By getting chromatin status in cells to revert back to its embryonic state, the large, intervening, non-coding RNA Hotair promotes cancer cell metastasis, and is a marker for breast cancer prognosis. In a series of experiments, Stanford University's Howard Chang and his team focused on Hox loci. "It's been called the Rosetta Stone of developmental biology," he says, since the genes there act to tell cells where they are located in the body. A few years ago, Chang's lab systematically assayed and identified a few hundred RNA-transcribed regions in Hox. They then started to look at the role of those regions in cancer progression. The expression levels of one of those regions, called Hotair, changed dramatically — by thousands of fold — during the course of cancer progression. Chang's team also noted that the levels of Hotair are heterogeneous in primary tumor samples. "That raised the idea that perhaps these different levels of Hotair in the primary tumors, in
really, really early stage tumors, could actually predict what's going to happen later on in these same patients," he says. "That was the case."
To examine the role of Hotair in cancer progression, the researchers turned to human cancer cell lines. In those lines, however, the levels of Hotair were not as high as they had seen in patients. They then increased the levels of Hotair in the cell lines, which led the cell lines to be more invasive and better able to metastasize, suggesting that Hotair promotes metastasis. "This is one of the — perhaps the — first examples that shows that lincRNAs are basically pathogenic molecules in human disease, specifically in cancer," Chang says. He adds that "by looking at [patients'] primary tumors at [the] time of diagnosis, the levels of Hotair provide information about what's going to happen five, 10 years down the line."
Chang's team set about studying how Hotair produces this effect. Over-expressed Hotair, they report in PNAS, mis-targets polycomb repressive complex 2, a histone H3 methylase involved in developmental gene silencing and cancer progression. Using a ChIP-chip approach, they looked at PRC2 occupancy throughout the genome under normal conditions as well as when Hotair is over-expressed. In the over-expressed state, H3K27me3 and two subunits of PRC2 were found on 854 new genes, though they left positions on only 37 genes. Many of the genes with over-expressed Hotair-induced PRC2 occupancy are involved in cell signaling and development, Chang found.
Using a pattern-matching algorithm, the team noticed that the Hotair-induced PRC2 occupancy pattern resembled the normal PRC2 occupancy pattern of embryonic fibroblasts. "The normal function of Hotair is basically to configure the chromatin state to tell the cells where they are located in the body — there's a specific chromatin pattern," Chang says. "Then when cancer cells mis-express Hotair, they basically reprogram themselves to think that they belong somewhere else in the body and access the genetic tools to then spread and metastasize. It's a very, I think, elegant way where disease states are actually subverting normal biology."
There's a host of ways that these findings could have an impact on clinical care, Chang says, pointing out that only the cancers that have high levels of Hotair would be sensitive to polycomb inhibition. "You can actually pick out the tumors that might be sensitive to polycomb in this way," he says. He also notes that when they targeted the PRC2 subunits with shRNA, the expression pattern reverted to normal.
Chang suspects that lincRNAs may even be involved in other diseases. "We are starting to find many other lincRNAs that are impacted by signaling pathways that are important in cancer as well in other developmental programs. It's very likely a very broad kind of mechanism," he says.