHIQUITA 1 (CHIQ1), a gene of unknown function involved in organ size control in Arabidopsis thaliana (Bossi et al., 2017). CHIQ1 physically interacts with other CHIQ-like proteins (Bossi et al., 2017) and the transcriptional repressor Polycomb repressive complex 2 (PRC2) via another member of the CHIQ family (Bossi et al., 2017); but we do not know whether CHIQ proteins alter PRC2 function. The predicted transcriptional regulator function of CHIQ1 and other CHIQ proteins has not yet been proven. More recently, CHIQ1 [named CONSTITUTIVELY STRESSED 1 (COST1)] was implicated in the regulation of autophagy in response to drought stress (Bao et al., 2020). However, the role of CHIQ1/COST1 during growth is independent of autophagy (Bao et al., 2020). Here, we show that CHIQ1 is a positive regulator of body size by measuring organ growth at the cellular, organ and organismal levels using genetic and chemical perturbations combined with live 4D imaging. Genetic interaction studies indicated that CHIQ1 works with three other previously uncharacterized CHIQ-like genes [CHIQUITA 1-LIKE 4 (CHIQL4), CHIQUITA 1-LIKE 5 (CHIQL5), and CHIQUITA 1-LIKE 6 (CHIQL6)] to control the timing of transition from proliferation to differentiation and final cell number in leaves. Kinematic and cell lineage-tracking analyses indicate that CHIQ1 maintains proliferative capacity in dividing cells and imply that CHIQ1 ensures that cells divide a certain number of times before entering differentiation. Finally, we show that four members of the CHIQ family act together as positive regulators of cell proliferation during organ growth and are required for proper timing of cell cycle exit in leaves.

CHIQ1 is a positive regulator of vegetative and reproductive organ size

CHIQ1 encodes a protein of unknown function, which is required for Arabidopsis thaliana to reach full size at maturity (Bossi et al., 2017). Previously, we showed that the lines carrying the null allele chiq1-1 had smaller leaves and stems than wild-type plants (Bossi et al., 2017). To determine whether other major organs, such as flowers, fruits and roots, exhibited similar size reduction, we examined them at various stages of development. At maturity, these organs were significantly smaller than those in the wild type (Fig. 1). However, during the first week of growth following germination, where most growth occurs by cell division, chiq1-1 organs were similar or even bigger than wild-type organs (Fig. 2, Fig. S1). Primary roots of chiq1-1 seedlings were ∼10% longer than those of wild-type plants 3-5 days after sowing. However, chiq1-1 roots became ∼20-40% shorter than wild-type roots as the seedlings aged (Figs 2A and 1B). During rosette growth, chiq1-1 leaves ranged from being larger than the wild type (1st and 2nd leaves) to being similar to the wild type (3rd-7th leaves), but all chiq1-1 leaves became smaller than wild type leaves as the plants aged (Fig. 2C,D, Fig. S1). Introducing wild-type CHIQ1 into the mutant background complemented the organ size phenotypes observed in the chiq1-1 null mutant (Figs 1 and 2).

Fig. 1.

chiq1-1 mature organs are smaller than wild type. (A) 21-day-old plants grown in soil using a rhizotron system (Rellan-Alvarez et al., 2015) [wild type (left), chiq1-1 (middle) and a complemented line (35Spro:CHIQ1-FLAG (right)]. Scale bar: 5 cm. (B) Length of roots of wild type (Col-0, blue), chiq1-1 (orange) and one complemented line (35Spro:CHIQ1-FLAG, gray) grown in soil for 21 days using a rhizotron system (Rellan-Alvarez et al., 2015) (n=20-21 plants per genotype; N=3 independent experiments). Letters represent significantly different groups (P<0.05) as determined by a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. (C) Rosette leaves from 40-day-old wild type (left) and chiq1-1 (right) plants. Scale bars: 2 cm. (D) Flowers from wild type (left), chiq1-1 (middle) and a complemented line (35Spro:CHIQ1-FLAG, right). Scale bars: 1 mm. (E) Fruits from wild type (top), chiq1-1 (middle) and a complemented line (35Spro:CHIQ1-FLAG, bottom). Scale bar: 2 mm. (F) Leaf blade area of mature leaves from soil-grown wild type (blue), chiq1-1 (orange) and two complemented lines (CHIQ1pro:CHIQ1-YFP line, dark-gray bar; 35Spro:CHIQ1-FLAG, light-gray bar) (n=6-71 plants per genotype per numbered leaf; N=2 independent experiments). The blade area of each leaf number (see Fig. 2B for the numbering scheme) was compared among the four genotypes using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. The red asterisks indicate statistical significance in the chiq1-1 leaf compared with the other three genotypes (P<0.05). (B,F) Graphs were generated using the ggplot2 package in R (Wickham, 2009). Boxes represent interquartile range (IQR) where the two edges of the box are 25th (Q1) and 75th (Q3) percentiles and the line in the box 50th percentile (median). The ends of the whiskers represent minimum (Q1−1.5×IQR) and maximum (Q3+1.5×IQR). Dots represent outliers.

Fig. 1.

chiq1-1 mature organs are smaller than wild type. (A) 21-day-old plants grown in soil using a rhizotron system (Rellan-Alvarez et al., 2015) [wild type (left), chiq1-1 (middle) and a complemented line (35Spro:CHIQ1-FLAG (right)]. Scale bar: 5 cm. (B) Length of roots of wild type (Col-0, blue), chiq1-1 (orange) and one complemented line (35Spro:CHIQ1-FLAG, gray) grown in soil for 21 days using a rhizotron system (Rellan-Alvarez et al., 2015) (n=20-21 plants per genotype; N=3 independent experiments). Letters represent significantly different groups (P<0.05) as determined by a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. (C) Rosette leaves from 40-day-old wild type (left) and chiq1-1 (right) plants. Scale bars: 2 cm. (D) Flowers from wild type (left), chiq1-1 (middle) and a complemented line (35Spro:CHIQ1-FLAG, right). Scale bars: 1 mm. (E) Fruits from wild type (top), chiq1-1 (middle) and a complemented line (35Spro:CHIQ1-FLAG, bottom). Scale bar: 2 mm. (F) Leaf blade area of mature leaves from soil-grown wild type (blue), chiq1-1 (orange) and two complemented lines (CHIQ1pro:CHIQ1-YFP line, dark-gray bar; 35Spro:CHIQ1-FLAG, light-gray bar) (n=6-71 plants per genotype per numbered leaf; N=2 independent experiments). The blade area of each leaf number (see Fig. 2B for the numbering scheme) was compared among the four genotypes using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. The red asterisks indicate statistical significance in the chiq1-1 leaf compared with the other three genotypes (P<0.05). (B,F) Graphs were generated using the ggplot2 package in R (Wickham, 2009). Boxes represent interquartile range (IQR) where the two edges of the box are 25th (Q1) and 75th (Q3) percentiles and the line in the box 50th percentile (median). The ends of the whiskers represent minimum (Q1−1.5×IQR) and maximum (Q3+1.5×IQR). Dots represent outliers.

Fig. 2.

chiq1-1 young organs are of similar size or slightly bigger than in wild type. (A) Scatter plot of normalized primary root length against the mean wild-type value of the same age in wild type (Col-0, blue), chiq1-1 (orange) and a complemented line (35Spro:CHIQ1-FLAG, gray) plants grown in media plates from 3 to 14 days after sowing (n=92-131 plants per genotype per time point; N=3 independent experiments). Error bars represent 95% confidence interval. The regression line corresponds to a power regression. All genotypes were compared at each time point using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Red asterisks represent a significant difference between chiq1-1 and the other two genotypes (P<0.05). (B) Schematic representation of leaf growth in the Arabidopsis rosette over time. (C,D) Scatter plot of normalized leaf blade area against the mean wild-type value of the same age of the first pair of leaves (C) and the 7th (D) leaf in soil-grown plants ranging in age from 4 days to 36 days from wild type (blue), chiq1-1 (orange) and two complemented lines (CHIQ1pro:CHIQ1-YFP line, dark gray; 35Spro:CHIQ1-FLAG, light gray) [n=18-108 plants per genotype; N=1 or 2 independent experiments (1st leaf) and n=7-21 plants per genotype; N=1 independent experiment (7th leaf)]. Error bars represent 95% confidence interval. The regression line corresponds to a power regression. All genotypes were compared at each time point using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Red asterisks represent a significant difference between chiq1-1 and the other three genotypes (P<0.05). Graphs were generated using the ggplot2 package in R (Wickham, 2009).

