Increased levels of hypoxia-inducible factor-1α are associated with Bcl-xL expression, tumor apoptosis, and clinical outcome in chondrosarcoma

Authors


Abstract

Hypoxia-inducible factor (HIF)-1α is a key nuclear transcription factor that regulates the cellular response to hypoxia, and is important for solid tumor growth and survival. However, the underlying role of HIF-1α in human chondrosarcoma has not been well characterized. This study aims to investigate the expression patterns of HIF-1α in chondrosarcoma, and its association with clinicopathologic features, Bcl-xL expression, apoptosis index (AI), and overall survival of patients with chondrosarcoma. Our results shown that the protein levels of HIF-1α were increased, and the mRNA and protein levels of Bcl-xL were also increased in SW1353 cells under hypoxic conditions. In eight patients with chondrosarcoma, increased expression of HIF-1α and Bcl-xL was detected in chondrosarcoma tissues compared with the paired adjacent normal tissues. Of 34 archival specimens of chondrosarcomas, 20 (58.8%) showed high HIF-1α protein expression as compared to benign cartilage tumors. Increased HIF-1α expression was correlated with a higher pathological grade and MSTS stage of chondrosarcoma. Moreover, HIF-1α expression was significantly associated with Bcl-xL expression and AI. More significantly, the survival rate of patients with HIF-1α high tumors was significantly lower than that of patients with HIF-1α low tumors. These findings suggest that increased HIF-1α levels mediated up-regulation of Bcl-xL play a prominent role in evasion of apoptosis and tumor progression, and can be predictive for the prognosis in human chondrosarcoma. © 2010 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 29:143–151, 2011

Hypoxia, as the most important microenvironment, is an essential feature of solid tumors, mainly attributable to malformed vasculature and inadequate perfusion. It profoundly influences the biological behavior, response to therapy, and clinical prognosis of human cancers.1, 2 However, tumor cells have evolved multiple strategies to respond and adapt to the hypoxic microenvironment.1, 3 Notably, hypoxia-inducible factors (HIFs) have emerged as the critical molecular mediators in response to hypoxia. To date, three HIFs (HIF-1, -2, and -3) have been identified that regulate transcriptional programs in response to hypoxic conditions.3 They are heterodimeric transcription factors consisting of an oxygen-regulated HIF-α subunit, and a constitutively expressed HIF-β subunit (also known as aryl hydrocarbon receptor nuclear translocator, ARNT).4 In the presence of oxygen, HIF-1α is degraded by the tumor suppressor Von Hippel-Lindau (VHL) through the ubiquitin-proteasomal pathway.5, 6 In the absence of oxygen, HIFs binds to hypoxia-response elements (HREs), thereby transactivating expression of numerous hypoxia-response genes that regulate various biological processes, including cell proliferation, angiogenesis, metabolism, and apoptosis.1, 7

Tumorigenesis and tumor progression are thought to result from, not only limitless replicative potential, but evasion of apoptosis.8 Hypoxia can induce apoptosis through the HIF-1α-mediated activation of two pro-apoptotic proteins, NIX and NIP3.9, 10 Nevertheless, tumor cells have developed a variety of strategies for evasion of apoptosis under hypoxic conditions including by blocking Bax translocation11 and induction of the anti-apoptotic genes IAP212 and Bcl-xL.13 Hypoxia-induced apoptosis-resistant phenotype eventually predominates, although hypoxia can activate pro-apoptotic molecules to induce apoptosis. It has been widely accepted that suppression of apoptosis can facilitate accumulation of aberrant cells and may be a critical event in the pathogenesis of malignancy.

The expression patterns of HIF-1α have been characterized for human tumors such as breast cancer,14 head and neck cancer,15 epithelial ovarian cancer,16 and cervical cancer.17 These observations accumulated over the past decade suggest that increased HIF-1α expression has been associated with carcinogenesis, pathological grade, and overall survival. A recent report has illuminated that Bcl-xL, a hypoxia-responsive anti-apoptotic protein of the Bcl-2 family, is directly regulated by HIF-1α in human prostate cancer cells, which may be an important mechanism by which HIF-1α protects human prostate cancer cells from apoptosis and leads to treatment resistance.13 Additional studies have documented that overexpression of the anti-apoptotic proteins Bcl-xL is associated with chemotherapy, radiation resistance and clinical outcome in patient with tumors.18–20 Therefore, these findings provide strong support for the hypothesis that HIF-1α-mediated up-regulation of Bcl-xL have prominent roles in evasion of apoptosis, treatment resistance, and overall survival.

