Identification of transcription factors linked to cell cycle regulation in Arabidopsis

Abstract

Cell cycle is an essential process in growth and development of living organisms consists of the replication and mitotic phases separated by 2 gap phases; G1 and G2. It is tightly controlled at the molecular level and especially at the level of transcription. Precise regulation of the cell cycle is of central significance for plant growth and development and transcription factors are global regulators of gene expression playing essential roles in cell cycle regulation. This study has uncovered TFs that are involved in the control of cell cycle progression. With the aid of multi-parallel quantitative RT-PCR, the expression changes of 1880 TFs represented in the Arabidopsis TF platform was monitored in Arabidopsis synchronous MM2d cells during a 19 h period representing different time points corresponding to the 4 cell cycle phases after treatment of MM2d cells with Aphidicolin. Comparative TF expression analyses performed on synchronous cells resulted in the identification of 239 TFs differentially expressed during the cell cycle, while about one third of TFs were constitutively expressed through all time points. Phase-specific TFs were also identified.

Introduction

Cell cycle is one of the most fundamental and elaborate processes of living organisms consist of 4 different phases. In summary, in S phase DNA is replicated. This phase is separated from cell division (M phase) by 2 gaps, G1 and G2, phases. During G1 phase, cell grows and duplicates its organelles but its DNA is not replicated. In G2 phase, cell growth continues and proteins are synthesized in preparation for mitosis. In this stage, completion of DNA synthesis precisely is monitored too.^1,2 Cell cycle is driven by a complex series of events, each precisely regulated at multiple levels.³ This process is controlled by regulatory networks including a diverse group of proteins through protein-protein interactions or chromatin modification. Regulatory proteins involve conserved core set of proteins such as CDKs, cyclins, CDK activators and inhibitors and transcription factors (TFs).^4,5 Transcriptional regulation via specific DNA-binding TFs is the first step of gene expression and plays more important roles in plants than in animals.⁶ TFs are key regulatory proteins that bind to short conserved regions in their target promoters which are called cis-elements or motifs. These groups of proteins are responsible for enhancing or repressing the transcriptional rate of their target genes. They control temporal and spatial expression patterns of target genes in response to intrinsic and external signals.^7,8 In Arabidopsis, it has been estimated that 5–10% of the genes encode TFs that comprises over 2000 TFs.⁶ The identification of cell cycle-regulated genes requires highly synchronized cell cultures.⁹ Suspension cultures, in which plant cells grow uniformly in liquid medium, provide an ideal tool for functional genomics approaches particularly for transcript profiling of cell cycle regulation.¹⁰ In such systems, dividing cells are in an asynchronous state or different stages of the cell cycle that makes it difficult to study of cell cycle-regulated events. Synchronization is a technique commonly used to collect cells at a specific point of the cell cycle from an asynchronous culture. Then, synchronized suspension cultures are powerful tools enabling study of the cell cycle regulation. Synchronization can be induced through physical, chemical and nonchemical methods. In physical method, cells sort by flow-cytometry based on their size and DNA content.^11,12 Induction of synchrony can be achieved by starvation or nutritional deficiency of essential compounds for growth such as hormones, phosphate, nitrate or sucrose ^13-15 or by applying specific inhibitors blocking the progression of cell cycle events beyond a given point in the cycle.¹² Utilizing of DNA synthesis inhibitors have been frequent means to obtain synchronized plant cells. In Arabidopsis, cells accumulate in S phase using a treatment with Aphidicolin. Aphidicolin is a fungal toxin derived from Cephalosporium aphidicola that inhibits DNA polymerase α and δ and results in reversible arrest of the cells at S phase. After releasing from the block, cells resume DNA synthesis and synchronously pass S phase and enter G2 phase.^16,17 TF genes are generally expressed at relatively low levels. Due to technical limitations, DNA arrays technologies are unable to accurately detect and quantify transcripts of many TF genes.¹⁸ In contrast, sensitive and reproducible quantitative real-time reverse-transcription PCR (qRT–PCR) allows detection of even low-abundance transcripts.^19,20 To identify cell cycle-regulated genes, several genome-wide analysis have been performed using microarray technologies that led to identify rather few TFs. Studies on cell cycle-regulated genes using qRT-PCR have not been widely reported. In this study, expression levels of thousands of Arabidopsis TFs whose regulatory activities fluctuate periodically across the cell cycle were measured simultaneously by qRT-PCR. For doing such large-scale of profiling experiment, Arabidopsis synchronized cell cultures using Aphidicolin and available Arabidopsis TFs qRT-PCR platform were used. For the first time, synchrony was evaluated using 3 separate procedures including flow-cytometery, analysis of metaphase/anaphase index (M/AI) and expression pattern of 150 well known cell cycle marker genes which differentiates different phases.

Results

Despite genome wide expression analysis of cell cycle-regulated genes in Arabidopsis MM2d cells using microarray technology,¹⁰ just a few TFs involved in cell cycle regulation were identified so far. Hence I sought to use Arabidopsis MM2d cells as well for the analysis of cellular responses at the transcriptional level by qRT-PCR technique.

MM2d cell suspension culture

The Arabidopsis fast growing MM2d cells were maintained by weekly sub-culturing and cell density was determined daily from sub-culture (0) to 8 d later. In the present experiment the mode of MM2d cells growth was similar that of which described by.²¹ The growth of Arabidopsis cell cultures followed standard growth kinetics in which the number of cells increased exponentially due to high rate of cell division followed by slow rate of cell increase during stationary phase. Cells number reached a peak at day 5 and then declined. For the experiment, samples (MM2d cells) were harvested at 5 d after sub-culture representing early stationary phase. In fifth day MM2d cells showed the expansion and elongation characterization of early stationary phase (Fig. 1).

Figure 1.

Growth pattern of the MM2d cell line (left) and early stationary phase (D5) (right). Cells were sub-cultured and cell growth of cell lines MM2d monitored by determination of the number of cells (×10⁵) per mL culture medium by hemocytometer.

