Abstract
Background/Aim: The p38 family of mitogen-activated protein kinases (MAPK) includes four isoforms: p38α, -β, -γ and -δ. The aim of this study was to elucidate possible functions of p38α and p38β in human pancreatic cancer. Materials and Methods: Isoform expression was determined in seven human pancreatic cancer cell lines. After shRNA based selective knockdown of p38α and p38β, in vitro growth and migration as well as in vivo tumorigenicity were assessed. Results: All pancreatic cancer cells expressed p38 isoforms. Knockdown of p38α and p38β inhibited in vitro growth. Migration was markedly reduced in p38α shRNA expressing clones, but not altered by p38β knockdown. While in vivo inhibition of p38β decreased tumor formation and growth, the knockdown of p38α significantly enhanced tumorigenicity. Conclusion: p38 MAPKs may exert isoform specific functions in pancreatic cancer. Selective targeting may contribute to individualized treatment of pancreatic cancer in the future.
Pancreatic ductal adenocarcinoma (PDAC) is the fourth leading cause of digestive cancer mortality in Western countries (1). The overall 5-year survival rate of this dismal disease is still less than 5% despite improvements in surgical and multimodal approaches in the past decades (2, 3). Although the reasons for the aggressiveness of pancreatic cancer are not yet completely understood, numerous studies have identified that dysregulation of several growth factor pathways in addition to key genetic and epigenetic events play a role in its pathobiology (4). New approaches in addition to surgery combined with chemo- and radiotherapy are desperately being sought to overcome the poor outcomes associated with this disease.
Our group previously characterized the role of a fibroblast growth factor receptor-1 splice variant FGFR1-IIIb. In contrast to most other naturally-occurring FGFR isoforms, overexpression of FGFR1-IIIb inhibited the malignant phenotype of pancreatic cancer (5). These effects were associated with reduced p42/p44 (ERK 1/2) and increased p38 MAPK activity (5, 6). This observation prompted the investigation of the possible roles of p38 MAPK proteins in more detail.
The p38 mitogen-activated protein kinase (MAPK) signaling pathway was initially described to markedly induce inflammation, particularly upon extracellular cytokine or environmental stress exposure (7). The p38 MAPK family consists of four serine/threonine kinases with a sequence homology greater than 90%, termed p38α (MAPK14), p38β (MAPK11), p38γ (MAPK12) and p38δ (MAPK13) (8, 9). While p38α and p38β are ubiquitously expressed, p38γ is predominantly expressed in skeletal muscle and p38δ is found in the testis, pancreas, kidney and small intestine (10). Despite the close homology of the four isoforms, they each enact different upstream activating and downstream effector proteins (7). The different p38 isoforms are also activated by different upstream kinases (9). Hereby, MAPKK-4 almost exclusively activates p38α, while MAPKK-3 and MAPKK-6 can also phosphorylate the other three isoforms (11).
Regulation of p38 signaling is determined by many factors, including cell-specific differences in expression, subcellular localization, and substrate specificity (12). Consequently, p38 and the specific isoforms have been found to be responsible for a broad spectrum of cellular effects. For instance, p38α reduces cell proliferation by negative regulation of cell cycle progression, modulation of p53 expression, or by induction of apoptosis (8, 13). However, p38α has also been shown to induce anti-apoptotic effects mediated by inflammatory cytokines as well as participating in drug-resistance by enabling DNA-repair caused by cytotoxic drugs (8). Furthermore, p38α promotes cell invasion both by directly promoting the expression of matrix metalloproteinases (MMPs) and vascular endothelial growth factor A (VEGFA) and indirectly by contributing to an inflammatory milieu with the expression of cytokines such as IL-6 or TNFα (14, 15). These contrasting effects of the p38 MAPKs have been described in several clinical studies (16-18).
The results for human pancreatic cancer studies have thus far proven inconclusive. Handra-Luca et al. reported high overall non-specific expression of p38 in pancreatic tumor cells being associated with shorter recurrence-free survival, while Zhong et al. found p38 expression being correlated with improved survival (19, 20). Therefore, in our study we aimed to characterize the expression pattern and oncogenic potential of p38 MAPKs in cultured pancreatic cancer cells. Specifically, we sought to determine the effects of each isoform individually and elucidate possible functions of the p38α and p38β isoforms in the pathogenesis of pancreatic cancer.
