Role for E2F in Control of Both DNA Replication and Mitotic Functions as Revealed from DNA Microarray Analysis

Abstract

We have used high-density DNA microarrays to provide an analysis of gene regulation during the mammalian cell cycle and the role of E2F in this process. Cell cycle analysis was facilitated by a combined examination of gene control in serum-stimulated fibroblasts and cells synchronized at G₁/S by hydroxyurea block that were then released to proceed through the cell cycle. The latter approach (G₁/S synchronization) is critical for rigorously maintaining cell synchrony for unambiguous analysis of gene regulation in later stages of the cell cycle. Analysis of these samples identified seven distinct clusters of genes that exhibit unique patterns of expression. Genes tend to cluster within these groups based on common function and the time during the cell cycle that the activity is required. Placed in this context, the analysis of genes induced by E2F proteins identified genes or expressed sequence tags not previously described as regulated by E2F proteins; surprisingly, many of these encode proteins known to function during mitosis. A comparison of the E2F-induced genes with the patterns of cell growth-regulated gene expression revealed that virtually all of the E2F-induced genes are found in only two of the cell cycle clusters; one group was regulated at G₁/S, and the second group, which included the mitotic activities, was regulated at G₂. The activation of the G₂ genes suggests a broader role for E2F in the control of both DNA replication and mitotic activities.

Rapid progress has been made in the understanding of regulatory pathways that govern the transition of cells from a quiescent state into a cell cycle. Such studies have highlighted the critical role of the signaling pathway that involves the accumulation of D cyclin/cdk4 activity leading to the phosphorylation of the retinoblastoma protein, which then allows an accumulation of E2F transcription activity (21, 24). A variety of experiments have demonstrated the role of E2F proteins in the control of expression of genes important for DNA replication as well as further cell cycle progression (5, 18). In particular, E2F activity is responsible for the activation of genes encoding DNA replication proteins, enzymes responsible for deoxynucleotide biosynthesis, proteins that assemble to form functional origin complexes, and kinases that are involved in the activation of initiation.

Although much has been learned from these studies of E2F transcription control, important questions remain. For one, the scope of the gene-regulatory control by E2F proteins has not been addressed. In large part, the identification of target genes has followed from the initial studies of the DNA tumor virus oncoproteins, such as adenovirus E1A and simian virus 40 T antigen; previous work demonstrated that these proteins were capable of inducing quiescent cells to enter S phase, and associated with this induction was an activation of various genes encoding DNA replication activities (17). This activity coincides with an ability to inactivate the Rb tumor suppressor protein and thus allow an accumulation of E2F proteins. Analysis of promoters for genes such as DNA polymerase α, thymidine kinase, and others revealed the presence of E2F binding sites that were shown to be critical for the normal control of expression of these genes. As additional DNA replication genes have been identified, including those encoding proteins that recognize and establish a functional origin of replication, the majority have been shown to be targets for E2F control. As such, it now appears that a primary role of the G₁ cdk/Rb/E2F pathway is the control of genes that allow cells to enter S phase and begin DNA replication.

Despite these advances, the study of E2F gene control has been incremental, following from preconceived views of the role of the Rb/E2F pathway in cell proliferation. As one approach to better understanding the full extent of gene expression under the control of the Rb/E2F pathway, not influenced by the bias of previous work, we have analyzed the expression of a large number of genes using high-density DNA microarrays. The strength of this approach lies in the ability to assay a very large number of potential targets in an unbiased manner—no presumptions are made about the nature of the pathway(s) that might be affected or regulated by E2F activities. For these experiments, we have made use of Affymetrix GeneChip DNA microarrays that contain murine gene sequences and expressed sequence tags (ESTs) and then assayed the profile of gene expression following expression of E2F proteins in quiescent cells.

At the same time, and to serve as a basis for comparison with the E2F-induced genes, we have also profiled the pattern of gene expression changes that occur as cells are initially stimulated to proliferate as well as when cells cycle in the presence of growth factors. We find that many of the E2F-induced genes are normally regulated at G₁/S of the cell cycle, consistent with previous studies. Strikingly, however, we also find that a substantial number of the E2F-induced genes are normally regulated at G₂ of the cell cycle, suggesting a role for E2F activity in initiating a cascade of gene control during the cell cycle.

MATERIALS AND METHODS

Cells and viruses.

The mouse embryo fibroblast (MEF) cell line 2r15 was established from a wild-type 13.5-day embryo essentially as described (20). MEFs were grown in Dulbecco's modified Eagle's medium (DMEM) containing 15% fetal bovine serum (FBS). To bring cells to quiescence for the serum stimulation experiment, nearly confluent cells were split 1:5 and incubated overnight in DMEM containing 15% FBS. The medium was replaced with DMEM containing 0.2% FBS, and the cells were cultured for 30 h. These quiescent cells were stimulated by adding FBS at the final concentration of 15%. To bring cells to quiescence for the hydroxyurea (HU) experiment, almost-confluent cells were split 1:2 and incubated for 48 to 60 h in DMEM containing 15% FBS. Cells became quiescent due to contact inhibition during this period. These quiescent cells were released to grow by splitting 1:5 in DMEM containing 15% FBS. Three hours after splitting, HU was added to the medium at a final concentration of 0.5 mM, and cells were incubated for a further 18 h. Cells were washed twice with DMEM and refed with DMEM containing 15% FBS to release them from HU block. Cell synchrony in both experiments was assessed by flow cytometry (22).

The methods for preparation and determining the titer of viruses have been described (19). For infection with recombinant adenoviruses, 2r15 cells were brought to quiescence by serum starvation, and virus infection was carried out as described (13). Following infection, cells were cultured in DMEM containing 0.2% FBS for 18 h before harvesting for further treatment. Recombinant adenoviruses expressing E2F1 or E2F2 were titrated to identify multiplicities of infection that would achieve an equivalent level of production of the DNA-binding activities. For one experiment, multiplicities of 600 for E2F1 and 250 for E2F2 were used; a second experiment employed a multiplicity of 600 for E2F1 and 400 for E2F2.

Nuclear extract preparation and E2F DNA-binding assays.

Nuclear extracts from 2r15 cells were prepared as described (13). E2F DNA-binding assays were performed as described (9) using dihydrofolate reductase promoter DNA fragment as the probe.

RNA preparation.

Total RNA was prepared by treating cells with Trizol reagent (Gibco). mRNA was selected from total RNA with the polyATract mRNA isolation system (Promega) according to the manufacturer's instructions.

Northern analysis.

Northern analysis was performed as described (13).

DNA microarray analysis.

The targets for Affymetrix DNA microarray analysis were prepared according to the manufacturer's instructions. Either the Mu6500 or the Mu11K DNA Affymetrix GeneChip microarray was hybridized with the targets at 45°C for 16 h and then washed and stained using the GeneChip Fluidies station according to the manufacturer's instructions. DNA chips were scanned with the GeneChip scanner, and signals obtained by scanning were processed by the GeneChip expression analysis algorithm (version 3.2) (Affymetrix).

Cluster analysis of cell cycle expression patterns.

