Figure 1.
KLK4-expressing SKOV-3 cells are less migratory as 2D monolayers.
A. Western blotting with anti-V5 antibody shows KLK4 expression in the conditioned media (CM) and whole cell lysate (WCL) from stable KLK4 tranfectants (KLK4-1, KLK4-2 and KLK-3; lanes 1–3), vector (Vec-1, Vec-2 and Vec-3; lanes 4–6) and native SKOV-3 (lane 7) cells, and transiently expressed wildtype KLK4, mutant-KLK4S207A (KLK4S/A), vector and mock transfectants (lanes 8–11); GAPDH was used as a loading control for WCL. B. Western blotting with anti-KLK4 antibody shows relative levels of KLK4 protein in WCL of KLK4-1, KLK4-3 clones and OVCA432, and 10 ng of recombinant (r)KLK4 protein. C. IF microscopy with anti-V5/KLK4-N terminal antibodies (green) and phalloidin (red) in KLK4-1, Vec-1 clones, native SKOV-3 or negative control (IgG). Scale bar, 20 µm. D. Transwell migration assays with KLK4-1, KLK4-2, KLK4-3, Vec-1, Vec-2 and native SKOV-3 cells; n = 3, mean ± SEM, *P<0.05.
Figure 2.
KLK4-expressing SKOV-3 cells form compact MCAs in a 3D-suspension microenvironment.
A. MCA/spheroid formation conducted in 10% FCS containing media at 4 h, day 1, 4 and 7, with representative images of KLK4-1, KLK4-2, KLK4-3, Vec-1, Vec-2 clones, native SKOV-3 and endogenous KLK4 expressing OVCA432 cells. B. Quantitative analysis of percentage of 4 h (0 time point) that formed compact MCAs (≥30 µm) after 1, 4 and 7d from the 3 KLK4 clones combined, 2 vector clones combined, native SKOV-3 and OVCA432 cells. C. MCA formation under serum free conditions by KLK4-1 clone and native SKOV-3 cells treated with 50 ng/ml active recombinant (r)KLK4or mutant-KLK4S207A (KLK4S/A) and PBS control at day 1, 4 and 7. D. Quantitative analysis of compact MCA formation of KLK4-1, SKOV-3 treated with rKLK4, KLK4S/A and PBS as a control over 1, 4 and 7d. For panels A and C, scale bars, 200 µm; for B and D, the experiment was repeated 3 times with triplicates; mean ± SEM; *P<0.05; **P<0.01; and ***P<0.001.
Figure 3.
MCAs clear mesothelial monolayers mimicking invasion into the peritoneal membrane.
A. Bright-field and fluorescence (CellTracker492) images show KLK4-1, Vector-1, SKOV-3 and OVCA432 MCA (4 h and day 3) clearance of mesothelial monolayers. Discontinuous lines indicate perimeters of the spreading MCAs. B. Quantitative analysis shows the average diameter of 10 MCAs from 3 separate experiments for above cell lines at 4 h and day 3 respectively; mean ± SE, n = 3, **P<0.01. C. IF microscopy images show mesothelial monolayer clearance of MCAs formed by KLK4-1 labeled with CellTracker492, Vector-1, OVCA432 and SKOV-3 cells stained with an E-cadherin antibody (green); both MCAs and mesothelial LP9 cells were stained with Phalloidin for F-actin (Alexa Flour 568, red) and DAPI for nuclei (blue) respectively; discontinuous lines indicate positions of multiple Z sections shown as right and bottom panels. For panels A and C, scale bars, 50 µm.
Figure 4.
Inhibition of KLK4 increased paclitaxel sensitivity.
WST-1 assay shows cell survival of KLK4-1, KLK4-2, KLK4-3, Vec-1, Vec-2 clones and native SKOV-3 MCAs after paclitaxel (A) treatment in 3D-suspension. B. Left panel, representative images of KLK4-1 MCAs on day 4 with mouse IgG and a functional KLK4 blocking antibody, PBS as a control and KLK4 selective inhibitor SFTI-FCQR (SFTI) as indicated. Right panel, representative images of OVCA432 MCAs on day 4 with PBS as a control and KLK4 selective inhibitor SFTI-FCQR (SFTI) or aprotinin as indicated. Scale bars, 200 µm. C. Cell survival determined by WST-1 assay after treatment with paclitaxel (Pac) on 3D-suspension cultured KLK4-1 clone and OVCA432 cells +/−1 µM SFTI or 5 µM aprotinin (Aprot). Experiments in panels A and C were repeated 3 times in triplicate, bars represent Mean ± SEM. Statistical significance indicated as *P<0.05, **P<0.01.
Figure 5.
KLK4 induced uPA expression in SKOV-3 cells.
A. Western blot analysis shows expression of uPA, α5 integrin (ITN) and KLK4 (V5) of 3 KLK4 and 3 vector control clones with native SKOV-3 cells as a control with GAPDH as a loading control. B. Western blot analysis shows expression of KLK4 and uPA in KLK4-1 clone transfected with siRNA KLK4 exon 1 (psilK4Ex1), and both KLK4 exon 1 and 2 knockdown constructs (psilK4Ex1+2), p-silencing scramble (psil Ctl) and mock controls. GAPDH was used as a loading control. C. Western blotting shows expression of KLK4 and uPA in KLK4-1 cells cultured as 2D-monolayers (2D), 3D-collagen I (Col I), 3D-Matrigel (Matrigel), and 3D-suspension (Susp), with GAPDH as a loading control. D. Western blot shows expression of KLK4 and uPA in serous EOC cells of primary tumors (T) and ascites (A) from 6 patients. WCL of OVCA432 MCAs serves as a positive control and GAPDH as a loading control. E. Densitometry analysis of 3 Western blots indicative of that shown in D. **P<0.05 and ***P<0.001 indicate the significantly different levels of KLK4 and uPA in ascitic (A) and primary tumor cells (T).
Figure 6.
KLK4 is associated with poor outcome of patients.
A. Phase contrast images show the similar morphology of trypan blue stained KLK4-1 MCAs and ascites-derived MCAs from a serous EOC patient. Scale bar, 50 µm. B. IF confocal microscopy with an antibody against KLK4 (green), phalloidin (red) and DAPI (blue) in ascitic serous EOC MCAs from 2 patients; scale bars, 25 µm. C. Kaplan-Meier survival analysis shows the relationship between KLK4 mRNA levels in tumor tissue samples and survival status of a cohort of 38 serous EOC patients. Left panel, progression free survival (PFS) time for patients with low KLK4 (n = 25) and high KLK4 (n = 13) levels (χ2 = 8.3, p = 0.004). Right panel, overall survival time for patients with low KLK4 (n = 25) and high KLK4 (n = 13) levels (χ2 = 4.9, p = 0.03).