Abstract
The biological properties of CCN proteins include stimulation of cell proliferation, migration, and adhesion, as well as angiogenesis and tumorigenesis. We quantified CYR61, CTGF, WISP-1, and NOV mRNA expression levels in samples from sixty-six primary gliomas and five normal brain samples using quantitative real-time PCR assay. Statistical analysis was performed to explore the links between expression of the CCN genes and clinical and pathological parameters. Overexpression of CYR61, CTGF, WISP-1, and NOV occurred in 48% (32 of 66), 58% (38 of 66), 36% (24 of 66), and 15% (10 of 66) of primary gliomas, respectively. Interestingly, significant associations were found between CYR61 expression versus tumor grade, pathology, gender, and age at diagnosis. Also, a significant correlation existed between CTGF mRNA levels versus tumor grade, gender, and pathology. In contrast to CYR61 and CTGF, no significant association was found between expression of either WISP-1 or NOV versus any of the pathological features. Furthermore, Cox regression analysis showed that CYR61 and CTGF expression had a significant correlation with patient survival. These results suggest that CYR61 and CTGF may play a role in the progression of gliomas; their levels at diagnosis may have prognostic significance; and these proteins might serve as valuable targets for therapeutic intervention.
INTRODUCTION
Malignant gliomas are the most common primary brain tumors of the adult central nervous system (1) . They are divided into four stages by WHO classification scheme (2) . Grade I tumors are biologically benign and can be surgically cured; grade II tumors are low-grade malignancies that may have a long clinical course but are not curable by surgery; grade III tumors are malignant and lead to death within a few years; grade IV tumors are highly malignant and lethal within 9–12 months (3) . The malignant gliomas are thought to develop as the result of stepwise accumulation of multiple genetic alterations (4) . Anaplastic astrocytoma (grade III) typically has a p53 mutation, p16/ARF deletion, and ras pathway activation (3) . Glioblastoma multiforme (GBM), in addition to these alterations, also quite often has amplification and expression of the rearranged epidermal growth factor receptor (EGFR) and often has an alteration of PTEN leading to activation of the Akt pathway (5 , 6) . Therefore, the unique characteristics of each grade of astrocytoma may be associated with a set of genetic alterations. Recent microarray studies revealed that hundreds of gene transcripts may be expressed at significantly altered levels in high-grade gliomas compared with either low-grade gliomas or normal brain tissue (7, 8, 9) .
The recent discovery that CYR61 is highly expressed in breast cancers and associated with more advanced disease has brought to light an emerging family of conserved, modular proteins (10, 11, 12, 13) . This protein family now consists of six distinct members, including CTGF (connective tissue growth factor; Refs. 14 , 15 ), NOV (nephroblastoma overexpressed gene; Refs. 16 , 17 ), CYR61 (cysteine-rich protein; Ref. 18 ), WISP-1 (wnt-1 inducible gene; Ref. 16 ), and WISP-2 (also termed rCop-1) and WISP-3 (19) . CYR61, CTGF, and NOV were the first members of this group, resulting in all of these proteins being called the “CCN family” (20) . The primary translational products of CCN family members are 343–381 residues, which generate proteins of Mr 35,000–40,000 with homologies ranging from 60 to 90%. All of them possess a secretory signal peptide at the NH2 terminus, indicating that they are secreted proteins.
All of CCN genes contain four distinct structural modules: insulin-like growth factor-binding protein (IGFBP), Von Willebrand type C (VWC), thrombospondin type 1 (TSP1), and COOH-terminal domain (CT). These distinct modules exhibit homology to conserved regions in a variety of extracellular mosaic proteins (20) . Each module is involved in protein binding and contains conserved cysteine, hydrophobic, and polar residues. The biological properties of CCN proteins include stimulation of cell proliferation, migration, adhesion, and extracellular matrix (ECM) formation. They also regulate more complex biological processes such as angiogenesis and tumorigenesis (21, 22, 23) .
Recent studies have shown that CCN genes are involved in tumorigenesis. Consistent with its profibrotic properties, CTGF is overexpressed in pancreatic cancers (24) and melanomas (25) . WISP-1 is strongly expressed in the fibrovascular stroma of breast tumors developing in Wnt-1 transgenic mice (16) . Moreover, forced overexpression of WISP-1 in normal rat kidney fibroblasts (NRK-49F) was sufficient to induce their transformation (26) . Many of the human tumor cell lines express CYR61, suggesting that CYR61 expression may promote tumorigenesis (27) . Furthermore, CYR61-nonexpressing tumor cell lines tend to be less tumorigenic compared with those that do express the protein (10 , 13) .
