Involvement of evolutionarily plastic regions in cancer associated CHR3 aberrations

Abstract

A functional test to identify tumor antagonizing regions on chromosome 3 (chr3), called the Elimination Test, was developed in our group. It is based on microcell mediated transfer of human chr3 into mouse or human tumor cells and analysis of the monochromosomal hybrids after their growth in vivo. We identified three regions on 3p14-p22, which were frequently lost in derived tumors (common or frequently eliminated regions). In order to understand the role of these regions in tumor development, we continued the study following two leads: analysis of breakpoint clusters to identify and characterize possible instability features and search for tumor suppressor genes within the deleted regions.

First, we identified and characterized a common eliminated region 1 (CER1) homologous sequence in mouse, where it was divided into two syntenic blocks on chromosome 9. In these blocks, the gene order and content was maintained with exception of two mouse gene duplications. Comparative analysis helped us to characterize five previously not identified mouse genes (Kiss et al. 2002). A more extensive comparative study of CER1 showed that its border regions are characterized by evolutionary plasticity: synteny breaks in several species, recent tandem gene duplications, retroposed pseudogene insertions, and horizontal evolution of the genes. Thus we showed that the cancer associated breakpoint regions have features of evolutionary plasticity. These results and other publications from our group suggested structural instability at the borders of eliminated regions identified by the Elimination Test (Darai et al. 2005) (Kost-Alimova et al. 2003; Kost-Alimova et al. 2004).

As a next step, in order to analyze rearrangements of entire chr3 in human tumor cells, we developed and compared two high resolution methods, array-CGH and mpFISH. We proved that although our 1Mb chr3 BAC/PAC array could identify single copy number changes even in pentaploid cells, mpFISH provided a more accurate analysis in the dissection of complex karyotypes at high ploidy levels. In heterogeneous or normal cell contaminated samples the most precise analysis can be made by mpFISH due to its ability to give information at single cell level (Darai-Ramqvist et al. 2006). Using high resolution methods we analyzed ten carcinoma cell lines and identified two new hot spots of tumor breakpoints at 3p12-p13 and 3q21. These tumor breakpoint regions carried large segmental duplications, retrotransposable elements and satellite repeats, which participated in recent primate evolution and, as we suggest, are associated with structural chromosomal instability (CIN). CIN is an ongoing dynamic process. Therefore in order to prove that the instability at the breakpoint regions characterizes structural CIN phenotype and it is required for tumor development and progression, dynamic analysis of the tumors must be done. This may elucidate the mechanism of tumor development; and may help to develop CIN phenotype markers useful in choice of consequent treatment.

Following the second lead of our study, we have analyzed a putative tumor suppressor gene LIMD1, which is located within the deleted central part of CER1. We found that it binds specifically to pRb and suppresses E2F driven transcription. A tumor suppressor effect of this gene was proven in in vitro and in vivo experiments, as well as in tumor biopsies (Sharp et al. 2004). In another part of the study we analyzed in details chr3 rearrangements in human renal cell carcinoma and nasopharyngeal carcinoma derived monochromosomal (chr3) hybrids and showed that aneuploid tumors maintain a mandatory chromosomal segment balance with stringency concerning no gain of 3p14-21 and no loss of 3q26-27. We concluded that the mechanism of tumor suppression by chr3 transfer is based on the alternative quantitative model. According to this model the tumor cell does not tolerate an increased dosage of the relevant gene or segment, and the lost part can be either of normal cell derived exogeneous or tumor derived endogenous origin (Kost-Alimova et al. 2007).

First, we identified and characterized a common eliminated region 1 (CER1) homologous sequence in mouse, where it was divided into two syntenic blocks on chromosome 9. In these blocks, the gene order and content was maintained with exception of two mouse gene duplications. Comparative analysis helped us to characterize five previously not identified mouse genes (Kiss et al. 2002). A more extensive comparative study of CER1 showed that its border regions are characterized by evolutionary plasticity: synteny breaks in several species, recent tandem gene duplications, retroposed pseudogene insertions, and horizontal evolution of the genes. Thus we showed that the cancer associated breakpoint regions have features of evolutionary plasticity. These results and other publications from our group suggested structural instability at the borders of eliminated regions identified by the Elimination Test (Darai et al. 2005) (Kost-Alimova et al. 2003; Kost-Alimova et al. 2004).

As a next step, in order to analyze rearrangements of entire chr3 in human tumor cells, we developed and compared two high resolution methods, array-CGH and mpFISH. We proved that although our 1Mb chr3 BAC/PAC array could identify single copy number changes even in pentaploid cells, mpFISH provided a more accurate analysis in the dissection of complex karyotypes at high ploidy levels. In heterogeneous or normal cell contaminated samples the most precise analysis can be made by mpFISH due to its ability to give information at single cell level (Darai-Ramqvist et al. 2006). Using high resolution methods we analyzed ten carcinoma cell lines and identified two new hot spots of tumor breakpoints at 3p12-p13 and 3q21. These tumor breakpoint regions carried large segmental duplications, retrotransposable elements and satellite repeats, which participated in recent primate evolution and, as we suggest, are associated with structural chromosomal instability (CIN). CIN is an ongoing dynamic process. Therefore in order to prove that the instability at the breakpoint regions characterizes structural CIN phenotype and it is required for tumor development and progression, dynamic analysis of the tumors must be done. This may elucidate the mechanism of tumor development; and may help to develop CIN phenotype markers useful in choice of consequent treatment.

Following the second lead of our study, we have analyzed a putative tumor suppressor gene LIMD1, which is located within the deleted central part of CER1. We found that it binds specifically to pRb and suppresses E2F driven transcription. A tumor suppressor effect of this gene was proven in in vitro and in vivo experiments, as well as in tumor biopsies (Sharp et al. 2004). In another part of the study we analyzed in details chr3 rearrangements in human renal cell carcinoma and nasopharyngeal carcinoma derived monochromosomal (chr3) hybrids and showed that aneuploid tumors maintain a mandatory chromosomal segment balance with stringency concerning no gain of 3p14-21 and no loss of 3q26-27. We concluded that the mechanism of tumor suppression by chr3 transfer is based on the alternative quantitative model. According to this model the tumor cell does not tolerate an increased dosage of the relevant gene or segment, and the lost part can be either of normal cell derived exogeneous or tumor derived endogenous origin (Kost-Alimova et al. 2007).