The Diverse Roles of Nonsteroidal Anti-inflammatory Drug Activated Gene (NAG-1/GDF15) in Cancer

Abstract

Nonsteroidal anti-inflammatory drug (NSAID) activated gene-1, NAG-1, is a divergent member of the transforming growth factor-beta (TGF-β) superfamily that plays a complex but poorly understood role in several human diseases including cancer. NAG-1 expression is substantially increased during cancer development and progression especially in gastrointestinal, prostate, pancreatic, colorectal, breast, melanoma, and glioblastoma brain tumors. Aberrant increases in the serum levels of secreted NAG-1 correlate with poor prognosis and patient survival rates in some cancers. In contrast, the expression of NAG-1 is up-regulated by several tumor suppressor pathways including p53, GSK-3β, and EGR-1. NAG-1 expression is also induced by many drugs and dietary compounds which are documented to prevent the development and progression of cancer in mouse models. Studies with transgenic mice expressing human NAG-1 demonstrated that the expression of NAG-1 inhibits the development of intestinal tumors and prostate tumors in animal models. Laboratory and clinical evidence suggest that NAG-1, like other TGF-β family members, may have different or pleiotropic functions in the early and late stages of carcinogenesis. Upon understanding the molecular mechanism and function of NAG-1 during carcinogenesis, NAG-1 may serve as a potential biomarker for the diagnosis and prognosis of cancer and a therapeutic target for the inhibition and treatment of cancer development and progression.

Introduction

The use of aspirin and other cyclooxygenase (COX) inhibitors have been well established for the prevention and treatment of colorectal cancer. Our research and interest in NAG-1 arose from testing the hypothesis that changes in gene expression induced by COX inhibitors contributed to the prevention of colorectal cancer. From an indomethacin induced library from COX negative cells, we identified NAG-1, the most highly induced gene, by PCR based subtractive hybridization [1]. NAG-1 was identified by other groups using a variety of different cloning strategies and has several names, for example, macrophage inhibitory cytokine-1 (MIC-1) [2], placental transformation growth factor-β (PTGFB) [3], prostate-derived factor (PDF) [4], growth differentiation factor 15 (GDF15) [5], and placental bone morphogenetic protein (PLAB) [6]. NAG-1 has received considerable attention revealing a remarkable multifunctional role in controlling biological events. Not only does NAG-1 play a role in cancer development and progression, but NAG-1 also controls stress responses, bone formation, hematopoietic development, and adipose tissue function, as well as contributing to cardiovascular diseases [7]. The focus of this article is to discuss the diverse and conflicting roles of NAG-1 in cancer development and progression and to discuss if COX inhibitor-induced expression of NAG-1 can contribute to the cancer prevention observed with NSAID usage.

Biochemistry of NAG-1

NAG-1 is a divergent member of the TGF-β superfamily with an amino acid sequence similar to the bone morphogenic protein (BMP) genes. The human NAG-1 locus has been mapped to 19p12.1–13.1 [8] and the NAG-1 protein is encoded by two exons. The 309 bp Exon I contains a 71 bp 5′ untranslated region (UTR) and a 238 bp coding region, and the 647 bp Exon II contains a 3′ UTR. The gene contains a single 1820 bp intron [8]. The NAG-1 pro-domain consists of 167 amino acids and contains an N-linked glycosylation site at amino acid position 70 [9]. After dimerization of the full length pro-NAG-1 precursor by a specific disulfide linkage, the dimeric pro-protein undergoes proteolytic cleavage catalyzed by furin-like protease at the amino acid target sequence RXXR resulting in the release of a 112 amino acid C-terminal dimeric mature region. The mature dimer is then secreted into the extracellular media (Figure 1). Recently, it has been reported that the pro-domain selectively binds to an extracellular matrix [10]. NAG-1 may have multiple forms possibly present within the cell: the pro-NAG-1 monomer (~40kD), the pro-NAG-1 dimer (~80kD), the pro-peptide the N-terminal fragment after cleavage (~28kD) and the mature dimer (~30kD) (Figure 1). The presence of different forms in the cell, coupled with the resistance of the dimer to reduction, can often make analysis of the expression by western blot a challenge to correctly identify the forms expressed.

Figure 1. Dimeric formation of mature NAG-1 and different forms of NAG-1 in cells.

The pro-NAG-1 was cleaved at RXXR site and then secreted out the cells as a dimer. NAG-1 pro-peptide is also secreted out of cells.