The data acquired through absolute analysis of the Affymetrix GeneChip expression analysis algorithm (version 3.2) was imported to the GeneSpring analysis program (Silicon Genetics). The average difference value of each gene at each time point during the serum stimulation experiment, as well as the HU release experiment, was used. If the average difference value at a given time point was below the raw Q value, that number was replaced with the raw Q value. If the average difference value of a given gene was below the raw Q value at all time points, that gene was excluded from the clustering. The genes that showed substantial induction after serum stimulation were selected based on the following criteria: the maximum of the average difference value after serum stimulation should be 2.5-fold greater than the average difference value of quiescent state, and the difference of the maximum of the average difference value after serum stimulation and the average difference value of quiescent state should be greater than or equal to 50. A total of 578 of approximately 6,200 clones met both conditions. The expression pattern of each gene was normalized across the experiments by dividing the average difference value at each time point by the median of every average difference value through the serum stimulation and HU experiments with the same gene. Those genes were initially ordered hierarchically by applying the tree-making program (GeneSpring; Silicon Genetics) to the normalized expression patterns. The genes were then clustered into 16 sets by applying the k-mean clustering algorithm (GeneSpring; Silicon Genetics). The average at each time point of each set was calculated to generate the template patterns for the further clustering. Clones that showed an expression pattern similar to these 16 template patterns were then selected among the 578 genes described above. The similarity of the expression pattern to the template pattern was evaluated by calculating the standard correlation coefficient. Genes with a coefficient greater than or equal to 0.88 of the standard correlation coefficient were selected and clustered. If a given clone showed similarity to several template patterns, the pattern that gave the highest standard correlation coefficient was selected for that gene. To select the “growth” gene, the ratio of the standard deviation of the average difference values of HU experiment and the average of those values were calculated. Genes that had a ratio of less than or equal to 0.185 were selected. After this clustering and selection, clusters were grouped by eye. Finally, the expression pattern of each gene was examined, and a few outlying genes were excluded. The G₀ group was identified separately by applying a similar clustering approach but focusing on the genes expressed at a higher level in quiescent state than after serum stimulation.

Selection of E2F target genes.

An analysis of the data from the E2F expression samples and the control sample was performed using the comparison analysis of the Affymetrix GeneChip expression analysis algorithm (version 3.2). Genes that fit the following criteria were considered induced genes in a given experiment: the change call was either induced or marginally induced; the induction was greater than or equal to twofold: and the average difference value of E2F-expressing sample was at least 50. In order to determine how many calls were needed for statistical significance, we made the following statistical considerations. Let D denote the total number of genes on the chip, and let F denote the number of false-positive calls. Then q = F/D is the relative frequency of false-positives. For an arbitrary gene, the probability P that there are at least k false-positive calls for this gene out of the six comparisons can be directly derived from a binomial distribution with success parameter q, assuming q is an accurate estimate of the underlying false-positive probability. From this, we would expect, on average, D*p genes with at least k false-positive calls in the entire set of experiments. Since P is typically small, we assume that the number of these genes is roughly distributed according to a Poisson distribution with mean D*p. Hence, we find that the probability of identifying a gene with at least k false-positive calls in the list of D genes is approximately 1 − exp(−D*p). Using this formula, we conclude that four or more calls out of six cannot be explained by chance, with probabilities in the range of 10⁻⁴. Although we do not consider a single occurrence of a gene with three of six induced calls significant, it is likely that the majority of genes that are called as induced in three of six experiments are true positives, since we can assume that the number of false-positives with at least that many calls is Poisson distributed.

Supplementary material.

The entire dataset for both the cell cycle analysis and the E2F-induced gene analysis is available at http://cgt.duke.edu.

RESULTS

Analysis of gene expression during the cell cycle.

When cells are cultured in the absence of normal concentrations of growth factors, they enter a quiescent state usually referred to as G₀. Upon the addition of serum, the cells reenter a growth state and progress synchronously through G₁ into S phase and then G₂ and mitosis. Although a large number of studies have employed this experimental strategy to study the molecular events associated with a proliferative response, there are at least two limitations to this approach. First, gene expression changes that can be measured following the stimulation of quiescent cells to enter a proliferative cycle (serum stimulation) do not distinguish between regulation that is strictly related to growth stimulation versus cell cycle control. For instance, genes induced during G₁, including at G₁/S, may reflect the fact that the cells are reentering a cell cycle as opposed to passing through G₁ from a previous cell cycle; genes induced during this time might not be cell cycle regulated but rather growth regulated. Second, it is largely impossible to measure the events associated with continued cycling in serum-stimulated cultures, in particular the changes taking place at the second G₁/S transition, due to a loss of synchrony as the population of cells proceed into the cell cycle.

To address these issues, we have combined two forms of analysis to study the events associated with cell cycle reentry and cell cycle progression. In the first instance, MEFs were brought to quiescence by serum starvation and then stimulated to grow by the addition of serum. Samples were taken through 24 h after serum addition and analyzed by flow cytometry. Under the conditions of this experiment, cells began to enter S phase at 15 h following serum addition, as indicated by a determination of DNA content by flow cytometry (Fig. 1A). To analyze events specific to the cell cycle and apart from control related to stimulation out of a quiescent state, a second population of MEFs were synchronized at the beginning of S phase by arresting the cells in the presence of HU. Upon removal of the drug, these cells then progressed through S phase, G₂, and mitosis and into the next G₁ and second S phase. We have previously described the use of this experimental approach for the analysis of cell cycle regulation of E2F activity as well as certain E2F target genes (13). Flow cytometry analysis demonstrated that the cells completed the initial S phase by 6 h following release from the HU block and then entered the second S phase approximately 15 h following release (Fig. 1A).

FIG. 1.

Analysis of cell cycle progression in MEFs. (A) MEF cells were synchronized either by serum starvation or by HU block and brought back to the cell cycle progression either by adding serum (left panel) or by adding the fresh medium containing serum without HU (right panel). Cells were harvested at the indicated time points, stained with propidium iodide, and processed for flow cytometry. Percentage of cells in S phase at each time point is plotted. (B) E2F DNA-binding activity in the samples described in panel A. Nuclear extracts prepared from the indicated samples were assayed for E2F DNA-binding activity by electrophoretic mobility shift assay as described in the text. The identity of the indicated E2F binding activities is based on relative gel mobility and identification with specific antibodies. (C) Cyclin E expression during the cell cycle progression. RNA was prepared from the indicated samples and analyzed by Northern blotting, using a cyclin E cDNA probe. An equal amount of mRNA was loaded in each lane. (D) Comparison of gene expression measurement by the Affymetrix GeneChip cyclin E array to that obtained by densitometric scanning of a Northern blot of the same RNA sample. The average difference values of cyclin E gene calculated by the Affymetrix GeneChip expression analysis algorithm were normalized across the two experiments and plotted (■). The intensity of the cyclin E bands in the Northern blot shown in panel C was measured by densitometric scanning. The values are normalized across each experiment and plotted (●).

Aliquots of these samples from the two experiments were also assayed for E2F DNA-binding activity as a measure of progression through the proliferative response. As shown in Fig. 1B, E2F activities previously shown to accumulate at G₁/S, including E2F1 and E2F3a, were first observed at 12 h following serum stimulation and then peaked at 18 h, coinciding with G₁/S, defined by DNA synthesis measurements. These activities were also elevated in the HU-arrested cells and declined as the cells entered S phase, and then E2F3a activity reaccumulated at the second G₁/S transition. These observations parallel results described previously that demonstrate a cell cycle control of E2F3a activity (13, 14). In addition, an assay for cyclin E RNA accumulation by Northern blot revealed an accumulation at G₁/S that parallels the accumulation of E2F activity at G₁/S (Fig. 1C). As such, this experimental approach, which combines analysis of cells reentering a cell cycle from a quiescent state with analysis of proliferating cells leaving a G₁/S arrest, provides a comprehensive view of cell cycle progression.

