Downregulation of PAR-4, a pro-apoptotic gene, in pancreatic tumors harboring K-ras mutation



Oncogenic ras is known to inhibit cell death and growth inhibitory genes and activate prosurvival genes. Proapoptotic gene PAR-4, has been found to be downregulated by oncogenic ras. Since pancreatic tumors harbor a high incidence of K-ras point mutations, we hypothesized that oncogenic K-ras might influence the function and expression of PAR-4. PAR-4 expression levels were analyzed in 4 established pancreatic tumor cell lines, 10 normal pancreatic tissues, 44 frozen tumor tissues and 25 paraffin-embedded pancreatic adenocarcinoma samples by Real Time RT-PCR, Western blot analysis and immunohistochemistry. K-ras mutational status was analyzed by allele-specific oligonucleotide-hybridization. Expression levels of PAR-4 were correlated with the K-ras mutational status and clinical characteristics. Further, modulation of endogenous PAR-4 was tested by transiently expressing oncogenic ras in a wild-type K-ras pancreatic cancer cell line, BxPC-3. Three cell lines with K-ras mutations showed low levels of PAR-4 when compared to a normal pancreatic tissue. Of 44 frozen tumors, 16 showed appreciable upregulation of Par mRNA and 27 showed significant downregulation of PAR-4 mRNA when compared to normal pancreatic tissue and 1 had levels equivalent to normal pancreatic tissue. Of 25 paraffin-embedded tumors, 9 showed downregulation of PAR-4 protein and this downregulation of PAR-4 correlated significantly with K-ras mutational status (p < 0.00002). In addition, the presence of PAR-4 mRNA or protein expression in pancreatic tumors correlated with prolonged survival. Transient overexpression of oncogenic ras in wild-type K-ras BxPC-3 cells significantly downregulated the endogenous PAR-4 protein levels and conferred accelerated growth. Thus, downregulation or loss of PAR-4 expression by oncogenic ras may provide a selective survival advantage for pancreatic tumors, through inhibition of proapoptotic pathway mediated by PAR-4. © 2007 Wiley-Liss, Inc.

The process of apoptosis requires coordinate expression of specific genes. Gene function studies in diverse experimental models have identified several gene products as mediators of apoptosis which include: transcription factors (Egr-1, c-myc, fos, jun); bcl-2-related proteins (bax, bak, bad and bcl-x)1, 2; and cysteine proteases, interleukin-1-converting enzyme,3 and ICH-1L,4 in various mammalian cell systems. In addition to gene induction, apoptosis can result from downregulation of genes, such as bcl-2 and bag-1 that normally provide the cell with protection from apoptotic death.

Pancreatic cancer is one of the fourth most common cause of cancer death in the United States5 with an yearly incidence of 30,700 cases. The survival rates of pancreatic cancer patients are low compared with other malignancies and recurrence and metastisis after treatment remain high. Pancreatic tumors develop through sequential progression from Squamous (transitional) Metaplasia to Pancreatic Intraepithelial Neoplasia 3. Current studies have focussed on the role of p536, 7 members of the bcl-2 family1, 2 and APC gene8 in the dysregulation of apoptosis that forms a primary mechanism of pancreatic ductal cell carcinogenesis. A high proportion of pancreatic adenocarcinomas show mutations in p53 gene and may contribute towards the aggressiveness of the tumor.9 BCL-xL is overexpressed in these tumors suggesting that this may provide an altered threshold for the induction of apoptosis eventually permit clonal expansion of tumors, particularly, in cells containing mutant p53. A range of pancreatic abnormalities was observed (including adenocarcinoma to 22%) in p53 null mice with heterozygous to complete loss of APC gene. Although these studies indicate that the dysregulation of apoptosis may occur in pancreatic carcinogenesis, but the mechanism is not clearly understood.