Fig. 2.

chiq1-1 young organs are of similar size or slightly bigger than in wild type. (A) Scatter plot of normalized primary root length against the mean wild-type value of the same age in wild type (Col-0, blue), chiq1-1 (orange) and a complemented line (35Spro:CHIQ1-FLAG, gray) plants grown in media plates from 3 to 14 days after sowing (n=92-131 plants per genotype per time point; N=3 independent experiments). Error bars represent 95% confidence interval. The regression line corresponds to a power regression. All genotypes were compared at each time point using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Red asterisks represent a significant difference between chiq1-1 and the other two genotypes (P<0.05). (B) Schematic representation of leaf growth in the Arabidopsis rosette over time. (C,D) Scatter plot of normalized leaf blade area against the mean wild-type value of the same age of the first pair of leaves (C) and the 7th (D) leaf in soil-grown plants ranging in age from 4 days to 36 days from wild type (blue), chiq1-1 (orange) and two complemented lines (CHIQ1pro:CHIQ1-YFP line, dark gray; 35Spro:CHIQ1-FLAG, light gray) [n=18-108 plants per genotype; N=1 or 2 independent experiments (1st leaf) and n=7-21 plants per genotype; N=1 independent experiment (7th leaf)]. Error bars represent 95% confidence interval. The regression line corresponds to a power regression. All genotypes were compared at each time point using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Red asterisks represent a significant difference between chiq1-1 and the other three genotypes (P<0.05). Graphs were generated using the ggplot2 package in R (Wickham, 2009).

To determine whether the reduced organ size derives from a reduction in cell size or cell number, we first measured these parameters in plants at maturity. Fully expanded leaves in chiq1-1 had both fewer and smaller cells (Fig. 3A,B,E,G, Fig. S2). The same was observed in petals (Fig. 3C,D,F,H). As smaller cells could result from defects in endoreduplication (Melaragno et al., 1993; Katagiri et al., 2016), we asked whether endoreduplication was affected in chiq1-1 plants. The majority (59%) of chiq1-1 cells were diploid (2C) in mature leaves compared with 31% in wild-type cells (Fig. 3I). This indicates that the majority of chiq1-1 leaf cells did not enter endoreduplication after exiting the mitotic cell cycle, which could explain the reduced cell size observed in chiq1-1 leaves at maturity. Although most chiq1-1 cells skip endoreduplication, suggesting the transition from mitosis into endoreduplication is disrupted, pavement cells display the typical jigsaw puzzle shape in mature leaves, which suggests that terminal differentiation is not compromised in chiq1-1 plants. Trichome shape and size was not affected in chiq1-1 leaves (Fig. S3), suggesting CHIQ1 does not play a role during trichome terminal differentiation. However, the chiq1-1 seventh leaf had fewer trichomes (Fig. S3), which is consistent with the reduced number of the other epidermal cell types.

Fig. 3.

chiq1-1 mature organs have fewer and smaller cells, and cells undergo less endoreduplication in chiq1-1 leaves. (A-D) Scanning electron microscopy (SEM) images of the adaxial epidermis of the 7th leaf (A,B) and petal (C,D) from 51-day-old wild type (A,C) and chiq1-1 (B,D) plants. (E,G) Normalized average cell size (E) and total cell number (G) in the abaxial epidermis of the 7th leaf against the wild-type value of the same experiment from 36-day-old wild-type (Col-0, blue), chiq1-1 (orange) and a complemented line (35Spro:CHIQ1-FLAG, gray) plants (n=12-24 plants per genotype; N=4 independent experiments). (F,H) Normalized average cell size (F) and total cell number (H) of petal conical cells from wild type (Col-0, blue) and chiq1-1 (orange) open flowers (n=13 or 14 plants per genotype; N=3 independent experiments). Letters in E-H represent significantly different groups (P<0.05), as determined by two-way analysis of variance followed by a post-hoc Tukey's test. (I) Mean percentage of cells with each ploidy level in mature leaves from wild type (Col-0, blue), chiq1-1 (orange) and a complemented line (35Spro:CHIQ1-FLAG, gray) [n=4 samples per genotype (2 technical and 2 biological replicates, each biological replicate pooled from 60 plants per genotype); N=2 independent experiments]. Error bars represent 95% confidence interval. Letters represent significantly different groups (P<0.05), as determined by a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Graphs were generated using the ggplot2 package in R (Wickham, 2009). Boxes represent interquartile range (IQR) where the two edges of the box are 25th (Q1) and 75th (Q3) percentiles and the line in the box 50th percentile (median). The ends of the whiskers represent minimum (Q1−1.5×IQR) and maximum (Q3+1.5×IQR). Dots represent outliers.

Fig. 3.

chiq1-1 mature organs have fewer and smaller cells, and cells undergo less endoreduplication in chiq1-1 leaves. (A-D) Scanning electron microscopy (SEM) images of the adaxial epidermis of the 7th leaf (A,B) and petal (C,D) from 51-day-old wild type (A,C) and chiq1-1 (B,D) plants. (E,G) Normalized average cell size (E) and total cell number (G) in the abaxial epidermis of the 7th leaf against the wild-type value of the same experiment from 36-day-old wild-type (Col-0, blue), chiq1-1 (orange) and a complemented line (35Spro:CHIQ1-FLAG, gray) plants (n=12-24 plants per genotype; N=4 independent experiments). (F,H) Normalized average cell size (F) and total cell number (H) of petal conical cells from wild type (Col-0, blue) and chiq1-1 (orange) open flowers (n=13 or 14 plants per genotype; N=3 independent experiments). Letters in E-H represent significantly different groups (P<0.05), as determined by two-way analysis of variance followed by a post-hoc Tukey's test. (I) Mean percentage of cells with each ploidy level in mature leaves from wild type (Col-0, blue), chiq1-1 (orange) and a complemented line (35Spro:CHIQ1-FLAG, gray) [n=4 samples per genotype (2 technical and 2 biological replicates, each biological replicate pooled from 60 plants per genotype); N=2 independent experiments]. Error bars represent 95% confidence interval. Letters represent significantly different groups (P<0.05), as determined by a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Graphs were generated using the ggplot2 package in R (Wickham, 2009). Boxes represent interquartile range (IQR) where the two edges of the box are 25th (Q1) and 75th (Q3) percentiles and the line in the box 50th percentile (median). The ends of the whiskers represent minimum (Q1−1.5×IQR) and maximum (Q3+1.5×IQR). Dots represent outliers.

Taken together, our data show that CHIQ1 affects final organ size by modulating cell proliferation and expansion. We next investigated the role of CHIQ1 during cell proliferation and the transition from cell proliferation into cell expansion.

CHIQ1 maintains cell number and cell size during development

To determine what led to the late onset dwarfism in chiq1-1, we compared how cell proliferation and expansion changed over time between chiq1-1 and wild-type plants by employing a kinematic analysis on the first pair to leaves from 4 to 25 days after sowing. Total cell number was similar in both genotypes from day 4 to day 15 after sowing and became significantly lower at days 19 and 25 after sowing in chiq1-1 leaves compared with the wild type (Fig. 4A). The number of cells in the root apical meristem (RAM) of chiq1-1 decreased similarly over time (Fig. S4). The stomatal index (SI) – which represents the proportion of guard cells (specialized cells involved in gas exchange) among all cells in the leaf epidermis – was similar in both genotypes (Fig. S5). This indicates CHIQ1 does not affect the pattern of cell type differentiation in the epidermis.

Fig. 4.

