Twenty human CRC tissue specimens were obtained from 12 males and 8 females, (average age, 68.25 years; range, 21–87 years) by surgical resection between May 2011 and June 2012 in Jilin University Second Hospital (Jilin, China). Written informed patient consent was obtained and approval was acquired from the Jilin University Second Hospital Ethics Committee (no. 2012–43). Tissue microarrays (TMAs) were constructed. Histological tumor grade of colon and sigmoid colon tissue were included in the analysis. Expression of tTG was associated with patient demographics, adjuvant treatment regimens and histological parameters.

Specimens and cells were examined using light microscopy (Eclipse TE-2000-U equipped with an attached SXM1200F digital camera; Nikon, Tokyo, Japan) following haematoxylin and eosin (HE) staining.

Paraffin-embedded slices (4-μm thick) were probed with anti-human tTG monoclonal antibody (Sigma-Aldrich, St. Louis, MO, USA; dilution, 1:300) at 4°C overnight. Sections were immersed in 0.3% H₂O₂ in absolute methanol for 15 min to block endogenous peroxidase. Color was developed using chromagen 3,3′-diaminobenzidine (DAB substrate kit and Immunohistochemistry kit; Biosynthesis Biotechnology, Beijing, China). Slices were counterstained with hematoxylin, mounted on glass coverslips and sealed with neutral resin.

UCT-116 human CRC cell lines were donated from Jilin University Institute of Regenerative Medicine. UCT-116 cells were routinely cultured in Dulbecco’s modified Eagle’s media (DMEM; Gibco-BRL, Life Technologies, Carlsbad, CA, USA) supplemented with 20% fetal bovine serum (FBS; Gibco-BRL) and 50 U/ml antibiotics under the conditions of 5% CO₂ at 37°C. Following trypsinization, cells were incubated in DMEM with 0.5% FBS for 24 h. Cells were treated with 2.5 μmol/l cantharidinate, provided by Associate Professor Yang of the Second Hospital of Jilin University and 2.5 μmol/l fluorouracil (Shanghai Hanhong Group, Shanghai, China), with untreated cells used as a control, for 48 h.

Total RNA was extracted using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions. First strand cDNA was synthesized using PrimeScript™RT enzyme mix I, oligo dT primers and random hexamers (Takara Bio, Inc., Shiga, Japan). qPCR analysis was performed using first strand cDNA, forward and reverse primers and the SYBR premix Ex Taq™ Green II kit (Takara). Primers were synthesized by Sangon Biotech Company (Shanghai, China) and the sequences were as follows: Forward: 5′-GGCACAGTCAAGGCTGAGAATG-3′ and reverse: 5′-ATGGTGGTGAAGACGCCAGTA-3′ for glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and forward: 5′-GACAAGCGCATCACACAGACA-3′ and reverse: 5′-TCTTTCGTTAGAGCCAAGGCC-3′ for tTG. Reaction and signal detection were measured in triplicate independently by an iCyder iQ real-time PCR Detection System (Bio-Rad, Hercules, CA, USA). mRNA levels were calculated as the relative expression ratio compared with GAPDH.

Statistical analysis of data was performed using SPSS Version 11 for Windows (SPSS, Inc., Chicago, IL, USA). All data are presented as the mean ± SEM. Statistical comparisons were determined by Student’s t-test and P<0.05 was considered to indicate a statistically significant difference.

The present study was performed on a TMA, constructed from surgically resected samples of patients with varying grades of differentiation of CRC. The demographics of the patients are shown in Table I as well as the clinicopathological features of the colon and the grade of differentiation. The incidence of CRC of grades IIB, IIC, IIIB and IIIA in males was 8.3, 8.3, 8.3 and 0%, respectively. The incidence of CRC of grades IIB, IIC, IIIB and IIIA in females was 0, 0, 25 and 12.5%, respectively. The incidence of colon and sigmoid colon cancer was 75 and 25% in males and 40 and 60% in females, respectively.

Following standard H&E staining, normal colon tissue and the CRC TMAs were observed to have representative histological structures by microscopy (Fig. 1). The immunohistochemical staining of microarray samples was representative (Fig. 2A and B).

Immunohistochemical staining was used to determine the expression of tTG protein in human CRC tissue. The results showed that tTG was expressed in the membrane and cytoplasm of the normal tissue of patients with CRC (Fig. 2A). The expression of tTG markedly increased in CRC tissue. tTG was primarily expressed in the tumor and interstitial regions (Fig. 2B). The expression of tTG exhibited a significant difference between CRC and normal tissue (P<0.01; Fig 2C).

Pathological changes of UCT-116 human CRC cells in various groups are shown in Fig. 3A–C. The quantity of UCT-116 cells decreased 48 h following the application of fluorouracil and cantharidinate (Fig. 3B and C). This result suggested that cantharidinate may exert an accessorial effect and may reduce the chemotherapeutic time, subsequently decreasing the side-effects of the treatment of human CRC.

Fig. 4 shows tTG immunohistochemical staining in UCT-116 cells. Immunohistochemical staining detected a high level of tTG in the untreated UCT-116 cell control (Fig. 4A and D). Cantharidinate and fluorouracil treatment decreased the level of tTG in UCT-116 cells significantly (P<0.05; Fig. 4B–D). This result suggested that cantharidinate may inhibit the expression of tTG and block tumor growth.

To determine whether the tTG mRNA level in UCT-116 cells altered following the application of cantharidinate, qPCR analysis was performed. In Fig. 5, the expression of tTG mRNA in the cantharidinate group decreased to 0.69-fold that of the untreated cancer cell control (P<0.05). The level of tTG in the cantharidinate group was approximately the same as that in the fluorouracil group. This result suggested that cantharidinate may kill tumor cells by inhibiting tTG mRNA expression.