K-ras mutations occur very early10 and are the most frequent mutations11, 12, 13 in pancreatic cancer followed by mutation or silencing of p53,14 p1615 and DPC4/smad4.16 Addressing the contributions by these lesions to tumor cell survival after treatment may point to new ways to improve therapy for this disease. For pancreatic cancer, K-ras mutation have been reported as a negative prognostic factor after surgery and adjuvant chemoradiation or surgery alone.17 Activated ras (mutant or oncogenic) are known to inactivate genes, which are directly involved in the regulation of growth inhibition and apoptosis. One such example is that oncogenic ras inhibits TGF-β signaling by downregulating RII expression.18 A majority of pancreatic tumors show loss of expression of RII and are resistant to exogenous TGF-β mediated growth inhibition.19 In another example, oncogenic ras was also found to downregulate the proapoptotic gene, “PAR-4.”20 Using rat prostate model, gene expression studies revealed that a gene designated as PAR-4 was exclusively induced by apoptotic stimuli but not by growth promoting, growth-inhibitory or necrotic stimuli.21 Gene function studies indicated that PAR-4 sensitizes cells to the action of apoptotic stimuli; however, PAR-4 by itself does not cause apoptosis. The gene product PAR-4 is widely expressed in unaffected tissues.22 Presence of this gene product in the normal tissue may act as important physiological mediator of cell apoptosis in either health or disease. Since PAR-4 is causally required for apoptosis, absence of this gene expression may render pancreatic tumors a selective resistance to apoptosis-inducing agents. Par-4 is known to selectively inhibit the oncogenic Ras-dependent NF-κB function, the activation of which makes the tumors more resistant.23 In addition, Par-4 binds the atypical protein kinase C isoforms (aPKCs), which serves to inhibit their enzymatic activity.24 Therefore, one of the mechanisms whereby Par-4 induces apoptosis is through its ability to block the aPKC-IKK axis which in turn inhibits the antiapoptotic protein NF-κB. Based on the functional role of PAR-4 in cell death and since pancreatic tumors harbor high incidence of ras mutations, we hypothesized that the downregulation of PAR-4 due to mutations in K-ras will result in the impairment or dysregulation of apoptotic mechanism and thus render selective survival advantage for pancreatic tumors.

To test this hypothesis, established pancreatic tumor cell lines, normal pancreas and pancreatic tumors were analyzed for PAR-4 expression at the mRNA level. In addition, resected pancreatic adenocarcinoma tissue samples were analyzed for mRNA and protein expression by Real-time RT-PCR and immunohistochemistry, respectively. These results were correlated unilaterally with the clinical aggressiveness of the disease and K-ras mutational status. Our results demonstrate that there is a strong association with K-ras mutation and downregulation of PAR-4 and further downregulation of PAR-4 correlated with poor prognosis.


Cell lines

Four established pancreatic cancer cell lines (BxPC-3, Capan-2, MIA PaCa-2 and Panc-1), NIH 3T3 mouse fibroblast and a normal epithelial cell line Hs 578N were used in this study. These pancreatic cancer cell lines and normal cell lines were purchased from American Type Culture Collection (Rockville, MD).

Pancreatic specimens

Ten normal pancreatic tissues and 44 surgically resected and frozen ductal adenocarcinoma of pancreas were obtained. This study was approved by the Institutional Review Board of the University of Kentucky. Twenty-five resected paraffin-embedded samples of pancreatic tumors were also obtained to perform immunohistochemical analysis. The tumors were staged and graded according to the guidelines described by American Committee for Staging of Cancer.

Mutational analysis of K-ras by polymerase chain reaction and allele-specific oligonucleotide hybridization

DNA was isolated after deparaffinizing 10 sections of 10 μm as described by us previously.25 For frozen tissue, homogenized powder of tissue was prepared in liquid nitrogen and subjected to DNA extraction using phenol:chloroform method as described by us previously.25 Point mutations of K-ras (codon 12) oncogene were detected by PCR and allele-specific oligonucleotide hybridization, as per our previously published protocol.25, 26 Briefly, 1 μg of DNA was used to amplify the entire exon 1 of K-ras oncogene using 3′ and 5′ primers. The primers (ras point primer set) and allele-specific oligonucleotide probes (ras muta-probe set) were purchased from a commerical manufacturer. Using a core reagent kit, a 100 μl reaction containing 1 μg of DNA template, 10 μl of 10X PCR buffer II, 6 μl of 25 mM Magnesium Chloride (MgCl2), 2 μl of 200 μM of each dATP, dCTP, dGTP and dTTP, 100 pmol of each 3′ and 5′ primers, and 2.5 U of heat-stabile Taq polymerase was prepared in a 0.5 ml eppendorf tube and the reaction was performed in an automated thermocycler for denaturation at 94°C for 1 min, annealing at 55°C for 2 min and primer extension at 72°C for 1 min. This condition was repeated for 40 cycles. Products obtained after amplification resulted in 118 base pair size DNA and were separated by 3% agarose gel electrophoresis. The 50 μl of the PCR product was denatured and dot blotted using Bio-Rad Dot Blot apparatus on a nylon membrane. The membrane was hybridized with radiolabeled oligonucleotide probe with allele-specific mutation combinations. The membrane was washed and autoradiographed for analysis of mutations as described by us previously.25, 26