Cell cycle synchronization

MM2d cell suspension and a procedure for achieving highly synchronization at S phase using Aphidicolin have been previously described by.²¹ In order to study genes expression pattern during cell cycle progression, early stationary phase MM2d cells for inducing synchrony by Aphidicolin block/release treatment was used. Synchrony was evaluated using incorporation of flow-cytometery, analysis of metaphase/anaphase index (M/AI) and expression pattern of cell cycle marker genes which differentiates different phases.^11,12,22 Flow-cytometery analysis was done to monitor cell cycle synchronization over a 19 h period after treatment of MM2d cells with Aphidicolin for 24 h and subsequent removal of the block (totally 10 time points). Obtained data indicated that the majority of the cells proceed synchronous through S phase with an S-phase peak of 74% immediately after release (Fig. 2). In order to prove efficiency of Aphidicolin blocking effect on the treated cells and proper synchronization, M/AI was measured in the same samples using DAPI staining method and fluorescent microscopy. In the microscopic observations, the peak of the M/AI was reached to about 12% after 10 h (corresponding to M phase) which reflects relatively short length of metaphase and anaphase in mitosis (Fig. 3). The trend and values of M/AI in each time were in agreement with.¹⁰ It has been mentioned that small size of the Arabidopsis genome and late condensation of the chromosomes in prophase make scoring of prophase and telophase stages of mitosis difficult and lead to underestimate the number of cells going through mitosis.²¹ Flow-cytometry and M/AI results provided us 2 references for proving synchronization efficiency. It is known that certain cell cycle regulators show clear patterns of expression during the cell cycle. These genes may affect cell cycle progression directly or indirectly. For instance, based on previous studies, AtCPR, a cell cycle-related gene from Arabidopsis thaliana, was regulated in a cell cycle phase-dependent manner. The transcript of AtCPR accumulated at the G1 to S phase.²³ Three auxin-activated genes, AXR1, ARGOS and ANT, control cell number and modify organ size through elevating the expression of the CYCD3;1 gene in G1 to S phase transition.²⁴ Among many induced TFs in Aphidicolin block/release experiments, the expression of AT2G34140 (AtDOF2;3), AT2G34140 (both from DOF family), and AT2G40970 (MYBC1) was associated with the S phase and AT3G07340 (a member of bHLH family), AT2G40750 (AtWRKY54) and AT1G08810 (AtMYB60) have been identified as specific markers in G1, G2 and M phases respectively.¹⁰ Moreover, the interaction between E2F transcription factors and RETINOBLASTOMA-RELATED (RBRs) is essential for the control of G1 to S phase transition. The heterodimeric transcription factor E2FA/DP is one of the key activators of S phase entry.²⁴ In this regard Skirycz et al have described that the DOF transcription factor OBP1 modulates cell cycle activity in Arabidopsis and causes transition from G1 to S phase. OBP1 acts through affecting the expression of CDKA regulators and S phase-specific TFs.²⁵ Hence, cell cycle progression in Aphidicolin-synchronized cells was validated using various marker genes. To this aim, I focused on 150 well known marker genes including TFs and core cell cycle regulators with specific expression patterns in various cell cycle phases.^10,26 To sum up, expression pattern of given regulators of the cell cycle represented a high degree of synchrony in the experiment (see below).

Figure 2.

Flocytometry resault shows that majority of the cells proceed synchronous through S phase with an S-phase peak of 74% immediately after release of Aphidicolin.

Figure 3.

Microscopic observations. Peak of the M/AI was reached to about 12% in 10h after removal of Aphidicolin corresponding to M phase.

Identification of cell cycle-regulated gene expression profiles

To confirm synchronization efficiency more in addition to the results obtained from flow-cytometry and M/AI analysis, the expression patterns of several genes known to be cell cycle-related were determined by qRT-PCR technology and compared with the previously published microarray data. Here in, to establish a new reference of synchronization efficiency, expression patterns of the genes in different time points corresponding to specific cell cycle phase were monitored. To this aim, I focused on expression patterns of 150 cell cycle- related genes including 81 TFs and 69 core cell cycle genes. The genes involve in a number of cellular processes including hormone response, signal transduction, transcription control, and metabolic regulation. For the experiment, MM2d Arabidopsis cells were reversibly blocked in G1/S boundary by using Aphidicolin and sequential samples taken at 2 hourly intervals after releasing of the inhibitor till 16 h followed by a final sampling at 19 h (totally 10 time points). Expression pattern of the 150 cell cycle- related genes was detected using qRT-PCR in treated MM2d cell suspension. The comparative CT method²⁷ was the method of choice for calculating fluctuations in genes expression. Briefly, the CT value of a gene of interest was adjusted based on the CT value of an endogenous reference gene, UBQ10 (AT4G05320) and represented as ΔCT. The fold change (FC) value for the gene was calculated for samples treated with Aphidicolin in comparison with control samples in time immediately after removal of Aphidicolin (T0 as calibrator). To this purpose, ΔCT of each time point was subtracted from that of time zero, resulting in ΔΔCT as fold change (equations 1 and 2). Equation 1: ΔC_Tg = C_Tg–C_{T UBQ10} and equation 2: ΔΔC_Tg = ΔC_Tg–ΔC_{T, T0}

In order to uncover groups of co-regulated genes, that are genes with similar expression patterns, the 150 cell cycle–related genes were subjected to cluster analysis based on their expression level. Similarity in expression patterns of the 150 genes was assessed by creating a dendrogram based on normalized and comparative expression levels of the genes. This hierarchical cluster analysis clearly shows that different groups of genes show peaks of expression at specific time points throughout the time course (Fig. 4). Abridged branch analysis resulted in the creation of sub-branches or nodes (A-H), which reflects difference in expression pattern and timing. Although the hierarchical cluster analysis is a useful tool for grouping genes based on similarity of expression timing, it may not reflect the diversity of different regulatory patterns. Therefore, the data set also clustered using K-means clustering on Euclidean distance values. According to K-means procedure, 15 clusters were determined to be the optimal number of groups for the data. The expression profiles of the genes in each cluster are shown in Figure 5. I had aim using this clustering to examine whether known cell cycle co-regulated genes were assigned to the same or similar clusters and refer to proving synchronization.

Figure 4.

