Glucose-Regulated Protein 78 Is a Potential Serum and Imaging Marker for Early Detection of Ovarian Cancer

Abstract

Background: Understanding malignant transformation associated with ovarian cancer (OVCA) is important to establish early detection tests. This study examined whether expression of glucose-regulated protein 78 (GRP78, marker of cellular stress) increases during OVCA development, and whether GRP78 can be detected by targeted-transvaginal ultrasound (TVUS) imaging.

Methods: Normal ovaries (n = 10), benign (n = 10) and malignant ovarian tumors at early (n = 8) and late stages (n = 16), hens with and without ovarian tumors at early and late stages (n = 10, each) were examined for GRP78 expression during OVCA development by immunohistochemistry, immunoblotting, gene expression and immunoassay. Feasibility of GRP78-targeted TVUS imaging in detecting early OVCA was examined.

Results: Compared with normal ovaries and benign tumors, intensity of GRP78 expression was higher (p < 0.0001) in OVCA patients. Compared with normal (9007.76 ± 816.54 pg/mL), serum GRP78 levels were significantly higher (p < 0.05) in patients with early (12,730.59 ± 817.35 pg/mL) and late-stage OVCA (13,930.12 ± 202.35) (p < 0.01). Compared with normal (222.62 ± 181.69 pg/mL), serum GRP78 levels increased (p < 0.05) in hens with early (590.19 ± 198.18 pg/mL) and late-stage OVCA (1261.38 ± 372.85) (p < 0.01). Compared with non-targeted, GRP78-targeted imaging enhanced signal intensity of TVUS (p < 0.0001).

Conclusions: Tissue and serum levels of GRP78 increase in association with OVCA. GRP78 offers a potential serum and imaging marker for early OVCA detection.

Keywords: chronic inflammation; early detection; glucose-regulated protein 78; laying hen; ovarian cancer; oxidative stress; targeted-imaging agents; transvaginal ultrasound imaging.

Figure 1

Expression of GRP78 by normal ovaries and malignant ovarian tumors at early and late stages; sections were stained with hematoxylin & eosin (A–C) and with anti-GRP78 antibodies (D–F). (A) Ovarian section from a normal postmenopausal ovary showing ovarian surface epithelial (OSE) cells and stroma (S). (B) Section of an ovarian serous malignant tumor (Tu) at early stage showing a compact sheath-like tissue mass surrounded by fibromuscular layer. (C) Section of an ovarian serous malignant tumor at late stage showing a papillae-like appearance of malignant cells. (D) Normal postmenopausal ovarian section showing GRP78 staining by the OSE cells. (E,F) Ovarian serous malignant tumors at early and late stages, respectively, showing GRP78 staining by the cell surface (yellow arrows) and nucleus (red arrows). Magnification = 40×.

Figure 2

Changes in the intensity of GRP78 expression during OVCA development and progression in patients. Compared with normal ovaries and ovaries with benign tumors, expression of GRP78 was significantly higher in ovaries with early-stage HGSC (p < 0.001) and increased further in late-stage HGSC (p < 0.0001). * p < 0.05, ** p < 0.01, *** p < 0.0001.

Figure 3

GRP78 protein and gene expression by normal ovaries and ovaries with HGSC at early and late stages. (A) One-dimensional Western blotting (1-D-WB) showed GRP78 expression by the normal (N), early (E) and late (L) stage OVCA (n = 4 each). Anti-GRP78 antibodies showed multiple immunoreactive bands including one at approximately 78 kDa with strong intensity from early and late stage tumors. Normal ovary showed very weak immunoreactive signal for GRP78 expression. (B) 2-D-WB of a malignant tumor (late-stage HGSC) confirmed immunoreaction for GRP78 of approximately 78 kDa. (C) GRP78 expression by the normal human ovarian surface epithelial cells (HOSE, N) and OVCAR3 (O), SKOV3 (S), Caov3 (Ca) ovarian cancer cell lines (n = 4 each). Compared to the lysate from normal cells, strong immunoreaction for GRP78 was shown by all OVCA cell lines. (D) Compared with normal, gene expression study showed strong amplification for GRP78 mRNA in early- and late-stage OVCA (p < 0.01). (E) Relative signal intensity of GRP78 protein expression detected by 1-D-WB bands shown in panel A presented as an intensity ratio (mean ± SEM). (F) Relative signal intensity of GRP78 protein expression detected in 1-D-WB bands shown in panel C presented as an intensity ratio (mean ± SEM). ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Figure 4

Spontaneous incidence of ovarian cancer in hens. (A) Ovary in a healthy laying hen showing a hierarchy of different sizes of preovulatory follicles. (B) A regressed ovary and oviduct in a hen; no preovulatory follicle is seen in the ovary. (C) OVCA at early stage in a hen. The tumor mass appears like a cauliflower (dotted lines), limited to the ovary and is accompanied with little ascites. (D) OVCA at late stage in a hen. Tumor mass is metastasized to other organs.

