Planta

, Volume 223, Issue 6, pp 1315–1328

Nocturnal changes in leaf growth of Populus deltoides are controlled by cytoplasmic growth

Authors

    • Institut for Chemistry and Dynamics of the Geosphere: Phytosphere (ICG-III)Research Centre Jülich
  • Vaughan Hurry
    • Umeå Plant Science Centre, Department of Plant PhysiologyUmeå University
  • Nathalie Druart
    • Umeå Plant Science Centre, Department of Forest Genetics and Plant PhysiologySwedish University of Agricultural Sciences
  • Catherine Benedict
    • Umeå Plant Science Centre, Department of Plant PhysiologyUmeå University
  • Ingar Janzik
    • Institut for Chemistry and Dynamics of the Geosphere: Phytosphere (ICG-III)Research Centre Jülich
  • Andrés Chavarría-Krauser
    • Institut for Chemistry and Dynamics of the Geosphere: Phytosphere (ICG-III)Research Centre Jülich
  • Achim Walter
    • Institut for Chemistry and Dynamics of the Geosphere: Phytosphere (ICG-III)Research Centre Jülich
  • Ulrich Schurr
    • Institut for Chemistry and Dynamics of the Geosphere: Phytosphere (ICG-III)Research Centre Jülich
Original Article

DOI: 10.1007/s00425-005-0181-0

Cite this article as:
Matsubara, S., Hurry, V., Druart, N. et al. Planta (2006) 223: 1315. doi:10.1007/s00425-005-0181-0

Abstract

Growing leaves do not expand at a constant rate but exhibit pronounced diel growth rhythms. However, the mechanisms giving rise to distinct diel growth dynamics in different species are still largely unknown. As a first step towards identifying genes controlling rate and timing of leaf growth, we analysed the transcriptomes of rapidly expanding and fully expanded leaves of Populus deltoides Bartr. ex. Marsh at points of high and low expansion at night. Tissues with well defined temporal growth rates were harvested using an online growth-monitoring system based on a digital image sequence processing method developed for quantitative mapping of dicot leaf growth. Unlike plants studied previously, leaf growth in P. deltoides was characterised by lack of a base-tip gradient across the lamina, and by maximal and minimal growth at dusk and dawn, respectively. Microarray analysis revealed that the nocturnal decline in growth coincided with a concerted down-regulation of ribosomal protein genes, indicating deceleration of cytoplasmic growth. In a subsequent time-course experiment, Northern blotting and real-time RT-PCR confirmed that the ribosomal protein gene RPL12 and a cell-cycle gene H2B were down-regulated after midnight following a decrease in cellular carbohydrate concentrations. Thus, we propose that the spatio-temporal growth pattern in leaves of P. deltoides primarily arises from cytoplasmic growth whose activity increases from afternoon to midnight and thereafter decreases in this species.

Keywords

Cell cycleCytoplasmic growthLeaf growthMicroarrayPopulusRibosomal protein

Abbreviations

LPI

Leaf plastochron index

M

Mature leaves

RGR

Relative growth rate

RP

Ribosomal protein

Y

Young leaves

Introduction

Leaf growth involves two cellular growth modes: cytoplasmic growth and vacuolated growth. Cytoplasmic growth is accomplished by increasing cytoplasmic mass, and necessarily involves high metabolic activity and macromolecular synthesis. Cytoplasmic growth can result in cell proliferation (with cell division) or endoreduplication (without cell division; Sugimoto-Shirasu and Roberts 2003). In vacuolated growth, on the other hand, cells enlarge by expanding cell walls while taking up water and solutes into the vacuole. Vacuolated growth accounts for the bulk of rapid leaf expansion, and it is widely accepted that the key regulators of cell wall extension play an important role in controlling vacuolated growth (Cosgrove 2000).

Leaf growth does not progress at a constant rate but often exhibits diel fluctuations (Dale 1988; Walter and Schurr 2005), which is a prominent feature of leaf growth (Walter and Schurr 2005). In dicot leaves, growth rate is also not uniform across the lamina; a basipetal gradient in growth rate is often observed. Formation of such a growth gradient has been attributed to the earlier cessation of cell division and expansion in the distal part of the lamina compared to the basal part (Poethig and Sussex 1985; Granier and Tardieu 1998; Schmundt et al. 1998; Donnelly et al. 1999; Walter and Schurr 2005). Furthermore, both developmental and growth processes are coordinated but differently regulated in the different types of cells that constitute a leaf (mesophyll, epidermal, vascular or guard cells). Spatio-temporal variations, in combination with heterogeneity in cell types, complicate the elucidation of leaf growth mechanisms in dicot leaves.

Large-scale transcriptional profiling using microarrays has enabled studies of the genetic controls of complex and coordinated processes, such as spatial (Trainotti et al. 2004), cell-type-specific (Nakazono et al. 2003) or developmental (Beemster et al. 2005) aspects of leaf growth. In a study on base-tip gradients in expanding leaves of tobacco, Trainotti et al. (2004) found high expression of genes involved in cell proliferation and photosynthesis in basal and apical regions of leaves, respectively. The study by Nakazono et al. (2003) using laser-capture microdissection identified genes preferentially expressed in epidermal and vascular cells of maize coleoptiles. By combining kinematic analysis and microarray analysis of leaf development in Arabidopsis, Beemster et al. (2005) distinguished cell-cycle genes that are specifically expressed in proliferating cells from those that are constitutively expressed or expressed during endoreduplication. Thus far, however, there has not been a transcriptomic study approaching mechanisms giving rise to the temporal variations in diel leaf growth dynamics.

The objective of the work presented here is to identify the mechanisms controlling temporal growth dynamics in the dicot leaf. Eastern cottonwood, Populus deltoides Bartr. ex. Marsh, was used for the study as little is known about molecular mechanisms of leaf growth in trees compared with herbaceous plants. In addition, the genomic information that has become available for P. trichocarpa facilitates investigation of the regulatory mechanisms at the transcriptional level (Taylor 2002; Bhalerao et al. 2003; Brunner et al. 2004; Cronk 2005). Growth analysis was performed with an image sequence analysis technique developed for quantitative mapping of dicot leaf growth (Schmundt et al. 1998) to characterise temporal as well as spatial growth patterns in leaves of P. deltoides at high resolution. The transcriptomes of these samples were then analysed using the POP2 cDNA microarray (Sterky et al. 2004) to identify the genes and related mechanisms responsible for modulating temporal (nocturnal) leaf growth in P. deltoides. Finally, an independent experiment combining RNA analyses and carbohydrate analysis was carried out to validate the major finding from the microarray analysis. The results point to the importance of cytoplasmic growth in rapidly expanding leaves of this species and suggest its regulation by ribosome biosynthesis.