Hierarchical clustering of cell cycle- relaated genes including 69 core cell cycle genes and 81 transcription factors. The fold change values for each sample relative to time zero as control were log₂ transformed. The mean values for the 3 independent biological replications were subjected to complete linkage hierarchical clustering. The color saturation reflects the magnitude of the averaged log₂FC. Expression values higher and lower than those of the control are shown in red and green respectively. The vertical dendrogram (left) indicates the relationship among the genes regarding their expression patterns. The color scale (top) indicates the color assigned to each log₂FC. The pattern has been created using MeV software ver.4.3.

Figure 5.

Expression profiles of cell cycle- related genes including 69 core cell cycle genes and 81 transcription factors. Cell cycle-regulated gene expression profiles are plotted against the different time points after removal of Aphidicolin. The pattern has been created using MeV software ver.4.3.

Genes Show Co-Regulation in Cell Cycle Synchronization

Cyclin genes

The majority of cyclin genes being represented in nodes A and E (Fig. 4) showing similar trend in expression and group in same or near clusters (Fig. 5 clusters4, 5, 6, and 7). Expression of almost all cyclins type A and B declines in G1/S boundary but rises during S phase and shows one peak around 6 h corresponding to the early G2 phase. For this group of cyclins, a significant increase of expression is observable at 12 h, followed by a sharp peak at 14 h related to mitosis. Despite common pattern of regulation among CYCB, CYCA1 and CYCA2 cyclins, there are differences in their expression levels (Fig. 6A), approximately with comparative detected levels of CYCB1≥CYCB2>CYCB3 and CYCA2≥CYCA1. CYCA3;1 and CYCA3;2 genes along with CKL5 and 15 TFs group in cluster 1 (Fig. 5). This group of genes shows highest expression level in S phase and then a decline through cell cycle progression indicating sequential wave pattern. Based on presented data, D-type cyclins show oscillatory behavior which is in agreement with previous study ^26,28 (Fig. 4 nodes D and E; Fig. 2 clusters5 and 6). During cell cycle progression CYCD3;1 and CYCD4;1 transcripts accumulate early in G2 and G1 phases and decline twice thereafter; in G2/M transition and as cells move toward S phase. In MM2d cells, CYCD3;2 and CYCD6;1 genes having similar expression profiles express in higher level in G1 phase in comparison with 2 former cyclin D. CYCD1;1 and CYCD3;3 are seen in cluster5, increases constitutively with expression peak in G1 phase whereas CYCD5;1 (Fig. 5 cluster5) together with CYCD2;1 (Fig. 5 cluster6) reveals almost same behavior with expression peaks in early G2 and G1 phases (Fig. 6B).

Figure 6.

Expression patterns of several cell cycle-related genes in Aphidicolin-induced synchrony (A-K). The fold change values for each sample relative to time zero as control were log₂ transformed. The mean values for the 3 independent biological replications were plotted as Log₂FC.

CDK genes

Expression of CDKB2;1 and CDKB2;2 decreases in S cells whereas increases from G2 phase continuously and show a significant increase of expression in 14 h corresponding to the late mitosis and then remains constant thereafter (Fig. 6C). Both CDKB2 genes fall into nodeA/cluster7 with cyclins A and B, which are thus co-regulated with mitotic cyclins (Fig. 4 nodeA; Fig. 5 cluster7). In spite of relatively low M/AI recorded for the reasons mentioned above, the robust identification of mitosis-specific genes validates the synchrony of the culture used. Transcript profiling analysis clearly shows that the expression of CDKA1 is highly fluctuating in the cell cycle. Expression of CDKA1 decreases in G1/S transition but reaches a peak in S phase, again declines in G2 phase and shows second peak at 14 h in late mitosis. CDKA1 clusters along with CKL7, DPa, CYCD4;1 and 4 TFs show a highly similar pattern of expression across cell cycle progression (Fig. 5 cluster10; Fig. 6D). Among all genes in this study, CDKD1 shows highest expression level with a distinct peak at 10 h related to mitosis. In contrast, CDKD2 is constantly expressed at a lower level and CDKC2 is up-regulated after Aphidicolin block release. This gene shows a local decrease and increase in late S and G2 respectively and expressed with low fluctuation thereafter, as CDKG2, CDKE1 and CDKF1 (Fig. 5 clusters3 and 6; Fig. 6E). Also in contrast to CDKD1, CDKC1 is upregulated in G1/S boundary and declines through G2 and M phases, shows a peak in late mitosis and decreases thereafter (Fig. 6E).

KRP genes

Among the KRP genes, KRP2 is expressed at high level in Aphidicolin-blocked cells. All KRP genes follow similar trend in expression except KRP4 which indicates higher degree of fluctuation through its expression. KRP genes show 2 distinct peaks in early G2 phase and M/G1 transition. In particular, KRP2 shows a clear peak of expression at late mitosis (Fig. 6F).

Other cell cycle regulators

A similar expression profile is observed for the E2F/DP/DEL families and RB gene (Fig. 4 nodeH). Their expression shows a decline through S phase while their levels increase in S/G2 boundary and continue to rise through M phase. This trend reaches to a peak in 14 h corresponding to late mitosis (Fig. 6G). The members of CKL gene family appeared to have similar pattern of expression especially closely related CKL5 and CKL6 genes and CKL10, 11, 12, 14 along with MCM5. They increase in S cells so that reaches a peak at S/G2 transition, increase toward M phase and show another peak in M/G1 boundary (Fig. 4 nodeH; Fig. 6H). In contrast, transcripts of CKL2, 3 and 15 together with WEE1 decline during S phase but show a peak in S/G2 transition. Their trend decrease through mitosis and are up-regulated when cells move toward G1 phase (Fig. 6H). As cells move from G2 into M phase, a major shift in the expression of CKS1 and CKS2 occurs. So that an increase in the expression followed by its decreasing through G1 phase is observed. In spite of slight differences, both genes are appeared as mitosis-specific genes (Fig. 6I).