We next used the RNA from each of these samples to hybridize to high-density DNA microarrays in order to provide a broader examination of the changes in gene expression as cells enter a proliferative state and also pass through a cell cycle. We made use of Affymetrix GeneChip DNA arrays that contained approximately 6,200 murine gene sequences and ESTs. RNA from each of the samples was converted to target following established procedures and then used to hybridize to the GeneChip arrays. The hybridized chips were then processed and analyzed as described in Materials and Methods. The hybridization quantified by the Affymetrix software is shown in Fig. 1D and compared to a densitometric analysis of the cyclin E Northern blot shown in Fig. 1C. It is evident from this analysis that the microarray analysis closely matches the Northern analysis.

In order to identify groups of genes with a similar pattern of expression within the cell cycle, the Affymetrix average difference values for each gene, as calculated by the GeneChip expression analysis algorithm, were plotted as a function of time following serum stimulation or time after HU release. Preliminary visual inspection of the data indicated the existence of distinct patterns of gene expression. We have clustered genes based on vectors of expression levels consisting of Affymetrix average difference values for all time points in both the growth stimulation and the cell cycle experiments. This was done using k-means clustering as implemented in the GeneSpring software (Silicon Genetics). This approach is a self-organization of the measured gene expression data and is hence not biased by any prior expectations of how genes might be regulated. Criteria were set to eliminate genes that failed to show significant induction in the serum stimulation experiment. Expression patterns of genes that met these criteria were normalized across the experiments and then clustered by a k-mean clustering algorithm. We have tested several values for the total number of clusters in the k-means clustering procedure. The final analysis was based on 16 clusters; with fewer clusters, we could not identify a unique course of up- and downregulation within each individual cluster, while a larger number of clusters led to distinct clusters with a similar course of gene expression. For this setup, we can summarize each cluster of genes by a characteristic sequence of up- and downregulation at specific time points in the experiments.

Delineation of multiple, distinct patterns of expression within the mammalian cell cycle.

Figure 2 displays the clusters as a function of the time of expression through the two experiments. As indicated in the figure, clusters could be identified that included genes expressed highly in quiescent cells and then turned off once the cells begin to proliferate (G₀); genes whose expression increased soon after the stimulation of growth and then fell to basal levels (early G₁); genes whose expression increased in G₁, declined, and then increased again during the second G₁ (G₁ cycle); genes whose expression increased in G₁ and then remained constant thereafter (G₁ growth); genes whose expression increased at the G₁/S transition, declined, and then increased again at the second G₁/S transition (G₁/S cycle); genes whose expression increased at G₁/S and then remained constant (G₁/S growth); and finally, genes whose expression increased at a time coincident with the end of S phase, declined, and then increased again at the second G₂ (G₂ cycle). Examples of patterns of expression for specific genes within each cluster are shown in Fig. 3A. The identities of the genes in these clusters, together with information regarding functional properties, were obtained from a search of the UniGene database and are listed in Table 1.

FIG. 2.

Identification of patterns of gene expression following growth stimulation and during the mammalian cell cycle. Expression profile of individual genes in the serum stimulation and HU release experiments, clustered according to the methods described in the text. The expression level of each gene in the two experiments was displayed by a pseudo-color visualization matrix (6). In each experiment, a vertical column represents all of the clustered genes for a given time point. The intensity of expression, as determined from the average difference values calculated by GeneChip expression analysis (Affymetrix), is depicted by the intensity of red color.

FIG. 3.

Specific examples of genes regulated during the cell cycle. (A) Representative example of expression profile among each cluster is shown with its identification. The average difference value at each time point across the cell cycle experiments is plotted for each gene. NGF, nerve growth factor; RXR, retinoid X receptor. (B) Northern analysis for selected G₂ cell cycle cluster genes. RNA samples prepared from the indicated times points of the HU release experiment were analyzed by Northern blotting, using probes for cdc2 and importin-α2. The cyclin E profile is shown for comparison. An equal amount of mRNA was loaded in each lane.

TABLE 1.