Materials and Methods
Cell culture. Cultured human PDAC cell lines AsPC-1, BxPC-3, Capan-1, MIA PaCa-2, and PANC-1 were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). COLO-357 and T3M4 cells were a gift from M. Korc (Indianapolis, IN, USA). p38α- and p38β-shRNA transfected clones were grown in medium containing Geneticin (G418) [2.4 mg/ml (MIA PaCa-2) and 1.2 mg/ml (PANC-1)]. Cell culture was performed as described previously (21).
Immunoblot analysis. Total cell lysates were prepared and subjected to immunoblot analysis as previously described (5). Rabbit anti-human p38α, p38β, p38γ and p38δ polyclonal antibodies (#9218, #2339, #2307, and #9214, respectively, from Cell Signaling Technology (Danvers, MA, USA) were used (1:1,000) to detect p38 isoforms. To confirm equal loading, membranes were stripped and re-probed with an anti-β-actin antibody (5).
Establishment of cell clones expressing p38α and p38β shRNA. Validated SureSilencing human p38α (MAPK14), p38β (MAPK11) and control plasmids were purchased from SuperArray Bioscience Corp. (Frederick, MD, USA). MIA PaCa-2 and PANC-1 cells were transfected in a stable manner using lipofectamine (Invitrogen, Carlsbad, CA, USA) following the manufacturer's protocol. Transfected cells were selected with G418 (2.4 and 1.2 mg/ml for MIA PaCa-2 and PANC-1, respectively) for 14 days before isolation of individual clones as previously described (22).
p38 kinase activity assay. To determine p38 activity, a p38 MAPK assay kit (#9820, Cell Signaling Technology) was used according to the protocol of the manufacturer. In brief, total cell lysates (250 μg in 200 μl of lysis buffer) were incubated for 20 h at 4°C with re-suspended immobilized phospho-p38 MAPK (Thr180/Tyr182) primary antibody (20 μl). After washing, the pellet was re-suspended in 50 μl of kinase buffer supplemented with 200 μM ATP and appropriate quantity of kinase substrate (ATF-2 Fusion Protein, 1 μl/assay). Following incubation of the mix for 30 min at 30°C, immunocomplexes were captured by centrifugation and subjected to immunoblot analysis using phospho-ATF-2 (Thr71) antibodies.
Anchorage-dependent growth assay. Basal growth of different clones expressing p38α or p38β shRNA was determined by the 3- (4,5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay and cell counting. To assess the basal cell growth, the cells (10,000 per well) were propagated for 48 h in complete medium before initiation of the MTT assay. For cell counting, cells (60,000 per dish) were seeded in 6-cm dishes and propagated for 48 h in complete medium. Medium was changed daily until cell counting was performed (5, 22).
Anchorage-independent growth assay. Basal anchorage-independent cell growth was assessed by a double-layer soft agar assay as described (5). Briefly, cells were suspended in complete medium containing 0.3% agar and seeded in triplicate in 6-well plates onto a base layer of complete medium containing 0.5% agar. One milliliter of complete medium containing 0.3% agar was added every 5 days. After 14 days, 300 μg MTT/well was added to stain viable colonies for 24 h before counting by microscopy.
Single cell movement assay. Cells (50,000 per well) were seeded onto fibronetin-coated (5 μg/ml in PBS) six-well plates and grown for 20 h. Cell movement during the following 24 h was monitored by an Olympus IX81 motorized inverted microscope taking pictures every 10 min (6). The total distance of individual cells covered within 24 h was tracked and determined using the ImageJ 1.32 program (NIH, Bethesda, MD, USA).
