INTRODUCTION

Aldolase A gene expression is regulated by an intriguing mechanism that involves three alternative promoter regions (distal or pL, middle or pM, and proximal or pF). Transcription of multiple mRNA species (L, M, and F types), containing the same coding region but different untranslated 5-ends, is driven by three autonomous promoters and splicing of alternative leader exons (L, M, and F) (1-3). In man, two of these mRNA species (L and F types) are ubiquitously expressed, whereas the third one (M type) is muscle-specific (4-6).

We have demonstrated that the human genomic regions upstream from exons F and L are able to drive autonomous transcription (7, 8). We found that the basal transcriptional activity of pF is regulated by several binding sites for the ubiquitous factor Sp1 (7). Binding of nuclear trans-acting factors AP-1 and NF-1 in the 384/262 region increases pF transcriptional activity (7, 9).

The pL promoter is regulated by both positive and negative cis-elements (8). Within this promoter, we detected a negative regulatory element (hAldA-NRE)¹ and a protein factor from Hep3B cells that binds to this element. This complex down-regulates to one-fifth the transcription of the distal promoter in several cell types and furthermore regulates transcription of other cellular genes (10).

In rodent tissues, only two mRNAs, the ubiquitous AH type, corresponding to the human F type, and the muscle-specific type (M type), have been characterized (11, 12). We found that a third species, which corresponds to human aldolase A L-type mRNA, is expressed and correctly processed in rodent cell lines and that its expression is increased during differentiation and is associated with cell-growth arrest (13).

Here we report that the modulation of murine L-type mRNA expression is transcriptionally controlled and differentially regulated in proliferating and confluence-arrested cell populations. Within the murine distal promoter, we identified a negative regulatory element (mAldA-NRE), homologous to the human aldolase A silencer, that is recognized by a protein factor from NIH3T3 cells of ~97 kDa. This protein binds to the mAldA-NRE to form a complex that is present at higher levels in proliferating cells than in growth-arrested cells. Preliminary mutational analysis of the AldA-NRE suggests that at least two target sequences are required for specific protein binding and full silencer function.

MATERIALS AND METHODS

Mouse NIH3T3 fibroblasts were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (FCS). The medium was changed every 2-3 days, and the cultures were maintained strictly subconfluent. Cells were allowed to grow exponentially for 3 days, at which time the medium was changed to Dulbecco's modified Eagle's medium supplemented with 0.5% FCS. Cells were maintained in 0.5% FCS for 3 days to achieve quiescence and then stimulated with fresh medium containing 20% FCS for the indicated times. With this procedure, we found that 80-90% of the cells were arrested in the G₀/G₁ phase. Concentration of gas1 mRNA, a growth-arrested specific gene (14), and cytofluorometric measurements confirmed cell synchrony. Actinomycin D (5 µg/ml; Sigma) or cycloheximide (50 µg/ml; Sigma) was added to the cells at the times indicated.

Total RNA was isolated by the method of Chomczynski and Sacchi (15). RNase protection assays were performed as described (13). The aldolase A L-type probe was obtained by transcribing with T7 RNA polymerase, in the presence of [-³²P]CTP (400Ci/mmol; Amersham Corp.), a mouse genomic 180-bp fragment (including 49 bp of the 3-end of exon L1 and the 71 bp of exon L2) cloned in the pGem3Z vector (Promega). The protected fragments generated from this probe are 118 and 69 bp long. The hybridization aldolase A F-type probe was obtained by transcribing a mouse genomic fragment of 350 bp (which included the entire exon F) cloned in the pGem3Z vector. This probe generated two protected fragments of 141 and 95 bp. A 75-bp protected fragment of rat -actin cDNA was used as an internal control. In Northern blot experiments, 10 µg of total RNA was separated by electrophoresis on 1.5% formaldehyde-agarose gel and then transferred onto a Nytran membrane (Schleicher & Schuell). Filters were baked at 80 °C and then hybridized for 16 h at 65 °C in Church's buffer (7% SDS, 0.5 M NaH₂PO₄, and 1 mM EDTA). After washing for 45 min in 50 mM NaH₂PO₄ and 1% SDS, filters were autoradiographed with x-ray film (Eastman Kodak Co.). The hybridization probes used as controls were a cDNA fragment of mouse gas1 and a specific DNA fragment of 28 S RNA labeled with [-³²P]dATP (3000 Ci/mmol; Amersham Corp.) using a random priming kit (Promega) (specific activity of ~10⁹ cpm/µg).

