Apoptosis and Tumor Resistance Conferred by Par-4

Abstract

Par-4 is a tumor suppressor protein with a pro-apoptotic function. Epigenetic silencing of Par-4 is seen in diverse tumors, and Par-4 knockout mice develop spontaneous tumors in various tissues. Endogenous Par-4 is essential for sensitization of cells to diverse apoptotic stimuli, whereas ectopic expression of Par-4 can selectively induce apoptosis in cancer cells. The cancer-specific pro-apoptotic action of Par-4 resides in its centrally located SAC domain. This chapter reviews a novel mouse model with ubiquitous expression of the SAC domain. These SAC transgenic mice display normal development and life span, and, most importantly, are resistant to spontaneous, as well as oncogene-induced, autochthonous tumors. The tumor resistant phenotype and undetectable toxicity of SAC in vivo suggests the SAC domain possesses tremendous therapeutic potential.

INTRODUCTION

Tissue homeostasis is regulated by an intricate balance between the rate of cell proliferation and cell death. Cell death occurs throughout the life span of multicellular organisms, and represents the only irreversible cell fate decision. Apoptosis, or programmed cell death, is by far the predominant form of physiological cell death, which is used for the coordinated elimination of excess, hazardous, or damaged somatic cells. Apoptosis is a regulated process, controlled by diverse extracellular and intracellular signals, and depends on both extant and de novo protein synthesis (1, 2). Tremendous pathological implications are associated with deregulation of the delicate balance between cell life and death, including neurodegenerative diseases, immunodeficiency, infertility, autoimmune diseases, and cancer (3),.

Cancer recently passed cardiovascular disease as the major cause of death in the United States (4). Cancer results from an imbalance between the functions of cell survival proteins and apoptotic proteins, with defects in the apoptotic machinery playing an important role in tumor pathogenesis, allowing neoplastic, as well as genetically unstable cells, to survive (5). Moreover, deregulation of apoptosis impacts chemo- and radio-resistance, as it increases the threshold for cell death and facilitates metastasis (6). Many tumor suppressor proteins have been identified and studied. Here, we describe both a pro-apoptotic tumor suppressor protein, prostate apoptosis response 4 (Par-4), which executes apoptosis selectively in cancer cells, and the tumor resistant phenotype of transgenic mice carrying Par-4 or its effecter domain. In particular, this chapter reviews Par-4 identification and function, and details the tumor resistant phenotype of transgenic mice expressing either Par-4 or its effector domain, SAC.

STRUCTURE AND FUNCTION OF PAR-4

Identification and characterization of Par-4

Prostate cancer (PCa), the third leading cause of cancer death in the United States (4), is usually treated by androgen-ablation, which results in the elevation of intracellular Ca²⁺ and subsequent induction of apoptosis. Tumors of the prostate are a mixed population of androgen-dependent and -independent cells (7). Unlike androgen-dependent PCA, androgen-independent PCA do not exhibit elevated intracellular Ca²⁺ levels upon androgen ablation. These androgen-independent cells, therefore, are refractory to androgen-ablation therapy, and are usually the basis for relapse of aggressive PCa. However, forced elevation of intracellular calcium, using the ionophore inomycin, can cause apoptosis in androgen-independent cells (8). Par-4 was identified by differential hybridization between rat AT-3 androgen-independent prostate cancer cells that were either unexposed or exposed to ionomycin in the presence of cycloheximide (9). Par-4 was an immediate early apoptotic gene induced in response to Ca2+ elevation in both androgen-dependent and -independent cells. Human Par-4 was later discovered by yeast two hybrid studies as a partner of Wilms’ tumor 1 (WT1) (10) and atypical protein kinase C (aPKC) (11).

Human Par-4 gene is located on the minus strand of chromosome 12q21.2 (12). The gene encompasses 99.06kb of DNA, and consists of 7 exons and 6 introns. The initiation codon, ATG, is located on exon 2. Par-4 is evolutionarily conserved in vertebrates (13), and is ubiquitously expressed in all the tissues and cell types in humans, mice, horses, pigs, and cows. Low expression of Par-4 is found in certain terminally differentiated cells, such as neurons, lymphocytes, and epithelial cells of the mammary gland, and specific retinal cells, ductal cells of the prostate, and smooth muscle cells, thereby indicating Par-4 is down-regulated during differentiation. Consistent with its pro-apoptotic functions, Par-4 levels are generally higher in dying cells, for example, in prostate ductal cells of castrated rats and degenerating neurons.

