Disturbed balance of expression between XIAP and Smac/DIABLO during tumour progression in renal cell carcinomas

Dysregulation of apoptosis plays an important role in tumour progression and resistance to chemotherapy. The X-linked inhibitor of apoptosis (XIAP) is considered to be the most potent caspase inhibitor of all known inhibitor of apoptosis-family members. Only recently, an antagonist of XIAP has been identified, termed Smac/DIABLO. To explore the relevance of antiapoptotic XIAP and proapoptotic Smac/DIABLO for tumour progression in renal cell carcinomas (RCCs), we analysed XIAP and Smac/DIABLO mRNA and protein expression in the primary tumour tissue from 66 RCCs of all major histological types by quantitative real-time PCR, Western blot and ELISA. X-linked inhibitor of apoptosis and Smac/DIABLO mRNA expression was found in all RCCs. Importantly, the relative XIAP mRNA expression levels significantly increased from early (pT1) to advanced (pT3) tumour stages (P=0.0002) and also with tumour dedifferentiation (P=0.04). Western blot analysis confirmed the tumour stage-dependent increase of XIAP expression on the protein level. In contrast, mRNA and protein expression levels of Smac/DIABLO did not significantly change between early and advanced tumour stages or between low and high tumour grades. Consequently, the mRNA expression ratio between antiapoptotic XIAP and proapoptotic Smac/DIABLO markedly increased during progression from early (pT1) to advanced (pT3) tumour stages. Moreover, RCCs confined within the organ capsule (pT1 and pT2) exhibited a significantly lower XIAP to Smac/DIABLO expression ratio when compared with RCCs infiltrating beyond the kidney (pT3; P=0.01). Thus, our investigation demonstrates that the delicate balance between XIAP and Smac/DIABLO expression is gradually disturbed during progression of RCCs, resulting in a relative increase of antiapoptotic XIAP over proapoptotic Smac/DIABLO, thereby probably contributing to the marked apoptosis resistance of RCC.

