Complementary role of MEK activation, ERK hyperexpression, and PAC1 downregulation in determining the activation of ERK in human acute leukemias. Normal unselected BM (N), purified CD34⁺ (P), and acute leukemia samples were subjected to immunoblot (IB) and RT-PCR analysis as described in Materials and Methods. For immunoblot analysis, lysates containing the same amount of protein were applied to each lane; 25 μg of protein for phosphorylated ERK1/2 (ERK-P); 50 μg of protein for phosphorylated MEK1/2 (MEK-P); 25 μg of protein for total ERK1/2 (ERK); 25 μg of protein for -tubulin. In all immunoblot analyses, K562 samples (K) were loaded together as the evidence of same exposure time in autoradiogram. Separated proteins were transferred to the immobilon-P membrane and stained with phospho-specific ERK1/2 antibody or phospho-specific MEK1/2 antibody or total ERK1/2 antibody or -tubulin antibody, respectively. In immunoblot analysis of total ERK using 25 μg protein, the ERK1 band was not visualized because ERK1 detected by total ERK1/2 antibody was much less abundant than ERK2. For RT-PCR analysis, PCR was performed with PAC1 primer for 22 cycles, with β-actin primers for 25 cycles. PCR products derived from 20 ng of total RNA were applied to each lane. Leukemia samples above the lanes from (A) to (D) correspond with those in Table 1.