Real-time RT-PCR analysis of Par-4 mRNA expression

The expression of Par-4 mRNA was determined by quantitative real-time RT-PCR and normalized to expression of GAPDH. Isolated RNA from frozen tissue was stored at −80°C until used in a 2-step process of reverse transcription (RT) and real-time PCR. All reagents used, including primers and FAM-labeled probes were purchased from Applied Biosystems (Foster City, CA). To eliminate the possibility of measuring any contaminating DNA, primer or probe sets were designed such that at least 1 element of the set spanned an intron and RNA preparation included a DNase digestion step. Using the gene bank sequence for Par-4 mRNA (NM_002583), we designed primers flanking the boundary of exon 3 and 4 and the PCR yielded 366 base pair product.27 The product length, the primers and probes were validated by testing for linearity over a range of concentrations as well as reproducibility of CT (threshold cycle) values at any one concentration when repeated at least 3 times. For random-primed RT, we used TaqMan® RT Reagents (part No. N808-0234) and a Perkin Elmer DNA thermal cycler following kit protocol. RT reactions were used immediately for RT-PCR. Real-time PCR was performed using TaqMan Universal Master Mix, No AmpErase® UNG (part No. 4324018) and an ABI PRISM® 7700 Sequence Detection System according to manufacturer's protocol. CT was determined for each message. Final results for Par-4, expressed as %RNA, are normalized to GAPDH according to the following formula: [2-(CT message − CT GAPDH)] × 100. RNA isolated from the 10 normal pancreatic tissues was used as controls for detecting to measure the actual expression levels in pancreatic tumor tissue and calculate the fold changes.

Transient transfections and western blot analysis

Transient transfections were performed using oncogenic Ki-ras plasmid construct (pCB6+ containing CMV-Ki-rasVal12) in wild-type K-ras BxPC-3 by calcium phosphate precipitation method as described previously.28 For NIH 3T3, cells, transient transfections were performed using both wild type CMV-Ki-ras and mutant CMV-K-rasVal12 For control, pCB6+ empty vector was used. After the transfection, whole-cell protein extracts were prepared and 25 μg amounts of each protein were separated by SDS-polyacrylamide gel electrophoresis. The proteins were transferred to membranes (millipore) and subsequently probed using anti-PAR-4 or anti-pan-Ras (Santa-Cruz, CA) or anti-β-actin (Sigma, MO) antibodies. Using BIORAD gel documentation system and Multi-analyst software, the density of the PAR-4, pan-Ras and β-actin bands were measured and densitometric values were calculated as PAR-4/β-actin and pan-Ras/β-actin ratios.

Generation of growth curves

For growth (proliferation) curve determination, stable transfectants of BxPC-3 (stable transfection was achieved by electroporation using the above plasmids) were seeded into 36,100-mm tissue culture plates at 10,000 cells per plate in high-glucose DMEM. Eighteen plates were treated with ZnSO4 and 18 plates were left untreated. Cells were allowed to grow for a total period of 12 days. Cell numbers were determined on days 2, 4, 6, 8, and 12 using Coulter counter (Beckman) and proliferation rate of the cells treated with BxPC-3/Ki-ras (Val 12) was compared with proliferation rate of BxPC-3/Vector cells.


PAR-4 protein expression in paraffin embedded tissues were determined by immunohistochemistry using rabbit polyclonal antibody R334 (SantaCruz Biotechnology, CA). Deparaffinized sections of formalin-fixed, paraffin-embedded tumor tissue were treated with 0.5% H2O2 in methanol for 20 min. The avidin–biotin peroxidase complex (ABC) method in combination with an antigen retrieval system was used to visualize PAR-4 expression. As a control for this method, adjacent sections will be stained with the primary antisera replaced by normal rabbit serum at the same dilution. Nuclear staining for PAR-4 was assessed and quantification of the immunoreaction was accomplished using image analysis technique (Cell Analysis System, Elmhurst, IL). Fifty randomly chosen tumor cells in 10 different fields were analyzed, including those with the highest and lowest immunoreactions. The percentage of positive staining was automatically calculated by the software.

