Elsevier

Gene

Volume 567, Issue 2, 10 August 2015, Pages 103-131
Gene

Gene Wiki Review
The biological functions of Naa10 — From amino-terminal acetylation to human disease

https://doi.org/10.1016/j.gene.2015.04.085Get rights and content

Abstract

N-terminal acetylation (NTA) is one of the most abundant protein modifications known, and the N-terminal acetyltransferase (NAT) machinery is conserved throughout all Eukarya. Over the past 50 years, the function of NTA has begun to be slowly elucidated, and this includes the modulation of protein–protein interaction, protein-stability, protein function, and protein targeting to specific cellular compartments. Many of these functions have been studied in the context of Naa10/NatA; however, we are only starting to really understand the full complexity of this picture. Roughly, about 40% of all human proteins are substrates of Naa10 and the impact of this modification has only been studied for a few of them. Besides acting as a NAT in the NatA complex, recently other functions have been linked to Naa10, including post-translational NTA, lysine acetylation, and NAT/KAT-independent functions. Also, recent publications have linked mutations in Naa10 to various diseases, emphasizing the importance of Naa10 research in humans. The recent design and synthesis of the first bisubstrate inhibitors that potently and selectively inhibit the NatA/Naa10 complex, monomeric Naa10, and hNaa50 further increases the toolset to analyze Naa10 function.

Introduction

Nα-terminal acetylation (NTA) is one of the most abundant modifications of eukaryotic proteins. Today it is believed that the majority of the proteome of higher organisms is fully or partially acetylated. In fact, recent large-scale proteomics analyses have identified peptides that were fully or partially acetylated at their designated N-terminus in the following percentages: 13–19% in Halobacterium salinarum and Natronomonas pharaonis (Falb et al., 2006, Aivaliotis et al., 2007), 29% in Haloferax volcanii (Kirkland et al., 2008), about 16% in 45 tested bacteria (Bonissone et al., 2013), 60–70% in Saccharomyces cerevisiae (Arnesen et al., 2009b, Van Damme et al., 2011c, Van Damme et al., 2014, Bonissone et al., 2013), 75% in Drosophila melanogaster (Goetze et al., 2009), 90% in Arabidopsis thaliana (Bienvenut et al., 2012), at least 4% in Caenorhadbitis elegans (Mawuenyega et al., 2003), 83% in mouse (Lange and Overall, 2011), 90% in human erythrocytes (Lange et al., 2014) and 85% in HeLa cells (Arnesen et al., 2009b, Van Damme et al., 2011c). However, these values do not necessarily reflect the whole proteomes. A recent computational analysis of large-scale proteome analyses was used to develop a prediction software for NTA in archae (Pyrococcus furiosus, Thermoplasma acidophilum, H. salinarum and N. pharaonis), animals (Homo sapiens, C. elegans, and D. melanogaster), plants (A. thaliana and Oryza sativa) and fungi (S. cerevisiae and Neurospora crassa). The analysis revealed a bias for N-terminal acetylated proteins in highly abundant cytosolic proteins (Martinez et al., 2008). This bias could indicate that the reported percentage of acetylation is higher than the percentage in the actual proteomes: archeae 1–6.5%, animal 58%, fungi and plants 60% (Martinez et al., 2008). Furthermore, in some studies only annotated N-termini were analyzed, others included N-termini derived from alternative translation initiation sites.

In vitro data suggests that NTA occurs mainly co-translationally on the emerging polypeptide chain at a length of approximately 25–80 residues (Strous et al., 1973, Strous et al., 1974, Filner and Marcus, 1974, Driessen et al., 1985, Gautschi et al., 2003), either on the initiating methionine (iMet) or on the second amino acid after methionine cleavage, also known as N-terminal methionine excision (NME) (Kendall and Bradshaw, 1992, Xiao et al., 2010, Bonissone et al., 2013). The removal of the iMet is the first occurring widespread protein modification and involves peptide deformylases and methionine aminopeptidases (MetAPs) (Giglione et al., 2014). In addition to co-translational acetylation, accumulating evidence also supports the occurrence of post-translational Nα-acetylation. For example, the ribosomal protein L7/L12 in Escherichia coli becomes acetylated post-translationally depending on the availability of nutrients (Gordiyenko et al., 2008). Furthermore, proteomic analyses identified NTA of internal peptides, further supporting the idea of post-translational acetylation (Helbig et al., 2010, Helsens et al., 2011). This is especially interesting for many proteins that are imported into organelles, after which the cleaved mature N-terminus of the protein (now missing its target/transit peptide) is acetylated by dedicated NATs that reside in the respective target organelle as shown for yeast mitochondrial localized proteins (Van Damme et al., 2014) or chloroplast proteins in Chlamydomonas reinhardtii and A. thaliana (Zybailov et al., 2008, Bienvenut et al., 2011, Bienvenut et al., 2012).