In this study, we used real-time reverse transcription-PCR to quantify expression of CNN genes in gliomas. The mRNA levels of four genes (CYR61, CTGF, WISP-1, and NOV) were measured from 66 primary gliomas and five normal human brain tissue samples. Furthermore, we determined whether overexpression of one of the CCN genes was correlated with clinical and pathological parameters of the gliomas as well as survival data, using several models of statistical analysis.
MATERIALS AND METHODS
Patients and Samples
We analyzed tissue from excised primary gliomas of 66 patients treated at Cedars-Sinai Medical Center from 1996 to 2000 after their informed consent. The samples were examined histologically for the presence of tumor cells.
RNA Extraction and cDNA Synthesis
Total RNA was extracted from glioma specimens by using TRIzol reagent (Life Technologies, Inc.) according to the standard protocol. The quality of the RNA samples was determined by electrophoresis through agarose gels and staining with ethidium bromide, and the 18S and 28S RNA bands were visualized under UV light. Two μg of total RNA were processed directly to cDNA by reverse transcription with Superscript II (Life Technologies, Inc.) according to the manufacturer’s protocol in a total volume of 50 μl.
Real-Time Reverse Transcription-PCR
Theoretical Basis.
Reactions were characterized at the point during cycling when amplification of the PCR product was first detected, rather than the amount of PCR product accumulated after a fixed number of cycles. The parameter Ct was defined as the fractional cycle number at which the fluorescence generated by cleavage of the probe passes a fixed threshold above baseline. The CCN target message in unknown samples was quantified by measuring Ct and by using a standard curve to determine the quantity of starting target message. We also quantified transcripts of β-actin as the endogenous RNA control, and each sample was normalized on the basis of its β-actin content. For each experimental sample, the amount of the targets and endogenous reference is determined from the standard curve. The amount of target was divided by the endogenous reference amount to obtain a normalized target value. The relative target gene expression level was also normalized to a mean value (value = 1) from five normal brain tissue samples (calibrator). Final results, expressed as N-fold difference in CCN gene expression relative to the β-actin and the calibrator, termed ΔCCN, were determined as follows:
Another housekeeping gene, 18S, was used as a second endogenous reference gene to determine the consistency of normalization.
Primers and Probes.
Primers and probes for the CCN and β-actin genes were designed using software PRIMER35 , as described previously (11) . We conducted BLAST searches against dbEST and nr (the nonredundant set of GenBank, EMBL, and DDBJ database sequences) to confirm the total gene specificity of the nucleotide sequences chosen for the primers and probes and the absence of DNA polymorphisms. To avoid amplification of contaminating genomic DNA, one of the two primers or the probe was placed at the junction between two exons or in a different exon. Primers were purchased from Life Technologies, Inc., and probes were from Perkin-Elmer Applied Biosystems.
Standard Curve Construction.
The standard curve was constructed with 10-fold serially diluted total RNA extracted from the MDA-MB-231 cells. Fig. 1⇓ shows the real-time reverse transcription-PCR standard curve for the CTGF gene. A strong linear relationship between the Ct and the log of the starting copy number was always demonstrated. The efficiency of the reaction (E), calculated by the formula: E = 101/m − 1, where m is the slope of the standard curve, ranged from 90 to 100% in the different assays. The standard curves for WISP-1, CYR61, and NOV were also constructed (data not shown).
Standard curves of CTGF mRNA by real-time PCR. A, amplification plots for reactions with five points of the CTGF standard curve (10-fold serially diluted human breast cancer cell MDA-MB-231 cDNAs, (1.5 × 10−1, 1.5 × 10−2, 1.5 × 10−3, 1.5 × 10−4, and 1.5 × 10−5 μg/ml). Cycle number is plotted versus change in normalized reporter signal (ΔRn). ΔRn represents the normalized reporter signal (Rn) minus the baseline signal established in the first 18 PCR cycles. Ct represents the fractional cycle number at which a significant increase in Rn above a baseline signal (thick horizontal line) can first be detected. B, standard curve generated after determination of Ct values plotted against starting quantity of target DNA. The data from the Ct values were plotted to give a log-linear regression standard curve. ○, data for standard curve samples; □, data for glioma patients performed in triplicate.
