Figure 1.
β1i is frequently expressed in colorectal and pancreatic cancer tissues.
(a) Immunoblotting analysis for β1i using protein lysates prepared from nonmalignant (N) and cancerous (T) colonic tissues from the same donors (n = 4 pairs). β-actin was used as a loading control. The band intensities for β1i and β-actin were densitometrically analyzed and used to obtain the relative β1i expression normalized to β-actin levels. (b) Immunohistochemical staining for β1i using a tumor array containing 153 evaluable tumor colon tissue specimens. The intensities of β1i positive staining were evaluated on a scale of 0 to 3. Approximately 70% (46 specimens had intensity of grade 1, 38 with grade 2, and 8 with grade 3, out of 153 total tumor specimens) of colorectal tissues have positive (the staining intensity ≥1) β1i staining. (c) Immunohistochemical staining for β1i using a tumor array containing 43 evaluable tumor pancreatic tissue specimens. The intensities of β1i positive staining were evaluated on a scale of 0 to 2. Approximately 53% (9 specimens with intensity grade 1, and 14 with grade 2, out of 43 total tumor specimens) of pancreatic cancer tissues had ≥1 β1i staining intensity.
Figure 2.
Ac-PAL-AMC is hydrolyzed selectively by β1i.
(a) Chemical structure of Ac-PAL-AMC. (b) Ac-PAL-AMC hydrolysis over time by purified constitutive proteasome (cross), purified immunoproteasome (closed circle), or purified immunoproteasome pre-treated with UK101 (open square). A linear increase was seen in immunoproteasome-containing wells only. (c) The rate of hydrolysis of Ac-PAL-AMC or Suc-LLVY-AMC in the presence of Panc-1 cell extract (10 µg). Hydrolysis of Ac-PAL-AMC was almost completely inhibited by UK101 pre-treatment (10 µM). Hydrolysis of Suc-LLVY-AMC was partially inhibited by UK-101 pre-treatment. Epoxomicin (Epx, 10 µM), a broadly-acting proteasome inhibitor, was used as a positive control.
Figure 3.
The catalytic activity of β1i is not affected by the PSMB9 codon 60 polymorphisms.
(a) The catalytic activity of β1i measured by rate of hydrolysis of Ac-PAL-AMC in human cancer cell lines with H/H or R/H genotype (blank bars) and R/R genotype (filled bars). Variable activity was observed among cell lines. (b) Rate of Ac-PAL-AMC hydrolysis in cell lines with H/H or R/H genotype (open circles) or R/R genotype (closed squares). Rate of hydrolysis did not show any statistically significant difference between H+ and H- cell lines.
Figure 4.
Determination β1i expression levels across multiple human cancer cell lines.
An equivalent amount (10 µg) of cell extracts were subjected to immunoblotting with an antibody to β1i along with serially diluted H23 cell extracts (calibration standards). Densitometric analyses of band intensities were performed to quantify relative expression levels of β1i. Cell lines expression high levels of β1i are shown in (A) and cell lines with low expression of β1i are shown in (B).
Figure 5.
The expression levels of β1i are not affected by the PSMB9 codon 60 polymorphisms, but associated with the β1i activity.
(a) β1i expression in cell lines with H/H or R/H (open circle) and R/R genotypes (closed square). β1i expression did not show statistically significant difference between H+ and H- cell lines. (b) Correlation of Ac-PAL-AMC hydrolysis with β1i expression in all cell lines tested. A highly significant correlation was noted between the expression levels of β1i and its catalytic activity (R2 = 0.86, p<0.0001).