Materials and methods

Plant material

Plants of P. deltoides Bartr. ex. Marsh (clone S7c8), propagated from cuttings, were grown in 12–l pots in a glasshouse and watered daily. Experiments for microarray analysis were conducted using rapidly expanding (young, Y) and fully expanded leaves (mature, M) of five plants in February and March, 2004. Another set of experiments were conducted for RNA and carbohydrate analyses using rapidly expanding leaves of seven plants in February and March, 2005. The photoperiod in the glasshouse was maintained at 15 h/9 h and 16 h/8 h day/night in 2004 and 2005, respectively. Illumination in the glasshouse (SON-T AGRO 400, Philips) was automatically turned on when the ambient light intensity outside the glasshouse became <400 μmol m−2 s−1 during daytime. The temperature in the glasshouse typically ranged from 19 to 25°C between night and day.

For definition of developmental stages of leaves, we used leaf plastochron index (LPI; Erickson and Michelini 1957):
$$ {\text{LPI}}_{a} = n - a + \frac{{\ln L_{n} - \ln {{20}} }} {{\ln L_{n} - \ln L_{{{(n + 1)}} }}}, $$
(1)
where n and a are the serial numbers of the index and the sample leaves, respectively, Ln is the lamina length (mm) of the index leaf, and L(n+1) is the lamina length of the leaf just above the index leaf n. A leaf just exceeding 20 mm in length was selected as the index leaf (LPI=0) as it seems appropriate for P. deltoides grown under constant conditions (Larson and Isebrands 1971). Leaves at LPI=4 or 5 (growing at about 30% day−1) were chosen as Y, and those at LPI=10 or 11 (fully expanded) as M. The lamina length was 100–150 mm for Y and 200–250 mm for M.

Leaf growth analysis

Leaf growth was measured by a digital image sequence processing method (Schmundt et al. 1998). During the image acquisition, leaves were fixed to the focal plane of a CCD-camera (640×480 pixels) by using weights and illuminated with infrared diodes (880 nm). Near-infrared images were captured every minute on the abaxial side of the leaves using an interference filter (880 nm, Schott, Mainz, Germany). Infrared-cut-off filters (>800 nm, Edmund Scientific, Tucson AZ, USA; >870 nm, Prinz Optics, Stromberg, Germany) were mounted over the sample leaves to reduce the infrared light coming from the surroundings (sunlight and/or the glasshouse illumination) and thus minimise fluctuations in the image brightness and enhance signal-to-noise ratio of the analysis.

Image sequences were evaluated with algorithms based on a structure-tensor approach (Bigün and Granlund 1987; Schmundt et al. 1998) that calculates velocities from all structures that are consistently moving within an image sequence of a growing leaf. Relative growth rates (RGR) were obtained by tracking time-dependent deformation of a polygonal area of interest (AOI) selected within the image. Subsequently, RGR was calculated as
$$ {\text{RGR}}(t) = \frac{1} {{A(t)}}\frac{{{\text{d}}A(t)}} {{{\text{d}}t}}, $$
(2)
where A(t) is the area of AOI at time t. The mean RGR-maps were obtained by calculating a 30-min mean of RGR for every rectangle defined by four neighbouring pixels in the image. The resolution of RGR on a given image depends on the selected filter set for interpolation of velocities from different positions within the scene (Scharr 2005) and on the number of sequential images used for calculation of velocities. Mean RGR-maps calculated from a sequence of 30 min (as shown in Fig. 1b) have a resolution of several hundred spots per image.
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-005-0181-0/MediaObjects/425_2005_181_Fig1_HTML.gif
Fig. 1

a–d Analysis of relative growth rates (RGR, % h−1) in leaves of P. deltoides. a Diel changes in RGR in rapidly expanding (Y, LPI=4–5) and fully expanded leaves (M, LPI=10–11) measured on the second day after leaf fixation. The hourly mean of RGR was calculated for each leaf by averaging the RGR values over an hour. Symbols are a mean of >6 different leaves (error bars, ± SD). Light↔dark transitions caused a sharp increase and decrease in RGR in Y at 22:00 and 7:00, respectively. During the night of the third day, samples for the microarray experiment were harvested at the time points of high (t1) or low (t2) RGR in Y. b Colour-coded maps of RGR across the lamina of a typical Y leaf at t1 and t2 on the day of sampling. The RGR was averaged over 30 min to obtain these images. c Base-tip comparison of RGR in the Y leaf shown in (b). The 30-min average of RGR was calculated for base, middle, and tip regions of the images to examine base–tip gradient within the area used for the growth analysis (error bars, ± SD). d Base–tip comparison of epidermal cell size in Y leaves (n=19–21; error bars, ± SD). The illustration also shows the approximate positions of the three regions in a leaf

Diel growth patterns were monitored for 3 days from the time of leaf fixation. The RGR (% h−1) was calculated for each image as described above and the hourly average obtained. For the tissue sampling for microarray analysis, growth rate was monitored online during the sampling on the third day by calculating RGR upon acquisition of every new image. The estimated time lag between capturing of the last image for the online RGR calculation and sampling was ca. 8 min for Y and 10 min for M. This time lag was mainly due to the requirement of at least several images for calculation of RGR from the movement of structures within a given set of images. During the online monitoring for tissue sampling, only a mean RGR of a selected AOI was calculated (without spatial information). Samples were immediately frozen in liquid nitrogen after removal of the midvein and stored at −80°C until RNA isolation.

Measurement of epidermal cell size

Three leaves of Y were harvested in the morning and two discs (38.5 mm2) were removed from base, middle, and tip regions of each leaf. Leaf discs were incubated in 1 ml 90% phenol for 2 days. Epidermal cell size was then analysed on the adaxial surface of the samples with a microscope (Olympus BX 40) connected to a CCD-camera (Sony XC 003P). Data were collected from a total of 19–21 sections, avoiding the guard cells and veins, from six discs for each region.

Microarray analysis

Samples of Y and M were collected from the same plants during the night at time points of high (t1, 0:00–0:30) and low (t2, 4:00–4:30) RGR in Y. For M, only the samples at t1 were included in the microarray experiment (Mt1). Total RNA was isolated separately from each of the Y samples collected at t1 (Yt1, n=5) and t2 (Yt2, n=5) as well as Mt1 (n=2) according to Chang et al. (1993) with the following modifications: no spermidine was used in the extraction buffer and 2.6% (v/v) β-mercaptoethanol was used instead of 2%. After the extraction, RNA samples were cleaned up using the RNeasy kit (Qiagen). The RNA samples from different leaves were pooled for Yt1, Yt2 and Mt1. The cDNA synthesis, probe preparation, hybridisation conditions, data collection and data analysis were done according to Smith et al. (2004). The cDNA was prepared from 30 μg of total RNA. For each sample, about 500 ng of cDNA were subsequently labelled with fluorescent dyes (Cy3 and Cy5, Amersham). The labelled samples were heated to 95°C for 3 min, chilled on ice for 30 s and applied to the prehybridised slides. The POP2 arrays containing 24,912 clones representing over 100,000 poplar ESTs (Sterky et al. 2004) were used for transcript profiling. A total of 20 separate hybridisations to POP2 microarrays (8 replicates for Yt1 vs. Yt2, 6 for Yt1 vs. Mt1, and 6 for Yt2 vs. Mt1, including dye swaps) were performed.