Fewer cells divide in chiq1-1, and chiq1-1 cells exit proliferation prematurely in growing organs. (A,B) Kinematic studies in the first pair of true leaves from wild-type (Col-0, blue) and chiq1-1 (orange) seedlings grown on soil. Estimated total cell number (A) and average cell area (B) in the abaxial epidermis from wild-type (Col-0, blue) and chiq1-1 (orange) seedlings at 4, 6, 8, 11, 13, 15, 19 and 25 days after sowing (n=5-30 plants per genotype per time point; N=1-3 independent experiments). Letters represent significantly different groups (P<0.05) as determined by a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Error bars represent 95% confidence interval. (C) Expression of CYCD3.3pro:GUS in wild-type (Col-0, left) and chiq1-1 (right) first true leaves grown for 7, 8 and 9 days. Every image is representative of two independent lines. Ten leaves were analyzed per genotype per experiment. Scale bars: 0.5 mm. (D) Expression of CDKB1.1pro:GUS in the 4th and 5th leaves of wild-type (Col-0, left) and chiq1-1 (right) plants at days 16, 20 and 24 after sowing. Every leaf is a representative image of two independent experiments and four leaves were analyzed per genotype per experiment. Scale bars: 2 mm. (E) Expression of the cell cycle marker CYCBpro:CYCBdestruction box-GFP in the root apical meristem of 5-day-old wild-type (left panel) and chiq1-1 (right panel) seedlings stained with propidium iodide. Scale bars: 100 um. (F) Left panel: schematic representation of the experimental design and cell lineage tracking analysis in leaves. The first pair of leaves was imaged daily from day 6 to 8 after sowing. Images were taken from the whole leaf and cells in the middle region of the leaf blade were traced and followed for 2 days (more details in text). In this representation, cells with the same color belong to the same lineage. Right panel: cell cycle length from day 6 to day 8 in leaves from wild-type (Col-0, blue), chiq1-1 (orange) and a complemented line (35Spro:CHIQ1-FLAG, gray) seedlings (n=17-35 plants per genotype; N=3-7 independent experiments). Letters represent significantly different groups (P<0.05) as determined by two-way analysis of variance followed by post-hoc Tukey's test. Graphs were generated using the ggplot2 package in R (Wickham, 2009). Boxes represent interquartile range (IQR) where the two edges of the box are 25th (Q1) and 75th (Q3) percentiles and the line in the box 50th percentile (median). The ends of the whiskers represent minimum (Q1−1.5×IQR) and maximum (Q3+1.5×IQR). Dots represent outliers.

Fig. 4.

Fewer cells divide in chiq1-1, and chiq1-1 cells exit proliferation prematurely in growing organs. (A,B) Kinematic studies in the first pair of true leaves from wild-type (Col-0, blue) and chiq1-1 (orange) seedlings grown on soil. Estimated total cell number (A) and average cell area (B) in the abaxial epidermis from wild-type (Col-0, blue) and chiq1-1 (orange) seedlings at 4, 6, 8, 11, 13, 15, 19 and 25 days after sowing (n=5-30 plants per genotype per time point; N=1-3 independent experiments). Letters represent significantly different groups (P<0.05) as determined by a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Error bars represent 95% confidence interval. (C) Expression of CYCD3.3pro:GUS in wild-type (Col-0, left) and chiq1-1 (right) first true leaves grown for 7, 8 and 9 days. Every image is representative of two independent lines. Ten leaves were analyzed per genotype per experiment. Scale bars: 0.5 mm. (D) Expression of CDKB1.1pro:GUS in the 4th and 5th leaves of wild-type (Col-0, left) and chiq1-1 (right) plants at days 16, 20 and 24 after sowing. Every leaf is a representative image of two independent experiments and four leaves were analyzed per genotype per experiment. Scale bars: 2 mm. (E) Expression of the cell cycle marker CYCBpro:CYCBdestruction box-GFP in the root apical meristem of 5-day-old wild-type (left panel) and chiq1-1 (right panel) seedlings stained with propidium iodide. Scale bars: 100 um. (F) Left panel: schematic representation of the experimental design and cell lineage tracking analysis in leaves. The first pair of leaves was imaged daily from day 6 to 8 after sowing. Images were taken from the whole leaf and cells in the middle region of the leaf blade were traced and followed for 2 days (more details in text). In this representation, cells with the same color belong to the same lineage. Right panel: cell cycle length from day 6 to day 8 in leaves from wild-type (Col-0, blue), chiq1-1 (orange) and a complemented line (35Spro:CHIQ1-FLAG, gray) seedlings (n=17-35 plants per genotype; N=3-7 independent experiments). Letters represent significantly different groups (P<0.05) as determined by two-way analysis of variance followed by post-hoc Tukey's test. Graphs were generated using the ggplot2 package in R (Wickham, 2009). Boxes represent interquartile range (IQR) where the two edges of the box are 25th (Q1) and 75th (Q3) percentiles and the line in the box 50th percentile (median). The ends of the whiskers represent minimum (Q1−1.5×IQR) and maximum (Q3+1.5×IQR). Dots represent outliers.

Differences in cell size between wild-type and chiq1-1 leaves showed similar dynamic patterns over time as the cell number. The average cell area was greater in chiq1-1 at 4-8 days after sowing (Fig. 4B, Fig. S6A), which could explain the ephemerally larger leaf size of the mutant relative to the wild type at early developmental time points (Fig. 2C). Later in development, the average cell area was similar in both genotypes at 11-13 days after sowing, and became smaller in chiq1-1 seedlings, starting at day 15 after sowing (Fig. 4B). Because the leaf epidermis is composed of cells that are orders of magnitude different in size, depending on cell type (e.g. guard versus pavement) and differentiation state (e.g. dividing versus differentiated), we wondered whether examining only one cell type in much more detail would provide further insight. Therefore, a more-detailed analysis of the cell area distribution of epidermis cells, excluding guard cells, was performed. The proportion of small cells – which likely represent dividing cells based on previous studies (Andriankaja et al., 2012; Jones et al., 2017) – was smaller in 8-day-old chiq1-1 leaves compared with wild-type leaves (Fig. S6B). This suggests that the proportion of dividing cells is smaller in 8-day-old chiq1-1 leaves while the proportion of expanding cells is larger than the wild type, which would explain the ephemerally larger leaves in 8-day-old chiq1-1 plants compared with the wild type. These data suggest that cell proliferation is decreasing faster in chiq1-1 organs and that cells may be transitioning prematurely into a differentiated state.

CHIQ1 keeps the timing of exit from cell proliferation

To test the hypothesis that CHIQ1 may be involved in controlling the exit from cell proliferation, we followed expression patterns of cell cycle markers over time in actively growing leaves and roots. For cell cycle markers, we used CYCLIN D3.3 (CYCD3.3), CYCLIN B1.1 (CYCB1.1) and CYCLIN-DEPENDENT KINASE B1.1 (CDKB1.1), which encode components of the cell cycle machinery expressed during cell proliferation (Dewitte et al., 2003; Boudolf et al., 2004; Aki and Umeda, 2016). The domain of expression of CYCD3.3 and CDKB1.1 became more restricted earlier in chiq1-1 leaves (Fig. 4C,D). Similarly, fewer cells expressed CYCB1.1 in the RAM of 5-day-old chiq1-1 seedlings (Fig. 4E). Together, these results suggest that fewer cells were dividing in chiq1-1 organs, supporting the hypothesis of earlier exit from proliferation. Alternatively, as CYCB1.1 is expressed in late-G2/M, chiq1-1 cells may be undergoing a longer cell cycle due to being arrested in the cell cycle phases G1 or S.

Cells divide fewer times before exiting proliferation in chiq1-1 leaves

To distinguish between fewer cells dividing or a longer cell cycle in chiq1-1 leaves, individual cells expressing the fluorescent epidermis-specific plasma membrane marker RCI2A (Roeder et al., 2010) were tracked from epidermis images of intact first true leaves from 6-day-old seedlings every 24 h for 2 days (Fig. 4F). Cell cycle length was estimated from the proliferation rate, which was calculated by counting the cells that divided during the course of the experiment and their progeny. Cells that did not divide during the experiment were not included in the calculation of proliferation rate. The cell cycle length was not significantly different between wild-type, chiq1-1 and the complemented line 35Spro:CHIQ1-FLAG (Fig. 4F), indicating that the rate of cell cycle progression is similar in all genotypes. To further confirm that CHIQ1 does not affect cell cycle progression, we treated plants with hydroxyurea (HU), which inhibits the enzyme ribonucleotide reductase, reduces the amount of dNTPs available for DNA synthesis and delays entry into mitosis (Singh and Xu, 2016). HU increased the cell cycle length in wild-type and chiq1-1 leaves indistinguishably (Fig. S7). All these data together indicate that CHIQ1 does not compromise the rate of cell cycle progression and, more importantly, that the decrease in the cell proliferation rate at the leaf level is not due to a longer cell cycle but to a defect in the timing of proliferation exit. This premature proliferation exit decreases the population of dividing cells in chiq1-1 organs, leading to smaller mature organs with fewer cells in chiq1-1 than in wild-type plants.

