Livin expression is an independent factor in rectal cancer patients with or without preoperative radiotherapy

The current study got approval from the ethics committee of Linkoping University and was in compliance with the Helsinki Declaration. The present study included 144 rectal cancer patients from the Southeast Swedish Health Care region who participated in the Swedish Rectal Cancer Trial between 1987 and 1990 [3]. Of the 144 patients, 77 underwent tumor resection alone and 67 underwent preoperative RT and tumor resection. None of the patients had received chemotherapy before surgery. Besides the tumor specimens, matched normal mucosa adjacent to the tumor tissue were collected from 71 cases, distant (4-35 cm from the primary tumor) normal mucosa were collected from 110 cases, and metastases in the regional lymph nodes were collected from 47 cases. The mean patient age was 66 years (range, 36-85 years). The median follow-up was 85 months. RT was given to 25 Gy in five fractions during a median of 8.5 Days (range, 6-18 days) [3]. Surgery was then performed in a median of 3.4 days (0-11 days) after RT. The mean tumor distance to the anal verge was 7.4 cm in the surgical group and 8.6 cm in the surgery plus RT group (P = 0.10). Other patient and tumor characteristics were presented in Table 1. No statistically significant differences were found between the two groups.

Representative paraffin-embedded tissue blocks were selected for the tissue microarray. Three morphologically representative regions were chosen in each block and three cyclindrical core core tissue specimens (0.6 mm in diameter) were taken from these areas, inserted in another paraffin block. Sections from the second block were cut into 5 μm chips using a microtome, mounted on microscopic slides. The tissue microarrays were constructed using a manual arrayer (Beecher Inc., WI).

IHC for Livin expression was done on 5-μM tissue microarray sections from paraffin-embedded surgical specimens. The sections were baked in an oven at 60°C for over 6 hrs and then deparaffinized with xylene and rehydrated with a series of decreasing concentrations of ethanol. To demask antigen epitopes, the sections were soaked in DIVA solution (Biocare Medical, CA) in a high pressure cooker at 125°C for 30 sec after which the sections were cooled to 90°C for 10 sec and then kept in room temperature for 30 min followed by washing in phosphate buffered solution (PBS, pH 7.4). To inhibit endogenous peroxidase activity, the sections were incubated with 3% H2O2-methanol for 20 min. After blocking with power block solution (Spring Bioscience, CA) for 10 min, the sections were incubated with goat anti-Livin antibody (RnD, MN) at a concentration of 2.5 μg/ml at 4°C overnight. EnVision anti-goat Polymeric conjugate (Dako, Carpinteria, CA) was subsequently applied for 30 min. The slides were washed in PBS and the peroxidase reaction was performed using 3,3’-diaminobenzidin (Sigma Chemical, St. Louis, MO) and 3% H2O2. Finally, hematoxylin was used for counterstaining. Sections known to show positive staining for Livin were included for each turn, receiving either the primary antibody or control isotype Ig, as positive or negative controls, whereas there was no staining in the negative controls.

The IHC results of the Livin in tissue specimens were the mean of scores by two independent authors (Z.-Y. D., and H. Z., who is a pathologist) in a blinded fashion without knowledge of the clinicopathological or biological information. Each investigator estimated the proportion of cells stained and the intensity of staining in the whole section. The intensity in epithelial cells or tumor cells was scored as 0 (negative staining), 1 (weak staining exhibited as light yellow), 2 (moderate staining exhibited as yellow brown), and 3 (strong staining exhibited as brown). If there was a discrepancy in individual scores, then both investigators re-evaluated the slides together to reach a consensus before combining the individual scores. To avoid an artificial effect, the cells on the margins of the sections and in areas with poor morphology were not counted.

Proliferation in the cancer cells was measured using IHC for Ki-67 as an indicator (n = 115). Low and high proliferation were defined in sections where <32% or ≥32% of cancer cells expressed Ki-67 [16]. The data for P53 (n = 139), MAT8 (n = 124) and ATM (n = 66) of primary rectal cancers determined by IHC were taken from previous studies performed with the same cases used in the present study at out laboratory [17‐19]. Apoptosis was detected by the terminal deoxynucleotidy transferase-mediated dUTP-biotin nick end labeling (TUNEL) assay [20].