Cell migration assay. The ability of cells to migrate through filters was measured using a BioCoat Matrigel invasion chamber (BD Biosciences, San Jose, CA. USA). Cell culture inserts with an 8 μm pore size PET membrane were used according to the protocol of the manufacturer. The bottom chamber included medium (0.75 ml) containing 10% FCS, whereas cells (2×104 suspended in 0.5 ml of medium containing 1% FCS) were seeded into the upper chamber and incubated for 36 h at 37°C in a humidified atmosphere containing 5% CO2. Remaining cells on the upper surface were mechanically removed. Membranes were then washed, fixed, and stained by Diff-Quik (Medion Diagnostics, Düdingen, Switzerland). The number of cells that migrated to the lower surface of the filters was determined by counting stained cells under a light microscope as described (5, 22).
In vivo tumorigenicity assay. To assess the effect of inhibition of p38α and p38β on xenograft formation, 106 cells per site were injected subcutaneously (sc) into two sites of 4- to 6-week-old female athymic mice. Animals were monitored for tumor formation every 4 days. Tumor size was measured in three dimensions. Tumor volume was determined by the equation vol=l×w×d×0.5, where l is the length, w is the width, and d is the diameter. Animals were sacrificed 12 weeks after injection according to our animal protocol (#718) if neither tumor volume (>2 cm3) nor skin ulcerations prompted earlier termination. Tumors were excised and fixed in formalin for histological analysis.
Microscopic analysis of xenograft tumors. To assure a standardized analysis, explanted tumors were cut in half through the largest diameters of each tumor. Formalin-fixed and paraffin-embedded 5-μm sections were prepared from the central areas of all xenograft tumors followed by H&E staining and immunohistochemical analysis using a Vectastain avidin-biotin complex kit (5, 23).
Kaplan–Meier plots for patient survival correlated with p38α and p38β expression. The Cancer Genome Atlas Database (TCGA) and OncoLnc were used to create Kaplan–Meier survival plots correlated with p38α and p38β expression. Lower and upper quartile were used to determine low versus high mRNA expression. A total of 43 patients were included in the analysis (24, 25).
Statistics. Results were expressed as mean expression levels (±SD or SEM). Data were analysed using SigmaPlot 12.0. Student's t-test, rank sum test, and chi-square test were used for statistical analysis. A p-value <0.05 was considered to indicate statistical significance (two-sided).
Results
Expression of p38 MAPKs in cultured pancreatic cancer cells. Protein expression of the four p38 MAPKs was determined in seven cultured PDAC cell lines (Figure 1A). p38α was the predominant p38 isoform expressed in all cell lines. AsPC-1, COLO-357, and PANC-1 cells lines expressed equally high levels of p38β, while the other four cell lines demonstrated lower levels of p38β compared to p38α levels. p38γ protein was detected at low levels in all lines and p38δ was expressed at low levels in 3 of 7 cell lines.
MIA PaCa-2 and PANC-1 cells were used to establish knockdown clones expressing either p38α or p38β shRNA in a stable manner. These cell lines were chosen because MIA PaCa-2 cells predominantly express p38α (low p38β, low p38γ, no p38δ) and PANC-1 cells express equal levels of p38α and p38β (low p38γ, no p38δ). Immunoblot analysis revealed that the 40 kDa band corresponding to p38α was decreased significantly in MIA PaCa-2 (M) and PANC-1 (P) clones expressing p38α shRNA (M14-25 and M14-26, and P14-42 and P14-43, respectively) compared to wild-type and control-transfected cells (MNeo and PNeo, respectively) (Figure 1B, upper panels). The 43 kDa band corresponding to p38β was decreased in MIA PaCa-2 clones M11-34 and M11-47 and in PANC-1 clones P11-12 and P11-15, respectively, expressing p38β shRNA (Figure 1B, second panels). Total p38 MAPK activity determined as ATF-2 phosphorylation activity was decreased in PANC-1 cells (equivalent levels of p38α and p38β) by inhibition of either p38α or p38β expression (Figure 1B, right lower panel), while total p38 MAPK activity was only decreased in MIA PaCa-2 cells (high level of p38α, low level of p38β) by inhibition of p38α and not by p38β expression (Figure 1B, left lower panel).