CHIQ1 works with other CHIQ-like proteins to control the timing of proliferation exit

Previous work showed that CHIQ1 physically interacts with other CHIQ-like proteins (Bossi et al., 2017). In addition, CHIQ1, CHIQL4, CHIQL5 and CHIQL6 proteins have a similar length and domain composition (Bossi et al., 2017). To test whether CHIQ1, CHIQL4, CHIQL5 and CHIQL6 work together genetically and to understand the role of CHIQ proteins during organ growth, we generated a quadruple mutant called chiq-quad, which lacks CHIQ1, CHIQL4, CHIQL5 and CHIQL6, and analyzed leaf area, cell number and cell size during development and at maturity. Mature leaf area in both the single and quadruple mutants was smaller than that in the wild type (Fig. 5A), but did not differ between chiq1-1 and chiq-quad (Fig. 5A). However, cell number in chiq-quad leaves was further reduced compared with that in chiq1-1 leaves (Fig. 5B, Fig. S8A,B), which is consistent with these CHIQ1-like proteins participating in cell proliferation. Interestingly, the average cell area in chiq-quad leaves was greater than that in chiq1-1 leaves and similar to the wild type (Fig. 5C, Fig. S8C,D), which explains why the leaf area of chiq-quad and chiq1-1 plants was indistinguishable, despite having fewer cells. This supports a compensatory mechanism that could have been triggered as chiq-quad failed to reach a certain cell number threshold (Fujikura et al., 2009). In the growing leaf, a larger proportion of cells exited proliferation earlier in chiq-quad compared with chiq1-1 and wild type, as seen by the earlier decrease in cell number (Fig. 5E). In addition, cell size (Fig. 5F) and stomatal index (Fig. 5G) increased earlier in chiq-quad compared with chiq1-1, suggesting that cells in chiq-quad undergo cell differentiation even earlier than chiq1-1 mutants. These data indicate that CHIQ proteins may be part of a complex that regulates the duration of cell proliferation at the organ level and the timing of cell differentiation onset.

Fig. 5.

CHIQ proteins control the timing of the transition into differentiation. (A) Leaf blade area of mature leaves from soil-grown wild type (Col-0, blue), chiq1-1 (orange) and a quadruple mutant (chiq-quad: chiq1-1;chiql4-1;chiql5-1;chiql6-1, red) (n=17-22 plants per leaf per genotype; N=3 independent experiments). The blade area of each leaf was compared among the three genotypes using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. The red asterisk indicates statistical significance (P<0.05) compared with the wild type. (B,C) Total cell number (B) and average cell size (C) in the abaxial epidermis of the 7th leaf from 36-day-old wild type (Col-0, blue), chiq1-1 (orange) and two independent quadruple lines carrying the plasma membrane marker RCI2A [chiq-quad (8A) (light pink) and chiq-quad (3B) (pink)] (n=11 plants per genotype; N=3 independent experiments). Letters represent significantly different groups (P<0.05), as determined by two-way analysis of variance followed by a post-hoc Tukey's test. (D-G) Kinematic studies in the first pair of true leaves from wild type (Col-0, blue), chiq1-1 (orange) and two independent quadruple lines carrying the plasma membrane marker RCI2A [chiq-quad (8A) (light pink) and chiq-quad (3B) (pink)] grown on soil. Leaf blade area (D), estimated total cell number (E), average cell area (F) and stomatal index (G) in the abaxial epidermis from seedlings at 6, 8, 11 and 25 days after sowing (n=10-23 plants per genotype per time point; N=3 independent experiments). Letters represent significantly different groups (P<0.05), as determined using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Error bars represent 95% confidence interval. Graphs were generated using the ggplot2 package in R (Wickham, 2009). Boxes represent interquartile range (IQR) where the two edges of the box are 25th (Q1) and 75th (Q3) percentiles and the line in the box 50th percentile (median). The ends of the whiskers represent minimum (Q1−1.5×IQR) and maximum (Q3+1.5×IQR). Dots represent outliers.

Fig. 5.

CHIQ proteins control the timing of the transition into differentiation. (A) Leaf blade area of mature leaves from soil-grown wild type (Col-0, blue), chiq1-1 (orange) and a quadruple mutant (chiq-quad: chiq1-1;chiql4-1;chiql5-1;chiql6-1, red) (n=17-22 plants per leaf per genotype; N=3 independent experiments). The blade area of each leaf was compared among the three genotypes using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. The red asterisk indicates statistical significance (P<0.05) compared with the wild type. (B,C) Total cell number (B) and average cell size (C) in the abaxial epidermis of the 7th leaf from 36-day-old wild type (Col-0, blue), chiq1-1 (orange) and two independent quadruple lines carrying the plasma membrane marker RCI2A [chiq-quad (8A) (light pink) and chiq-quad (3B) (pink)] (n=11 plants per genotype; N=3 independent experiments). Letters represent significantly different groups (P<0.05), as determined by two-way analysis of variance followed by a post-hoc Tukey's test. (D-G) Kinematic studies in the first pair of true leaves from wild type (Col-0, blue), chiq1-1 (orange) and two independent quadruple lines carrying the plasma membrane marker RCI2A [chiq-quad (8A) (light pink) and chiq-quad (3B) (pink)] grown on soil. Leaf blade area (D), estimated total cell number (E), average cell area (F) and stomatal index (G) in the abaxial epidermis from seedlings at 6, 8, 11 and 25 days after sowing (n=10-23 plants per genotype per time point; N=3 independent experiments). Letters represent significantly different groups (P<0.05), as determined using a Kruskal–Wallis test followed by pairwise comparison using the Wilcoxon test. Error bars represent 95% confidence interval. Graphs were generated using the ggplot2 package in R (Wickham, 2009). Boxes represent interquartile range (IQR) where the two edges of the box are 25th (Q1) and 75th (Q3) percentiles and the line in the box 50th percentile (median). The ends of the whiskers represent minimum (Q1−1.5×IQR) and maximum (Q3+1.5×IQR). Dots represent outliers.

In this study, we show that CHIQ1 maintains the proliferative capacity of cells and keeps the timing of cell cycle exit during organ development by combining cell population (kinematic analysis) and single cell (cell lineage tracking) studies. Kinematic studies involve assessing cell size and cell number over time in growing organs, and are a powerful framework in which to characterize how cellular parameters change during organ development at the organ level (Rymen et al., 2010). However, during a kinematic study, proliferation rate and the length of the cell cycle are calculated from the estimated cell number at each time point assuming that all cells divide. This is not accurate based on how proliferation occurs in Arabidopsis leaves (Donnelly et al., 1999; Andriankaja et al., 2012). Nevertheless, 21 genes have been reported to affect proliferation rate according to kinematic studies (Table S1). As cell cycle length was not measured directly in any of these studies, it is not possible to distinguish whether these genes are involved in controlling the proportion of dividing cells or cell cycle length at this time.

Here, we calculated cell cycle length of wild-type and chiq1-1 leaves by tracking individual cells over time with live imaging and considering only the cells that underwent division. We found no difference in the cell cycle length between genotypes in both control conditions and in response to a G1/S cell cycle checkpoint trigger using HU. This discovery revealed that the proliferation rate at the organ level in chiq1-1 was reduced due to the number of dividing cells decreasing, rather than to an increase in cell cycle length.

