To understand the physiological function of the mammalian heterotrimeric CCAAT binding factor CBF, also known as NF-Y, we have generated a conditional Cbf-b mouse mutant by introducing loxP sites in the murine Cbf-b/Nf-ya gene. Controlled expression of Cre recombinase deletes the gene in vivo, which leads to a loss of DNA binding by the CBF complex and hence CBF-mediated transcription. Deletion of both Cbf-b alleles causes early embryo lethality, indicating that CBF activity is essential for early mouse development. In primary cultures of mouse embryonic fibroblasts, conditional inactivation of CBF results in a block in cell proliferation and inhibition of S phase or DNA synthesis, which is followed by induction of apoptosis. We conclude that the CBF transcription factor complex is essential for cell proliferation and viability.

The mammalian CCAAT motif-binding factor, CBF (also called nuclear factor-Y or NF-Y) is an evolutionarily conserved transcription factor present from yeast to human. It consists of three different subunits, CBF-A, CBF-B, and CBF-C, which are all required for formation of a specific CBF-DNA complex. Both the CBF-A and CBF-C subunits contain a histone-fold motif and interact with each other to form a CBF-A/CBF-C heterodimer. The A/C heterodimer then interacts with a 21-amino acid stretch in CBF-B to form the heterotrimeric CBF transcription factor. Thus, the absence of any one of the CBF subunits results in loss of binding of the CBF complex to DNA and CBF-directed transcription (1, 2).

The CCAAT motif is present in promoters of many mammalian genes including genes expressed in specific cell types as well as genes regulated during the cell cycle, such as topoisomerase IIα, cyclin B1, CDC25C, E2F1, CDC2, and thymidine kinase genes (1, 2, 3, 4, 5, 6). Even though CBF activity is found to be present in all mammalian tissues, the genes that are regulated by this transcription factor complex in vivo are still unknown. The physiologically relevant target genes, whose transcription is highly dependent on CBF activity, can only be identified using an animal model where endogenous CBF activity can be abrogated in a specific tissue or at a specific developmental stage.

Previously, expression of the dominant negative CBF-B mutant in mouse NIH 3T3 fibroblasts caused the cells to grow slower, with a modest increase in the time required for the cells to complete the S phase of the cell cycle (7). Because transformed cell lines, such as 3T3 fibroblasts, have aberrant expression of many proteins, the effects of dominant negative CBF-B expression in these cells could be masked or accentuated, which would not allow for a clear understanding of the physiological role of CBF activity.

To establish the role of CBF-mediated transcription in vivo, we generated a conditional allele of the gene for the CBF-B subunit, which could be deleted using the Cre-lox system. Using this mouse model, this study clearly demonstrates that CBF-dependent transcription is essential during early mouse development. Conditional deletion of both Cbf-b alleles in primary cultures of MEF1 cells caused a complete block in the progression of the cells into S phase. Subsequently, Cbf-b null cells underwent apoptosis. The results described in this study clearly show that CBF-mediated transcription is required for cell proliferation and viability.

Construction of the Cbf-bneo+flox Targeting Vector, Generation of Btargeted Allele in ES Cell Clones, and Southern and PCR Analysis of Mouse Tail DNA.

The seventh intron of the murine Cbf-b gene was cloned by amplification of genomic DNA using published sequences (8). The internal probe (Fig. 1,A) was used in screening a λ FIXII library comprising SV129 genomic DNA. The unique EcoRV site in the second intron was altered to a BamHI site, and a loxP sequence (I) was cloned at the 3′ end of this restriction site by ligating an oligonucleotide cagcctcccgggggatccataacttcgtataatgtatgctatacgaagttatcccgggtgttga (Btargeted allele). The neomycin resistance gene, flanked by loxP sites II and III from pPGKneopAloxP(9), was cloned into the unique SphI site in the eighth intron with a BamHI site also introduced at the 3′ of the third loxP site. The human thymidine kinase gene was cloned 5′ region to the SpeI site, as shown in Fig. 1 A.

Using published methods, the targeting vector was introduced into ES cells (9). Southern analysis of the BamHI and KpNI-digested genomic DNA, using the 5′ and 3′ external probes, was used to confirm homologous recombination in selected ES clones. Individual ES clones were used to obtain chimeric mice, which were then crossed with C57BL6 mice to obtain animals that have the Btargeted allele.

Genomic DNA from mouse tails, cells, or embryonic tissue was isolated by digestion in a buffer containing 0.5 μg/ml proteinase K, NaCl, EDTA, and SDS at 55°C overnight. PCR primer 1 (gtaagtcaggctccaggg), in intron 2 and 5′ of the EcoRV site of the CBF-B gene, and primer 2 (aggcaaggcagatttaggaaggtc), which is in intron 2 and 3′ of the EcoRV site, were used to distinguish between the 200-bp Bwt allele and the 250-bp Bflox allele. Primer 1 and primer 3 (gggttgtcaggatgttcgcag), in intron 8 and 3′ of the SphI site, were used to amplify the 400-bp product to genotype a Bdel allele.

MEF Cultures.

We used 13.5 d.p.c. Bflox/del and Bwt/wt embryos to isolate MEFs using previously published procedures (10). MEFs were used at passage 4 and 5 for the experiments.