The mature NAG-1 has 7 cysteine residues with 6 cysteines likely forming a cysteine knot, a key structural characteristic of members of the TGF-β superfamily. The seventh cysteine forms a disulfide linkage to a second molecule of NAG-1 forming a homodimer. The secreted dimer is present in the serum and secreted into the media of cultured cells that expressing NAG-1. The mature dimer is highly glycosylated and shares very little of its identity with other TGF-β superfamily proteins. There is some evidence for the presence of the pro-form of NAG-1 as well as the pro-peptide in the media of cultured cells [10]. Molecular modeling based on the known structure of other TGF-β members suggests that the three dimensional structure is most like GDF-8 or myostatin, however the NAG-1 crystal structure has not been reported [11].

TGF-β members bind to form a complex between Type-I and Type-II receptors. Although seven Type-I and five Type-II receptors have been identified for the TGF-β superfamily, the specific receptor for NAG-1 remains to be identified. Some studies suggest that the mature dimer can activate TGF-β response elements [12]. In addition, the activation of other intracellular signaling pathways, for example, the MAPK and EGFR/ErbB signaling pathways [13, 14], are reported to be activated by NAG-1. Some evidence suggests that the active form of NAG-1 is the mature secreted dimer. However with all the different forms biosynthesized and the potential for interactions between these forms (binding partners), it is likely the mature dimer, pro-forms, and pro-peptides of NAG-1 play a central role in modulating the biological activity of NAG-1.

The murine NAG-1 gene was also identified and characterized [5]. The human NAG-1 and murine NAG-1 genes both contain two exons, which encode 308 amino acids protein (human) and 303 amino acid protein (mouse), respectively. However, the tissue distribution of mouse NAG-1 protein is different from human [15]. The human NAG-1 is expressed in the prostate, colon, placenta, and poorly or not at all expressed in the liver [3], whereas the mouse NAG-1 is highly expressed in the liver but not in the prostate, colon and placenta [15]. In addition, sequence comparison between the human and mouse NAG-1 promoters in the ~700 bp region revealed only 39% homology [16], possibly explaining the different expression pattern of NAG-1 at the transcriptional level between human and mouse. Further, the differences in the N-terminal region of NAG-1 peptide sequences in human and mouse may contribute to different regulation of expression and even alter the biological activity of NAG-1. Since the C-terminal region of the NAG-1 peptide sequence is conserved in the cysteine residues of human and mouse gene, it is assumed that mouse NAG-1 may also form a dimer. The crystal structures of human or mouse NAG-1 have not been solved although modeling predicts the structure of human NAG-1 would be similar to the structure of GDF-8 [11]. Further investigations are needed to elucidate whether different expression pattern and structure between human and mouse NAG-1 may have an impact on its biological functions.

Yamaguchi et al. cloned the canine NAG-1 gene and investigated its expression in canine tissues [17]. The predicted canine NAG-1 amino acid sequence revealed nine cysteine residues and an RXXR sequence that was conserved in the human, mouse, and rat, suggesting that canine NAG-1 protein may have similar biological activity to other species. The canine NAG-1 is induced by several NSAIDs, with the most robust up-regulation by piroxicam in osteosarcoma cells [17]. The secreted and pro-forms of NAG-1 were detected in canine tissues by western blot analysis. Canine cancer models could be a useful tool to study NAG-1 expression as related to cancer development.

Regulation of expression

Transcriptional regulation of NAG-1 has been extensively investigated by our group. NAG-1 is up-regulated in human colorectal cancer cells by several NSAIDs [18], as well as by dietary compounds, including resveratrol [19], genistein [20], diallyl disulfide [21], conjugated linoleic acid [22], green tea catechins [23], epigallocatechin-3-gallate (EGCG) [24], indole-3-carbinol [25], capsaicin [26], damnacanthal [27], PPARγ ligands [28, 29], and 1,1-Bis(3′-indolyl)-1-(p-substituted phenyl) methanes [30]. NAG-1 expression was also seen in other cancer cells by anti-cancer compounds [31–33]. A very diverse number of chemicals with a wide range of chemical structures induce the expression of NAG-1, suggesting multiple mechanisms responsible for the increase in expression. We have characterized the human NAG-1 promoter, which contains several cis-acting and trans-acting elements [16]. Sp1 transcription factors regulate the basal transcription of NAG-1 through the GC box located within −133 bp of the NAG-1 promoter, whereas p53 sites play a pivotal role in dietary compound-induced NAG-1 expression. Two p53 sites are located within the −133 bp promoter with a third site located in the 5′ UTR [3, 19]. Furthermore, several COX inhibitors and PPARγ ligands induce NAG-1 expression at the transcriptional level via EGR-1 transcription factors [34, 35]. Recently, we have identified that the transcriptional factor C/EBPβ contributes to NAG-1 induction mediated by capsaicin and damnacanthal [27, 36]. Figure 2 summarizes the transcriptional regulation of NAG-1 by NSAIDs and dietary compounds through different transcriptional factors. Collectively, NAG-1 is regulated by multiple mechanisms suggesting that NAG-1 could be a molecular target for cancer chemoprevention.