Duration of cell proliferation in an organ is controlled by both cell intrinsic (cell-level) and extrinsic (organ-level) mechanisms. How these mechanisms interact to control proliferation exit and final cell number in leaves is unknown. Several lines of evidence support that the duration of cell proliferation is controlled by cell extrinsic mechanisms that depend on cell-cell communication via the movement of proteins or small molecules (Helariutta et al., 2000; Serralbo et al., 2006; Anastasiou et al., 2007; Marcotrigiano, 2010; Nobusawa et al., 2013). Among the three proposed cell-intrinsic mechanisms (counter, timer and sizer), the cell cycle counter mechanism has more supporting evidence than a timer or sizer mechanism in plants (Sun et al., 2009; Kumpf et al., 2014). A cell cycle counter mechanism controlling stem cell division has been proposed for stem cells in floral and root meristems (Sun et al., 2009). The stem cell activity in the floral meristem terminates when the expression of the homeotic transcription factor WUSCHEL (WUS) is stably repressed by the transcriptional regulators KNUCKLES (KNU) and AGAMOUS (AG) (Sun et al., 2009). This repression depends on the dilution of the repressive chromatin mark H3K27me3 on the promoter of KNU across successive replication cycles, which allows AG to increase KNU gene expression. This DNA replication-dependent delay in the activation of KNU expression provides the proper number of cells required to form floral organs by a cell division counter mechanism (Sun et al., 2009). Similarly, the amount of the chromatin mark H3K36me at specific target genes after each DNA replication round is controlled by the SET-domain protein ASH1-RELATED 3 (ASH1R3) (Kumpf et al., 2014). This epigenetic modification at each cell cycle could act as a cell division counter in the root apical meristem (Kumpf et al., 2014). In both cases, the dilution of a chromatin mark sets the timing of cell differentiation (and therefore may force cells to exit proliferation), indicating chromatin-level regulation. Although a similar mechanism could operate in leaves to control the number of cell division rounds within the amplifying cell population (i.e. dividing cells that are not stem cells), it would have to depend on other, yet undiscovered, genes, as WUS and KNU are stem-cell specific and ASH1R3 is root specific, and none of these genes has been implicated in controlling the proliferation-differentiation transition in leaves. In this work, we show that CHIQ1, and the CHIQ-like proteins CHIQL4, CHIQL5 and CHIQL6, control the proper timing of transition from proliferation into differentiation during organ development and, previously, we found that CHIQ1 interacted with CHIQL6, which interacted with a member of the chromatin repressor complex PRC2 (Bossi et al., 2017). These results raise the possibility that CHIQ-like proteins, including CHIQ1, might also control the number of times a cell divides by modulating the level of chromatin markers. However, we currently have no evidence that supports or contradicts the hypothesis that CHIQ1 is a transcriptional regulator acting via PRC2.

Most studies of cell proliferation and how different genes act to control proliferation rate have been performed at the cell population level using kinematic studies. Although this approach is accurate in describing a general cellular phenotype (De Veylder et al., 2001; Andriankaja et al., 2012), it explains phenomena for an ‘average ideal cell’, disregarding valuable information contained in the heterogeneous population. This study highlights the need to study development at the individual cell level in order to understand fundamental rules of biology, such as what cell-intrinsic and -extrinsic mechanisms control cell proliferation exit in plant organs.

Plant material and growth conditions

Arabidopsis thaliana plants were grown either in PRO-MIX HP Mycorrhizae potting soil (Premier Tech Horticulture, Quakertown, PA, USA) or in 0.5× Murashige and Skoog basal salt mixture (MS) media (PhytoTechnologies Laboratories) (pH 5.7), supplemented with 0.8% agar (Difco) and 1% sucrose (SIGMA). Plants were grown at 22°C in a 16:8 light:dark photoperiod, 40% RH and full-spectrum LED lights outputting ∼115 μmol m−2 s−1 Photosynthetic Photon Flux Density (PPFD) measured at pot level. Seeds were stratified at 4°C for 4 nights to break dormancy. The following SALK lines were used: SALK_064001 (chiq1-1), SALK_116702 (chiql 4-1), SALK_105421 (chiql 5-1) and SALK_086603 (chiql 6-1). The quadruple mutant was generated by crossing SALK_064001, SALK_116702, SALK_105421 and SALK_086603.

To construct binary vectors expressing CHIQ1, a 1900 bp genomic region containing the promoter and the protein-coding sequence of CHIQ1 gene and a 1278 bp region containing the CHIQ protein coding sequence were amplified by PCR from Col-0 genomic DNA and cloned into the entry vector pDONR221 (Life Technologies) to create pDONR221-CHIQ1pro:CHIQ1 and pDONR221-CHIQ1 vectors, respectively. pDONR221-CHIQ1pro:CHIQ1 was transferred to the binary vector pGWB640 (Nakamura et al., 2010) using Gateway cloning (Life Technologies) to create the vector CHIQ1pro:CHIQ1-YFP, which expresses the CHIQ1-YFP protein fusion under the endogenous promoter of CHIQ1. pDONR221-CHIQ1 was transferred to the binary vector pB7HFC3_0 (Huang et al., 2016), using Gateway cloning (Life Technologies) to create the vector 35Spro:CHIQ1-FLAG, which expresses the wild-type allele of CHIQ1 under the constitutive promoter 35S.

To obtain complemented lines, chiq1-1 plants were transformed with the transgene 35Spro:CHIQ1-YFP or 35Spro:CHIQ1-FLAG by dipping their flowers into an Agrobacterium (strain pGVl101 pMP90) cell suspension containing the transgene construct. Briefly, a single colony harboring each construct was cultured in LB media with 50 μg/ml gentamicin, 25 μg/ml rifampicin and 50 μg/ml kanamycin or 50 μg/ml spectomicin until the culture reached an optical density of 0.5-1.0 at 600 nm. When the culture was ready, the LB was removed and the cell pellet was resuspended in plant growth media (0.43% MS (m/v), 0.05% MES (m/v), 5% sucrose (m/v) and 0.02% Silwett L77 (v/v) at pH 5.8-6) and transferred into a wide-mouthed container. Inflorescences of chiq1-1 plants were dipped into this solution for a few minutes. Treated plants were placed horizontally and covered, left in the dark overnight, and moved to greenhouse-growing conditions the next day.

Transgenic lines were selected with BASTA (glufosinate ammonium). The adult phenotype was complemented in all transgenic lines analyzed (∼10 independent lines per construct). The lines 3089x (CHIQ1pro:CHIQ1-YFP) and HFC10.4 (35Spro:CHIQ1-FLAG) were chosen for detailed organ size studies during development.

For marker gene studies, we used the following lines: (1) a transcriptional fusion containing the promoter of the cell cycle gene CDKB1.1 linked to the GUS-encoding gene uidA (CDKB1.1pro:GUS) from Dr Kathryn Barton (Carnegie Institution for Science, Washington DC, USA); (2) a transcriptional fusion containing the promoter of the cell cycle gene CYCD3.3 linked to the GUS-encoding gene uidA (CYCD3.3pro:GUS) from Dr Jose Dinneny (Stanford University, CA, USA); (3) a fusion containing the promoter of the cell cycle gene CYCB1.1 and its destruction box linked to the fluorescent protein GFP (CYCBpro:CYCBdestruction box-GFP) from Dr Masaaki Umeda (Nara Institute of Science and Technology, Ikoma Japan); and (4) a translational fusion containing the promoter of the epidermis-specific gene ATML1 and the plasma membrane localized protein RCI2A linked to the fluorescent protein mCitrine from Dr Adrienne Roeder (Cornell University, Ithaca, NY, USA). These transgenic lines were crossed with chiq1-1 plants to obtain siblings carrying the corresponding marker gene in the wild-type background or in the homozygous mutant background.

Phenotypic analyses

Leaf area

Leaf area was measured from plants grown in soil for 4, 5, 6, 7, 8, 11, 13, 15, 19, 22 and 25 days after sowing. Leaves were dissected and photographed with a compound microscope (Nikon) or a dissecting scope (Leica MZ6 microscope), or scanned, depending on their size. Leaf blade area was measured from the images using ImageJ (Schneider et al., 2012).