AdCre Infection of MEFs and Detection of β-Gal Activity in ROSA MEFs.

AdCre was a kind gift from Dr. Frank Graham (McMaster University, Ontario, Canada), and the adenovirus was amplified by the Adenovirus Core facility at our institute. MEFs (4 × 105), plated on 10-cm dishes, were rinsed with PBS containing Mg2+ and Ca2+. The appropriate amount of AdCre for a MOI of 10, in 1 ml of PBS, was added to the cells and placed at 37°C for 70 min with rocking every 15 min, after which PBS was removed and replaced with 10 ml of fresh MEF medium.

Forty-eight h after infection, mock-infected and AdCre-infected ROSA MEFs were stained for β-gal activity using previously published procedures (9).

Cell Synchronization, BrdUrd Labeling, Immunofluorocytochemistry, and Confocal Microscopy.

Aphidicolin was added at 2 μg/ml to MEFs plated at 4 × 105 cells/10-cm dish in regular MEF medium containing 15% FCS for 40 h. Cells were labeled with 10 μm BrdUrd for 1 h, fixed with paraformaldehyde, and processed for immunofluorescent detection with anti-BrdUrd antibody using a FLUOS kit following the suggested protocol (Roche Molecular Biochemicals).

MEFs grown on gelatin-coated coverslips were fixed for 10 min with 1% paraformaldehyde, permeabilized in 0.5% Triton X-100 for 30 min, and blocked in 80 mm PIPES (pH 6.8), 5 mm EGTA, 2 mm MgCl2 (PEM buffer) containing 1.5% nonfat dry milk. A polyclonal antibody directed against the COOH-terminal region of CBF-B was used at 1:200. Mouse anti-phosphorylated serine 10 histone-3 (5-598; Upstate Biotechnology, Lake Placid, NY) and anti-MPM-2 antibody (Upstate Biotechnology) were used at the recommended dilution. Secondary antibodies conjugated to fluorescein (Santa Cruz Biotechnology, Santa Cruz, CA) were used at a 1:200 dilution. Nuclear DNA was stained with propidium iodide at 0.1 μg/ml in PBS.

Apoptosis Assays.

TUNEL staining to determine apoptotic cells was done with cells grown on coverslips, and cells were analyzed at 96, 120, and 144 h after infection using the Apodirect kit (Serologicals Co.) according to the recommended protocol.

cDNA Arrays.

Total RNA was isolated from cells in culture using the RNeasy kit (Qiagen) using the manufacturer’s recommended protocol from uninfected and infected MEFs at 112 and 120 h after infection. Individual apoptosis gene array 1 and 2 filters and mouse cell cycle array filters (Superarray, Inc.) were hybridized to 32P-labeled reverse-transcribed total RNA (5 μg) from uninfected and infected Bwt/wt and Bflox/del MEFs, respectively, according to the manufacturer’s protocol. Radioactive signal intensities were analyzed on a PhosphorImager (Molecular Dynamics). For each filter, the background signal (puc18) was subtracted, the intensity of the positive control (GAPDH) signal was taken as 100%, and the signal intensity of all other genes was calculated as a percentage of the GAPDH signal. Relative expression levels of 135 genes were calculated, and the names of the genes are listed online2 in Mouse Apoptosis gene array 1 and 2 and Mouse Cell cycle, Q series (Superarray, Inc.).

Real-Time PCR Analysis.

Five μg of total RNA were reverse transcribed using random hexamers and Superscript II enzyme (Invitrogen) following the manufacturer’s suggested protocol. Relative levels of Cbf-b mRNA, eukaryotic elongation factor 1G mRNA, and caspase-2 mRNA were determined in the four samples of RNA isolated at 112 h from uninfected and infected Bwt/wt and Bflox/del MEFs, respectively. The Roche Light Cycler instrument and the LightCycler Fast Start DNA Master SYBR Green I kit were used according to the manufacturer’s protocol (Roche Molecular Biochemicals). The 5′ and 3′ primers used for mouse EEF IG that amplify a 183-bp product are tccaatgaggacaccctctc and agcaaaggcattcttcctca, respectively. The 5′ and 3′ primers for CBF-B that amplify a 210-bp product are tggtgcaagtcagtggaggccagctta and atgatctgctgggtttgacc, respectively. These primers are within the region of the Cbf-b gene targeted for deletion. The 5′ and 3′ primers for caspase-2, which amplify a 178-bp product, are tgacaatgctaactgtccaa and gtctcatcttcatcaactcc, respectively.

Generation of a Conditional Bflox Allele.

To disrupt the endogenous mouse gene for the CBF-B subunit, we used the Cre recombinase-loxP system (11). The 346-amino acid-long murine CBF-B polypeptide is encoded in nine exons, of which the DNA binding region and the subunit interaction domains are encoded in exons 7 and 8 (8). To create a conditional Cbf-b allele (Btargeted), which would remove all of the functional domains of CBF-B, loxP sites flanking exon 3 and exon 8 were inserted in the targeting vector. Targeted ES clones were used to generate chimeric mice, and mice heterozygous for the targeted Bflox+neo allele were obtained. Crosses with CMVCre C57BL6 transgenic strain, in which a cytomegalovirus promoter directed Cre expression during early mouse embryogenesis, generated all three possible recombinant B alleles. Recombination between loxP sites II and III created the Bflox allele, which left the B gene flanked by two loxP sites for future tissue-specific targeted deletion of the Bflox gene (Fig. 1 B). Recombination between loxP sites I and III in the Bflox+neo allele created a Bdel allele, which removed both the functional DNA binding and transactivating regions of the protein. Recombination between loxP sites I and II deleted the B sequences and left the neo gene. This Bdel/neo+ allele, which was identified by PCR amplification of the neo sequence, was not used.

