MKK4 suppresses metastatic colonization by multiple highly metastatic prostate cancer cell lines through a transient impairment in cell cycle progression

Abstract

Metastatic dissemination in prostate cancer is often early, but not all cancer cells form clinical metastases. Map kinase kinase 4 (MKK4) suppresses metastasis in a preclinical prostate cancer model. We hypothesize that MKK4 will specifically inhibit metastatic colonization through impaired proliferation. Three highly metastatic rat prostate cancer cell lines (AT6.1, Mat-Lu and AT3.1) were employed. Stably over-expressing HA-MKK4 or vector control lines were injected into immunocompromised mice. These experiments validated that HA-MKK4 specifically affects metastatic colonization and increases survival. Median survival (days) with HA-MKK4 vs. vector was 42 vs. 28 (p < 0.0001) for AT6.1, 25 vs. 19 (p < 0.0001) for Mat-Lu and 27 vs. 20 (p < 0.0001) for AT3.1. HA-MKK4 suppresses colonization within 14 days post dissemination, after which exponential proliferation resumes. Although overt metastases retain HA-MKK4, it is inactive within these lesions. Nonetheless, metastasis-derived cell lines were shown to retain functional HA-MKK4 and like their parental HA-MKK4 line are suppressed for experimental metastasis formation in vivo. Disseminated AT6.1-HA-MKK4 cells were analyzed and were found to have an alteration in cell cycle. Specifically, there was an accumulation of cells in G1-phase (p = 0.024) and decrease in S-phase (p = 0.037) compared with vector. In multiple prostate cancer lines, HA-MKK4 suppresses an early step in metastatic colonization. These data support a model in which MKK4 activation at the metastatic site causes a cell-cycle arrest, which is eventually overcome despite presence of functional HA-MKK4. Further studies will specifically interrogate the regulation of MKK4 activation within the metastatic microenvironment and the down-stream molecular events critical for metastasis suppression.

Figure 1. Summary of the derivation of the family of Dunning rat prostatic cancers used in this study

The Dunning rat prostate cancer lines include multiple cell lines of various histologies, aggressiveness, and pattern of spread derived from a spontaneously originating prostate adenocarcinoma. Adapted from Ref. .

Figure 2. MKK4-mediated suppression of spontaneous metastasis formation is due to impaired metastatic colonization

A. Schema for spontaneous metastasis assay. Upper Panel Injection of 2 ×10⁵ AT6.1-Vector or AT6.1-HA-MKK4 cells subcutaneously into the flank of immunodeficient mice or syngeneic rats yields progressively growing tumors. Lower Panel The location and disposition of cancer cells at specific time points during spontaneous metastasis: disseminated cells (DC), microscopic metastases (MM), and overt metastases (OM) found in lungs at characteristic days post injection (dpi). B. Quantitation of the number of disseminated cells within the lungs at 21 dpi using q-RT PCR. Data is shown as the mean ± standard error [i.e. AT6.1-HA-MKK4 (8680±4174 cells) vs. AT6.1-Vector (19684±5842 cells) p=0.156]. C. Histologic appearance of microscopic metastases 21 dpi. Arrows denote foci of metastatic AT6.1-Vector and AT6.1-HA-MKK4 cells.

Figure 3. Ectopic expression of MKK4 suppresses metastatic colonization of AT6.1 prostate cancer cells in an experimental metastasis assay

A. Schematic of the results from an experimental metastasis assay in which 1×10⁴ cells are injected intravenously via tail vein venopuncture. B. Box plot of the number of overt surface lung lesions present in mice injected with either AT6.1-Vector or AT6.1-HA-MKK4 cells via experimental metastases assay. Boundaries of box are 25^th and 75^th percentiles, with solid line indicating median, dashed line indicating mean, whiskers indicating 10^th and 90^th percentiles. The mean (± standard error of the mean) number of overt metastases AT6.1-Vector was 54.9 ± 6.0 as compared to 6.8 ± 1.6 for AT6.1-HA-cells (p<0.001). C. Representative gross pathology of lungs harvested from mice 28 dpi of either AT6.1-Vector or AT6.1-HA-MKK4 cells. D. Kaplan-Meier survival analyses showed that the median survival for AT6.1 Vector animals was 28 days vs. 42 days for AT6.1-HA-MKK4 animals (p<0.0001).