The potential association of HIF-1α with chondrosarcoma provides the rationale for deeply exploring the underlying role of HIF-1α in the pathogenesis of chondrosarcoma in vivo, and determines whether HIF-1α can be predictive for overall survival of patients with chondrosarcoma. In this study, we initially investigated the mRNA and protein levels of HIF-1α and Bcl-xL in chondrosarcoma cells under hypoxic conditions, and chondrosarcoma tissues, and further analyzed the association of HIF-1α with clinicopathologic parameters, AI, Bcl-xL expression, and clinical outcome of patients with chondrosarcoma. Our findings suggest that increased levels of HIF-1α-mediated up-regulation of Bcl-xL play a prominent role in decreased AI and the worse outcome of patients with chondrosarcoma.

MATERIALS AND METHODS

Cell Culture

SW1353 cells were obtained from the American Type Culture Collection (Bethesda, MD) (ATCC no. HTB-94) and grown in Dulbecco's modified Eagle's medium (DMEM, Hyclone, Logan, UT) supplemented with 10% fetal bovine serum (Hyclone, Logan).21 SW1353 cells (2–3 × 106) in 100 mm disk were maintained in a humidified cell incubator with 5% CO2 at 37°C. Cells were passed when they were 80–90% confluent.

Hypoxia Mimetic Treatment of SW1353 Cells

3–5 × 105 SW1353 cells were cultured in six-well plates overnight, and then treated with 200 µM of CoCl2 (Sigma, St. Louis, MO) for 6 h. Cells were collected and processed for RT-PCR and Western blot analysis of HIF-1α and Bcl-xL expression as described below.

Patients and Tumor Specimens

Fresh chondrosarcoma tissues and adjacent normal tissues were collected from eight patients treated in the Department of Orthopaedics of Peking University Third Hospital. Paraffin-embedded tissue specimens from patients with pathologically established chondrosarcomas were also obtained from the Peking University Third Hospital. A total of 34 conventional chondrosarcoma specimens, 17 osteochondromas and enchondromas were selected based on accepted clinicopathological and radiological criteria.22 Dedifferentiated, mesenchymal, juxtacortical, and clear-cell chondrosarcomas were excluded based on their distinctly different clinicopathological features. Histological grading was performed according to previously published criteria.23 The time of follow-up was calculated from the date of the operation. The median follow-up was for 34 months (4–98 months). This study was in full compliance with national legislation and the ethical standards of the Chinese Medical Association (IRB00001052-08044).

RT-PCR and Real-Time PCR Analysis

Total RNA was extracted from chondrosarcoma specimens, the paired adjacent normal tissues, and treated SW1353 cells using TRIzol reagent (Invitrogen, Carlsbad, CA, Life Technologies, Inc.) according to the manufacturer's instructions. First-strand cDNA was generated using the total RNA in a reverse transcriptase reaction using a poly (dT) oligonucleotide as a primer and SuperScript II reverse transcriptase (Invitrogen, Life Technologies, Inc.). cDNAs were then subjected to RT-PCR and real-time PCR analysis. Two oligomers of primers were synthesized based on the reported sequences of HIF-1α (5′-CCT GCA CTC AAT CAA GAA GTT GC-3′ and 5′-TTC CTG CTC TGT TTG GTG AGG CT-3′), BCL-xL (5′-GCA GGC GAC GAG TTT GAA CT-3′, 5′-CTC GGC TGC TGC ATT GTT C-3′), and GAPDH (5′-ATC ATC CCT GCC TCT ACT GG-3′ and 5′-CCC TCC GAC GCC TGC TTC AC-3′).13, 24 RT-PCR was performed using the obtained RNA and the oligomers as templates and primers, respectively. The cDNA was amplified by 27–30 PCR cycles and the thermal cycle profile was as follows: denaturation for 1 min at 94°C, annealing for 1 min at 60°C, and extension for 1 min at 72°C. The gene of GAPDH was adopted as an internal control in these RT-PCR reactions.