Transcription factors show multiple patterns of expression

In nodeA/cluster7, there are 4 TFs (AT1G63100, AT3G16280, AT4G11080 and AT5G11510) which group with mitotic peaking genes including CDKBs and mitotic cyclins (Fig. 4 nodeA, Fig. 5 cluster7). Searching for the central core of the MSA element (AACGG) in the promoter of genes classified in cluster7 indicate presence of MSA site in most of them. Due to having the similar expression profile of mitotic genes and presence of MSA element in their promoters, potential regulation of mentioned TFs in M phase is proposed. Expression of SCARECROW (AT1G63100) was increased from S phase toward mitosis cells and continued to rise through G1 phase. This gene is a member of the GRAS family of TFs which affect cell division rate both in roots and leaves by stimulating S-phase progression of the cell cycle.^29,30 TINY (AT3G16280) is a member of AP2 family and its over-expression in Arabidopsis leads to plants of reduced organ size. This effect is a consequence of reduction in cell elongation.³¹ Some TFs were known to play a role in cell cycle regulation. For instance, it was reported that CYCD3 and RB genes are upregulated in Arabidopsis plants over-expressing BOL (AP2 family).³² AT4G11080 classified with mitotic peaking genes including CDKBs and mitotic cyclins and was predicted to be plastid-targeted that could provide a link between nuclear and plastid division²¹ sequentially. AtMYB3R4 (AT5G11510) is most closely related to the tobacco genes NtmybA2 and NtmybA1. The later genes mediate transcription of G2/M phase-specific genes through binding to the MSA element in their promoters.^33,34 Interestingly, there are 3 copies of the MSA elements in the promoter of AtMYB3R4 which may be responsible for expression of this group of MSA-dependent genes in Arabidopsis.²⁶ AtMYB3R4 shows close homology to the mitotic spindle-associated kinases³⁵ and may function in metaphase/anaphase control Auxin is essential for cell division and cell cycle progression. Expression of CYCD is subjected to transcriptional regulation by growth promoting-factors such as auxin.³⁶ In presented data, the auxin-induced transcriptional regulators IAA12 (AT1G04550), IAA13 (AT2G33310), IAA2 (AT3G23030) and IAA30 (AT3G62100) are grouped in clusters4, 5 and 9 (Fig. 5) together with D type cyclins. They are showing similarity in expression pattern during cell cycle progression. This group of genes oscillates with expression peaks in early G2 and late M phases (Fig. 6J). Interestingly, comparison all TFs expression patterns with the CYCA3;1 and CYCA3;2 expression profiles identified 14 TFs with common trends in expression and strong likelihood of role in S phase (Fig. 5 cluster1; Fig. 6K). Many of these TFs carry E2F site in their promoters and widely proposed as conferring S phase regulation. E2F genes are counted as S phase regulated genes³⁷ and show also similar expression pattern with cluster1 genes (Fig. 5 cluster1). The data set includes TFs from different families follow distinct patterns during cell cycle progression. Node F includes several TFs belonging to NAC, WRKY, zinc-finger and MYB families whose expression is up-regulated at M and G1 phases (Fig. 5 cluster1) and node E includes genes encoding AP2 and MADS-box proteins, express during S and G2 phases (Fig. 4 nodeE). Totally, a close match was found for known regulated genes between the previously reported microarray data^38,39 and the qRT-PCR data. The results proved efficient synchronization caused by Aphidicolin. The outcome of this part of my study which was repeated 3 times was production of highly synchronized MM2d cells necessary for doing TFs expression profiling.

Transcription factors expression profiling

A platform for Arabidopsis TFs was primarily designed by⁴⁰ which its current version covers 1880 TF genes. This platform was used to monitor changes in expression level of TFs during cell cycle progression in synchronous MM2d cells by qRT-PCR technique. Gene expression was monitored during a 19 h period representing different time points which are corresponding to specific cell cycle phases. The whole experiment was repeated 3 times in the form of independent biological replications. For each experiment, samples were taken from synchronous suspension cultures having been evaluated by 3 separate procedures including flow-cytometry analysis, M/AI and expression of various cell cycle marker genes. Using available TF platform, the expression level of 1880 TFs was scored with high confidence. As explained above the comparative CT method was used for calculating fluctuations in genes expression (equations 1 and 2). For the first biological replication genes which showed |Log₂FC|≥2 were selected as differentially expressed genes during cell cycle progression. After measuring the expression level of the whole set of selected genes for all time points in the second and third replications, a gene was assigned as being cell cycle-regulated if the corresponding Log₂FC appeared with the same sign in all 3 biological replications. In this case the FC value was represented as average values on 3 independent experiments. It has to be notice that finally only the genes with |Log₂FC|≥3 value were kept and the others omitted.

Constitutively expressed TF genes during cell cycle progression

Analysis of the expression profile revealed a total of 239 genes corresponding to 13% of the genes presented in the Arabidopsis TF platform. This might indicate that a certain part of TFs were either not expressed during cell cycle or expressed below the detection limit of qRT-PCR system. Sixteen percent of 239 genes (38 genes) were found to be upregulated and 59% (141 genes) to be down-regulated. The remaining TFs were detected in one or 2 phases, among these 6% were induced and 19% repressed. Of the 239 TFs, 87 (∼36%) genes expressed constitutively through all time points from which 12 genes showed up-regulated expression and the others were downregulated. Since these 87 genes were either induced or repressed during cell cycle and kept the expression status through all phases, they were annotated as constitutively expressed TFs (Table 1). The expression patterns of some constitutively expressed TFs during cell cycle progression were shown in Figure 7.

Figure 7.

The expression patterns of some constitutively expressed TFs during cell cycle progression.