To identify the different alleles, Southern analysis of BamHI-digested genomic DNA was performed with two probes (Fig. 1 C). The 5′ probe identified the 9-kb Bwt (Lanes 1–4) and the 7-kb Btargeted alleles, respectively (Lanes 1–3). In addition, an internal probe, which lies within the sequences targeted for deletion, identified the Bwt, Bflox, and the Bflox+neo alleles. Indeed, addition of the neo cassette in the targeted locus introduced a BamHI site at the 3′ end of the neo gene, which persists even after the neo sequence is deleted. Hence, BamHI restriction fragments of 5.5, 4.5, and 3.5 kb identified Bflox+neo (Lane 3), Bwt (Lanes 1–4), and Bflox (Lane 2), respectively. Because this probe lies within Cbf-b sequences targeted for deletion, the absence of a band corresponding to a “targeted’ allele, as detected by the 5′ probe, indicated a Bdel allele (Lane 1).

Two sets of PCR amplification from mouse tail DNA were also used for genotyping, as shown in Fig. 1, D and E. PCR primers 1 and 2 amplified a 200-bp Bwt band or a 250-bp Bflox band (Fig. 1,D, Lanes 1–3). Primers 1 and 3 amplify a 400-bp band of the Bdel allele (Fig. 1 E, Lanes 2 and 3).

Heterozygous animals, with one allele of Cbf-b deleted, were normal and fertile. In crosses between heterozygous mutants, no homozygous Bdel/del mutants were obtained, as early as the 8.5 d.p.c. stage of embryonic development. The distribution of the three genotypes, detected at the specific developmental stages, is shown in Table 1. These results indicate that CBF activity is essential for early mouse embryo development.

Generation of Cbf-b Null, or Bdel/del, Cells in Culture.

The role of CBF-B in cell cycle progression was tested in primary cultures of MEFs from heterozygous 13.5 d.p.c. embryos, containing one Bflox allele and one Bdel allele, that were infected with adenovirus expressing Cre recombinase (AdCre). Efficiency of virus-mediated delivery of Cre recombinase (Cre) and its subsequent enzymatic activity were tested in MEFs isolated from ROSA26R embryos, where β-gal is expressed after the removal of a stop cassette flanked by loxP sites (11). These Bwt/wt cells served as a control to rule out nonspecific effects of Cre. Around 80–90% of Bwt/wt cells were positive for β-gal activity 48 h after AdCre infection at a MOI of 1:10, whereas uninfected cells were negative. For detection of Cre-mediated recombination and deletion of the second Cbf-b allele in Bflox/del cells, Southern analysis of NdeI-digested genomic DNA was done, using the 5′ probe (Fig. 2,A). The 9-kb band corresponds to the Bwt or Bflox locus, and the 11.5-kb band corresponds to the Bdel locus (Fig. 1 B). A scan of these autoradiogram indicated that more than two-thirds of the cells in culture were B-null (Bdel/del) at 72 and 96 h after infection with AdCre. To maximize recombination and increase the percentage of B-null cells in culture, higher MOIs with AdCre were tested. Increasing the level of Cre protein had a deleterious effect on cell growth of the control Bwt/wt MEFs, probably due to nonspecific recombination as reported by other investigators (12), and was not used in this study. Because we used the optimal MOI for all our experiments, we consider that any effect in the Bflox/del MEFs that we describe here was caused specifically by the deletion of both Cbf-b genes.

Subsequent loss of CBF-B protein at 96 h after viral infection was monitored by immunofluorescence and confocal microscopy. The top row of Fig. 2,B shows the presence of CBF-B in all nuclei of Bwt/wt cells infected with AdCre. All nuclei of uninfected Bflox/del (flox cells) also show the presence of CBF-B (second row of Fig. 2 B). The bottom row shows absence of anti-CBF-B staining in six of seven nuclei at 96 h postinfection with AdCre, whereas one nucleus is still positive for CBF (white arrow). These fields, which are representative of many that were examined, reflect the results of the Southern analysis, which indicated that 70% of the Bflox/del MEFs had undergone recombination and were null for the Cbf-b gene. Loss of CBF-B protein also caused a marked decrease in CCAAT motif binding activity when tested in DNA binding experiments using nuclear extracts from these MEFs (data not shown). We concluded that the Bflox allele of Cbf-b was deleted in about 70% of the cells after adenoviral delivery of Cre recombinase, that no CBF-B polypeptide was present in Bdel/del-null cells at 96 h postinfection, and that CBF-DNA binding activity was lost in this population of cells.

Decrease in Growth Rate of the Bdel/del-Null Cells in Culture.