Real-time PCR reactions were carried out using iCycler iQTM Real-Time PCR Detection System (Bio-Rad, Hercules, CA). The conditions of gene amplification were similar to RT-PCR described above. Data were analyzed according to the relative standard curve method with normalizing the values of GAPDH expression in each sample.25 Melting curves for each PCR reaction were generated to ensure the purity of the amplified products.

Western Blot

Detailed experimental procedures for Western blot have been described previously.25 Briefly, cells were washed in ice-cold PBS then resuspended in lysis buffer containing 50 mM Tris–HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 1% sodium deoxycholate, and freshly added protease inhibitor cocktail for 30 min on ice. Lysates were centrifuged at 16,000g for 20 min and the protein concentration of the supernatant was quantified using a protein assay kit.25 An equal amount (50 µg) of each cell lysate was separated by 12.5% SDS–PAGE and transferred to a nitrocellulose membrane. Membranes were blocked in TBS containing 5% nonfat dried milk and 0.05% Tween-20 for 1 h, then incubated with a monoclonal antibody against HIF-1α or Bcl-xL overnight. After washing, bound primary antibody was detected using an IRDye 800CW-conjugated secondary antibody and visualized using infrared fluorescence. Images were obtained using an Odyssey infrared imaging system (Li-Cor Bioscience, Lincoln, NE).

Immunohistochemistry and Scoring

All specimen slides (5 µm) were deparaffinized and rehydrated. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 9 min. For antigen retrieval, slides were incubated with 50 mM Tris–EDTA (pH 9.0) in a pressure cooker for 2 min. Primary antibodies and the corresponding dilution were as follows: HIF-1α (Abcam, Cambridge, UK; 1:50), Bcl-xL (Cell Signaling, Beverly, MA; 1:75). All primary antibodies were incubated with slides at 4°C overnight. After washing thrice in PBS, the sections were then incubated with EnVision+, peroxidase, mouse or rabbit (DAKO Diagnostics, Giostrup, Denmark) for 30 min each at 37°C. The slides were developed with DAB substrate (DAKO Diagnostics) and counterstained with hematoxylin. Then, the slides were dehydrated following standard procedure and sealed with coverslips. Negative controls were included by omitting primary antibody during the primary antibody incubation.

The stained slides were simultaneously scored blinded by two, independent observers without clinicopathologic information, and a consensus score was reached for each slide. For the assessment of cytoplasmic staining, the positive reaction of Bcl-xL was scored into four grades according to the intensity of the staining: 0, 1+, 2+, and 3+. The percentages of Bcl-xL-positive cells were also scored into four categories: 0 (0–5%), 1 (5–20%), 2 (21–50%), and 3 (51–100%) according to a previous report.20 In the cases with a discrepancy between duplicated scores, the average score from the two tissue cores was taken as the final score. The sum of the intensity and percentage scores is used as the final staining score. The staining pattern of the slides was defined as follows: 0, negative; 1–2, weak; 3–4, moderate; 5–6, strong. In addition, negative to weak expression is considered as negative expression, whereas moderate to strong expression is recognized as positive expression.

In particular, scoring results for HIF-1α staining were available as previously described with minor modifications.26, 27 Briefly, only cells with completely and darkly stained nuclei were regarded as positive, and this nuclear staining was interpreted as an increased level. Cytoplasmic staining, observed occasionally, was ignored because active HIFs are located only in the nucleus. The extent of staining was scored as follows: 0 (negative), no reactivity; 1+ (weak), nuclear reactivity in up to 10% of cells; 2+ (moderate), nuclear reactivity 10–50% of cells; 3+ (strong), nuclear reactivity in >50% of cells. For all analysis, more than 10% positive nuclei of cells involved per tissue sample was identified as positive expression (high expression). When independent scoring of a case differed, the case was rechecked, and the final score was determined by recounting HIF-1α-positive cells using a multiheaded microscope with all of the two observers simultaneously viewing the slide.

TUNEL Assay

An In Situ Cell Death Detection kit, POD (Roche, Penzberg, Germany), was used to perform TUNEL assays according to the manufacturer's instructions. Briefly, after staining, 10 tumor specimens were randomly selected from each group. The number of TUNEL-positive tumor cells among 1,000 tumor cells from 10 randomly selected fields in each section was counted, and the AI was determined as the average of the percentage of positive nuclei.