Human Par-4 is a 38 kDa protein containing 342 amino acids, whereas rat Par-4 has 332 amino acids and mouse Par-4 has 333 amino acids. Par-4 has two putative nuclear localization sequences (NLS) in the N-terminal region, and a leucine zipper domain (LZ) and a nuclear export sequence (NES) in the C-terminal portion. These domains are 100% conserved in the human, rat, and mouse homologs (14). In addition, Par-4 possesses a number of conserved sites for phosphorylation by kinases, such as PKA and AKT. The schematic structure of Par-4 is presented in Figure 1. The presence of these motifs suggests that the function of Par-4 may be tightly regulated by post-translational modification, localization, or dimerization with partners of biological consequence.

Figure 1. The schematic structure of Par-4 protein.

The nuclear translocation signal (NLS2), leucine zipper (LZ), the centrally located SAC domain, and PKA- and Akt1-phosphorylation sites are presented.

Par-4 is a tumor suppressor and a pro-apoptotic protein

Tumor-suppressor genes play pivotal roles in maintaining genomic integrity and regulating cell proliferation, differentiation, and apoptosis, thus, loss-of-function mutations in these genes are directly related to tumorigenesis. The criteria for tumor suppressor gene designation are (1) identification of loss of function in the development of a cancer, and (2) demonstration that inactivation of the gene in vivo enhances tumor initiation, growth, or progression (15).

Par-4 is located in chromosome 12q21, a region that is unstable and often deleted in pancreatic and gastric cancer (16, 17). 12q21 is also one of the regions reorganized in Wilm’s tumorigenesis (18). Down-regulation of Par-4 is seen in a variety of cancers, such as renal-cell carcinomas (19), neuroblastoma (20), acute lymphoblastic, leukemia, chronic lymphocytic leukemia (17) and endometrial cancer (21). Although Par-4 mutation is rare, a recent study identified a single base mutation at aa189 [Arg (CGA) to Stop (TGA)] localized in exon 3, or the effector domain of Par-4; this mutation causes premature termination of Par-4 in human endometrial carcinoma (21). The same study also demonstrated Par-4 promoter hypermethylation in 32% of the analyzed tumors; in this study hypermethylation correlated with low levels of Par-4 protein. Par-4 promoter hypermethylation and silencing were also detected in endometrial cancer cell lines SKUT1B and AN3CA (21).

Par-4 knockout mice develop spontaneous tumors in various tissues, including endometrium, liver, and lung, and exhibit prostatic intraepithelial neoplasia (PIN). These mice also show an increased incidence of chemical- or hormone-inducible tumors of the bladder and endometrium (22). Heterozygous loss of Par-4 yields the same frequency of tumors/PIN as homozygous loss. The immunological profile of Par-4 −/− knockout mice reveals an increased proliferative response of peripheral T cells, inhibition of apoptosis, elevated NF-κB activity, and decreased JNK activity (23).

Ras genes are the most frequently mutated oncogenes that are detected in human tumors. (24) Oncogenic Ras induces both pro- and anti-apoptotic signaling pathways, but the predominant oncogenic effects of Ras result from the down-modulation of pro-apoptotic pathway components (25–27). Par-4 protein is down-regulated by oncogenic Ras in a variety of cell types through the MEK-ERK pathway (28, 29). Restoration of Par-4 protein levels inhibits oncogene-induced transformation of cells, whereas over-expression of Par-4 effects apoptosis in cells expressing oncogenic Ras.

Par-4 over-expression is sufficient to induce apoptosis in most cancer cells in the absence of a second apoptotic signal, but does not cause apoptosis in normal or immortalized cells (14, 30, 31). In normal or immortalized cells, over-expression of Par-4 sensitizes the cells to a wide range of pro-apoptotic stimuli, such as growth factor withdrawal, agents that elevate intracellular Ca²⁺, TNF, UV, γ-irradiation, or IFNγ (32–34). Par-4 has also been found to be an essential downstream regulator of cell-death programs initiated by various exogenous signals, such as TRAIL, vincristine, doxorubicin, and radiation (33). In addition, endogenous Par-4 also plays an important role as an apoptotic agent under certain conditions, as in degenerative neuronal diseases (35–38).