Phase-specific TF genes

On the other hand, identified genes were categorized in groups based on the presence in each phases including 168 genes in S phase (corresponding to 0–4 h), 149 genes in G2 phase (6–8 h), 157 genes in M phase (10–14 h) and 153 genes in G1 phase (16–19 h). According to this classification, a total of 152 genes were detected in one or two phases that annotated as specific cell cycle-associated TFs. Of the 152 genes about 21% (32) were found differentially expressed in S phase and called S phase-specific genes (Table 2), most of those appeared to be down-regulated. From five G2 phase-specific genes (Table 3), expression of four genes was detected positively in comparison with the expression level at time zero. Expression of 13 genes of the 152 genes was exclusively obtained in M phase that majority (8 genes) were downregulated (Table 4). Up to 18 genes of the 152 genes were identified to be expressed only in G1 phase including 13 up and 5 down-regulated genes (Table 5). TFs link signals to the downstream genes that execute different cellular processes. Of the 30,000 genes encoded by the Arabidopsis genome, around 6% are TF encoding genes⁴¹ where approximately a fraction of 13% were identified in my research to be cell cycle-related genes. TFs involved in cell cycle regulation belonged to AP2/EREBP, bZIP, C2H2, C3H, MYB, bHLH, WRKY, NAC, MADS-box, LOB, DOF, HB, GATA, HSF, YABBY, ABI3/VP1 and other families. The most overrepresented TFs were members of the AP2/EREBP, C2H2, MYB, WRKY and bHLH families.

Table 2.

S phase-specific genes with up- and down regulation during cell cycle progression. The C_T values for all genes were normalized based on the C_T value of UBQ10 gene and were log₂ transformed relative to time zero as control. The mean FC values for the 3 independent biological replications were represented here

Log2FC			Log2FC
Mean S	Gene Family	Gene Name	Mean S	Gene Family	Gene Name
2.5	GATA	AT3G16870	−3.2	ABI3VP1	AT4G01580
2.6	GATA	AT4G32890	−2.6	AP2-EREBP	AT1G12630
−2.2	HB	AT5G59340	−2.0	AP2-EREBP	AT5G65130
−2.2	LOB	AT1G67100	3.1	AP2-EREBP	AT1G12980
2.6	MADS	AT1G71692	−2.8	bHLH	AT2G40200
−3.1	MYB	AT1G68320	−2.3	bHLH	AT1G02340
−3.1	MYB	AT3G06490	−2.1	bHLH	AT2G22770
−2.8	MYB	AT5G16600	2.8	bHLH	AT3G47640
−2.4	MYB	AT1G74430	−3.3	bZIP	AT5G60830
−2.0	MYB	AT5G04760	−2.5	bZIP	AT1G75390
−2.5	NAC	AT3G15170	−4.2	C2H2	AT2G01940
−2.5	NAC	AT1G79580	−3.2	C2H2	AT4G02670
−2.5	NAC	AT1G71930	−2.9	C2H2	AT1G66140
2.0	NAC	AT3G15500	−3.3	CO-like	AT1G25440
−3.1	RWP-RK	AT1G18790	−2.8	DOF	AT1G26790
−2.6	WRKY	AT3G62340	2.4	DOF	AT2G46590

Table 3.

G2 phase-specific genes with up- and down regulation during cell cycle progression. The C_T values for all genes were normalized based on the C_T value of UBQ10 gene and were log₂ transformed relative to time zero as control. The mean FC values for the 3 independent biological replications were represented here

		Log2FC
Gene Name	Gene Family	Mean G2
AT4G31660	ABI3VP1	2.9
AT5G56960	bHLH	−3.7
AT1G68810	bHLH	2.6
AT3G29340	C2H2	2.7
AT1G14580	C2H2	2.1

Table 4.

M phase-specific genes with up- and down regulation during cell cycle progression. The C_T values for all genes were normalized based on the C_T value of UBQ10 gene and were log₂ transformed relative to time zero as control. The mean FC values for the 3 independent biological replications were represented here

		Log2FC
Gene Name	Gene Family	Mean M
AT1G77640	AP2-EREBP	−2.6
AT5G25890	AP2-EREBP	−2.9
AT1G34410	AP2-EREBP	−2.5
AT4G25470	AP2-EREBP	−2.8
AT1G29600	C3H	2.8
AT1G73870	CO-like	2.6
AT1G69570	DOF	2.3
AT3G60460	MYB	−2.3
AT5G62380	NAC	2.4
AT1G30135	Tify	−2.0
AT1G69180	YABBY	−2.4

Table 5.

G1 phase-specific genes with up- and down regulation during cell cycle progression. The C_T values for all genes were normalized based on the C_T value of UBQ10 gene and were log₂ transformed relative to time zero as control. The mean FC values for the 3 independent biological replications were represented here

Log2FC			Log2FC
Mean G1	Gene Family	Gene Name	Mean G1	Gene Family	Gene Name
-3.1	CO-like	AT1G68520	2.4	AP2 EREBP	AT1G16060
2.8	HB	AT1G62990	2.1	AP2-EREBP	AT1G12890
2.7	LOB	AT2G45420	2.1	AP2-EREBP	AT1G24590
2.4	MYB	AT2G39880	3.2	ARP	AT1G80390
3.2	MYB	AT1G73410	2.6	ARP	AT2G46530
3.2	MYB	AT5G12870	2.2	ARR B	AT5G49240
-2.6	MYB	AT2G47460	3.2	bHLH	AT5G15160
-2.0	RWP-RK	AT5G66990	3.3	C2H2	AT2G28200
-2.0	WRKY	AT2G40750	−2.9	C2H2	AT1G75710