Nuclear protein extracts from NIH3T3 cells were prepared according to Dignam et al. (16) with some modifications. Briefly, cellular pellets were obtained in cold phosphate-buffered saline and resuspended in solution I (10 mM Hepes, pH 7.9, 10 mM KCl, 1.5 mM MgCl₂, 0.1 mM EGTA, 0.5 mM phenylmethylsulfonyl fluoride, and 0.1 mM dithiothreitol). Cells were broken 10-fold through a 1-ml syringe with a 25-gauge needle. The nuclear pellet was obtained by centrifugation for 30 s in an Eppendorf centrifuge at 4 °C and resuspended in solution II (solution I plus 400 mM NaCl and 5% glycerol and without KCl). The pellet was swollen for 30 min at 4 °C in solution II and then centrifuged at 14,000 rpm at 4 °C in an Eppendorf centrifuge for 45 min. The supernatant containing nuclear extracts was aliquoted and stored at 80 °C. Gel shift assays were performed as described (8).

The plasmid construct clones AldAL555, RBP9-CAT+AldANRE1 (where RBP is retinol-binding protein), and RBP9-CAT+AldANRE2 are described elsewhere (10). The clones AldAL555MutN1 and AldAL555MutN2 were obtained from clone AldAL555 using the MutN1 and MutN2 oligonucleotides, respectively, in accordance with the instructions provided with the ExSite polymerase chain reaction-based site-directed mutagenesis kit (Stratagene).

Subconfluent NIH3T3 cells were transfected with 20 µg of supercoiled plasmid DNA using the calcium phosphate precipitation method (17). After transfection, cells were cultured for 72 h in Dulbecco's modified Eagle's medium with 10% FCS to maintain them in their logarithmic phase of growth or were allowed to grow in Dulbecco's modified Eagle's medium containing 10% FCS for 12 h and replaced with serum-free medium for another 48 h to force them into a quiescent state. The cell extracts were assayed for CAT activity by thin-layer chromatography using [¹⁴C]chloramphenicol as a substrate (18). Transfection efficiency was normalized against -galactosidase activities.

Equal amounts (40 µg) of nuclear extracts from growing and serum-deprived NIH3T3 cells were loaded onto 10% SDS-polyacrylamide gel and electrophoresed at 25 mA for 1 h. The gel was soaked for 30 min in transfer buffer (25 mM Tris, 250 mM glycine, and 20% methanol) and then transferred onto a nitrocellulose filter (Amersham Corp.) using the Bio-Rad transfer cell system. The filter was blocked for 2 h at room temperature in a solution containing 5% nonfat dry milk and binding buffer (50 mM Hepes, pH 7.9, 50 mM KCl, 5 mM MgCl₂, 1 mM EDTA, 1 mM dithiothreitol, and 5 µg/ml salmon sperm DNA). The filter was washed for 15 min at room temperature in binding buffer, and binding was carried out at room temperature in the same buffer plus 0.25% nonfat dry milk with at least a 10⁶ cpm/ml concentration of the appropriate labeled probe. The oligonucleotides used in the Southwestern experiments were as follows: M3 oligonucleotide, CTGCCAGAGCCTCAACTGCCTCTGTTTCGAGATC; Neg3 oligonucleotide, CTGCCGAGACCTCAACTGTCTCTGCTTCGAGATC; Neg1 oligonucleotide, TCCCCTTAGAGAGCAACAGACGTGT; MutN1 oligonucleotide, TCCCCTTCCGCGGCAACAGACGTGT; and MutN2 oligonucleotide, TCCCCTTAGAGAGCGGCAGACGTGT. After three 5-min washes in binding buffer, the filter was dried and placed on x-ray films for visualization.

RESULTS

We have previously demonstrated that the expression of L-type mRNA in cultured murine cells increases during the growth-arrested state (13). We have extended this work to investigate whether the increased expression is cell cycle-regulated and whether the expression of F-type mRNA is similarly regulated.