Apoptosis has been recognized to be an essential process during neuron development, where it appears to be fundamental for the control of the final number of neurons and glial cells in the central and peripheral nervous system (39). Excessive death of one or more populations of neurons results in disease or injury. For example, death of hippocampal and cortical neurons results in Alzheimer’s disease (AD), death of mid brain neurons results in Parkinson’s disease (PD), death of neurons in the striatum results in Huntington’s disease (HD), and, finally, death of lower motor neurons results in amyotrophic lateral sclerosis (ALS) (40). Par-4 plays a role in embryonic neuronal development, and serves to prevent hyper-proliferation of nerve tissues (41). This is achieved by the asymmetric distribution of Par-4 protein during the mitosis of neuronal progenitor cells; the daughter cells lacking Par-4 differentiate into neurons, while those with high levels of Par-4 undergo apoptosis (42).

Several studies have demonstrated endogenous Par-4 is up-regulated in different neurodegenerative diseases. Par-4 protein levels are increased 8- to 20-fold in the hippocampus and inferior parietal cortex in AD patients compared to the same regions in healthy individuals of the same age group (43, 44). In Parkinson’s disease models, administration of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine to monkeys and mice dramatically increased Par-4 levels in the substantia nigra. This increase precedes mitochondrial dysfunction and cell death, both of which are inhibited in human neuroblastoma dopaminergic cells (SK-N-MC) that are pretreated with Par-4 antisense oligodeoxynucleotides (36). Par-4 is also over-expressed in lumbar spinal cord samples from patients with ALS and ALS transgenic mice. Primary mouse spinal cord motor neurons or NSC-9 motor neuron cells are rescued from apoptosis induced by oxidative insults when pretreated with Par-4 antisense oligonucleotides (37).

A variety of signaling pathways involved in neuronal apoptosis are being identified in cell culture studies and knockout mouse models, including down-regulation of the TNF death receptor pathway (45), NF-κB inactivation (46), and mitochondrial dysfunction caused by down-regulation of Bcl2 and BclxL. Par-4 has been reported to down-regulate Bcl2 (28) and NF-κB (47, 48), and is responsible for inducing mitochondrial dysfunction and caspase activation in synaptosomes following exposure to oxidative insults (49).

Mechanism of apoptosis by Par-4

Apoptotic strategies to kill tumor cells can involve direct induction of pro-apoptotic molecules, modulation of either anti-apoptotic or survival proteins, or restoration of tumor suppressor gene functions. The anti-cancer function of Par-4 is achieved by two distinct means: activation of molecular components of the cell-death machinery, and inhibition of pro-survival factors.

The nuclear factor-kappaB (NF-κB) is one of the key contributing factors to the pathogenesis and chemoresistance of many human tumors (50). NF-κB confers resistance to cell death by activating the expression of anti-apoptotic/pro-survival genes. These genes include the inhibitor of apoptosis proteins XIAP, c-IAP1 and c-IAP2, and the Bcl2 family proteins like Bcl-xL, Bfl-1/A1, NR13, and Bcl2. In many cases, NF-κB is persistently activated in tumors owing to constitutive IKK kinase activity, but there are several examples where NF-κB activity results from over-expression and nuclear accumulation of the c-Rel protein (51, 52). In addition to its crucial protective activity within tumor cells, NF-κB plays an important role in the external tumor milieu, where it acts in a paracrine fashion to accelerate tumor cell growth (53). One of the essential apoptotic functions of Par-4 is the inhibition of NF-κB pathway. Studies show Par-4 inhibits the Ras- or Raf-induced transcriptional activity of NF-κB in nucleus (14). By inhibiting PKC, Par-4 also interferes with IκB phosphorylation, as well as subsequent TNFα-induced NF-κB nuclear translocation (54). This latter mechanism of Par4-mediated inhibition of NF-κB is carried out without disrupting the DNA-binding capacity of the NF-κB complex (14, 54).

Programmed cell death usually occurs through one of the two major signaling pathways, known as the extrinsic and the intrinsic pathways. The extrinsic pathway works through the binding of specific ligands to cell surface death receptors, such as Fas/CD95, TNF-R1, or TRAIL, which in turn activate the death domain protein, FADD, and caspase-8. The intrinsic pathway functions exclusively within the cell, and is activated by factors that lead to alterations in the mitochondria, such as ionizing or UV radiation, or agents that increase the intracellular levels of Ca²⁺. Par-4 activates the extrinsic death pathway by enabling the trafficking of Fas and Fas ligand (FasL) to the plasma membrane. The membrane translocation of Fas and FasL leads to the formation of the DISC complex, which consists of trimeric Fas, the adaptor protein FADD, and pro-caspase 8; the DISC complex, in turn, induces apoptosis. Par-4 translocation of Fas and FasL to the plasma membrane is limited in hormone-independent cancer cells in which over-expression of Par-4 leads directly to death. This effect is not observed in the hormone-dependent cancer cells, which are resistant to the over-expression of Par-4-induced apoptosis (30).