Discussion

In the present study utilizing 3 ways for confirming synchronization, achieving a high level of synchrony was confirmed. It is the first time which a combinatorial method comprising genes expression patterns, mitotic index and flow-cytometry were used to judge about cell cycle synchronization. The data indicated that instead of combinatorial method, cell cycle-related gene expression patterns may be used as a method of choice for synchronization confirmation. Although global gene expression analysis using Affymetrix gene chips revealed presence of many TFs, but a small fraction of them has been studied with respect to cell cycle progression. Quantitative RT-PCR to screen the whole set of Arabidopsis TFs for monitoring their expression patterns during cell cycle progression was used. Then the results of a wide scale analysis of TF genes expression profiling in Arabidopsis synchronized cell culture were presented here. For this aim, I conducted time-course qRT-PCR experiment and subsequently the TFs were clustered based on expression patterns so that co-expressed genes were grouped into the same cluster. Four large groups correspond to the G1, S, G2 and M phases consist of genes that are likely to show cell cycle-dependent regulation of their expression. Based on literatures some of these TFs have been already characterized but the majority was identified in this study as being cell cycle-associated genes. In addition to the genes expressed constitutively during cell cycle progression, there were genes detected to be expressed specifically in one phase. It is interesting that the number of TFs with S phase-specificity was around 13 percentages (32 TFs) of whole TFs were detected here. These TFs might involve in DNA replication and modification and in chromatin assembly which is tightly control demanding. Expression of 5 TF genes was detected only in G2 phase. These genes probably involve in determining if the cell can proceed to enter mitosis and divide. Up to 18 TFs specifically expressed in G1 phase. Transition from G1 to S phase involves activation of the genes encoding enzymes used in nucleotide metabolism and DNA synthesis as well as histone genes encoding proteins required for the package of newly replicated DNA. Expression of 13 TFs was detected in M-phase. These genes may have role in functional regulatory mechanisms in mitosis such as RNA processing and specific chromatin decondensation.⁴² The majority of the M-specific TFs appeared as downregulated, which may be a result of either degradation in their related mRNA or lack of transcription in Mitosis. Decline in expression of these TFs may be due to having repressing effects on entering the cell into the M phase or because of controlling the genes having catalytic activities against cell division instruments. Although gene expression that regulates cell cycle progression forms only part of the control networking, transcriptional regulation plays a major role in the cell cycle control which modulated by TFs via interaction with their corresponding binding sites on the DNA sequences.^26,43 The transcriptional regulation of core cell cycle genes which is a key element in the cell cycle control is mediated and regulated by TFs. Prominent example is MYB proteins that bind to mitosis specific elements (MSA) in promoters of mitotic cyclins and confer specific regulation in G2/M.³⁴ For example, AT3G60460 (DUO1, AtMYB125) which in this study was as M phase-specific gene controls male gamete formation in flowering plants and integrates cell specification with cell cycle progression. It promotes generative cell division through activating G2/M regulators such as cyclin genes during pollen mitosis II. This supports the view that cell cycle regulation and specific developmental processes are tightly coupled and regulated by R2R3 MYB genes in plants.^44,45 Another example is AT1G24590 (also named DRNL, SOB2, ESR2 and BOLITA) an AP2, having been identified as up regulated gene in G1 phase in present study. This gene normally has role in suppressing cell division and cell expansion in part due to perturbations of cell cycle activators and inhibitors such as histone H4, RBR1, CYCD3;1, CYCD1;1, OBP1, KRP4 and TCP genes.^32,46,47 Simultaneous activation of both cell cycle activators and inhibitors is characteristic for TFs involved in the control of cell proliferation and is believed to provide a feedback mechanism for the tight control of the cell division.²⁵ In conclusion, I was able to identify TFs linked to cell cycle regulation in Arabidopsis. Most of them appeared to exhibit variation in expression level throughout the cell cycle. Detailed characterization of the genes of unknown function will help further unravel regulatory networks underlying the cell cycle and to understand the specific role of these genes in regulating cell division during development. The function of TFs and their integration into regulatory networks can only be fully exploited if the target genes of the TFs are known. Growth and hence progression through the cell cycle is regulated by many physiological and developmental parameters and is also known to be influenced by environmental factors, including abiotic stress. An important aspect will be to understand the expression pattern of the cell cycle-regulated TF genes in more detail. It is necessary to study the expression behavior of the genes after a number of treatments known to trigger or slow down growth processes such as high and low light conditions, dark-light transitions, varying nitrogen supplies and hormone sorts. It will allow unraveling the integration of these TFs into the regulatory networks that link growth-affecting stimuli (e.g. nutrient supply or hormone treatment) to cell cycle progression. Cell cycle can be manipulated by affecting on duration of different cell cycle stages or by alteration of the cell cycle through switching from mitotic cycles to endoreduplication cycles and increasing in cell ploidy level. So, cell cycle genes are interesting targets for the modification of plant growth, architecture and yield.⁴⁸ Because of plants responding to fluctuations in their environment by changes in cell division rates in meristems, alteration of normal cell division allows the control of aspects of plant environment responses. Cell proliferation closely linked to the cell cycle. Phytohormones such as auxin control organ cell proliferation via altering cell numbers as well as determination of organ size.^49-51 Auxin is required for the G1/S phase transition in cell cycle and initiation of DNA replication. Auxin together with cytokinin activates cell cycle regulatory genes which promote cell division, expansion and differentiation in root and shoot apical meristems^52,42;(Bhalerao and Bennett, 2003). Auxin-binding proteins such as ARF transcription factors bind auxin responsive promoter elements and may either activate or repress transcription. Among the known ARF proteins, ARF2 has been characterized to be a general repressor of cell division in many aerial organs of the plant. ARF2 prolongs expression of ANT and CYCD3;1 in mature stems and leaves.⁵³ It has been reported that ARF7 and ARF19 are involved in the control of cell division patterns during lateral root development.⁵⁴ From this group of genes, changes in expression of ARF21 (AT1G34410) in M phase were detected here. This gene gives a chance to be involved in cell division which has to be investigated. Expression of one AP2 gene CRF (cytokinin response factor), CRF6 (AT3G61630), expressed constitutively throughout the cell cycle was detected (Table 1). CRFs (CRF1–CRF6) have been previously examined and shown to be regulated by cytokinin and affect cell expansion in leaf and cotyledon.⁵⁵ From DREB subfamily AT4G25470 (CBF2, DREB1C) appeared to be down-regulated in M phase. This gene functions in developmental processes as extends life span of plants and thus delays onset of leaf senescence by 2 weeks.⁵⁶ Expression of AT5G25890 (IAA28) was detected in M phase (Table 4). IAA28 protein is transcriptional repressor of auxin-responsive gene expression and inhibits initial cell divisions associated with lateral root initiation.^57,54 In Arabidopsis C2H2-type zinc finger genes are involved in a wide range of functions.^58-60 In present study 2 members of C2H2 family; AT3G19580 (AZF2) and AT1G27730 (ZAT10, STZ) were downregulated during cell cycle (Table 1). It has been demonstrated that JA application arrests cell cycle in G2 by repressing expression of a large group of M phase associated genes. Transcriptional regulators of JA biosynthesis included both activators and repressors. The JA repressors were encoded by ZAT10 and AZF2 proteins.⁶¹ Interestingly, in my data the expression trend of AZF2 and ZAT10 increased in 2 points in G2 phase and G2/M transition. The expression of AT3G06490 (AtMYB108, BOS1) was appeared only in S phase (Table 2) ⁶² reported that BOS1 is a JA-inducible TF which regulates final stages of stamen development and male fertility. Expression of AT2G47190 (AtMYB2) gene was monitored constitutively during overall phases of cell cycle (Table 1). Overexpression of AtMYB2 lead to reduce cell size and dwarfed phenotype in Arabidopsis.⁶³ In this experiment, expression of 5 genes from bZIP family was detected through cell cycle progression. This group of TF genes is supposed to be linked with core cell cycle genes. Most histone genes are cell cycle-dependently expressed during the S phase. The promoter regions of these genes contain cis-acting elements to bind proteins such as bZIPs and confer S phase-specific expression of histone genes.⁶⁴

Table 1.