Identification of genes regulated during the cell cycle

Accession no.	Gene description	Functional category
G0 cluster
W30101	EST; similar to calcium/calmodulin-dependent protein kinase type 1 (R. norvegicus)	Signal transduction
L07924	Ral guanine nucleotide dissociation stimulator	Signal transduction
W09590	EST; protein-tyrosine phosphatase zeta precursor (R. norvegicus)	Signal transduction
U40576	SIM2	Transcription control
U80036	Paired-like homeodomain transcription factor 2	Transcription control
M74570	Aldehyde dehydrogenase	Metabolism
W14823	EST; similar to integrin beta 5	Cell adhesion
W87253	EST; integrin beta 5	Cell adhesion
U16162	Prolyl 4-hydroxylase alpha-1 subunit	Collagen synthesis
AA000961	EST; legumain	Cysteine endopeptidase
D85414	NEDD-4	E3 ubiquitin-protein ligase (by similarity)
W62849	EST; similar to 4F2 cell surface antigen heavy chain	Amino-acid transport
AA014996	EST; similar to apolipoprotein B (R. norvegicus)	Lipoprotein catabolism
X66976	Procollagen, type VIII, alpha 1	Major constituent of Descemet's membrane
M26270	Acyl-CoA desaturase 2
X97755	MSI
M32240	Peripheral myelin protein, 22 kDa
U00674	Syndecan 2
D16262	Mesoderm-specific transcript
AA109109	EST; similar to myosin regulatory light chain 2 (H. sapiens)
AA062342	EST; similar to RAB-28 (R. norvegicus)
AA014656	EST
AA117286	EST
AA061086	EST
Early G1 cluster
U59876	mGCM1	Transcription control
M82874	Helix loop helix 1	Transcription control
M83380	RelB	Transcription control
X75014	Phox2	Transcription control
M95200	Vascular endothelial growth factor	Signal transduction
X62701	Urokinase plasminogen activator receptor	Signal transduction
D30782	Epiregulin	Signal transduction
Z80941	Semaphorin H	Signal transduction
V00836	Nerve growth factor beta	Signal transduction
X69619	Inhibin beta-A	Signal transduction
U43186	MAPK/ERK kinase kinase 2	Signal transduction
AA107252	EST; phosphatidylinositol 4-kinase (R. norvegicus)	Signal transduction
L01695	Calcium/calmodulin-dependent 3′,5′-cyclic nucleotide phosphodiesterase 1B	Signal transduction
M14872	Gonadoliberin I	Signal transduction
K02108	Keratin complex 2, gene 6a	Filament assembly
M88242	Prostaglandin G/H synthase 2 precursor	Prostaglandin synthesis
M91443	Gap junction beta-5	Transmembrane channels
U66865	Vacuolar protein sorting-associated protein 45	Vesicular transportation
X73985	Calretinin	Calcium binding protein
X79003	Integrin alpha 5	Cell adhesion
L42293	Sterol O-acyltransferase	Cholesterol synthesis
M55154	Transglutaminase 2, C polypeptide	Protein cross link
M55669	Neuroendocrine convertase 2	Prohormone processing
M73696	Glvr-1	Inorganic phosphate transporter
Z30970	TIMP-3	Regulation of matrix metalloproteinase activity
D10727	NPC-derived proline-rich protein
D50263	DAN
U68267	Myosin binding protein H
X04123	DNA nucleotidylexotransferase
AA028547	EST; hypothetical protein
AA051231	EST; similar to zinc finger protein
AA097292	EST; beta-tubulin
AA137667	EST; similar to gamma-filamin (H. sapiens)
W09925	EST; gamma-filamin (H. sapiens)
W97775	EST; similar to gamma-filamin (H. sapiens)
AA002759	EST
AA023458	EST
AA062328	EST
AA106256	EST
AA106913	EST
W08057	EST
W34915	EST
G1 cycle cluster
X66225	Retinoid X receptor γ	Nuclear hormone receptor/transcription control
M26053	T-cell receptor germline beta-chain gene constant region	Signal transduction
U70622	Lysophosphatidic acid receptor	Signal transduction
M21532	Purkinje cell protein 2	Signal transduction
M12836	T-cell receptor gamma, variable 4	Signal transduction
W83658	EST; similar to G protein gamma 2 subunit	Signal transduction
L36135	T-cell receptor delta chain	Signal transduction
M26417	T-cell receptor beta chain	Signal transduction
D13867	Integrin alpha 3	Cell adhesion
W50665	EST; similar to T-cadherin	Cell adhesion
AA028501	EST; cytochrome c oxidase, subunit VIIIb	Mitochondrial electron transport
M60429	Immunoglobulin heavy chain C region	Immunoglobulin
K02785	cDNA induced by PDGF with some homology to c-fos
M96430	Acrosin
M35667	Lysozyme-binding Ig κ chain (HyHEL-10) V23-J2 region
M23236	Proline-rich protein M14
W62585	EST; plenty-of-prolines-101
W65899	EST; similar to guanine nucleotide-binding protein g(i)/g(s)/g(t) beta subunit (H. sapiens)
W89939	EST
W91095	EST
W98265	EST
W98496	EST
G1 growth cluster
U27457	ORC2	Replication initiation
U20225	adl	DNA synthesis
M11310	APRT	DNA synthesis
W34756	EST; uridine kinase (R. norvegicus)	DNA synthesis
U44426	D52	Signal transduction
U28151	p130Cas	Signal transduction
U02313	MAST205	Signal transduction
W15568	EST; similar to U1 snRNP-specific protein C	Pre-mRNA processing
D87691	eRF1	Transcription termination
U51002	DLX-2	Transcription control
J04179	HMG-Y	Transcription control
W13191	EST; similar to thyroid hormone receptor alpha-2	Nuclear hormone receptor/transcription control
AA062397	EST; similar to RAB-5B	Protein trafficking
U07951	GDP-dissociation inhibitor	Protein trafficking
X01237	P53	Cell cycle checkpoint
D63784	Id-associated protein 1	Cell growth
W85270	EST; similar to inorganic pyrophosphatase (H. sapiens)	Lipoprotein catabolism
D85904	APG-2	Heat shock protein
U21673	T complex testis-specific protein 2
U77415	Bop1
AA111149	EST; similar to tubulin alpha 2
U43918	Proliferation-associated protein 1
U49720	Blue-sensitive opsin
W18778	EST; similar to tubulin Mbeta 1
W41072	EST; similar to ubiquitin conjugating enzyme (H. sapiens)
AA138226	EST; clathryn light chain (R. norvegicus)
AA051583	EST
AA062365	EST
AA073743	EST
AA088121	EST
AA110046	EST
W35742	EST
W49178	EST
W56929	EST
G1/S cycle cluster
X15666	Ribonucleotide reductase M2	DNA replication
U63337	Cdk2	DNA replication
X75888	Cyclin E	DNA replication
X62154	MCM3	DNA replication
D26089	MCM4	DNA replication
D26090	MCM5	DNA replication
D86726	MCM6	DNA replication
D26091	MCM7	DNA replication
X57800	PCNA	DNA replication
X72711	Replication factor C, 140 kDa	DNA replication
W20901	EST; similar to replication factor C 38-kDa subunit (H. sapiens)	DNA replication
D17384	DNA polymerase alpha 1, 180 kDa	DNA replication
U19604	DNA ligase I	DNA replication
L26320	Flap endonuclease-1	DNA replication
D13473	Rad51	DNA repair
U42190	EST; MSH6	DNA repair
X70472	B-myb	Transcription control
U75680	Histone stem-loop binding protein	Pre-mRNA processing
U27177	p107	Cell cycle regulation
X14805	DNA methyltransferase (cytosine-5)	DNA synthesis
AA096762	EST; similar to NASP
AA154451	EST
G1/S growth cluster
K02927	Ribonucleotide reductase M1	DNA replication
U05823	Pericentrin	Mitois/meiosis
D12646	KIF4	Mitosis
X72310	DP-1	Transcription control
M60523	ID-3	Transcription control
AA023287	EST; similar to TFIIB chain A (H. sapiens)	Transcription initiation
U60530	MAD homolog 2	Signal transduction
X56045	RAN binding protein 1	RNA export/protein import
X63100	Gap junction alpha-6	Transmembrane channels
X13752	Delta-aminolevulinate dehydratase	Porphyrin and heme biosynthesis
U70494	Histone H2A.Z
Y09419	Spermine synthase
K03235	Proliferin 2
X16151	T-lymphocyte activation protein
M72394	Phospholipase A2
AA138105	Similar to NOVEL ANTIGEN 2 (H. sapiens)
L15429	L6 antigen
AA118151	EST
G2 cycle cluster
U58633	Cdc2	Mitosis
L16926	Cdc25C	Mitosis
AA000468	EST; similar to p55CDC (cdc20)	Mitosis
AA008043	EST; similar to p55CDC (cdc20) (R. norvegicus)	Mitosis
Z26580	Cyclin A2	Mitosis
X64713	Cyclin B1	Mitosis
U20636	Cyclin F	Mitosis
D55720	Importin alpha 2	Mitosis
X82786	Ki-67	Mitosis
W70733	EST; similar to KIFC1	Mitosis
W98100	EST; similar to KIF3B	Mitosis
L06144	Polo-like kinase	Mitosis
AA050055	EST; Ayk1	Mitosis/melosis
Z46757	HMG2	Transcription control
L11316	Ect2	Signal transduction
M86377	TTK	Signal transduction
D21099	STK-1	Signal transduction
X64550	Intracellular hyaluronic acid binding protein	Signal transduction
AA023099	EST; similar to dUTPase (R. norvegicus)	DNA synthesis
U20497	p19-INK4D	Cell cycle regulation
AA032836	Similar to cyclin-dependent kinases regulatory subunit 1 (H. sapiens)	Cell cycle regulation
Y09632	Rab6	Protein trafficking
W99981	EST; similar to peptidyl-protyl cis-trans isomerase a	Protein folding
AA067813	Glutamate cysteine ligase	Glutathione biosynthesis
AA028398	EST; tubulin beta-3
X58069	Histone H2A.X

Although there were clusters identified in the k-means clustering analysis whose biological relevance was not immediately apparent, other clusters clearly related to known functional properties. For instance, the G₁/S and G₂ clusters included a number of genes encoding replication and mitotic activities, respectively. The relationship between the time of RNA accumulation and the time when the gene product functions, at least for replication activities controlled at G₁/S, has not always been seen in past work studying yeast cell cycle control. In addition, past experiments have not clearly detailed a role for transcriptional regulation during G₂ in mammalian systems. In large part this is a reflection of the experimental strategy, which generally examines gene expression following serum stimulation of quiescent cells. Simply examining the pattern of gene expression following stimulation of cell growth does not reveal a clear pattern of gene control at G₂, a situation most likely due to loss of cell synchrony. That is, such genes are induced by serum addition, but whether they are activated at G₁/S, in S phase, or later is difficult to discern (for instance, compare the G₁/S and G₂ clusters in Fig. 3A).