Figure 2. Transcriptional regulation of NAG-1 by NSAIDs and dietary compounds.

NAG-1 promoter contains several cis-acting and trans-acting elements. Both Sp1 and EGR-1 transcription factors have been identified to regulate the basal transcription of NAG-1. Two p53 sites that located within the -133 bp promoter play a pivotal role in dietary compound-induced NAG-1 expression.

Epigenetic regulation of expression

Whether NAG-1 expression is epigenetically regulated has been studied in glioblastoma cell lines. We first examined whether histone modification plays a role in NAG-1 expression. We found that the histone deacetylase inhibitor, trichostatin A (TSA), induces NAG-1 promoter activity and induces NAG-1 expression [37]. Further studies suggested that TSA-induced NAG-1 expression not only involves the interaction with the transcriptional factors Sp1 and EGR-1 at transcriptional level, but also the increase of mRNA stability at post-transcriptional level [37].

Aberrant promoter hypermethylation is a common mechanism for silencing tumor suppressor genes in cancer cells. Previous work shows that the NAG-1 promoter has several CpG islands [38]. In glioblastoma cell lines, basal NAG-1 expression was increased by the demethylating agent, 5-aza-2′-deoxycytidine. The NAG-1 promoter was densely methylated in several glioblastoma cell lines as well as in primary oligodendroglioma tumor samples, which have low basal expression of NAG-1 [38]. DNA methylation at two specific sites (−53 and +55 CpG sites) in the NAG-1 promoter was strongly associated with lower NAG-1 expression. The methylation of the NAG-1 promoter at the −53 site blocks EGR-1 binding and thereby suppresses NAG-1 induction. Pre-incubation with 5-aza-2′-deoxycytidine increased NAG-1 basal expression, and subsequent incubation with a NAG-1 inducer increased NAG-1 expression [38]. Thus, methylation of specific promoter sequences may cause transcriptional silencing of the NAG-1 locus in gliomas and may ultimately contribute to tumor progression. However, many other tumors and cells are reported to highly express NAG-1. While the methylation status is unknown in other tumors, this may be due to the lack of CpG island methylation in NAG-1 overexpresing tumors. Further studies are necessary to clarify the conflicting data on the expression of NAG-1 in tumors and the possible link to CpG island methylation.

Determining NAG-1 expression in tissue

NAG-1 expression in normal and transformed tissue has been reported in a number of publications as reviewed by Mimielle and Batra [7]. However, there is no clear consensus about the expression levels in tumors compared to normal tissue although most data indicate higher expression in tumors relative to normal tissues. One consideration is the different methodologies used to measure NAG-1 expression by different investigators. The specificity of antibodies used to measure expression in many reports is frequently not clearly stated. For example, the use of an antibody that detects the monomer form but poorly reacts with the dimer form could yield conflicting expression data as compared to the use of an antibody that reacts well with the dimer but poorly with the monomers. Because pro-NAG-1 is cleaved at the RXXR site, the activity of the cleaving enzyme can influence the level of NAG-1 inside the cell as the cleaved NAG-1 is rapidly secreted. It is possible that analysis of the cell lysate would yield higher NAG-1 expression than in cells where cleaving activity is higher. However, recent studies did not examine the activity of the cleaving enzyme when analyzing NAG-1 expression. Thus, reports of NAG-1 expression by measurement of protein expression should be viewed with caution.

Determination of gene copy number can be used to compare the expression of NAG-1 between different cells in culture and to determine the expression level in normal and tumor tissues. In a recent publication we measured the expression of NAG-1 in glioma cell lines and in normal and glioblastoma tumor samples [38]. In 11 out of 12 tumor samples the gene copy number was significantly lower than the gene copy number observed in the normal tissue and was in general agreement with the expression of the pro-NAG-1 protein expression in the tissue as measured by Western analysis. For the low grade glioma cell lines the gene copy number was 5 to 10 times higher than the gene copy number for the glioblastoma cell lines. The correlation between the gene copy number and the expression of the pro-NAG-1 in the cells and concentration of secreted NAG-1 as determined by ELISA was inconsistent. In some cells most of the NAG-1 was the secreted NAG-1 in the media while in other cells most of the NAG-1 remained as the pro-NAG-1 inside the cells. Thus, we propose the measurement of gene copy number is a better estimate of NAG-1 expression in tissues.

Complex roles of NAG-1 in cancer development and progression

The role NAG-1 plays in the development and progression of cancer is complex and poorly understood. Some experimental evidence suggests that NAG-1 has tumor suppressor activity, while other data suggests that it has oncogenic activity. The anti-tumorigenic and pro-tumorigenic effects of NAG-1 on tumor growth appear to be dependent on the type of cancer and the stage of the cancer. The following is a summary of the experimental evidence supporting the anti- and pro-tumorigenic activities of NAG-1.