Resistance to apoptosis plays a critical role during tumorigenesis and tumour progression, facilitating the accumulation of transforming mutations, promoting evasion of tumour cells from immunosurveillance and contributing to resistance against anticancer drugs Reed, 1999;Ramp et al, 2000Ramp et al, , 2003aIgney and Krammer, 2002). The molecular pathways of apoptotic cell death are controlled by genes that either promote or inhibit the activation of the caspase cascade and, consequently, an imbalance between proapoptotic and antiapoptotic regulators may significantly contribute to apoptosis resistance.
Recently, the activity of caspases has been shown to be inhibited by members of the inhibitor of apoptosis protein (IAP) family (Salvesen and Duckett, 2002). Up to now, eight human IAP proteins have been identified: X-linked inhibitor of apoptosis (XIAP), survivin, c-IAP1, c-IAP2, NAIP, BRUCE, ILP-2 and ML-IAP (Livin) (Mahotka et al, , 2002aReed, 1999;Salvesen and Duckett, 2002). X-linked inhibitor of apoptosis has been shown to be the most potent caspase-inhibitory IAP family member that is able to inhibit caspase-3, -7 and -9 (Deveraux et al, 1999;Huang et al, 2001;Srinivasula et al, 2001). Moreover, XIAP exhibits E3 ligase activity and is able to ubiquitinate caspase-3 for degradation by the proteasome (Suzuki et al, 2001). X-linked inhibitor of apoptosis, therefore, acts as a potent suicide inactivator, effectively inhibiting apoptosis induced by a variety of extracellular stimuli, including anticancer drugs and ionising radiation (Li et al, 2000;Asselin et al, 2001;Holcik et al, 2001;Matsumiya et al, 2001).
Only recently, three inhibitors of XIAP have been identified, namely Second Mitochondria-derived Activator of Caspases/Direct IAP-Binding Protein with Low PI (Smac/DIABLO), Omi/HtrA2 (Hedge et al, 2002;Martins et al, 2003;Srinivasula et al, 2003) and XIAP-Associated Factor 1 (XAF1) (Fong et al, 2000;Liston et al, 2001;Verhagen and Vaux, 2002). The mechanism by which XAF1 antagonises the effects of XIAP is not very well known, but might affect the redistribution of XIAP from the cytoplasm to the nucleus (Liston et al, 2001). Omi/HtrA2 and Smac/ DIABLO are localised within the mitochondria and are released into the cytosol upon apoptotic stimuli. Cytosolic Omi/HtrA2 results in displacement of XIAP from caspases and loss of their suppressive effect on caspase activity (Hedge et al, 2002;Martins et al, 2003). Moreover, Omi/HtrA2 was shown to be able to degrade XIAP protein (Martins et al, 2003;Srinivasula et al, 2003). Cytosolic and proteolytically processed Smac/DIABLO can also disrupt the interaction between XIAP and caspases, thereby relieving the inhibition of caspase-3, -7 and -9 by XIAP (Srinivasula et al, 2001;van Loo et al, 2002;Verhagen and Vaux, 2002). Since overexpression of Smac/DIABLO was able to sensitise tumour cells against anticancer drug-and TRAIL-induced apoptosis (Zhang et al, 2001;Guo et al, 2002;Ng et al, 2002;Mc Neish et al, 2003), Smac/DIABLO is thought to be a major antagonist of XIAP.
Importantly, inhibition of apoptosis by IAPs also seems to promote tumour progression in vivo, suggesting that IAPs might be used as novel prognostic factors. Thus, AML patients with low XIAP levels exhibited a significantly longer survival time (Tamm et al, 2000). Conflicting observations, however, have been reported between XIAP expression and tumour stage or patient's survival in carcinomas of the cervix as well as non-small-cell lung carcinomas of early or advanced stages (Ferreira et al, 2001a, b;Liu et al, 2001;Hofmann et al, 2002). Preliminary investigations demonstrated expression of Smac/DIABLO in 62% of various carcinoma types (Yoo et al, 2003), but the relation between Smac/DIABLO expression and tumour staging, grading or survival has not been analysed yet. Moreover, since XIAP and Smac/DIABLO are major opponents, it is reasonable to assume that the expression of bothrather than the expression of a single regulator -may be relevant for regulating apoptosis susceptibility.
The aim of the present investigation, therefore, was to explore the relevance of XIAP and Smac/DIABLO expression for tumour progression in a large cohort of renal cell carcinomas (RCCs) of all major histological types.

Patients and specimens
Tumour tissue samples of 66 consecutive RCCs of different histological types, grades and stages (Table 1) were obtained from patients who had undergone nephrectomy. The tumour specimens were immediately flash frozen in liquid nitrogen and stored at À801C until RNA extraction. Tumour typing, grading and staging were performed according to the principles outlined by the WHO (Mostofi and Davis, 1998), Störkel et al (1989) and the UICC (Wittekind and Wagner, 1997). The proportion of the different RCC types in our study reflects the frequency of each tumour type in a large RCC series (Mostofi and Davis, 1998).

RNA extraction
Total RNA was isolated by the guanidium thiocyanate extraction method as described by Chomczynski and Sacchi (1987). Afterwards, the total RNA was purified using the RNeasy minikit with DNAse treatment according to the manufacturer's instructions (Qiagen, Hilden, Germany) to exclude DNA contamination. The integrity of all tested total RNA samples was verified by intact 18S/ 28S rRNA bands in agarose gel electrophoresis.