Although Par-4 resides in both the cytoplasm and the nucleus, it is primarily cytoplasmic in normal tissues, as judged by immunocytochemistry (10). The NLS2 domain of Par-4 is essential for nuclear translocation, as Par-4 constructs lacking an intact NLS2 domain fail to translocate into nucleus. Interestingly, both normal and hormone-dependent cells are resistant to nuclear translocation of Par-4, and, by extension, resistant to Par-4-mediated apoptosis; this is due to the fact nuclear translocation of Par-4 is essential for inhibition of NF-κB transcriptional activity (14).

Over-expression of Par-4 selectively induces apoptosis in cancer cells, but not in normal or immortalized cells. The cancer cell-specific apoptotic action of Par-4 is attributed to its selective activation via phosphorylation at the T155 residue by protein kinase A (PKA) (55). PKA phosphorylates a wide range of substrates that are involved in the regulation of metabolism, cell growth, and differentiation. PKA isoform I is over-expressed in primary tumors, human cancer cells, and transformed cells (55). Phosphorylation by PKA is a crucial event that triggers the apoptotic function of Par-4, and, accordingly a mutation at the T155 residue completely abolishes Par-4-mediated apoptosis (55). A model (56) of selectively induced, Par-4-mediated apoptosis in cancer cells states that Par-4 requires two distinct events for activation: nuclear entry and phosphorylation by PKA. In normal or immortalized cells, owing to the low levels of active PKA, Par-4 is not phosphorylated at the T155 residue, and, in addition, is retained in the cytoplasm by an unknown mechanism. Conversely, in cancer cells sensitive to its effects, Par-4 is phosphorylated by PKA at T155, and is readily translocated to the nucleus. Interestingly, in cancer cells that are resistant to apoptosis by Par-4, Par-4 is phosphorylated at the T155 residue, yet is inhibited from entering the nucleus by pro-survival factors (57). These observations support the model of Par-4-induced apoptosis in cancer cells being dependent on phosphorylation and nuclear entry (56).

Par-4 associated proteins and their roles in apoptosis

Par-4 associates with several proteins in the cytoplasm and the nucleus. These binding partners either regulate Par-4, or facilitate sensitization of Par-4 resistant cells to apoptotic stimuli. The leucine zipper domain located at the carboxy-terminus of Par-4 is essential for Par-4 binding to its partners. One such partner for Par-4 is the Wilm’s tumor protein, WT-1, which functions as a trans-activator or -repressor for many growth factor genes associated with the etiology of Wilms’ tumor (58). WT-1 induces the transcription of the anti-apoptotic protein Bcl2. Par-4 binds to the zinc finger domain of WT-1 through its leucine zipper domain, and down-regulates the anti-apoptotic gene Bcl2 at the promoter level (10, 28, 59).

Akt, as recently reported, is another binding partner which impacts Par-4 activity in cancer cells. Unlike ectopic expression of Par-4, which induces apoptosis in cancer cells, endogenous Par-4 by itself does not cause apoptosis, yet is necessary for apoptosis induced by diverse exogenous insults. This implies endogenous Par-4 must be inactive, and its apoptotic potential is unleashed in response to apoptotic insults. A recent study demonstrated the cell survival kinase, Akt1 (or protein kinase B), binds to endogenous Par-4 and inactivates it by phosphorylation (human Par-4 at S230, rat par-4 at S249, and mouse Par-4 at S231) (57). The binding and phosphorylation of Par-4 by Akt1 makes Par-4 a substrate for the chaperone protein 14-3-3 (57), which effectively sequesters Par-4 in the cytoplasm. This may explain why endogenous Par-4 (even if phosphorylated at 155T) by itself does not cause apoptosis. Akt is functionally involved in anti-apoptosis in diverse cancer cells (60, 61), and its activity is elevated in cancer due to the loss of PTEN tumor suppressor gene function and/or the activation of upstream lipid kinase, phosphoinositide 3-kinase (PI3K) (62).

