Centrosome clustering and cyclin D1 gene amplification in double minutes are common events in chromosomal unstable bladder tumors

Abstract

Background: Aneuploidy, centrosome abnormalities and gene amplification are hallmarks of chromosome instability (CIN) in cancer. Yet there are no studies of the in vivo behavior of these phenomena within the same bladder tumor.

Methods: Twenty-one paraffin-embedded bladder tumors were analyzed by conventional comparative genome hybridization and fluorescence in situ hybridization (FISH) with a cyclin D1 gene (CCND1)/centromere 11 dual-color probe. Immunofluorescent staining of alpha, beta and gamma tubulin was also performed.

Results: Based on the CIN index, defined as the percentage of cells not displaying the modal number for chromosome 11, tumors were classified as CIN-negative and CIN-positive. Fourteen out of 21 tumors were considered CIN-positive. All T1G3 tumors were included in the CIN-positive group whereas the majority of Ta samples were classified as CIN-negative tumors. Centrosome clustering was observed in six out of 12 CIN-positive tumors analyzed. CCND1 amplification in homogeneously staining regions was present in six out of 14 CIN-positive tumors; three of them also showed amplification of this gene in double minutes.

Conclusions: Complex in vivo behavior of CCND1 amplicon in bladder tumor cells has been demonstrated by accurate FISH analysis on paraffin-embedded tumors. Positive correlation between high heterogeneity, centrosome abnormalities and CCND1 amplification was found in T1G3 bladder carcinomas. This is the first study to provide insights into the coexistence of CCND1 amplification in homogeneously staining regions and double minutes in primary bladder tumors. It is noteworthy that those patients whose tumors showed double minutes had a significantly shorter overall survival rate (p < 0.001).

Figure 1

Chromosome 11 copy number variability. (A) Negative-CIN tumors. (B) Moderate-CIN tumors. (C) High-CIN tumors. (D) CIN index vs. tumour grade correlation.

Figure 2

Centrosome abnormalities. Immunolabeling was performed for γ-tubulin (red) and α and β-tubulin (green). DNA staining was performed with DAPI (blue). Black and white images correspond to DAPI reverse staining. (A-B) Normal/bipolar spindle. (C-D) Bipolar metaphase with string-like centrosome. (E) Tumour cells with abnormally long centrosomes, close to the adjacent normal urothelium. (F-G) Multipolar spindle. (H-I) Pseudo-bipolar metaphase. (J) Tumour cells with supernumerary centrosomes. Scale bar, 3 μm.

Figure 3

CCND1 amplification behaviour in bladder tumors. FISH identification of chromosome 11 centromere (green) and CCND1 gene (red) in paraffin-embedded tumors. DNA staining was performed with DAPI (blue). Black and white images correspond to DAPI reverse staining. (A-H) Metaphasic cells showing the proposed sequence of 11q13 amplicon fragmentation from HSRs to DMs. (I-K) Sample U-364 showed a complex pattern of CCND1 amplification. Three sub-populations were detected in this sample. (I) Sub-population with gain of whole chromosome 11. (J) Sub-population containing HSR with high-level amplification of CCND1 (K) Sub-population containing amplification of CCND1 and undetermined flanking material in HSR. (L-O) Peripheral location of DMs in metaphasic cells. (P-Q) CCND1-positive micronuclei, see arrows. In Q, note the elimination in the micronucleus of whole CCND1 copies, except those attached to the centromere. (R) Metaphasic cells containing a dicentric chromosome with two centromeric signals of chromosome 11 and CCND1 amplification, see asterisks. (S and T) CCND1 with HSRs appears to be forming internuclear bridges, see arrows. (U and V) Nuclear blebs as nuclear protrusions with high CCND1 signal.