…your Histone Acetyl Transferase HAT, that is. In yesterday’s Neuron, two independent laboratories report that the HAT activity of CREB-binding protein (CBP) is crucial for learning and memory. Both groups came to this conclusion by studying models of Rubenstein-Taybi syndrome (RTS); this inherited disorder is caused by mutations in CBP and marked by skeletal malformation and mental retardation. Links between CBP, memory, and Alzheimer’s disease have been growing, most recently with data from Jie Shen’s lab suggesting that γ-secretase, the enzyme which releases amyloid-β peptides, maintains memory thorough a CBP-linked pathway (see ARF related news story).

In the first paper, Angel Barco, Eric Kandel and colleagues from Columbia University, New York, and RIKEN, Tsukuba, Japan, used a mouse model of RTS in which one copy of CBP is missing. This models the haploinsufficiency that causes some human forms of RTS. When first author Juan Alarcon characterized these mice, he found that they had normal levels of activity, motivation, anxiety, and short-term memory. However, long-term memory was another matter. Though in a number of tests, including the Morris water maze, the mice performed just as well as those with two copies of CBP, the heterozygous mice performed poorly in fear conditioning and novel object recognition, tasks that rely on storing long-term memories of a single prior event. When re-exposed to mild shock, for example, the RTS mice “froze” less than 15 percent of the time—in contrast, wild-type mice go rigid about 40 percent of the time in the same experiment.

To investigate the reason for this memory deficit, Alarcon and colleagues measured long-term potentiation (LTP), a physiological response in neurons that is both easily measured and essential for memory. Alcaron found that neurons in CBP-heterozygous animals have normal early LTP, but late LTP—which requires transcription—was attenuated, thus linking the memory defect to protein synthesis.

This connection makes sense because CBP is a co-factor for the transcriptional activation of many proteins. But is it the HAT activity of CBP that it particularly essential for LTP and memory, or something else? To address this, Alarcon examined the acetylation state of histones in the CBP+/- animals, finding that histone H2B in particular was reduced by about 50 percent. This suggests that the acetylation activity might be the missing component, a finding that was supported by giving the mice drugs that block deacetylation. When the histone deacetylase inhibitor suberoylanilide hyroxamic acid (SAHA) was introduced into the brain ventricles a few hours before a contextual fear conditioning experiment, the mice scored as well as normal animals.

In the second paper, Mark Mayford and colleagues at The Scripps Research Institute, La Jolla, and the University of San Diego, arrived at almost the same conclusions, though based on a slightly different model. By generating transgenic mice that express CBP with an inactive HAT domain, first author Edward Korzus directly tackled the question of which CBP activity is linked to RTS. In characterizing these mice, he also found that the animals had normal short-term memory but found them lacking when it came to the long-term memory needed for novel object recognition. Animals missing the CBP HAT domain performed about 30 percent less well than did wild-type mice. Korzus, too, found no difference between the HAT-negative mice and wild-type in the Morris water maze.

Significantly, the model Korzus chose eliminates developmental defects as an explanation for the memory impairment. He placed expression of the transgene under the control of a doxycyclin-sensitive promoter, and only after doxycycline was withheld from the diet of adult mice did they run into trouble with their long-term memories.

As in the Alcaron et al. study, Korzus and colleagues used a histone deacetylase inhibitor, in this case trichostatin A, to try to reverse the memory impairment. Injecting the inhibitor two hours before testing leveled the playing field for the transgenic mice , as SAHA had done in the first paper. However, it is interesting to note that Korzus found that histone H3 was preferentially acetylated in the presence of trichostatin, not histone H2B. This suggests that acetylation of different histones can rescue long-term memory deficits.

“Together, Alarcon et al. and Korzus et al. illuminate a central role for the histone acetyltransferase activity of CBP in long-term memory, providing strong support for the idea that chromatin remodeling serves to maintain memories,” write Kelsey Martin and Yi Sun from the University of California, Los Angeles, in an accompanying preview. The studies also shed light on the potential new “epigenetic therapeutic” approaches to treating mental retardation and other neurologic diseases, Martin and Sun write. Alzforum readers may already be familiar with the potential of histone deacetylase inhibitors, which have been shown to slow neurodegeneration in fruit fly models of polyglutamine diseases (see ARF related news story).—Tom Fagan


  1. Tom Fagan has done a very nice job of calling our attention to these exceptionally important and concordant sets of results supporting Robin Holliday's original proposal for a role for epigenetic modifications in long- term memory (Holliday, 1999). Robin emphasized methylations and demethylations of CpGs in his 1999 paper; less was known at that time about alterations in gene expression associated with chromatin modifications by histone acetylases and deacetylases.