PCR Amplification.
Amplification reactions contained 5 μl of cDNA, 12.5 μl of the Universal Taqman 2× PCR mastermix (Applied Biosystems), and 2.5 μl of each of the specific primers and the probe. Primer and TaqMan probe concentrations in the final volume of 25 μl, were 500 nm and 100 nm, respectively. All of the reactions were performed in triplicate in an iCycler iQ system (Bio-Rad, Hercules, CA), and the thermal cycling conditions were as follows: 2 min at 50°C, 10 min at 95°C, followed by 45 cycles of 95°C for 15 s and 60°C for 1 min.
Immunohistochemical Staining
Immunohistochemical staining for CYR-61 was performed with polyclonal antiserum from Santa Cruz Biotechnology. Heat-induced epitope retrieval was performed with a pressure cooker and TRIS buffer (pH 9.0) for 2 min. Localization was performed with DAKO ENVISION (Carpinteria, CA) conjugated to horseradish peroxidase followed by the diaminobenzidine reaction. Negative controls consisted of substitution of the primary antiserum with normal rabbit serum at the same dilution. The slides were counterstained with hematoxylin. Immunohistochemical staining for glial fibrillary acidic protein (GFAP) was performed as above using monoclonal antibody from DAKO.
Statistical Analysis
χ2 test and t test were used to study the association of each gene with single clinical factors (age, gender, pathology, grade). For each gene, Kaplan-Meier survival curves for patients with positive versus negative gene expression were plotted and log-rank test was used for comparing the equality of the two survival curves. Cox proportional hazard model was also developed to correlate the clinical characteristics, survival, and the expression of the four CCN genes. Stepwise procedure was used for covariate selection. Kappa (κ) statistical analysis was used to assess the relationship between all pairs of the four genes.
RESULTS
The relative levels of expression of CCN genes were quantified in 66 gliomas and five normal brain tissues by performing real-time PCR. The expression levels were determined as a ratio between CYR61, CTGF, WISP-1, or NOV and the reference gene β-actin to correct for variation in the amounts of RNA. The relative target gene expression was also normalized to a mean value (value = 1) for the five normal brain tissue samples (calibrator). Each of the normalized target values was divided by the calibrator normalized target value to generate the final relative expression levels. Because the expression values of the five normal brain tissues were between 0.3 and 2.5, we set values of 3 or more as the cutoff point for overexpression of CCN genes at the RNA level in gliomas.
To determine whether β-actin is suitable for the calibrator of normalization, a second housekeeping gene, 18S, was also used as a reference calibrator for CYR61, CTGF, WISP-1, and NOV in five normal brain and six glioma samples. The levels of expression of the CCN genes were comparable with those when β-actin was used as the reference gene (data not shown).
Expression of CYR61 in Primary Gliomas.
Overexpression of CYR61 was found in 32 (48%) of 66 glioma samples (Fig. 2A)⇓ ⇓ . Univariate analysis showed a significant association between tumor grade, pathology, and gender, as well as age at onset of disease, compared with the level of expression of CYR61 by the primary brain tumor samples (Table 1)⇓ . Overexpression of CYR61 occurred predominantly in the most malignant samples (GBM). Twenty-seven (68%) of 40 GBM patients overexpressed CYR61; however, overexpression of CYR61 was detected in only 4 (20%) of 19 astrocytomas and 1 (14%) of 7 oligodendrogliomas (P < 0.0001). Interestingly, analysis of tumor grade showed that only 1 (3%) of 29 patients with either grade II or grade III brain tumors had high levels of CYR61; in marked contrast, 31 (66%) of 47 of those with grade IV tumors overexpressed CYR61 (P < 0.0001). Furthermore, gender analysis showed that 70% (14 of 20) of female patients had tumors with high levels of CYR61 mRNA compared with only 39% (18 of 46) male patients whose gliomas overexpressed CYR61 (P = 0.021).
CCN gene expression in glioma tissues. Relative expression levels of CYR61 (A), CTGF (B), WISP-1 (C), and NOV (D) are shown in 5 normal brain tissues and 66 primary gliomas samples. Expression levels are displayed as a ratio between the target genes and a reference gene (β-actin) to correct for variation in the amounts of RNA.
(cont’d) The relative expression level has been normalized in such a manner that the mean ratio of the five normal brain samples equals a value of 1. □, normal; ▪, glioblastoma multiforme (GBM); , astrocytoma;
, oligodendroglioma.