We performed RNase protection analysis to examine aldolase A L-type mRNA expression in proliferating or serum-starved mouse NIH3T3 fibroblasts. Total RNA was isolated from NIH3T3 cells in exponential growth, in quiescence, and at various times after serum restimulation. The antisense RNAs used as probes in the RNase protection assay are shown in Fig. 1: riboprobe L includes mouse exons L1 and L2, whereas riboprobe F contains the entire mouse exon F. Using riboprobe L, we detected two protected fragments of 69 and 118 nucleotides, which correspond to different transcription initiation sites of mouse L-type mRNA (Fig. 2A). The hybrid band, corresponding to the most abundant 69-nucleotide transcript, was much more intense in serum-starved cells (Fig. 2A, lane 2) than in proliferating cells (lane 1). Densitometric scanning showed that the concentration of L-type mRNA increased ~8-fold during quiescence (data not shown). When the serum-starved NIH3T3 cells were stimulated to reenter the cell cycle by the mitogenic action of the serum, L-type mRNA expression was markedly repressed already after 2 h of serum stimulation (Fig. 2A, lane 3). The 118-bp mRNA, when detectable, also increased in serum-starved cells (Fig. 2A, lane 2) compared with proliferating cells (lane 1) and decreased considerably after 2 h of serum stimulation (lane 3). Both mRNA (118 and 69 bp) levels progressively decreased up to 12 h after serum stimulation (Fig. 2A, lanes 4-6), and as demonstrated by densitometric scanning (data not shown), they slightly increased between 12 and 20 h of serum restimulation, when the cells, although not completely synchronized, went through the cell cycle. The modulated expression of 118- and 69-bp transcripts suggests that L-type mRNAs in NIH3T3 fibroblasts are modulated in a cell cycle-dependent manner. To determine whether the arrest of cellular proliferation affects F-type mRNA expression, we performed RNase protection analysis using riboprobe F (see Fig. 1). Two specific bands of 95 and 141 nucleotides, corresponding to the major transcription initiation sites of mouse F-type mRNA, were detected (Fig. 2B). Identical levels of F-type mRNA expression were observed in proliferating cells (Fig. 2B, lane 1), in serum-starved NIH3T3 cells (lane 2), and in cells refed for 2, 4, 6, 8, 12, 16, 20, and 24 h (lanes 3-10). In this case, there was no difference between proliferating and growth-arrested cells or after serum restimulation. An antisense fragment complementary to rat -actin mRNA, which gives a hybrid band of 75 nucleotides, served as a internal control of the amounts of RNA used in each sample.

[View Larger Version of this Image (12K GIF file)]

-act.

[View Larger Version of this Image (55K GIF file)]

Northern blot analysis with the growth-arrested gas1-specific probe was used as a control of cell quiescence after serum deprivation (Fig. 2C). The same filter was hybridized to a specific 28 S probe to ensure comparable RNA loading (Fig. 2D).

To examine whether the induction of L-type mRNA expression in growth-arrested cells is due to increased mRNA stability, we analyzed the expression of L-type mRNA in proliferating and serum-starved NIH3T3 cells after treatment with actinomycin D, a specific inhibitor of RNA polymerase II. Total RNA from proliferating and serum-starved NIH3T3 cells, extracted at different times after the addition of actinomycin D to the culture medium, was analyzed in RNase protection experiments using riboprobe L (Fig. 3). L-type mRNA levels in proliferating NIH3T3 cells were unchanged up to 24 h of actinomycin D treatment (Fig. 3A, lanes 1-6), whereas they decreased substantially in serum-starved NIH3T3 cells after 14 and 24 h of treatment (Fig. 3B, lanes 5 and 6). In contrast, the gas1 transcript decreased significantly within 3 h in both proliferating and serum-starved cells (Fig. 3, C and D). The actinomycin D treatment of the proliferating and serum-deprived cells affected mainly the 69-bp transcript. These results indicate that the increased levels of L-type mRNA observed in serum-starved cells (Fig. 2A, lane 2) are not due to increased mRNA stability, thus indirectly suggesting a transcriptionally regulated control.

-act.

[View Larger Version of this Image (46K GIF file)]

To determine whether de novo protein synthesis is responsible for the modulation of L-type mRNA expression during the growth state, we performed RNase protection assays on total RNA from proliferating NIH3T3 cells treated for 4, 8, and 12 h with cycloheximide (Fig. 4A, lanes 2-4). The levels of L-type mRNA transcripts increased significantly after 4 h of cycloheximide treatment (Fig. 4A, lane 2), with the increases being more evident for the most abundant species (69 nucleotides long). Cycloheximide treatment, even after 8 h, did not affect the F-type transcripts (Fig. 4B, lanes 2 and 3). These results indicate that the expression of F- and L-type mRNAs is regulated by different transcriptional mechanisms. The decreased mRNAs in Fig. 4 (A and B, lanes 4) are probably caused by cycloheximide-induced toxicity. Taken together, the experiments with actinomycin D and cycloheximide suggest the existence of a protein that acts as a transcriptional repressor of L-type mRNA synthesis in proliferating NIH3T3 cells.

