AcK-knowledge Reversible Acetylation

See allHide authors and affiliations

Science's STKE  10 Aug 2004:
Vol. 2004, Issue 245, pp. pe42
DOI: 10.1126/stke.2452004pe42


In 1966, the histone was identified as the first protein subject to reversible acetylation. The ensuing 30 years of research on histone acetylation has been critical for elucidating how gene transcription and chromatin remodeling are regulated at the molecular level. This central focus on histones, however, has also restricted our understanding of reversible acetylation, and therefore the enzymes that catalyze this reaction, to cellular processes predominantly associated with chromatin. The study of reversible acetylation has become more or less synonymous with histone acetylation. Recent developments—including increased ability to detect acetylated proteins, the characterization of novel acetyltransferases and deacetylases, and the identification of specific inhibitors for these enzymes—have revealed that this histone-central paradigm probably reflects only a fraction of the cellular processes regulated by reversible acetylation. New studies have uncovered unexpected roles for reversible acetylation in many diverse areas, thereby establishing protein acetylation as a highly versatile signaling modification that has functions beyond gene transcription and chromatin remodeling.

Protein acetylation on lysine residues (Ac-K) is a dynamic, reversible, and highly regulated chemical modification (1). Through a large body of research that focused almost exclusively on histone modifications, reversible acetylation has been categorically linked to gene transcription and chromatin dynamics (2). When it comes to protein acetylation, histones were thought to be the canonical substrate. However, several recent studies have begun to challenge this simplistic paradigm and implicate reversible protein acetylation in a surprisingly diverse array of cellular processes including cell motility, protein trafficking, immune synapse formation, and apoptosis (36). These observations suggest that reversible acetylation has numerous functions that are just now being discovered. Here, we discuss two recent reports by Cohen et al. and Serrador et al. that expand the functional repertoire of protein acetylation to the regulation of apoptosis and T cell immune synapse formation.

By searching for proteins with sequence motifs similar to those surrounding the acetylated lysine residues found in the tumor suppressor p53, which had previously been identified as a substrate for acetylation, Cohen et al. (6) identified the DNA damage repair protein Ku70 as a target of reversible acetylation. Ku70 has been extensively characterized as a critical factor in the repair of DNA double-strand breaks (7). However, subsequent work revealed an unsuspected function for Ku70 in regulating apoptosis (8, 9). It was found that Ku70 can inhibit apoptosis by binding and sequestering the apoptotic activator Bax in the cytoplasm. Because apoptosis is a tightly controlled process, it might be expected that the Ku70-Bax interaction must somehow be regulated. Cohen and colleagues now report that the Ku70-Bax interaction is regulated by Ku70 acetylation. More specifically, acetylation of Ku70 is stimulated by ultraviolet (UV) irradiation, a known apoptosis-inducing agent. The acetylation of Ku70 results in its dissociation from Bax.

These findings lead to a simple model: In an unstressed state, Ku70 is hypoacetylated and sequesters Bax in the cytoplasm. In response to apoptotic stimuli, Ku70 becomes acetylated and subsequently releases Bax from its sequestration, thereby enabling Bax to translocate to the mitochondria and initiate an apoptotic cascade. This model is supported by the finding that apoptosis is augmented either by overexpression of the acetyltransferases CREB-binding protein (CBP) and CBP-associated factor (P/CAF), both of which can efficiently acetylate Ku70 in vitro. Additionally, treatment with deacetylase inhibitors such as trichostatin A (TSA) and nicotinamide can similarly mimic this effect on apoptosis. Therefore, either acetyltransferase overexpression or exposure to deacetylase inhibitors can enhance Ku70 acetylation, thereby leading to Bax dissociation and subsequent apoptosis. These results reveal a novel function for acetylation in the regulation of Bax-mediated apoptosis.