Root length

Primary root length was measured in seedlings grown on 0.5× MS agar media supplemented with 1% sucrose for 3, 5, 7, 10, 12 and 14 days after stratification. Seedlings were imaged and the length of the primary root was measured using ImageJ (Schneider et al., 2012). Root length in 21-day-old plants was measured using a rhizotron system (Rellán-Álvarez et al., 2015). Assembled rhizotrons were placed in a box with water, and sown with seeds that had been stratified at 4°C for 4 days to break dormancy. These boxes were sealed with a transparent lid and packing tape. Plants were grown under a 16:8 light:dark photoperiod in a growth chamber at 22°C under full-spectrum LED lights outputting ∼150 PPFD at pot level. Three days after sowing, the lids were removed, and the rhizotrons were watered once a day for 7 days with 2 ml of distilled water using a micropipette. Twenty-one days after sowing, the rhizotrons were opened, and individual plants were removed, washed and mounted onto a clear plastic sheet using double-sided tape. These plants were scanned and the root length of each plant was measured using ImageJ (Schneider et al., 2012).

Cell area and number in leaves

Cell area was measured from confocal images of the 1st, 4th and 7th leaves from 25- to 35-day-old plants (mature leaf) and the 1st true leaves from 4-, 6-, 8-, 11-, 13-, 15-, 19- and 25-day-old plants (Col-0 and chiq1-1, growing leaf) or 6-, 8-, 11- and 25-day-old plants [quadruple mutant chiq-quad (chiq1-1;chiql4-1;chiql5-1;chiql6-1), growing leaf] grown on soil. All lines (Col-0, chiq1-1, the complemented line 35Spro:CHIQ1-FLAG and chiq-quad) express the plasma membrane fluorescent marker RCI2A in the epidermis (Roeder et al., 2010). The fluorescent marker RCI2A was introgressed into chiq1-1 and the 35Spro:CHIQ1-FLAG line; and introduced by Agrobacterium transformation into the quadruple background. To measure leaf area, leaves were dissected and photographed using a Leica SP8 confocal microscope, dissecting microscope (Leica MZ6 microscope) or scanner, depending on their size. Three confocal images (20×/0.7, glycerine, 512×512) of the abaxial epidermis were taken from each leaf, corresponding to the 25th, 50th and 75th percentile region of the blade, respectively. Excitation was performed using a 516 nm laser line, and mCitrine was collected at 526-580 nm. Maximum projections were calculated in ImageJ and all complete cells were traced (pencil width=1 pixel for day 6 and day 8, pencil width=2 pixels for day 11 and older plants). Total leaf area and individual cell areas were measured using ImageJ (Schneider et al., 2012). Average cell size, total cell number, cell production and stomatal index were calculated according to Rymen et al. (2010).

Trichome morphology, size and number in leaves

The number of trichomes on the 4th and 7th leaf were counted by hand under a dissecting microscope (Leica MZ6 microscope). To measure the branch size of a trichome, leaves were dissected and photographed using a dissecting microscope (Leica MZ6 microscope). The average size per trichome branch was calculated by dividing the whole area of each trichome measured from images using ImageJ software (Schneider et al., 2012) by the number of branches in each trichome.

Cell number in the root apical meristem (RAM)

Roots of 8 and 12 day-old wild-type (Col-0) and chiq1-1 seedlings were stained with 10 μg/ml propidium iodide (PI) for ∼5 min and imaged with a Leica SP8 confocal microscope. Excitation was performed using a 488 nm laser line, and PI was collected at 570-670 nm. ImageJ software was used to count epidermal cells along a cell file. The RAM zone was defined as described in the literature (Pavelescu et al., 2018): the region in the root encompassed by the first cell in the epidermal file after the quiescent center to the first cell that is, at most, half the length of the adjacent older cell within the same cell file.

Cell cycle marker gene studies

Expression of the reporter gene uidA (GUS) driven by the promoter of the cell cycle gene CYCD3.3 (CYCD3.3pro:GUS) was analyzed in the 1st true leaf of plants grown for 7, 8 and 9 days on 0.5× MS agar plates supplemented with 1% sucrose. Expression of GUS driven by the promoter of the cell cycle gene CDKB1.1 (CDKB1.1pro:GUS) was analyzed in the 4th and 5th leaf of soil-grown plants at 16, 20 and 24 days after sowing (DAS). Plants were stained in GUS staining solution (Jefferson et al., 1987) at 37°C overnight, and were destained in 70% ethanol at room temperature for 24 h. Images were taken with a Leica MZ6 microscope. Expression of GFP driven by the promoter and destruction box of the cell cycle gene CYCB1.1 (CYCBpro:CYCBdestruction box-GFP) was analyzed in the root apical meristem of 5-day-old seedlings grown on plates. Roots were stained in 10 μg/ml propidium iodide for ∼5 min before imaging with a Leica SP8 confocal microscope. Excitation was performed using a 488 nm laser line, and PI and GFP were collected at 570-670 and 498-548 nm, respectively.

Endoreduplication analysis

Mature 6th and 7th leaves were collected from plants grown in soil. Leaves were chopped into thin slices and incubated in an enzyme solution for 90 min to obtain protoplasts (Yoo et al., 2007). Protoplasts were filtered with a 100 μm cell strainer and washed several times with W5 buffer (Yoo et al., 2007). After the last wash, protoplasts were resuspended in lysis buffer [45 mM magnesium chloride, 30 mM sodium citrate, 20 mM MOPS, 0.5% Triton and 2% 4′,6-diamidino-2-phenylindole (DAPI, Thermo Scientific) (pH 7)] (Galbraith, 2014) to obtain nuclei. Nuclei solution was filtered with a 40um cell strainer before flow cytometry analysis. For each experiment, leaves from 60 plants were pooled and nuclei extracts were split into two technical replicates. Two independent experiments were conducted. Data were collected on the LSR II.UV analyzer (NIH S10 Shared Instrument Grant S10RR027431-01) at the Stanford Shared FACS Facility (SSFF) and analyzed using the flowCore, flowViz, flowStats and flowDensity libraries within the R package Bioconductor (https://www.bioconductor.org/).

Analysis of cell cycle parameters

Six-day-old wild-type and chiq1-1 seedlings carrying a plasma membrane fluorescent marker RCI2A in the epidermis (Roeder et al., 2010) and grown on agar media were mounted on slides with 400 μl of 0.5× MS media with 1% sucrose, after removal of one cotyledon. For HU treatments, seedlings were mounted on slides with 400 μl of 0.5× MS media with 1% sucrose plus 0.5 mM or 1 mM HU. Exposed young leaves were imaged using a Leica SP8 confocal microscope through a multi-immersion 20×/0.7 objective using glycerine immersion media. Excitation was performed at 516 nm, and mCitrine was collected at 526-580 nm. Leaves were imaged taking a z-stack once a day for 2 days. Maximum projections were calculated in ImageJ (Schneider et al., 2012).

The mid-region of the leaf, where active division is happening, was determined and an area of ∼40 cells was selected. Cells within the selected region were traced (pencil width=1 pixel) and cell sizes were measured using ImageJ (Schneider et al., 2012). To measure cell division, the cells within the region were numbered and colored using ImageJ (Schneider et al., 2012). The mother and daughter cells were manually tracked through subsequent images. Cells that divided were identified and their progeny counted. The number of cells produced per mother cell was used to calculate proliferation rate and cell cycle length.

We thank the ABRC and Drs K. Barton, M. Umeda, A. Roeder and J. Dinneny for providing mutant seeds and seeds of transgenic lines expressing cell cycle stage-specific markers or a plasma membrane localized-fluorescent protein; Drs D. Nusinow and S. Ishiguro for providing binary plasmids; T. LaRue, A. Srinivas and Dr. J. Vilarasa-Blasi for technical assistance on the rhizotron and root imaging studies; H. Nam, L. Mai and J. A. Kim for technical assistance; T. Knaak and the Stanford Shared FACS Facility for technical assistance on endoreduplication studies; A. Malkovskiy for microscopy assistance; G. Materassi-Shultz for maintenance and watering of all plants used in this study; G. Huntress and M. Lopez for IT assistance; I. Villa and T. Van de Sande for general facilities and equipment support; and Drs D. Ehrhardt, H. Meyer, S. Xu, L. de Veylder, L. Willis and members of the Rhee lab and the Carnegie-Stanford plant research community for helpful discussion. This work was carried out on the ancestral land of the Muwekma Ohlone Tribe, which was and continues to be of great importance to the Ohlone people.