Statistical analysis

Statistical analyses of data were performed by using the student ‘t’ test or one-way analysis of variance, depending on the number of groups in comparison and χ2 statistics as well as Wilcoxon Rank-sum test. Survival of patients was calculated from the day of treatment using the standard Kaplan–Meier Method and the Log-Rank Method was used to test for statistical significance.


Low levels of PAR-4 protein and mRNA in mutant K-ras pancreatic cancer cell lines

Since oncogenic ras was found to downregulate endogenous PAR-4 protein,20 we analyzed 4 pancreatic cancer cell lines for PAR-4 expression and these levels were compared with the normal pancreatic tissue. A total of 10 normal pancreatic tissues were analyzed for Par-4 mRNA expression. The CT for Par-4 gene ranged from 11.2 to 14 and GAPDH CT ranged from 9.25 to 12. The CT of each normal pancreatic tissue was compared to each cell line and as well as each pancreatic tumor tissue Par-4 CT data. Based on this analysis, the normal epithelial cell line HS 578N showed expression equal to normal pancreatic tissue Par-4 gene expression profile. The cell line BxPC-3 which has wild-type K-ras showed appreciable levels of PAR-4 mRNA and protein expression that was comparable with the PAR-4 expression profile of the normal epithelial cell line as well as normal pancreatic tissue (Fig. 1a). The Panc-1 cell line that harbor mutation in codon 12 of K-ras showed modest downregulation levels of PAR-4 mRNA when compared to the expression profile of the normal epithelial cell line HS 578N, as well as normal pancreatic tissue (Fig. 1a). Significant downregulation of Par-4 mRNA expression was found in Capan-2 and MIA PaCa-2 (both are mutant for K-ras) when compared to the expression profile of the normal epithelial cell line HS 578N, as well as normal pancreatic tissue (Fig. 1a). The mutant K-ras cell lines, Capan-2 and MIA PaCa-2, showed low levels of Par-4 protein when compared to wild-type K-ras cell line, BxPC-3 (Fig. 1b). These data indicate that mutations in the K-ras gene may have an influence on the normal levels of PAR-4 expression.

Figure 1.

Par-4 mRNA and protein levels in pancreatic cancer cell lines. (a) 3D Profile: Cell lines vs. normal tissues. The 3D Profile graphs the fold difference in expression of Par-4 between pancreatic cancer cell lines and normal tissues obtained from real time RT-PCR. Columns pointing up (with z-axis values >1) indicate an upregulation of Par-4 expression, and columns pointing down (with z-axis values <1) indicate a downregulation of Par-4 expression in the cell lines relative to the normal tissue samples. Four pancreatic cancer cells lines and 1 normal epithelial cell were compared with the mean of Par-4 expression observed in 10 normal tissues. (b) Western blot analysis of Par-4 protein expression in pancreatic cancer cell lines BxPC-3 (Lane 1), MIA PaCa-2 (Lane 2) and Capan-2 (Lane 3). The densitometric values were calculated as par-4/β-actin ratio.

PAR-4 mRNA is downregulated or absent in frozen pancreatic adenocarcinoma tissue

Based on the PAR-4 expression found in pancreatic tumor cell lines as shown in Figure 2 and since pancreatic tumors harbor high incidence of K-ras mutations, we analyzed frozen pancreatic tumor and normal specimens collected over the last 2 years for PAR-4 mRNA expression by Real-Time RT-PCR. In addition, K-ras mutational status of both normal and tumor specimens were determined using dot blot hybridization. All the 10 normal pancreatic tissue specimens showed appreciable levels of mRNA, as indicated above in CTs. Of 44 frozen tumors, 16 showed appreciable upregulation of Par mRNA and 27 showed significant downregulation of PAR-4 mRNA when compared to normal pancreatic tissue and 1 had levels equivalent to normal pancreatic tissue (Figs. 2a and 2b). Thirty-two out of 44 cases showed mutations in codon 12 of K-ras (72.7%) and no mutations were detected in the normal pancreatic tissues. Out of 32 cases with mutations in K-ras, 26 showed downregulation of Par-4 expression and 1 showed no change in Par-4 expression when compared with the profile of normal pancreatic tissues. All the 3 cases with wild-type K-ras showed normal levels of PAR-4 mRNA when compared to PAR-4 mRNA levels in normal pancreatic tissue. A significant association with K-ras mutation with downregulation of Par-4 expression was observed (p < 0.0002). These results strongly suggest that K-ras oncogenic mutation might lead to downregulation of PAR-4 expression (Table I).