This straightforward model, however, leaves us with an important question: How do diverse environmental signals impinge on the acetylation status of Ku70? Is the activity of the Ku70 acetylase or deacetylase regulated in response to apoptotic stress? Cohen and colleagues provide one potential clue. They have identified P/CAF and CBP as the leading candidates for the Ku70 acetyltransferase. Intriguingly, they demonstrated an accumulation of CBP in the cytoplasm in response to UV irradiation and suggested that CBP may regulate the cytoplasmic Ku70-Bax complex. This observation may be of great importance because CBP is widely believed to function as a nuclear acetyltransferase involved in gene transcription and other chromatin-related processes (10). If correct, the hypothesis that CBP regulates the cytoplasmic Ku70-Bax complex would argue that CBP actually has cytoplasmic, nontranscriptional functions, at least in response to apoptotic stress. In agreement with this possibility, CBP can also be found in cytoplasmic polyglutamine-associated inclusion bodies that have been implicated in the pathogenesis of neurodegenerative diseases (11). Determining whether CBP can function by acetylating specific cytoplasmic protein substrates could lead to a drastically different understanding of this intensively studied but probably often misunderstood acetyltransferase. It could also provide new insight into the role of protein acetylation in neurodegeneration, which is intimately linked to apoptosis.

The other side of the equation concerns how cells keep Ku70 in a deacetylated state under normal conditions. Ku70 acetylation is enhanced by both TSA and nicotinamide—which inhibit class I/II HDAC (histone deacetylase) family and class III SIR2 family deacetylases, respectively—and therefore both HDAC and SIR2 members appear to be involved in Ku70 deacetylation. Because sequestration of Bax by Ku70 is thought to occur in the cytoplasm, it is tempting to speculate that HDAC6, a microtubule-associated cytoplasmic deacetylase, may be responsible for deacetylating Ku70, thereby negatively regulating Bax activity. HDAC6 has been implicated in several cellular processes, including the misfolded protein stress response. Through a novel nongenomic mechanism, HDAC6 assists with the disposal of toxic protein aggregates by facilitating their transport to a specialized proteolytic center, known as the aggresome (12). In doing so, HDAC6 promotes cell survival by preventing toxic protein aggregates from accumulating in the cytoplasm. In light of this prosurvival role for HDAC6, it seems plausible that this deacetylase could participate in additional environmental stress pathways as well, possibly regulating apoptosis through modulation of the Ku70-Bax interaction. In addition to HDAC6, HDAC4 and HDAC7 have been cytoplasmically localized and therefore could be potential candidates for regulation of the Ku70-Bax interaction (13).

In addition to the HDACs, the deacetylation of Ku70 apparently involves the class III nicotinamide adenine dinucleotide (NAD)–dependent SIR2 deacetylases, which have been implicated in aging and cellular longevity (14). It is noteworthy that members of this family have also been detected outside the nucleus, possibly indicating nongenomic roles for these deacetylases as well (15, 16). For example, the cytoplasmic localization of SIRT2 places it in the correct location to regulate Ku70 acetylation. Another SIR2 family member, SIRT3, has recently been characterized as a mitochondrial-specific deacetylase. Although it seems unlikely that this deacetylase would contact Ku70 directly, it is thought-provoking to speculate that a mitochondrial deacetylase might be directly involved in the regulation of apoptosis.

The identification of multiple acetyltransferases and deacetylases that regulate Ku70 acetylation supports the idea that Ku70-Bax complex formation is highly regulated. Such complexity is not surprising because abnormal regulation of Bax could lead to either inappropriate entry into apoptosis or an inability to enter apoptosis even under severe stress conditions. Interesting parallels can be drawn between Ku70 and p53, another key apoptotic regulator that is subject to reversible acetylation (17). Acetylation of p53 is thought to prevent MDM2-mediated ubiquitination, leading to enhanced p53 stability and activity (18). Similar to Ku70, p53 can be acetylated on multiple lysine residues, which appears to be a common feature among many acetylated proteins. Again, similar to Ku70, acetylation of p53 is catalyzed by both CBP and P/CAF, whereas deacetylation involves both the class I and class III deacetylases. This combinatorial regulation at multiple lysine residues by multiple enzymes may allow p53 and Ku70 to fine-tune a response to various environmental signals. In this manner, apoptosis might be regulated by the stepwise progression of p53 and Ku70 acetylation, whereby the hypoacetylated state corresponds to inhibition of apoptosis, and the hyperacetylated state corresponds to stress-induced activation of cell death through both p53-dependent and Bax-mediated pathways. Such a model would allow for a cooperative apoptotic response to environmental stresses through specific protein acetylation events.