-act.

[View Larger Version of this Image (51K GIF file)]

We have previously demonstrated that a negative regulatory cis-element (hAldA-NRE) within the distal promoter (pL) of the human aldolase A gene is involved in the negative regulation of pL transcription (10). To determine if the different L-type mRNA levels expressed in growth-arrested and proliferating NIH3T3 cells result from a negative transcriptional mechanism that involves specific DNA-protein interactions, we performed transient transfection experiments in mouse cells using previously described human recombinant constructs (10) (Fig. 5). The transcriptional activity of the AldAL555 clone, which includes the hAldA-NRE plus exons L1 and L2 fused to the CAT transcriptional unit, was much lower in proliferating NIH3T3 cells (6.4%) than in growth-arrested cells (12.1%). This indicates that starved cells lack a negative transcriptional control.

[View Larger Version of this Image (17K GIF file)]

Two clones in which the AldA-NRE was cloned in opposite directions (RBP9-CAT+AldANRE1 and RBP9-CAT+AldANRE2) upstream from the basal promoter of the human retinol-binding protein gene fused to the CAT gene showed a reduced transcriptional activity in proliferating NIH3T3 cells (1.5 and 7.5%, respectively) compared with the levels obtained with the RBP9-CAT clone without the AldA-NRE (22.6%). In contrast, the RBP9-CAT+AldANRE1 and RBP9-CAT+AldANRE2 clones showed increased levels of CAT transcriptional activity in growth-arrested cells (5.1 and 15.5%, respectively). The transcriptional activity of the RBP9-CAT clone alone was the same in both proliferating and starved cells. A pGM2CAT clone containing 500 bp of the Rous sarcoma virus promoter region fused to the CAT gene (not shown in Fig. 5) was used as a positive control of the transcription assay (taken as 100% activity) in both culture conditions. These results indicate that also in mouse cells, as already demonstrated in human cells (10), there is a negative nuclear protein able to repress transcription of at least two different promoters and that the repression observed in proliferating cells was abolished when cells were cultured in serum-deprived medium.

The human and mouse silencer sequences within the L promoter contain a highly homologous fragment from 430 to 338 bp upstream from exon L1 (Fig. 6). Within this fragment, we previously identified two protected DNA cis-elements (called Neg1 and Neg3) that are specifically recognized by the same nuclear proteins extracted from a human hepatoma cell line (Hep3B) (10). Here, an oligonucleotide (M3) based on the murine sequence and homologous to human Neg3 (see boxed sequence in Fig. 6) was used in gel retardation experiments together with nuclear extracts from proliferating and starved NIH3T3 cells. A retarded DNA-protein complex was found in both proliferating (Fig. 7A, lane 2) and growth-arrested (lane 3) NIH3T3 cells, although the interaction was diminished in growth-arrested cells. When growth-arrested cells were refed with 20% FCS for different times (2, 4, 6, 8, 12, 16, 20, and 24 h) and allowed to reenter the cell cycle, the binding activity appeared to be cell cycle-regulated. This finding is shown by the time-dependent peak-and-trough profile of the DNA-protein complex after serum restimulation, i.e. an increase after 2-12 h of restimulation (Fig. 7A, lanes 4-8) followed by a decrease after 16 and 20 h (lanes 9 and 10). A bar graph demonstrating the intensity of the complex in each lane is shown in Fig. 7B. A retarded complex of identical electrophoretic mobility was observed with the human Neg1 and Neg3 oligonucleotides and nuclear extracts from NIH3T3 cells (data not shown). These results demonstrate that a similar nuclear factor present in both human and murine cells recognizes the same sequence motif and binds to it.

