Anti-Akt, anti-phospho-specific Akt (p-Akt; Ser473), anti-ERK, and anti-phospho-specific ERK (p-ERK) antibodies were purchased from Cell Signaling Technology (Beverly, MA, USA). Anti-β-actin antibody was purchased from Sigma.

The human umbilical vein endothelial cells (HUVECs) were obtained from Lonza (Walkersville, MD, USA) and maintained in EGM-2 medium containing the bullet kit according to the supplier’s instructions. All the cells were incubated at 37°C in a humidified atmosphere of 5% CO₂.

Recombinant REG Iα protein was generated in insect cells using the Bac-to-Bac expression system (Invitrogen, Carlsbad, CA, USA) by Kitayama Labes (Ina, Japan). Full length human REG Iα cDNA was cloned and inserted into the pFastBac vector (Invitrogen). The constructed vector was then transformed into E. coli DH 10Bac, and recombinant Bacmid-REG Iα was produced by transposition. Then, Spodoptera frugiperda (Sf9) insect cells were infected with Bacmid-REG Iα to generate the recombinant baculoviruses carrying human REG Iα cDNA. The recombinant baculovirus particles were harvested in the culture supernatant, and used to infect Sf9 insect cells in a large volume (1 L) of culture medium. The supernatant (crude extract) including the secreted REG Iα protein was then incubated with Ni-NTA agarose (Qiagen) and purified by elution through SP-Sepharose (GE Healthcare Life Science).

Total RNA was extracted from each cell line using Trizol reagent (Invitrogen). Five micrograms of total RNA were reverse-transcribed using oligo dT primer (Applied Biosystems, Branchburg, NJ, USA) and 200 U of Superscript™ II reverse transcriptase (Invitrogen) in a total volume of 20 μl. For the following PCR, pairs of oligonucleotide primers for human EXTL3 (11) and human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (9) were prepared. Human EXTL3: 5′-CAACCGATTCTTACCCTGG-3′ (sense) and 5′-GGAAGTTCATGGCAATGTCC-3′ (antisense); human GAPDH: 5′-GGCTGCTTTTAACTCTGGTA-3′ (sense and 5′-ATGCCAGTGAGCTTCCCGT-3′ (antisense). One microliter of RT product (cDNA) was amplified by PCR as previously described (11).

Following treatment with or without a reagent, the cells were lysed in protein extraction buffer as previously reported (12). Protein extract (25 μg) was fractionated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane. The membrane was incubated with a primary antibody and then with a peroxidase-conjugated secondary antibody. Proteins were detected using an enhanced chemiluminescence system (Amersham Biosciences, Buckinghamshire, UK).

A total of 31 gastric cancer tissues was obtained from specimens that were resected surgically at Dokkyo University School of Medicine. The tissue specimens were fixed in a 10% formalin solution and embedded in paraffin. This study was approved by the Dokkyo University Surgical Pathology Committee and an informed consent was obtained from all the participant patients.

Multiple hematoxylin and eosin-stained sections of all 31 lesions were examined (Table I). The following factors were determined for all the patients and lesions; age, sex, tumor size, tumor location, Lauren’s histological classification, tumor invasion, lymph node metastases and tumor stage according to the system of the American Joint Committee on Cancer.

Immunohistochemical staining for CD34 and REG Iα was performed with an Envision kit (DAKO, Kyoto, Japan) as described previously (3,13), using anti-CD34 antibody (1:200; Santa Cruz Biotechnology, Santa Cruz, CA, Japan) and anti-REG Iα antibody (1:500). Finally, the sections were incubated in 3,3′-diaminobenzide tetrahydrochloride with 0.05% H₂O₂ for 3 min and then counterstained with Mayer’s hematoxylin. To evaluate the immunoreactivity of REG Iα protein, at least 500 tumor cells were counted in five different visual fields for each sample of the cancerous tissues. A specimen was considered positive for REG Iα protein if 20% of the tumor cells were positively stained (9). To evaluate angiogenesis in the tumors, MVD was assessed by immunostaining with the anti-CD34 antibody as described above. Five different fields (×200) were digitally photographed with a high-resolution microscope (DP20, Olympus, Tokyo, Japan), and the obtained images were analyzed using NIH ImageJ1.47 image analysis software (http://rsbweb.nih.gov/ij). MVD was quantified as the percentage of the microvascular area relative to the tumor stroma in each image and the results were averaged (14).