NTA is catalyzed by distinct Nα-acetyltransferases (NATs) that belong to the GCN5-related N-acetyltransferase (GNAT) family, a diverse family that catalyze the transfer of an acetyl group from acetyl-CoA to the primary amine of a wide variety of substrates from small molecules to large proteins (Vetting et al., 2005). Besides the NATs, this protein family also includes lysine acetyltransferases (KATs) and histone acetyltransferases (HATs) (Marmorstein and Zhou, 2014).

In 2009, a new nomenclature for the Nα-acetyltransferases was introduced (Polevoda et al., 2009), in which the concept of multi-protein complexes for NATs was formalized. In humans, six NATs, NatA-F, were defined that specifically co-translationally catalyze the acetylation of the Nα-terminal amino group of a well-defined subset of proteins, although Nε-acetylation of internal lysines has also been reported (Kalvik and Arnesen, 2013). NatA consists of the catalytic subunit Naa10 and the auxiliary subunit Naa15 and acetylates small side chains such as Ser, Ala, Thr, Gly, Val and Cys after the initiator methionine has been cleaved by methionine aminopeptidases (via NME) (see Fig. 1). NatB and NatC are defined as multimeric complexes containing the catalytic subunits Naa20 and Naa30 and the auxiliary subunits Naa25 and Naa35/Naa38, respectively. They acetylate proteins with their methionine retained. The only known substrates for NatD (Naa40) are histone H2A and H4. Naa50 is the catalytic subunit of NatE, with a substrate specificity for N-termini starting with methionine followed by Leu, Lys, Ala and Met (Van Damme et al., 2011b). NatF is composed of Naa60 and has a substrate specificity that partially overlaps with NatC and NatE. It is important to note that this might not be the complete picture, as there are possibly other proteins binding and interacting with proteins in these NATs as currently defined [for reviews see Arnesen, 2011, Van Damme et al., 2011a, Starheim et al., 2012, Aksnes et al., 2015a].