Reverse transcription
For cDNA synthesis, 1 mg of total RNA was reversed transcribed in a final volume of 20 ml containing 15 mM of each dNTP (Roche, Germany), 10 U of recombinant RNasin RNase inhibitor (Promega, Heidelberg, Germany) and 2.5 U of AMV reverse transcriptase (Promega, Heidelberg, Germany) with the corresponding 1 Â RT buffer. The following sets of primers were used: an XIAP-specific oligonucleotide primer (10 pmol; 5 0 -TTC CTC GGG TAT ATG GTG TCT GAT-3 0 ) or a Smac/DIABLO-specific primer (10 pmol; 5 0 -TTC AAT CAA CGC ATA TGT GGT CTG-3 0 ) that specifically hybridise at the 3 0 regions of XIAP and Smac/DIABLO, respectively. The sequence-specific antisense GAPDH primer was: 5 0 -TGT CAT CAT ATT TGG CAG GTT T-3 0 (10 pmol). Reverse transcriptions (RTs) of both of XIAP and GAPDH as well as both of Smac/DIABLO and GAPDH were performed in the same reactions to avoid variations in the efficiency of cDNA synthesis. The RT reaction mixtures were incubated at 551C for 1 h.
PCR products were additionally checked by electrophoresis on 3% agarose gels containing ethidium bromide and visualised under UV transillumination. PCR products were confirmed by DNA sequencing. Briefly, products were excised from agarose gels and isolated by QIAquick gel extraction kit (Qiagen, Hilden, Germany). The purified PCR products were sequenced using the ABI-Prism BigDye Terminator Cycle Sequencing kit (ABI, Weiterstadt, Germany) with the respective forward and reverse primers used in the real-time PCR according to the manufacturer's protocol. Sequence analysis was carried out using an ABI-Prism 310.

Quantification of XIAP and Smac/DIABLO by an external standard
A cDNA fragment of the housekeeping gene GAPDH was cloned. Plasmids (pGEM-T-easy; Promega, Heidelberg, Germany) were isolated, linearised by SacI restriction and cleaned. Plasmid concentration was measured and the number of GAPDH copies was calculated. Serial dilutions of the linearised GAPDH plasmid from 6 Â 10 8 -3 Â 10 6 copies were prepared. For relative quantification of XIAP, Smac/DIABLO and GAPDH copy numbers, a standard curve was created by plotting the crossing point (CP) values against the number of copies. Every dilution was run at least in duplicate. The number of gene copies was calculated by the LightCycler Software (Version 3.5) according to the second derivative maximum method.

Enzyme-linked immunosorbent assay (ELISA)
For quantification of Smac/DIABLO protein expression in RCCs, quantitative sandwich ELISA was carried out using the DuoSet s IC assay kit (R&D Systems, Wiesbaden, Germany) as described by the manufacturer's manual. Briefly, proteins were isolated from 24 arbitrarily selected clear cell RCCs (pT1: n ¼ 8, pT2: n ¼ 8, pT3: n ¼ 8) and diluted to a concentration of 400 mg ml À1 . A standard curve was generated by using two-fold serial dilutions and a high standard of 5000 pg ml À1 . A 96-well microtitre plate was coated overnight at room temperature by capture antibody and, after washing with washing buffer (0.05% Tween 20 in PBS, pH 7.3), blocked with blocking buffer (1% BSA, 5% sucrose in PBS, pH 7.3, with 0.05% NaN 3 ) for 2 h, followed by washing. Samples or standards of 100 ml were added in duplicate and incubated for 2 h. After washing, 100 ml of biotinylated rabbit anti-human Smac/DIABLO antibody (R&D Systems, Wiesbaden, Germany; concentration 150 ng ml À1 ) was added to each well and incubated for 2 h. After washing again, 100 ml of streptavidin conjugated to horseradish peroxidase (1 : 200 dilution) was added and incubated for 20 min, followed by washing. In all, 100 ml of substrate solution (1 : 1 mixture of H 2 O 2 and tetramethylbenzidine) was added and incubated for 20 min, and then stopped by 2 N H 2 SO 4 . Optical density was determined immediately at 450 and 570 nm in a spectrophotometer and the values at 570 nm were subtracted from the values at 450 nm. The absorbances after subtraction were directly used for statistical analysis.

Statistical analysis
GAPDH copy numbers (external standard) were used to normalise the copy numbers of XIAP and Smac/DIABLO, calculating ratios of relative mRNA levels of XIAP, Smac/DIABLO and XIAP to Smac/ DIABLO. Statistical analysis was performed using the Mann -Whitney and Wilcoxon tests with the statistical software package SPSS 11.0 (SPSS, Chicago, USA). A P-value of less than 0.05 was considered to indicate the statistical significance.