HUVECs were seeded in complete medium in 96-well plates (1×10⁴ cells/well) and 6-well plates (2×10⁵ cells/well) for cell growth and apoptosis assay, respectively. After 24 h, the cells were washed in a serum-free medium and then incubated with or without REG Iα protein for the indicated time. For the cell growth assay, the treated cells were incubated with Premix WST-1 reagent (Takara, Tokyo, Japan) for 1 h and the plates were read at 450 and 600 nm in a spectrophotometer (Molecular Devices, Sunnyvale, CA, USA). For the apoptosis assay, the treated cells were collected, washed with PBS, and incubated with Annexin V-FITC and propidium iodide (PI) in binding buffer in accordance with the manufacturer’s protocol (MEBCYTO-Apoptosis Kit; MBL, Ina, Japan). The stained cells were analyzed on a FACScalibur flow cytometer (Becton-Dickinson, Franklin Lakes, NJ, USA) and the data obtained were analyzed using CellQuest software (Becton-Dickinson).

All values were expressed as the mean ± SEM. The data for MVD were analyzed using unpaired two-tailed t-test. Chi-squared analyses were performed to determine the correlation between various pathological parameters and Fisher’s exact test was also performed when necessary. P-values of <0.05 were considered to indicate statistical significance.

EXTL3 was ubiquitously expressed not only in the epithelial cells, but also in the endothelial cells in the normal gastric mucosa (Fig. 1A and B). In gastric cancer tissues, EXTL3 was expressed in tumor vascular cells as well as cancer cells (Fig. 1C and D).

Before examining the effect of REG Iα protein on the endothelial cells, we tested the expression of EXTL3 in HUVEC. Subsequently, we confirmed that expression of EXTL3 and its gene product was detectable in the cells by RT-PCR and western blot analysis (Fig. 1E), suggesting that HUVECs have the capability of responding to REG Iα stimulation.

The effect of REG Iα protein on intracellular signaling was investigated in the HUVEC cells. The expression of p-ERK and p-Akt was enhanced by REG Iα stimulation (10–100 nM) (Fig. 2A). The expression of p-ERK peaked in the HUVEC cells at 15 min after REG Iα stimulation, and that of p-Akt was enhanced from 15 min and sustained for 60 min (Fig. 2B). We then examined whether anti-REG Iα antibody inhibits the REG Iα-induced signaling in HUVEC cells. As shown in Fig. 2C, the basal level of p-ERK and p-Akt expression was decreased by treatment with REG Iα antibody. Moreover, the increased expression of p-ERK and p-Akt in REG Iα-treated HUVEC cells was attenuated by concomitant administration of anti-REG Iα antibody.

To clarify the role of REG Iα protein in angiogenesis, we examined the growth and anti-apoptosis effects of REG Iα protein on HUVEC cells in vitro. The rate of WST-1 cleavage was significantly and dose-dependently increased in REG Iα-treated HUVEC cells (Fig. 3A). Conversely, the increase in the level of WST-1 cleavage in REG Iα-treated cells was significantly reduced to almost the control level by addition of anti-REG Iα antibody (Fig. 3B).

In control preparations, deprivation of growth factors in complete culture medium induced cell apoptosis and death. However, HUVEC cells treated with REG Iα protein (10 nM) showed significantly lower Annexin V positivity, than the control cells (Fig. 3C and D). Similarly, the percentage of PI-positive cells was significantly lower in the REG Iα-treated preparations than in the controls (Fig. 3C and D). On the other hand, the decrease of Annexin V or PI positivity in REG Iα-treated HUVEC cells was restored to the control level by concomitant administration of anti-REG Iα antibody (Fig. 3D).

Among 31 samples of gastric cancer tissues, 19 (61.3%) were positive for REG Iα expression. Expression of REG Iα was significantly associated with the prevalence of lymph node metastasis and tended to correlate with the tumor stage (Table II). MVD was significantly higher in gastric cancers at an advanced stage. In addition, MVD tended to be higher in gastric cancers with lymph node metastasis. Furthermore, we investigated the relationship between REG Iα expression and MVD and observed that MVD was significantly higher in REG Iα-positive gastric cancers (Fig. 4).