Introduction

Elucidating apoptotic pathways has been instrumental in understanding various aspects of cancer biology. Apoptosis is the end point for a range of tumor-suppressive mechanisms, evolved to prevent the emergence of autonomous cells. Programmed death pathways feature prominently in the DNA damage response, where they serve to prevent the accumulation of mutations. In this regard, it is a long-held view that the key determinant of DNA damage response is the activation of p53 and p53-dependent pathways, which in turn results in cell cycle arrest and/or apoptosis (Vousden and Lu, 2002). While the canonical position of p53 as the guardian of the genome cannot be overemphasized, there is an emerging field of evidence to suggest that a number of alternative pathways not involving p53 also play crucial, albeit less dominant roles in apoptosis. These include both factors that modulate the p53 response without themselves being dependent on p53 (such as c-myc, p300 and E2F), as well as factors that bypass the p53 pathway and directly influence apoptosis (such as bcl2) (Chao and Korsmeyer, 1998; Seoane et al., 2002; Iyer et al., 2004; Zhao et al., 2005). Importantly, the latter group of p53-independent modulators of apoptosis includes several factors (for example, Akt, Pim and pp90RSK) that make up the survival pathways (Amaravadi and Thompson, 2005; Hammerman et al., 2005). These are especially important in coupling oncogenic pathways with apoptosis. In an environment replete of p53 function, which is the scenario in most cancers, p53-independent mechanisms may play vital, yet under-recognized roles in the cellular response to DNA damage. Therefore, understanding these pathways may provide valuable insight into hitherto unknown aspects of tumor biology, with consequent prognostic and therapeutic implications.

In this study, we focus our attention on a transcription factor, Kruppel-like factor 5 (KLF5), which has not previously been implicated in the DNA damage response. Kruppel-like factors belong to a family of Sp1-like zinc-finger proteins, with over twenty members identified to date (Black et al., 2001; Kaczynski et al., 2003; Safe and Abdelrahim, 2005). These proteins regulate gene expression by binding GC-rich regulatory elements within promoters. There is increasing evidence to suggest that KLFs play a central role in the regulation of cell growth, proliferation, differentiation and tumorigenesis. KLF5 is the fifth member of this family, and has been demonstrated to mediate various aspects of cardiovascular physiology, including remodeling, differentiation and development (Shi et al., 1999; Nagai et al., 2003; Ghaleb et al., 2005; Suzuki et al., 2005). Recent studies also suggest that KLF5 plays a role in tumor progression in breast and prostate cancers (Chen et al., 2002, 2003). While functional assays in untransformed cells show KLF5 to be associated with cellular proliferation, experimental data in cancer cells have been contradictory as to whether KLF5 acts as an oncogene or a tumor suppressor (Bateman et al., 2004). Overexpression of KLF5 in bladder cancer cells was shown to promote tumorigenesis in mice, increase G₁ to S-phase transition, activate mitogen-activated protein kinase and Akt pathways and increase cyclin D1 transcription (Chen et al., 2006). In contrast, KLF5 reduced cell viability and inhibited cell proliferation in esophageal and breast cancer cell lines (Chen et al., 2002; Yang et al., 2005). Tumor analyses have demonstrated the loss of KLF5 expression in numerous cancers, either through hemizygous deletion, or through activation of ubiquitination–proteosome pathways (Chen et al., 2003, 2005a).

In this report, we show that KLF5 plays an important role in modulating apoptosis secondary to DNA damage in a p53-independent manner, and suggest a mechanism through which this occurs.

Results

KLF5 is activated in response to DNA damage in a p53-independent manner

Previous microarray data suggest that KLF5 transcription is induced after DNA damage in a p53-dependent manner (Kho et al., 2004). To verify these findings, we performed western blots on HCT116 cells (which are p53 wild type) treated with 5-fluorouracil (5FU) (Figure 1a) and ultraviolet irradiation (UVR) (data not shown). These confirmed that KLF5 protein levels increase soon after DNA damage: levels were maximal 8 h after 5FU treatment. When similar experiments were carried out on p53-deficient derivatives of HCT116 (HCT116 p53 -/-), KLF5 levels increased in identical fashion. Treatment with different drug doses demonstrates that KLF5 induction occurs at lower 5FU doses than required to induce apoptosis, but is maximal above concentrations of 300 nM (Figure 1b). These experiments show that DNA damage induces KLF5 expression regardless of p53 status.