RESULTS
Quantification of XIAP and Smac/DIABLO expression X-linked inhibitor of apoptosis, Smac/DIABLO and the housekeeping gene GAPDH could be detected at cycles 19 -42, 24 -38 and 17 -26, respectively ( Figure 1A -C). The specificity of the amplification products was confirmed by agarose gel electrophoresis, which revealed distinct bands for all PCR products (Figure 1d) and by DNA sequencing (data not shown). Cloning of a GAPDH cDNA fragment permitted the generation of a standard curve by serial dilution (Figure 2). The standard curve with a correlation coefficient of 0.99 was used for a relative quantification of the copy numbers obtained from XIAP, Smac/DIABLO and GAPDH in each tumour sample. GAPDH copy numbers were used as an external standard to normalise the copy numbers of XIAP and Smac/ DIABLO, calculating the relative mRNA levels (i.e. GAPDHnormalised mRNA expression level).
Tumour stage-and grade-dependent increase of XIAP expression mRNA expression of XIAP was found in all RCCs (n ¼ 59), irrespective of their tumour stages, grades or histological types. As can be seen in Figure 3A and B, the relative mRNA expression levels of XIAP significantly increased from early (pT1) to advanced tumour stages (pT3). Importantly, RCCs infiltrating beyond the kidney (pT3) exhibited significantly higher relative XIAP mRNA expression levels when compared with tumours confined to the organ (pT1 and pT2; P ¼ 0.002; Figure 3A and B). Quite similarly, XIAP mRNA expression significantly increased with tumour dedifferentiation ( Figure 3C). Thus, the relative expression level of XIAP was significantly higher in poorly differentiated RCCs (G3) than in well and moderately differentiated tumours (G1 and G2; P ¼ 0.04).
Moreover, RCCs of the clear cell type, which were reported to exhibit a poorer prognosis than papillary RCCs (Cheville et al, 2003) showed higher XIAP expression levels when compared with XIAP levels in papillary RCCs ( Figure 3D), although the difference was of borderline significance only (P ¼ 0.05).
Since XIAP expression is known to be regulated transcriptionally and post-transcriptionally (Tamm et al, 2000;Holcik et al, 2001;Hofmann et al, 2002), we also performed Western blot analysis in 12 arbitrarily selected RCC tumour samples of early and advanced tumour stages (pT1: n ¼ 6, pT3: n ¼ 6). These investigations revealed a distinct XIAP signal of the expected size (53 kDa) in all RCCs (Figure 4). Remarkably, a stage-dependent increase of XIAP expression became evident also on the protein level (Figure 4), thereby further confirming our results on increased XIAP mRNA expression in advanced tumour stages.
Our results, therefore, demonstrate an increase in antiapoptotic XIAP mRNA and protein expression during tumour progression.

Constant expression of Smac/DIABLO irrespective of tumour stage and grade
Since Smac/DIABLO is known to be a major antagonist of XIAP (van Loo et al, 2002;Verhagen and Vaux 2002), we also analysed the stage-and grade-dependent expression of Smac/DIABLO in RCCs. mRNA expression of Smac/DIABLO was also found in all RCCs (n ¼ 66), irrespective of their staging, grading or histological typing.
Importantly, as shown in Figure 5A -D, the relative mRNA expression levels of proapoptotic Smac/DIABLO did not significantly change between early and advanced tumour stages ( Figure  5A and B) or between low and high tumour grades ( Figure 5C). Similar Smac/DIABLO mRNA levels were observed in RCCs of the clear cell and papillary types ( Figure 5D).
To analyse whether Smac/DIABLO expression remains constant also at the protein level, we investigated the corresponding protein expression in 24 arbitrarily selected RCCs of different tumour stages by ELISA. As can be seen in Figure 6, Smac/DIABLO protein levels showed no significant differences between different tumour stages.
Our results, therefore, demonstrate that expression of Smac/ DIABLO mRNA and protein remains constant during RCC progression.