Previous work has suggested that some of the genes in the G₂ cluster are induced at either G₁/S of the cell cycle or late in G₁ (8). In order to confirm that the microarray analysis did indeed reflect the true behavior of these genes, we assayed the samples from the HU release experiment by Northern analysis, using probes for several genes categorized as G₂ regulated. As shown in Fig. 3B, it is apparent that both the cdc2 gene and the importin-α2 gene are indeed activated at G₂, consistent with the microarray assays. These patterns are in sharp contrast to the pattern for cyclin E expression, which is regulated at G₁/S. We believe that the discrepancy between these data and previous studies very likely reflects the method of cell synchronization and the ambiguity of cell cycle position when only a serum stimulation experiment is employed.

The importance of combining the HU-synchronized samples with the serum-induced samples is clearly illustrated by the last three clusters identified in Fig. 2. An analysis of only the serum-induced samples would not distinguish these genes. Rather, they would be grouped together as genes induced late in G₁. But by combining these data with the HU-synchronized analysis, it becomes readily apparent that there are in fact three distinct clusters—genes induced late in G₁ that remain constant, genes induced late in G₁ that cycle, and genes induced in G₂ that cycle.

Finally, a particularly revealing relationship can be seen in those genes that are activated at G₁/S. One group includes genes activated during G₁ whose expression levels remain high as cells continue to proliferate (G₁/S growth cluster). This group includes genes encoding a variety of proteins that function in transcription, signal transduction, and RNA metabolism (Table 1). In contrast, a second group is also activated at G₁/S, but expression of this group oscillates as the cells continue to cycle in the presence of growth factors (G₁/S cycle cluster). This group includes genes whose function is distinct from the other G₁/S-induced group of genes in that these genes encode proteins that are almost exclusively involved in DNA replication. We do note that there is some discrepancy between these results and past experiments that identified several of these DNA replication genes as showing constant expression following G₁/S (13). In particular, the previous work suggested that the expression of a subset of the Mcm genes was constant following the initial G₁/S, whereas the analyses performed here with DNA microarrays revealed an oscillation in the expression of each of the Mcm genes, as shown for mcm7 in Fig. 3A. Although we cannot identify a clear distinction in the two analyses that would explain this difference other than a cell type difference, the fact that a substantial number of additional genes encoding replication proteins are coordinately regulated in this manner leads us to believe that the G₁/S oscillating pattern of expression may be a common aspect of control of replication activities.

Identification of genes induced by expression of E2F activities.

We have previously described the use of recombinant adenovirus vectors as a means to efficiently produce proteins in otherwise quiescent cells (4). The strategy takes advantage of the ability of adenoviruses to infect quiescent cell populations and do so with an efficiency that allows a biochemical analysis of the entire population of cells. Given the fact that the E2F1, E2F2, and E2F3a activities normally only accumulate at G₁/S of the cell cycle, as demonstrated previously and as shown by the data here in Fig. 1, overproduction of these proteins in a quiescent cell allows an analysis of the induction of potential target genes by these E2F proteins in the absence of other growth regulation activities. Indeed, we have made use of this approach in past experiments to study the induction of various E2F target genes (3, 4, 12). We have now extended this work through the use of DNA microarrays to facilitate the assay of large numbers of genes in order to gain a more comprehensive view of the pathway of gene control involving E2F activities. Moreover, by performing these analyses in conjunction with the cell cycle determinations, they provide an opportunity to establish a context for understanding previously characterized as well as uncharacterized E2F-regulated genes.

MEFs were brought to quiescence by serum starvation and then infected with either a control adenovirus that expresses green fluorescent protein (GFP) or with viruses that express the E2F1, E2F2, or E2F3 gene products. As shown in Fig. 4A, these conditions allowed an accumulation of E2F1 or E2F2 activity that, at least for E2F1, was at a level similar to that observed when cells normally pass through G₁/S. Thus, the experimental approach does not represent a gross overproduction of the proteins but rather an accumulation to near physiological levels in the absence of the other events normally associated with a proliferative response. In contrast to the accumulation of E2F1 and E2F2 activity, the production of E2F3 activity was markedly reduced compared to the others despite the use of a substantial multiplicity of infection (data not shown). Indeed, an increase in E2F3 activity was only clearly evident upon treatment of extracts with deoxycholate, suggesting that the majority of the ectopically expressed protein was bound to Rb. Given this reduced level of E2F3 activity, we have chosen to focus primarily on the analysis of gene induction by E2F1 and E2F2. A virus titration was used to determine the multiplicities of infection needed to achieve an equivalent level of E2F1 and E2F2 activity.

FIG. 4.

Expression of E2F activities in quiescent fibroblasts. (A) Production of E2F binding activity in cells infected with recombinant adenoviruses containing either the E2F1 or E2F2 gene. Quiescent MEF cells were infected with either E2F1- or E2F2-expressing recombinant adenovirus or with GFP-expressing recombinant adenovirus as a control. E2F2 virus was infected at three different multiplicities to obtain the same E2F DNA-binding activity as with E2F1. E2F binding activities were analyzed by E2F DNA-binding assay with the nuclear extracts prepared from the infected cells. (B) Induction of cyclin E RNA accumulation in cells expressing E2F activities. Expression of cyclin E mRNA was analyzed by Northern analysis in E2F1- and E2F2-expressing cells. mRNA was prepared from the same cells analyzed for E2F DNA-binding activities in panel A and subjected to Northern analysis using a cyclin E probe.

Measurement of the expression of cyclin E, a previously demonstrated E2F target, demonstrated that the production of the E2F1 and E2F2 activities did lead to an induction of cyclin E expression (Fig. 4B). We then used the RNA from these infections to generate target for GeneChip analysis. Targets prepared using the RNA from Ad-E2F-infected cells were hybridized to sets of the Affymetrix murine 11K GeneChips and compared to the hybridization pattern obtained with a control (target prepared from RNA from control-virus-infected cells).

We set the following criteria based on the Affymetrix GeneChip expression analysis software as the basis for identifying genes induced by E2F activities: an intensity of expression (average difference value) that was greater than or equal to 50 in the E2F-expressing cells; the gene was considered increased or marginally increased by comparison analysis using the Affymetrix GeneChip expression analysis algorithm; the fold change, as reported by the Affymetrix comparison analysis, was greater than or equal to 2.0. Of the approximately 11,000 sequences scored in the hybridization assays, a small fraction in any given experiment met these criteria. For instance, in one experiment in which the 11,000 sequences were scored for expression using RNA from E2F1- or E2F2-expressing cells, a total of 255 genes exhibited an induction of at least twofold.

It was also clear from an inspection of the data that there was variation from experiment to experiment in the genes scored as induced in the E2F-expressing cells. Such variation could represent differences in the actual experimental manipulations; alternatively, variations in the hybridization analysis could contribute to the variation. To address the basis for the variation, RNA expression was analyzed from two independent experiments. In addition, the RNA samples from one of these experiments were assayed twice independently. Samples obtained from each of these experiments were used to prepare targets and then used for hybridization to the 11,000 murine gene DNA microarray. Reproducibility was assessed by comparing the duplicate hybridization of a given sample. A comparison of the expression profiles of any given gene sequence in the duplicate hybridizations should, in principle, yield the same value. However, we observed 83 genes scoring as induced in the second hybridization over the first, using the criteria described above for the case of two E2F1-expressing samples, and 69 genes scored as induced in the second hybridization over the first for the E2F2 sample. These false-positives constitute different genes for the E2F1 and E2F2 comparisons, and they do not cluster into any known functional group. In contrast, they appear to represent a random sample from a uniform distribution of the set of genes on the chip.