Despite its discovery more than 50 years ago, very little is known about the biological function of NTA. For many years, it has been generalized from a few examples that NTA broadly protects many proteins from degradation. This was supported by the fact that acetylation of globin and lysozyme prevents their degradation by the ubiquitin proteolytic system from reticulocytes (Hershko et al., 1984). In line with this, the NTA of enkephalins diminishes their proteolytic cleavage by aminopeptidase M (Jayawardene and Dass, 1999), improves the chemical stability of guinea pig myelin basic protein (de Haan et al., 2004) and protects glucagon-like peptide (GLP-1) from dipeptidyl peptidase IV (DPP-IV)-mediated degradation (John et al., 2008). Analysis of the half-life of β-galactosidase in a split-ubiquitin system showed that proteins having N-terminal amino acids that are prone to acetylation (Met, Ser, Ala, Thr, Val, or Gly) have a relatively long half-life, whereas Arg, Lys, Phe, Leu, or Asp at the amino-terminus have very short half-lives (Bachmair et al., 1986). One way that NTA could contribute to protein stability is by blocking the access for N-terminal ubiquitination as shown for p16 and p14/p19ARF (Ben-Saadon et al., 2004, Ciechanover and Ben-Saadon, 2004, Kuo et al., 2004). Also, p21Cip1 is acetylated (most likely by NatA) at its N-terminus, whereas an N-terminally tagged variant that abolishes Nα-acetylation becomes ubiquitylated (Chen et al., 2004). However, other studies suggest that NTA has no effects on protein stability (Greenfield et al., 1994, Yi et al., 2011) and recent studies even showed that NTA might rather promote protein ubiquitination and degradation, depending on the cellular availability of interaction partners. The ubiquitin ligases Doa10 and Not4 recognize Nt-acetylated Ala, Val, Ser, Thr, and Cys and earmark acetylated substrates for degradation (Hwang et al., 2010, Shemorry et al., 2013). NTA therefore creates protein degradation signals (AcN-degrons) that are targeted by the Ac/N-end rule pathway resulting in ubiquitylation and proteasome-mediated degradation by the Doa10 E3 N-recognin, in conjunction with the Ubc6 and Ubc7 E2 enzymes (Varshavsky, 2011). Conversely, multiple Doa10 substrates do not require Nα-acetylation for their degradation, and acetylation has only mild effects on the stability of the tested substrates (Zattas et al., 2013). These discrepancies can be explained with the findings that Nα-acetylation can have stabilizing effects when an interaction partner is involved. In this case, the acetylated N-termini recruits the interaction partners that then shields the AcN-degron, preventing ubiquitinylation and degradation to regulate subunit stoichiometries (Shemorry et al., 2013, Park et al., 2015). In agreement with this, it is widely accepted that Nα-terminal acetylation can act as an avidity enhancer within protein complexes (Deakin et al., 1980, Scott et al., 2011, Nazmi et al., 2012). In this context, since the recent identification of an E2 ubiquitin ligase Ube2w that targets the α-amino group of proteins, it was speculated that – in this case – NTA could protect proteins from being proteasomally degraded by blocking of the target site (Aksnes et al., 2015a). NTA may also regulate NEDDylation, since Nα-terminal acetylation of the E2 enzymes UBC12 and UBE2F has been found to be required for recognition by DCN-like co-E3s which promote ligation of NEDD8 to cullin targets (Scott et al., 2011, Scott et al., 2014, Monda et al., 2013). This is also another good example where NTA acts as an avidity enhancer.

Other studies suggest that NTA may play a role in the structural stabilization of N-terminally flexible proteins, as bioinformatics analysis of the yeast proteome showed that proteins with N-terminal-disordered regions are more likely to be acetylated (Holmes et al., 2014). In line with that, many studies have shown that NTA can stabilize an N-terminal α-helix (Shoemaker et al., 1987, Fairman et al., 1989, Chakrabartty et al., 1993, Doig et al., 1994, Greenfield et al., 1994, Jarvis et al., 1995, Fauvet et al., 2012, Kang et al., 2012, Kang et al., 2013).

Another function of NTA has been implicated in protein sorting and secretory processes. In yeast, the ARF-like GTPase Arl3p/ARP is acetylated by NatC and this modification is required for its targeting to the Golgi apparatus, possibly through the acetylation-dependent interaction with the integral membrane protein Sys1p (Setty et al., 2004). Simultaneously, a different group confirmed that Sys1p is the receptor for Arl3p and knockout of NatC or mutation of the NatC complex to abrogate its acetyltransferase activity resulted in failure to target Arl3p to the Golgi (Behnia et al., 2004). Furthermore, targeting of the human homologue of Arl3p, ARFRP1, is dependent on Sys1p and mutation of the N-terminus of ARFRP1, that would abrogate acetylation by NatC, induced its mis-localization in COS cells (Behnia et al., 2004). This and the fact that the N-terminus of Arl3p is a potential NatC substrate in S. cerevisiae, D. melanogaster, C. elegans and plants indicate that this system is well conserved. Two other human lysosomal Arf-like GTPases, Arl8a and Arl8b (also known as Arl10b/c and Gie1/2), and their single homologue in Drosophila are potential substrates of NatC, and mass spectrometric analyses confirmed that human Arl8b is N-terminally acetylated (Hofmann and Munro, 2006). Later in vitro acetylation assays showed that Arl8b is acetylated by NatC and knockdown of the catalytic subunit of NatC (Starheim et al., 2009a) or replacement of the leucine in position 2 with alanine (Hofmann and Munro, 2006) resulted in a loss of its lysosomal localization. It should be mentioned that the protein was still found to be acetylated, presumably by NatA following removal of the initiator methionine (Hofmann and Munro, 2006), indicating that specifically the acetylated methionine rather than acetylation itself is important for lysosomal targeting of Arl8b. Also, the inner nuclear membrane protein Trm1-II was found to be mislocalized to the nucleoplasm, when NatC was knocked out or when the second amino acid was mutated to inhibit NatC-dependent NTA (Murthi and Hopper, 2005). On the other hand, systematic analysis of predicted N-terminal processing in yeast showed that cytoplasmic proteins are typically acetylated, whereas those destined for secretion via the ER are largely unmodified (Forte et al., 2011). Mutation of the N-terminal amino acid of the secretory protein carboxypeptidase Y, which allowed acetylation of this protein, inhibited targeting to the ER (Forte et al., 2011). However, fluorescence microscopy analysis in yeast indicated unaltered subcellular localization patterns for all 13 studied NatC substrates, after disruption of the NatC catalytic subunit (Aksnes et al., 2013). Furthermore, no disruption of the nuclear membrane, endoplasmic reticulum, Golgi apparatus, mitochondria, or bud neck was observed upon NatC deletion, suggesting the intactness of these organelles and subcellular structures as judged by the unchanged shape, number, size and distribution in the cell (Aksnes et al., 2013). A follow-up study showed that not NatC but rather NatF (Naa60) is associated with Golgi membranes and the ER, and disruption of this Naa60 induces Golgi fragmentation (Aksnes et al., 2015b). Taken together, this indicates that NatC is not – at least not in general – a determinant for substrate subcellular localization (Aksnes et al., 2013). Similarly, fluorescence microscopy analysis of 13 NatB substrates in wild type and NAA20Δ yeast cells revealed that acetylation by NatB is not a general signal for protein localization (Caesar et al., 2006).