Inhibition of tumor formation

1. The overexpression of NAG-1 in cancer cells HCT116 [1], MCF-7 [39], PC-3 [40], and glioblastoma [41] inhibits the growth of tumors in nude mice in xenograft models. Furthermore, the expression of NAG-1 induces apoptosis in several cancer cells in vitro [42]. Many drugs and chemicals including COX inhibitors with documented cancer prevention activity induce the expression of NAG-1 in a number of different cells in vitro [42]. Investigations to determine the mechanisms for this increased expression reveal that known tumor suppressors may regulate expression of NAG-1. Activation of the tumor suppressor genes p53 [3], EGR-1 [35], GSK-3β and C/EBPβ [36] are required to increase NAG-1 expression as mentioned above. This is indirect evidence supporting the notion of NAG-1 acting as a tumor inhibitor.

2. However, evidence supporting inhibition of cancer formation by NAG-1 comes from experiments with a transgenic mouse expressing hNAG-1 ubiquitously. After treating mice with the intestinal carcinogen AOM, a reduced number of foci were observed in the hNAG-Tg mice as compared to wild type mice [43]. Furthermore, NAG-1 Tg mice bred to the Apc^min mice also had a lower number of observed polyps. Collectively, both chemically and genetically induced intestinal cancer is lower in the NAG-1 Tg mice. In addition, the NAG-1 Tg mice were also reported to be resistant to urethane induced lung tumors [44]. More recently, an inhibition of prostate tumorigenesis was observed in the transgenic adenocarcinoma of the mouse prostate (TRAMP) model of prostate cancer after crossing the TRAMP mouse with NAG-1 transgenic mouse [45]. These findings indicate that NAG-1 may act as a tumor suppressor in the early stages of tumor development.

3. The Apc^min mouse was also bred to a mNAG-1 knockout mouse to yield a Apc^min mouse not expressing mNAG-1 [46]. The deletion of the NAG-1 gene did not alter the spontaneous development of intestinal polyps observed in the Apc^min mouse. However, inhibition of polyp formation was only observed in the wild type Apc^min expressing mNAG-1 after treatment of these mice with the COX inhibitor sulindac suggesting that tumor inhibition by sulindac was dependent on the expression of mNAG-1 [46]. This finding also supports the hypothesis for NAG-1 potentially acting to inhibit tumor growth at the early stages of cancer.

Pro-tumorigenic activity

1. The expression of NAG-1 has been reported to be highly regulated in tumors of human cancer samples [7]. Furthermore, the serum concentration of NAG-1 in human cancer patients is high, with the serum levels associated with declining patient survival. Measurement of the secreted form of NAG-1 has been proposed as a marker for cancer progression and risk assessment [7].

2. In several mouse xenograft studies, human NAG-1 is reported to enhance tumor growth. For example, Boyle et al. showed inhibition of NAG-1 expression by shRNA inhibits melanoma growth in xenografts [47]. Orthotopically implanted PC-3 cells engineered to express NAG-1 developed more metastases than PC3 vector cells [48]. Furthermore, overexpression of NAG-1 in these cells enhanced migration and invasion of PC-3 cells [48]. Another study with LNCaP androgen-independent variants indicated that NAG-1 acts to promote cancer development [49]. Recently, the expression of NAG-1 in the TRAMP model was reported to inhibit prostate tumor growth but the expression of NAG-1 increased metastases to distant organs [45]. These findings suggest NAG-1 may act to promote cancer growth and progression.

3. The addition of recombinant NAG-1 or the forced expression of NAG-1 can stimulate cell proliferation. NAG-1 is reported to stimulate the growth of several gastric cell lines mediated by the activation of the ERK1/2 pathway [50]. Also, NAG-1 was reported to activate the AKT and ERK1/2 pathways in human breast and gastric cells by the transactivation of ErbB2/ HER2 oncogene [14]. These studies suggest that NAG-1 may act as a positive regulator of cell growth in HER-2 over-expressing tumors.

Collectively, both the anti-tumorigenic and pro-tumorigenic activity of NAG-1 is supported by experimental evidence. The intuitive response after considering the high serum levels observed in cancer patients is to conclude that the highly expressed protein may be a driving in tumor growth. However, an explanation may be that tumor cells resistant to NAG-1 expression during the events of progression. These resistant cells proliferate in the developing tumor, and because NAG-1 expression increases with stress, higher secreted NAG-1 is observed in the serum. Therefore, serum NAG-1 levels may be a reflection of tumor size and stage, indicating patient prognosis and survival.