Figure 1.

Kruppel-like factor 5 (KLF5) levels after DNA damage and in KLF5-depleted cells. Western blot showing KLF5 levels in nuclear extracts (with PCNA loading controls) in (a) HCT116 and p53 -/- cells at various time points after 5-fluorouracil (5FU) treatment. (b) HCT116 cells treated with indicated doses of 5FU 8 h after treatment. (c) KLF5-depleted (KLF5 shRNA) and negative control (NC shRNA) HCT116 cells 24 h after 5FU treatment.

Full figure and legend (67K)

Loss of KLF5 promotes apoptosis in response to DNA-damaging agents regardless of p53 status

To analyse further the role of KLF5 in the DNA damage response, we generated KLF5 stable knockdowns (KLF5 short hairpin RNA (shRNA)) from HCT116 using shRNA technology. Cells stably transfected with a scrambled shRNA sequence (NC shRNA) were used as controls. All subsequent experiments were performed on two independently derived pairs of cells, and average expression levels of KLF5 in these cells are shown (Figure 1c). When treated with 5FU and UVR, KLF5-depleted cells demonstrated a doubling of apoptotic fractions compared to controls (Figures 2a and b). Western blots of apoptotic markers caspase-9 and PARP confirmed these findings (Figure 2c).

Figure 2.

Loss of Kruppel-like factor 5 (KLF5) promotes DNA damage-dependent apoptosis. (a) Flow cytometry profile showing percentage of apoptotic cells (SubG₁) in KLF5shRNA and NC shRNA cells, after 5-fluorouracil (5FU) treatment. (b) Graph showing percentage of apoptotic cells in KLF5 shRNA and NC shRNA cells, after 5FU and UV treatment. (c) Western blots showing caspase-9, PARP and -actin levels in KLF5 shRNA and NC shRNA cells, after 5FU treatment. (d) Graph showing percentage of apoptotic cells in KLF5-depleted (KLF5 siRNA) and negative control (NC siRNA) A549, HCT15 and HCT116 p53 -/- cell lines, after 5FU treatment. In (b and d), apoptotic fractions were derived from determining subG₁ fractions in flow cytometry profiles. Error bars denote one standard deviation.

Full figure and legend (112K)

To demonstrate that this phenotype was neither peculiar to the HCT116 cell line nor to the p53 status of cells, we carried out similar DNA damage experiments on a number of different cell lines. In these analyses, we used A549 (p53 wild type), HCT15 (p53 mutant) and HCT116 p53 -/- cells. KLF5 depletion was achieved using commercially obtained small interfering RNA (siRNA) molecules (both KLF5-specific and negative controls) from Dharmacon Inc. (Lafayette, CO, USA), and depletion confirmed at the start of all experiments by western blots (data not shown). In every case, apoptotic fractions were doubled in KLF5-depleted cells after 5FU treatment (Figure 2d). However, the effect of KLF5 depletion on apoptosis was more dramatic in cells wild type for p53. These experiments show that loss of KLF5 sensitizes cells to DNA-damaging agents.

Increased apoptosis in KLF5-depleted cells is not associated with abnormalities in p53-dependent pathways, but occurs secondary to reduced bad phosphorylation

Western blots were performed to determine levels of p53 and p53-dependent factors in KLF5 shRNA and NC shRNA cells after 5FU treatment (Figure 3a). These showed that there were no differences in p53, p21, PUMA and NOXA levels that could account for the increased level of apoptosis in KLF5-depleted cells. Levels of E2F1 and E2F4, which have previously been implicated in apoptosis control, were also unaffected by KLF5 status.

Figure 3.

p53-dependent and -independent factors involved in the DNA damage response. Western blots showing levels of: (a) p53, p53-dependent and -independent factors, (b) factors involved in apoptosis and (c) bad and bad phosphorylation in Kruppel-like factor 5 (KLF5) shRNA and NC sh RNA cells, after 5-fluorouracil (5FU) treatment (with -actin loading control).