TF genes constitutively up- or down-regulated during cell cycle progression. The C_T values for all genes were normalized based on the C_T value of UBQ10 gene and were log₂ transformed relative to time zero as control. The mean FC values for the 3 independent biological replications were represented here

		Log2FC*
Gene Name	Gene Family	Mean S	Mean G2	Mean M	Mean G1
AT1G68840	ABI3VP1	−3.2	−3.5	−3.0	−3.0
AT1G25560	ABI3VP1	−2.5	−2.7	−3.2	−3.2
AT1G71520	AP2-EREBP	−7.5	−8.5	−8.1	−7.3
AT1G22810	AP2-EREBP	−5.8	−5.5	−5.9	−5.8
AT2G20880	AP2-EREBP	−5.7	−6.1	−6.3	−6.3
AT4G28140	AP2-EREBP	−5.2	−5.0	−5.5	−4.4
AT1G12610	AP2-EREBP	−5.1	−4.0	−5.9	−4.0
AT1G64380	AP2-EREBP	−4.7	−4.1	−4.6	−4.4
AT4G34410	AP2-EREBP	−4.7	−4.0	−3.7	−4.3
AT1G80580	AP2-EREBP	−4.2	−3.5	−4.3	−3.8
AT5G47230	AP2-EREBP	−3.7	−3.4	−2.7	−2.5
AT2G33710	AP2-EREBP	−3.7	−3.3	−3.4	−3.3
AT5G67000	AP2-EREBP	−3.6	−3.0	−3.7	−3.9
AT4G25490	AP2-EREBP	−3.2	−2.8	−3.7	−3.6
AT3G61630	AP2-EREBP	−3.0	−2.8	−2.8	−2.7
AT3G50260	AP2-EREBP	−3.0	−2.8	−2.6	−3.5
AT5G52020	AP2-EREBP	−2.3	−2.4	−2.7	−3.0
AT3G16280	AP2-EREBP	−2.2	−2.1	2.0	2.9
AT3G60490	AP2-EREBP	3.3	3.0	2.9	3.5
AT4G32280	ARP	2.3	3.0	2.5	3.1
AT4G38910	BBR/BPC	−4.8	−3.9	−2.3	−2.3
AT2G31210	bHLH	−4.1	−3.4	−4.1	−3.2
AT4G20970	bHLH	−3.9	−3.6	−3.0	−2.9
AT3G59060	bHLH	−2.5	−2.7	−2.4	−2.1
AT3G56980	bHLH	5.8	5.4	−2.7	−2.4
AT3G49760	bZIP	−4.1	−3.1	−3.6	−3.9
AT5G07160	bZIP	−2.9	−3.8	−5.8	−6.5
AT1G06850	bZIP	2.5	3.3	2.8	3.0
AT5G67450	C2H2	−6.4	−6.9	−5.8	−5.4
AT2G37430	C2H2	−5.5	−6.6	−8.7	−7.8
AT2G28710	C2H2	−5.3	−6.5	−6.2	−6.3
AT3G19580	C2H2	−4.6	−4.3	−4.4	−4.1
AT3G53600	C2H2	−4.4	−5.4	−6.5	−6.6
AT1G49900	C2H2	−4.3	−4.6	−4.6	−4.9
AT3G46090	C2H2	−3.9	−3.8	−4.3	−4.4
AT5G04340	C2H2	−3.1	−3.0	−2.6	−2.5
AT1G27730	C2H2	−3.0	−3.9	−3.9	−2.9
AT3G46070	C2H2	−2.7	−2.6	−3.3	−3.7
AT5G60470	C2H2	−2.4	−2.6	−2.3	−2.6
AT3G55980	C3H	−3.9	−3.4	−3.4	−3.1
AT2G19810	C3H	−3.3	−3.6	−2.5	−3.0
AT2G21810	CHP-rich	−2.2	−3.3	−2.8	−2.7
AT1G68190	CO-like	−3.3	−2.9	−3.5	−3.9
AT4G37650	GRAS	3.2	3.3	2.9	3.1
AT5G46010	HB	−3.5	−3.3	−3.3	−3.5
AT3G50890	HB	−3.5	−2.8	−4.2	−3.2
AT2G19510	LOB	−3.1	−4.6	−4.2	−5.0
AT4G00210	LOB	−3.0	−2.6	−2.9	−2.5
AT4G22700	LOB	−2.6	−2.2	−3.4	−2.7
AT2G42430	LOB	2.2	3.2	3.9	4.8
AT4G37540	LOB	5.0	4.0	4.8	4.8
AT1G59810	MADS	−4.8	−4.5	−3.3	−2.8
AT1G60040	MADS	−4.6	−4.4	−3.4	−3.0
AT2G21650	MEE3	−3.5	−6.2	−4.4	−4.2
AT2G47190	MYB	−6.4	−7.3	−6.1	−5.6
AT3G23250	MYB	−3.3	−5.5	−4.7	−4.0
AT1G13300	MYB	−2.7	−3.2	−2.4	−2.5
AT5G37260	MYB	−2.7	−2.2	−2.8	−2.4
AT1G22640	MYB	2.4	2.4	2.9	2.6
AT1G74080	MYB	2.9	3.1	3.4	3.4
AT5G61620	MYB-like	−4.3	−4.6	−5.5	−6.0
AT1G25550	MYB-like	−3.1	−2.7	−2.6	−2.6
AT5G45580	MYB-like	2.8	3.1	2.3	2.4
AT3G44350	NAC	−4.5	−3.7	−3.6	−3.1
AT1G77450	NAC	−3.9	−3.7	−3.8	−3.8
AT5G41090	NAC	−3.4	−4.3	−5.0	−5.8
AT5G22380	NAC	−3.3	−3.2	−3.1	−3.4
AT4G38340	RWP-RK	−3.2	−2.6	−2.9	−2.5
AT5G51910	TCP	2.5	2.4	2.7	3.0
AT4G11070	WRKY	−7.5	−7.9	−7.5	−7.2
AT4G18170	WRKY	−5.4	−5.2	−4.6	−4.4
AT5G24110	WRKY	−5.4	−9.1	−8.4	−7.7
AT4G23810	WRKY	−5.1	−6.2	−5.7	−4.6
AT2G38470	WRKY	−4.9	−5.7	−4.9	−4.3
AT4G23550	WRKY	−4.8	−5.3	−4.8	−5.1
AT1G66560	WRKY	−4.5	−3.0	−2.0	−2.4
AT2G40740	WRKY	−4.2	−4.4	−5.0	−6.0
AT4G01250	WRKY	−3.8	−5.6	−5.7	−5.5
AT5G15130	WRKY	−3.6	−3.9	−4.7	−3.5
AT4G31550	WRKY	−3.3	−3.5	−3.5	−3.1
AT1G55600	WRKY	−2.6	−2.9	−3.3	−3.9
AT4G39070	Zinc finger	−3.1	−4.8	−4.2	−4.3