NAG-1 and colorectal cancer

Colorectal cancer is the third most common cancer and leading cause of cancer death in the United States. The role of NAG-1 in colorectal cancer tumorigenesis is by far unclear. It has been reported that NAG-1 levels are increased in the serum of colorectal cancer patients [51]. The serum level of NAG-1 correlates with the development of adenomatous polyps and was proposed as a prognostic marker for disease progression and recurrence [51]. However, patients who had used NSAIDs also had a higher serum level of NAG-1 associated with protection from the recurrence of colon adenoma [52]. Most in vitro studies related to NAG-1 function in colorectal cancer suggest a tumor suppressor role of NAG-1. For example, many anti-cancer compounds increased NAG-1 expression in colorectal cancer cells [53–55] and tumor suppressor proteins control the NAG-1 expression [56]. Recent data suggests that NAG-1 is a downstream target of ER stress, mediating ER-stress-induced apoptosis [57].

The anti-tumorigenic activity of NAG-1 in colon cancer was more evident in in vivo studies using NAG-1 transgenic mice. To examine if NAG-1 expression provides resistance to colorectal carcinogens or genetic colorectal cancer models, we developed transgenic mice (NAG-1 Tg, C67/BL6 background) expressing human NAG-1. NAG-1 Tg mice and control siblings were treated with azoxymethane (AOM) and aberrant crypt foci (ACF) were counted. An approximate 50% reduction in ACF was observed after AOM treatment in NAG-1 Tg mice, indicating that NAG-1 expression suppresses AOM-induced tumorigenesis [43]. NAG-1 Tg mice were also crossed with Apc^Min mice, to generate mice expressing NAG-1 on the Apc^Min background, and polyp formation was measured in their intestines. 40% inhibition of polyp formation in the intestine, compared to control littermates was found. These results indicate that NAG-1 is a potential tumor suppressor gene in carcinogenic- and genetic-induced colorectal cancer animal models. Our understanding of the molecular pathways and mechanism responsible for the apparent paradoxical action of NAG-1 in colorectal cancer has been examined. Many in vitro studies show the expression of NAG-1 induces apoptosis in colorectal tumors. Often colorectal tumors contain non-cancerous cells, including immune cells and vascular cells that are important in inflammation. Chronic colitis is associated with an increased risk of developing colorectal cancer, and the susceptibility to cancer increases when the tissue is chronically inflamed [58]. The link between inflammation and colorectal cancer is strong with interplay between the inflammatory cells to the development and progression of cancer is critical. NAG-1 is reported to inhibit inflammatory cytokines from Lipopolysaccharide (LPS) treated macrophages, suggesting NAG-1 may exert an anti-tumorigenic effect by lowering inflammation [2]. In preliminary experiments we found the serum levels of inflammatory cytokines after treatment with LPS were lower in the NAG-1 Tg mice as compared to wild type littermates, suggesting NAG-1 Tg mice have a lower inflammatory response. However, in contrast to the previous report [2], we could not confirm the inhibition of LPS-induced cytokine formation by NAG-1 in macrophages, suggesting other more complex mechanisms are involved..

NAG-1 and lung cancer

Lung cancer is the leading cause of cancer-related death in men and women in US and pulmonary adenocarcinoma (PAC) is the most common type of lung cancer. Unlike colorectal cancer, NAG-1’s role in lung cancer has not been studied well. Newman et al. reported for the first time that NAG-1 is increased in the presence of retinoids [32]. Subsequently, other researchers reported that NAG-1 plays an important role in retinoid-induced anti-tumorigenesis [59], isochaihulactone-triggered apoptotic pathway [60], and sodium salicylate-induced apoptosis [61] in lung cancer cells. These results suggest that NAG-1 exhibits anti-tumorigenic activity in lung cancer, as assessed by in vitro assays. In vivo assays were performed and confirmed NAG-1’s anti-tumorigenic properties in lung cancer. Urethane (ethylic ester of carbamic acid) is a carcinogen which specifically promotes the development of lung tumors from alveolar type II pneumocytes in rodents [62]. Among the many animal models available for the analysis of human lung adenocarcinoma, urethane-induced lung tumorigenesis in mice is thought to be one of the most useful because of its faithful reproducibility, histological similarity, and time-dependent progression from hyperplasia through adenoma and eventually to adenocarcinoma [63]. Our group has developed the over-expressing NAG-1 Tg mice on the FVB background as FVB strains, compared to other strains of mice, are very sensitive to urethane. Lung tumors were induced by urethane injection and lung tissues were histologically examined. Control mice exhibited many lung tumors in their lungs. However, NAG-1 Tg mice had fewer lung tumors, suggesting that NAG-1 can act as a tumor suppressor in this model [44]. Interestingly, NAG-1 Tg mice had a lower frequency of inflammatory cells in the lung tissue as assessed by lysozyme expression. The reduced inflammation and tumor burden in the lung of NAG-1 Tg mice may be mediated by the down-regulation of the p38 MAPK signaling pathway. We also found higher activation of caspase 3/7 in the NAG-1 Tg mice in lung tissue [44]. Consistent with these findings, Yu et al. suggested that NAG-1 is a molecular target for isochaihulactone-induced anti-tumorigenesis in lung cancer, as assessed by in vitro and in vivo assays [64]. These data suggest that NAG-1 plays an important role in inflammation and lung tumorigenesis in vivo.