Array slides were scanned with a ScanArray 4000 (PerkinElmer Sverige AB, Sweden) at 5-μm resolution at the wavelengths 543 and 633 nm for Cy3 and Cy5, respectively. The scanned images were analysed by GenePix Pro4.1 software (Axon Instruments, California, USA). Variation in gene expression was assessed by using a Limma (Linear Models for Microarray data) package (Wettenhall and Smyth 2004) in Bioconductor R 1.9.0 (http://www.bioconductor.org). Data were treated by loess normalisation after background subtraction (Edwards 2003) in Bioconductor. The changes in transcript abundance were analysed by calculating the in silico values of the average transcript level of all samples (Diaz et al. 2003). Expression intensities relative to this in silico average were then log2 transformed. For statistical analysis, B values (B-statistics, Wettenhall and Smyth 2004) and P values (Student’s t test) were used in combination. A total of 4,553 genes having B values >1 were included in the hierarchical cluster analysis using the average linkage method (Eisen et al. 1998). Genes in each cluster were considered highly significant and used for further analysis only when P<0.0015.

Northern blot analysis

Rapidly expanding leaves (LPI 4-5) were sampled at 14 different time points from seven plants in total, including the five plants used for the microarray experiment. At each of the 14 time points, sample leaves were taken from three randomly chosen plants while growth analysis was performed on a comparable leaf of the fourth plant. Samples were collected at two time points a day and sampling was repeated on different days in February and March, 2005. Leaves were frozen in liquid nitrogen after removal of leaf discs for carbohydrate analysis and the midvein and stored at −80°C until RNA isolation.

Isolation of total RNA was done as described for the microarray analysis. For Northern blot analysis, RNA from three replicates was pooled for each time point. Approximately 10 μg of total RNA were denatured by heating for 10 min at 65°C in a solution containing 50% (v/v) formamide, 6% formaldehyde (v/v) and 20 mM MOPS and fractionated by electrophoresis on a 1.3% agarose gel containing 6% formaldehyde (v/v) and 40 mM MOPS. Gels were blotted onto a nylon hybridisation transfer membrane (Roche) and rRNA bands were stained with methylen blue to check loading amounts. RNA was then cross-linked to the membrane with a UV Stratalinker 1800 (Stratagene).

Fragments coding a conserved region of RPL12 gene (212 bp, PU04952) and H2B gene (190 bp, PU10158) were amplified by PCR with cDNA as template. Primers used were 5′-CCAGAGAGAGACAGGAAG-3′ and 5′-CCATCAGTAATCTCCTGCTG-3′ for RPL12 and 5′-CAAGGCGATGGGTATCAT-3′ and 5′-AGTGACCGCTTTAGTTCCCTC-3′ for H2B. PCR products were purified with the QIAquick PCR purification Kit (Qiagen) and labelled with digoxygenin using PCR DIG Probe Synthesis Kit (Roche). Protocols and reagents for the chemiluminescent visualisation of RNA using CDP-Star (Roche) were according to the method supplied by the company. Signals from hybridised mRNA were detected with a luminescence image analyser LAS-3000 (Fuji Film).

Real-time RT-PCR

Real-time RT-PCR was performed on individual night-time samples used in the Northern blot analysis. For real-time RT-PCR experiments, mRNA was isolated from total RNA using GenElute mRNA Miniprep Kit (Sigma) to eliminate possible effects from changes in rRNA concentrations. Then cDNA was prepared from approximately 0.1 ng mRNA of each sample using iScript cDNA Synthesis Kit (Biorad) and labelled with iQ SYBRgreen (Biorad) using the method provided by the company. PCR reactions using the primers described above were run by iCycler equipped with MyiQ Detection System (Biorad) as follows: after the initial heating step of 95°C for 5 min, the plates were run at 95°C for 1 min for denaturation at the beginning of each cycle, then at 55°C for 1 min for annealing and 72°C for 1 min for extension. Absence of secondary products was checked at the end of each run by melting curves. The PCR efficiencies were calculated to be >95%. Relative expression was calculated from the differences in threshold cycle values between the sample with the highest value (i.e. the lowest expression level) and all samples. The sample with the highest threshold cycle value was found at 09:00 and 04:30 for H2B and RPL12, respectively.

Carbohydrate analysis

Soluble sugars and starch were extracted from leaf discs removed from the sample leaves for RNA analyses in February and March 2005 and analysed as previously described (Walter et al. 2005).

Results

Characterisation of spatio-temporal patterns in leaf growth

Leaves of P. deltoides maintain growth both in the light and in the dark (Stiles and Van Volkenburgh 2002). Figure 1a illustrates the typical diel variations in relative growth rate (RGR, % h−1). The RGR of young leaves increased from early morning towards the end of the light period, and thereafter, gradually decreased during the night. In other words, the diel growth pattern of P. deltoides was the reverse of those found in other dicot plants, such as castor bean (Schmundt et al. 1998) and tobacco (Walter and Schurr 2005), having the maximum and the minimum RGR at around dawn and dusk, respectively.

The abrupt light–dark (or dark–light) transition in the glasshouse caused a sharp transient increase (or decrease) in RGR in young leaves. Such rapid changes occur in growing tissues of many plants during light↔dark transitions, and this is associated with light-induced changes in apoplastic pH and/or turgor on leaf expansion (Christ 1978; Bertram and Lercari 1997; Stahlberg and Van Volkenburgh 1999; Stiles and Van Volkenburgh 2002). Avoiding these two sharp peaks as well as effects of changing sunlight, the samples for microarray analysis were harvested from young, rapidly expanding leaves (Y) and mature, non-expanding leaves (M) during the night at the time points of high (t1) and low (t2) RGR in young leaves (Fig. 1a). The average RGR (% h−1) at the time of sampling was 1.98±0.37 (SD) for Yt1 and 0.18±0.3 for Yt2 (differences are significant at P<0.0001). The leaves for M were not growing (RGR≈0).

In P. deltoides, regions of high and low growth rates were distributed in patches across the leaf lamina, and the shapes and spots of these patches changed with time (Fig. 1b). Such heterogeneous spatial growth patterns have also been documented in other dicot species (Walter and Schurr 2005). Notably, the spatial distribution of leaf growth in P. deltoides was characterised by the absence of a clear base-tip gradient (Fig. 1c). Moreover, we did not find any significant base-tip gradient in epidermal cell size (Fig. 1d), indicating that there was no obvious developmental gradient across the lamina. This spatial pattern, again, contrasts with previous observations in other dicot plants showing a base-tip gradient (Schmundt et al. 1998; Walter and Schurr 2005).

