MMP7 is required to mediate cell invasion and tumor formation upon Plakophilin3 loss

Abstract

Plakophilin3 (PKP3) loss results in increased transformation in multiple cell lines in vitro and increased tumor formation in vivo. A microarray analysis performed in the PKP3 knockdown clones, identified an inflammation associated gene signature in cell lines derived from stratified epithelia as opposed to cell lines derived from simple epithelia. However, in contrast to the inflammation associated gene signature, the expression of MMP7 was increased upon PKP3 knockdown in all the cell lines tested. Using vector driven RNA interference, it was demonstrated that MMP7 was required for in-vitro cell migration and invasion and tumor formation in vivo. The increase in MMP7 levels was due to the increase in levels of the Phosphatase of Regenerating Liver3 (PRL3), which is observed upon PKP3 loss. The results suggest that MMP7 over-expression may be one of the mechanisms by which PKP3 loss leads to increased cell invasion and tumor formation.

Fig 1. PKP3 loss leads to the generation of an inflammation associated signature in cell lines derived from stratified epithelia.

The Y-axis in all panels reflects the fold change in transcription, which is calculated as described in the Materials and Methods. The X-axis indicates the clone name. mRNA was prepared from the vector controls (vec) or PKP3 knockdown clones (sh-pkp3-1 and shpkp3-2) derived from either HCT116, FBM or HaCaT cells as indicated. (A) Real time PCR assays were performed using oligonucleotides specific to PKP3 and GAPDH. Relative expression of PKP3 in the HCT116, HaCaT and FBM derived PKP3 knockdown clones (shpkp3-1 and shpkp3-2) was compared to the respective vector control clones (vec). Expression of GAPDH has been used for normalization. (B) Real time PCRs were performed using oligonucleotides specific for IL6, SAA1, S100A8, S100A9, CCL2, CBS and GAPDH, in HaCaT and FBM derived PKP3 knockdown clones and the respective vector controls. Expression of GAPDH has been used for normalization. (C) Real time PCR was performed using oligonucleotides specific to MMP7 and GAPDH, with cDNA obtained from the vector control and PKP3 knockdown clones derived from the three cell types under study. Expression of GAPDH has been used for normalization. The standard errors are plotted and student’s t test was performed (* indicates a p value <0.01).

Fig 2. Loss of MMP7 leads to a decrease in transformation in cells lacking PKP3.

(A) mRNA prepared from HCT116 derived PKP3 knockdown cells transfected with the vector control (shpkp3-2 + vec) or the MMP7 knockdown construct (shpkp3-2 + shMMP7-1 and shpkp3-2 + MMP7-2) was used as a substrate for reverse transcriptase followed by real time PCR reactions using oligonucleotides specific for MMP7. All expression was normalized to the levels of GAPDH. The fold change is graphed on the Y-axis and the clone name is on the X-axis. Note that MMP7 levels are lowered in the double knockdown clones as compared to the vector control. The standard errors are plotted and student’s t test was performed (* indicates a p value <0.01). (B) 75μg of a whole cell extract (WCE) was resolved on 12% SDS PAGE gels followed by Western blotting with antibodies specific to PKP3. Note that PKP3 levels are lower in clones with a PKP3 knockdown. Western blots for β actin served as a loading control (upper panels). 100μg of acetone precipitated cell supernatants were resolved on 12% SDS PAGE gels followed by Western blotting with antibodies specific to MMP7. Note that MMP7 levels are higher in supernatants prepared from the PKP3 knockdown cells as compared to the vector controls and the levels are lower in the double knockdown clones. The same blot was stained with Ponceau stain to demonstrate equal loading of proteins (lower panels). (C) Scratch wound healing assays were performed on the HCT116 derived vector control (vec), PKP3 knockdown clones (shpkp3-1 and shpkp3-2), shpkp3-2 derived vector control clone (shpkp3-2+vec) and shpkp3-2 derived MMP7 knockdown clones (shpkp3-2+shMMP7-1 and shpkp3-2+shMMP7-2) as described. (D) Matrigel invasion assays were performed in Boyden’s chambers for HCT116 derived vector control cells, PKP3 knockdown clones and the double knockdown clones. The number of cells observed in ten random fields of the membrane for each clone was determined as described in Materials and Methods, representative images of for each clone are shown. The mean and standard deviation of three independent experiments are plotted. Note that loss of PKP3 leads to an increase in invasion as compared to the vector control and this phenotype is reversed in the double knockdown clones. (E) Soft agar colony formation assay was performed with the HCT116 derived vector control and pkp3 knockdown clones, the shpkp3-2 derived double knockdown clones and the shpkp3-2 derived vector control clone. The mean and standard deviation of three independent experiments is plotted. (F) 1x10⁶ cells of the shpkp3-2 derived vector control or double knockdown clones were injected sub-cutaneously into nude mice and allowed to develop tumors. The table shows the number of mice injected with the respective clones and the number of mice among them which were able to develop tumors. Wherever indicated the p value was calculated using a student’s t-test.

Fig 3. MMP7 expression decreases upon inhibition of PRL3 activity.

(A) The HCT116 derived PKP3 knockdown clones (shpkp3-1 and shpkp3-2) or the vector control (vec) were treated with either the vehicle control (DMSO) or 5 or 10 μM PRL-3 inhibitor-1(PRL-3i) for 24 hours. The mRNA prepared from the treated cells was used as a substrate for reverse transcriptase followed by real time PCR reactions using oligonucleotides specific for MMP7. All expression was normalized to the levels of GAPDH. The fold change is graphed on the Y-axis and the clone name is on the X-axis. The standard errors are plotted and student’s t test was performed. Note that MMP7 levels are lowered upon treatment with PRL-3 inhibitor. (B) The HCT116 derived vector control (vec) and PKP3 knockdown clones (shpkp3-1 and shpkp3-2) were treated with either DMSO or 10 μM PRL3 inhibitor-1(PRL3i) for 24 hours or 48 hours. The cell supernatants were collected and a100μg of acetone precipitated protein was resolved on 12% SDS PAGE gels followed by Western blotting with antibodies to MMP7. The same blot was stained with Ponceau stain to indicate equal loading. (C) The whole cell lysates of HCT116 derived vector control and plakophilin3 knockdown clones treated with DMSO or PRL3i for 24 hours were resolved on 12% poly-acrylamide gel. This was followed by Western blotting with antibodies to PKP3, β actin and PRL3. The molecular weights of these proteins are indicated in brackets.