NAG-1 and Pancreatic cancer

Pancreatic cancer is a major cause of cancer-related deaths in developed countries and has the highest mortality rate among major cancers. Pancreatic cancers may cause only vague symptoms before being detected and chemotherapeutic regimens for this disease have provided very limited improvements in tumor regression and overall survival rates after diagnosis [65]. Although the precise pathogenesis of pancreatic cancer remains unclear, common mutations in several cell proliferation-related genes have been described: mutation of K-ras, p16, p53, and Smad4 genes have been identified in sporadic pancreatic tumors [66]. Since conventional therapeutic approaches do not decrease mortality of this deadly cancer, we have paid more attention to alternative research including identification of molecular target approaches to increase survival rate.

NAG-1 is induced by several anti-cancer compounds such as PPARγ ligands in pancreatic cancer cells [67, 68]. NAG-1 expression plays an important role in synthetic triterpenoids derived from glycyrrhetinic acid-induced anti-tumorigenesis in pancreatic cancer cells [69]. Consistently, many other compounds also increase NAG-1 expression in pancreatic cancer cells including NSAIDs (NS-398 and tolfenamic acid) [70]. The mechanisms by which these compounds increase NAG-1 expression include activation of KLF4, EGR-1, and GSK-3β pathways. For example, MCC-555 increases the tumor suppressor KLF4, which binds to the NAG-1 promoter, thereby initiating NAG-1 transcription [68]. In conclusion, NAG-1 could play an important role in pancreatic tumorigenesis. The induction of NAG-1 by various compounds in pancreatic cancer cells suggests NAG-1 can be an attractive target for pancreatic cancer chemoprevention.

NAG-1 and prostate cancer

Prostate cancer is the most commonly diagnosed cancer and the second leading cause of cancer-related deaths in men in the United States. Despite the clinical importance of prostate cancer, the molecular mechanisms underlying the development and progression of this disease remain unknown. Many efforts have been made to establish the role of NAG-1 in prostate cancer development and progression. However, reports in the literature are contradictory and thus make the role of NAG-1 in prostate carcinogenesis elusive. In general, significant data from in vitro and in vivo laboratory studies have shown that NAG-1 exhibits anti-proliferative, pro-apoptotic, and thus anti-tumorigenic activities, but clinical data suggest that NAG-1 maybe pro-tumorigenic [42]. NAG-1 has been shown to induce growth arrest in DU145 human prostate carcinoma cells [12] and induce apoptosis involving caspase-3 activation in DU145 cells but with no effects on proliferation [71]. Forced expression of NAG-1 in PC-3 prostate carcinoma cells inhibited proliferation and the growth of these cells in a xenograft tumor model [40]. Chiu et al. found that NAG-1 induction by isochaihulactone is responsible for isochaihulactone-induced cell death in LNCaP prostate cancer cells [31]. Wynne and Djakiew found that NSAID inhibition of prostate cancer cell migration is mediated by NAG-1 induction through the p38 MAPK pathway in PC-3 cells [72]. More recently, TRAMP mice were crossed with NAG-1 overexpressing mice to examine the effects of NAG-1 on tumor development and progression. The study showed that overexpression of NAG-1 in TRAMP mice significantly reduced tumor size and lowered tumor grades compared to the TRAMP control mice. However, the NAG-1overexpressing TRAMP mice developed more distant organ metastases suggesting a complex role of NAG-1 in prostate tumorigenesis in the TRAMP model [45].

Although most laboratory studies suggest an anti-tumorigenic role of NAG-1 in prostate carcinogenesis, a few studies showed that NAG-1 may promote prostate tumorigenesis. Overexpression of NAG-1 in prostate cancer PC-3 cells has been shown to increase migration and invasion of these cells [48]. The authors also found that NAG-1 expression increases metastases to distant organs in PC-3 orthotopic prostate model in the nude mouse [48]. Chen et al. found that NAG-1 promotes cell proliferation of LNCaP cells through ERK activation [73]. Thus, NAG-1 seems to also work as a pro-tumorigenic protein. While laboratory studies in general suggest an anti- tumorigenic activity of NAG-1that induces growth arrest or apoptosis, clinical studies demonstrate that NAG-1 expression is up-regulated in human prostate cancers which may also correlate with tumor grade and progression [42]. Studies examined the association between plasma levels of NAG-1 and status of prostate tumor progression. The plasma levels of the secreted mature protein are greatly elevated in patients from several studies [51, 74]. In particular, the plasma level of NAG-1 has been positively associated with prostate cancer metastasis [48, 74-76]. Therefore, measurement of the secreted NAG-1 in the blood has then been proposed as a diagnostic marker for prostate cancer [77] and a measure of prostate cancer progression [74, 78]. However, a possibility exists that higher NAG-1 concentrations in the blood are a reflection of tumor size and not an indicator that NAG-1 is acting to enhance tumor growth.

