Specifically targeting ERK1 or ERK2 kills melanoma cells

Abstract

Background: Overcoming the notorious apoptotic resistance of melanoma cells remains a therapeutic challenge given dismal survival of patients with metastatic melanoma. However, recent clinical trials using a BRAF inhibitor revealed encouraging results for patients with advanced BRAF mutant bearing melanoma, but drug resistance accompanied by recovery of phospho-ERK (pERK) activity present challenges for this approach. While ERK1 and ERK2 are similar in amino acid composition and are frequently not distinguished in clinical reports, the possibility they regulate distinct biological functions in melanoma is largely unexplored.

Methods: Rather than indirectly inhibiting pERK by targeting upstream kinases such as BRAF or MEK, we directly (and near completely) reduced ERK1 and ERK2 using short hairpin RNAs (shRNAs) to achieve sustained inhibition of pERK1 and/or pERK2.

Results and discussion: Using A375 melanoma cells containing activating BRAFV600E mutation, silencing ERK1 or ERK2 revealed some differences in their biological roles, but also shared roles by reduced cell proliferation, colony formation in soft agar and induced apoptosis. By contrast, chemical mediated inhibition of mutant BRAF (PLX4032) or MEK (PD0325901) triggered less killing of melanoma cells, although they did inhibit proliferation. Death of melanoma cells by silencing ERK1 and/or ERK2 was caspase dependent and accompanied by increased levels of Bak, Bad and Bim, with reduction in p-Bad and detection of activated Bax levels and loss of mitochondrial membrane permeability. Rare treatment resistant clones accompanied silencing of either ERK1 and/or ERK2. Unexpectedly, directly targeting ERK levels also led to reduction in upstream levels of BRAF, CRAF and pMEK, thereby reinforcing the importance of silencing ERK as regards killing and bypassing drug resistance.

Conclusions: Selectively knocking down ERK1 and/or ERK2 killed A375 melanoma cells and also increased the ability of PLX4032 to kill A375 cells. Thus, a new therapeutic window is open for future clinical trials in which agents targeting ERK1 and ERK2 should be considered in patients with melanoma.

Figure 1

ERK1 and/or ERK2 shRNA treatment of A375 cells selectively reduces total and phosphorylated levels of respective proteins at days 2, 4 and 6; accompanied by killing of melanoma cells. A. Western blot showing total and phosphorylated protein levels at indicated time intervals following exposure to shRNAs as indicated. SC: scrambled control; E1: ERK1; E2: ERK2; E1/2: ERK1 and ERK2. Actin levels confirm equal loading (left side panel). Data presented is representative of three independent experiments. Right side panel shows the relative density measurement of day 4 treatment levels for indicated proteins. B. Relative killing percentages on days 2, 4, and 6 following exposures to indicated shRNAs. Percentage of dead cells as described in Methods section using Annexin-PI and FACS analysis. Histograms represent the means +/- SEM from four independent experiments. Statistical significance portrayed as follows: *p < 0.002; **p < 0.0001. C. Relative killing percentages on days 2, 4, and 6 following exposures to indicated shRNAs. Percentage of cells undergoing apoptosis (cells with sub-Go DNA content) as described in Methods section Histograms represent the means +/- SEM from three independent experiments. Statistical significance portrayed as follows: *p < 0.03; **p < 0.001.

Figure 2

ERK1 and/or ERK2 shRNA treatment of A375 cells reduces clonogenic expansion and growth of colonies. A. Colony formation in soft agar after 21 days using A375 cells exposed to shRNAs (upper panel-phase contrast microscopic appearance after crystal violet staining). Quantification of clone number per 10x microscopic field for cells exposed to shRNAs for duplicate experiments performed with triplicate wells. (lower panel; all values using ERK1 and/or ERK2 shRNA compared to SC shRNA, p < 0.001). B. Appearance of A375 cells after exposure (9 days) to indicated shRNAs revealing crystal violet stained clones readily visible in the SC shRNA treated cultures, but rare to absent in the other wells (upper panel). Quantitation of clone number per well for A375 cells exposed for 9 days to indicated shRNAs for duplicate experiments performed with triplicate wells (lower panel; all values using ERK1 and/or ERK2 shRNA compared to SC shRNA, p < 0.001).

