Introduction

There are some evidences indicating that apoptosis resistance may play an important role in development and progression of endometrial carcinoma (EC) [1–4]. Cells resistant to apoptosis are likely to be resistant to therapy and be responsible for relapses.

Apoptosis can be initiated by two main mechanisms: the intrinsic pathway, which has its origin in the mitochondria and the extrinsic apoptotic pathway, triggered by the activation of death receptors situated in the cell surface [5]. A final common feature for execution of the apoptotic program is the activation of a cascade of caspases, which are proteases that have a cysteine containing active site that cleaves protein substrates at specific amino acid motifs containing an aspartic acid residue.

The extrinsic pathway is activated by the members of the tumor necrosis factor (TNF) superfamily [6–8]. After ligand binding, the activated receptors recruit an adaptor protein called Fas-Associated protein with Dead Domain (FADD). FADD binds to the receptor through interactions with pro-caspase-8 to form a complex at the receptor called DISC. In turn, active caspase-8 activates effector caspases, such as caspase-3, causing the cell to undergo apoptosis. The tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and the Fas ligand belong to the pro-apoptotic cytokines of the TNF superfamily. TRAIL induces apoptosis in many types of cancer cells. However, some neoplastic cells are resistant to TRAIL [9]. TRAIL can bind four different receptors DR4/TRAIL-R1, DR5/TRAIL-R2, DcR1/TRAIL-R3, and DcR2/TRAIL-R4. DR4 and DR5 are functional receptors that induce apoptosis upon ligation with TRAIL. DcR1 and DcR2 (TRAILR3 and TRAILR4), known as decoy receptors, lack the intracellular domains required to induce apoptosis [10].

The different steps of the apoptotic process are deregulated in cancer [11, 12]. Down-regulation of death receptors, such as Fas, and loss-of-function mutations and deletions on receptors DR4 and DR5 have been detected in some tumors. Moreover, up-regulation of decoy receptors (DcR1 and DcR2) that compete with the functional receptors for the ligand binding may result in a diminished binding of TRAIL to their functional receptors [10]. Such a mechanism has been suggested as a potential cause for apoptosis resistance in some types of tumor.

The main aim of the present study was to assess the expression of DcR1 (TRAILR3) in a series of normal endometrium (NE) and EC in order to assess the possible role of DcR1 expression in apoptosis resistance in EC.

Materials and methods

Tissues

The material was composed of formalin-fixed, paraffin-embedded tissue samples from 142 patients for immunohistochemical evaluation and an independent series of fresh frozen tissue samples from 47 patients for quantitative reverse transcription-polymerase chain reaction (RT-PCR) analysis.

Tissue microarrays

Two tissue microarrays (TMA) were constructed. The first TMA was constructed from 80 paraffin-embedded samples of NE in different phases of the menstrual cycle (20 proliferative, 50 secretory, ten menstrual) obtained from the surgical pathology files from Hospital Universitari Arnau de Vilanova de Lleida. The second TMA was composed of 62 EC obtained from the surgical pathology files of Hospital Universitari Arnau de Vilanova de Lleida and Hospital de Sant Pau, Barcelona, Spain. The specimens were collected between 2003 and 2006; they were consecutively obtained, without any criteria for selection. They included 22 endometrioid carcinomas (EEC) grade I, 24 EEC grade II, seven EEC grade III, three mixed endometrioid and non-endometrioid carcinomas (all of them with a serous component, which in one case also coexisted with clear cell elements), two non-endometrioid carcinomas (NEEC; all of them serous carcinomas) grade III, and the epithelial components of four malignant mixed Müllerian tumors (all of them with a non-endometrioid component of serous type). Forty-three tumors were stage I, 11 were stage II, and eight were stage III. Staging information was complete in all cases.

The study was approved by the local ethics committee, and the samples were obtained with a specific consent.

A tissue arrayer device (Beecher Instrument, MD) was used to construct the TMA. Briefly, all the samples were histologically reviewed, and representative areas were marked in the corresponding paraffin blocks. Two selected cylinders (0.6 mm in largest diameter) from two different areas were included in each case. Control normal tissues from the same EC specimens were also included.