Aside from the above, NTA has been shown to affect protein function and/or activity in a variety of cases, including hemoglobin isoforms (Scheepens et al., 1995, Ashiuchi et al., 2005), phospholamban (PLB) (Starling et al., 1996), N-TIMPs (N-terminal inhibitory domains of TIMPs/inhibitors of metalloproteinases) (Van Doren et al., 2008), parvalbumin (Permyakov et al., 2012), melanocyte-stimulating hormone (MSH) in the barfin flounder (Verasper moseri) (Kobayashi et al., 2009), the contractile proteins actin and tropomyosin in fission and budding yeast (Polevoda et al., 2003, Singer and Shaw, 2003, Coulton et al., 2010) as well as the stress-induced carboxypeptidase Y inhibitor Tfs1p in yeast (Caesar and Blomberg, 2004).

In addition, NTA has been linked to various diseases, including apoptosis and cancer (Kalvik and Arnesen, 2013), host parasite interaction in malaria (Chang et al., 2008), and has been discussed to play a role in neurodegenerative disorders (see below). As pointed out in earlier reviews: “Although…[NTA]…is essential for cell viability and survival, very little is known about the physiological reasons associated with this crucial role” (Giglione et al., 2014) and “there may be a variety of acetylation-dependent functions depending on the target protein, rather than one general function [and] there is even the possibility that this modification affects the function of only very few proteins” (Arnesen, 2011). The cellular phenotypes observed by disruption of the different NATs have been summarized in a very recent review (Aksnes et al., 2015a).

The best studied Nα-acetyltransferase NatA consists of the catalytic subunit Naa10 and the auxiliary subunit Naa15. In this review we mainly concentrate on Naa10 structure and function and discuss recent developments in the field.

As mentioned above, the NatA complex consists at least of the auxiliary and catalytic subunits, Naa15 and Naa10, respectively and is evolutionarily conserved from yeast to vertebrates (Mullen et al., 1989, Park and Szostak, 1992, Sugiura et al., 2003, Arnesen et al., 2005a). We adopt here the nomenclature of inserting letters to indicate the species about which we are discussing, so yNatA refers to NatA in yeast, where we are specifically referring to S. cerevesiae, hNatA refers to NatA in humans, and mNatA refers to NatA in mice. However, this nomenclature in 2009 did not address other species, and it might be worth updating the nomenclature at some future international meeting focused on the NATs.