[View Larger Version of this Image (41K GIF file)]

[View Larger Version of this Image (40K GIF file)]

To characterize the nuclear protein that binds the AldA-NRE, we conducted Southwestern experiments with proliferating and serum-starved NIH3T3 nuclear extracts (Fig. 8). Using as probe the mouse M3 oligonucleotide, a specific complex of ~97 kDa appeared in both extracts. The DNA-protein interaction was clearly stronger in proliferating NIH3T3 cells (Fig. 8A, lane 1) than in starved NIH3T3 cells (lane 2). The same result was obtained using the human Neg1 oligonucleotide as probe (Fig. 8B, lanes 1 and 2). To explore DNA sequence requirements for both the binding activity and the silencer function of the AldA-NRE, two mutant oligonucleotides were designed and analyzed by Southwestern assay and by transient transfections. In the MutN1 oligonucleotide, the AGAGA motif was replaced by a CCGCG sequence, and in the MutN2 oligonucleotide, the dinucleotide AA was mutated to GG (Fig. 8C). In Southwestern experiments, both MutN1 and MutN2 displayed a significant decrease in binding activity (Fig. 8B, lanes 3 and 4 and lanes 5 and 6, respectively) compared with Neg1 (lanes 1 and 2). The binding activity was also reduced when the MutN1 and MutN2 mutant oligonucleotides were used as probes in the electrophoretic mobility shift assays (Fig. 9, A and B). Furthermore, both MutN1 and MutN2 showed a higher binding activity in proliferating cells (Fig. 9, A and B, lanes 1) than in growth-arrested cells (A and B, lanes 2). Fig. 9C shows the competition of the wild-type complex with unlabeled Neg1 oligonucleotide (lanes 2 and 4 versus lanes 1 and 3, respectively, in both proliferating and starved nuclear extracts) and the competition with increasing concentrations of unlabeled MutN1 (lanes 6-8) and MutN2 (lanes 9-11) mutant oligonucleotides. The MutN1 oligonucleotide shows a lower binding activity than MutN2 (Fig. 9, A and B). Accordingly, it is a less efficient competitor (Fig. 9C, lanes 6-8) compared with MutN2 (lanes 9-11) for Neg1 complex formation.

[View Larger Version of this Image (20K GIF file)]

[View Larger Version of this Image (78K GIF file)]

In transient transfection experiments (Fig. 8C), the mutations in the AGAGA motif (MutN1) or in the dinucleotide AA (MutN2) caused a loss of silencer function. In fact, the CAT activity increased to 18.9% in cells transfected with the AldAL555MutN1 clone and to 8.6% in cells transfected with the AldAL555MutN2 clone compared with cells transfected with the wild-type AldAL555 clone (4.6%). The CAT activities are normalized to the value of a positive control pGM2CAT clone taken as 100%. The transcriptional activity of the AldAL555 clone observed in NIH3T3 cells is lower than that found in Hep3B cells (10).

These data indicate that the human Neg1 and mouse M3 oligonucleotides bind the same 97-kDa protein and that their binding activity is higher in proliferating cells than in serum-starved cells. Mutations in the hAldA-NRE (MutN1 and MutN2) cause a reduced silencing function (detected by transient transfections) and a diminished DNA-protein binding activity (revealed by Southwestern and electrophoretic mobility shift assays). The binding activity is differently affected by the mutations in the AGAGA (MutN1) and AA (MutN2) motifs contained in the Neg1 site. In fact, the highly reduced binding activity of MutN1 (compared with MutN2) was responsible for the higher transcriptional activity (Fig. 8, B and C).

DISCUSSION

Regulation of the aldolase A gene involves multiple cis-elements located within the promoter regions upstream from leader exons L1, M, and F that interact with several nuclear proteins (7-11). The differential expression of various aldolase A mRNAs (L, M, and F types) is controlled by an interplay of interactions between DNA and proteins according to cell and stage specificity (20). We have previously identified a negative regulatory element (hAldA-NRE) within the human aldolase A distal promoter (pL) that acts as a true "silencer" because it represses the transcription in its own configuration and the transcription of a reporter gene driven by a heterologous promoter (10). We have also demonstrated that the third promoter (pL) within the aldolase A gene exists also in rat and mouse cells and that it is regulated by transcription factors in a proliferation/quiescence-dependent manner (13).

We have now identified within the murine distal promoter a cis-element (mAldA-NRE), highly homologous to the human silencer, that binds a nuclear protein from growing NIH3T3 cells, thus down-regulating transcription of the L-type promoter. Gel retardation assays and Southwestern experiments indicate that the binding efficiency of the repressor-silencer complex is greater in proliferating NIH3T3 cells than in serum-deprived cells. These data suggest a mechanism of cell cycle-dependent regulation. The down-regulation exerted by the murine repressor is removed when proliferating NIH3T3 cells are treated with cycloheximide (i.e. absence of active protein synthesis), and thereafter, the distal promoter region (pL) is transcribed already 4 h after the treatment at levels comparable to those of resting cells. Transcriptional regulation of the aldolase A L-type promoter gene in proliferating and growth-arrested NIH3T3 cells was conclusively demonstrated by transient transfections. Mutations in the hAldA-NRE showed that at least two DNA motifs strongly inhibited the silencer activity in proliferating cells. (Fig. 8C). The same mutated sequences display a reduced binding activity for the repressor protein in Southwestern experiments (Fig. 8B) and in electrophoretic mobility shift assays (Fig. 9), thus suggesting that these nucleotides play a role in conferring a silencer-competent DNA structure and in binding to silencer protein.