Hierarchical cluster analysis of differentially expressed genes

Transcriptome analysis of young and mature leaves, collected at two time points (Fig. 2), was used to identify genes controlling nocturnal leaf growth. Our experimental design was based on the following assumptions: (1) the comparison Yt1 versus Yt2 (young leaves at two different time points with different growth rates) reflects genes controlling temporal growth together with those regulated by the circadian clock; (2) the comparison Yt1 versus Mt1 (young and mature leaves at the same time point) contains genes in temporal leaf growth as well as leaf ontogeny; and (3) the comparison Yt2 versus Mt1 (young and mature leaves at two different time points) consists of circadian-regulated genes and genes in leaf ontogeny. Therefore, genes controlling the processes in temporal growth could be identified as the common group of genes between Yt1versus Yt2 and Yt1versus Mt1, those related to leaf ontogeny will appear in the common part between Yt1 versus Mt1 and Yt2 versus Mt1, and those under the circadian control between Yt1 versus Yt2 and Mt1 versus Yt2.
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-005-0181-0/MediaObjects/425_2005_181_Fig2_HTML.gif
Fig. 2

Schema of the experimental design for the microarray analysis. Rapidly expanding (Y) and fully expanded leaves (M) were harvested at t1 (0:00–0:30) and t2 (4:00–4:30) when high and low growth was detected in Y, respectively. (See Fig. 1 for the growth rates)

A total of 4,553 genes with B values >1 (B-statistics, Wettenhall and Smyth 2004) were included in the hierarchical cluster analysis (Eisen et al. 1998) shown in Fig. 3. The genes were clustered into two major subtrees containing (a) down- or (b) up-regulated genes in Mt1. Each of these subtrees was subdivided into three clusters (A1-3 and B1-3, respectively). According to our experimental design (Fig. 2), the clusters A1 and B1 comprise genes whose expression varies during leaf ontogeny (Y≠M). Clusters A2 and B2 show changes in gene expression characteristic of nocturnal growth in Y (Yt1≠both Yt2 and Mt1). The genes in A3 and B3 are likely to be under the control of the circadian clock (t1t2).
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-005-0181-0/MediaObjects/425_2005_181_Fig3_HTML.gif
Fig. 3

a, b Hierarchical cluster analysis of 4553 genes with differential expression in rapidly expanding (Y) and fully expanded leaves (M) of P. deltoides. Fold changes relative to the in silico average of all samples were colour-coded from red to green. Samples were harvested during the night at the time of low and high growth rates in Y (Yt2 and Yt1 in the first and second column from the left, respectively). For M (the third column from the left), only the samples collected at t1 were used for the microarray experiment. (See Fig. 2 for definition of Y, M, t1 and t2.) The constructed tree is presented in two large subtrees according to the expression ratios in Mt1: genes that were (a) down- or (b) up-regulated in Mt1. Clusters in both subtrees were named as indicated: A1–3 for (a) and B1–3 for (b). The number of the genes in each cluster was: A1, 1325; A2, 864; A3, 391; B1, 1349; B2, 480; B3, 144

Genes associated with leaf ontogeny

Table 1 shows a set of genes that were differentially expressed between Y and M (clusters A1 and B1 in Fig. 3). Different genes in organ development and morphology were up-regulated in Y and M, presumably reflecting a shift in developmental stage. Lipid metabolism and secondary metabolism are the two major functional groups up-regulated in young leaves of P. deltoides (Table 1). A gene encoding a B-like cyclin was found to be expressed in Y at both t1 and t2, possibly indicating ongoing proliferation activities (Beemster et al. 2005). In mature leaves (Table 1), on the other hand, we found marked up-regulation of genes encoding enzymes in cell wall modification, starch degradation (β-amylase) and synthesis of raffinose-family oligosaccharides (galactinol synthase). Mature leaves of P. deltoides also exhibited high transcript levels of metallothioneins, as has been found in mature leaves of P. trichocarpa x deltoides (Kohler et al. 2004).
Table 1

Some genes that were differentially expressed between expanding leaves (Y) and fully expanded leaves (M)

Clone ID

Description

Yt1/Mt1

Yt2/Mt1

Up-regulated in Y (Cluster A1)

 Morphogenesis, development

  PU03239

Homeobox-leucine zipper protein 5 (HAT5)

13.0

13.1

  PU25576

Argonaute protein (AGO1)

6.4

5.9

 Lipid metabolism

  PU09305

GDSL-motif lipase/hydrolase family protein, similar to family II lipase

25.5

28.1

  PU12468

GDSL-motif lipase/hydrolase family protein, similar to family II lipase

11.0

10.8

  PU27165

GDSL-motif lipase/hydrolase family protein, similar to family II lipase

24.2

30.1

 Transport

  PU08635

Plasma membrane intrinsic protein (SIMIP) (GI:2306917) [Arabidopsis thaliana]

6.1

4.0

  PU28531

Plasma membrane intrinsic protein (SIMIP) (GI:2306917) [Arabidopsis thaliana]

6.9

4.6

 Secondary metabolisms

  PU20140

Phenylalanine ammonia-lyase 1 (PAL1)

9.4

15.7

  PU24917

3-deoxy-d-arabino-heptulosonate 7-phosphate synthase 1 / DAHP synthetase

5.1

7.0

  PU12625

Chalcone synthase

6.9

12.2

  PU12470

Cytochrome P450 family protein

11.6

15.3

  PU12736

Flavonoid 3’-monooxygenase (EC 1.14.13.21)

8.0

11.2

 Others

  PU08741

Cyclin, similar to B-like cyclin (GI:780267) [Medicago sativa]

2.1

2.5

  PU10821

Protease inhibitor, putative

6.5

8.7

  PU12055

Protease inhibitor/lipid transfer protein (LTP) family protein

7.3

11.0

  PU12170

UDP-glucoronosyl/UDP-glucosyl transferase family protein

5.4

8.2

  PU24606

Peroxidase 3 precursor (EC 1.11.1.7)

9.8

11.3

  PU26508

Apyrase (APY2)

9.8

5.9

 

Mt1/Yt1

Mt1/Yt2

Up-regulated in M (Cluster B1)

 Morphogenesis, development

  PU10362

No apical meristem (NAM) family protein

2.8

2.7

  PU22203

No apical meristem (NAM) family protein

3.6

3.5

  PU10786

SEUSS transcriptional co-regulator (GI:18033922) [Arabidopsis thaliana]

10.6

9.1

 Protein degradation

  PU03249

Polyubiquitin (UBQ14)

10.5

7.6

 Cell-wall related

  PU08008

Pectinesterase family protein, similar to pectin methylesterase

6.4

3.6

  PU08930

Pectinesterase family protein

28.4

24.8

  PU11514

Endo-xyloglucan transferase (GI:2244732) [Gossypium hirsutum]

18.7

7.4

  PU26628

Glycosyl hydrolase family 3 protein

14.1

13.9

 Sugar, starch

  PU09978

Beta-amylase enzyme (GI:6065749) [Arabidopsis thaliana]

8.1

6.8

  PU09085

Galactinol synthase (GI:5608497) [Ajuga reptans]

21.1

11.7

  PU10261

Galactinol synthase (GI:5608497) [Ajuga reptans]

18.8

9.4

 Chloroplast

  PU10714

RuBisCO small subunit 1A chloroplast precursor (EC 4.1.1.39)

11.0

8.2

  PU10691

Lhcb2 protein (GI:4741946) [Arabidopsis thaliana]

5.0

4.9

  PU07331

DNAJ heat shock protein (J11) (GI:9843641) [Arabidopsis thaliana]

6.3

5.4

  PU10218

DNAJ heat shock protein (J11) (GI:9843641) [Arabidopsis thaliana]