Figure 5

Microscopic presentation of ovarian adenocarcinomas in hens. (A) Section of a normal ovary in a hen showing a follicle (F) embedded in the stroma (S). (B) Section of a serous OVCA in a hen. The tumor (Tu) appears like a compact sheath of malignant cells surrounded by fibromuscular layers in the stroma. (C) Section of an ovarian mucinous carcinoma in a hen. The tumor consists of malignant cells of columnar morphology arranged as a single layer in glandular structure containing mucin-like secretion in the lumen. (D) Section of an ovarian endometrioid carcinoma in a hen showing back-to-back confluent tumor glands consisting of a single layer of malignant cells. G = granulosa cell layer, T = theca layer in stromal follicle, Magnification = 40×.

Figure 6

Expression of GRP78 by normal ovary and ovarian malignant tumors at early and late stages in hens. (A) Normal ovarian section showing GRP78 staining by few ovarian surface epithelial (OSE) cells (red arrows), while no staining was observed in others (black arrows). (B) Ovarian endometrioid carcinoma at early stage showing intense staining for GRP78. (C) Ovarian serous carcinoma at late stage showing malignant cells stained strongly for GRP78. The nucleus as well as surface of malignant cells showed strong immunoreactivity for GRP78 staining. (D) Compared to normal ovaries, the expression of GRP78 was significantly high in ovaries with tumor at early stage (p < 0.03). However, significant differences were not observed in the intensity of GRP78 expression among different histological types of OVCA (serous n = 4, endometrioid n = 3, mucinous n = 3). Endo = Endometrioid, Muc = Mucinous, S = Stroma, Magnification = 40×. * p < 0.05.

Figure 7

Changes in the serum expression of GRP78 during the development and progression of OVCA in patients and hens. (A) Box-plot graph shows the serum levels of GRP78 were significantly (p < 0.05) higher in patients with early-stage OVCA compared to women without OVCA. Serum level increased further in women with late-stage OVCA (p < 0.01). (B) Box-plot graph shows the serum levels of GRP78 were significantly (p < 0.05) higher in hens with early stage OVCA than normal hens and increased further in late-stage OVCA (p < 0.01). * p < 0.05, ** p < 0.01.

Figure 8

Detection of GRP78 protein by 1-dimensional and 2-dimensional Western blotting in normal ovary or ovaries with tumor in hens. (A) 1-D-WB detected GRP78 in normal ovary (N) and ovaries with early- (E) and late-stage (L) serous OVCA (n = 5 each). Compared with normal, strong expression of GRP78 is seen in early-stage and late-stage OVCA. (B) Homogenates of malignant cells from early (E) and late (L) stage HGSC showed strong immunoreactivities for GRP78. (C) Relative signal intensity of 1-D-WB bands shown in panels A and B, presented as a detected intensity ratio (mean ± SEM). (D–F) 2-D-WB using normal ovarian or tumor ovarian extracts showed specific staining for GRP78 (approximately 78 kDa) for normal (D), early-stage endometrioid (E) and late-stage serous (F) OVCA. (G) Nuclear matrix protein (NMP) from a late-stage HGSC also showed strong GRP78 staining in 2-D-WB. * p < 0.05, *** p < 0.001.

Figure 9

Enhancement in signal intensity of TVUS scanning by GRP78-targeted imaging of ovarian tumors in hens. (A) Pre-targeted scan of an ovary suspected to have solid mass (arrows) and ascites-like fluid (star). (B) Sonogram of the same ovary presented in (A) imaged after injection with GRP78-targeted imaging agents. Compared with pre-targeted imaging, post-targeted scan showed remarkable increase in imaging signals confirming the binding of imaging agents with the tumor (arrows). (C) Gross examination confirmed the presence of solid mass in the ovary accompanied with ascites. (D) Immunohistochemical examination showed intense expression of GRP78 by the serous tumor. (E) The intensity of ultrasound signals was significantly (p < 0.0001) higher in post-targeted imaging than pre-targeted imaging. OV = ovarian; S = stroma; Tu = tumor; 40× = Magnification. * p < 0.0001.