At 72, 96, 120, and 144 h postinfection, cells were counted to estimate growth rate in the presence or absence of CBF-B. As shown in Fig. 2 C, Bwt/wt cells continue to increase in number and undergo ∼3 rounds of doubling in the span of a 144-h culture period. Infection of the Bwt/wt cells with AdCre did not cause much difference in their growth rates when compared with that of uninfected cells. However, uninfected Bflox/del cells grew slower than uninfected Bwt/wt cells. In addition, at 72 h after infection of Bflox/del cells, there was no increase in the number of cells in culture. This absence of cell proliferation in the population of infected Bflox/del cells (mainly Bdel/del cells) at this time in culture indicated a delay or block in the cell cycle. A similar result was observed in multiple experiments using MEF cultures established from four different Bflox/del embryos. This showed that complete loss of CBF-B protein in MEFs resulted in growth arrest.

Analysis of Cell Cycle Progression of Bwt/wt, Bflox/del, and Bdel/del MEFs.

To identify cells in S phase of the cell cycle, infected or uninfected Bwt/wt and Bflox/del MEFs were BrdUrd-labeled for 1 h and processed for immunohistochemistry (Fig. 3, top panel; Table 2). At 96 h after infection with AdCre, 18% of the Bwt/wt nuclei stained positive with the BrdUrd antibody, whereas Bflox/del MEFs infected with AdCre had less than 2% of their nuclei positive for BrdUrd labeling. Thus, in the absence of CBF-B protein, there is a marked decrease in the number of cells in S phase. This could be due to a G1-S block or an inability to replicate DNA. Interestingly, uninfected Bflox/del cells, which proliferate more slowly in culture than wild-type cells, had 12% of cells in S phase compared with 25% of BrdUrd-labeled nuclei in uninfected cells with wild-type levels of CBF-B protein. Thus, heterozygous level of CBF-B protein causes a 50% decrease in S phase-positive MEFs, and B-null cells had an almost complete block in DNA synthesis.

Similarly, at 120 h after infection with AdCre, cells were analyzed for the presence of histone H3 phosphorylation at serine 10, which is a marker for M phase (13). Uninfected Bwt/wt cells and uninfected Bflox/del cells have 16% and 14% of cells positive for histone H3 phosphorylation (Fig. 3, bottom panel; Table 2). On the other hand, the Bwt/wt cells that were infected with AdCre have 12% of nuclei positive for phosphorylated histone H3. Surprisingly, around 7% of nuclei of AdCre-infected Bflox/del cells, which are mainly Bdel/del cells at this time, are positive for phosphorylated histone H3. In Bwt/wt cells infected or uninfected with AdCre, the S phase:M phase ratio is about 1.5, which indicates that there are about one-third more cells in S phase than in M phase. In contrast, in uninfected Bflox/del cells, the ratio of S phase:M phase is 0.86, indicating that fewer cells are in S phase as in M phase. Interestingly, after Bflox/del cells were infected with AdCre, the ratio of S-phase nuclei:M-phase-positive nuclei drops further to 0.28. Thus, in the population of cells where at least 70% of MEFs have no detectable CBF-B protein, even though there is a further drastic drop in cells able to enter S phase, nuclei still show positive labeling for M phase. This indicates that in the absence of CBF-B, there is an aberrant distribution of cells in specific cell cycle stages. At these times, or later, there is also no apparent increase in cell number in Bdel/del MEFs. These results support the view that in the absence of CBF-B, most cells are blocked from either entering into or progressing through S phase. In addition, some cells are also stalled in the M phase of the cell cycle or may be prematurely entering mitosis.

The slower growth of uninfected Bflox/del cells and the lower proportion of these cells in S phase compared with Bwt/wt cells could be due to CBF-B haploinsufficiency in the primary fibroblasts. To further analyze these differences, aphidicolin was used to arrest the Bflox/del and Bwt/wt cells at the G1-S boundary. After cells were released from the block, entry into S phase was assayed by BrdUrd incorporation. Interestingly, the Bflox/del cells had only 4% of cells in S phase, whereas 40% of Bwt/wt cells were progressing through the S phase at this time (Table 2). This discrepancy between the Bflox/del and the Bwt/wt cells demonstrates that the lower level of CBF-B protein compromises the ability of cells to recover after a block in DNA synthesis.

Bflox/del Cells, Infected with AdCre, Undergo Apoptosis.

Nuclei of many B-null cells had a smaller size, showing compaction of their DNA. Other nuclei were large and undergoing abnormal blebbing with formation of micronuclei. These abnormal morphologies suggested that the cells could be undergoing programmed cell death and/or mitotic catastrophe. To further document the hypothesis of apoptosis, at 96, 120, and 144 h after AdCre infection, Bwt/wt MEFs and Bflox/del cells were analyzed by TUNEL assay. At 96 h after infection, when there was already an effect on cell growth in infected Bflox/del cells, there were almost no TUNEL-positive cells in either uninfected or AdCre-infected Bwt/wt and Bflox/del MEFs (data not shown). At 120 and 144 h after infection with AdCre, only 1% of nuclei of Bwt/wt cells were positive by TUNEL (Fig. 4). Similarly, only 4% of nuclei of uninfected Bflox/del cells were positive for apoptosis. In contrast, at 120 and 144 h after AdCre infection, about 75% of Bflox/del nuclei are positive for TUNEL staining (Fig. 4, bottom three rows). The bottom row in Fig. 4 shows a higher magnification view of two nuclei that are TUNEL positive; one is a condensed nucleus, and the other has a normal size. Taken together, these results demonstrate that in the absence of CBF-B, progression through the cell cycle is inhibited, and apoptosis is initiated.

