Allelic Losses at 1p36 and 19q13 in Gliomas: Correlation with Histologic Classification, Definition of a 150-kb Minimal Deleted Region on 1p36, and Evaluation of CAMTA1 as a Candidate Tumor Suppressor Gene

Allelic loss at 1p is seen in 70% to 85% of oligodendrogliomas (typically in association with 19q allelic loss) and 20-30% of astrocytomas. Because most 1p deletions in gliomas involve almost the entire chromosome arm, narrowing the region of the putative tumor suppressor gene has been difficult. To better define the histologic correlates of different patterns of 1p and 19q loss, we evaluated 1p/19q status in a large group of gliomas. This also allowed us to define a very small minimal deleted region (MDR) on 1p36.Among 205 consecutive cases of glioma studied for 1p loss of heterozygosity (LOH), 112 tumors were evaluated for both 1p and 19q LOH using at least three polymorphic markers on 1p and 19q each. The latter group included both low-grade tumors (oligodendroglioma, diffuse astrocytoma, and "oligoastrocytoma") and high-grade tumors (anaplastic oligodendrogliomas, anaplastic astrocytomas, anaplastic oligoastrocytomas). Tumors with small segmental 1p losses (defined as LOH at some loci with retention of heterozygosity at other loci) were studied using a more extensive panel of markers to define the 1p MDR. The candidate gene was screened for mutations and its expression was studied by qualitative and quantitative reverse transcriptase-PCR and Northern blotting.Allelic losses on 1p and 19q, either separately or combined, were more common in classic oligodendrogliomas than in either astrocytomas or oligoastrocytomas (P < 0.0001). Classic oligodendrogliomas showed 1p loss in 35 of 42 (83%) cases, 19q loss in 28 of 39 (72%), and these were combined in 27 of 39 (69%) cases. There was no significant difference in 1p/19q LOH status between low-grade and anaplastic oligodendrogliomas. In contrast, no astrocytomas and only 6 of 30 (20%) oligoastrocytic tumors had combined 1p/19q loss. Although rare, 1p deletions were more often segmental in astrocytomas (5 of 6, 83%) than in oligodendrogliomas (3 of 35, 9%; P = 0.006). Eleven tumors (6 oligodendrogliomas or having oligodendroglial components, 5 purely astrocytic) with small segmental 1p losses underwent further detailed LOH mapping. All informative tumors in the oligodendroglial group and 2 of 3 informative astrocytomas showed LOH at 1p36.23, with a 150-kb MDR located between D1S2694 and D1S2666, entirely within the CAMTA1 transcription factor gene. Mutation analysis of the exons encoding conserved regions of CAMTA1 showed no somatic mutations in 10 gliomas, including 6 cases with and 4 cases without 1p LOH. CAMTA1 is normally expressed predominantly in non-neoplastic adult brain tissue. Relative to the latter, the expression level of CAMTA1 was low in oligodendroglial tumors and was further halved in cases with 1p deletion compared with those without 1p deletion (Mann-Whitney, P = 0.03).Our data confirm the strong association of combined 1p/19q loss with classic oligodendroglioma histology and identify a very small segment of 1p36 located within CAMTA1 that was deleted in all oligodendroglial tumors with 1p LOH. This MDR also overlaps the neuroblastoma 1p36 MDR. CAMTA1 shows no evidence of inactivation by somatic mutations but its expression is reduced by half in cases with 1p LOH, suggesting that the functional effects of CAMTA1 haploinsufficiency warrant further investigation.