Relationship between levels of expression of CYR61 in gliomas and the clinical and pathological features of the individuals
To determine whether overexpression of Cyr61 mRNA increased the protein level of Cyr61, proteins extracted from one normal brain tissue and several human primary glioma samples with different expression levels of Cyr61 mRNA were examined by Western blot assay. Levels of Cyr61 protein paralleled levels of Cyr61 mRNA as measured by real-time PCR in both normal brain and human gliomas (data not shown). Immunohistochemical staining for Cyr61 was evaluated in three GBM tumors and three normal brain tissues. The normal brain tissue was negative except for sparse cytoplasmic staining in a few glial cells and neurons. In contrast, there was strong staining for Cyr61 in the neoplastic astrocytoma cells. Glial fibrillary acidic protein was used for glial cell-specific staining. (Fig. 3)⇓ .
Immunohistochemical staining for Cyr61. Normal brain tissue shows negative for Cyr61(A). High-grade astrocytoma (glioblastoma multiforme) showing staining of neoplastic astrocytes for CYR.61 (B) and glial fibrillary acidic protein (C)
Expression of CTGF in Primary Gliomas.
Overexpression of CTGF occurred in 38 (58%) of 66 gliomas (Fig. 2B)⇓ ⇓ . The correlations of CTGF mRNA levels with clinical and pathological parameters were similar to those of CYR6 at diagnosis (Table 2)⇓ . A strongly significant association existed between pathology, tumor grade, and gender versus CTGF expression. Thirty-one (78%) of 40 samples from patients with GBM overexpressed CTGF; in contrast, only 6 (30%) of 19 individuals with astrocytoma and 1 (14%) of 7 cases with oligodendroglioma had increased levels of CTGF (P < 0.001). For tumor grade, only 2 (10%) of 19 brain tumors with either grade II or grade III overexpressed CTGF, compared with 36 (77%) of 47 brain tumors that were grade IV and also overexpressed CTFG (P < 0.001). Similar to CYR61, a significant association was noted between patients who were female and whose tumor overexpressed CTGF [16 (80%) of 20] versus those who were male with high levels of CTGF [22 (48%) of 46; P = 0.015]. Statistical analysis showed no significant difference between age (P = 0.39) versus CTGF expression in the primary brain tumors of 66 individuals.
Relationship between levels of expression of CTGF in gliomas and the clinical and pathological features of the individuals
Expression of WISP-1 and NOV in Primary Gliomas.
Although a relatively high percentage [24 (36%) of 66] of samples overexpressed WISP-1 (Fig. 2C)⇓ ⇓ , no significant correlations were found between WISP-1 expression versus the clinical and pathological parameters of these individuals and their tumors (Table 3)⇓ . In contrast to the other CCN gene family members, only 10 (15%) primary brain tumors had high levels of NOV mRNA among 66 patients (Fig. 2D)⇓ ⇓ . Similar to WISP-1, statistical analysis showed no difference between the clinical and pathological features and the level of expression of NOV in the primary brain tumors (Table 4)⇓ .
Relationship between levels of expression of WISP-1 in gliomas and the clinical and pathological features of the individuals
Relationship between levels of expression of NOV in gliomas and the clinical and pathological features of the individuals
Correlation of Expression among CYR61, CTGF, WISP-1, and NOV in Gliomas.
κ statistical analysis showed that expression of CYR61 and CTGF are highly correlated (κ = 0.82; 95% confidence interval, 0.68–0.95; Table 5⇓ ). Significant concordance also occurred between CYR61 and WISP-1 expression, and CTGF and WISP-1 expression, in the gliomas (κ = 0.39 and 0.42; 95% confidence interval, 0.17–0.61 and 0.22–0.61, respectively). The association of NOV with the other three genes was significant (κ = 0.25) but was much lower than the correlation among CYR61, CTGF, and WISP-1 (Table 5)⇓ .
Pair-wise comparisons of expression of the CYR61, CTGF, WISP-1, and Nov genes in gliomas
Relationship between Survival and Expression of CCN Genes in the Gliomas.