Full figure and legend (141K)

To determine whether KLF5 affects factors involved in the apoptotic response, we performed western blots to determine levels of Bax, Bcl2, Bid, BclX_L, BclX_S (Figure 3b), Bad and phospho-Bad, in treated cells (Figure 3c). KLF5 loss specifically resulted in reduced Bad phosphorylation at both ser112 and ser136 residues. Densitometric analysis confirmed that reduced Bad phosphorylation was significant across triplicate blots (Supplementary Data). All other factors examined were unaffected. Interestingly, there was no difference in phosphorylation levels of BclX_L, Bax, Bid (migration patterns in Figure 3b) or Bcl2 (data not shown) between the two cell lines. These experiments confirm that KLF5 modulates the apoptotic response in a p53-independent manner and suggest an association between KLF5 and Bad phosphorylation.

Loss of KLF5 promotes apoptosis through Pim1 kinase

Recent data have shown that apoptosis can be controlled by a number of survival pathways, which converge to affect phosphorylation of Bad (Bergmann, 2002). Of these, the best characterized are the Akt, Mek/Erk and Pim kinases. Our western blots showed that 5FU-treated KLF5 shRNA and NC shRNA cells have identical levels of Akt, phospho-Akt, Mek1/2, phospho-Mek1/2, Erk1/2 and phospho-Erk1/2 (Figure 4a). However, levels of Pim1 kinase differed significantly depending on KLF5 status. In NC shRNA cells, Pim1 levels increased soon after 5FU-induced DNA damage (Figure 4b). However, in KLF5-deficient cells, Pim1 protein levels did not increase and remained at levels lower than those seen in NC shRNA cells. Similar results were obtained when cells were treated with UVR (Figure 4c). Real-time reverse transcription (RT)–PCR demonstrated that loss of KLF5 results in a failure to induce Pim1 expression after 5FU or UVR-induced damage, compared to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and Pim2 controls (Figure 4d). These results suggest that KLF5 loss results in a failure to induce Pim1 in response to DNA damage.

Figure 4.

Involvement of Pim1 in Kruppel-like factor 5 (KLF5)-dependent modulation of apoptosis. Western blots of KLF5 shRNA and NC shRNA cells (with -actin loading control) showing levels of (a) factors involved in bad phosphorylation after 5-fluorouracil (5FU) treatment. (b) Pim1 after 5FU treatment and (c) Pim1 and bad phosphorylation after UV irradiation. (d) Graph showing relative levels of Pim1 and Pim2 mRNA in KLF5 shRNA and NC shRNA cells, 12 h after treatment with 5FU and UV irradiation (UVR). Expression levels were determined using real-time RT–PCR, in triplicate experiments with GAPDH control. Error bars denote one standard deviation. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; RT, reverse transcription; shRNA, short hairpin RNA.

Full figure and legend (127K)

Overexpression of Pim1 reverses the phenotype induced by KLF5 loss

To confirm that the phenotype seen in KLF5-depleted cells is mediated through Pim1, we transiently transfected NC shRNA and KLF5 shRNA cells with constructs overexpressing Pim1, and treated cells with 5FU. In these experiments, cells were also transfected with constructs expressing Pim2 and a kinase-deficient Pim1 mutant (mutant Pim1) as controls. Expression levels of Pim1 and Pim2 were confirmed by western blots (data not shown). Cell cycle profiling showed that only cells transfected with wild-type Pim1 had apoptotic fractions similar to NC shRNA cells (Figure 5a); overexpression of Pim2 and mutant Pim1 failed to rescue the apoptotic phenotype. Overexpression of all three factors had no effect on apoptosis in NC shRNA cells. Altogether, these experiments show that the increased sensitivity of KLF5-depleted cells to DNA damage is mediated through Pim1-survival kinase.

Figure 5.

Pim1 overexpression reverses the apoptotic phenotype. (a) Graph showing percentage of apoptotic cells in NC shRNA and Kruppel-like factor 5 (KLF5) shRNA cells transiently transfected with Pim1, Pim2 and Pim1 kinase-deficient mutant (Pim1 mut), after 5-fluorouracil (5FU) treatment. Apoptotic fractions were derived from determining subG₁ fractions in flow cytometry profiles. Error bars denote one standard deviation. (b) ChIP and PCR of Pim1 two promoter segments containing Sp1-binding sites, in HCT116 before (unrx) and after 5FU treatment. Promoter location as indicated based on Pim1 sequence. Pulldowns were done using KLF5 and IgG (negative control) antibodies. Input lane shown as loading control.

