Phosphoglycerate kinase 1 a promoting enzyme for peritoneal dissemination in gastric cancer

Abstract

Peritoneal carcinomatosis is a frequent finding in gastric cancer associated with a poor prognosis. The features that enable gastric tumors to disseminate are poorly understood until now. Previously, we showed elevated mRNA levels of phosphoglycerate kinase 1 (PGK1), an adenosine triphosphate-generating enzyme in the glycolytic pathway, the chemokine receptor 4 (CXCR4), the corresponding chemokine ligand 12 (CXCL12) and beta-catenin in specimens from gastric cancer patients with peritoneal carcinomatosis. In this study, the influence of PGK1 on CXCR4 and beta-catenin was assessed as well as the invasiveness of PGK1 overexpressing cancer cells. In this current study, we found that PGK1 regulates the expression of CXCR4 and beta-catenin at the mRNA and protein levels. On the other hand, CXCR4 regulates the expression of PGK1. Plasmid-mediated overexpression of PGK1 dramatically increased the invasiveness of gastric cancer cells. Interestingly, inhibition of CXCR4 in cells overexpressing PGK1 produced only a moderate reduction of invasiveness suggesting that, PGK1 itself has a critical role in tumor invasiveness. Immunohistochemistry in specimens from diffuse gastric cancer patients also revealed an overexpression of PGK1 in patients with development of peritoneal carcinomatosis. Therefore, PGK1 may be a crucial enzyme in peritoneal dissemination. Together these findings suggest that the enhanced expression of PGK1 and its signaling targets CXCR4 and beta-catenin in gastric cancer cells promote peritoneal carcinomatosis. Thus, PGK1 may serve as prognostic marker and/or be a potential therapeutic target to prevent dissemination of gastric carcinoma cells into the peritoneum.

Figure 1. small interfering RNA-mediated PGK1 gene silencing reduces CXCR4 and β-catenin expression in MKN45 cells

MKN45 cells treated with control siRNA or PGK1 siRNA pool (A) or CXCR4 siRNA pool (B) were assessed for PGK1, CXCR4 and β-catenin mRNA levels by qRT-PCR at 28, 34 and 40 h (A) or 52, 82 and 92 h (B) post transfection, mRNA levels are given as relative changes compared to corresponding MKN45 cells treated with non-targeting control siRNA resulting in 100%. (C), Cell lysates of the MKN45 control cells or PGK1-siRNA cells were examined for PGK1 protein expression by immunoblot using β-actin as a reference. Lysates were prepared 46, 64 and 76 h post transfection. (D), Cell lysates of the MKN45 control cells or MKN45 CXCR4-siRNA cells were examined for PGK1 protein expression by Western blot using β-actin as a reference. Lysates were prepared 70, 82 and 106 h post transfection. (E), Cell proliferation and viability of PGK1siRNA and control siRNA treated cells were determined by the trypan blue dye exclusion method. Total cell numbers of living and dead cells are given at 24, 48, 72, 96 and 120 h and represent the mean numbers of three independent experiments. (F), CXCR4 and β-catenin protein levels in MKN45 siControls or MKN45 siPGK1 cells were determined by FACS analysis 76h post transfection. Representative FACS profiles are shown, and the respective medians of fluorescence and SFI values are shown within the histograms.

Figure 2. PGK1 overexpression increases CXCR4 and β-catenin levels in MKN45 cells

(A), PGK1 mRNA overexpression in MKN45 cells stably transfected with PGK1 pEF-IRES plasmid normalized to MKN45 pEF-IRES control pools (equals to 100%) as well as CXCR4 mRNA levels were assessed by qRT-PCR. Mean expression levels of PGK1, CXCR4 and β-catenin mRNA levels in pools #1 – #3 are presented in the right panel (*p<0.05). The left panel depicts the single pools. (B), Cell lysates of the MKN45 control pools or the MKN45 PGK1 pools were examined for PGK1 protein expression by immunoblot using β-actin as a reference. (C), MKN45/PGK1pools #1 – #3 or MKN45/control pools were assessed for CXCR4 and β-catenin protein by flow cytometry. Representative flow cytometry profiles of the MKN45/PGK1pools and MKN45/control pools are shown and the respective medians of fluorescence and SFI values are displayed within the histograms.

Figure 3. PGK1 overexpression dramatically increases the invasiveness of gastric cancer cells: modulation by inhibition of CXCR4

(A), Invasion for 24h was assessed in three independent MKN45 puromycin control pools and in MKN45/PGK1pools #1–#3. Data are expressed as means of the control pools #1–#3 and MKN45/PGK1pools #1–#3 and SD (*p<0.05) (B), MKN45/PGK1pool#3 cells were treated with isotype control or anti-CXCR4 antibody. (A), (B), Representative photomicrographs of the results of the invasion assays are depicted.

Figure 4. PGK1 overexpression in human diffuse gastric cancer with peritoneal carcinomatosis

Immunohistochemical staining for PGK1 in diffuse gastric cancer patients with (a and b) and without (c and d) peritoneal carcinomatosis. a) and b) show a strong staining in most of the tumor cells whereas c) and d) reveal only a weak or no staining for PGK1. All images are 50-times magnified.

Figure 5. Pathway-model: PGK1 a promoter for peritoneal cancer dissemination

The figure represents a potential model of peritoneal dissemination in gastric cancer, via PGK1 signaling. PGK 1 overexpression leads to increased CXCR4/CXCL12 signaling which supports metastatic homing and tumor dissemination. Enhanced PGK1 expression influences tumor invasion by elevated CXCR4 and β-catenin levels. CXCR4 is reciprocally regulated by PGK1 and overexpression of CXCR4 allows a link to CXCL12 signaling and thus immune evasion mechanisms apart from direct influences. β-catenin furthermore exhibits its function in cell proliferation and metastasis. Therefore PGK1 might be a crucial enzyme in peritoneal dissemination and metastasis.