DNA Packaging Defects in Breast Cancer

Institution: Lawrence Berkeley National Laboratory
Investigator(s): Terumi  Kohwi-Shigematsu , Ph.D. -
Award Cycle: 2000 (Cycle VI) Grant #: 6JB-0134 Award: $333,261
Award Type: IDEAS II
Research Priorities
Pathogenesis>Mistakes on the master blueprint: molecular genetics and gene regulations



Initial Award Abstract (2000)
Recently the human genome was reported to be approximately 3.15 billion pairs of DNA. The amount and length of this DNA is staggering, especially considering the incredibly tiny size of the cell's nucleus. Apparently, there are highly regulated ways that the DNA in the 23 pairs of chromosomes is packaged. One aspect of this packaging is already well known. It involves the histone proteins, and they combine with DNA to form a microscopic structure that looks likes "beads- on-a-string" (nucelosomes)- each nucleosome contains 146 base pairs of DNA, and it is repeated every 200 base pairs on average. Our laboratory studies a higher level of DNA organization that involves the attachment of specific stretches of DNA (called matrix attachment region or MARs) to a skeletal framework made of proteins and RNA molecules inside the nucleus. This attachment of DNA, which results in separating DNA into a series of loop domains, is mediated by a specific group of proteins in the nucleus. This loop organization is thought to be important not only for DNA compaction; it may also allow specific sequences to be "read" by enzymes in a timely manner when making proteins or replicating the DNA. In other words, this "packaging" may facilitate bookkeeping of the DNA's information.

Using previous support from the BCRP we have found a protein that binds to MARs is greatly increased in breast cancer cells. This protein was identified as poly (ADP-ribose) polymerse (PARP), which is known to play a role in DNA damage repair. Our new results strongly suggest that PARP also plays a key role in the DNA packaging of breast cancer cells through binding to MARs. Our hypotheses are that (i) PARP plays an essential role in the genome organization in breast cancer, and (ii) PARP is involved in the activation of genes related to growth and metastasis characteristic of breast cancer. Using the present BCRP support we will test these ideas by deleting PARP from aggressive breast cancer cells and determining the effect on both cell invasion properties and cancer-specific gene expression. In addition, we will use the newly developed techniques in gene microarrays to compare breast cancer cells having PARP present or those having PARP deleted. This will give us an initial read-out of the global effects of PARP in breast cancer. Our idea is that cancer cells may express these proteins that bind to MARs at different levels compared to normal cells, to organize their DNA to maintain their growth rate and invasive characteristics. Finally, we will use a mouse model system for human breast cancer. These mice are engineered to overproduce an oncogene (a gene that turns normal cells cancerous) so as to spontaneously form breast cancer and metastasizing to the lung. We will further engineer these mice to not produce PARP. We will then ask whether breast cancer can still develop and/or metastasize under these conditions.

These studies will provide important information as to how invasive breast cancer cells manage to maintain their aggressive phenotype. This information will also be useful for developing breast cancer therapies in the future.


Final Report (2002)
Cancer cells, generally characterized by uncontrolled cell growth and the ability to invade surrounding tissues, may have their own DNA organization to support these activities. Specialized DNA sequences, which have a high propensity to become base unpaired, are found to be tethered to the nuclear skeletal framework. These DNA sequences are serving as “anchoring” bases for chromatin loop domains. There is a short list of proteins identified to date that selectively bind to this set of DNA sequences. We have identified poly (ADP)-ribose polymerase (PARP) to be one such protein and it is expressed at a high level in breast cancer cells, but at a very low level in normal cells. When we depleted PARP from human aggressive beast cancer cells in culture, the breast cancer cells reverted to a much less invasive phenotype.

The purpose of this CBCRP funded research has been to study the roles of PARP in breast cancer. We aimed to identify a group of genes that are expressed differentially between aggressive breast cancer cells and the same cells with PARP being depleted. We also studied the tumor formation in mice genetically engineered to overproduce an oncogene, c-neu, either in the presence (wild type background) or the absence of PARP (PARP-depleted background).

We cloned 96 genomic sequences that bind to PARP in vivo in the human MDA-MB-231 breast cancer cells and sequenced them. Most sequences were identified to be repetitive sequences such as satellite 2, LINE 1, and LTR. Satellite 2 sequences are located at the major heterochromatin block of human chromosome 1. PARP may be involved in organization of heterochromatin in breast cancer cells. The human microarray analysis containing 20,000 genes revealed that several genes for signal tansduction, protein modification, cytoskeleton and chromosome condensation are differentially regulated by extracellular matrix (ECM) signals between PARP-depleted versus wild type MDA-MB-231 cells. Although the effect of PARP in breast tumor formation using a mouse model is still ongoing, we observed a promising preliminary data suggesting that mice overproducing c-neu, which are normally highly prone to grow breast tumors, become resistant to the tumor formation when PARP is genetically depleted.

The data described above strongly suggest that the nuclear protein PARP is important for sustaining an aggressive phenotype by organizing chromatin, mediating signals from extracellular matrix, and regulating multiple genes. These data also suggest that targeting PARP may be beneficial for cancer treatment.

Note: Dr. Kohwi-Shigematsu is continuing this topic of research with a new CBCRP grant in 2002.
Linking chromatin architecture to cellular phenotype: BUR-binding proteins and cancer
Periodical:Journal of Cellular Biochemistry. Supplement
Index Medicus: J Cell Biochem Suppl
Authors: Galande S and Kohwi-Shigematsu T
Yr: 2000 Vol: Suppl 35 Nbr: Abs: Pg:36-45
Caught in the act: binding of Ku and PARP to MARs reveals novel aspects of their functional interaction.
Periodical:Critical Reviews in Eukaryotic Gene Expression
Index Medicus: Crit Rev Eukaryot Gene Expr
Authors: Galande S, Kohwi-Shigematsu T.
Yr: 2000 Vol: 10 Nbr: 1 Abs: Pg:63-72