Statistical Analysis

Data are presented as the mean ± SEM (the standard error of the mean). Differences between groups were analyzed using the Student's t-test for continuous variables, or the Pearson's Chi-square test/Fisher's exact test and Spearman's rank correlation analysis for categorical variables. Curves for overall survival were drawn according to the Kaplan–Meier method, and differences between the curves were analyzed by applying the log-rank test. Statistical analyses were performed using SPSS (version 17.0; SPSS, Inc., Chicago, IL) and p < 0.05 was considered statistically significant. All reported p values are two-sided.

RESULTS

Hypoxic Induction of HIF-1α and Bcl-xL Expression in SW1353 Cells

To determine HIF-1α and Bcl-xL expression in SW1353 cells under hypoxic conditions, we performed RT-PCR and Western blot to analyze the mRNA and protein levels of HIF-1α and Bcl-xL in response to hypoxia mimetic treatment (200 µM CoCl2). As shown in Figure 1A,B, the mRNA levels of HIF-1α did not change in response to hypoxic treatment whereas increased amounts of HIF-1α protein were significantly detected, suggesting that hypoxia influences HIF-1α at the protein level in SW1353 cells. We also found that the mRNA and protein levels of Bcl-xL were remarkably increased under hypoxic conditions.

Figure 1.

HIF-1α and Bcl-xL expression are up-regulated in SW1353 cells under hypoxic conditions. (A) SW1353 cells were treated with the hypoxia mimetic CoCl2 (200 µM) for 6 h. RT-PCR analyzed the mRNA levels of HIF-1α and Bcl-xL. GAPDH was used as an internal control. (B) SW1353 cells were treated as (A), Western blot analyzed the protein levels of HIF-1α and Bcl-xL, β-actin was used as the loading control.

Levels of HIF-1α and Bcl-xL Were Increased in Chondrosarcoma Tissues

The expression profiles of HIF-1α and Bcl-xL have been reported for many cancers, although the expression patterns of HIF-1α and Bcl-xL in chondrosarcoma have not been well characterized. Therefore, the mRNA and protein levels of HIF-1α and Bcl-xL in eight patient samples of chondrosarcoma, and the corresponding adjacent normal tissues, were analyzed by real-time PCR and Western blot, respectively. Our results shown that the mRNA levels of HIF-1α and Bcl-xL were significantly increased in the chondrosarcoma specimens compared to the paired adjacent normal tissues (**p < 0.01, Fig. 2A). Similarly, increased amounts of HIF-1α and Bcl-xL protein were detected in the chondrosarcoma samples versus the adjacent normal tissues (Fig. 2B,C). The diagram (Fig. 2C) presented was the statistical analysis of HIF-1α and Bcl-xL protein (HIF-1α and Bcl-xL/β-actin, respectively) of eight cases of chondrosarcomas (*p < 0.05, **p < 0.01).

Figure 2.

HIF-1α and Bcl-xL expression are increased in human chondrosarcoma tissues. (A) Real-time PCR analyzed the HIF-1α and Bcl-xL mRNA levels in chondrosarcoma tissues and the paired adjacent normal tissues. N, adjacent normal tissues; CS, chondrosarcoma tissues. **p < 0.01. (B) Western blot detected HIF-1α and Bcl-xL protein levels in chondrosarcoma tissues and the paired adjacent normal tissues. Four representative pairs of tissues were presented. (C) The diagram presented was the statistical analysis of HIF-1α and Bcl-xL protein (HIF-1α or Bcl-xL/β-actin) of eight cases of chondrosarcomas. *p < 0.05, **p < 0.01.