In any case, there is good in vitro and in vivo evidence that yNatA acetylates the N-termini of small side chains like serine, alanine, glycine and threonine (Arnold et al., 1999, Polevoda et al., 1999) and NatA from humans has identical or nearly identical specificities, acetylating proteins starting with small side chains like serine, glycine, alanine, threonine and cysteine (Arnesen et al., 2009b, Van Damme et al., 2011b, Van Damme et al., 2011c), after the removal of the initiator methionine by methionine aminopeptidases. It was found that heterologous combinations of human and yeast Naa10p and Naa15p are not functional in yeast, suggesting significant structural subunit differences between the proteins from the different species (Arnesen et al., 2009b). (Met)-Ala-N-termini are more prevalent in the human proteome, whereas (Met)-Ser-N-termini are more abundant in the yeast proteome (Van Damme et al., 2011c). Accordingly, hNatA displays a preference towards these Ala-N-termini whereas yNatA seems to be the more efficient in acetylating Ser-starting N-termini, indicating that NatA substrate specificity/efficiency of NTA has co-evolved with the repertoire of NatA type substrates expressed (Van Damme et al., 2014).

Size-exclusion chromatography and circular dichroism showed that purified human Naa10 consists of a compact globular region comprising two thirds of the protein and a flexible unstructured C-terminus (Sánchez-Puig and Fersht, 2006). The recent X-ray crystal structure of the 100 kDa holo-NatA (Naa10/Naa15) complex from Schizosaccharomyces pombe revealed that the auxiliary subunit Naa15 is composed of 37 α-helices ranging from 8 to 32 residues in length, among which 13 conserved helical bundle tetratricopeptide repeat (TPR) motifs can be identified (Liszczak et al., 2013). These Naa15 helices form a ring-like structure that wraps completely around the Naa10 catalytic subunit (Liszczak et al., 2013). TPR motifs mediate protein–protein interactions, and it was speculated that TPR might be important for interaction with other NatA-binding partners such as the ribosome, Naa50/NatE and the HYPK chaperone (Liszczak et al., 2013), but this needs to be proven in future experiments. We discuss the possible interaction with NatE in more detail below. Naa10 adopts a typical GNAT fold containing a N-terminal α1–loop–α2 segment that features one large hydrophibic interface and exhibits the most intimate interactions with Naa15, a central acetyl CoA-binding region and C-terminal segments that are similar to the corresponding regions in Naa50 (Liszczak et al., 2013). The X-ray crystal structure of archaeal Thermoplasma volcanium Naa10 has also been reported, revealing multiple distinct modes of acetyl-CoA binding involving the loops between β4 and α3 including the P-loop (Ma et al., 2014). To our knowledge, there is not yet any published cryo-electron microscopy data regarding larger complexes between the ribosome, nascent polypeptide chain and any NATs.

Besides acting in a complex, it has been shown that a fraction of human Naa10 exists independent of Naa15 in the cytoplasm and is able to acetylate acidic side chains like aspartate and glutamate in γ- and β-actin (Van Damme et al., 2011b, Foyn et al., 2013a). These Type I actins are natural NatB substrates (iMet followed by an amino acid with an acidic side chain) and are therefore initially acetylated by NatB in yeast and humans (Van Damme et al., 2012). However, further processing/cleavage by an Nα-acetylaminopeptidase (ANAP), which specifically removes the N-terminal Ac-Met or Ac-Cys from actin exposes the acidic N-terminal residue (Polevoda and Sherman, 2003b), which can be subsequently acetylated by Naa10. This substrate switching of Naa10 from small side chains towards acidic side chains could be explained by comparing the X-ray crystal structures of complexed (Naa15-bound) and uncomplexed Naa10 of S. pombe. The complexed form of Naa10 adopts a GNAT fold containing a central acetyl CoA-binding region and flanking N- and C-terminal segments that allows the acetylation of conventional substrates (Liszczak et al., 2013). In the uncomplexed form, Leu22 and Tyr26 shift out of the active site of Naa10, and Glu24 is repositioned by ~ 5 Å resulting in a conformation that presumably allows for the acetylation of acidic N-termini (Liszczak et al., 2013). However, some proteins starting with an N-terminal acidic amino acid are usually further modified by arginyl-transferases and targeted by the Arg/N-end rule pathway for degradation (Varshavsky, 2011). Therefore, further studies have to show if Type I actins are unique substrates of non-complexed Naa10 and/or if more in vivo substrates with acidic N-termini exist. Such studies also need to explore whether the NTA of actin does trigger any downstream processing in the Arg/N-end rule pathway.