Immunohistochemical study

TMA blocks were sectioned at a thickness of 3 μm, dried for 16 h at 56º before being dewaxed in xylene and rehydrated through a graded ethanol series, and washed with phosphate-buffered saline. Antigen retrieval was achieved by heat treatment in a pressure cooker for 2 min in EDTA (pH 8.9). Before staining the sections, endogenous peroxidase was blocked. A DcR1 polyclonal rabbit antibody was used (1:100 dilution; PA1-21270, Affinity Bioreagents). After incubation, the reaction was visualized with the EnVision Detection Kit (DAKO) using diaminobenzidine chromogen as a substrate. Sections were counterstained with hematoxylin. Appropriate positive and negative controls were also tested. Immunohistochemical results were evaluated by two pathologists, by following uniform pre-established criteria. DcR1 immunostaining was graded semi-quantitatively by considering the percentage and intensity of the staining. A histological score was obtained from each sample, which ranged from 0 (no immunoreaction) to 300 (maximum immunoreactivity). The score was obtained by applying the following formula, \( {\text{Histoscore}} = 1 \times \;\left( {\% \;{\text{light}}\;{\text{staining}}} \right) + 2 \times \;\left( {\% \;{\text{moderate}}\;{\text{staining}}} \right) + 3 \times \;\left( {\% \;{\text{strong}}\;{\text{staining}}} \right) \). The reliability of such score for interpretation of immunohistochemical staining in EC TMAs has been shown previously [2–4, 13, 14]. Since each TMA included two different tumor cylinders from each case, immunohistochemical evaluation was done after examining both samples.

Endometrial epithelial tissue explant isolation

Fresh frozen (normal and neoplastic) endometrial tissue was analyzed. The normal endometrial tissue samples were classified according to the phase in the menstrual cycle (six proliferative, 12 secretory, one inactive). Moreover, 28 tumor samples were obtained from patients with EC; they corresponded to 24 EEC and four NEEC. Ten tumors were grade I, 13 were grade II, and five were grade III. Twelve tumors were stage IA, four were stage IB, six were stage IC, and six were stage II. Age of the patients ranged from 49 to 88 years of age (mean, 68.4 years). The samples were collected in Dulbecco's Modified Eagle Medium (DMEM), chopped in 1-mm pieces, and incubated with collagenase in DMEM for 1.5 h at 37°C with periodic mixing. Afterwards, tissue was mechanically dissociated through a 10-ml pipette and a 1-ml blue tip and resuspended in 2 ml of fresh DMEM medium. To separate endometrial epithelial cells from the stromal fraction, the dissociated tissue was seeded on top of 8 ml of DMEM medium, and tissue was allowed to sediment by gravity. This step was repeated three times. Finally, tissue explants were resuspended in DMEM supplemented with 10% fetal bovine serum, 1 mM sodium pyruvate, 2 mM l-glutamine, and 1% of penicillin/streptomycin (Sigma) at 37°C with saturating humidity and 5% CO2. They were frozen until use. Immunostaining for keratin was used to confirm that the resulting tissue explants were exclusively composed of epithelial cells.

Real-time PCR analysis

Total cellular RNA was extracted using Trizol reagent (Life Technologies, Switzerland) and 1 μg of total RNA was retrotranscribed into cDNA using Taqman Reverse Transcription Reagents (P/N N808-0234). The RT reaction was then used as template for a 25-μl reaction for real-time detection of DcR1 using TaqMan Technology on an Applied Biosystems 7000 sequence detection system. Applied Biosystems assays-on-demand primers and TaqMan MGB probes (FAM™ dye-labeled reporter and no fluorescent quencher) for the target gene and predeveloped glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (VIC-dye-labeled probe) were used. Gene expression quantification was performed in separate tubes (singleplex) for both the target gene and endogenous control gene using the primer and probe sequences for DcR1 and GAPDH (assay ID: HS 00174949 m1 and HS 99999905 m1, Applied Biosystems). The thermal cycler conditions were UNG activation 2 min at 50°C, AmpliTaq activation 95°C for 10 min, denaturation 95°C for 15 s, and annealing/extension 60°C for 1 min (repeated 40 times). Duplicate control values from two independent RNA extractions were analyzed with quantitative relative software using the comparative CT (ΔΔCT) method as described by the manufacturer. The amount of target \( \left( {{2^{ - \Delta \Delta {\text{CT}}}}} \right) \) was obtained by normalizing to an endogenous reference (GAPDH) and relative to a calibrator.

Statistical analysis

Statistical analysis was performed on a database using SPSS for Windows (version 11.5; SPSS Inc., Chicago, Ill). Immunohistochemical results were compared by the Mann–Whitney U and Student’s t tests when applicable. Associations between qualitative variables were assessed by χ 2 or Fisher exact tests. Correlations between quantitative variables were established through Pearson’s and Spearman’s rho tests. Statistical significance was set at p ≤ 0.05.

