Hypoxic regulation of cytoglobin and neuroglobin expression in human normal and tumor tissues

Cygb mRNA was detected in all aerobic GBM cell lines (controls), and when cultured under hypoxic conditions (0.6% O₂ × 48 h), all cell lines showed significantly increased expression of Cygb mRNA (M006x, M010b and U87T, p < 0.05; M006xLo and M059J, p < 0.001) (Figure 1). Two cell lines also showed significant increases in Cygb transcript at earlier time points (M006xLo at 6 and 24 h, p < 0.001; M059J at 24 h, p < 0.01). Consistent with qRT-PCR results, Cygb protein was detected in all aerobic GBM cell lines with hypoxia-tolerant cells (M006x, M006xLo) showing the highest basal levels of Cygb. After 48 h of hypoxia, Cygb protein was significantly increased in four of five cell lines (M006xLo, M010B, and M059J, p < 0.001; U87 T, p < 0.01), with the greatest relative increases seen in hypoxia-sensitive cells M010b and M059J (Figure 2). Cygb protein was also significantly increased in M006xLo cells after 24 h of hypoxia (p < 0.01). M006x cells showed a modest increase in Cygb protein after hypoxia, but this was not significant (Figure 2A). There was no correlation between the magnitude of Cygb protein increase after hypoxia and the respective basal levels of each cell line.

Immunohistochemical staining of TMAs showed that Ngb and Cygb were variably expressed in both normal tissues and tumors. As reported previously [39], Ngb was found primarily in the cytoplasm, whereas Cygb was detected in both nucleus and cytoplasm. Among normal tissues, Ngb expression was greatest in bone, liver, sigmoid colon, rectum and kidney, with low or negligible levels in skin, muscle and lung (Table 1). Tissues with highest Cygb expression were stomach fundus, kidney tubules and cerebellum. However, compared to Ngb, Cygb levels were reduced overall and were low or absent in approximately half of the tissues. There was no significant correlation between levels of Cygb and Ngb protein in these normal tissues (p < 0.07). However, two features of Cygb and Ngb expression in normal tissues were of particular interest. First, in several normal tissues, expression of Cygb and Ngb was not uniform throughout the tissue. Rather, distinct cell types or tissue structures frequently showed strong positive staining while the remainder of the tissue showed weak or absent staining. Among the structures strongly stained for Cygb were ductal cells of the breast and kidney, secretory cells of the salivary gland, white pulp/lymph of the spleen, and tips of desquamating cells found in several types of normal tissue. Ngb staining was also prominent among ductal cells of breast, endometrium, testis, prostate and salivary gland. Both Ngb and Cygb were generally absent from stroma. Second, most normal tissues showed low or absent CA IX immunoreactivity. However, stomach, small bowel, and gallbladder showed nearly identical patterns of Ngb, Cygb and CA IX staining (Figure 3).

Tumor sections showed three distinct staining patterns for Cygb/Ngb: (1) a uniform expression throughout the non-stromal tissue; (2) distinct regions of positively staining cells among otherwise negative or weakly staining tissue; or (3) densely staining focal areas frequently including intensely positive foci. Among individual tumor sections included in the 'tumor/matched normal tissue' TMA, Ngb and Cygb expression was generally increased compared to levels observed in corresponding normal tissues. As well, Cygb and Ngb immunoreactivity showed similar patterns of distribution. Most tumor sections (24/29) showed low or absent CA IX immunoreactivity. However, the breast, liver, bladder and thyroid tumors that contained regions strongly positive for CA IX also showed in matching sections similar staining patterns for Cygb and Ngb (Figure 4).

Both Cygb and Ngb were detected in all human brain tumors, including grades I-IV astrocytomas (Figure 5), and ependymoblastomas, gangliogliomas and oligodendrogliomas (Figure 6). Among low grade astrocytomas (Grades I and II), Ngb staining intensity was significantly greater than that of Cygb (p < 0.05). Grade III astrocytomas and GBMs also showed relatively greater Ngb staining but these differences were not significant (Tables 2 and 3). In tumors that showed distinct patterns of Ngb and Cygb expression, a comparison of tissue sections showed that these proteins were expressed in the same regions of the tumor. Large, positively staining Ngb and Cygb foci were commonly observed in GBM tumors. Approximately 60% of the GBM sections were also positive for CA IX staining which was confined to regions of the tumor also positive for Cygb and Ngb.

The origin and characterization of the GBM cell lines have been published previously: the M059J (ATCC number CRL2366) and M010b cell lines are hypoxia-sensitive; the M006x and M006xLo cell lines are hypoxia-tolerant [46, 57, 58]. The U87T cell line was established following serial in vivo selection in a glioma mouse model [59] and was kindly provided by Dr. Donna Senger (University of Calgary). Its relative sensitivity to hypoxia has not been determined. All cells were maintained as monolayer cultures in DMEM/F12 media supplemented with 10% fetal calf serum and 1 mM L-glutamine in a humidified atmosphere of 5% CO₂ in air at 37°C. All tissue culture supplies were purchased from GIBCO.

A de-gassing manifold was used to generate hypoxia in vitro[60]. Exponential phase cells (~2 × 10⁵) were seeded onto 60-mm glass plates and then incubated under standard laboratory culture conditions (5% CO₂ in air) for 4 days. The medium was then replenished and the plates were transferred to aluminum chambers from which the air was evacuated and then replaced with 5% CO₂/balance N₂ until an O₂ tension of 0.6% was achieved. The sealed, air-tight aluminum chambers were then incubated at 37°C for 6-48 h. At the end of each incubation interval, the aluminum chambers were unsealed, the tissue culture plates removed, RNA and protein isolated from the cells as described in Section 2.4 and 2.5.