Based on the surgical staging system of the Musculoskeletal Tumor Society (MSTS),28 lesions were correspondingly identified as: stage IA (12), stage IB (10), stage IIA (6), and stage IIB (6). The clinical parameters of patients with chondrosarcoma included in this study are presented in Table 1. To investigate the expression patterns of HIF-1α and Bcl-xL in benign cartilage tumors and chondrosarcomas, we next performed immunohistochemistry on 34 chondrosarcoma samples, 17 benign cartilage tumors consisting of 9 osteochondromas, and 8 enchondromas. We observed that the positive signal of Bcl-xL was identified at the cytoplasm (Fig. 3A,B), whereas the immunoreactive signal of HIF-1α predominately distributed at the nucleus (Fig. 3C–F). Statistical analysis data (Table 2) display that high HIF-1α expression was 58.5% (20/34) in chondrosarcoma samples as compared to 11.8% (2/17) in benign cartilage tumors (p < 0.01). By contrast, high Bcl-xL expression was 58.5% (20/34) in chondrosarcoma samples compared with 17.6% (3/17) in benign cartilage tumors (p < 0.01). These results suggest that increased HIF-1α and Bcl-xL expression may contribute to the development of chondrosarcoma.

Table 1. Distribution of Chondrosarcomas Based Upon Patient Demographics, Anatomic Location, Histological Grade, and MSTS Stage (Enneking Classification)
Demographic DataChondrosarcoma
  1. MSTS, Musculoskeletal Tumor Society.

Male:female16:18
Age (years) (median; range)47.9; 18–72
Anatomic location
 Cervical vertebrae7
 Thoracic vertebrae4
 Lumbar vertebrae5
 Sacrum4
 Rib1
 Humerus4
 Proximal fibula2
 Distal femur3
 Proximal tibia2
 Phalanx2
Histological grade
 I22
 II6
 III6
MSTS stage
 I A12
 I B10
 II A6
 II B6
Figure 3.

Representative images of HIF-1α and Bcl-xL immunohistochemical staining in benign cartilage tumors and chondrosarcomas. (A) Enchondroma with negative cytoplasmic Bcl-xL expression. (B) Chondrosarcoma with strong cytoplasmic Bcl-xL expression. (C) Positive nuclear HIF-1α immunostaining in osteochondroma. (D) Negative nuclear HIF-1α immunostaining for a well-differentiated (grade I) chondrosarcoma. (E) Malignant tumor cells of a moderately differentiated (grade II) chondrosarcoma exhibited moderate nuclear immunopositivity for HIF-1α. (F) Diffuse and strong nuclear expression of HIF-1α was detected in a poorly differentiated (grade III) chondrosarcoma. Original magnification ×400. [Color figure can be viewed in the online issue, which is available at http://www.interscience.wiley.com]

Table 2. Expression of Hypoxia-Inducible Factor-1α (HIF-1α) and Bcl2-Like 1 (Bcl-xL) in Benign Cartilage Tumors and Chondrosarcomas
GroupsTotal No. of CasesHIF-1α ExpressionBcl-xL Expression
LowHighp-ValueNegativePositivep-Value
Benign bone tumors17152 143 
Chondrosarcomas3414200.00214200.007

Association Among HIF-1α Expression and Clinicopathologic Parameters

We further analyze the association of HIF-1α expression as individual variables with standard chondrosarcoma clinicopathologic parameters. Associations of HIF-1α expression with gender, age, anatomical location, histological grade, and MSTS stage are presented in Table 3. HIF-1α expression was significantly associated with tumor grade (p = 0.009) and MSTS stage (p = 0.009) of chondrosarcoma. Increased levels of HIF-1α were found in tumor grade I of chondrosarcoma specimens (9 specimens had ≥10% immunopositive cells of 22 specimens tested), with the number of positive specimens further increasing in tumor grades II and III (11 of 12). Similarly, increased levels of HIF-1α were detected in tumor MSTS grade IA + IB of chondrosarcoma specimens (9 of 22), with the number of positive specimens further increasing in tumor MSTS grade IIA + IIB (11 of 12). However, there was no significant correlation between the level of HIF-1α expression and other clinicopathological parameters such as gender, age, and anatomical location. Our findings demonstrated that increased levels of HIF-1α were associated with a higher pathologic grade and MSTS stage of chondrosarcoma, indicating that increased levels of HIF-1α have a prominent role in the more aggressive potential of chondrosarcoma.