In the other report, Serrador et al. (5) provide a link between HDAC6 and T cell immune synapse formation. The immune synapse is a specialized cell-cell junction between T cells and antigen-presenting cells (APCs) that is important for T cell activation (19). Serrador and colleagues show that, upon contact between T cells and APCs, HDAC6 quickly translocates to and concentrates at the immune synapse. HDAC6 translocation to the nascent immune synapse is correlated with a transient deacetylation of microtubules, consistent with the proposed microtubule deacetylase activity for HDAC6. Overexpression of wild-type HDAC6, but not a mutant HDAC6 that lacks deacetylase activity, can perturb organization of the immune synapse and the activation of T cells. These results reveal that HDAC6-regulated protein acetylation—possibly α-tubulin acetylation or modification of other components of the immune synapse—is important for the control of immune synapse formation. Although the exact function of α-tubulin acetylation remains obscure, it is most often associated with stable microtubules and is absent from more dynamic microtubule populations (20). Local α-tubulin acetylation could influence immune synapse formation by affecting the dynamics of microtubules surrounding the synapse or by affecting protein trafficking to and from the synapse. The latter possibility is particularly interesting. Although HDAC6 was initially identified as a microtubule-associated tubulin deacetylase, more recent studies demonstrate that HDAC6 can also dynamically interact with the microtubule motor dynein and has additional substrates other than α-tubulin. Indeed, the unique association with microtubules, dynein motors, and ubiquitinated proteins confers on HDAC6 the capacity to regulate the intracellular transport of misfolded protein aggregates. Because immune synapse formation involves the concentration of specific signaling molecules, it is logical to speculate that HDAC6 might regulate this process by modulating protein transport. Although the mechanism of how HDAC6 translocates to and regulates the immune synapse remains unclear, the findings of Serrador and colleagues potentially place the action of HDAC6-regulated protein acetylation close to the cell membrane and reveal a new location where reversible acetylation might play an important regulatory function.

Historically, modification of histones was perceived as the most important purpose of acetylation, if not its sole purpose. Recent studies, however, hint at a major change in the landscape of reversible acetylation. In addition to α-tubulin and Ku70, important nontranscriptional regulators such as importin α, ACS [acetyl–coenzyme A (CoA) synthetase], and the molecular chaperone Hsp90 were recently found to be regulated by reversible acetylation (2123). These new findings highlight a diverse role for acetylation in protein transport and metabolism as well as protein folding and assembly (Fig. 1). The discovery that specific members of the SIR2 family (such as SIRT3) are localized to mitochondria suggests the possibility that reversible acetylation is involved in modulating mitochondrial function. Undoubtedly, many acetylated protein substrates await discovery, and we are probably only at the beginning of an exciting era of acetylation biology. In the future, advanced proteomics techniques and the further characterization of acetyltransferases and deacetylases will pave the way for a deeper understanding of this highly versatile signaling modification.

Fig. 1.

Diverse cellular processes regulated by reversible acetylation. HDAC6 and SIRT2 can modulate the acetylation status of tubulin, thereby regulating cell motility. The tubulin acetyltransferase (TAT) for this process is currently unknown. Cytoplasmic deacetylases including both the HDACs and the Sir proteins can inhibit apoptosis through the Ku70-Bax interaction. Apoptosis is enhanced through the activity of CBP and P/CAF, although how these acetylases might respond to apoptotic insults is unknown. HDAC6 regulates the misfolded protein stress response in association with microtubule-based motor proteins. Potential acetylated substrates in this pathway have yet to be identified (X in yellow circle). In the nucleus, SIRT1 and HDAC1 acetylate and regulate p53 transcriptional activity, further modulating the apoptotic response. The mitochondrial-specific deacetylase SIRT3 may regulate apoptosis through acetylation of unidentified mitochondrial substrates (X in pink circle). ACS has been detected as an acetylated protein, although the role for such a modification in mitochondrial metabolism has yet to be explored. All acetyltransferases are depicted in blue, deacetylases in green, and known acetylated protein targets in red.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
View Abstract

Navigate This Article