5.6

4.1

 Others

  PU10656

Mitochondrial substrate carrier family protein

11.2

9.2

  PU03887

Metallothionein-like protein 2A (MT-2A) [Arabidopsis thaliana]

11.3

8.8

  PU05089

Metallothionein protein, putative

7.8

6.7

  PU25170

Metallothionein-like protein 2A (MT-2A) [Arabidopsis thaliana]

9.6

7.4

Fold changes are expression ratios between Yt1 (or Yt2) and Mt1, which were used for the cluster analysis presented in Fig. 3

Significant at P<0.0015 by Student’s t test

Genes associated with rapid growth at night

All genes in clusters A2 and B2 (Fig. 3) with P values <0.0015 are listed in Table 2. Central in the up-regulated genes in Yt1 is a series of ribosomal protein (RP) genes. The fact that replicate probes of some RP genes produced similar results demonstrates the reproducibility of the hybridisation. The concerted up-regulation of many different RP genes in Yt1 prompted us to look at expression patterns of these genes in different tissues and organs of poplar. The library distribution analysis done with Populus DB (http://www.populus.db.umu.se) indicated that EST (expressed sequence tag) clones of the RP genes with >2-fold up-regulation in Yt1 were most frequently isolated from meristematic tissues (Table 3). In particular, clones of 60S RPL12 have so far been obtained only from tissues with a high potential for cell proliferation. Although some cell-cycle genes (e.g. histones) were also clustered in A2 in the hierarchical cluster analysis (Fig. 3), most of them did not have P values <0.0015 in our statistical analysis, and therefore, they were not included in Table 2. In contrast to the long list of RP genes, we did not find any specific functional group that was down-regulated in Yt1.
Table 2

Genes that were up- or down-regulated in expanding leaves (Y) at a time point of rapid expansion (t1) compared with a time point of minimal expansion (t2) or fully expanded leaves (M) at t1

Clone ID

Description

Yt1/Yt2

Yt1/Mt1

Up-regulated in Yt1 (Cluster A2)

 Protein synthesis

  PU01870

40S ribosomal protein S3 (RPS3C)

1.7

1.9

  PU07127

40S ribosomal protein S5 (RPS5B)

2.0

2.0

  PU30290

40S ribosomal protein S5 (RPS5B)

1.7

1.8

  PU00739

40S ribosomal protein S7 (RPS7B)

1.8

1.9

  PU29333

40S ribosomal protein S7 (RPS7B)

1.8

2.0

  PU07461

40S ribosomal protein S9 (RPS9C)

1.6

1.8

  PU05294

40S ribosomal protein S11 (RPS11B)

2.1

2.3

  PU02737

40S ribosomal protein S15 (RPS15A)

1.8

2.0

  PU24357

40S ribosomal protein S15A (RPS15aA)

2.0

2.2

  PU01922

40S ribosomal protein S16 (RPS16A)

1.9

1.8

  PU20046

40S ribosomal protein S18 (RPS18A)

2.2

2.3

  PU00740

40S ribosomal protein S19 (RPS19C)

1.8

2.0

  PU13024

40S ribosomal protein S23 (RPS23B)

1.8

1.9

  PU02190

40S ribosomal protein S25 (RPS25E)

1.6

1.7

  PU07790

40S ribosomal protein S25 (RPS25E)

1.7

1.8

  PU25274

40S ribosomal protein S25 (RPS25E)

1.6

1.7

  PU00718

60S ribosomal protein L3 (RPL3A)

2.3

2.9

  PU30227

60S ribosomal protein L3 (RPL3A)

2.1

2.5

  PU07827

60S ribosomal protein L4/L1 (RPL4A)

1.9

1.9

  PU00933

60S ribosomal protein L5 (RPL5B)

1.6

1.8

  PU30997

60S ribosomal protein L7 (RPL7B)

1.8

1.7

  PU04527

60S ribosomal protein L7 (RPL7D)

2.1

2.4

  PU02516

60S ribosomal protein L8 (RPL8C)

1.7

1.8

  PU22081

60S ribosomal protein L8 (RPL8C)

1.8

2.0

  PU00746

60S ribosomal protein L9 (RPL90B)

1.9

2.0

  PU01466

60S ribosomal protein L9 (RPL90B)

1.8

2.1

  PU01936

60S ribosomal protein L9 (RPL90B)

2.2

2.5

  PU03023

60S ribosomal protein L9 (RPL90B)

2.0

2.0

  PU04125

60S ribosomal protein L9 (RPL90B)

2.1

2.5

  PU04725

60S ribosomal protein L9 (RPL90B)

2.2

2.6

  PU02554

60S ribosomal protein L10 (RPL10B)

2.0

2.4

  PU02356

60S ribosomal protein L10A (RPL10aB)

2.0

2.4

  PU04952

60S ribosomal protein L12 (RPL12C)

2.2

2.2

  PU02530

60S ribosomal protein L13A (RPL13aD)

1.8

2.0

  PU29421

60S ribosomal protein L13A (RPL13aD)

2.0

2.2

  PU00539

60S ribosomal protein L15 (RPL15A)

1.6

1.6

  PU05129

60S ribosomal protein L23 (RPL23A)

1.8

1.9

  PU29377

60S ribosomal protein L23 (RPL23A)

2.3

2.6

  PU31004

60S ribosomal protein L23 (RPL23A)

2.5

2.7

  PU02691

60S ribosomal protein L23A (RPL23aA)

1.8

1.7

  PU20388

60S ribosomal protein L24 (RPL24A)

1.8

1.9

  PU31026

60S ribosomal protein L24 (RPL24A)

1.8

1.9

  PU20650

60S ribosomal protein L27 (RPL27C)

2.1

2.3

  PU30941

60S ribosomal protein L27A (RPL27aC)

1.6

1.7

  PU01724

60S ribosomal protein L28 (RPL28A)

1.9

2.0

  PU06868

60S ribosomal protein L30 (RPL30A)

2.0

2.0

  PU25232

60S ribosomal protein L30 (RPL30A)

1.9

2.0

  PU29445

60S ribosomal protein L30 (RPL30A)

2.0

2.1

  PU01345

60S ribosomal protein L32 (RPL32A)

2.0

2.0

  PU01869

60S ribosomal protein L34 (RPL34A)

1.8

1.7

  PU02485

60S ribosomal protein L35a (RPL35aA)

1.8

1.8

  PU10393

60S ribosomal protein L35a (RPL35aA)

1.8

1.8

  PU22635

60S ribosomal protein L35a (RPL35aA)

1.9

1.9

  PU13104

60S acidic ribosomal protein P1 (RPP1A)

1.7

1.8

  PU22543

Elongation factor 1-alpha

1.8

1.9

  PU30819

Elongation factor 1-alpha

1.7

1.9

  PU09508

Dyskerin, putative / nucleolar protein NAP57, putative

1.7

1.8

  PU09216

Nascent polypeptide-associated complex (NAC) domain-containing protein

1.6

1.6

  PU06743

Protein disulfide isomerase, putative

2.4

2.4

  PU07520

Protein disulfide isomerase, putative

2.4

2.4

  PU03002

Heat shock protein 70, putative

1.6

1.7

  PU20344

Heat shock protein 70, putative

1.6

1.9

 Others

  PU02015

Phagocytosis and cell motility protein ELMO1-related

1.6

1.6

  PU02480

Adenine phosphoribosyltransferase 1 (APT1)