Other efforts have been made to determine the role of NAG-1 polymorphism during prostate carcinogenesis. Three nonsynonymous single nucleotide polymorphisms (nsSNPs) exist in the gene that causes amino acid changes in the coding region. A common C to G (Exon 2+2423) substitution (histidine to aspartic acid) at codon 202 of the precursor protein is commonly called H6D because the amino change is at position 6 of the mature NAG-1 protein (rs1058587) [79]. A large study of 1340 prostate cancer cases and 765 controls in Sweden suggested the G allele (the H6D/NAG-1) is associated with decreased risk of developing prostate cancer [80]. A second large study involving 819 cases and 731 controls in Australia had similar findings, although these findings were not statistically significant [81]. However, results from this study also suggest a higher mortality rate from prostate cancer for patients carrying the G allele relative to men with the CC genotype. Similarly, a case control study (506 controls and 506 cases) in the United States found that the G allele is marginally associated with a lower prostate cancer incidence, although this was statistically insignificant [82]. Recently, our laboratory examined the tumor inhibitory effects of H6D/NAG-1 on DU145 xenografts in nude mice. We found that mice with tumors expressing the H6D/NAG-1 have significantly smaller tumor weights and slower growth compared to the control mice, [11] suggesting a potential anti-tumorigenic role of H6D/NAG-1 during prostate cancer development. A few studies also examined the association of other NAG-1 SNPs with prostate cancer risk and mortality. However, these data in general did not support an association like H6D/NAG-1. Collectively, NAG-1 polymorphisms, especially the H6D/NAG-1, may play an important role in human prostate cancer carcinogenesis. However, the function of NAG-1 in prostate cancer remains controversial and its signaling pathways remain poorly understood. NAG-1 may play an anti-tumorigenic role at the early stages of carcinogenesis, but a pro-tumorigenic one during cancer progression. The exact mechanism of this apparent dichotomy of NAG-1 during prostate carcinogenesis is not clear at present and needs to be elucidated in future studies.

NAG-1 and gastric cancer

Unlike the extensive studies of NAG-1 in prostate and colorectal cancers, studies in gastric cancer are limited. However, similar to findings from prostate cancer studies, the role of NAG-1 in gastric cancer carcinogenesis is also controversial. Few clinical studies found that NAG-1 expression is up-regulated in the serum of gastric cancer patients and its expression is strongly associated with cancer metastasis, suggesting an oncogenic role for NAG-1 during gastric cancer progression [83]. Interestingly, in vitro studies using stably transfected cells suggest that the overexpression of NAG-1 induces the invasiveness of gastric cancer SNU-216 cells through the upregulation of urolinase-type plasminogen activator system and the ERK1/2 MAPK kinase pathway [50]. Kim et al. also showed that the overexpression of NAG-1 induces the transactivation of ErbB2 in gastric cancer SNU-216 cells and activates ERK1/2 and AKT signaling cascades [14]. In contrast, NAG-1 induction upon NSAID treatment has been reported to induce apoptosis in gastric cancer cells [84, 85], suggesting a tumor suppressor role for NAG-1 in gastric cancer development. In addition, administering celecoxib to gastric cancer patients significantly induced NAG-1 expression in tumor samples and inhibited gastric adenocarcinoma growth compared to the control patients [86]. Huang et al. also found that treating gastric cancer patients with celecoxib significantly induced NAG-1 expression in tumor samples which may be responsible for celecoxib induced apoptosis and the reduction of microvessel density in the tumor samples of treatment groups compared to the control patients [87]. These findings raise the question of whether NAG-1 induction plays a role in NSAID-induced inhibition of cancer development which will be addressed below.