Full figure and legend (63K)

KLF5 binds to Pim1 promoter consequent to DNA damage

Previous data have shown that the Pim1 gene promoter contains a number of putative Sp1 sites, suitable for KLF5 binding (Meeker et al., 1990). We performed chromatin immunoprecipitation (ChIP) in HCT116 cells subject to DNA damage: KLF5 antibodies were used to pull down chromatin fractions bound to KLF5, and PCR was performed using primers flanking two segments on the Pim1 promoter known to contain putative Sp1 consensus sequences. PCR showed that treatment with 5FU results in increased binding of KLF5 to one of the two Sp1-binding regions (located between nucleotide –953 and –789 of the Pim1 promoter region) compared to untreated cells (Figure 5b). This suggests that in response to DNA damage, KLF5 binds to the Pim1 promoter and regulates expression.

Discussion

Previous studies have implicated the KLF5 protein not only in the regulation of proliferation, cell cycle progression, cellular proliferation and differentiation, but also in carcinogenesis (Sun et al., 2001; Chen et al., 2002, 2003; Nandan et al., 2004; Ghaleb et al., 2005). However, the underlying molecular mechanism for KLF5 in cancer pathogenesis remains largely unclear. Given its role in cell cycle control and proliferation, some controversy exists as to whether KLF5 functions as a tumor suppressor, as there is an equal degree of evidence suggesting otherwise. The apparent contradiction in experimental data suggests that the role of KLF5 in these pathways is largely context dependent. The possibility that KLF5 expression is induced by DNA-damaging agents prompted us to question if KLF5 functions in an alternative capacity in damage-response and/or apoptotic pathways.

In addition to phorbol esters and mitogenic stimuli, previous data suggest that KLF5 expression could also be induced by chemotherapeutic agents (Kho et al., 2004). It was hypothesized that this is p53 dependent, and that KLF5 functions as a transcriptional target of p53. Hence, we were not surprised to find that KLF5 levels increased soon after UV and 5FU induced DNA damage. However, our experiments also indicate that this induction occurred even in a p53-null background, and at sub-apoptotic doses of DNA-damaging agents.

Our subsequent findings were derived from experiments performed in cells where KLF5 has been depleted, either stably or transiently, using RNA-interference technology. Loss of KLF5 consistently results in an increased sensitivity to DNA damage, regardless of cellular p53 status. Unlike the role of KLF5 in cell cycle and proliferative pathways, the apoptotic phenotype consequent to KLF5 depletion appears to be independent of cellular context. Substantiating a p53-independent mechanism of control, we showed that neither p53 nor the main p53-dependent targets analysed were affected by KLF5 depletion. Indeed, preliminary microarray analyses comparing transcriptional profiles of 5FU treated, KLF5-deficient cells are consistent with these findings, showing no differences between the major p53 transcriptional targets (MS Hamza and NG Iyer, manuscript in preparation). These results raise the distinct possibility that KLF5 modulates DNA damage-induced apoptosis through a p53-independent mechanism.

Despite the significant increase in cell death secondary to DNA damage in KLF-deficient cells, most of the key mediators controlling intrinsic apoptotic pathways (Bax, Bcl2, BclXL, PUMA and NOXA) were unaffected. Instead, the increased sensitivity seen was associated with markedly reduced levels of Bad phosphorylation, at two specific residues: ser112 and ser136. Analyses of Bad kinases showed that only Pim1 levels were reduced, both at transcript and protein levels. Furthermore, overexpressing Pim1 (but not Pim2 or a kinase-deficient Pim1 mutant) was sufficient to reverse the apoptotic phenotype in KLF5-deficient cells. Finally, ChIP demonstrated that occupancy of KLF5 on the Pim1 promoter increases consequent to DNA damage.

The role of Bad in apoptotic control is undisputed (Chao and Korsmeyer, 1998). More importantly, the control of Bad activity appears to be at the junction where various signal transduction pathways are coupled to apoptosis (Bergmann, 2002). When unphosphorylated, active Bad induces apoptosis by sequestering anti-apoptotic members of the Bcl2 family (for example, Bcl2 and BclXL), permitting aggregation of pro-apoptotic members (for example, Bax and Bak) (Ruvolo et al., 2001). The latter then drives cytochrome C release and caspase activation. In contrast, Bad activity is repressed by phosphorylation at three specific residues by several cytoplasmic serine/threonine kinases, and these include survival kinases such as Akt and members of the Pim family (Mitsiades et al., 2004; Macdonald et al., 2006).