1.7

1.7

  PU02523

WD-40 repeat family / auxin-dependent (ARCA) / guanine nucleotide-binding protein

1.8

2.0

  PU07564

Nuclear RNA-binding protein (RGGA)

1.6

1.7

  PU07832

Fringe-related protein (Notch signalling)

1.8

1.8

  PU08348

CAAX amino terminal protease family protein

2.0

1.9

  PU26463

10 kDa chaperonin, putative

2.1

2.1

  PU29500

Nucleoside diphosphate kinase 1 (NDK1)

1.6

1.8

  PU30171

Synaptobrevin family protein, similar to vesicle-associated membrane protein 722

1.6

1.6

  PU05984

Unknown

1.6

1.6

  PU05675

Unknown

1.6

1.7

  PU02007

Unknown

1.7

1.8

  PU02271

Unknown

1.6

1.6

  PU03113

Unknown

1.9

1.9

  PU06771

Unknown

1.7

1.7

  PU08015

Unknown

1.6

1.7

 

Yt2/Yt1

Mt1/Yt1

Down-regulated in Yt1 (Cluster B2)

 Chloroplast

  PU08963

Malate dehydrogenase [NADP], chloroplast, putative

1.6

1.8

  PU09185

LytB family protein

2.3

2.1

  PU10136

LytB family protein

2.4

2.5

  PU10314

Thioredoxin, putative, similar to thioredoxin F-type, chloroplast precursor (TRX-F)

1.9

2.3

  PU10585

ABC transporter family protein

2.3

2.1

  PU10983

DNAJ heat shock protein

2.6

2.9

  PU10980

Zeaxanthin epoxidase (ZEP) (ABA1)

2.7

2.8

  PU25789

Zeaxanthin epoxidase (ZEP) (ABA1)

2.3

2.6

  PU10181

Unknown

2.6

2.2

  PU10202

Unknown

2.7

2.5

 Others

  PU03262

SOUL heme-binding family protein

2.1

2.1

  PU04549

Beta-amylase, putative (BMY3)

3.1

3.6

  PU05423

MutT/nudix family protein, putative

1.8

1.8

  PU12878

MutT/nudix family protein

3.0

2.2

  PU05436

Ferric reductase-like transmembrane component family protein

2.6

2.5

  PU30317

Ferric reductase-like transmembrane component family protein

2.2

2.4

  PU06126

Rieske [2Fe-2S] domain-containing protein, similar to cell death suppressor protein lls1

1.7

1.9

  PU06886

Glyoxalase II, putative / hydroxyacylglutathione hydrolase, putative

1.7

1.9

  PU09014

Phosphoglycolate phosphatase, putative

1.5

1.6

  PU09019

Basic helix-loop-helix (bHLH) family protein

1.7

1.8

  PU09750

ABC transporter family protein

2.0

1.7

  PU10255

Zinc finger (C3HC4-type RING finger) family protein (RMA1)

3.4

2.8

  PU06353

Unknown

2.4

2.7

  PU09087

Unknown

1.9

1.9

  PU10620

Unknown

1.9

2.1

  PU29264

Unknown

2.4

3.0

Fold changes are expression ratios between Yt1 and Yt2 (or Mt1), which were used for the cluster analysis presented in Fig. 3

Significant at P<0.0015 by Student’s t test

Table 3

Frequency of ribosomal protein genes cloned in the tissue-specific poplar cDNA libraries

Gene

Library (number of times cloned in the library)

Total number of times cloned

40S ribosomal protein S11

Shoot meristem (26)

95

60S ribosomal protein L3

Cambial zone (6)

12

60S ribosomal protein L7

Cambial zone (12)

22

60S ribosomal protein L9

Cambial zone (13)

22

60S ribosomal protein L10

Cambial zone (8)

14

60S ribosomal protein L10A

Cambial zone (4)

15

60S ribosomal protein L11

Root (5), shoot meristem (4)

18

60S ribosomal protein L12

Cambial zone (1), apical shoot (1), cambium (1), dormant bud (1), catkin (1)

5

60S ribosomal protein L17

Apical shoot (8), cambium (7), root (7), shoot meristem (7)

55

60S ribosomal protein L30

Shoot meristem (33)

121

60S ribosomal protein L32

Cambial zone (4)

6

Protein disulfide isomerase

Shoot meristem (7), tension wood (5)

28

Only clones with >90% sequence identity of >70% of the EST were considered

(Populus DB, http://www.populus.db.umu.se)

Genes associated with the circadian clock

Many genes that are proposed to be controlled by the circadian clock appear in Table 4. LATE ELONGATED HYPOCOTYL, a morning gene encoding one of the components in the central oscillator of the circadian clock in plants (Schaffer et al. 1998; Hayama and Coupland 2003), was highly expressed at t2. Concomitantly, transcripts of aquaporins and some chloroplast proteins were also abundant in Yt2. We found few genes that were up-regulated at t1 by >2-fold (cluster B3 in Fig. 3, Table 4). On the whole, the data in Table 4 are consistent with the reports in other microarray studies on circadian gene regulation in plants (Harmer et al. 2000; Schaffer et al. 2001).
Table 4

Some genes that were up-regulated by >2-fold at a time point of minimal (t2) or rapid expansion (t1)

Clone ID

Description

Yt2/Yt1

Yt2/Mt1

Up-regulated at t2 (Cluster A3)

 Transcription

  PU13388

MYB family transcription factor, LATE ELONGATED HYPOCOTYL (GI:3281845)

12.4

6.2

 Chloroplast

  PU07325

Photosystem I reaction center subunit psaK

2.2

2.4

  PU08421

Photosystem I reaction center subunit psaK

2.4

2.4

  PU21417

Chlorophyll a oxygenase (CAO)

2.8

2.8

  PU09033

NADP-dependent glyceraldehyde-3-phosphate dehydrogenase, putative

2.2

2.6

  PU09037

NADP-dependent glyceraldehyde-3-phosphate dehydrogenase, putative

2.6

3.3

  PU12091

Starch synthase, putative

4.7

3.5

  PU13521

Starch synthase, putative

5.9

4.2

  PU20746

Chloroplast inner envelope membrane protein, putative

2.4

2.1

 Transport

  PU02076

Plasma membrane intrinsic protein 2A (PIP2A) / aquaporin PIP2.1 (PIP2.1)

2.9

3.0

  PU02888

Plasma membrane intrinsic protein 2A (PIP2A) / aquaporin PIP2.1 (PIP2.1)

3.2

3.2

  PU12492

Plasma membrane intrinsic protein 2A (PIP2A) / aquaporin PIP2.1 (PIP2.1)

3.2

3.7

  PU25215

Plasma membrane intrinsic protein, putative

2.5

2.4

  PU28958

Plasma membrane intrinsic protein, putative

3.2

2.8

  PU09309

Nitrate transporter (NTP3)