Increased Expression of Caspase-2 Gene in CBF-B-Null Cells.

Because Bdel/del cells had a block in cell proliferation and were undergoing apoptosis, we studied changes in gene expression using cell cycle-related and apoptosis-related cDNA arrays. Total RNA was isolated at 112 h after infection with AdCre from Bwt/wt and Bflox/del MEFs and reverse transcribed, and the cDNAs were hybridized to filters containing sequences of 39 apoptosis-related genes.

Expression level of the proapoptotic gene, initiator caspase-2, was increased 2-fold in CBF-B-null cells but unchanged in Bwt/wt cells either infected or uninfected with AdCre (Fig. 5) in two separate experiments. In contrast, expression of all other genes present in the apoptosis-related cDNA array was not significantly changed in all four sets of RNA samples. The increase in caspase-2 mRNA was further substantiated as a 3-fold change using real-time PCR analysis (data not shown).

Expression of Cre recombinase during early mouse embryogenesis resulted in deletion of a Cbf-bflox allele in vivo. Heterozygous mice, with deletion of one Cbf-b allele, developed normally, similar to mice with both wild-type Cbf-b alleles. However, when heterozygous mice were bred, no Cbf-b-null mouse embryos were detected, as early as the 8.5 d.p.c. stage, indicating that CBF-B is required for early embryonic development. Expression of Cre in CBF-B heterozygous mutant MEFs carrying one functional Bflox allele created B-null cells, which were unable to proliferate in culture and eventually underwent apoptosis. Inhibition of cell growth correlated with the deletion of the remaining Cbf-b allele and the loss of CBF-B polypeptide in the nucleus. In the conditions used in our experiments, there was no significant alteration of cell growth in Bwt MEFs. Thus deletion of both Cbf-b alleles caused a specific block in cell proliferation and caused cell death. We hypothesize that the early embryo lethality in mice is also due to an inability of cells to proliferate and survive.

We have consistently observed that the growth rate of heterozygous Bflox/del fibroblast cells is significantly lower than that of Bwt/wt cells in culture. This is further supported by the observation that the proportion of these cells in S phase is clearly lower than that of the Bwt/wt cells. Hence, in primary cultures of fibroblast cells, loss of half the amount of CBF-B decreases the number of cells in S phase. Furthermore, Bflox/del fibroblast cells are sensitive to cell cycle arrest by aphidicolin. On removal of aphidicolin treatment, Bwt/wt cells reentered S phase, but Bflox/del cells were unable to synthesize DNA, as ascertained by BrdUrd incorporation. Aphidicolin inhibits DNA polymerase α, δ, and ε and causes stalled replication forks, and removal of aphidicolin allows cells to restart DNA synthesis. Prolonged stalling of replication in the presence of aphidicolin causes the loss of specific MCM proteins from chromatin in Xenopus cells. These proteins are required for reinitiation of replication and are regulated stringently to avoid multiple replications of DNA within a single cell cycle (14). We hypothesize that loss of half the amount of CBF-B protein, in a specific cell type such as fibroblasts, might be responsible for the expression of lower amounts of proteins required to reinitiate replication.

Analysis of MEFs by TUNEL assay showed that loss of CBF-B induces apoptotic cell death. The time of cell death occurred after the arrest of the cell cycle because almost no cells were TUNEL positive at 96 h after AdCre infection, when entry into or progression through S phase was already inhibited. However at 144 h after AdCre infection, the majority of MEFs lacking CBF-B were positive for TUNEL staining. Thus CBF is a survival factor for mammalian cells and is required for cell cycle progression. By analyzing the expression of genes identified previously that lead to or prevent apoptosis, we found that expression of caspase-2 mRNA was increased in the absence of CBF-B but that the expression of other genes present in the array was unchanged in the time interval between the S-phase inhibition and the detection of TUNEL-positive cells. Examination of the proximal promoter region of both the mouse and human caspase-2 gene shows the presence of two CCAAT motifs within 250 bp of the transcription start site. Conservation of the putative CBF binding sites indicates that transcription of caspase-2 might be directly regulated by CBF. The increase in caspase-2 mRNA in CBF-null cells suggests that the transcription of this gene could be negatively regulated by CBF, which would require further experimental analysis. Recently, an increase in the expression of caspase-2 mRNA, specifically in response to DNA damage, was reported in a human cell line (15). Caspase-2, which is one of the initiator caspases, triggers the processing of downstream effector caspases, is localized mainly in the nucleus, is autoactivated by oligomerization and binding to its adapter protein RAIDD, and has been shown to be transcriptionally up-regulated in response to certain hormonal conditions or stress (15, 16, 17). Expression levels of other initiator caspases, which are localized in other specific nonnuclear compartments such as the endoplasmic reticulum, Golgi, and mitochondria, were not increased. We hypothesize that inhibition of the S phase or premature entry into the mitotic phase in B-null cells could cause DNA damage and result in the induction of caspase-2 mRNA.