Figure 3

ERK1 and/or ERK2 shRNA triggers caspase cascade including PARP cleavage, and caspase inhibitor partial block of killing of A375 cells. A. Western blot analysis of caspase 3, caspase 8 and PARP including both intact (inactive) and cleaved (active) forms in A375 cells on days 2, 4 and 6 following exposure to shRNAs. GAPDH confirms equivalent loading. B). Addition of pan-caspase inhibitor, ZVAD (20 mM; last 2 days) reduces killing of A375 cells triggered by ERK1 and/or ERK2 shRNA. Histograms represent the means +/- SEM from 3 independent experiments. Statistical significance portrayed as follows: *p < 0.01, **p < 0.02, ***p < 0.002.

Figure 4

Decreased mitochondrial outer membrane potential by silencing ERK1 and/or ERK2; accompanied by increased levels of activated Bax. A. Treatment with ERK1 and/or ERK2 shRNAs at indicated time points increase outer mitochondrial membrane permeability as assessed by TMRE labeling as described in Methods (upper panel). Increased presence of activated Bax detected in A375 cells treated for 4 days with ERK1 and/or ERK2 shRNA (lower panel). B. Western blot analysis of selected proteins mediating pro-survival (upper panel) versus pro-apoptosis (lower panel) in A375 cells exposed for 4 days with either shRNAs. GAPDH levels confirm equivalent protein loading.

Figure 5

PD0325901 or PLX4032 treatment of A375 melanoma cells triggers rapid reduction in pERK1 and pERK2 levels but these levels are restored after 4 days with PLX4032; accompanied by increased killing, and alterations in protein levels regulating survival and apoptosis. A. Kinetic changes of selected protein levels by immunoblot analysis altered by treatment with PD0325901 (1 μM) or PLX4032 (5 μM) for 2, 4, 8, 24, 48 hr of drug exposure (left side panel). B. Similar analysis of protein levels treated with PD0325901 or PLX4032 after 4 days of drug exposure (right side panel). GAPDH levels confirm equal protein loading. C. Relative killing percentages of A375 cells on days 2, 4, and 6 following exposure to PD0325901 (1 μM) or PLX4032 (5 μM). Percentage of dead cells as described in Methods section. Killing at 2 days by either drug (p < 0.01), after 4 days (p < 0.001) and after 6 days (PD0325901, p < 0.001; PLX4032, p < 0.01) based on 3 independent experiments. Statistical significance portrayed as follows: *p < 0.01, **p < 0.001. D. Western blot analysis of selected proteins mediating pro-survival (left side panel) versus pro-apoptosis (right side panel) in A375 cells exposed for 4 days to PD0325901 (1 μM) or PLX4032 (5 μM). GAPDH indicates equivalent protein loading.

Figure 6

Silencing ERK1 and/or ERK2 reduces protein levels of upstream kinases after 4 days including BRAF, CRAF, and phospho MEK; while PD0325901 or PLX4032 exposure variably influences these kinases, with less to no effects of any treatments on BRAF mRNA levels. A. Western blot of upstream kinases in A375 cells 4 days after treatment with shRNAs (left side panel), or after exposure to PD0325901 (1 μM) or PLX4032 (5 μM) on right side panel. GAPDH confirms equivalent protein loading. B. BRAF mRNA levels after 4 days of exposure to PD0325901 orPLX4032 (upper panel); or after treatment with shRNAs (lower panel). Histograms represent the means +/- SD based on three independent experiments.

Figure 7

Silencing ERK1 and/or ERK2 increase sensitivity of A375 cells to killing by PLX4032. A. Relative killing percentages of A375 cells by PLX4032 (5 mM; 4 days) alone or when combined with indicated shRNAs. A375 cells were infected with shRNAs for a total of 4 days and PLX4032 added for the last 2 days. Note compared to either PLX4032 alone or the shRNAs alone, when combined there is increased killing (p < 0.001). Histograms represent the means +/- SEM based on 3 independent experiments. Statistical significance portrayed as follows: **p < 0.001.