Role of NAG-1 in the prevention of cancer by NSAIDs

NSAIDs are the most widely used drugs for treatment of inflammatory diseases and long-term use of NSAIDs prevents the development of several types of cancer [88, 89]. Both COX-dependent and COX-independent mechanisms have been proposed for the chemopreventive and anti-tumorigenic activities of NSAIDs. NAG-1 expression is up-regulated by several NSAIDs in a COX-independent manner in human cancer cells. As mentioned above, NAG-1 was first identified by our laboratory from indomethacin-treated COX-deficient human colorectal cancer HCT116 cells [1]. Celecoxib has been shown to induce apoptosis in COX-2-deficient human gastric cancer cells via activation of NAG-1 expression and inhibition of AKT/GSK3β [85]. Sulindac sulfide significantly induced NAG-1 expression in gastric cancer SNU601 cells that are devoid of COX-2 expression, increased apoptosis and decreased cell viability in this cell line [84]. In addition, neither COX expression nor the level of PGE₂ and/or arachidonic acid affects NSAID-induced NAG-1 expression [18]. These studies suggest a COX-independent mechanism for the anti-tumorigenic effects of NSAIDs which may mediated by increased NAG-1 expression. Other than the induction of NAG-1 by NSAIDs in cell culture models, a number of studies reported NAG-1 expression was induced in animal models. Feeding C57/BL6 mice sulindac induced mNAG-1 expression in liver and colon tissues [90, 91]. Indomethacin treatment has been shown to induce the expression of NAG-1 mRNA in human sinonasal cancer cell AMC-HN5 xenograft tumors in mice in a dose-dependent manner [92]. The volume of the xenograft tumors in indomethacin-treated nude mice was reduced compared to that in control mice. In another study, celecoxib treatment increased NAG-1 expression in a dose-dependent manner in COX-2 knockout mice and wild type littermates (COX-2^+/+) [93]. NAG-1 induction upon NSAID treatment in animal models suggests that NAG-1 may be important for the anti-tumorigenic activity of NSAIDs in humans.

Many studies have shown that sulindac fed to Apc^Min mice inhibits polyp formation. However, the contribution of NAG-1 expression to the prevention of polyp formation by sulindac has not been determined. Zimmers et al. crossed Apc^Min mice with NAG-1(−/−) mice, which did not alter polyps formation [46]. In this study, sulindac was effective in reducing the polyp formation in Apc^Min mice that express wild-type NAG-1. However, sulindac did not reduce polyp formation in NAG-1(−/−) crossed with Apc^Min mouse [46]. This finding suggests that NAG-1 is critical for anti-tumorigenic activity of sulindac in the Apc^Min mouse model.

Knocking down NAG-1 in cell culture models was also used to elucidate the role of NAG-1 in NSAID-induced inhibition of cancer cell growth. In one study, sulindac sulfide induced NAG-1 expression in ovarian cancer SKOV3 cells and significantly suppressed cell growth [94]. Transfecting SKOV3 cells with the NAG-1 small interfering RNA (siRNA) construct restored SKOV3 cell viability, suggesting sulindac sulfide-induced NAG-1 expression is responsible for this sulindac sulfide-induced cell growth arrest [46]. By treating human prostate cancer PC-3 cells with NAG-1 siRNA, Wynne and Djakiew demonstrated that NAG-1 plays an important role in NSAID-induced inhibition of PC-3 cell migration [72]. Indomethacin induced apoptosis and NAG-1 expression in human sinonasal carcinoma AMC-HN5 cells, in which the indomethacin-induced apoptosis was suppressed by transfecting the cells with NAG-1 siRNA [95]. Collectively, these reports suggest that a major part of NAG-1’s function is to provide for NSAID-induced inhibition of tumorigenesis both in vivo and in vitro. Recently, results from one clinical study found that NSAID users have higher serum NAG-1 level which was related to preventing adenoma recurrence in cancer patients [52]. In two other studies, NAG-1 expression was significantly induced in tumors upon celecoxib treatment in gastric cancer patients. They observed increased apoptosis and microvessel density reduction [87] and that NAG-1 induction might be responsible for the inhibition of gastric adenocarcinoma growth [86]. However, further studies are needed to confirm this finding in clinical studies. In conclusion, the prevention of tumor growth by NSAIDs is very complex, targeting both COX-prostaglandin pathway and NSAIDs-induced gene expression as illustrated by increased expression of NAG-1.

Prospective and future directions

Considerable advancement has been made in understanding the biological actions of NAG-1 and the roles this unique member of the TGF-β family plays in physiological processes and in the development and progression of cancer. Despite these advances, the mechanisms responsible have not been elucidated. One underlying problem that impedes progress is a complete understanding of the biological activity of the multiple forms of this protein that are present in and secreted from the cell. Studies with the purified secreted protein have often yielded conflicting data and results that could not be confirmed in a second laboratory. Determining the biological activity of the different NAG-1 forms needs to be done in future studies. Also, it may be necessary to determine if the secreted dimer requires a binding partner for its activity. This has been observed with other members of the TGF-β family. Future studies are also needed to identify the NAG-1 receptor(s). Some evidence suggests the receptor may be related to the TGF-β receptor, which is a complex of type I and type II receptors. Other future studies will need to determine the downstream signaling pathways once NAG-1 binds to its receptor. Better understanding the nature of the receptor(s) and the downstream signaling pathways may provide the insight to the how the protein can inhibit cancer at the early stages, yet promote cancer progression in the later stages of cancer.