2.4

2.7

 Others

  PU03292

Beta-galactosidase, putative/lactase, putative

3.9

3.4

  PU08469

Cell division control protein 2 homolog A (CDC2A), CDKA;1

2.3

2.2

  PU10139

Mitogen-activated protein kinase (MPK17), putative

4.3

4.4

  PU10930

Pyrophosphate-fructose-6-phosphate 1-phosphotransferase-related

2.3

2.5

  PU12494

Subtilase family protein, similar to meiotic serine proteinase

2.4

2.3

  PU25137

Haloacid dehalogenase-like hydrolase family protein

3.3

3.1

  PU08943

Unknown

3.9

4.5

  PU11540

Unknown

2.6

2.4

  PU31010

Unknown

2.2

2.2

 

Yt1/Yt2

Mt1/Yt2

Up-regulated at t1 (Cluster B3)

  PU09373

CONSTANS-like protein-related

2.4

2.3

  PU11305

Thiazole biosynthetic enzyme, chloroplast (ARA6) [Arabidopsis thaliana]

2.8

3.7

  PU08396

Thiazole biosynthetic enzyme, chloroplast (ARA6) [Arabidopsis thaliana]

2.9

3.7

  PU31080

Glycine-rich RNA-binding protein (GRP7) [Arabidopsis thaliana]

2.6

3.4

  PU01755

Glycine-rich RNA-binding protein (GRP7) [Arabidopsis thaliana]

2.3

2.4

Fold changes are expression ratios between Yt2 and Yt1 (or Mt1), which were used for the cluster analysis presented in Fig. 3

Significant at P<0.0015 by Student’s t test

Nocturnal expression patterns of RPL12 and histone H2B

In order to verify the up-regulation of RP genes at a point of high nocturnal RGR, additional experiments were conducted using leaves of the same developmental stage as Y under a similar growth condition that reproduced the growth rates and patterns found in the leaf growth analysis for the microarray experiment (Figs. 1a, 4a). Expression of RPL12, presumably characteristic of proliferative activities (Table 3), and histone H2B, a cell-cycle gene that appeared in cluster A2 in Fig. 3 but was not included in Table 2 due to the P value cutoff, were closely analysed in this experiment comprised 14 time points (Fig. 4a).
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-005-0181-0/MediaObjects/425_2005_181_Fig4_HTML.gif
Fig. 4

ac Diel time-course analysis of relative growth rates (RGR, % h−1) and expression levels of ribosomal protein L12 (RPL12) gene and histone H2B gene in rapidly expanding leaves of P. deltoides. a Diel changes in hourly mean of RGR in rapidly expanding leaves (LPI=4–5). Symbols are a mean of seven different leaves (error bars, ± SD). Light↔dark transitions caused a sharp increase and decrease in RGR at 22:00 and 6:00, respectively. Bars on the x-axis denote sampling points for RNA analysis. b Northern blot analysis of the diel expression patterns of RPL12 and H2B. Total RNA was isolated from three replicate leaves collected at 14 time points indicated in (a). RNA samples were pooled for each time point and ca. 10 μg of total RNA were loaded. Two bands of rRNA visualised by methylen blue staining show the loading amount. c Nocturnal changes in the expression levels of RPL12 and H2B as quantified by real-time RT-PCR. mRNA was isolated from the total RNA of each night-time sample to eliminate possible effects from variations in rRNA levels. Each bar is a mean of three samples (error bars, ± SD) that were pooled for the Northern-blot analysis in (b)

First, Northern blot analysis was performed using total RNA samples pooled from three replicate leaves collected at each time point. The results clearly indicated down-regulation of these two genes with decreasing RGR from midnight to dawn (Fig. 4b). Although it seems that both genes responded to the dark–light transition by transient up-regulation (07:00), the expression levels remained mostly low during the phase of low RGR. The marked temporal variations of RPL12 expression, however, also suggested possible changes in concentrations of ribosomal RNAs (rRNAs) that, together with RPs, form the ribosome complex. In an attempt to minimise the effects, if any, of varying rRNA levels, which normally correspond to 98–99% of total RNA, mRNA was isolated from the total RNA of the night-time samples and used for a quantitative analysis with real-time RT-PCR.

The results from the real-time RT-PCR experiment, using the same amounts of mRNA and measuring individual replicates separately instead of pooling them together, confirmed the expression patterns of these genes found in the Northern blot analysis (Fig. 4c). The transcript levels of H2B were relatively high until 00:45 but thereafter decreased rapidly (Fig. 4c). Likewise, the samples collected at 00:45 showed high expression levels of RPL12, but the post-midnight decline was not as rapid as for H2B. When the expression levels between the two time points at around t1 and t2 were compared with each other (00:45 and 04:30 in this experiment, respectively), H2B was up-regulated at t1 by 2.2-fold and RPL12 by 2.6-fold (both significant at P<0.05).

Variations in glucose, fructose, sucrose and starch

The energy and structural components required for leaf growth are supplied by carbohydrates, and the concentrations of carbohydrates often undergo pronounced diel variations in leaves. Furthermore, the cell cycle, especially at the transition from G1 to S phase, is sensitive to sugar concentrations (den Boer and Murray 2000; Stals and Inzé 2001). Hence, concentrations of soluble sugars (glucose, fructose and sucrose) and starch were determined in the leaves used in the Northern blot and real-time RT-PCR analyses shown in Fig. 4. Carbohydrates accumulated during the day and decreased in the evening and at night (Fig. 5) in much the same way as has been found in other studies on growing poplar leaves (Mayrhofer et al. 2004; Walter et al. 2005). High concentrations of carbohydrates coincided with high RGR in the afternoon and in the evening (Figs. 4a, 5). However, the nocturnal decrease in RGR, which started after the peak expression of H2B at around midnight (Fig. 4), was delayed for 2–3 h compared with the corresponding changes in the carbohydrate concentrations (Fig. 5).
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-005-0181-0/MediaObjects/425_2005_181_Fig5_HTML.gif
Fig. 5

a–c Diel changes in concentrations of glucose and fructose (a), sucrose (b) and starch (c) extracted from the leaf samples used for the Northern blot analysis and the real-time RT-PCR analysis shown in Fig. 4. Each symbol is a mean of three samples (error bars, ± SD)

Discussion

Spatio-temporal growth pattern in leaves of P. deltoides

Leaf growth rates vary dynamically over the course of a day (Dale 1988; Walter and Schurr 2005), with maximal rates of growth occurring at dawn in some plants and at dusk in others (Walter and Schurr 2005; and references therein). Our growth analysis indicated that P. deltoides belongs to the latter group (Figs. 1a, 4a). In the study by Stiles and Van Volkenburgh (2002), leaves of P. deltoides maintained higher growth rates in the dark than P. trichocarpa (because light strongly stimulated leaf growth in P. trichocarpa). The leaves of P. trichocarpa might grow mostly by cell expansion after completing cell proliferation at 10–20% of the final size whereas cell proliferation continues in P. deltoides until 80–90% of the final size (Van Volkenburgh and Taylor 1996). Prolonged cell proliferation has also been documented for leaves of sunflower, in which a base-tip gradient did not appear during the exponential growth phase involving active cell proliferation but did appear with basipetal cessation of cell division (Granier and Tardieu 1998). Thus, the lack of a base-tip gradient in P. deltoides leaves during the exponential growth phase is indicative of a continuation of cell proliferation.