Author contributions

Conceptualization: F.B., S.Y.R.; Methodology: F.B., B.J., E.L., H.C.; Formal analysis: F.B., B.J., E.L., S.Y.R.; Resources: F.B., S.Y.R.; Data curation: F.B., S.Y.R.; Writing - original draft: F.B., Y.D.; Writing - review & editing: F.B., B.J., E.L., H.C., Y.D., S.Y.R.; Visualization: F.B., B.J., H.C.; Supervision: F.B., S.Y.R.; Project administration: F.B., S.Y.R.; Funding acquisition: S.Y.R.

Funding

This work was supported in part by Carnegie Institution for Science Endowment and grants from the National Science Foundation (IOS-1546838 and IOS-1026003) and the U.S. Department of Energy, the Office of Science and the office of Biological and Environmental Research (Genomic Science Program grants DE-SC0018277, DE-SC0008769, DE-SC0020366 and DE-SC0021286).

The peer review history is available online at https://journals.biologists.com/dev/article-lookup/doi/10.1242/dev.200565.

Aarssen
,
L. W.
(
2015
).
Body size and fitness in plants: revisiting the selection consequences of competition
.
Perspect. Plant Ecol. Evol. Syst.
17
,
236
-
242
.
Aki
,
S. S.
and
Umeda
,
M.
(
2016
).
Cytrap marker systems for in vivo visualization of cell cycle progression in Arabidopsis
.
Methods Mol. Biol.
1370
,
51
-
57
.
Anastasiou
,
E.
,
Kenz
,
S.
,
Gerstung
,
M.
,
MacLean
,
D.
,
Timmer
,
J.
,
Fleck
,
C.
and
Lenhard
,
M.
(
2007
).
Control of plant organ size by KLUH/CYP78A5-dependent intercellular signaling
.
Dev. Cell
13
,
843
-
856
.
Andriankaja
,
M.
,
Dhondt
,
S.
,
De Bodt
,
S.
,
Vanhaeren
,
H.
,
Coppens
,
F.
,
De Milde
,
L.
,
Mühlenbock
,
P.
,
Skirycz
,
A.
,
Gonzalez
,
N.
,
Beemster
,
G. T. S.
et al. 
(
2012
).
Exit from proliferation during leaf development in Arabidopsis thaliana: a not-so-gradual process
.
Dev. Cell
22
,
64
-
78
.
Bao
,
Y.
,
Song
,
W.-M.
,
Wang
,
P.
,
Yu
,
X.
,
Li
,
B.
,
Jiang
,
C.
,
Shiu
,
S.-H.
,
Zhang
,
H.
and
Bassham
,
D. C.
(
2020
).
COST1 regulates autophagy to control plant drought tolerance
.
Proc. Natl. Acad. Sci. USA
117
,
7482
-
7493
.
Bossi
,
F.
,
Fan
,
J.
,
Xiao
,
J.
,
Chandra
,
L.
,
Shen
,
M.
,
Dorone
,
Y.
,
Wagner
,
D.
and
Rhee
,
S. Y.
(
2017
).
Systematic discovery of novel eukaryotic transcriptional regulators using sequence homology independent prediction
.
BMC Genomics
18
,
480
.
Boudolf
,
V.
,
Vlieghe
,
K.
,
Beemster
,
G. T. S.
,
Magyar
,
Z.
,
Torres Acosta
,
J. A.
,
Maes
,
S.
,
Van Der Schueren
,
E.
,
Inzé
,
D.
and
De Veylder
,
L.
(
2004
).
The plant-specific cyclin-dependent kinase CDKB1;1 and transcription factor E2Fa-DPa control the balance of mitotically dividing and endoreduplicating cells in Arabidopsis
.
Plant Cell
16
,
2683
-
2692
.
Burton
,
P. B.
,
Raff
,
M. C.
,
Kerr
,
P.
,
Yacoub
,
M. H.
and
Barton
,
P. J.
(
1999
).
An intrinsic timer that controls cell-cycle withdrawal in cultured cardiac myocytes
.
Dev. Biol.
216
,
659
-
670
.
Chen
,
H.
,
Einstein
,
L. C.
,
Little
,
S. C.
and
Good
,
M. C.
(
2019
).
Spatiotemporal patterning of zygotic genome activation in a model vertebrate embryo
.
Dev. Cell
49
,
852
-
866.e7
.
Conlon
,
I.
and
Raff
,
M.
(
1999
).
Size control in animal development
.
Cell
96
,
235
-
244
.
De Veylder
,
L.
,
Beeckman
,
T.
,
Beemster
,
G. T.
,
Krols
,
L.
,
Terras
,
F.
,
Landrieu
,
I.
,
van der Schueren
,
E.
,
Maes
,
S.
,
Naudts
,
M.
and
Inzé
,
D.
(
2001
).
Functional analysis of cyclin-dependent kinase inhibitors of Arabidopsis
.
Plant Cell
13
,
1653
-
1668
.
Dewitte
,
W.
,
Riou-Khamlichi
,
C.
,
Scofield
,
S.
,
Healy
,
J. M. S.
,
Jacqmard
,
A.
,
Kilby
,
N. J.
and
Murray
,
J. A. H.
(
2003
).
Altered cell cycle distribution, hyperplasia, and inhibited differentiation in Arabidopsis caused by the D-type cyclin CYCD3
.
Plant Cell
15
,
79
-
92
.
Donnelly
,
P. M.
,
Bonetta
,
D.
,
Tsukaya
,
H.
,
Dengler
,
R. E.
and
Dengler
,
N. G.
(
1999
).
Cell cycling and cell enlargement in developing leaves of Arabidopsis
.
Dev. Biol.
215
,
407
-
419
.
Durand
,
B.
,
Fero
,
M. L.
,
Roberts
,
J. M.
and
Raff
,
M. C.
(
1998
).
p27Kip1 alters the response of cells to mitogen and is part of a cell-intrinsic timer that arrests the cell cycle and initiates differentiation
.
Curr. Biol.
8
,
431
-
440
.
Fujikura
,
U.
,
Horiguchi
,
G.
,
Ponce
,
M. R.
,
Micol
,
J. L.
and
Tsukaya
,
H.
(
2009
).
Coordination of cell proliferation and cell expansion mediated by ribosome-related processes in the leaves of Arabidopsis thaliana
.
Plant J.
59
,
499
-
508
.
Galbraith
,
D. W.
(
2014
).
Flow cytometry and sorting in arabidopsis
.
Methods Mol. Biol.
1062
,
509
-
537
.
Gázquez
,
A.
and
Beemster
,
G. T. S.
(
2017
).
What determines organ size differences between species? A meta-analysis of the cellular basis
.
New Phytol.
215
,
299
-
308
.
Gonzalez
,
N.
,
Vanhaeren
,
H.
and
Inzé
,
D.
(
2012
).
Leaf size control: complex coordination of cell division and expansion
.
Trends Plant Sci.
17
,
332
-
340
.
Hayflick
,
L.
and
Moorhead
,
P. S.
(
1961
).
The serial cultivation of human diploid cell strains
.
Exp. Cell Res.
25
,
585
-
621
.
Helariutta
,
Y.
,
Fukaki
,
H.
,
Wysocka-Diller
,
J.
,
Nakajima
,
K.
,
Jung
,
J.
,
Sena
,
G.
,
Hauser
,
M. T.
and
Benfey
,
P. N.
(
2000
).
The SHORT-ROOT gene controls radial patterning of the Arabidopsis root through radial signaling
.
Cell
101
,
555
-
567
.
Huang
,
H.
,
Alvarez
,
S.
,
Bindbeutel
,
R.
,
Shen
,
Z.
,
Naldrett
,
M. J.
,
Evans
,
B. S.
,
Briggs
,
S. P.
,
Hicks
,
L. M.
,
Kay
,
S. A.
and
Nusinow
,
D. A.
(
2016
).
Identification of evening complex associated proteins in arabidopsis by affinity purification and mass spectrometry
.
Mol. Cell. Proteomics
15
,
201
-
217
.
Jefferson
,
R. A.
,