Table 3. Expression of HIF-1α in Chondrosarcomas and Its Correlation With Clinicopathologic Parameters
Clinicopathologic ParametersTotal No. of CasesHIF-1α Expressionp-Value
LowHigh
No.%No.%
Gender     0.738
 Male16637.51062.5 
 Female18844.41055.6 
Age (years)     1.000
 ≥40241041.71458.3 
 <4010440660 
Anatomical location     0.728
 Limb bone13646.2753.8 
 Axial bone21838.11361.9 
Histological grade     0.009
 Well-differentiated (I)221359.1940.9 
 Moderately or poorly (II, III)1218.31191.7 
MSTS stage     0.009
 I A + I B221359.1940.9 
 II A + II B1218.31191.7 

Correlation Among Increased HIF-1α Levels With Bcl-xL Expression and Apoptosis Index

Of the 34 chondrosarcomas involved in this study, 20 (58.8%) showed high HIF-1α expression (Table 2). In the 20 chondrosarcomas regarded as high HIF-1α tumors, the Bcl-xL-positive and Bcl-xL-negative tumor frequencies were 16/20 (80%) and 4/20 (20%), respectively (Table 4). In contrast, in the remaining 14 chondrosarcomas classified as low HIF-1α tumors, 4/14 (28.6%) were Bcl-xL-positive, and 10/14 (71.4%) were Bcl-xL-negative. There was a strong positive correlation between the increased expression of HIF-1α and Bcl-xL (Table 4, p < 0.01).

Table 4. Correlation Between Bcl-xL Expression, Apoptosis Index, and HIF-1α Expression in Chondrosarcomas
HIF-1α ExpressionNo. of CasesBcl-xL ExpressionApoptosis Index (Mean ± SEM)#p-Value
PositiveNegative*p-Value
  • SEM, standard error of the mean.

  • *

    Fisher's Exact test, p < 0.01.

  • #

    Spearman's rank correlation analysis, p < 0.01.

High20164 0.81 ± 0.32 
Low144100.0051.89 ± 0.890.002

To investigate whether increased HIF-1α and Bcl-xL expression were associated with the AI in chondrosarcoma, we performed the TUNEL assay. The mean of AI in the HIF-1α high and HIF-1α low groups was 0.81 and 1.89, respectively (Table 4 and Fig. 4A). This difference was statistically significant (p < 0.01), indicating that there was lower apoptosis in the high HIF-1α expressing tumors. By contrast, the mean of AI in the Bcl-xL-positive and Bcl-xL-negative groups was 0.7 and 1.74, respectively. AI in the Bcl-xL-positive group was also significantly lower than that in the Bcl-xL-negative group (Fig. 4B).

Figure 4.

Relationship between expression of HIF-1α and Bcl-xL, and apoptosis index in human chondrosarcoma tissues. The diagram presented was the statistical analysis of TUNEL assay in HIF-1α expression (A) and Bcl-xL (B). **p < 0.01, p values were determined by Spearman's rank correlation analysis.

Relationship of HIF-1α and Bcl-xL as Individual Determinants With Clinical Outcome

To determine whether the protein expression of HIF-1α and Bcl-xL was associated with clinical outcome of patients with chondrosarcoma, a Kaplan–Meier analysis was performed. As judged by the Kaplan–Meier analysis (Fig. 5A–C), the survival rate of patients with HIF-1α high tumors was significantly lower than that of patients with HIF-1α low tumors (p < 0.05) (Fig. 5A). Similarly, the survival rate of patients with Bcl-xL high tumors was significantly lower than that of patients with Bcl-xL low tumors (p < 0.05) (Fig. 5B). On the basis of the combination with HIF-1α and Bcl-xL expression, we further classified the chondrosarcoma patients into two groups: group I (HIF-1α high and Bcl-xL-positive tumors, 16 patients), group II (HIF-1α low and Bcl-xL-negative tumors, 10 patients). Our results show that the survival rates of HIF-1α high and Bcl-xL-positive tumors were remarkably lower than that those of patients in HIF-1α low and Bcl-xL-negative tumors (p < 0.01) (Fig. 5C).

Figure 5.

Kaplan–Meier survival curves of patients with chondrosarcoma with regard to the protein expression of HIF-1α and Bcl-xL. (A) The survival rate of patients with HIF-1α high tumors was significantly worse than that of patients with HIF-1α low tumors (p < 0.05). (B) The survival rate of patients with Bcl-xL-positive tumors was significantly worse than that of patients with Bcl-xL-negative tumors (p < 0.05). (C) Kaplan–Meier survival curves with regard to the combination of HIF-1α and Bcl-xL expression. Two groups were classified according to the expression status of HIF-1α and Bcl-xL: HIF-1α high and Bcl-xL-positive tumors (n = 16), HIF-1α low and Bcl-xL-negative tumors (n = 10). The survival rates of HIF-1α high and Bcl-xL-positive tumors were significantly worse than that those of patients in HIF-1α low and Bcl-xL-negative tumors (p < 0.01). p values were determined by the log-rank test.