Results

Immunohistochemistry

In NE, DcR1 was evaluated in 64 cases. Three cases were excluded because there was no representative normal tissue left during the construction of the array. Moreover, the number of cylinders missed in the construction of the TMA was thirteen. Overall, positive DcR1 cytoplasmic staining was demonstrated in 51 cases (79.6%), and the H score varied from ten to 200 (mean, 90). DcR1 immunostaining varied according to the different phases of the menstrual cycle. DcR1 immunoexpression was significantly higher in the secretory phase (mean H score, 82.4; Fig. 1) than in the proliferative endometrium (mean H score, 31.3; Fig. 2; p < 0.001, Fig. 1). The staining was cytoplasmic.

Fig. 1
figure 1

DcR1 expression in normal endometrium during the secretory phase (early stage)

Full size image
Fig. 2
figure 2

DcR1 immunostaining in normal endometriun during the proliferative phase

Full size image

In EC, DcR1 was evaluated in 54 of the 62 cases (Fig. 3). Two cases were excluded because there was no representative tumor tissue in the TMA sections (only non-neoplastic stromal tissue or necrosis was present), and six cases were missed in the TMA construction. Overall, DcR1 cytoplasmic immunostaining was detected in 53 of the 54 cases (98.1%). The staining was cytoplasmic and heterogeneous, and the H score ranged from five to 290 (mean, 83.4).

Fig. 3
figure 3

DcR1 immunohistochemical staining in endometrial carcinoma

Full size image

Overall, there was no significant correlation between DcR1 immunostaining between NE (all phases during the endometrial cycle) and EC (p = 0.215). However, when we compared the differences between EC and the proliferative endometrium, we found that the mean H scores were different (83.4 and 31.3, respectively) and that difference had statistical significance (p = 0.001)

No significant differences were seen between DcR1 expression and histological type, grade, and stage. The mean H score for EEC and NEEC were 93.6 (range, 5–290), and 39.28 (range, 15–70), respectively, and such difference did not reach statistical significance (p = 0.231). Moreover, mean H scores were 94 for grade I tumors (range, 5–220), 92.1 for grade II tumors (range, 20–290), and 51.3 for grade III tumors (range, 15–125), and such a difference was not statistically significant (p = 0.107). Finally, there was no significant difference between DcR1 expression and stage. Mean H scores were 91 (range, 30–180) for stage IA, 81.8 (range. 30–2220) for stage IB, 85 (range, 5–240) for stage IC, 146.6 (range, 20–290) for stage IIA, 78.7 (range, 15–150) for stage IIB, and 74.2 (range, 15–170) for stage IIIA.

Quantitative RT-PCR

By quantitative real-time PCR, all normal endometrial tissues (proliferative, secretory, or inactive) had similar levels of DcR1expression (0.8–1.7 RQ), which were considered the basal levels of DcR1 expression in normal tissue. Increased DcR1 expression (≥5-fold higher than the basal levels) was detected in 13 of 28 EC (46.4%). High DcR1 expression levels were found in ECs of different stages: IA, four of 12 (33%); IB, two of four (50%); IC, four of six (66%); and IIA and IIB, three of six (50%; Table 1). No statistical significance was obtained when evaluating DcR1 expression with clinicopathological parameters, such as histological type, grade, or stage. However, EC infiltrating the myometrium deeply (stage IC) showed the highest proportion of positivity for DcR1. Moreover, it was clear that one third of the EC samples exhibited increased DCR1 mRNA levels in comparison with normal endometrial tissue.

Table 1 DcR1 expression by QRT-PCR
Full size table

Discussion

Deregulation of apoptosis plays an important role in development and progression of cancer [9, 11, 12]. Moreover, cells resistant to apoptosis are likely to escape the immune surveillance, but they may also be resistant to therapy. The “extrinsic pathway” is activated by ligand-bound death receptors such as TNF, Fas, or TRAIL receptors [5–8]. Different steps of the extrinsic apoptotic pathway have been found to be deregulated in many tumors. Down-regulation of death receptors, like Fas, has been detected in colorectal and hepatocellular carcinomas; and loss-of-function mutations and deletions on receptors DR4 and DR5 occur in metastatic breast carcinomas and lung carcinomas. In other tumors, there is up-regulation of decoy receptors (DcR1 and DcR2) that compete with the functional receptors for the ligand binding and may result in a diminished binding of TRAIL to their functional receptors and decreased apoptosis [10].

ECs can be classified into two main clinicopathological variants: type I, EEC and type II, NEEC [15]. Recent molecular studies have shown that these two forms of EC follow different molecular pathways [15]. Type I, EEC, characteristically shows five different molecular alterations that may coexist in some cases: microsatellite instability [16] and mutations of the PTEN [17–19], PIK3CA [20, 21], K-RAS [22], and β-catenin genes [23]. In contrast, type II, NEEC, is usually associated with mutations in the p53 tumor suppressor gene and loss of heterozygosity at many different loci [24, 25].