    The important phenotypic consequences of haploinsufficiency for the CREB binding protein provides a rationale for an investigation of potential roles of genetic polymorphisms, or better, of haplotype variations, in the modulation of long-term memory in human subjects.

    PS: I also want to give special thanks to Tom Fagan for having highlighted the excellent accompanying commentary by Kelsey C. Martin and YE Sun. (Kelsey Martin is my daughter!)


    . Is there an epigenetic component in long-term memory?. J Theor Biol. 1999 Oct 7;200(3):339-41. PubMed.

  2. I see that many of the signs of Rubenstein-Taybi syndrome are also reported in Down's syndrome.

    The study by Branchi et al. (1) finding that overexpression of DYRK1A results in increased phosphorylation of FKHR, high levels of cyclin B1 and increased phosphorylation of CREB is of great interest.

    They report increased brain weight associated with DYRK1A overexpression, yet reduced brain weight is reported in Down's syndrome.

    Of further interest is the study by Daitoku et al. (2) finding that CREB-binding protein binds and acetylates FKHR and that Sir2 binds and deacetylates residues acetylated by CREB-binding protein.

    Do the CBP mutations reported in RTS allow for FKHR binding?

    Ironic that the CREB pathway required for learning and memory is also affected by the longevity gene, Sir2.

    Is this one reason for the mental retardation and reduced lifespan reported in Down's syndrome?


    . Transgenic mouse in vivo library of human Down syndrome critical region 1: association between DYRK1A overexpression, brain development abnormalities, and cell cycle protein alteration. J Neuropathol Exp Neurol. 2004 May;63(5):429-40. PubMed.

    . Silent information regulator 2 potentiates Foxo1-mediated transcription through its deacetylase activity. Proc Natl Acad Sci U S A. 2004 Jul 6;101(27):10042-7. PubMed.

  3. Which of the human sirtuins might target FKHR? Might we suspect SIRT2, as North et al. (1) find that it is a tubulin deacetylase.

    Hempen et al. (2) report decreased acetylated alpha-tubulin in neurofibrillary tangle-bearing neurons in AD.

    Does the overexpression of DYRK1A reported by Funakoshi et al. (3), which has resulted in chromosome missegregation, suggest that it may be a downstream effect on FKHR/SIRT2 and resultant deacetylated alpha-tubulin?

    Might we suspect a SIRT1/FKHR association? Takata and Ishikawa (4) report that SIRT1 associates with HES1 and HEY2.

    Does the fact that DYRK1A overexpression resulting in increased phosphorylated CREB also implicate this gene product in the increased presenilin-1 levels found in Down's syndrome?

    Mitsuda et al. (5) report that CREB controls expression of presenilin-1.

    I see that the Sambamurti group (6) find a CREB binding site on BACE.


    . The human Sir2 ortholog, SIRT2, is an NAD+-dependent tubulin deacetylase. Mol Cell. 2003 Feb;11(2):437-44. PubMed.

    . Reduction of acetylated alpha-tubulin immunoreactivity in neurofibrillary tangle-bearing neurons in Alzheimer's disease. J Neuropathol Exp Neurol. 1996 Sep;55(9):964-72. PubMed.

    . Overexpression of the human MNB/DYRK1A gene induces formation of multinucleate cells through overduplication of the centrosome. BMC Cell Biol. 2003 Sep 10;4:12. PubMed.

    . Human Sir2-related protein SIRT1 associates with the bHLH repressors HES1 and HEY2 and is involved in HES1- and HEY2-mediated transcriptional repression. Biochem Biophys Res Commun. 2003 Jan 31;301(1):250-7. PubMed.

    . Activated cAMP-response element-binding protein regulates neuronal expression of presenilin-1. J Biol Chem. 2001 Mar 30;276(13):9688-98. PubMed.

    . Gene structure and organization of the human beta-secretase (BACE) promoter. FASEB J. 2004 Jun;18(9):1034-6. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. The Senility-Presenilin Connection Turned Upside Down
  2. Drugs Slow Neurodegeneration in Fly Model of Huntington's

Further Reading

No Available Further Reading

Primary Papers

  1. . Chromatin acetylation, memory, and LTP are impaired in CBP+/- mice: a model for the cognitive deficit in Rubinstein-Taybi syndrome and its amelioration. Neuron. 2004 Jun 24;42(6):947-59. PubMed.
  2. . CBP histone acetyltransferase activity is a critical component of memory consolidation. Neuron. 2004 Jun 24;42(6):961-72. PubMed.
  3. . To learn better, keep the HAT on. Neuron. 2004 Jun 24;42(6):879-81. PubMed.