^*FC stands for fold change which is defined as 2 ^−ΔΔCT and is represented as average values on 3 independent experiments.

Materials and Methods

Arabidopsis cell suspension culture

A suspension culture of the fast growing cell line MM2d, originally derived from Landsberg erecta stem explants⁶⁵ was obtained from Prof. James Murray (university of Cambridge, UK). Cultures were maintained by weekly sub-culturing of 3.5 ml saturated culture into 100 ml fresh MSS media containing Murashige and Skoog salt supplemented with 3% w/v sucrose, 0.5 mg/l NAA and 0.05 mg/l kinetin, with pH adjusted to 5.8 using 1N KOH. Cells were grown at 26°C in continuous darkness while rotated at 130 rpm in 300 ml narrow-necked Erlenmeyer flasks covered with aluminum foil. For growth analysis, samples were taken in the first day after sub-culturing at 24h intervals until 8th day of sub-culturing. Cells were microscopically observed and counted using a counting chamber (Neubauer, Germany) and Olympus BX51 microscope (Olympus, Tokyo, Japan).

Synchronization induction by Aphidicolin

MM2d cells were reversibly blocked in late G1/early S phase with Aphidicolin according to.²¹ Briefly, Aphidicolin was dissolved in dimethyl sulphoxide (DMSO) and added to 100 ml fresh MSS medium containing 20 ml aliquot of cell culture in the early stationary phase (5 d after previous sub-culturing) to give final concentration of 4 μg/ml Aphidicolin. MM2d cultures were incubated for 24 h under conditions described above and then cells were gently washed with 2 L of MSS medium through a nylon net (mesh size 47 μm) to remove the drug. Cells were re-suspended in a total volume of 120 ml of fresh MSS medium, incubated under the mentioned cultivation condition, and samples were taken hourly for validation procedures. A control set of cultures was treated with DMSO (1 μg/ml) and handled in the same way as Aphidicolin-treated cells.

QRT-PCR analysis

For gene expression analysis, total RNA was isolated from Aphidicolin-treated cells using TRIZOL kit (Invitrogen) according to the manufacturer´s instructions. After treating the extracted RNA with RNase-free DNaseI (Sigma), absence of the residual genomic DNA was checked by qRT-PCR, make uniform using primers spanning an intron of the LATE ELONGATED HYPOCOTYL gene. Purity and concentration of RNA were evaluated by spectrophotometer and its integrity was checked on a 1% (w/v) formaldehyde agarose gel before and after DNaseI treatment. RNA (2 μg) was reverse transcribed with Superscript III reverse transcriptase kit (Invitrogen) in a reaction volume of 20 μl to generate first-strand cDNA. Quantity of the synthesized cDNA was evaluated using qRT-PCR amplification of reference gene (UBQ10, AT4G05320) and its quality was assessed with primer pairs designed for 3’ and 5’ ends of GAPDH gene (AT1G13440). Quantitative RT-PCR was performed in an ABI PRISM 7900 HT sequence detection system (Applied Biosystems) adopting 394 well-plates. Reactions in a final volume of 5 μl contained 0.5 μl of template (cDNA or RNA), 2 μl of mixed 0.5 mM forward and reverse primers and 2.5 μl of SYBR green master mix (Applied Biosystems). The standard thermal profile was used for all reactions as: 50°C for 2 min, 95°C for 10 min; 40 cycles of 95°C for 15s and 60°C for 1 min followed by generation of dissociation curve. Data were analyzed using the SDS2.2.1 software (Applied Biosystems). C_T values for all genes (C_Tg) were normalized based on the C_T value of the UBQ10 gene (ΔC_T = C_Tg –C_TUBQ10). For comparative expression analysis relative to the expression in time zero (T0; immediately after removal of Aphidicolin used as calibrator), the normalized expression data were divided by the normalized expression data in time zero. For doing that 2^−ΔΔCt method was used as fold change in gene expression where ΔΔC_{T g =} ΔC_{T g}−ΔC_{Tg, T0} which C_T refers to the number of cycles at which SYBR Green fluorescence in a PCR reaches an arbitrary value during the exponential phase of DNA amplification and was set at 0.2 in all experiments.

Key Message:

I’d like to draw your attention to innovation aspect of this work because of incorporation of 3 separate procedures for checking cell synchronization used in the present study for the first time.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

I appreciate all staffs in Max Planck Institute (Golm/Germany) for their help during this study.

Funding

This work was funded by the Ministry of Science, Research and Technology of Iran.