Increased expression ratio of XIAP to Smac/DIABLO during tumour progression
Recent observations demonstrated an inverse expression pattern of XIAP and Smac/DIABLO after TRAIL-, UV-B-, and druginduced apoptosis (Chow et al, 2003;Griffin et al, 2003;Takasawa and Tanuma, 2003), suggesting that the relative ratio between XIAP and Smac/DIABLO is crucial to determine apoptosis susceptibility.
We, therefore, calculated the ratio between antiapoptotic XIAP and proapoptotic Smac/DIABLO mRNA expression levels in different tumour stages. As demonstrated in Figure 7, the mRNA expression ratio between XIAP and Smac/DIABLO markedly increased during progression from early (pT1) to advanced (pT3) tumour stages. Importantly, RCCs confined to the organ capsule (pT1 and pT2) exhibited a significantly lower expression ratio between XIAP and Smac/DIABLO when compared with advanced RCCs infiltrating beyond the kidney (pT3; P ¼ 0.01).
Thus, our investigation demonstrates that the delicate balance between XIAP and Smac/DIABLO expression is gradually disturbed during progression of RCCs, resulting in a relative increase of antiapoptotic XIAP over proapoptotic Smac/DIABLO.

DISCUSSION
The present investigation demonstrates for the first time a stageand grade-dependent increase of antiapoptotic XIAP expression in human RCCs. In contrast, the mRNA and protein levels of its antagonist Smac/DIABLO remained constant, resulting in an increased expression ratio between XIAP and Smac/DIABLO during tumour progression. Suppression of apoptosis promotes tumour progression, immune evasion of neoplastic cells as well as resistance to chemotherapy and irradation Mahotka et al, 1999;Reed, 1999;Ramp et al, 2000Ramp et al, , 2003aIgney and Krammer 2002;Mahotka et al, 2002a, b). Several genes critical in the regulation of apoptosis have been identified, including XIAP -a member of the IAP family. X-linked inhibitor of apoptosis is thought to act as a key determinant of apoptosis resistance by effectively inhibiting the activation of caspase-3, -7 and -9 (Deveraux et al, 1999;Holcik et al, 2001;Srinivasula et al, 2001;Suzuki et al, 2001). Thus, high expression of XIAP has been reported in many malignant tumour types, such as carcinomas of the breast, ovaries, lung, pancreas, cervix and prostate (Ferreira et al, 2001a, b;Liu et al, 2001;Gerhard et al, 2002;Hofmann et al, 2002;Mc Eleny et al, 2002;Parton et al, 2002;Sui et al, 2002) as well as leukaemias (Tamm et al, 2000). Moreover, upregulation of XIAP expression has been found to result in apoptosis resistance after exposure to anticancer drugs and ionising radiation (Li et al, 2000;Asselin et al, 2001;Holcik et al, 2001;Matsumiya et al, 2001), whereas downregulation of XIAP by antisense vectors increased the apoptosis sensitivity of carcinoma cell lines Holcik et al, 2001).
In our study, we found XIAP expression in all RCCs of all major histological types. This observation further confirmed the ubiquity of XIAP expression in human cancers, as reported earlier for other tumour types (Tamm et al, 2000). Importantly, we could demonstrate for the first time in RCCs that XIAP mRNA expression levels significantly increased from early (pT1) to advanced tumour stages (pT3) and -quite similarly -also with tumour dedifferentiation. Moreover, RCCs of the clear cell type, which were reported to exhibit a poorer prognosis than papillary RCCs (Cheville et al, 2003), showed a tendency for higher XIAP expression. Since XIAP expression is known to be also posttranscriptionally regulated, we additionally performed Western Similarly, a correlation between XIAP mRNA and protein levels was previously reported in non-small-cell lung carcinomas (Hofmann et al, 2002). Since tumour stage and grade are considered as major prognostic parameters in RCC (Srigley et al, 1997), our results   (Tamm et al, 2000). It is reasonable to assume, therefore, that in RCCs -and probably also in AML -increased expression of antiapoptotic XIAP contributes to reduced apoptosis susceptibility of tumour cells, thereby providing an important survival advantage during tumour progression. Conflicting