Human Par-4 was cloned using the yeast two-hybrid system with the regulatory domain of PKCζ as bait (11). PKCζ belongs to the atypical PKC (aPKC) family, which positively regulates cell proliferation and cell survival by activating the transcription factors, AP-1 and NFκ-B. aPKCs also block death receptor induced apoptosis by phosphorylating the pro-apoptotic protein FADD, thus preventing DISC formation. Exposure to apoptotic stimuli, such as UV irradiation, ceramide, or TNF, triggers an interaction between endogenous Par-4 and PKCζ, leading to a dramatic reduction of PKCζ enzymatic activity, and an increase in apoptosis (11, 54). The interaction between Par-4 and PKCζ induces a conformational change in PKCζ that attenuates its catalytic activity. This reduced activity of PKCζ inhibits IKK activation and IkB phosphorylation in cytoplasm, and NF-κB phosphorylation in the nucleus. PKCζ-induced activation of NF-κB is lost in the presence of equal or excess amounts of Par-4 (63). However, PKCζ and Par-4 are part of a ternary complex that includes an adapter protein, p62 (63); binding of p62 to the PKCζ-Par-4 complex blocks Par-4-mediated inhibition of PKCζ kinase activity (63).

Interestingly, Par-4 associates with other pro-apoptotic proteins, such as the nuclear ZIP kinase (ZIPK or DLK) and THAP-1 in the PML bodies. In response to IFNγ and As₂O₃, the interaction of Par-4 and ZIPK enhances the association of ZIPK and the pro-apoptotic protein, DAXX, in PML bodies, and activates DAXX-induced apoptosis (64, 65). The binding of Par-4 and THAP-1, a novel nuclear pro-apoptotic factor in PML bodies, enhances apoptosis induced by either serum withdrawal or TNF (66).

In summary, over-expression of Par-4 alone selectively induces apoptosis in hormone-independent cancer cells, but not in hormone-resposive, normal, or immortalized cells. Over-expression of Par-4, however, sensitizes hormone-responsive, normal, or immortalized cells to apoptotic stimuli. Endogenous Par-4 also effects apoptosis, but requires exogenous cell-death inducing agents. Whether by over-expression or activation of endogenous Par-4, Par-4 requires both nuclear entry and phosphorylation at T155 to execute its apoptotic function, which involves activation of the Fas/FasL death receptor pathway and inhibition of NF-κB survival pathway.

SAC TRANSGENIC MICE

Identification and characteristics of SAC

Par-4 protein is not significantly homologous to any other protein in GenBank. Serial deletion of Par-4 from both the amino and carboxyl termini led to the identification of a core domain (amino acids 137–195) that, when over-expressed, is sufficient for apoptosis in cancer cells. This 58 amino acid region contains both the NLS2, which allows nuclear entry, and the T155 phosphorylation site, which is essential for Par-4 activation. To explore the apoptotic potential of this 58 amino acid core domain, we studied a broad panel of cancer, immortalized, and primary normal cells (48; and our unpublished data). The results indicate this core domain induces apoptosis in cancer cells regardless of their sensitivity or resistance to full-length Par-4, yet does not induce apoptosis in normal cells (48), our unpublished data). By extension, intratumoral injection of a core domain-expressing adenovirus into xenografts (derived from Par-4 sensitive and resistant cancer cells) in nude mice caused rapid inhibition of tumor growth (30, 67). Apoptosis by the 58 amino acid domain can be provoked in normal cells by artificially elevating PKA activity with exogenous cAMP or over-expressing the catalytic subunit of PKA (55). Owing to the selective apoptosis properties of this Par-4 core domain, we designated it SAC, which denotes selective apoptosis of cancer cells (48).

The SAC sequence is 100% conserved in mammals and rodents (Figure 2), and our functional analysis reveals the SAC domain, similar to Par-4, induces Fas and FasL membrane translocation (48). Co-transfection of SAC with dominant-negative FADD inhibits SAC-induced apoptosis, while over-expression of the SAC domain inhibits NF-κB activity in cancer cells in the same manner as Par-4 (48). These observations demonstrate the molecular directive for Par-4-mediated apoptosis is encoded in the SAC domain, and that the leucine zipper domain is dispensable both for Par-4 induced apoptosis, and inhibition of NF-κB by SAC, as well as Par-4. This observation is consistent with a previous study showing a leucine zipper deletion mutant of Par-4 functionally inhibits NF-κB in cancer cells. Nonetheless, SAC resembles neither the death domains (68), nor the death effecter domains, of other pro-apoptotic proteins.

Figure 2. The protein sequence alignment of the SAC domain.

The uppercase letters in blue color indicate the identical amino acids to the consensus sequences, while those in red color indicate divergence from the consensus sequence. The amino acid sequences are from Genbank.