DISCUSSION

Chondrosarcoma represents a heterogeneous group of bone malignant neoplasms, characterized by the synthesis of pure hyaline cartilage. It occurs principally in the third to sixth decade of adults, which remains the third most common primary malignant bone tumors and accounts for ∼20% of bone malignancies.22 Conventional chondrosarcoma can be further categorized into primary central tumors that locate within the medullar cavity and constitute ∼85% of conventional chondorsarcomas, or secondary peripheral tumors that arise from benign, preexisting lesions such as osteochondroma or enchondroma.29 Hypoxia is an important characteristic of solid tumors such as chondrosarcoma. Hypoxic signaling remains intact and appears to be implicated in angiogenesis during chondrosarcoma development.30, 31 Moreover, increased HIF-1α expression significantly correlated with tumor angiogenesis and cell proliferation in chondrosarcoma.27 In addition, HIF-1α is a significant positive regulator of tumor progression and metastatic potential.32 These findings imply that hypoxia and HIFs could have a critical role in the development and progression of chondrosarcoma.

In the in vitro segment of this study, we detected up-regulated expression of HIF-1α at the protein level rather than the mRNA level in chondrosarcom cells under hypoxic conditions, which is consistent with the previous report,33 indicating that hypoxia could regulate posttranscriptional level of HIF-1α. To our knowledge, we present the first evidence that the mRNA and protein levels of HIF-1α are higher in chondrosarcoma tissues compared with the corresponding adjacent normal tissues. Immunohistochemical analysis further confirmed that chondrosarcoma cells exhibit higher levels of HIF-1α expression, mainly in the nuclei, whereas benign bone tumors correlate with lower levels of HIF-1α expression (Fig. 2). It has been widely accepted that chondrosarcomas can develop de novo (primary) or from a preexisting benign cartilage tumor, such as enchondroma or osteochondroma (secondary).34 Given the ability of HIF-1α to facilitate tumorigenesis in tumor tissues, increased levels of HIF-1α in chondrosarcoma would be consistent with a role for HIF-1α in the development of chondrosarcomas in vivo.

Emerging evidence indicates that tumorigenesis is characterized by the severe imbalance between cell proliferation and programmed cell death (apoptosis). Our findings have demonstrated that increased levels of HIF-1α were significantly correlated with decreased AI in chondrosarcoma in vivo (Fig. 4A). Evading apoptosis is an important feature of all malignant tumors, including chondrosarcoma.35 Thus, increased levels of HIF-1α associated with decreased AI are hypothesized to be an important mechanism for evasion of apoptosis by chondrosarcoma cells, and implicated in the pathogenesis of chondrosarcoma. A recent publication based on DNA microarrays provides evidence that the transcriptional response to hypoxia varies among human cells, and some of this variation is traceable to variation in expression of the HIF-1α gene.36 Similarly, several previous reports by means of DNA-microarray analysis37, 38 have also documented that hypoxia and hypoxia-mediated up-regulation of HIF-1α gene and other apoptosis-regulated genes, such as Hsp70, Bag-1, galectin-3, and gelsolin, which are involved in stress resistance and anti-apoptotic signaling, have been suggested to be implicated in increased apoptosis resistance. Therefore, the combination of the findings in this study, with the results of previous studies, further clarify a pivotal role for HIF-1α in defects in apoptosis, and the pathogenesis of chondrosarcoma.