Clearly, the variation described above leads to statistical significance problems for “calls” of induced genes if they are based on a single comparison. To address this issue, we examined all six analyses of gene expression comparing E2F1 or E2F2 against the control. While we would expect a substantial number of false-positive calls for each individual comparison caused by chance variation in measurement, we do not expect these false-positive calls to refer to the same genes in several comparisons. For instance, cyclin E met the criteria in all six possible comparisons, and there were many more genes that met the criteria in more than one comparison. To ensure maximal confidence in the identification of genes as truly induced by E2F activity, we have combined the data for the E2F1 and E2F2 expression analysis and used a criterion of induction being called in four of the six assays to identify genes as induced by E2F activities (see Materials and Methods for a description of the statistical analysis).

It is evident from the list detailed in Table 2 that many previously identified E2F target genes, including cyclin E, cdk2, and thymidylate synthase, were found in this group. But additional genes were evident as well, including other activities known to function in conjunction with DNA replication, such as DNA primase, DNA ligase, flap endonuclease, and topoisomerase. In addition to these, we also identified a number of E2F-induced genes that encode activities not involved in DNA replication, such as several transcriptional regulatory proteins (HMG proteins, enhancer of zeste), DNA repair (RAD51), and cell cycle control (p18). The largest group of E2F-induced genes apart from those encoding replication activities was, however, a collection of genes that encode proteins that function in mitosis. These include kifC1, cdc2, cyclin B, and cdc20.

TABLE 2.

Identification of E2F-induced genes

Accession no.	Gene description	Functional category	Mean induction (fold) ± SD (n = 3)
			E2F1	E2F2	Northern	Reference
M13352	Thymidylate synthase	DNA replication	2.5 ± 0.3	3.3 ± 0.1	Yes	3
M19438	Thymidine kinase 1	DNA replication	5.3 ± 0.8	5.9 ± 0.9	Yes	4
AA023099	EST; similar to dUTPase (R. norvegicus)	DNA replication	15.2 ± 2.8	13.7 ± 5.3
K02927	Ribonucleotide reductase M1	DNA replication	2.5 ± 0.3	3.9 ± 0.4
M14223	Ribonucleotide reductase M2	DNA replication	4.3 ± 0.5	4.4 ± 0.3	Yes	4
X15666	Ribonucleotide reductase M2	DNA replication	19.6 ± 5.4	20.7 ± 3.3	Yes	4
U63337	CDK2	DNA replication	1.9 ± 0.5	2.3 ± 0.3	Yes	4
X75888	Cyclin E	DNA replication	14.4 ± 3.6	22.3 ± 6.2	Yes	3; this study
X62154	MCM3	DNA replication	2.2 ± 0.2	3.1 ± 0.8	Yes	13
D26091	MCM7	DNA replication	1.9 ± 0.1	2.4 ± 0.1	Yes	13
AA690055	EST; similar to DBF4-related protein	DNA replication	2.0 ± 0.1	2.0 ± 0.1
AA692716	EST; similar to DBF4-related protein	DNA replication	5.3 ± 2.9	4.8 ± 1.2
X53068	PCNA	DNA replication	2.2 ± 0.4	2.6 ± 0.1	Yes	2
W20901	EST; similar to replication factor C 38-kDa subunit	DNA replication	3.0 ± 0.5	3.6 ± 0.8
D13544	DNA primase small subunit	DNA replication	2.1 ± 0.3	2.3 ± 0.2
D12513	DNA topoisomerase II alpha	DNA replication	3.1 ± 1.1	3.9 ± 1.3
D12513	DNA topoisomerase II alpha	DNA replication	3.5 ± 1.8	6.5 ± 4.0
U04674	DNA ligase I	DNA replication	2.2 ± 0.3	3.1 ± 0.5
L26320	Flap endonuclease-1	DNA replication	2.8 ± 1.1	3.0 ± 1.1
D13473	Rad51	DNA repair	2.6 ± 0.9	3.6 ± 0.5
U58633	cdc2	Mitosis	1.9 ± 0.4	2.2 ± 0.4	Yes	3; this study
AA008043	EST; similar to p55CDC (cdc20) (R. norvegicus)	Mitosis	3.3 ± 0.4	4.5 ± 0.5	Yes	This study
Z26580	Cyclin A2	Mitosis	2.2 ± 0.1	2.7 ± 0.2	Yes	3
X58708	Cyclin B1	Mitosis	3.9 ± 0.6	4.5 ± 1.7	Yes	This study
X66032	Cyclin B2	Mitosis	1.9 ± 0.2	2.5 ± 0.5
X82786	Ki-67	Mitosis	2.2 ± 0.1	2.7 ± 0.3
D49544	KIFC1	Mitosis	5.3 ± 1.1	10.7 ± 9.9
U42385	FIN16	Mitosis	2.9 ± 0.9	4.5 ± 1.4
Z31235	EST; stathmin	Mitosis	2.1 ± 0.2	2.9 ± 0.4
AA267955	EST; similar to HEC	Mitosis	1.8 ± 0.2	2.3 ± 0.0
AF002823	BUB1	Mitosis	2.2 ± 1.0	3.3 ± 0.3
D55720	Importin alpha-2 subunit	Mitosis	3.0 ± 1.5	3.7 ± 2.7	Yes	This study
Z46757	HMG2	Transcription control	6.2 ± 1.8	7.7 ± 1.5
U52951	Enhancer of zeste	Transcription control	2.7 ± 0.0	3.1 ± 0.6
D21099	STK-1	Signal transduction	6.1 ± 0.9	8.1 ± 2.2
AA711028	EST; similar to pituitary tumor-transforming 1	Signal transduction	2.2 ± 0.4	2.6 ± 0.5
U75680	Histone stem-loop binding protein	Pre-mRNA processing	2.7 ± 0.2	2.4 ± 0.2
M26391	RB	Cell cycle regulation	2.2 ± 0.6	2.7 ± 0.6	Yes	3
U19596	p18-INK4C	Cell cycle regulation	2.2 ± 0.3	2.0 ± 0.1	Yes	4
AA059527	EST; similar to brain neuron cytoplasmic protein 2		2.5 ± 0.3	2.9 ± 0.3
U69488	Viral envelope like protein		3.5 ± 1.0	2.8 ± 0.5
AJ002390	Annexin VIII		6.8 ± 3.4	9.9 ± 1.8
AA002747	EST		2.9 ± 0.1	3.3 ± 0.4
AA002925	EST		5.3 ± 2.4	7.2 ± 4.3
AA051276	EST		3.7 ± 1.0	6.0 ± 2.7
AA117100	EST		1.9 ± 0.2	2.2 ± 0.1
AA154451	EST		2.7 ± 0.9	3.5 ± 0.4
AA184798	EST		2.2 ± 0.6	2.9 ± 0.1
AA189300	EST		3.2 ± 1.1	3.4 ± 1.0
AA189313	EST		3.0 ± 0.6	2.8 ± 0.2
AA200970	EST		2.2 ± 0.3	2.4 ± 0.0
AA266783	EST		2.4 ± 0.3	3.1 ± 0.7
AA285553	EST		3.9 ± 1.4	6.3 ± 3.3
AA407737	EST		7.6 ± 2.6	4.6 ± 1.3
AA426917	EST		2.6 ± 0.3	2.7 ± 0.6
AA516966	EST		11.5 ± 3.5	11.2 ± 6.0
AA561108	EST		2.1 ± 0.2	2.7 ± 0.2
AA590750	EST		2.6 ± 0.2	3.1 ± 0.3
AA590750	EST		2.4 ± 0.5	2.6 ± 0.2
AA592163	EST		8.5 ± 1.0	9.3 ± 0.9
AA673176	EST		2.3 ± 0.2	2.4 ± 0.2
AA673431	EST		2.0 ± 0.1	2.6 ± 0.6
C76791	EST		1.8 ± 0.3	2.7 ± 0.1
C77497	EST		2.8 ± 0.5	3.3 ± 1.5
C81593	EST		3.2 ± 0.7	3.3 ± 0.5