Due to the absence of the leucine zipper domain, the SAC domain neither binds Akt1, nor is it phosphorylated/inactivated by Akt1. This lack of negative regulation by Akt may explain why SAC induces apoptosis not only in Par-4 sensitive cells, but also in Par-4 resistant cells with elevated Akt activity. Moreover, although SAC localizes to the nucleus in normal and cancer cells, it effects apoptosis only in cancer cells since SAC requires T155 phosphorylation by PKA (which is elevated in cancer cells) (56). The apoptosis function of SAC is illustrated in Figure 3. Thus, as the SAC domain selectively induces apoptosis in cancer cells without negatively impacting normal cells, SAC is an ideal candidate for molecular therapy.

Figure 3. The mechanism for cancer selective apoptosis by SAC.

SAC localizes to the cytoplasm and nucleus when it is over-expressed in cells. SAC is activated by PKA, which is elevated in cancer cells. The activated SAC induces apoptosis by activation of Fas/FasL death receptor pathway and inhibition of NF-κB cell survival pathway.

Generation of SAC transgenic mice

The similarity between the mouse and human genomes, and, particularly, between the histological stages and genetic pathways underlying tumor development in mice and humans, make mice an ideal model system for the study of human cancer-relevant genes. We decided to make transgenic mice of SAC to determine whether (1) sustained and ubiquitous expression of the SAC domain is tolerated by normal tissues over the entire life span in mice, and (2) the physiologic levels of transgenic SAC attained in various tissues were adequate for tumor-suppressive function in vivo.

The pCAGGS (69) plasmid vector, which has a chicken β-actin promoter and a CMV enhancer, was chosen as the transgene-bearing plasmid vector for constitutive and ubiquitous expression of the desired transgenes. A 174-bp DNA fragment containing the SAC domain (amino acid 137-195) of rat Par-4 tagged with eGFP at the c-terminus was cloned into the pCAGGS vector. The eGFP coding regions were cloned into pCAGGS as a control. The SAC-GFP and GFP transgenic mice were made on B6C3F1 background, and three transgenic mouse clones of each DNA construct were generated. Ubiquitous expression of the transgenic protein products was confirmed with whole cell lysates isolated from nine different tissues, and, accordingly, transgene protein expression was relatively low compared to the endogenous Par-4 levels. Southern hybridization indicated a low level of transgene copy numbers (67).

Normal development of SAC mice

Tumor suppresser genes respond to a variety of stress signals to trigger cell cycle arrest, apoptosis, or senescence, thereby protecting against malignant transformation. Although an increase in tumor suppressor activity can diminish the risk of developing cancer, it may simultaneously accelerate age-related loss of tissue cellularity, since enhancing tumor suppressor function can compromise the proliferation of stem cells, enhance apoptosis, and/or increase senescence (70). For example, a premature aging phenotype was seen in an amino-terminal truncated p53 transgenic mouse model (despite the mice showing increased tumor resistance) (71, 72). Although SAC had been tested in normal cells in tissue culture, in vivo tissue tolerance of near-physiologic expression of SAC over an extended time period had to be determined.

We explored quantitative indicators of fertility, viability, body weight, and aging in the SAC transgenic mice. The birth ratio of the SAC transgenic mice was similar to that of the GFP transgenic mice and littermate control mice. All of the animals expressing the transgenes were developmentally normal, and there were no gender-based differences in the average body weight of the transgenic and control animals. Notably, during a 2.5 year observation period, the SAC and GFP transgenic mice maintained the corresponding transgene throughout their life span (our unpublished data), and the SAC-GFP transgenic mice lived a few months longer (P < 0.01) than the GFP transgenic mice or littermate control mice. These findings indicate ubiquitous expression of the SAC domain is well tolerated in transgenic mice, and does not interfere with the development, fertility, or life span of the animals.

SAC prevents oncogene-induced transformation in murine embryonic fibroblasts (MEFs)

Our previous studies indicated over-expression of the SAC domain or Par-4 prevented cellular transformation by oncogenic Ras: apoptosis was induced in transformed cells due to phosphorylation of T155 residue by elevated levels of endogenous PKA activity (14, 55). Although the MEFs from SAC, GFP, and littermate control mice passaged normally, we sought to analyze the anti-tumor and apoptotic function of SAC on the MEFs under the conditions of oncogenic transformation.

We first tested the susceptibility of SAC MEFs to oncogenic transformation by colony formation assays in soft agar. The MEFs were transduced with adenoviral constructs encoding oncogenic Ras and c-Myc, or GFP for control. The oncogenes produced a 5–6 times more foci in the MEFs from the GFP transgenic mice and littermate control mice relative to the very few foci observed in the MEFs derived from the SAC transgenic mice (67). Quantitative analysis of apoptosis revealed infection with oncogenic Ras or co-infection with oncogenic Ras and c-Myc adenovirus resulted in 10 fold higher levels of spontaneous apoptosis in the MEFs from SAC transgenic mice relative to MEFs from GFP transgenic or littermate control mice (67). These results confirmed the anti-transformation effects of SAC protein. Given that SAC MEFs undergo apoptosis in the presence of oncogenes, we inferred the introduction of initiating oncogenic lesions prompts the SAC domain to induce apoptosis as a safeguard against cellular transformation.