Coordinated up-regulation of RP genes in rapid leaf expansion

The gene expression pattern in young leaves at the time of high RGR highlighted the concerted up-regulation of genes associated with protein synthesis, including a number of ribosomal protein (RP) genes (Table 2). Based on these results and the observation from the Northern blot and the real-time RT-PCR analyses of RPL12 (Fig. 4b, c), we propose that the nocturnal changes in RGR in young leaves of P. deltoides (Fig. 1a) could engage regulation of ribosome biosynthesis. Biosynthesis of ribosome is a process involved in cytoplasmic growth during proliferation and endoreduplication (Sugimoto-Shirasu and Roberts 2003). In fact, many RP genes were found in tissues with high meristematic activities (Table 3, Sterky et al. 2004).

The diel expression patterns of H2B and RPL12 did not quite parallel each other (Fig. 4b, c). While H2B, whose expression is specific to S phase of the cell cycle as reported for tobacco and Arabidopsis (Breyne et al. 2002; Menges et al. 2002), peaked at around midnight followed by an abrupt decrease, nocturnal changes in the transcript levels of RPL12 were rather gradual (Fig. 4c). It has been suggested that biosynthesis of ribosomes, expected at the G1/S transition, may be a determinant for the attainment of a critical cell size to proceed from G1 to S phase (Cuadrado et al. 1985). During cytoplasmic growth, cells must increase the cytoplasmic mass, which is achieved inter alia by accumulating macromolecules in the cytoplasm, such as numerous ribosomes. Some RP genes may play regulatory roles in cytoplasmic growth and proliferation, as has been manifested by mutation of RP genes in Arabidopsis and tobacco plants (RPS18, Van Lijsebettens et al. 1994; RPS13, Ito et al. 2000; RPS5, Weijers et al. 2001; RPL3, Popescu and Tumer 2004). These RP-gene mutants share a common phenotype of displaying delayed or disturbed cell division and development. Ribosome biosynthesis is controlled at the transcriptional and translational level in yeast and mammalian cells, respectively (Jorgensen et al. 2004). Our microarray analysis, revealing a concerted up-regulation of many RP genes at the time of high RGR in growing leaves of P. deltoides (Table 2), suggests a transcriptional control of RP synthesis and its association with temporal regulation of cytoplasmic growth in plants.

Nocturnal leaf growth and cell cycle control

Diel growth rhythms with the maximum from dusk to midnight, such as those found in leaves of P. deltoides (Figs. 1a, 4a), have also been documented for other higher plant species (Bünning 1952). Furthermore, the unicellular algae Chlamydomonas reinhardtii and Euglena gracilis preferentially divide in the dark and this has been associated with circadian gating of cell division (Goto and Johnson 1995; Bolige et al. 2005). It was postulated that this growth rhythm in C. reinhardtii may be an adaptive strategy to avoid deleterious effects of UV radiation on DNA replication (Nikaido and Johnson 2000), which takes place during S phase of the cell cycle. Notably, the temporal pattern of H2B transcription, an S phase-specific gene (Breyne et al. 2002; Menges et al. 2002), indicated that the population of cells entering S phase may have been dramatically reduced after midnight and during the morning in rapidly expanding leaves of P. deltoides (Fig. 4b, c). Unlike in monocots, in which the meristematic region of growing leaves is protected from the environment, proliferating cells in dicot leaves are often fully exposed to the sunlight. In addition to other protective mechanisms against UV-radiation, such as DNA repair and accumulation of UV-absorbing compounds (Stapleton and Walbot 1994; Britt 1996; Krause et al. 2003), UV-avoidance of S-phase entrance could be advantageous for rapidly expanding dicot leaves during the stage of high proliferation activities.

Diel changes in sugar concentration and cell growth

The synthesis of new ribosomes, which consumes a large fraction of the cell’s resources, represents an investment for future growth (Warner 1999). Because this investment will be profitable only if the other essentials are also available for growth (Warner 1999), cells may have mechanisms to perceive nutrient availability and regulate ribosome biosynthesis accordingly. Transcription of RP genes is regulated by carbon and nitrogen availability in yeast (Jorgensen et al. 2004; Schawalder et al. 2004). In plants, nitrogen deprivation results in a coordinated decline of cell division and cell expansion in growing leaves (Roggatz et al. 1999) and that sugar levels are sensed during G1 phase in the cell cycle (den Boer and Murray 2000; Stals and Inzé 2001). We note that the nocturnal decline in leaf growth (Figs. 1a, 4a) was preceded by a reduction in carbohydrate concentrations (Fig. 5). Although a cause–effect relationship cannot be inferred from this study, the results conform to a scenario in which active growth in the evening contributes to decreasing cellular carbohydrate levels, which in turn may induce a feedback down-regulation of RP genes, leading to deceleration of cytoplasmic growth.

Concluding remarks

Whole plant growth activity (biomass production) is closely related to the total leaf area in many plants (Van der Werf 1996). Understanding of leaf growth processes and their responses to intrinsic and extrinsic stimuli is hence a key to improve crop and forest management as well as breeding strategies. An emerging picture of diel leaf growth mechanisms in P. deltoides highlights cytoplasmic growth, requiring up-regulation of ribosome biosynthesis (Table 2), as the primary process underlying the nocturnal variations in the leaf growth of this species. The diel growth rhythms which favour maximal RGR in the evening and at night (Figs. 1a, 4a) may be an adaptive strategy of timing DNA replication out of phase with UV-radiation (Fig. 4b, c; Nikaido and Johnson 2000). Superimposed on the circadian gating of cell cycle (Goto and Johnson 1995; Bolige et al. 2005), diel variations in some growth factors, such as sugars (Fig. 5), could also influence the rate and/or the duration of temporal growth processes. The fact that circadian clock-period mutants of Arabidopsis exhibited reduced rates of growth and photosynthesis (Dodd et al. 2005) underpins the role of the circadian clock in coordinating growth and carbohydrate metabolism in the terrestrial day–night cycle. Further studies are needed to elucidate the circadian regulation of diel leaf growth patterns, including another type with maximal RGR from dawn to early morning.

Acknowledgements

Kind assistance by Ingela Sandström, Andreas Sjödin (UPSC, Umeå) and Volker Wendisch (FZJ, Jülich) for the microarray analysis, Katharina Klug and Kate Morrissey (FZJ, Jülich) for RNA analyses, and Maja Christ and Kerstin Nagel (FZJ, Jülich) for carbohydrate assay is greatly acknowledged. The microarray experiment at UPSC was supported by Forest Research Environment (FORE, Umeå), the European Community’s programme “Improving the Human Research Potential and the Socio-Economic Knowledge Base”, the “Project-based Personnel Exchange Programme” of Deutscher Akademischer Austauschdienst (DAAD), and by a grant from the Swedish Forestry and Agricultural Research Council to VH.

Copyright information

© Springer-Verlag 2005