Figure 2.

Par-4 mRNA expression in normal and tumor frozen specimens of pancreatic tissue by real-time quantitative RT-PCR. (a) 3D Profile: Tumor tissues vs. normal tissues. The 3D Profile graphs the fold difference in expression of par-4 gene between tumor and normal tissues obtained from real time RT PCR. Columns pointing up (with z-axis values >1) indicate an upregulation of Par-4 expression, and columns pointing down (with z-axis values <1) indicate a downregulation of Par-4 expression in the tumor tissue samples relative to the normal tissue samples. Tumor tissues obtained from 44 patients were compared with the mean Par-4 expression observed in 10 normal tissues. (b) Scatter plot: Tumor tissues vs. normal tissues. This graph depicts a log transformation plot of the expression level of Par-4 gene between normal tissue samples (x-axis) and tumor tissue samples (y-axis). The black line indicates fold changes of 1. The pink lines indicate the 4-fold-change in gene expression threshold. Out of 44 tumor samples, 43 showed >4-fold change; 16 up-and 27 downregulated compared to normal tissues.

Table I. Correlation of C-KI-RAS Mutant Status with Down-Regulation of PAR-4
No.StageGradeSurvivalGene alterations
  1. Par-4 expression only in the normal ducts and not in the tumor area.


Downregulation of PAR-4 protein expression in retrospective paraffin-embedded pancreatic tumor specimen correlated with poor survival

To further investigate the influence of PAR-4 expression on survival, we analyzed 25 paraffin-embedded specimens obtained from patients treated at the University of Kentucky for pancreatic adenocarcinoma. The stage grade survival, K-ras status and PAR-4 expression status are shown in Table II. Out of 25 retrospective paraffin-embedded pancreatic tumor samples, 9 samples showed absence of PAR-4 protein expression in both normal and tumor areas of the pancreatic tissue. In the left hand panel of Figure 3, a pancreatic tumor specimen showed intense PAR-4 expression and had wild-type K-ras status. In the center panel, a tumor specimen showed complete absence of PAR-4 expression and had a K-ras mutation (Fig. 3). Interestingly, in the right hand panel a specimen showed clearly PAR-4 expression in the normal pancreatic area but not in the tumor area and this sample had a K-ras mutation. In general, PAR-4 protein was localized in the cytosol of ducts and acinar cells, whereas, no staining was observed in islet cells. Fifteen tumor samples showed intense cytoplasmic staining of PAR-4 immunoreaction in the tumor areas and 1 sample showed immunoreaction only in normal ducts with absence of PAR-4 protein expression in tumor area (Table II).

Figure 3.

Par-4 protein expression in pancreatic intraepithelial neoplasia. Immunohistochemical analysis of Par-4 protein expression in 3 different pancreatic tissue specimens. Upper panel is H&E staining and the lower panel shows Par-4 staining. [Color figure can be viewed in the online issue, which is available at]

Table II. Clinical, Pathological and Gene Alteration Characteristics of Paraffin-Embedded Pancreatic Tumors Specimens
  • Survival percentages include 95% confidence intervals.

  • 1

    Wilcoxon Rank-sum test.

  • 2

    Log-rank test.