While previous experiments have implicated KLF5 in pathways that promote Akt activity, we found that levels of Akt and phospho-Akt were unaffected in KLF5-deficient cells (Chen et al., 2006). Instead, Pim1 expression levels were reduced in these cells. In fact, microarray experiments in drug-treated cells show that Pim1 is one of the most significantly downregulated genes in KLF5-depleted cells (MS Hamza and NG Iyer, manuscript in preparation). Pim1 belongs to a family of survival kinases that function downstream of JAK/STAT activation, and regulates both apoptosis and cellular metabolism (Hammerman et al., 2005). Pim1 phosphorylates a range of cellular substrates, which includes Bad at two distinct residues: ser112 and ser 136 (Aho et al., 2004; Macdonald et al., 2006). Interestingly, unlike all other serine/threonine kinases known to phosphorylate bad, Pim1 activity is regulated at the transcription level, that is, by controlling expression levels of the active protein. It is evident in our experiments that the increase in Pim1 levels mirrors KLF5 induction (in cells wild type for KLF5), in response to DNA damage. Moreover, previous data have established that there are several Sp1-binding sites on the Pim1 promoter, where KLF5 could potentially bind (Meeker et al., 1990). ChIP experiments described here certainly support this prediction: KLF5 binding to one of these Sp1 consensus sites, between nucleotides –953 and –789 on the Pim1 promoter, increased soon after damage. We hope that future experiments using promoter assays will verify these findings and narrow down the specific Sp1 sites responsible for KLF5-induced transactivation of Pim1. We therefore hypothesize that in response to DNA damage, KLF5 is induced, and in turn functions as a transcriptional activator of Pim1 expression. Conversely, loss of KLF5 results in a failure to transactivate Pim1, reduced bad phosphorylation and increased apoptosis.

The potential clinical implications of these findings are intriguing. Our data suggest that KLF5 depletion increases chemosensitivity in a cell culture environment. KLF5 expression has been shown to be reduced in several types of cancers (Dong, 2001; Chen et al., 2002, 2003). Future studies should establish whether KLF5 expression in primary tumor samples correlates with chemotherapeutic response in patients. Interestingly, it has been shown that Pim1 levels serve as a good predictor of response to 5FU-based chemotherapy in patients with gastric cancer (Chen et al., 2005b). Equally provocative is the notion that inhibitors of KLF5 and Pim1 could be used in a therapeutic setting to augment tumor sensitivity to DNA-damaging agents.

While there is little mechanistic overlap between the control of cellular proliferation and apoptosis, the two processes are coupled at various levels by key molecules (Felsher, 2003). Factors responsible for orchestrating these diverse, yet interdependent pathways are often targets for oncogenic mutations. There is little doubt that KLF5 acts as an important regulator of cellular proliferation, affecting cell survival, transformation and angiogenic pathways. In this report, we show that KLF5 also plays a critical role in modulating apoptosis through Pim1 signaling.

Materials and methods

Cell culture

HCT116, A549 and HCT15 cell lines were obtained from ATCC (Manassas, VA, USA), while HCT116 p53 -/- were kind gifts from Dr B Vogelstein (Polyak et al., 1996). Cells were cultured in Dulbecco's modified Eagle's medium, supplemented with 10% fetal calf serum (Sigma, St Louis, MO, USA). All drugs were obtained from Sigma unless otherwise stated. DNA damage was induced with 5FU (375 ng/ml) and UV irradiation (15 J/m²); the latter delivered using a Stratalinker UV-crosslinker (Stratagene, La Jolla, CA, USA). KLF5-deficient (KLF5 shRNA) and negative controls (NC shRNA) were generated by transfecting HCT116 cells with KLF5 shRNA and negative shRNA constructs, respectively. Stable transfectants were selected with 50 g/ml puromycin, and individual clones were isolated and expanded under selective conditions. KLF5 expression levels were determined before each experiment to ensure adequate knockdown.

Plasmid constructs

KLF5 shRNA and negative control shRNA constructs were generated using the pRetroQ vector system according to the manufacturer's protocol (BD Biosciences, San Jose, CA, USA). KLF5 target sequence used for RNA-interference was 5'-GCTCACCTGAGGACTCA-3'. Wild-type and kinase-deficient (NT81, where N terminus truncated of 81 amino acids) mutants of pSV-pim-1 and wild-type pSV-pim-2 constructs were generous gifts from Dr Koskinen (described in Aho et al., 2004; Rainio et al., 2005).

