Annexin A1 contributes to pancreatic cancer cell phenotype, behaviour and metastatic potential independently of Formyl Peptide Receptor pathway

Abstract

Annexin A1 (ANXA1) is a Ca(2+)-binding protein over-expressed in pancreatic cancer (PC). We recently reported that extracellular ANXA1 mediates PC cell motility acting on Formyl Peptide Receptors (FPRs). Here, we describe other mechanisms by which intracellular ANXA1 could mediate PC progression. We obtained ANXA1 Knock-Out (KO) MIA PaCa-2 cells using the CRISPR/Cas9 genome editing technology. LC-MS/MS analysis showed altered expression of several proteins involved in cytoskeletal organization. As a result, ANXA1 KO MIA PaCa-2 partially lost their migratory and invasive capabilities with a mechanism that appeared independent of FPRs. The acquisition of a less aggressive phenotype has been further investigated in vivo. Wild type (WT), PGS (scrambled) and ANXA1 KO MIA PaCa-2 cells were engrafted orthotopically in SCID mice. No differences were found about PC primary mass, conversely liver metastatization appeared particularly reduced in ANXA1 KO MIA PaCa-2 engrafted mice. In summary, we show that intracellular ANXA1 is able to preserve the cytoskeleton integrity and to maintain a malignant phenotype in vitro. The protein has a relevant role in the metastatization process in vivo, as such it appears attractive and suitable as prognostic and therapeutic marker in PC progression.

Figure 1

(A) Western blot showing B11, D6 and G5 KO clones for ANXA1. ANXA1 expression has been compared with WT and PGS MIA PaCa-2 and normalized on tubulin levels. (B) Proteins belonging to annexin superfamily identified by LC-MS/MS. ***p < 0.001. (C) Pie chart showing the absolute number of proteins identified as differentially expressed in ANXA1 KO MIA PaCa-2 cells by LC-MS/MS. The proteins were grouped based on their main reported function according to UniProtKB. (D) Proteins identified as differentially expressed by LC-MS/MS and involved in the process of cytoskeleton organization. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2

(A) Cell surface expression of CD44 was analyzed by flow cytometry. The violet area in the plot is relative to human IgG1; CD44 signals are showed in green for WT MIA PaCa-2, in pink for PGS MIA PaCa-2 and in black for ANXA1 KO MIA PaCa-2. (B) RT-PCR for K18 mRNA expression measured on levels of HPRT. Values are expressed using the delta-delta Ct method to derive relative fold change. ***p < 0.001. (C,D) Western blots showing K8, ANXA1, lamin A/C, vimentin and GAPDH expression in WT, PGS and ANXA1 KO MIA PaCa-2 cells. (E,F) Vimentin and lamin A/C relative expression was analyzed by densitometry. The optical density of the protein bands was normalized on GAPDH levels giving to the control band an arbitrary value of 100. The blots were exposed to Las4000 (GE Healthcare Life Sciences) and the relative intensities of bands were determined using ImageQuant software (GE Healthcare Life Sciences). (G) Immunofluorescence analysis to detect ANXA1 (red; panels a, b, c), vimentin (yellow; panels d, e, f), lamin A/C (purple; panel g, h, i) and F-actin (green; panels l, m, n) in WT, PGS and ANXA1 KO MIA PaCa-2. Nuclei were stained with DAPI. Magnification 63 × 1.4 NA. Bar = 10 μm. (H) Fluorescence intensity for ANXA1, vimentin and lamin A/C signals (arbitrary units, A.U.) using ImageJ software; determined on 150 cells (for three independent experiments). The results relative to ANXA1 KO MIA PaCa-2 are representative ± SEM of almost three analyzed clones with a similar behaviour.