Effect of p38α and p38β inhibition on pancreatic cancer cell proliferation, colony formation, motility, and invasion in vitro. The anchorage-dependent (MTT assay and cell counting) and -independent (soft agar assay) growth of p38α and p38β shRNA-expressing clones was significantly inhibited in both cell lines in comparison to wild-type and control transfected cells, resulting in an inhibition of anchorage-dependent growth of up to 43% (Figure 2A-D). Colony-formation was also reduced in both cell lines by up to 25% (Figure 2E, F).
p38α inhibition resulted in a marked reduction in single cell motility in M14-25 and M14-26. In contrast, inhibition of p38β expression did not show changes in motility of M11-34 and M11-47 (Figure 3A). The same phenotype was found in PANC-1 (Figure 3B).
Inhibition of p38α protein expression reduced cell invasion in both cell lines up to 38%. Inhibition of p38β protein expression did not affect invasion (Figure 3C-F).
Inhibition of p38β decreased tumorigenicity and inhibition of p38α enhanced tumorigenicity in nude mice. To assess the effect of p38α and p38β expression on tumor formation and growth in vivo, MIA PaCa-2 and PANC-1 cells with knockdown of each isoform and their respective controls were injected subcutaneously (sc) into nude mice. Xenograft tumors developed at 13 of 16 (81.25%) and 13 of 16 (81.25%) sites injected with wild-type and control transfected MIA PaCa-2 cells, respectively. Fourteen of 16 (87.5%) and all 16 (100%) sites injected with p38α shRNA-expressing cell clones M14-25 and M14-26, respectively, developed tumor nodules. In contrast, only 3 of 16 (18.75%) and 8 of 16 (50%) sites injected with p38β shRNA-expressing cells M11-34 and M11-47 respectively, developed tumors, unveiling a significantly reduced tumor formation capability (p=0.006). Tumor volume assessment revealed no growth difference between wild-type and control-transfected cells, but a significant reduction in the mean tumor volume in mice transplanted with p38β shRNA expressing M11-34 and M11-47 clones (Figure 4A). A significant increase in mean tumor volume was found in animals transplanted with M14-25 and M14-26 clones expressing p38α shRNA (Figure 4A).
In PANC-1 cells, 6 of 8 (75%) and 7 of 8 (87.5%) sites injected with wild-type and control transfected cells, respectively, developed tumors. In contrast, only 2 of 8 (25%) and 1 of 8 (12.5%) sites injected with p38β shRNA-expressing clones P11-12 and P11-15, respectively, showed tumors (p<0.05), while all the sites (8 of 8, 100%) injected with p38α shRNA-expressing clones P14-42 and P14-43 developed tumor nodules. Although smaller in size, tumor volumes of PANC-1 cell clones revealed similar results as observed for MIA PaCa-2 cells. Growth at the three sites of p38β shRNA expression clones was virtually undetectable even after 48 days and thus not displayed with the other clones (Figure 4B). Alongside our findings in MIA PaCa-2, animals transplanted with p38α shRNA-expressing P14-42 and P14-43 clones rapidly developed tumors with a large mean tumor volume compared with animals transplanted with wild-type or control-transfected PANC-1 cells (Figure 4B). Thus, animals carrying tumors arising from p38α shRNA expressing P14-42 and P14-43 clones were sacrificed after 6 to 7 weeks due to the occurrence of skin ulcerations.
Expression of p38 MAPK proteins in cultured human pancreatic cancer cells and the effect of p38α and p38β inhibition on total p38 MAPK activity. (A) Protein expression of the four p38 MAPK proteins. β-actin (lower panel) served as loading control. (B) Effect of p38α and p38β shRNA expression in MIA PaCa-2 (M – left panels) and PANC-1 (P – right panels) cells. Immunoblot analysis is shown for p38α, p38β, and β-actin for the respective wild-type cells (M WT, P WT), control transfected clones (M Neo, P Neo), p38β shRNA expressing clones (M11-34 and M11-47, P11-12 and P11-15), and p38α shRNA-expressing clones (M14-25 and M14-26, P14-42 and P14-43). The lower panel shows the results of an in vitro total p38 MAPK activity assay determining phosphorylation of ATF-2.