Another important point elucidated in this study is that Bcl-xL, a hypoxia-responsive and anti-apoptotic molecule of the Bcl-2 family, was increased in both the mRNA and protien levels under hypoxic conditions in chondrosarcoma cells (Fig. 1B), and chondrosarcoma tissues (Fig. 2). Previous studies have documented that elevated expression of Bcl-xL is detected in solid tumors such as oropharyngeal cancer, and hypoxia-mediated selection of Bcl-xL may provide a mechanism for tumor expansion and resistance to cancer therapy.39 Our results are consistent with this report. Amazingly, there is the significant association between increased expression of HIF-1α and Bcl-xL in chondrosarcoma, but supported by the following evidence: hypoxia and hypoxia-mediated increased expression of HIF-1α can induce up-regulation of Bcl-xL in SW1353 cells (Fig. 1). Moreover, Bcl-xL is transcriptionally regulated by HIF-1α and HIF-1α-dependent up-regulated expression of Bcl-xL may be an important mechanism for evasion of apoptosis by tumor cells.13, 18 Additional studies have demonstrated that hypoxia can increase the expression of anti-apoptotic proteins such as Bcl-xL, Bcl-2, and IAP family members, which is believed to play a protective role in TRAIL-induced apoptosis in response to hypoxia.40 It is noteworthy that Bcl-2, another anti-apoptotic protein of the Bcl-2 family, is also implicated in the regulation of the programmed cell death of hypertrophic chondrocytes. Further studies have demonstrated that Bcl-2 lies downstream of PTHrP in a pathway that controls chondrocyte maturation and skeletal development.41 A recent report has illuminated that up-regulation of PTHrP and Bcl-2 characterizes progression of osteochondroma towards grade I secondary peripheral chondrosarcoma.42 And therefore Bcl-2 was recently confirmed as a valuable immunohistochemical diagnostic marker for the distinction between osteochondroma and low-grade secondary peripheral chondrosarcoma.43 These findings suggest that the Bcl-2 family such as Bcl-xL and Bcl-2 may be involved in the pathogenesis of chondrosarcoma. More importantly, we also found that increased expression of Bcl-xL was significantly associated with decreased AI and the worse outcome of chondrosarcoma, which was consistent with the previous study.20 Taken together, our findings suggest that increased levels of HIF-1α-mediated up-regulated Bcl-xL could be implicated in defects in apoptosis of human chondrosarcoma in vivo.

The clinical value of elevated HIF-1α levels in chondrosarcoma has not been well characterized. We, thus, initially investigated the potential association between HIF-1α expression and clinicopathological characteristics of 34 patients with chondrosarcoma. We found that increased expression of HIF-1α strongly correlated with tumor grade and MSTS stage of chondrosarcoma (Table 3). There is now a general consensus that tumor hypoxia is an adverse prognostic factor. Therefore, we further observed patients with chondrosarcoma via follow-up examinations to determine whether elevated HIF-1α levels can be predictive for the prognosis of human chondrosarcoma. Our findings show that HIF-1α and Bcl-xL expression were significantly associated with the clinical prognoses of patients with chondrosarcoma (Fig. 5A,B). This finding is consistent with the previous report.27 More significantly, based on the combination of HIF-1α and Bcl-xL, the survival rates of HIF-1α high and Bcl-xL-positive tumors were remarkably lower than that those of patients in HIF-1α low and Bcl-xL-negative tumors (Fig. 5C). Our findings are consistent with the previous reports that HIF-1α is associated with the prognosis of patients with several malignancies.14–17 Taken together, these findings clearly hightlight a role for HIF-1α-mediated up-regulation of Bcl-xL in defects in apoptosis, and could help the clinician to determine the prognosis of human chondrosarcoma associated with increased HIF-1α levels.

In conclusion, we present evidence for the first time that increased levels of HIF-1α and Bcl-xL are remarkably detected in chondrosarcoma cells in response to hypoxic conditions. Elevated HIF-1α levels are also observed in chondrosarcoma tissues as compared to the adjacent normal tissues and benign cartilage tumors, which are significantly associated with a higher histological grade and MSTS stage of human chondrosarcoma. More significantly, increased levels of HIF-1α are associated with elevated Bcl-xL expression, decreased AI and the worse overall survival of human chondrosarcoma. Our findings suggest that increased HIF-1α levels mediated up-regulation of Bcl-xL have a prominent role in evasion of apoptosis and the pathogenesis of chondrosarcoma, which can be predictive for the overall survival in human chondrosarcoma. Further investigation of the possible application of targeting HIF-1α in chondrosarcoma management is also of great interest.

Acknowledgements

This work was supported by grants from the National Natural Sciences Foundation of China (30571870).

Ancillary