Sample sizeN = 16N = 27 
c-K-ras-2  0.00021
 Wild type11 (68.75%)1 (3.7%) 
 Mutant5 (31.25%)26 (96.3%) 
Survival (months)  0.0112
 663% [43%, 91%]44% [29%, 68%] 
 1231% [15%, 65%]4% [1%, 25%] 
 1819% [7%, 52%]0% 
 2413% [3%, 46%]0% 
 366% [1%, 42%]0% 

All the 25 samples showed amplified product for K-ras PCR and thus all the samples were included in the study. Eleven of the 25 carcinomas (44%) analyzed had a point mutation in codon 12 of K-ras oncogene. Of the 11 mutations, 6 (54.5%) were guanine to adenine transitions, 2 (18%) were guanine to thymine transversions, 1 (9%) was a guanine to cytosine transversion and 2 (18%) had both guanine to adenine transitions and guanine to cytosine transversions. Out of 14 with wild-type K-ras, 2 showed absence of PAR-4 expression. Interestingly, out of 11 with mutant K-ras, 7 showed absence of PAR-4 expression (p < 0.0002).

Presence of PAR-4 overexpression correlated with better survival in pancreatic cancer

In the paraffin-embedded tumor tissue cohort of patients, no correlations were observed between the presence and the absence of PAR-4 expression with grade or stage of the tumor. However, presence of PAR-4 expression in tumors suggested better prognosis in terms of survival. The actuarial survival at 18 months was 25% and at 4 years was 12.5% in patients with PAR-4 protein expression as compared to 0% at 18 months for patients with no expression of PAR-4 protein (Fig. 4a).

Figure 4.

Par-4 expression and survival in pancreatic tumor cohorts. Kaplan–Meier survival curve for patients with and without PAR-4 expression. (a) A 25 sample analysis from the formalin-fixed paraffin-embedded tumor tissue specimen; (b) A 43 sample analysis from the frozen tissue cohort.

In the frozen tumor cohort, data from 43 patients were included in the analysis since 1 patient show normal levels of PAR-4 expression similar to the profile of normal pancreatic tissue. The actuarial survival at 18 months was 19% and at 3 years was 6% with PAR-4 upregulation as compared to 0% at both 18 months and 3 years with downregulation of PAR-4 expression (Fig. 4b and Table I). These analysis demonstrate that PAR-4 may potentially be an pivotal prognostic indicator for survival.

Transient overexpression of oncogenic K-ras in wild-type K-ras BxPC-3 cells downregulated the endogenous PAR-4 protein expression and promotes growth

It is evident from the above observations that there is an inverse relationship between K-ras mutational status and expression of proapoptotic gene Par-4 in both clinical samples and cell lines. Mutant K-ras pancreatic cancer cells show downregulation of Par-4 expression, whereas, wild-type K-ras pancreatic cancer cells show no change in Par-4 expression when compared to normal. To ascertain the correlation of oncogenic K-ras and downmodulation of Par-4 expression, BxPC-3 (wild-type K-ras cells) were overexpressed with oncogenic K-ras plasmid that harbor Val12 mutation and the endogenous Par-4 levels were assessed after 48 hr. Overexpression of CMV-Ki-rasVal12 in BxPC-3 cells caused a 50% increase in total pan-ras p21 expression. Interestingly, this overexpression caused a 50% decrease in endogenous PAR-4 protein expression suggesting an inverse relationship between mutant K-ras and Par-4 (Fig. 5a).

Figure 5.

Western analysis of Par-4 and pan-Ras expression in BxPC-3/NIH 3T3 transfectants and impact on growth. (a) Cells were transfected with vector alone or CMV-Ki-rasval12 and after 48 hr, total proteins were isolated. Western blots were probed using antibodies for Par-4, b-actin or pan-ras. Densitometric values were calculated as ratios. (b) Proliferation rate of BxPC-3 with vector alone or CMV-Ki-rasval12. Data represent a mean of 3 experiments. Error bars represent standard errors. (c) Western Blot Analysis of Par-4 expression in NIH 3T3 transient transfectants. Cells were transfected with vector alone, K-ras mutant or K-ras wild type. Western blot was probed using antibodies for Par-4, or β-actin.

Then, the growth kinetics of BxPC-3 cells in absence or presence of activated mutant Ki-ras was evaluated based on the cell number of the cells harvested on days 2, 4, 6, 8, and 12. In the vector alone transfectant, BxPC-3 cells proliferate with the estimated doubling time of 27 hr. Activated expression of mutant Ki-ras increased the rate of proliferation significantly (Fig. 5b). These results indicate that overexpression of CMV-Ki-rasVal12 resulted in an increased proliferation of BxPC-3 cells. Thus, these findings demonstrate that oncogenic K-ras has a direct role in regulation of proapoptotic Par-4 gene expression and thereby promote growth.