Previous studies show the SAC domain induces apoptosis by inhibiting the pro-cell survival activity of NF-κB in cell cultures (14, 55). We therefore performed NF-κB reporter assays (CAT assay) to determine basal and oncogenic Ras- and c-Myc inducible NF-κB activity in MEFs isolated from the SAC, GFP, and littermate control mice. Our results showed similar basal NF-κB activity in the SAC, GFP, and control MEFs. However, upon introduction of oncogenic Ras and c-Myc, the GFP and control MEFs showed 10-fold increased NF-κB activity compare to the NF-κB in the SAC MEFs. These findings indicate that the SAC domain inhibits oncogene induced NF-κB activity in MEFs. We further tested if NF-κB pathway inhibition alone was able to induce apoptosis in transformed cells. When construct expression a phosphorylation-defective mutant of IκB-α, a known inhibitor of NF-κB (73), was introduced into normal MEFs along with oncogenic Ras and c-Myc, there was a significant increase in apoptosis in IκB/Ras/c-Myc transfected cells relative to control GFP/Ras/c-Myc transfected cells. Thus, SAC induces apoptosis in oncogene transformed cells through inhibition of the NF-κB pathway (67).

In summary, low level SAC expression in MEFs is sufficient to thwart oncogene-induced transformation. Subsequent to the transforming event, elevated PKA levels activate SAC (unpublished data), which in turn induces apoptosis. NF-κB inhibition is responsible for SAC-induced apoptosis, further proving NF-κB inhibition by SAC does not depend on proteins, such as PKCζ, that bind to the leucine zipper domain of Par-4.

SAC transgenic mice have increased resistance to spontaneous tumors

Depending on the genetic background, mice develop spontaneous tumors in distinct tissues during aging; indeed, B6C3F1 mice are reported to develop lymphomas and hepatocarcinomas (74). The majority of spontanous cancers derive from sporadic mutations in the tumor suppressor genes or protooncogenes of somatic cells, which, in turn, become neoplastic in aging mice. These spontaneous tumors more naturally mimic human cancers, which are thought to develop from a single mutated cell residing in an otherwise normal organ.

Given the transformation-preventive properties of SAC observed in vitro, we examined SAC transgenic mice for tumor incidence in their livers and spleens, as well as several other tissues. We noted a high incidence of hepatocarcinomas (liver) and lymphoma (spleen) in aging (18 months or older) GFP transgenic and littermate control mice, but not in the SAC transgenic mice. Sections of liver from control and GFP animals showed the typical histopathology of cancer: replacement of normal liver architecture by nests and cords of atypical hepatocytes; slightly enlarged hepatocytes with distended irregular nuclei; necrotic zones within the sheets of tumor cells; and prominent vascular invasion in some tumors. Conversely, liver sections from the SAC animals showed preserved hepatic sinusoidal architecture. Most of the spleens from control and GFP animals were enlarged (4–10 times) relative to the spleens from the SAC mice (our unpublished data). Sections of spleens from control and GFP animals showed effacement of splenic architecture by a diffuse infiltrate of intermediate sized lymphocytes. Most of these lymphocytes had irregular nuclear contours, and moderately condensed chromatin. There was widespread destruction of splenic white pulp by the lymphocytic infiltrate. These lymphoma cells did not have a lymphoblastic appearance. Alternatively, spleen sections from SAC animals showed preserved splenic architecture with distinct red and white pulp zones, with the white pulp areas showing preserved marginal zones. There was some variability in the size of white pulp zones, but the underlying architecture was preserved in all cases. These findings imply the SAC transgene suppresses the development and growth of spontaneous tumors in the spleen and liver of the SAC transgenic mice (67). Importantly, several generations of the SAC transgenic mice displayed resistance to developing spontaneous tumors (unpublished data), implying stable inheritance of the tumor-suppressive SAC domain.