observations, however, have been reported between XIAP expression and tumour stage or survival in other tumour types. In these reports, no correlation between XIAPexpression levels and tumour stage, grade or survival was observed in patients with carcinomas of the cervix, non-small-cell lung carcinomas and advanced nonresectable non-small-cell lung carcinomas treated by chemotherapy alone (Liu et al, 2001;Ferreira et al, 2001a;Hofmann et al, 2002), whereas high XIAPexpression levels were associated with longer survival in patients with radically resected non-small-cell lung carcinomas (Ferreira et al, 2001b). Susceptibility for apoptosis is controlled by a multitude of anti-and proapoptotic regulators. Whereas antiapoptotic XIAP has been shown to be a potent caspase inhibitor (Deveraux et al, 1999;Holcik et al, 2001;Srinivasula et al, 2001;Suzuki et al, 2001), its antagonists Smac/DIABLO and Omi/HtrA2 promote apoptosis by binding to XIAP, thereby preventing them from inhibition of caspases (Verhagen et al, 2000Martins et al, 2003). Importantly, overexpression of Smac/DIABLO sensitises tumour cells against anticancer drug-and TRAILinduced apoptosis (Zhang et al, 2001;Guo et al, 2002;Ng et al, 2002;Mc Neish et al, 2003). Only recently, an inverse relation between XIAP expression and mitochondrial release of Smac/ Diablo has been observed after apoptosis induction in colon cancer cells, lymphoma cells and keratinocytes (Chow et al, 2003;Takasawa and Tanuma, 2003;Tillman et al, 2003). Moreover, XIAP protein was shown to function as an E3 ligase in the ubiquitination of Smac/DIABLO, thereby promoting its rapid degradation (Mac Farlane et al, 2002). These observations indicate a tight regulation of these opponents and might also suggest that the expression ratio -rather than the expression of a single regulator -effectively controls susceptibility for apoptosis. Investigations, however, analysing the expression of XIAP and Smac/DIABLO in relation to tumour stage, grade and type are lacking up to now. In our study, the mRNA and protein expression levels of Smac/DIABLO remained stable during RCC progression and did not significantly change between early and advanced tumour stages, low and high tumour grades as well as different types of RCCs. Since XIAP expression significantly increased during tumour progression, the expression ratio between XIAP and Smac/DIABLO also markedly increased during progression from early to advanced RCC stages. Thus, the delicate balance between XIAP and Smac/DIABLO is gradually disturbed during progression of RCCs, probably resulting in an increased inability for Smac/DIABLO to antagonise the antiapoptotic effects of XIAP. Although investigations concerning the expression of Omi/HtrA2 in relation to tumour stage and grade are lacking, it is reasonable to assume that the disturbed balance between XIAP and Smac/DIABLO during progression of RCCs cannot be adjusted by Omi/HtrA2. Importantly in this context, TNF-and folic acid-induced apoptosis in tubular epithelial cells, considered to be the cells of origin for clear cell and papillary RCC, has recently been shown to be associated with increased mRNA and protein expression of Smac/DIABLO (Justo et al, 2003). Carcinogenesis in tubular epithelial cells, therefore, might be associated with an inability to upregulate Smac/DIABLO expression, thereby generating an important selective growth advantage. Moreover, the inappropriate increase of antiapoptotic XIAP over proapoptotic Smac/DIABLO during progression of RCCs promotes the acquisition of apoptosis resistance, which might also contribute to the clinically known resistance of RCCs to anticancer drugs and irradation.
In conclusion, our investigation demonstrates for the first time a stage-and grade-dependent increase of antiapoptotic XIAP expression in RCCs, whereas the mRNA and protein levels of its antagonist Smac/DIABLO remained constant. As a consequence, the delicate balance between XIAP and Smac/DIABLO expression is gradually disturbed during progression of RCC, probably contributing to the marked apoptosis resistance of RCCs.