Figure 3

(A) Results of Wound healing assay on ANXA1 KO MIA PaCa-2; ***p < 0.01 vs WT and PGS MIA PaCa-2. The migration rate was determined by measuring the distances covered by individual cells from the initial time to the selected time-points (bar of distance tool, Leica ASF software). The data are representative of 5 independent experiments ± SEM. (B) Representative images captured by TIME LAPSE microscope of WT, PGS and ANXA1 KO MIA PaCa-2 at 0 h and 24 h from produced wounds. Magnification 10×. (C) Results of the invasion assay on ANXA1 KO, PGS and WT MIA PaCa-2 cells. Data represent mean cell counts of 12 separate fields per well ± SEM of 5 experiments. ***p < 0.001 vs WT and PGS MIA PaCa-2. (D) Representative images of analyzed fields of invasion assay. Magnification 20x. Bar = 150 μm

Figure 4

(A) MTT assay at 24, 48 and 72 h on WT, PGS and ANXA1 KO MIA PaCa-2 (clones B11, D6 and G5). (B) Hemocytometer cell counts of WT, PGS and ANXA1 KO at 24, 48 and 72 h of cell culture. **p < 0.01; ***p < 0.001.(C) Cell cycle analysis with PI staining, the graph is representative of 72 h of culture, after 24 h of serum starvation. (D) Quantitation of BrdU-positive cells. Representative forward scatter histograms and statistical analyses of BrdU incorporation by WT, PGS and ANXA1 KO cells. ***p < 0.001 vs control; ^▲▲▲p < 0.001 vs WT cells. (E) Western blot of Cyclin A, ALDH7A1, Phospho-ERK and ERK on WT, PGS and ANXA1 KO MIA PaCa-2 clones as B11, D6 and G5. All protein levels are normalized on GAPDH levels. Data are representative of 5 experiments with similar results. (F) Densitometry for cyclin A, ALDH7A1 and p-ERK expression in WT, PGS and ANXA1 KO MIA PaCa-2. Protein bands were normalized on GAPDH levels. The blots were exposed to Las4000 (GE Healthcare Life Sciences) and the relative intensities of bands were determined using ImageQuant software (GE Healthcare Life Sciences). ***p < 0.001. (G) Analysis of hypodiploid (apoptotic) nuclei by cytofluorimetric assay of the effect of gemcitabine 10 μM at 24, 48 and 72 h on WT, PGS and ANXA1 KO MIA PaCa-2 (three clones B11, D6, G5). The data are representative of 5 experiments with similar results. (H) Western blot for pro- and cleaved caspase-3. WT, PGS and ANXA1 KO MIA PaCa-2 clones were treated or not with gemcitabine 10 μM for 72 h. Protein bands were normalized on tubulin levels. Image is representative of three independent experiments performed on WT, PGS, B11, D6 and G5 ANXA1 KO clones.

Figure 5

(A) Cell surface expression of FPR-1 and FPR-2 in WT, PGS and ANXA1 KO MIA PaCa-2 cells was analyzed by flow cytometry. The violet areas in the plots are relative to secondary antibody alone. FPR-1 and FPR-2 signals are showed in green. (B) Wound Healing assay on ANXA1 KO, PGS and WT MIA PaCa-2 treated or not with fMLP (50 nM), Ac2-26 (1 μM), Boc-1 (10 μM) fMLP + Boc-1 and Ac2-26 + Boc-1. The migration rate has been calculated as reported in Methods section, *p < 0.05, **p < 0.01 and ***p < 0.001 vs untreated controls; ^▫p < 0.05, ^▫▫p < 0.01 and ^▫▫▫p < 0.001 vs WT and PGS MIA PaCa-2 relative experimental points. Invasiveness rates were measured in (C) *p < 0.05, **p < 0.01 and ***p < 0.001 vs untreated control; ▫p < 0.05, ▫▫p < 0.01 and ▫▫▫p < 0.001 vs WT and PGS MIA PaCa-2 relative experimental points. Data represent mean cell counts of 12 separate fields per well ± SEM of 5 independent experiments. The data relative to ANXA1 KO are representative of three analyzed clones.

Figure 6

(A) Average body weight of mice measured weekly from the implantation until the sacrifice. (B) Histogram of tumour weights. (C) An exemplar image, including also the spleens, of the tumour volumes generated in pancreas by WT, PGS and ANXA1 KO MIA PaCa-2. (D) Photos of mice liver. Metastatic lesions were indicated by yellow arrows. (E) Liver sections have been stained through H&E. Metastatic foci were labelled by stars; normal tissue areas were marked by triangles. Bar = 100 μm. (F) Quantitative analyses of metastasis number on liver surface. ***p < 0.001. (G) Metastasis foci number on H&E tissue sections. ***p < 0.001.