Figure 4. Ectopic HA-MKK4 specifically inhibits Mat-Lu and AT3.1 metastatic colonization

A. Biochemical characterization of Mat-Vector, Mat-Lu-HA-MKK4, AT3.1-Vector and AT3.1-HA-MKK4 cell lines. Left: Lysates of three independent Vector and HA-MKK4 clones from each cell line were immunoblotted (IB) for the expression of a HA-tagged protein, p38, and JNK. Actin was used as a loading control. Right: Kinase assays (KA) showing functionality of ectopically expressed HA-MKK4. B. Representative gross pathology of lungs harvested at 21 dpi from mice injected via tail vein with Mat-Lu-Vector and Mat-Lu-HA-MKK4 cells. C. Box plot representation of the mean number ± standard error of the mean of overt surface experimental metastases formed by Mat-Lu-Vector was 25.3 ± 7.7 as compared to 4 ± 1.4 for Mat-Lu-HA-MKK4 cells (p=0.045). D. Kaplan-Meier survival analyses showed that the median survival for Mat-Lu-Vector animals was 19 days vs. 25 days for Mat-Lu-HA-MKK4 animals (p<0.0001). E. Representative gross pathology of lungs harvested at 21 dpi from mice injected with AT3.1-Vector and AT3.1-HA-MKK4 cells. F. Box plot representation of the mean number ± standard error of the mean of overt surface metastases formed by AT3.1-Vector was 38.3 ± 7.6 as compared to 2.3 ± 1.1 for AT3.1-HA-MKK4 cells (p<0.001). G. Kaplan-Meier survival analyses showed that the median survival for AT3.1-Vector animals was 20 days vs. 27 days for AT3.1-HA-MKK4 animals (p<0.0001).

Figure 5. Transient suppression of metastatic colonization imparted by MKK4 not due to clonal loss or dysfunction of HA-MKK4 construct

A. Non-linear regression analysis of overt surface lung lesions as a function of time. Each point represents the mean ± SEM surface metastases from a minimum of 5 animals per time point. The Adjusted R²=0.67. The calculated t₀ delay for the AT6.1-HA-MKK4 curve is 15.7 days relative to the AT6.1-Vector curve. B. Representative immunoblots (top) from>30 AT6.1-MKK4 overt experimental metastases and (bottom) from multiple independent metastasis derived cell lines (MDLs) demonstrating presence of HA-MKK4 construct within overt metastases and MDL’s. C. Representative in vitro kinase assay (top) AT6.1-MKK4 overt metastases and (bottom) from multiple independent metastasis derived cell lines (MDLs) showing inactive HA-MKK4 kinase within overt metastases; however HA-MKK4 remains functionally intact and activatable in cell lines derived from overt metastases. D. Persistent suppression of metastatic colonization at 28 DPI with MDL’s. The mean and SEM for surface metastases were 80.5±9.7 for Vector, 19.8±6.3 for HA-MKK4 parental, and 19.4±3.5 for HA-MKK4 MDLs (p=0.97, MDL versus parental AT6.1-vector p<0.0001).

Figure 6. Ectopic expression of HA-MKK4 leads to a shift in cell cycle from S to G1 within disseminated AT6.1 prostate cancer cells

Disseminated AT6.1-Vector and AT6.1-HA-MKK4 cells were isolated from lungs at 7 and 14 days post injection and their cell cycle distribution was assessed via flow cytometry. A. Tabular details of the number of cells analyzed and percentage of cells in each of the cell cycle. B. Graphical representation of the cell cycle data. Bars represent the percentage of cells in each phase of the cell cycle with error bars representing standard error of the mean.