Total RNA from cultured cell lines was isolated using the RNeasy mini kit (QIAGEN). Reverse transcription (RT) was carried out with 1 μg total RNA. per 20 ml reaction volume using QuantiTect reverse transcription kit (QIAGEN). RT experiments were performed with GeneAmp PCR system 9700 (Applied Biosystems). Quantitative real-time PCR (qRT-PCR) analysis was performed with a 7900 HT Fast Real-Time PCR System (Applied Biosystems) using TaqMan fast universal PCR master mix and a validated TaqMan Gene Expression assay (Applied Biosystems) for the human CYGB gene (assay ID: Hs00370478_ml). Human 18 S rRNA gene (part no.: 4333760T, Applied Biosystems) was used as endogenous control. Amplification data were analyzed with SDS RQ Manager 1.2. Relative quantities of Cygb transcripts were normalized against relative quantities of the 18 S rRNA transcripts, and fold-expression changes calculated using the expression 2^-ΔΔCT.

Whole-cell lysates were prepared using complete Lysis-M buffer (Roche Diagnostics) and protein content determined using a protein assay kit (Pierce). Equal amounts of protein (50 μg) were resolved using 13% SDS -PAGE under reducing condition and electro-transferred to polyvinylidene difluoride membranes (Bio-Rad Laboratories, CA). Membranes were blocked with 5% skim milk then incubated with Cygb antibody (1:60 rabbit anti-human polyclonal antibody [sc-66855] Santa Cruz Biotechnology, Inc, CA) or α-tubulin antibody (1:5000, mouse anti-human α-tubulin monoclonal antibody, SIGMA) as loading control. Membranes were incubated with polyclonal goat anti-rabbit IgG (H+L) horseradish peroxidase (HRP)-conjugated (1:25000, Jackson ImmunoResearch Laboratories, Inc. PA) or polyclonal goat anti-mouse IgG HRP-conjugated (1:6000, DakoCytomation Denmark A/S, Demark). Bound proteins were detected using chemiluminescence reagents (SuperSignal West Pico Chemiluminescent Substrate, Thermo Scientific, IL) and visualized by exposing to X-ray film (Fuji Photo Film, Japan) that was developed using a Kodak X-OMAT 2000A processor (Eastman Kodak Company, Japan). X-ray films were scanned using an Artixscan 1800f scanner (MICROTEK, TAIWAN) and bands were analyzed using Quantity One 1-D analysis software (Bio-Rad Laboratories). The integrated areas of bands were determined and expressed in arbitrary units (AU).

Tissue microarrays (TMA) constructed from tissue sections obtained from various human solid tumors and normal adjacent tissues (IMH-346), GBMs (IMT-01256), and various brain cancers and normal adjacent tissues (IMT-01257) were purchased from IMGENEX Corporation (San Diego, CA). The manufacturer confirmed the glial origin of the brain tumors included on the arrays using immunoreactivity to anti-glial fibrillary acid protein. Immunohistochemical staining was performed according to manufacturer's protocol with modifications. TMA slides were deparaffinized by incubating in a dry oven at 62°C for 1.5 h, dewaxed in xylene, and then hydrated. TMAs were processed for antigen retrieval, incubated with blocking buffer (0.5 M glycine in PBS-0.2% brij35) and then with antibodies to Ngb (1:150, mouse anti-human Ngb monoclonal antibody, BioVendor Laboratory Medicine Inc., Czech Republic), Cygb (1:150, mouse anti-human Cygb monoclonal antibody, Abnova Corporation, Taiwan) or Ca IX (1:600, rabbit anti-human CA IX, Novus Biologicals, Inc., USA). The specificity of the primary antibodies was demonstrated by using Cygb or Ngb peptides to block positive immunostaining [Additional file 3]. Tissue endogenous peroxidase was quenched by incubation with 3% H₂O₂ for 15 min. Tissue sections were incubated with HRP-conjugated goat anti-mouse antibodies (DakoCytomation) and positive staining visualized by the chromogenic reaction of HRP with DAB (3,3' diaminobenzidine tetrahydrochloride) (DakoCytomation). Relative Cygb and Ngb expression in each tissue core was assessed by two observers who noted the overall tissue staining and the presence of positively stained focal regions. Each section was assigned an overall score, where 0 = absence of positively stained cells; 1 = a few positive cells/regions (< 10%); 2 = weak positive staining, occasional positive foci; 3 = intermediate staining, occasional positive foci (< 50%); 4 = strongly positive in most of the section, several positive foci; and 5 = strongly positive throughout with many intensely positive foci.

Data from four replicate experiments were expressed as mean ± standard error. Statistical analyses were performed using SigmaPlot 11 software (Systat Software, Inc). Differences between groups were compared using one-way ANOVA or ANOVA on ranks (Kruskal-Wallis) based on the normality and equal variance tests. To determine exactly which groups are different and the size of the difference, multiple comparisons versus control group were carried out using Bonferroni t-test and Dunnett's or Dunn's test for one-way ANOVA and ANOVA on ranks (Kruskal- Wallis), respectively, as post-hoc tests.