There are many evidences suggesting that alteration of apoptosis is important in development and progression of endometrioid adenocarcinomas of the endometrium. They frequently show activation of the PI3K pathway, either by inactivating mutations in PTEN [17–19] or by activating mutations in the PIK3CA gene [20, 21]. Activation of the PI3K pathway may lead to enhanced survival by different mechanisms that ultimately cause phosphorylation of BAD, inhibition of caspase-9, or down-regulation of Fas ligand. p53 alterations, which are characteristic of non-endometrioid carcinomas, may also occur in endometrioid tumors, particularly in those neoplasms showing overlapping features between types I and II tumors, and they may have an impact in apoptosis at several different levels [24, 25]. Also, members of the Bcl-2 family of genes, like BAX [1] and other proteins involved in apoptotic control, like NF-kB or survivin, are involved in endometrial carcinoma [2, 3]. Moreover, an important protein responsible for apoptosis resistance in endometrial carcinoma is FLICE-like inhibitory protein (FLIP), which shares a high degree of homology with caspase-8 but lacks protease activity and [4]. FLIP is frequently expressed in endometrial carcinomas and is regulated by CK2 [13, 26]. However, it is evident that FLIP overexpression does not explain, alone, by itself, resistance of TRAIL-induced apoptosis in endometrial carcinoma. For that reason, it is reasonable to look for other proteins involved in such a phenomenon.

Decoy receptors (DcR1 and DcR2) can bind TRAIL but are deficient in signaling to caspase-8. Up-regulation of decoy receptors (DcR1 and DcR2) competing with the functional receptors for the ligand binding and resulting in a diminished binding of TRAIL to their functional receptors has been regarded as a mechanism of inhibition of TRAIL-induced apoptosis in cancer cells [10]. However, it has recently been suggested that, rather than competing for binding for TRAIL, decoy receptors may inhibit the apoptotic signal by forming ligand-independent protein complexes, which would be deficient in mediating a robust death signal [27]. Moreover, in one study, formation of these heterocomplexes was ligand-dependent but correlated with reduced TRAIL-mediated cell death [28].

In this study, we have tried to assess the expression in DcR1 in normal endometrial tissue and endometrial carcinoma. By immunohistochemistry, DcR1 expression was higher in the secretory phase in comparison with the proliferative phase. This is not unusual for a protein which may be responsible for resistance for TRAIL-induced apoptosis, since the many proteins involved in the negative control of apoptotic cell death show elevated levels during the secretory phase. Moreover, DcR1 was also frequently expressed in endometrial carcinoma. Although there was no significant difference between endometrial carcinoma and normal endometrium, when considering all phases of the menstrual cycle, there were some differences between endometrial carcinoma and the normal endometrial tissue during the proliferative phase, which is probably the best control among normal endometrial tissues. When we checked the samples by quantitative real-time RT-PCR, there was a significant difference between normal endometrial tissue and endometrial carcinoma, and the levels of DcR1 in endometrial carcinoma showed some correlation with stage, since myoinvasive tumors (stage IC) showed the highest proportion of high DCR1 expression. Interestingly, in normal endometrium, we did not see differences in mRNA DcR1 among the different phases of the menstrual cycle. Lack of correlation between mRNA levels and protein expression are not unusual when evaluating gene expression in tumor samples. Technical problems could be responsible for the lack of correlation between mRNA and protein levels, but they have been minimized by the design of the methodological strategy. It is worth mentioning that mRNA was evaluated in tissue explants that were exclusively composed of epithelial cells, minimizing the possibility of stromal contamination that would interfere with the molecular interpretation. The discrepancy between the data obtained by immunohistochemistry with that of quantitative RT-PCR could be also explained by the presence of posttranslational processing of DcR1, which would modify protein expression. In summary, the results suggest that DcR1 expression occurs in normal endometrium and endometrial carcinoma and could play a role in resistance to TRAIL-induced apoptotic cell death during menstrual cycle and endometrial carcinoma tumor progression.

The study has some limitations. For example, it is not possible to assess the effect of DCR1 on apoptosis by correlating its expression with the apoptotic index or the expression of some apoptotic markers. It has long been observed that, frequently, those tumors that show higher levels of apoptosis resistance are precisely those that exhibit higher apoptotic indexes. Such a paradox is due to the fact that apoptosis resistance is usually associated with high grade and high cell proliferation, and these two features are, by themselves, associated with high apoptotic indexes. Moreover, further functional studies assessing cell viability after DCR1 down-regulation by transfection with DCR1 siRNA will be required to confirm the role of DCR1 on apoptosis resistance in endometrial carcinoma.