Flow cytometry analysis

Cells were harvested, fixed and stained with propidium iodide (50 g/ml) after treatment with RNase (100 g/ml). Stained cells were analysed for DNA content by flow cytometry using a FACScalibur (BD Biosciences). Cell cycle fractions were quantified using the CellQuest software.

Western blotting

Western blots were performed as described previously, using the following antibodies: goat anti-KLF5/BTEB2 (A-16), mouse anti-p53 (DO-1), mouse anti-Bcl-2, ADP-ribose anti-p21, rabbit anti-bcl-X_L, rabbit anti-bcl-X_S, mouse anti-bax, mouse anti-pim1, mouse anti-E2F1, rabbit anti-E2F4, mouse anti-PARP (Santa Cruz Biotechnologies, Santa Cruz, CA, USA), rabbit anti-Akt, rabbit anti-phospho-ser473-Akt, rabbit anti-MEK1/2, rabbit anti-phospho-Ser217/221-MEK1/2, rabbit anti-ERK1/2, rabbit anti-phospho-Thr202/Tyr204-Erk, rabbit anti-bad, rabbit anti-phospho-ser112-Bad, rabbit anti-phospho-ser136-bad and rabbit anti-caspase-9 (Cell Signaling Technologies, Danvers, MA, USA) (Zhao et al., 2005). Results were quantified by densitometric analysis using AlphaEase software. All western blots were performed in triplicate; blots shown are representative examples.

RNA interference

SMARTpool KLF5 small interfering RNAs (siRNA) and negative control siRNA (siCONTROL nontargeting pooled) were purchased from Dharmacon Inc. Cells were transfected using the LipofectAMINE 2000 reagent (Invitrogen, Singapore, Singapore) according to the manufacturer's protocol.

Real-time RT–PCR

Total cellular RNA was isolated using an RNeasy mini-kit (Qiagen, Valencia, CA, USA), and cDNAs synthesized using the one Step RT–PCR Kit (Clontech, Mountain View, CA, USA) according to the manufacturer's protocol. Quantification of KLF5, pim1, pim2 and -actin transcript levels was performed as reported previously (Iyer et al., 2004). Quantitative real-time PCR was performed on cDNA products using the DNA Engine Opticon Real-time Detection System (MJ Research, Waltham, MA, USA) and the QuantiTect SYBR Green RT–PCR kit (Qiagen). Copy numbers were estimated from the threshold amplification cycle numbers using software supplied. Primer sequences are available upon request.

Chromatin immunoprecipitation

ChIP was performed as previously described (Zhao et al., 2005). Antibodies used were goat anti-KLF5 (A-16) and normal goat immunoglobulins (Santa Cruz Biotechnologies). PCR primers were designed to amplify two separate regions of the Pim1 promoter known to contain Sp1 consensus sequences (Meeker et al., 1990). Primer pairs used were: pim1-prom forward1 5'-TGACACACATCCCTTCCC-3' and pim1-prom reverse1 5'-TGGAACAGAACTGGAGGC-3', flanking a region between nucleotide -953 and -789 of the Pim1 promoter region; pim1-prom forward2 5'-CGGCGTAGAGACCATTCTGA-3' and pim1-prom reverse2 5'-C TGCAGCAGACGCCCGGCTC-3' flanking a region between nucleotide -309 and +37 of the Pim1 promoter region (location based on nucleotide upstream of Pim1 start codon).