^aEvidence for induction based on Northern analysis of RNA from Ad-E2F-infected cells. 

Relationship of E2F-induced genes to cell cycle control—role for E2F in control of expression of G₂ genes.

The finding that many of the genes induced by either E2F1- or E2F2-encoded proteins known to function during mitosis was surprising given the fact that E2F activity, particularly E2F1-3, normally accumulates at G₁/S of the cell cycle. As such, it raised the possibility either that the effect of E2F activation on these genes was indirect or that these genes are normally regulated at G₁/S even though the products function in mitosis. The latter scenario has precedence, since a number of yeast DNA replication genes are induced in mitosis, well before S phase (1, 23). To address this question, we have examined the relationship between the control of transcription by E2F proteins and the control during the cell cycle.

As shown by the data in Fig. 5A, the E2F-induced genes did not distribute uniformly over all clusters derived from the cell cycle analysis. Rather, the majority accumulated in only two of these cell cycle clusters. Most of the E2F-induced genes fell into either the G₁/S cell cycle cluster, genes whose expression peaks at the initial G₁/S transition upon stimulation of cell proliferation and whose expression then continues to oscillate during the cell cycle with a peak at G₁/S, or the G₂ cell cycle cluster. The clustering of E2F-induced genes within the G₁/S group of cell cycle-regulated genes is consistent with previous work that demonstrates an accumulation of E2F activities at this time of the cell cycle. In contrast, the finding that a number of genes induced by E2F proteins are normally regulated at G₂ is surprising in light of the fact that these E2F activities are essentially undetectable at this time of the cell cycle. Although it is possible that there is an accumulation of E2F activity in G₂ that has gone undetected in previous work or that there is a role for other E2F activities, such as E2F4 or E2F5, which are present at this time, in transcription activation, it is also possible that the activation of these genes that are normally regulated at G₂ during the cell cycle is a secondary effect of E2F accumulation at G₁/S.

FIG. 5.

Relationship of E2F-regulated genes to control in the cell cycle. (A) Genes identified as E2F-induced, plotted as a function of the number of experiments in which the gene was induced, are compared to the cell cycle-regulated clusters shown in Fig. 3. For this comparison, we have included those genes that were scored as induced by either E2F1 or E2F2 in at least three of the six assays that were performed. Although there is some probability that a gene induced in only three of six assays could represent a false-positive, the probability is very small. (B) Northern analysis of E2F gene induction. RNA samples prepared from cells infected with either Ad-Con, Ad-E2F1, or Ad-E2F2 virus were analyzed by Northern blotting using probes for RRM2, Cdc20, cyclin B1, and importin-α2. An equal amount of mRNA was loaded in each lane.

To provide further verification of the induction of genes by E2F, particularly those in the G₂ category, we analyzed RNA samples by Northern blot assays. As shown by the data in Fig. 5B, the E2F-mediated induction of one of the G₁/S-regulated genes (RRM2) was clearly evident, similar to the induction of cyclin E, as seen in the analysis shown in Fig. 4B and consistent with the induction of many others in this category (Table 2). In addition, we also assayed several of the genes identified in the G₂ cluster, including cdc20, cyclin B1, and importin-α2. It is evident from these assays that each of these genes was indeed induced by E2F, either E2F1 or E2F2, similar to the induction of the G₁/S genes, thus confirming the results of the DNA microarray analysis.

The fact that cells expressing E2F1, E2F2, or E2F3 do not complete S phase or enter mitosis (data not shown) argues that the induction of these G₂-specific genes is not the simple consequence of induced cell cycle progression. But whether the G₂-specific genes such as cyclin B are directly or indirectly activated by E2Fs is not clear and must await a determination of the promoter elements that are critical for the induction of these genes in G₂.

DISCUSSION

A considerable body of work has detailed the transcriptional control properties of the E2F proteins, including the fact that E2F activities are critically important for the activation of genes that encode proteins important for DNA replication. Nevertheless, progress to this point has been incremental and driven largely by prior knowledge. The approach that we describe here represents an unbiased examination of the genes that are subject to E2F control, particularly as they relate to the normal control of the cell cycle. We believe that two important observations derive from these data. First, the logic of gene control during the mammalian cell cycle largely reflects an activation of genes at the time the gene products are required to function. Second, although E2F activity primarily accumulates at G₁/S, genes that are normally activated at G₂ of the cell cycle are also subject to E2F control.

Gene control during the cell cycle.

Although cell cycle control of gene expression has been studied in detail in yeast, studies in mammalian systems have generally been limited to the initial events following the stimulation of cell proliferation, including recent studies that have employed DNA microarrays to measure the expression of large numbers of genes (7, 10). In general, the experimental approach employed in these studies uses cells synchronized in a quiescent state as a result of growth factor deprivation. When growth factors are then added to such cultures, the cells reenter the cell cycle and maintain a reasonable degree of synchrony through the initial cell cycle. Studies of such cell populations for changes in gene expression have revealed waves of gene expression as the cells move from the quiescent state through G₁ and into S phase. This includes genes transcribed in the quiescent cell that are shut off when proliferation is stimulated, genes that are activated early in the proliferation process, and genes that are activated later in G₁. The genome-scale analyses recently performed have characterized the regulation of genes involved in fibroblast-specific processes such as wound healing but also a variety of genes involved in events such as cytoskeletal remodeling.

Our analysis of cell cycle control of gene expression extends these studies by combining the assay of gene expression in cells stimulated to reenter a cell cycle by addition of growth factors together with the assay of cells synchronized at G₁/S by HU block that are then released and allowed to pass through another cell cycle. This has allowed us to distinguish genes activated following the stimulation of cell growth that either remain constant in their expression as the cells continue to proliferate or oscillate in expression as cells begin to cycle. Two examples of the cell cycle clusters that derive from these analyses are particularly informative. First, for the genes activated at G₁/S, two distinct subgroups can be identified—those whose expression remains constant and those that oscillate, with peak expression occurring at the following G₁/S transition. Strikingly, this distinction in expression pattern of genes activated at G₁/S reflects a distinct grouping of functional activities, at least for the genes that oscillate as cells continue to cycle, since this group largely encode the DNA replication activities as well as DNA repair genes.