In summary, our study establishes that an important in vivo function of CBF is to control cell proliferation, particularly entry in and progression through the S phase of the cell cycle. Although the mechanism by which CBF regulates cell cycle progression remains to be clarified, recent preliminary RNA expression analysis (data not shown) indicates the decreased expression of the Cullin family of genes as well as cyclin A2 in Cbf-b-null cells; these genes are important for progression from G1to S phase of the cell cycle (18, 19). This analysis suggests that inactivation of CBF results in changes in the expression of multiple genes. Future experiments will identify CBF-dependent genes that are important for cell proliferation and early embryogenesis.

Our mouse model will be useful in determining whether other sites of in vivo proliferation, such as skin, hair, and regenerating livers, are dependent on CBF-mediated transcription. In addition, by using an inducible Cre recombinase and a mouse model for tumorigenesis, we should be able to test whether tumor cells require functional CBF for survival.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Sankar N. Maity, Department of Molecular Genetics, Box 11, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas 77030. Phone: (713) 792-8943; Fax: (713) 794-4295; E-mail: smaity@mdanderson.org

1

The abbreviations used are: MEF, mouse embryonic fibroblast; ES, embryonic stem cells; d.p.c., days post conception; MOI, multiplicity of infection; β-gal, β-galactosidase; BrdUrd, bromodeoxyuridine; TUNEL, terminal deoxynucleotidyltransferase-mediated nick end labeling; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; EEF IG, Mus musculus elongation factor like protein mRNA.

2

www.superarray.co.

Fig. 1.

A, restriction map of the murine wild-type Cbf-b genomic locus, targeting construct, and the resultant targeted Bflox+neo allele. The relative position of the coding exons within specific restriction fragments in the genomic clone is shown. Restriction sites for BamHI (B), KpnI (K), NdeI (N), SpeI (Sp), SalI (Sa), EcoRV (E), SphI (Sph) are shown. The loxP sites are indicated as *I, *II, and III. Positions of the 5′ and 3′ external probe and internal probe and PCR primers 1–3 used in mouse genotyping are shown here. B, restriction map of the Bflox and Bdelalleles generated in vivo from the targeted Bflox+neo gene. Positions of BamHI and NdeI restriction sites are shown, which were used to genotype Bflox and Bdel alleles. C, Southern analysis of mouse genotypes. BamHI-digested mouse tail genomic DNAs were used for Southern analysis using both the 5′ external and the middle probes. D, PCR analysis of genomic DNA to detect Bwt and Bflox alleles. PCR primers 1 and 2, which flank the 5′ and 3′ of the EcoRV site in the Cbf-b genomic locus, amplify a 200-bp Bwt band and a 250-bp Bflox band. E, PCR detection of Bdel allele. Primers 1 and 3 amplify a 400-bp band only when Bdel allele is present. Location of the primers is shown in A.

Fig. 1.

A, restriction map of the murine wild-type Cbf-b genomic locus, targeting construct, and the resultant targeted Bflox+neo allele. The relative position of the coding exons within specific restriction fragments in the genomic clone is shown. Restriction sites for BamHI (B), KpnI (K), NdeI (N), SpeI (Sp), SalI (Sa), EcoRV (E), SphI (Sph) are shown. The loxP sites are indicated as *I, *II, and III. Positions of the 5′ and 3′ external probe and internal probe and PCR primers 1–3 used in mouse genotyping are shown here. B, restriction map of the Bflox and Bdelalleles generated in vivo from the targeted Bflox+neo gene. Positions of BamHI and NdeI restriction sites are shown, which were used to genotype Bflox and Bdel alleles. C, Southern analysis of mouse genotypes. BamHI-digested mouse tail genomic DNAs were used for Southern analysis using both the 5′ external and the middle probes. D, PCR analysis of genomic DNA to detect Bwt and Bflox alleles. PCR primers 1 and 2, which flank the 5′ and 3′ of the EcoRV site in the Cbf-b genomic locus, amplify a 200-bp Bwt band and a 250-bp Bflox band. E, PCR detection of Bdel allele. Primers 1 and 3 amplify a 400-bp band only when Bdel allele is present. Location of the primers is shown in A.

Close modal
Fig. 2.

A, in vivo excision of Bflox allele in Bflox/del MEFs. Genomic DNA was isolated from Bwt or Bflox/del MEFs after 72- or 96-h infection with AdCre. The DNAs were digested with NdeI for Southern analysis with the 5′ probe. B, loss of CBF-B protein in AdCre-infected Bflox/del MEFs. Cells were stained with propidium iodide for DNA. A polyclonal CBF-B antibody was used with a fluorescein-tagged secondary antibody, and the analysis was done by confocal microscopy (×40 or ×63 magnification). CBF-B protein is detected in all nuclei of Bwt cells infected with AdCre (Wt +AdCre) and of uninfected Bflox/del cells (Flox). CBF-B is detected in one of seven nuclei (indicated by white arrow) of AdCre-infected Bflox/del cells (Flox + AdCre). C, cessation of growth in Bflox/del MEFs infected with AdCre. Numbers of Bwt and Bflox/del MEFs at 0, 72, 96, 120, and 144 h after AdCre infection are shown as a bar graph. Both parts of this figure are representative of six different experiments, which showed similar results.