References

1. Aho TL, Sandholm J, Peltola KJ, Mankonen HP, Lilly M, Koskinen PJ. (2004). Pim-1 kinase promotes inactivation of the pro-apoptotic Bad protein by phosphorylating it on the Ser112 gatekeeper site. FEBS Lett 571: 43–49. | Article | PubMed | ISI | ChemPort |
2. Amaravadi R, Thompson CB. (2005). The survival kinases Akt and Pim as potential pharmacological targets. J Clin Invest 115: 2618–2624. | Article | PubMed | ISI | ChemPort |
3. Bateman NW, Tan D, Pestell RG, Black JD, Black AR. (2004). Intestinal tumor progression is associated with altered function of KLF5. J Biol Chem 279: 12093–12101. | Article | PubMed | ISI | ChemPort |
4. Bergmann A. (2002). Survival signaling goes BAD. Dev Cell 3: 607–608. | Article | PubMed | ISI | ChemPort |
5. Black AR, Black JD, Azizkhan-Clifford J. (2001). Sp1 and Kruppel-like factor family of transcription factors in cell growth regulation and cancer. J Cell Physiol 188: 143–160. | Article | PubMed | ISI | ChemPort |
6. Chao DT, Korsmeyer SJ. (1998). BCL-2 family: regulators of cell death. Annu Rev Immunol 16: 395–419. | Article | PubMed | ISI | ChemPort |
7. Chen C, Benjamin MS, Sun X, Otto KB, Guo P, Dong XY et al. (2006). KLF5 promotes cell proliferation and tumorigenesis through gene regulation in the TSU-Pr1 human bladder cancer cell line. Int J Cancer 118: 1346–1355. | Article | PubMed | ISI | ChemPort |
8. Chen C, Bhalala HV, Qiao H, Dong JT. (2002). A possible tumor suppressor role of the KLF5 transcription factor in human breast cancer. Oncogene 21: 6567–6572. | Article | PubMed | ISI | ChemPort |
9. Chen C, Bhalala HV, Vessella RL, Dong JT. (2003). KLF5 is frequently deleted and down-regulated but rarely mutated in prostate cancer. Prostate 55: 81–88. | Article | PubMed | ISI | ChemPort |
10. Chen C, Sun X, Ran Q, Wilkinson KD, Murphy TJ, Simons JW et al. (2005a). Ubiquitin-proteasome degradation of KLF5 transcription factor in cancer and untransformed epithelial cells. Oncogene 24: 3319–3327. | Article | ISI | ChemPort |
11. Chen CN, Lin JJ, Chen JJ, Lee PH, Yang CY, Kuo ML et al. (2005b). Gene expression profile predicts patient survival of gastric cancer after surgical resection. J Clin Oncol 23: 7286–7295. | Article | PubMed | ISI | ChemPort |
12. Dong JT. (2001). Chromosomal deletions and tumor suppressor genes in prostate cancer. Cancer Metastasis Rev 20: 173–193. | Article | PubMed | ISI | ChemPort |
13. Felsher DW. (2003). Cancer revoked: oncogenes as therapeutic targets. Nat Rev Cancer 3: 375–380. | Article | PubMed | ISI | ChemPort |
14. Ghaleb AM, Nandan MO, Chanchevalap S, Dalton WB, Hisamuddin IM, Yang VW. (2005). Kruppel-like factors 4 and 5: the yin and yang regulators of cellular proliferation. Cell Res 15: 92–96. | Article | PubMed | ISI | ChemPort |
15. Hammerman PS, Fox CJ, Birnbaum MJ, Thompson CB. (2005). Pim and Akt oncogenes are independent regulators of hematopoietic cell growth and survival. Blood 105: 4477–4483. | Article | PubMed | ISI | ChemPort |
16. Iyer NG, Chin SF, Ozdag H, Daigo Y, Hu DE, Cariati M et al. (2004). p300 regulates p53-dependent apoptosis after DNA damage in colorectal cancer cells by modulation of PUMA/p21 levels. Proc Natl Acad Sci USA 101: 7386–7391. | Article | PubMed | ChemPort |
17. Kaczynski J, Cook T, Urrutia R. (2003). Sp1- and Kruppel-like transcription factors. Genome Biol 4: 206. | Article | PubMed |
18. Kho PS, Wang Z, Zhuang L, Li Y, Chew JL, Ng HH et al. (2004). p53-regulated transcriptional program associated with genotoxic stress-induced apoptosis. J Biol Chem 279: 21183–21192. | Article | PubMed | ISI | ChemPort |
19. Macdonald A, Campbell DG, Toth R, McLauchlan H, Hastie CJ, Arthur JS. (2006). Pim kinases phosphorylate multiple sites on Bad and promote 14-3-3 binding and dissociation from Bcl-XL. BMC Cell Biol 7: 1. | Article | PubMed | ChemPort |
