Pdgfr‐α in 1p/19q LOH Oligodendrogliomas
Christian Hartmann,Xiangli Xu,Gesine Bartels,Nikola Holtkamp,Isis Atallah Gonzales,Gesche Tallen,Andreas von Deimling
DOI: https://doi.org/10.1002/ijc.20525
2004-01-01
International Journal of Cancer
Abstract:PDGF-A stimulates O2-A oligodendrocyte precursor cells to proliferate in developing rat brain while differentiation is inhibited,1 and 2 oligodendroglioma mouse models rely on PDGF-B overexpression.2, 3 Marked expression of both PDGFR-α receptor and its ligand PDGF-A has been described in up to 94% of human oligodendrogliomas, suggesting autocrine and paracrine stimulation.4, 5 PDGFRA amplification has been identified in only a small number of anaplastic oligodendrogliomas WHO grade III with histologic features of WHO grade IV gliomas.6 Recently, gastrointestinal stromal tumors (GISTs) with increased expression of PDGFR-α have been shown to carry activating mutations in the intracellular part of PDGFRA.7 We investigated 108 gliomas (20 oligodendrogliomas WHO grade II, 32 anaplastic oligodendrogliomas WHO grade III, 18 astrocytomas WHO grade II, 17 anaplastic astrocytomas WHO grade III, 21 glioblastomas WHO grade IV) for point mutations in exons 2 to 21 of PDGFRA by SSCP analysis and subsequent sequencing. Primer sequences, amplification and sequencing conditions are available on request. We detected no mutations in PDGFRA, but found single nucleotide polymorphisms (SNPs) in exon 7 (14 cases, GGT-to-GGG, codon 360, silent), exon 10 (14 cases, TCC-to-CCC, codon 478, Ser-to-Pro) and exon 11 (1 case, GTT-to-TTT, codon 546, Val-to-Phe). The SNP in exon 11 has not been described before. No association between these SNPs and tumor type, grade, or location could be identified. Although the 5′ UTR and 3′ UTR exons 1, 22 and 23 were not analyzed so far, activating point mutations outside the coding region have been detected only rarely in receptor tyrosine kinases.7 In fact, PDGFRA mutations in GISTs were detected only in exon 12 and 18, coding for the activation loop and juxtamembrane domain, respectively.7 Therefore, we assume that activating mutations are an uncommon mechanism of genetic transformation in oligodendroglial and astrocytic gliomas. Similarly, PDGFRA amplification was already excluded before.6 A typical PDGFR-α Western blot using 4 LOH and 3 non-LOH oligodendrogliomas and normal cortex as control; in 2 of 4 LOH samples and in 1 of 3 non-LOH samples, PDGFR-α protein was detected, indicating no association between expression and 1p/19q genotype. Approximately 70% of oligodendrogliomas exhibit combined allelic loss of heterozygosity (LOH) on the short arm of chromosome 1 (1p) and the long arm of 19 (19q), indicating the involvement of tumor suppressor genes.8, 9 Up to now, these genes have not been identified.10 Upon coexpression with PDGF-B, the candidate gene p190ARhoGAP on 19q13.3 was found to reduce the incidence of oligodendrogliomas in a mouse model in comparison to PDGF-B expression alone.11 Therefore, we investigated whether LOH 19q implying reduced p190ARhoGAP gene dosage is responsible for increased PDGFR-α expression in human oligodendrogliomas. LOH 1p/19q was analyzed with microsatellite markers D1S1608, D1S548, D1S1597, D1S1592, D19S433, D19S431, D19S559 and D19S601 in 28 oligodendrogliomas. Amplification conditions and primer sequences are based on corresponding Genome Database entries (www.gdb.org). PCR products were separated on 8% denaturing acrylamide gels and visualized by silver staining. Eleven of 13 (85%) WHO grade II oligodendrogliomas exhibited LOH 1p/19q; 2 tumors retained both alleles. Eleven of 15 (73%) oligodendrogliomas WHO grade III showed allelic loss on 1p and 19q and 4 tumors retained both copies. This LOH 1p/19q distribution pattern is similar to results previously reported in oligodendrogliomas.8, 10 Proteins from these 28 oligodendrogliomas (13 WHO grade II and 15 WHO grade III) were denatured, separated by electrophoresis (7.5% Tris-glycine gel) and transferred to nitrocellulose membranes. Membranes were incubated overnight at 4°C with an antibody against PDGFR-α (C-20; dilution 1:200; rabbit polyclonal antibody; Santa Cruz Biotechnology, Heidelberg, Germany). Filters were incubated with a secondary peroxidase-labeled antibody (antimouse HRP, dilution 1:2,000, and the antirabbit HRP, dilution 1:10,000; Amersham Biosciences, Freiburg, Germany). Visualization was performed with enhanced chemiluminescence (Amersham Biosciences). No correlation between LOH 1p/19q and PDGFR-α expression was observed. Fourteen tumors expressed PDGFR-α; 14 oligodendrogliomas did not. Six of 11 oligodendrogliomas WHO grade II with LOH 1p/19q exhibited PDGFR-α; 5 grade II tumors revealed no expression. Six of 11 oligodendrogliomas WHO grade III with LOH 1p/19q showed PDGFR-α; 5 grade III LOH tumors were found without expression. Representative data are shown in Figure 1. In conclusion, we compile the central findings of both experimental series. First, our data suggest that mutations in PDGFRA are not common in gliomas; however, elevated protein levels occur in approximately half of all oligodendrogliomas. Second, our data in oligodendrogliomas showed no correlation between PDGFR-α expression and deletions on 19q that include p190ARhoGAP, which presumably coregulates PDGFR-α activity. Therefore, PDGFR-α regulation in oligodendrogliomas seems to be independent of a gene dosage effect exerted on p190ARhoGAP by LOH 19q. The role of p190ARhoGAP needs to be further evaluated. Yours sincerely, Christian Hartmann, Xiudan Xu, Gesine Bartels, Nikola Holtkamp, Isis Atallah Gonzales, Gesche Tallen, Andreas von Deimling.