The second clear example is the group of genes activated at G₂, which then oscillate in expression as cells continue to grow. Once again, these genes, which are clustered according to expression pattern, constitute a functional group. As is evident from our work and consistent with a recently published study that also examined cell cycle-specific gene control (2), genes activated at G₂ encode proteins involved in mitotic functions. Cho and colleagues also noted the regulation of genes involved in cell motility and remodeling of the extracellular matrix (2), suggesting a balance between cell proliferation and cell invasion. Taken together, it would appear that the expression of activities during the mammalian cell cycle coordinates synthesis with the time at which the activities are required to function.

Relationships of cell cycle control in yeast and mammalian cells.

The most extensive analyses of cell cycle-regulated gene expression, particularly through the use of DNA microarrays that include the entire set of open reading frames, has been carried out in S. cerevisiae. Two previous studies have detailed the gene expression changes during the S. cerevisiae cell cycle (1, 23). When comparing the results described here for the analysis of mammalian cell growth to these previous studies, it is apparent that there are many similarities in the program of cell cycle regulation in the two systems. For instance, many of the genes that encode the activities directly or indirectly involved in DNA replication are regulated near the G₁/S transition in both systems (1, 23). In addition, several DNA repair activities, including Rad51 and Msh6, are similarly controlled at G₁/S in yeast and mouse cells. Nevertheless, it is also evident that there are differences. The sharpest contrast between control in yeast and mouse cells is seen for the genes encoding DNA replication initiation proteins. Although each of the genes encoding proteins involved in replication initiation, such as Cdc6, Orc1, and the Mcm proteins, is regulated at G₁/S in mammalian cells, the majority of these are regulated either at mitosis or early in G₁ in yeast cells. Presumably, this difference in timing of expression of genes encoding the initiation complex proteins reflects a distinction in the mechanisms of prereplication complex assembly in the two systems.

E2F gene control and the cell cycle—role for E2F in control of mitotic activities.

Consistent with previous work, many of the genes newly identified as induced by the E2F proteins include those encoding DNA replication activities such as replication protein C, DNA ligase, DNA primase, topoisomerase, and flap endonuclease (Fig. 6). In addition, other E2F targets include genes encoding proteins that function in DNA metabolism, such as DNA repair enzymes. As such, it seems possible that the majority of the DNA synthetic machinery, including the apparatus that assembles at origins of replication, is regulated at G₁/S by E2F activities. Another recent study using DNA microarrays to analyze E2F-induced gene expression also identified DNA replication and cell cycle genes as induced by E2F proteins (15), but this study also identified a large number of additional genes with roles in apoptosis, differentiation, and development, the majority of which were not scored in our assays. Several reasons could explain the differences, but possibly they reflect differences in the cell type used for the expression of E2Fs as well as the use of actively growing cells instead of quiescent cells in our study.

FIG. 6.

E2F gene regulatory pathway. Schematic representation of the pathway initiated upon stimulation of cell growth that leads to an induction of cyclin D/cdk4 activity. A primary target for D/cdk4 is the Rb protein, inactivating its ability to regulate the accumulation of E2F proteins. Previously identified E2F targets include replication enzymes, proteins that form the initiation complex, and activities that regulate the function of the origin complex. The data presented here now identify additional genes (indicated in bold) that fall within these previously defined groups as well as new groups as indicated. pol, polymerase; Enh, enhancer; Rib. red., ribonucleotide reductase; thy., thymidine.

Perhaps of most interest in the analysis reported here is the finding that many of the E2F-induced genes are normally regulated at G₂ in the cell cycle and encode proteins that function in mitosis. Past work has documented changes in E2F activity as cells leave G₀ and then as cells pass through G₁/S, but there is no evidence for alterations in E2F activity as cells pass through the G₂ phase of the cell cycle. In several cases, the E2F-mediated control of these genes has already been recognized, since past work has shown that cyclin A, cyclin B, and cdc2 are regulated by E2F. With the exception of cyclin B, previous work has characterized the cell cycle control of these genes as occurring at G₁/S, not G₂. We believe this is largely the result of the method of cell synchronization and analysis, making it difficult to discern a peak of induction either in late G₁ or in G₂.

Although the vast majority of work has focused on the role of E2F in controlling expression of genes at G₁/S, it is true that previous work has provided evidence of a connection between E2F activity and the control of mitotic activities, at least in Drosophila melanogaster. In particular, the work of Edgar and colleagues has shown that the cdc25 string product, a rate-limiting activity for progression through mitosis, is a target for E2F in Drosophila cells (16). Moreover, overexpression of E2F was shown to accelerate both G₁/S and G₂/M, consistent with the ability of E2F to induce both cyclin E and string, rate-limiting activities for transition through these two cell cycle transitions. However, whereas the mammalian cdc25 gene is transcribed at G₂/M, the Drosophila cdc25 gene (string) appears to be expressed in G₁ (11).

Although it remains possible that there is a particular E2F activity or modified form of an E2F activity that is specifically operational at G₂, it is also possible that the induction of these genes normally regulated at G₂ is a secondary effect of the E2F activities. A trivial explanation would be that activation of these genes reflects an E2F-induced cell cycle progression. As such, the induction of the mitotic genes would simply reflect the stimulation of cell cycle progression. We believe this possibility is unlikely, since under the conditions of this experiment, there is little evidence for cells progressing through S phase. There is an induction of DNA synthesis, and this does appear to reflect true DNA replication, but the extent of this replication is quite limited. This is perhaps best seen by a cell sorting analysis that measures the DNA content of the cell population following expression of the E2F activities; these assays reveal an increase in the DNA content of the cell population but no evidence for progression to a G₂ DNA content. In addition, there is no indication for the appearance of any mitotic cells in the population.

Given these observations, we can envision at least two alternative explanations for the E2F-mediated induction of genes such as cyclin B, cdc2, and Bub1. One possibility is that these genes are activated by transcription factors whose expression is controlled at G₁/S by E2F activities. In this scenario, E2F gene control would establish a cascade of events, initially activating the genes encoding DNA replication activities and then secondarily activating genes encoding mitotic activities. Simple kinetic experiments to measure the timing of activation of genes following E2F induction, to determine if the induction of genes such as cyclin E precedes the induction of cdc2, have been inconclusive (data not shown). A second possibility could relate to the recent studies of Dean and colleagues, which provide evidence for two forms of E2F/Rb-mediated transcription repression (25). One repressor complex, which is inactivated by cyclin D/cdk4, appears to control genes normally expressed at G₁/S, including cyclin E. A second repressor, which is not affected by cyclin D/cdk4 but is inactivated by cyclin E/cdk2, persists longer in the cell cycle and appears to control genes such as cyclin A. This is thus consistent with the G₂ regulation seen in the experiments reported here. Thus, the induction of both groups of genes by E2F overexpression in our experiments could reflect a relief of two distinct types of repression that are normally temporally regulated in the cell cycle. Ultimately, the answer to this question will require a determination of the factors normally responsible for the G₂-specific control of genes such as cdc2.

Finally, although the complexity of the E2F family would suggest the potential of specificity in the activation or repression of transcription by the individual E2F family members, there are only hints of such from previous work and from the data generated in the present studies. For instance, previous work employing recombinant adenoviruses to express each of the E2F proteins demonstrated differences in gene induction (4), suggesting the potential for gene-specific activation events. Nevertheless, it is also true that the differences in gene induction by any one member of the E2F family are minimal. Moreover, the loss of function of individual E2F family members also has minimal consequences for gene regulation, with the disruption of E2F3 function appearing to have the most dramatic effect (8). Thus, either there is substantial overlap in gene induction by the individual E2F proteins or the specific targets have not yet been clearly identified.