Morphology of xenograft tumors. Xenograft tumors were paraffin-embedded and characterized by microscopy. Tumors of wild-type (n=13) and control transfected (n=13) MIA PaCa-2 cells showed a homogenous pattern of papillary structure of tumor cells displaying interspersed areas of necrosis. Overall, there was no striking morphological difference between control tumors and p38α [M14-25 (n=14) and M14-26 (n=16)] and p38β [M11-34 (n=3) and M11-47 (n=8)] shRNA-expressing cells. Proliferation rates (Ki-67 labeling) were without marked differences between control cells and p38α knockdown cells, while p38β knockdown cells showed slightly less Ki67 positive cells (Figure 5 right panel).
Effect of p38α and p38β MAPK inhibition on basal adherent cell growth and colony formation of MIA PaCa-2 and PANC-1 cells. Basal cell proliferation of wild-type (M WT/P WT), control-transfected (M Neo/P Neo), p38β shRNA expressing clones (M11-34 and M11-47/P11-12 and P11-15), and p38α shRNA-expressing clones (M14-25 and M14-26/P14-42 and P14-43) was determined by MTT assay (A, B) and cell counting (C, D). (A, B) Cells (10,000 cells/well) were cultured in 96-well plates in complete medium for 48 h and then analyzed by MTT. Results are shown as means (±SEM) compared to M WT and P WT of twelve determinations from six separate experiments. (C, D) 60,000 cells/well were grown in 6 cm dishes for 48 h before counting. Results are means (±SD) of four separate experiments carried out in duplicate. (E, F) Soft agar assay. Indicated cells (5000 cells/well) were seeded in 6 well plates and cultured in a soft agar double layer system. Colony formation was determined after 14 days by an inverted light microscope after staining viable colonies with MTT solution. Results are shown as mean number (±SEM) of colonies per well and are means of four separate experiments of triplicate determinations. **p<0.01, ***p<0.001 compared to WT and Neo.
Influence of p38α and p38β expression on in vitro single cell movement and migration. (A, B) Single cell movement: The total distance (μm in 24 h) of individual cells (n=50) covered within 24 h was evaluated using the ImageJ 1.32 and Simple Track program. Results are shown as mean distances in μm (±SD) covered within 24 h. *p<0.05 compared to M WT and M Neo. (C-F) Boyden chamber assay: The number of cells that moved to the lower side of the chamber within 36 h was determined. Migrated cells were stained and counted under an inverted brightfield microscope. (C, D) Mean number (±SD) of migrated cells within 36 h of three independent experiments (*p<0.05 compared with wild type cell and Neo clones). (E, F) Representative images of Boyden chamber migration.
Immunohistochemical analysis using specific antibodies for p38α and p38β confirmed the presence of both proteins in the cytoplasm of wild-type and control transfected cells. p38β expression was consistently lower than p38α and showed high expression in infiltrating stromal cells. Staining of the knockdown cells confirmed stably reduced isoform expression in vivo (Figure 5 left and middle panel).
Patient survival correlated with p38α and p38β expression. p38α and -β expression from 43 patients of The Cancer Genome Atlas Database (TCGA) was correlated with their survival. High overall p38α expression in human pancreatic cancer tissue was associated with shorter survival (Figure 6A), while high p38β expression showed a trend towards longer survival (Figure 6B), although in both cases not on a statistically significant level.
Discussion
Increased p38 activity has been previously demonstrated in a variety of solid tumors, including prostate, bladder, and lung (17, 26, 27), while reduced expression levels were found in hepatocellular carcinoma (28) and elevated p38 activity has been associated with a better prognosis in breast cancer (29).
Many studies investigating p38 MAPKs used pyridinyl-imidazole inhibitors such as SB203580 (22) and SB202190 to identify substrates for p38 MAPK proteins and characterize their role in signal transduction. However, because these substrates inhibit both p38α and p38β MAPK with similar IC50 values, it is not possible to distinguish between the effects of individual p38 MAPK proteins (30). In accordance to the expression levels in other tissues, our results revealed that p38α was the predominant protein expressed in cultured pancreatic cancer cells followed by p38β, while p38γ and p38δ were expressed at low levels or even absent (12, 31). We therefore focused our study on p38α and p38β.