In the human L promoter, we demonstrated that an AGAGA motif is the target for the binding with a nuclear factor from Hep3B cell extracts (10). In the present study, we show that this DNA motif plays an important role in binding and silencing functions also in murine NIH3T3 cells. Furthermore, we demonstrate that nucleotides AA, adjacent to the GA-rich motif, are required for silencing activity.

Taken together, these results indicate that (i) a nuclear protein present in mouse cells recognizes the sequences contained in the murine and in human AldA-NREs, (ii) the AldA-NRE is a more general silencer because it represses also a heterologous promoter, and (iii) the mAldA-NRE is involved in the cell cycle-dependent modulation of aldolase A L-type promoter activity. These observations raise the possibility that a negative trans-acting factor that binds to the murine and human AldA-NREs is a nuclear protein involved in a general mechanism of gene regulation during alternating stages of the cell cycle.

It has been suggested that negative regulatory elements, not yet completely characterized, play a role in modulating transcription of the aldolase A gene, particularly as regards the muscle-specific promoter in both the murine and human gene systems (19-21). In the mouse aldolase A gene, transient transfection experiments have demonstrated that a negative element, the muscle sequestering element, selectively sequesters the M promoter from nonspecific stimulatory activities. This indicates that the muscle sequestering element plays a pivotal role in the maintenance of muscle-specific mRNA expression (19). Furthermore, studies of the expression of the human aldolase A gene in transgenic mice indicated that removal of a putative negative element increases pM activity in fast muscles (20). This suggests that these sequences prevent constitutive activation of pM by enhancer-like sequences, located upstream from the proximal promoter (20). The above-described negative regulatory elements in the mouse and human aldolase A genes are located in the same genomic region of the AldA-NRE. In the present study, the AldA-NRE exerted its silencing effect on the modulation of a different promoter, i.e. on pL transcription during cell cycle stages. It has not yet been investigated if the murine and human AldA-NREs affect also the transcription driven by the muscle-specific promoter (pM). Should the AldA-NRE be the same cis-negative element reported by Stauffer and Ciejek-Baez (19) and by Salminen et al. (20), it is feasible that the interaction between the AldA-NRE and a nuclear protein is involved in the events that lead to the muscle differentiation program. Be that as it may, the interaction between a repressor protein and the human and mouse AldA-NREs reported in this study is the first well characterized case of negative control in aldolase A gene transcriptional activity. Moreover, by mutagenesis analysis, the specific sequences within the AldA-NRE that mediate the negative transcriptional effect have been defined.

A negative regulatory transcriptional element (reducing module), located upstream from exon M (clone 1066/731), has been found to attenuate the proximal promoter (AH1)-driven transcription in the rat thymocyte aldolase A gene (21). However, this negative "reducing module" is located in a position completely different from that of the AldA-NRE.

Transcriptional repression is involved in a wide variety of tissue-specific (22-24), developmental (25-27), cell cycle (28-30), and signal-responsive regulatory events (31-33). The specific protein that binds the AldA-NRE shows different binding affinity during the cell cycle. This protein could belong to a repressor family that is itself modulated during the cell cycle. The eukaryotic cell cycle consists of the continuous and discontinuous processes that are required for cell growth and proliferation. Most genes are expressed at roughly constant rates throughout the cell cycle, and thus, their transcription falls into the continuous class of cell cycle processes. A small number of genes show a discontinuous pattern of expression during the cell cycle. This group of genes includes those encoding products involved in DNA metabolism (thymidine kinase and dihydrofolate reductase), some structural proteins, components of the basic cell cycle regulatory machinery (cyclins and cyclin-dependent kinase), and transcription factors (E2F1, B-myb, and Jun), which themselves confer cell cycle-regulated expression to target genes. The isolation and cloning of the protein that binds the AldA-NRE will shed light on the molecular mechanisms that regulate its expression and the biological role played by this protein during cell cycle-regulated gene transcription.