Fig. 2.

A, in vivo excision of Bflox allele in Bflox/del MEFs. Genomic DNA was isolated from Bwt or Bflox/del MEFs after 72- or 96-h infection with AdCre. The DNAs were digested with NdeI for Southern analysis with the 5′ probe. B, loss of CBF-B protein in AdCre-infected Bflox/del MEFs. Cells were stained with propidium iodide for DNA. A polyclonal CBF-B antibody was used with a fluorescein-tagged secondary antibody, and the analysis was done by confocal microscopy (×40 or ×63 magnification). CBF-B protein is detected in all nuclei of Bwt cells infected with AdCre (Wt +AdCre) and of uninfected Bflox/del cells (Flox). CBF-B is detected in one of seven nuclei (indicated by white arrow) of AdCre-infected Bflox/del cells (Flox + AdCre). C, cessation of growth in Bflox/del MEFs infected with AdCre. Numbers of Bwt and Bflox/del MEFs at 0, 72, 96, 120, and 144 h after AdCre infection are shown as a bar graph. Both parts of this figure are representative of six different experiments, which showed similar results.

Close modal
Fig. 3.

Analysis of cell cycle stages in Bflox/del MEFs. Top panel, at 96 h after AdCre infection, Bwt and Bflox/del MEFS were analyzed for BrdUrd incorporation (S-phase marker). Bottom panel, at 120 h after AdCre infection, phosphorylated histone H3 (M-phase marker) was detected in nuclei of MEFS by immunofluorescence and confocal microscopy. These microscopy images are representative fields of two separate experiments that showed similar results.

Fig. 3.

Analysis of cell cycle stages in Bflox/del MEFs. Top panel, at 96 h after AdCre infection, Bwt and Bflox/del MEFS were analyzed for BrdUrd incorporation (S-phase marker). Bottom panel, at 120 h after AdCre infection, phosphorylated histone H3 (M-phase marker) was detected in nuclei of MEFS by immunofluorescence and confocal microscopy. These microscopy images are representative fields of two separate experiments that showed similar results.

Close modal
Fig. 4.

Apoptosis in CBF-B-null cells. TUNEL staining of Bwt or Bflox/del MEFs was performed at 120 and 144 h after AdCre infection. Confocal microscopy images, at ×40 (rows 1–6) and ×63 (row 7) magnification, are shown. These microscopy images are representative fields of three separate experiments that showed similar results.

Fig. 4.

Apoptosis in CBF-B-null cells. TUNEL staining of Bwt or Bflox/del MEFs was performed at 120 and 144 h after AdCre infection. Confocal microscopy images, at ×40 (rows 1–6) and ×63 (row 7) magnification, are shown. These microscopy images are representative fields of three separate experiments that showed similar results.

Close modal
Fig. 5.

Apoptosis-related gene expression analysis of Bwt and Bflox/del MEFs with or without AdCre infection. At 112 h after infection, total RNA isolated from uninfected or infected Bflox/del or Bwt/wt cells was reverse transcribed and hybridized to Apoptosis array filters 1 and 2 (Superarray, Inc.). Analysis by PhosphorImager was done to estimate changes in expression levels of 39 genes (only 5 genes are shown here). The table shows the relative percentage of each signal to the respective GAPDH signal on the corresponding filter, which was taken as 100%. This experiment was repeated twice with total RNA isolated from two independent experiments using cells from different sets of embryos at passage 6 that yielded similar results.

Fig. 5.

Apoptosis-related gene expression analysis of Bwt and Bflox/del MEFs with or without AdCre infection. At 112 h after infection, total RNA isolated from uninfected or infected Bflox/del or Bwt/wt cells was reverse transcribed and hybridized to Apoptosis array filters 1 and 2 (Superarray, Inc.). Analysis by PhosphorImager was done to estimate changes in expression levels of 39 genes (only 5 genes are shown here). The table shows the relative percentage of each signal to the respective GAPDH signal on the corresponding filter, which was taken as 100%. This experiment was repeated twice with total RNA isolated from two independent experiments using cells from different sets of embryos at passage 6 that yielded similar results.

Close modal
Table 1

Distribution of embryo genotypes

The table represents the number of 8.5 d.p.c., 13.5 d.p.c., and newborn pups that are (I) B wild-type that include Bwt/wt, Bflox/wt, Bflox/flox, (II) B heterozygote that include Bwt/del and Bflox/del, and (III) B-null embryos that are Bdel/del from timed crosses between Bwt/del and Bflox/del heterozygous mice.

B wild-typeB heterozygoteB- null
Newborn 41 83 
13.5 d.p.c. 16 
8.5 d.p.c. 10 
B wild-typeB heterozygoteB- null
Newborn 41 83 
13.5 d.p.c. 16 
8.5 d.p.c. 10 
Table 2

Comparison of Bwt/wt, Bflox/del, and Bdel/del MEFs that are in either S phase or M phase, as described in Fig. 3 

For each experiment, a total of 500 nuclei were counted to determine percentage of nuclei stained for BrdUrd incorporation or phosphorylated serine 10 of histone 3. In addition, a 40-h aphidicolin treatment was used to arrest uninfected Bwt/wt and Bflox/del cells at G1-S boundary. Six h after removing the block, BrdUrd labeling, immunocytochemistry, and microscopy were used to count 500 nuclei to estimate the percentage of BrdUrd-positive nuclei. These results were validated in two separate experiments.