20. Meeker TC, Loeb J, Ayres M, Sellers W. (1990). The human Pim-1 gene is selectively transcribed in different hemato-lymphoid cell lines in spite of a G + C-rich housekeeping promoter. Mol Cell Biol 10: 1680–1688. | PubMed | ISI | ChemPort |
21. Mitsiades CS, Mitsiades N, Koutsilieris M. (2004). The Akt pathway: molecular targets for anti-cancer drug development. Curr Cancer Drug Targets 4: 235–256. | Article | PubMed | ISI | ChemPort |
22. Nagai R, Shindo T, Manabe I, Suzuki T, Kurabayashi M. (2003). KLF5/BTEB2, a Kruppel-like zinc-finger type transcription factor, mediates both smooth muscle cell activation and cardiac hypertrophy. Adv Exp Med Biol 538: 57–65. Discussion 66. | PubMed | ISI | ChemPort |
23. Nandan MO, Yoon HS, Zhao W, Ouko LA, Chanchevalap S, Yang VW. (2004). Kruppel-like factor 5 mediates the transforming activity of oncogenic H-Ras. Oncogene 23: 3404–3413. | Article | PubMed | ISI | ChemPort |
24. Polyak K, Waldman T, He TC, Kinzler KW, Vogelstein B. (1996). Genetic determinants of p53-induced apoptosis and growth arrest. Genes Dev 10: 1945–1952. | Article | PubMed | ISI | ChemPort |
25. Rainio EM, Ahlfors H, Carter KL, Ruuska M, Matikainen S, Kieff E et al. (2005). Pim kinases are upregulated during Epstein-Barr virus infection and enhance EBNA2 activity. Virology 333: 201–206. | Article | PubMed | ISI | ChemPort |
26. Ruvolo PP, Deng X, May WS. (2001). Phosphorylation of Bcl2 and regulation of apoptosis. Leukemia 15: 515–522. | Article | PubMed | ISI | ChemPort |
27. Safe S, Abdelrahim M. (2005). Sp transcription factor family and its role in cancer. Eur J Cancer 41: 2438–2448. | Article | PubMed | ISI | ChemPort |
28. Seoane J, Le HV, Massague J. (2002). Myc suppression of the p21(Cip1) Cdk inhibitor influences the outcome of the p53 response to DNA damage. Nature 419: 729–734. | Article | PubMed | ISI | ChemPort |
29. Shi H, Zhang Z, Wang X, Liu S, Teng CT. (1999). Isolation and characterization of a gene encoding human Kruppel-like factor 5 (IKLF): binding to the CAAT/GT box of the mouse lactoferrin gene promoter. Nucleic Acids Res 27: 4807–4815. | Article | PubMed | ISI | ChemPort |
30. Sun R, Chen X, Yang VW. (2001). Intestinal-enriched Kruppel-like factor (Kruppel-like factor 5) is a positive regulator of cellular proliferation. J Biol Chem 276: 6897–6900. | Article | PubMed | ISI | ChemPort |
31. Suzuki T, Aizawa K, Matsumura T, Nagai R. (2005). Vascular implications of the Kruppel-like family of transcription factors. Arterioscler Thromb Vasc Biol 25: 1135–1141. | Article | PubMed | ISI | ChemPort |
32. Vousden KH, Lu X. (2002). Live or let die: the cell's response to p53. Nat Rev Cancer 2: 594–604. | Article | PubMed | ISI | ChemPort |
33. Yang Y, Goldstein BG, Chao HH, Katz JP. (2005). KLF4 and KLF5 regulate proliferation, apoptosis and invasion in esophageal cancer cells. Cancer Biol Ther 4: 1216–1221. | PubMed | ISI | ChemPort |
34. Zhao Y, Tan J, Zhuang L, Jiang X, Liu ET, Yu Q. (2005). Inhibitors of histone deacetylases target the Rb-E2F1 pathway for apoptosis induction through activation of proapoptotic protein Bim. Proc Natl Acad Sci USA 102: 16090–16095. | Article | PubMed | ChemPort |

Acknowledgements

We thank Dr B Vogelstein and P Koskinen for kindly providing reagents, and Drs M van der Heijden and A Thiagarajan for critically evaluating the manuscript. This project was funded by the Cancer Research and Education Fund from the National Cancer Centre, Singapore and through a Singhealth Foundation grant.

Supplementary Information accompanies the paper on the Oncogene web site (http://www.nature.com/onc).