Selected Publication:
Aleksic-Babic, K.
Evolution of Chromosomal Instability in Hepatocarcinogenesis
Studied in Two Different Mouse Models
[ Dissertation ] Medical University of Graz; 2011. pp. 173
[OPEN ACCESS]
FullText
- Authors Med Uni Graz:
-
Aleksic-Babic Kristina
- Advisor:
-
Speicher Michael
- Altmetrics:
- Abstract:
- In this study we used a diethylnitrosamine (DEN)-induced HCC mouse model, the most widely used model for liver tumorigenesis, which corresponds to human HCC with poor prognosis. However, despite the frequent use of this HCC model, a detailed, mechanistic characterization of the longitudinal changes in the respective tumor genomes has never been performed. Here we established the chronological order of genetic alterations during DEN carcinogenesis by examining mice at different points in time. Tumor samples were isolated by laser micro-dissection and subjected to array-comparative genomic hybridization (array-CGH) and sequencing analysis. Chromosomal gains and losses were already observed in tumors by week 32 and increased significantly by week 56. Loss of distal chromosome 4q, including the tumor suppressors Runx3 and Nr0b2/Shp, was a frequent early event and persisted during all tumor stages. This region is syntenic to human 1p36.33-p32.1 deleted in about 15% of human HCC cases. Surprisingly, sequencing revealed that ß-catenin mutations occurred late and were clearly preceded by chromosomal instability. Thus, contrary to common belief, ¿-catenin mutations and activation of the Wnt/ß-catenin pathway are not involved in tumor initiation in this model of chemical hepatocarcinogenesis.
Beside the DEN mouse model, we investigated longitudinal genetic changes during hepatocarcinogenesis in the Mdr2 knockout (KO) mouse model of HCC by the same methodological approach. The Mdr2 KO mouse model corresponds to inflammation-associated HCC. Mdr2 KO mice firstly develop cholestatic hepatitis, which progresses to HCC in a stepwise manner. Therefore, this model is suited for longitudinal studies and the investigation of precancerous lesions. Copy number changes were already observed in tumors by month 13 and increased significantly by month 24. Loss of distal chromosome 8q, including many tumor suppressors such as the well known Dlc1, was the most frequent early event and persisted during all tumor stages. Interestingly, loss of distal chromosome 4q observed in the DEN model was also very frequent in the Mdr2 KO model, confirming the relevance of this region in HCC development. In contrast to DEN tumors, both early and late Mdr2 KO tumors were characterized by frequent focal amplification of the chromosomal 18qA1-qA2 region syntenic to human 10p11-p12.3 gained in ~10% of human HCC. Amplification of this region is observed in human HCC associated with HBV infection and also in several human HCC cell lines. The region contains several cancer-related genes, one of which is the Crem gene, which encodes a group of transcription factors playing an important role in liver regeneration. Crem gene amplification on the genomic level was additionally confirmed by FISH analysis in both Mdr2 KO tumors and several human HCC cell lines. In these human HCC cell lines, expression analysis showed upregulation of the ICER transcription factors, which are encoded by the CREM gene. Further functional studies will reveal exact impact of CREM gene amplification on HCC progression.
In contrast to the DEN model, Mdr2 KO tumors are characterized by a very low incidence of ß-catenin gene mutations. Conversely, they are frequently positive for mutations in the Axin1 tumor suppressor gene. Axin1 mutations are frequently found in HBV-associated human HCC cases, characterized by high level of CIN. Hence, this finding is an additional confirmation that Mdr2 KO model corresponds to the most frequent type of human HCC and is therefore a very useful model for cancer research.