S phaseM phase
− aphidicolin   
Bwt/wtAdCre 25 16 
Bwt/wt+ AdCre 18 12 
Bflox/delAdCre 12 14 
Bflox/del+ AdCre 
+ aphidicolin   
B              wt/wt 40  
B              flox/del  
S phaseM phase
− aphidicolin   
Bwt/wtAdCre 25 16 
Bwt/wt+ AdCre 18 12 
Bflox/delAdCre 12 14 
Bflox/del+ AdCre 
+ aphidicolin   
B              wt/wt 40  
B              flox/del  
1
Maity S. N., de Crombrugghe B. Role of the CCAAT-binding protein CBF/NF-Y in transcription.
Trends Biochem. Sci.
,
23
:
174
-178,  
1998
.
2
Mantovani R. The molecular biology of the CCAAT-binding factor NF-Y.
Gene (Amst.)
,
239
:
15
-27,  
1999
.
3
Coustry F., Maity S. N., Sinha S., deCrombrugghe B. The transcriptional activity of the CCAAT-binding factor CBF is mediated by two distinct activation domains, one in the CBF-B subunit and the other in the CBF-C subunit.
J. Biol. Chem.
,
271
:
14485
-14491,  
1996
.
4
Ting J. P., Trowsdale J. Genetic control of MHC class II expression.
Cell
,
109(Suppl.)
:
S21
-S33,  
2002
.
5
Kao C. Y., Tanimoto A., Arima N., Sasaguri Y., Padmanabhan R. Transactivation of the human cdc2 promoter by adenovirus E1A.E1A induces the expression and assembly of a heteromeric complex consisting of the CCAAT box binding factor, CBF/NF-Y, and a 110-kDa DNA-binding protein.
J. Biol. Chem.
,
274
:
23043
-23051,  
1999
.
6
Hu Q., Bhattacharya C., Maity S. N. CBF binding mediates cell cycle activation of topoisomerase IIα: conventional CBF activation domains are not required.
J. Biol. Chem.
,
277
:
37191
-37200,  
2002
.
7
Hu Q., Maity S. N. Stable expression of a dominant negative mutant of CBF/NF-Y in mouse fibroblast cells resulting in retardation of cell growth and inhibition of transcription of various cellular genes.
J. Biol. Chem.
,
275
:
4435
-4444,  
2000
.
8
Li X. Y., Hooft van Huijsduijnen R., Mantovani R., Benoist C., Mathis D. Intron-exon organization of the NF-Y genes. tissue-specific splicing modifies an activation domain.
J. Biol. Chem.
,
267
:
8984
-8990,  
1992
.
9
Bi W., Deng J. M., Zhang Z., Behringer R. R., deCrombrugghe B. Sox9 is required for cartilage formation.
Nat. Genet.
,
22
:
85
-89,  
1999
.
10
Zindy F., Eischen C. M., Randle D. H., Kamijo T., Cleveland J. L., Sherr C. J., Roussel M. F. Myc signaling via the ARF tumor suppressor regulates p53-dependent apoptosis and immortalization.
Genes Dev.
,
12
:
2424
-2433,  
1998
.
11
Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain.
Nat. Genet.
,
21
:
70
-71,  
1999
.
12
Loonstra A., Voojis M., Beverloo H., Allak B., Drunen E., Kanaar R., Berns A., Jonkers J. Growth inhibition and DNA damage induced by Cre recombinase in mammalian cells.
Proc. Natl. Acad. Sci. USA
,
98
:
9209
-9214,  
2001
.
13
Hans F., Dimitrov S. Histone H3 phosphorylation and cell division.
Oncogene
,
20
:
3021
-3027,  
2001
.
14
Borel F., Lacroix F. B., Margolis R. L. Prolonged arrest of mammalian cells at the G1/S boundary results in permanent S phase stasis.
J. Cell Sci.
,
115
:
2829
-2838,  
2002
.
15
Lassus P., Opitz-Araya X., Lazebnik Y. Requirement for caspase 2 in stress-induced apoptosis before mitochondrial permeabilization.
Science (Wash. DC)
,
297
:
1352
-1354,  
2002
.
16
Shearwin-Whyatt L., Harvey N., Kumar S. Subcellular localization and CARD dependent oligomerization of the death adaptor RAIDD.
Cell Death Differ.
,
7
:
155
-165,  
2000
.
17
Monroe D. G., Berger R. R., Sanders M. M. Tissue-protective effects of estrogen involve regulation of caspase gene expression.
Mol. Endocrinol.
,
16
:
1322
-1331,  
2002
.
18
Furstenthal L., Swanson C., Kaiser B., Eldridge A. G., Jackson P. Triggering ubiquitination of a CDK inhibitor at origins of DNA replication.
Nat. Cell Biol.
,
3
:
715
-722,  
2001
.
19
Blanchard J. M. Cyclin A2 transcriptional regulation: modulation of cell cycle control at the G1/S transition by peripheral cues.
Biochem. Pharmacol.
,
60
:
1179
-1184,  
2000
.