We utilized shRNA vectors specifically targeting p38α or p38β to independently investigate the functions of p38α and p38β MAPK proteins in the pathogenesis of human pancreatic cancer, which has previously not been reported.
Isotype-specific knockdown was successful in both target cell lines without interference from other isoforms. Biological effects were confirmed by measuring overall p38 kinase activity, which was reduced in parallel to kinase knockdown. This effect was not observed in MIA PaCa-2, which demonstrate endogenously low levels of p38β.
Irrespective of the level of total p38 MAPK activity, inhibition of both p38α and p38β MAPK resulted in a significant decrease in in vitro cell proliferation in both anchorage dependent and independent growth.
Effect of p38α and p38β expression on xenograft formation. (A) Indicated MIA PaCa-2 wild-type (WT) and transfected clones (2×106 cells/site) were injected s.c. at each of two sites in athymic mice. Tumor size was measured every 4 days. Tumor volume was determined as described in the materials and methods section. Results are mean tumor volumes of M WT (●; n=13), M control transfected (Neo) (○; n=13), M 11-34 (▾; n=3), M11-47 (▵; n=8), M14-25(▪; n=14), M14-26 (□; n=16) cells. No tumor formation was observed at sites injected with M WT (3 of 16), M Neo (3 of 16), M11-34 (13 of 16), M11-47 (8 of 16), M14-25 (2 of 16). *p<0.05, compared with wild type cell and Neo clones. (B) Indicated PANC-1 wild-type and transfected clones (1×106 cells/site) were injected as in (A). Results are mean tumor volumes of P WT (●; n=6), P Neo (○; n=7), P14-42 (▾; n=8), M14-43 (▵; n=8) cells. No tumor formation was observed at sites injected with P WT (2 of 8), P Neo (1 of 8), P11-12 (6 of 8), P11-15 (7 of 8). *p<0.05, compared with wild-type cells and Neo clones.
Down-regulation of p38α expression had an inhibitory effect on single cell motility and invasion in vitro. p38 was reported to be a key signal transducer for cell motility and invasion of breast epithelial (32) and hepatocellular carcinoma (33) cells. Also, in the bladder carcinoma cell line HTB9, p38 may regulate invasiveness by stabilizing matrix metalloproteinases -2 and -9 (27). Activation of motility and invasion in breast epithelial cells was mediated by an H-Ras-dependent pathway followed by p38α activation (34).
Histology of xenograft tumors MIA PaCa-2. Wild type (WT) and control transfected (Neo) cells showed cytoplasmic staining of p38α and p38β with higher expression of the alpha isoform, while the knockdown clones lacked expression of the respective isoform (left and middle panel). Proliferation measured by Ki67 positive cells was generally low and did not differ between WT, control cells and p38α knockdown cells, while p38β knockdown cells showed a slight reduction (right panel).
The role of p38 MAPK in motility and invasion ability of pancreatic cancer cells is further supported by the fact that its upstream activator MKK4 was found to be associated with a higher frequency of recurrence/metastases in pancreatic cancer samples (35). Furthermore, the p38 signaling pathway was up-regulated in pancreatic cancer circulating tumor cells (36). These studies confirm the involvement of p38 MAPKs in processes such as motility and invasion, however, they did not explore the importance of specific p38 MAPK isoforms.
Kaplan–Meier survival plots in dependency of whole tissue mRNA expression of p38α (A) and p38β (B). Lower quartile was defined as low expression and upper quartile was defined as high expression (n=43, p=0.324 for p38α, p=0.0664 for p38β).
Our study revealed that inhibition of p38β MAPK in two different pancreatic cancer cell lines did not alter single cell motility or invasion, while inhibition of p38α displayed a marked effect on single cell motility as well as in vitro invasion. Based on these observations we conclude that p38β MAPK is not involved in the regulation of cell motility and invasion in pancreatic cancer cells, while p38α MAPK markedly influences these processes. Due to the abundant expression of p38α, the observed effects of other groups may also be attributed to p38α.