Kavanagh
,
T. A.
and
Bevan
,
M. W.
(
1987
).
GUS fusions: beta-glucuronidase as a sensitive and versatile gene fusion marker in higher plants
.
EMBO J.
6
,
3901
-
3907
.
Jones
,
A. R.
,
Forero-Vargas
,
M.
,
Withers
,
S. P.
,
Smith
,
R. S.
,
Traas
,
J.
,
Dewitte
,
W.
and
Murray
,
J. A. H.
(
2017
).
Cell-size dependent progression of the cell cycle creates homeostasis and flexibility of plant cell size
.
Nat. Commun.
8
,
15060
.
Karamat
,
U.
,
Sun
,
X.
,
Li
,
N.
and
Zhao
,
J.
(
2021
).
Genetic regulators of leaf size in Brassica crops
.
Hortic Res.
8
,
91
.
Katagiri
,
Y.
,
Hasegawa
,
J.
,
Fujikura
,
U.
,
Hoshino
,
R.
,
Matsunaga
,
S.
and
Tsukaya
,
H.
(
2016
).
The coordination of ploidy and cell size differs between cell layers in leaves
.
Development
143
,
1120
-
1125
.
Kumpf
,
R.
,
Thorstensen
,
T.
,
Rahman
,
M. A.
,
Heyman
,
J.
,
Nenseth
,
H. Z.
,
Lammens
,
T.
,
Herrmann
,
U.
,
Swarup
,
R.
,
Veiseth
,
S. V.
,
Emberland
,
G.
et al. 
(
2014
).
The ASH1-RELATED3 SET-domain protein controls cell division competence of the meristem and the quiescent center of the arabidopsis primary root
.
Plant Physiol.
166
,
632
-
643
.
Luan
,
L.
,
Jiang
,
Y.
,
Cheng
,
M.
,
Dini-Andreote
,
F.
,
Sui
,
Y.
,
Xu
,
Q.
,
Geisen
,
S.
and
Sun
,
B.
(
2020
).
Organism body size structures the soil microbial and nematode community assembly at a continental and global scale
.
Nat. Commun.
11
,
6406
.
Lui
,
J. C.
and
Baron
,
J.
(
2011
).
Mechanisms limiting body growth in mammals
.
Endocr. Rev.
32
,
422
-
440
.
Marcotrigiano
,
M.
(
2010
).
A role for leaf epidermis in the control of leaf size and the rate and extent of mesophyll cell division
.
Am. J. Bot.
97
,
224
-
233
.
Melaragno
,
J. E.
,
Mehrotra
,
B.
and
Coleman
,
A. W.
(
1993
).
Relationship between endopolyploidy and cell size in epidermal tissue of arabidopsis
.
Plant Cell
5
,
1661
-
1668
.
Nakamura
,
S.
,
Mano
,
S.
,
Tanaka
,
Y.
,
Ohnishi
,
M.
,
Nakamori
,
C.
,
Araki
,
M.
,
Niwa
,
T.
,
Nishimura
,
M.
,
Kaminaka
,
H.
,
Nakagawa
,
T.
et al. 
(
2010
).
Gateway binary vectors with the bialaphos resistance gene, bar, as a selection marker for plant transformation
.
Biosci. Biotechnol. Biochem.
74
,
1315
-
1319
.
Neurohr
,
G. E.
,
Terry
,
R. L.
,
Lengefeld
,
J.
,
Bonney
,
M.
,
Brittingham
,
G. P.
,
Moretto
,
F.
,
Miettinen
,
T. P.
,
Vaites
,
L. P.
,
Soares
,
L. M.
,
Paulo
,
J. A.
et al. 
(
2019
).
Excessive cell growth causes cytoplasm dilution and contributes to senescence
.
Cell
176
,
1083
-
1097.e18
.
Nobusawa
,
T.
,
Okushima
,
Y.
,
Nagata
,
N.
,
Kojima
,
M.
,
Sakakibara
,
H.
and
Umeda
,
M.
(
2013
).
Synthesis of very-long-chain fatty acids in the epidermis controls plant organ growth by restricting cell proliferation
.
PLoS Biol.
11
,
e1001531
.
Pavelescu
,
I.
,
Vilarrasa-Blasi
,
J.
,
Planas-Riverola
,
A.
,
González-García
,
M.-P.
,
Caño-Delgado
,
A. I.
and
Ibañes
,
M.
(
2018
).
A sizer model for cell differentiation in Arabidopsis thaliana root growth
.
Mol. Syst. Biol.
14
,
e7687
.
Penzo-Méndez
,
A. I.
and
Stanger
,
B. Z.
(
2015
).
Organ-size regulation in mammals
.
Cold Spring Harb. Perspect. Biol.
7
,
a019240
.
Pickering
,
J.
,
Rich
,
C. A.
,
Stainton
,
H.
,
Aceituno
,
C.
,
Chinnaiya
,
K.
,
Saiz-Lopez
,
P.
,
Ros
,
M. A.
and
Towers
,
M.
(
2018
).
An intrinsic cell cycle timer terminates limb bud outgrowth
.
eLife
7
,
e37429
.
Rellán-Álvarez
,
R.
,
Lobet
,
G.
,
Lindner
,
H.
,
Pradier
,
P.-L.
,
Sebastian
,
J.
,
Yee
,
M.-C.
,
Geng
,
Y.
,
Trontin
,
C.
,
LaRue
,
T.
,
Schrager-Lavelle
,
A.
et al. 
(
2015
).
GLO-Roots: an imaging platform enabling multidimensional characterization of soil-grown root systems
.
eLife
4
,
e07597
.
Roeder
,
A. H. K.
,
Chickarmane
,
V.
,
Cunha
,
A.
,
Obara
,
B.
,
Manjunath
,
B. S.
and
Meyerowitz
,
E. M.
(
2010
).
Variability in the control of cell division underlies sepal epidermal patterning in Arabidopsis thaliana
.
PLoS Biol.
8
,
e1000367
.
Rymen
,
B.
,
Coppens
,
F.
,
Dhondt
,
S.
,
Fiorani
,
F.
and
Beemster
,
G. T. S.
(
2010
).
Kinematic analysis of cell division and expansion
.
Methods Mol. Biol.
655
,
203
-
227
.
Schneider
,
C. A.
,
Rasband
,
W. S.
and
Eliceiri
,
K. W.
(
2012
).
NIH Image to ImageJ: 25 years of image analysis
.
Nat. Methods
9
,
671
-
675
.
Serralbo
,
O.
,
Pérez-Pérez
,
J. M.
,
Heidstra
,
R.
and
Scheres
,
B.
(
2006
).
Non-cell-autonomous rescue of anaphase-promoting complex function revealed by mosaic analysis of HOBBIT, an Arabidopsis CDC27 homolog
.
Proc. Natl. Acad. Sci. USA
103
,
13250
-
13255
.
Singh
,
A.
and
Xu
,
Y.-J.
(
2016
).
The cell killing mechanisms of hydroxyurea
.
Genes
7
,
99
.
Stanger
,
B. Z.
,
Tanaka
,
A. J.
and
Melton
,
D. A.
(
2007
).
Organ size is limited by the number of embryonic progenitor cells in the pancreas but not the liver
.
Nature
445
,
886
-
891
.
Sun
,
B.
,
Xu
,
Y.
,
Ng
,
K.-H.
and
Ito
,
T.
(
2009
).
A timing mechanism for stem cell maintenance and differentiation in the Arabidopsis floral meristem
.
Genes Dev.
23
,
1791
-
1804
.
Texada
,
M. J.
,
Koyama
,
T.
and
Rewitz
,
K.
(
2020
).
Regulation of body size and growth control
.
Genetics
216
,
269
-
313
.
Vollmer
,
J.
,
Casares
,
F.
and
Iber
,
D.
(
2017
).
Growth and size control during development
.
Open Biol.
7
,
170190
.
Wickham
,
H.
(
2009
).
ggplot2: Elegant Graphics for Data Analysis
.
New York, NY
:
Springer
.
Yoo
,
S.-D.
,
Cho
,
Y.-H.
and
Sheen
,
J.
(
2007
).
Arabidopsis mesophyll protoplasts: a versatile cell system for transient gene expression analysis
.
Nat. Protoc.
2
,
1565
-
1572
.

Competing interests

The authors declare no competing or financial interests.

Supplementary information