Human colon cancer cell lines RKO and SW620 were purchased from the American Type Culture Collection. Cells were maintained in Minimal Essential Medium (Sigma-Aldrich) supplemented with 10% FBS (GIBCO, Invitrogen, Carlsbad, USA), 1.5 mM L-glutamine (GIBCO) and 1X PEST (GIBCO) at 37°C in a 5% CO2 incubator. Mycoplasma contamination was excluded from the cells by using a commercially available PCR kit (PromoKine, Heidelberg, Germany). The endogenous Livin was knock-down with ON-TARGETplus^R SMARTpool siRNA against Livin (L-004391-00-5551, Thermo Fisher) by using DharmaFect 2 transfection reagent (Dharmacon) according to the manufacturer’ instructions. The transfection mixture consisted of antibiotic-free culture medium and 100 nM siRNA. The Dharmacon Non-Targeting scramble siRNA was used as a negative control.

The effect of radiotherapy on the proliferation of colon cancer cells was quantified by WST-1 assay (Roche) according to the manufacture’s instructions. Briefly, cells were seeded in 96-well tissue culture plate, in 100 μl culture medium. 10 μl of WST-1 assay solution was added to each well and the cells were further incubated at 37°C for 2-3 hrs. The absorbance was measured at a wavelength of 450 nm on a VersaMax microplate reader (Molecular Devices, Sunnyvale, CA). Untreated cells served as the indicator of 100% cell viability.

By immunostaining, Livin expression was predominantly detected in the cytoplasm of epithelial cells of normal mucosa, and tumor cells of primary cancers and lymph node metastases, with little staining in the nuclei (Figure 1A and Additional file 1). For further analysis of the study, only the staining of the cytoplasmic Livin was measured and presented. Of the 144 primary tumors examined, Livin expression was negative in 13 cases (9%), weak in 32 cases (22%), moderate in 72 cases (50%), and strong in 27 cases (19%). The frequency of high expression (moderate or strong staining) of Livin was significantly increased from distant normal mucosa (17%, 19/110; P = 0.028) or adjacent normal mucosa (17%, 12/71; P = 0.051) to primary tumors (69%, 99/144), while the expression was not different from primary tumors to lymph node metastases (40%, 19/47; P = 0.357). There was no significant difference between distant and adjacent normal mucosa (P = 0.367, Figure 1B).

Compared with the cases without preoperative RT, the frequency of Livin-positive expression in primary cancers with preoperative RT was decreased from 97% to 83% (P = 0.004), while in the distant normal mucosa, the frequency of Livin-positive expression was increased (71% VS 89%, P = 0.021). However, Livin-positive cases in adjacent normal mucosa were not significantly influenced by RT (75% VS 76%, P = 0.931). The frequency of Livin-positive cases in the lymph node metastases was unchanged in cases receiving preoperative RT (86% VS 90%, P =0.986).

In primary cancers, the expression of Livin was associated with lower frequency of stage I and higher frequency of stage II, III or IV cases (P = 0.044), and related to poor differentiation of cancers (P = 0.033, Figure 2A). Patients with Livin-positive or negative tumors had similar local (P = 0.647) and distant (P = 0.280) recurrence. The expression of Livin was not related to gender, or age (P > 0.05). 15.4% (2/13) and 39.7% (52/131) patients with Livin-negative or positive tumors died at 180 months after surgery, and the difference tended to be statistically significant (P = 0.091, Figure 2B). In multivariate analyses, the difference achieved statistical significance, independent of TNM stage, local and distant recurrence, grade of differentiation, gender, and age (P =0.048, Table 2).

In a subgroup analysis of the patients with preoperative RT, a statistically significant difference was observed between patients with Livin-positive or negative tumors (P = 0.047). We did not find any relationship of Livin expression to TNM stage, grade of differentiation, gender, or age in this subgroup (P >0.05). Due to a few cases with Livin-negative tumors in the subgroup of the patients without preoperative RT, we did not perform statistical analysis in the subgroup.

In the entire group, the frequency of expression of Livin was related to the expression of Ki-67 (P =0.027), p53 (P =0.012), MAT8 (P =0.020) and inversely related to ATM (P =0.007).

In the patients without preoperative RT, high expression of Livin was still related to MAT8 (P =0.032) and p53 (P =0.048), and inversely related to ATM (P =0.037, Figure 3). In the patients with preoperative RT, no trend was observed toward any of the relationship (P >0.05).

The endogenous Livin in SW620 and RKO colon cancer cells was knocked-down by siRNA. The non-targeting scramble siRNA was used as controls. The proliferation of cancer cells after irradiation was monitored continuously by WST-1 method. The colon cancer cells with Livin knock-down exhibited significant decrease in proliferation at each time point, compared scramble siRNA treated controls (P <0.05, Figure 4).