While the observed inhibitory effects of p38α and p38β MAPK inhibition on in vitro proliferation and colony-formation were similar, the in vivo experiments revealed a striking difference in their capability to induce and propagate subcutaneous xenograft tumors. Although p38α MAPK inhibition reduced in vitro proliferation, the contrasting induction of subcutaneous tumors and their growth enhancement was striking. Knockout experiments in mice revealed that p38α ablation not only resulted in enhanced proliferation of non-malignant cells such as cardiomyocytes, hepatocytes and haematopoietic cells (37, 38), but also rendered cells more susceptible to malignant transformation e.g. in a K-RasG12V-dependent manner in lung cancer (39, 40).
These studies suggest a tumor-suppressive role for p38α in human cancers. Nevertheless, immunohistochemical analysis of 99 human pancreatic cancer tissues showed a trend towards decreased survival in patients with high expression of p38, although only at a statistically significant level in patients without post-surgical treatment (19). The TCGA dataset analysis showed a seemingly contradictory finding with longer survival in patients with high p38β expression. Nevertheless, the TCGA dataset consists of bulk tumor RNA sequencing data and so the observed effects may be attributable to p38 expression and function in the tumor microenvironment (TME). In the TME, the TNF-mediated expansion of CD4+ regulatory T cells as a cellular mechanism of immune evasion was abolished by pharmacological p38 inhibition. This resulted in a reduction of lung metastases in a breast cancer model (41). Furthermore, CD4+ T cells with activated p38 signaling were shown to promote pancreatic cancer (42). These effects should also be considered when evaluating p38 as a therapeutic target.
Our findings from the specific inhibition of the two predominant p38 isoforms found in PDAC demonstrate opposing functions. We observed both a tumor growth-suppressive effect of p38α in pancreatic cancer as previously established in other malignancies, and strikingly, we have now shown that p38α functions can be opposed by p38β, promoting a malignant phenotype of pancreatic cancer by increasing both growth and invasion in vitro as well as in vivo tumor initiation and growth. Consistently, p38β protein serum levels were significantly increased in patients with pancreatic cancer compared to those diagnosed with chronic pancreatitis or with normal pancreas function (43), supporting our finding that p38β may act as an oncogenic promotor of this disease. Still, the precise mechanisms of action of the different isotypes remain to be determined. Our results and their possible application can be underlined by the findings of Koizumi et al. They've shown the importance of p38 activation for gemcitabine induced apoptosis in pancreatic cancer cells. Interestingly, a double negative mutant of p38α showed increased resistance towards gemcitabine treatment (44). Combined with our findings, a p38β directed therapy may be promising for pancreatic cancer treatment, while a nonspecific p38 inhibition might result in a reduced patient response towards chemotherapy.
These opposing functions may also explain the contrasting descriptions of p38 MAPKs as both tumor suppressors and tumor promotors in different cell lines and tissues (9, 11, 12). The knowledge of the specific functions of p38α and especially of p38β are scarce so far (9, 11, 12, 45, 46) and a recent review even suggested that their functions in development are redundant (47). Our present model system clearly demonstrates opposing functions of both kinases in in vivo tumorigenesis.
In conclusion, our study indicates that p38 MAPK proteins have different impacts on the biology of pancreatic cancer cells. p38α activity enhances motility and invasion in vitro, while the p38β isoform has no effect on these functions. Our results further demonstrate that p38α suppresses tumor growth in pancreatic cancer, while p38β may act as a tumor promoter in vitro and in vivo. In particular, inhibition of p38β may be a potent and attractive target for the treatment of this disease.
Acknowledgements
The Authors would like to thank Claudia Dreschler and Nadine Süβner for the technical assistance.
Footnotes
↵* These Authors contributed equally to this study.
Authors' Contributions
X.T. and M.K. designed the study, X.T., B.T. and X.X. performed the experiments, S.Z., D.H-B., U.K., and M.K. supervised the experiments, B.T. and M.K. wrote the paper.
Conflicts of Interest
The Authors have no conflicts of interest to declare regarding this study.
- Received July 2, 2020.
- Revision received July 27, 2020.
- Accepted July 28, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