Further we tested this hypothesis in nonepithelial normal cell background using NIH3T3 mouse fibroblast. NIH 3T3 cells were transiently transfected with vector alone or full length CMV K-Ras cDNA or mutant CMV-Ki-rasVal12 and expression of Par-4 was assessed after 12 hr. Presence of full length wild-type K-ras overexpression or vector control did not alter PAR-4 protein levels (Fig. 5c). Whereas K-ras (V12) mutant completely abrogated the PAR-4 protein levels after 12 hr of transient transfection. Together these findings indicate that irrespective of the cell type background, PAR-4 protein levels are downregulated in presence of mutant K-ras (V12).


Pancreatic tumors are the only type of solid tumors that show mutations of the K-ras gene at the highest incidence. The ras pathway in a cell is highly important for the transmission of growth promoting signals from the cell surface receptors to nuclear proteins, leading to a signaling cascade that affects the production and the regulation of other key proteins. Most mutations in genes lead to inactivation, whereas, mutations in ras genes lead to active signaling. This active signaling through mutation-activated or oncogenic ras causes specifically upregulation of prosurvival genes, such as NF-κB29 or transcription factors such as ETS, c-jun and c-myc.30, 31 On the other hand, oncogenic ras is a potent downmodulator of proapoptotic and growth inhibitory genes such as Par-423 and the TGF-β signaling cascade.32 Pancreatic tumors show accelerated cellular progression, indicating that mutations in K-ras may pave way for rapid tumor proliferation through inhibition of proapoptotic and growth inhibitory signaling and simultaneous upregulation of prosurvival signaling. Based on these reports, in this study, we hypothesized that Par-4 expression is downregulated in mutant K-ras pancreatic tumors. In addition to observed findings of this study in pancreatic tumor specimens demonstrating the downregulation of Par-4 in mutant ras tumors, this study also demonstrated that ectopic overexpression of oncogenic ras protein in wild-type K-ras pancreatic cancer cell line was sufficient to downregulate endogenous PAR-4 protein. Thus, it is evident that oncogenic K-ras is an important modulator of PAR-4 expression.

The next question that needs to be addressed is how does oncogenic ras downregulate the expression of par-4 gene. Recently Pruitt et al.33 demonstrated downregulation of Par-4 expression when rat epithelial cells were transfected with oncogenic K-ras (V12). This is in concordance with the observations reported in this study, further it was shown that the downregulation of Par-4 was due to the hypermethylation of Par-4 promoter through a MEK-dependent pathway and this methylation was reversed by 5-aza-2′-deoxycytidine and restored Par-4 expression. On the contrary, using mouse fibroblast NIH 3T3 cell is, Barradas et al. reported that the treatment with 5-aza-2′-deoxycytidine did not restore Par-4 expression by oncogenic K-ras, rather the authors of this study concluded that downregulation of Par-4 might be a result of an oncogenic Ras product that produces a potent signal through Raf/ζPKC-MEK pathway, independent of both p53 and p16/19. Further, it is important to note, that the study by Barradas et al. used NIH 3T3 fibroblast cells and the oncogenicity was induced by mutant H-ras construct. Based on these 2 reported studies (Barradas and Vivek) the mechanisms underlying PAR-4 downregulation in a Ras oncogenic background might be exclusively dependent on cell type and H-ras or K-ras mutational status. Since pancreatic tumors harbor K-ras mutations abundantly and the tumor is epithelial in origin, it might be speculated that PAR-4 downregulation in pancreatic tumors is due to hypermethylation of PAR-4 promoter.

In conclusion, immunohistochemical analysis and real-time RT-PCR analysis strongly suggested that presence of Par-4 expression correlated with prolonged survival in patients with pancreatic cancer. Based on this observation, it strongly inferred that the downregulation of Par-4 in pancreatic tumors might be advantageous for rapid tumor proliferation and for the emergence of a resistant tumor phenotype to evade from the killing effects of radiation and chemotherapy.


This work was supported by the Lucille P. Markey Trust for Cancer Research and NCI grant CA086937 (to M.M.A.). We sincerely thank Dr Vivek Rangnekar from University of Kentucky for his useful comments.