Cox proportional hazard model was used to investigate the correlation of expression levels of each of the CCN genes in the gliomas and survival of the patients, with the clinical characteristics (age, gender, pathology, and tumor grade) being controlled. Levels of CYR61, CTGF, and WISP-1, used as either continuous variables or dichotomized (positive versus negative), showed significant association with survival (Table 6⇓ ; Fig. 4⇓ ). In contrast, no significant association of NOV expression and survival was found. To compare independent prognostic impact, stepwise procedure was used for covariate selection. Multivariate (Cox regression) analysis showed that tumor pathology was the most significant predictor of clinical features for patient survival (Wald test: P = 0.0001; risk ratio 20.7; 95% confidence interval, 4.6–93). Among the four CCN genes, CYR61 was selected as the most significant independent prognostic factor for the risk of death (Wald test: P = 0.0001; risk ratio, 7.6; 95% confidence interval, 3.1–18.7).
Survival curves for glioma patients with high (+) or low (−) expression of CYR61 (A) CTGF (B), or WISP-1 (C) at the time of diagnosis.
Gliomas: relationship between survival and expression of CCN genes
Correlation between Expression of CCN Genes and Survival of GBM Patients.
Univariate analysis was used to study the association between expression of CCN genes and survival in 40 GBM patients. CYR61, CTGF and WISP-1, used as continuous variables or dichotomized (positive versus negative), showed significant association with patient survival (Table 7⇓ ; Fig. 5⇓ ). In contrast, no significant association of NOV expression and patient survival was found (data not shown). Statistical analysis failed to show significant association of age and gender with survival (Table 7)⇓ .
Survival curves for glioblastoma multiforme patients only with high (+) or low (−) expression of CYR61 (A), CTGF (B), or WISP-1 (C) at the time of diagnosis.
Glioblastoma multiforme: relationship between survival and expression of CCN genes
DISCUSSION
We performed real-time PCR to quantify the mRNA levels of CYR61, CTGF, WISP-1, and NOV. RNA quantitation using real-time PCR was made even more precise and reproducible by being based on Ct values established in the early exponential phase of the PCR reaction rather than the end point measurement of the amount of accumulated PCR product. Real-time PCR has high intra-assay and interassay reproducibility and provides statistical confidence. Despite the convenience and accuracy of real-time PCR, several factors have been considered to avoid potential errors that might affect the quality of the results. To exclude analysis of genomic DNA that might contaminate the RNA preparation, the probes were designed to cover an exon–exon junction of the gene of interest. All of the samples were quantified according to a standard curve, which was run on every PCR plate. In addition, the same standard samples were used for each PCR plate to standardize results among different plates. All of the samples were analyzed at least twice on independently prepared and analyzed reaction plates.
Expression of CYR61 was significantly correlated with gender and tumor grade, as well as patient survival. Overexpression of CYR61 was identified in 32 (48%) of the 66 glioma patients. Earlier studies have shown that CYR61 is overexpressed in breast cancers and may be involved in estrogen-mediated breast tumor development (10) . Significant associations were found between CYR61 expression and stage, tumor size, lymph node involvement, age, and estrogen receptor expression (11) . In this study, a significantly greater percentage of female patients (70%) had high levels of CYR61 in their gliomas compared with the percentage (39%) of males having high levels of CYR61. This divergence could possibly be explained because females have higher serum levels of estrogen and CYR61 is inducible by estrogen.
A previous study showed that overexpression of CYR61 was observed by Northern blot analysis in five (83%) of six glioma cell lines (28) . This small study did not examine primary glioma samples.
Recently, we observed that CYR61 acts as an oncogene through the integrin-linked kinase (ILK) to stimulate β-catenin-TCF/Lef and Akt signaling pathways.6 Forced expression of CYR61 in U343 glioma cells accelerated their growth in liquid culture, enhanced their anchorage-independent proliferation in soft agar, and significantly increased their ability to form tumors in nude mice. Overexpression of CYR61 resulted in β-catenin accumulation and nuclear translocation, leading to activation of β-catenin-TCF/Lef-1 signaling pathway. Furthermore, the data demonstrated that CYR61 can activate Akt by phosphorylation through a PI3 kinase-dependent manner, suggesting that CYR61 may stimulate several signaling pathways in the development of gliomas.