SAC transgenic mice have increased resistance to oncogene-inducible prostate tumor growth

A common approach to studying cooperative tumorigenesis in mice is to cross a specific tumor-prone transgenic/knockout mouse with a different transgenic/knockout mouse, and assess the effect of the combined genetic background on tumor development. To determine whether physiologic levels of SAC could inhibit the growth of aggressive autochthonous tumors, we crossed SAC or GFP transgenic mice with tumor-prone TRAMP mice that produce transgenic adenocarcinoma of the mouse prostate (75). The TRAMP model offered the advantage of testing tumor suppression by the SAC domain in the absence of cellular p53 or Rb function, which is inhibited by the large T antigen of SV40. We crossed the SAC, and GFP transgenic mice with the TRAMP mice and generated the SAC^−/−/TRAMP^+/−, SAC^+/−/TRAMP^+/−, SAC^+/−/TRAMP^−/−, and GFP^+/−/TRAMP^+/− offsprings. Histopathology of the prostates indicated that, by 6 months, 100% of the GFP^+/−/TRAMP^+/− and SAC^−/−/TRAMP^+/− mice developed adenocarcinoma of the prostate. In contrast, after 6 months, only 50% of the SAC^+/−/TRAMP^+/− mice developed high-grade PIN, 21.4% showed adenocarcinoma of the prostate and 28.6% still normal. The prostate weights indicate the majority of prostates (78.5%) from the SAC^+/−/TRAMP^+/− mice weighed under 2 g, and the remaining (21.5%) weighed 2 to 3 g, whereas a majority (63–78%) of the prostates from GFP^+/−/TRAMP^+/− and SAC^−/−/TRAMP^+/− exceeded 6 g. Collectively, these data indicate the SAC domain transgene inhibits TRAMP tumor progression. Immunohistochemical analysis of the prostate sections indicates the GFP^+/−/TRAMP^+/−mice express the GFP transgene in the PIN lesions and in the adenocarcinoma of the prostate at 3 and 6 months of age. Conversely, the SAC^+/−/TRAMP^+/− mice express the SAC transgene in normal cells of the prostate, but showed a loss of SAC domain expression within the PIN lesions, and in all of the adenocarcinoma sections. As expected, the SAC^+/−/TRAMP^−/− control mice continued to express the SAC transgene in the prostates at 3 and 6 months of age. These findings imply expression of the SAC transgene must be down-modulated prior to adenocarcinoma development in the prostate. This observation was further substantiated by the finding that, by 12 months of age, 80% of the SAC^+/−/TRAMP^+/− mice develop adenocarcinoma of the prostate, the remaining 20% had PIN lesions, and all tumors showed loss of SAC domain expression (our unpublished data). This finding implies the SAC transgene functioned as a tumor suppressor in the prostate tissue, and progression to advanced disease requires loss of its tumor-suppressor function. TUNEL staining revealed the SAC^+/−/TRAMP^+/− mice had a larger number of apoptotic cells in their PIN lesions relative to the PIN lesions and adenocarcinoma in the GFP^+/−/TRAMP^+/− and SAC^−/−/TRAMP^+/− mice (67). These data suggest the SAC domain may induce apoptosis in PIN cells to avert progression to adenocarcinoma.

Significance of SAC transgenic mice

In addition to normal development and life span, our SAC transgenic mice show significant resistance spontaneous and oncogene induced cancers. These findings are consistent with the role of the SAC domain in tumor suppression.

Susceptibility to cancer is a quantitative genetic trait involving complex interactions among a large number of genes (76, 77). As variation in both the alleles and expression of tumor-suppressor genes may differentially determine cancer susceptibility in different individuals (78, 79), the SAC transgenic mice clarify the effect of a modest increase in tumor-suppressor gene expression on tumor development. Several mouse models carrying tumor suppressor genes were previously generated (71, 72, 80). Two of the p53 models are based on truncations of p53 that show anti-cancer properties, but result in shortened a lifespan (71, 72). The present SAC mice provide another independent demonstration of the concept that it is possible to increase cancer resistance without negatively impacting viability. Moreover, the SAC transgenic mouse is the first example of a cancer-resistant mouse that evades oncogene-mediated tumorigenesis by initiating apoptosis via the suppression of a defined pro-cell survival pathway.

Most of the anticancer agents used today, including chemotherapy and radiotherapy fulfill their anti-cancer effect by triggering apoptosis in cancer cells. Since normal cells have intact programmed cell death mechanisms, anti-cancer therapies often induce cell death in normal tissues as well. Therefore, achieving tumor cell specificity is a major challenge in the development of cancer treatment. As SAC selectively induces apoptosis in cancer cells upon transformation, and its apoptotic function is inactive in normal cells, SAC makes an ideal anti-cancer therapeutic agent.