Joint first authors on the research article, Hiroyuki Aizawa and Shu-Ching Hu, used a technique called transactivator trap to clone CREST. The trap, like yeast two-hybrid assays, takes advantage of the fact that the DNA binding and transactivation domains of most transcription factors can function independently. Aizawa and Hu made a hybrid library in which the DNA binding domain of the transcription factor GAL 4 is fused to cDNAs from the rat cerebral cortex. They then transfected these hybrids into embryonic cortical neurons carrying an enzyme (chloramphenicol acetyltransferase, or CAT) reporter complete with upstream GAL 4 binding site. The idea is that hybrids with a transactivation domain may activate expression of CAT. To look for transactivators that are calcium dependent, the authors tested for CAT activity in the absence and presence of potassium chloride, which induces calcium influx into the neurons.

Using this method, the authors identified four putative calcium-regulated transcription factors; they characterize only CREST in this paper, so more interesting developments may be in store. The protein has 402 amino acids and is most closely related to the transcription coactivator SYT, a proto-oncogene that has been linked to synovial sarcomas. By making deletion mutants, the authors found that the C-terminal end of CREST is essential for calcium activation, while the N-terminal end of the molecule is necessary to suppress basal activation in the absence of the cation.

Because SYT is known to interact with the histone acetyltransferase p300, the authors wondered if CREST has a similar role. To test this, they immunoprecipitated CREST from 293 cells, and found that both p300, and the p300 relative CBP, coprecipitated with the transactivator. Using the deletion mutants, Aizawa and Hu were able to show that CBP binds to the N-terminal end of CREST.

So what is the role of the calcium-driven transactivator? The authors used Northern blots to show that the protein is expressed not only in the brain, but also in heart, liver, kidney, and testes. In the brain, the protein is expressed mostly during development, declining quickly after birth and barely detectable in the adult. To test its function, the scientists made homozygous mutant mice that had the transactivation domain removed. Fewer than 20 percent of these mice live to adulthood. The most obvious phenotype of the mutant mice is their small cortex and cerebellum. This may be due to failure of neuronal dendrites to spread, as Aizawa and Hu found that these mutant dendrites were up to 80 percent shorter than normal.

Though CREST seems essential for the brain and other organs to develop normally, its downstream targets have not been identified. One hint: The authors have found that, in addition to CBP and p300, it interacts with chromatin remodeling proteins. Its mechanism of action also remains a mystery, though the CBP interaction may offer an important clue in this regard. As pointed out in an accompanying News Focus by Gregory Jeffries, Takaki Komiyama, and Liqun Luo from Stanford University, CREST itself does not have a calcium binding site, so it must be activated by another player, such as CBP, which is a known target of Ca2+-calmodulin-dependent kinase IV (CaMKIV).—Tom Fagan.

References:
Aizawa H, Hu S-C, Bobb K, Balakrishnan K, Ince G, Gurevich I, Cowan M, Ghosh A. Dendrite development regulated by CREST, a calcium-regulated transcriptional activator. Science 2004 January 9;303:197-202. Abstract

Jefferis GS, Komiyama T, Luo L. Neuroscience. Calcium and CREST for healthy dendrites. Science. 2004 Jan 9;303(5655):179-81. No abstract available. Abstract