CTGF is transcriptionally activated with rapid kinetics in fibroblasts by serum growth factors (29) and TGF-β (30 , 31) . Moreover, CTGF has been implicated in cellular proliferation, migration, and tube formation of vascular endothelial cells in vitro and in angiogenesis in vivo (32 , 33) . Consistent with these properties, CTGF is often overexpressed in melanomas, sarcomas, chondrosarcomas, and pancreatic cancer cells (24 , 25 , 34 , 35) . Moreover, overexpression of CTGF was also noted in five of six glioma cell lines and some cells derived from primary gliomas by Northern analysis (28) . Recently, overexpression of CTGF has also been found in acute lymphoblastic leukemia and pediatric myofibroblastic tumors (36 , 37) .
In the present study, high levels of CTGF mRNA were noted in 38 (58%) of 66 glioma samples. Univariate analysis showed that gender, tumor stage, and pathology were significantly associated with overexpression of CTGF in the primary tumors. Thirty-one (78%) of the 40 GBM, the most malignant brain tumor, had high levels of CTGF compared with 6 (30%) of 19 astrocytomas and 1 (14%) of 7 oligodendregliomas, suggesting that overexpression of CTGF is correlated with tumor progression in gliomas. Consistent with our observation, a recent gene microarray study suggested that CTGF was one of the genes involved in the progression of gliomas (9) . CTGF has four identical structural domains and is closely related to CYR61. Both of them trigger downstream events via a signaling pathway through distinct integrins (27 , 38) .
WISP-1 was identified as a gene up-regulated in Wnt-1-transformed C57MG mouse mammary epithelial cells (16) . WISP-1 encodes a protein with a secretory signal peptide and has complete conservation of all 38 cysteine residues with those of CYR61 and CTGF. Overexpression of WISP-1 induced morphological transformation, increased cellular saturation density, promoted growth in normal rat kidney fibroblasts, and induced their tumor formation in nude mice (26) . WISP-1 was found to be highly expressed in breast cancer, and its expression was correlated with tumor stage, and lymph node status (11) . We found that it was highly expressed in 24 (36%) of 66 primary gliomas. WISP-1, used as continuous variable predictor, showed a significant association with survival of the patients. However, statistical analysis showed no significant association between other clinical and pathological parameters versus the level of WISP-1 in the gliomas. The role of WISP-1 in gliomas remains to be elucidated.
NOV was initially identified as an aberrantly expressed gene in chicken nephroblastomas induced by myeloblastosis-associated virus (39) . In our study, overexpression of NOV occurred in only 10 (15%) of 66 gliomas; and no significant correlation was found between expression of this gene and the clinical and pathological features. These findings suggest that NOV is not involved in either the development or progression of gliomas. Furthermore, a recent study showed that forced expression of Nov in the C6 glioma cell line inhibited cell growth and tumorigenic potential both in vitro and in vivo (40) .
In summary, our data indicate that overexpression of CTGF, WISP-1, and CYR61 may be involved in the process of development of gliomas and points to an association between expression of these proteins and several clinical and pathological features of these tumors. This comprehensive study of CCN gene expression in gliomas is an important first step in exploring the mechanism and function of these genes in the development of this malignancy. Clinically, our studies showed that prominent expression of the genes coding for CYR61 and CTGF is associated with gliomas being at an advanced stage at diagnosis. Statistical analysis showed that the level of expression of CYR61 and CTGF in the glioma at the time of diagnosis provided compelling prognostic information. Understanding the aberrant signaling pathways that are activated by high levels of expression of these CCN proteins, may offer useful therapeutic targets.
Footnotes
Grant support: Supported in part by NIH Grant NSFC303T0690, Marcia Schwartz Trust, and the Parker Hughes Research Fund. K. Black holds the Ruth and Lawrence Harvey Chair in Neurosciences. H. Koeffler is a member of the Jonsson Comprehensive Cancer Center and holds the endowed Mark Goodson Chair of Oncology Research at Cedars-Sinai Medical Center/UCLA School of Medicine.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Dong Xie, Division of Hematology/Oncology, Cedars-Sinai Medical Center, University of California-Los Angeles School of Medicine, Los Angeles, CA 90048. Phone: (310) 423-7740; Fax: (310) 423-0225; E-mail: xied{at}ucla.edu
↵5 http://www.genome.wi.nit.edu/cgi-bin/primer/primer3_www.cgi.
↵6 D. Xie, D. Yin, X. Tong, H. P. Koeffler. Cyr6l is overexpressed in gliomas and involved in integrin-linked kinase-mediated Akt and β-catenin-TCF/LEF signaling pathways, manuscript submitted.
- Received April 21, 2003.
- Revision received November 19, 2003.
- Accepted December 12, 2003.