Apart from its function as an N-terminal acetyltransferase, NatA has been shown at least in vitro to possess N-terminal propionyltransferase activity (Foyn et al., 2013b) and lysine acetylation activity (Jeong et al., 2002, Lin et al., 2004, Lim et al., 2006, Lim et al., 2008, Yoo et al., 2006, Lee et al., 2010b, Shin et al., 2014). Autoacetylation at an internal lysine K136 in human and mouse Naa10 has been shown to regulate its enzymatic activity (Seo et al., 2010, Seo et al., 2014). Particularly, in in vitro acetylation assays with recombinantly expressed and purified wt Naa10 and subsequent detection with anti-acetyl-lysine antibody revealed acetylation of mNaa10225, mNaa10235 and hNaa10235 (see below for isoforms), whereas Naa10 with K82A and/or Y122F mutations, that inhibit acetyl-CoA binding, or Naa10 K136 (a mutation of the putative acetyl-acceptor site) were not found to be acetylated (Seo et al., 2014). In strong contrast to this, LC/MS/MS analyses on human Naa10 expressed and purified from E. coli did not show acetylation at any of the internal 16 lysines after incubation with Acetyl-CoA (Murray-Rust et al., 2006). Generally, ε-acetylation of internal lysines by Naa10 and other NATs is quite controversial in the field, and structural comparisons suggests that specific loops (β6–β7 hairpin loops in NatA and NatE or an extended α1–α2 loop in NatD) which are absent in KATs (lysine acetyltransferase) would block internal lysines from being inserted into the active site (Magin et al., 2015). Thus it is not yet known how much lysine is directly acetylated by Naa10 in vivo, and the degree to which autoacetylation of Naa10 occurs in vivo is also currently not well characterized.

Naa10 (Nα-acetyltransferase 10; NatA catalytic subunit; ARD1, arrest-defective protein 1 homologue; DXS707; TE2), the catalytic subunit of NatA, has an apparent molecular weight of 26 kDa and contains a typical Gcn5-related N-acetyltransferases (GNAT) domain. In mouse, NAA10 is located on chromosome X A7.3 and contains 9 exons. Two alternative splicing products of mouse Naa10, mNaa10235 and mNaa10225, were reported in NIH-3T3 and JB6 cells that may have different activities and function in different subcellular compartments (Chun et al., 2007). The human NAA10 is located on chromosome Xq28 and is encoded by 8 exons (Tribioli et al., 1994). According to RefSeq (NCBI) (Pruitt et al., 2007), three different isoforms, Naa10235, Naa10220 and Naa10229, derived from alternate splicing (see Fig. 2) exist. Recently, an additional putative splice variant, hNaa10131 (GenBank accession no. BC063377), was identified by Sanger sequencing in a single clone after amplification of Naa10 from HeLa cDNA (Seo et al., 2015). However, this isoform is identical to Naa10 isoform 2, except that the splice acceptor site is shifted by 1 bp, which results in a frameshift in the resulting mRNA and a premature translational stop. Since this variant is neither annotated at NCBI, UniProt or Ensembl and was not detected on the protein level, we will not further discuss implications of this putative variant. Additionally, a processed NAA10 gene duplicate NAA11 (ARD2) has been identified that is expressed in several human cell lines including Jurkat, HEK293 and NPA (Arnesen et al., 2006b). However, later studies have revealed data arguing that Naa11 is not expressed in the human cell lines HeLa and HEK293 or in cancerous tissues, and NAA11 transcripts were only detected in testicular and placental tissues (Pang et al., 2011). Therefore, the functional role of Naa11 might be constricted to certain tissues only. Naa11 has also been found in mouse, where it is mainly expressed in the testis (Pang et al., 2009). NAA11 is located on chromosome 4q21.21 or 5 E3 for human or mouse, respectively, and only contains two exons.

Human Naa15 (Nα-acetyltransferase 15; NatA auxiliary subunit, NMDA Receptor-Regulated Protein; NARG1; Tubedown-100; Tbdn100; tubedown-1, NATH) is well described as the auxiliary subunit of NatA, although it may have NatA-independent functions. The human NAA15 gene is located on chromosome 4q31.1 and contains 23 exons. Initially, 2 mRNA species were identified, 4.6 and 5.8 kb, both harboring the same open reading frame encoding a putative protein of 866 amino acids (~ 105 kDa) protein that can be detected in most human adult tissues (Fluge et al., 2002). According to RefSeq/NCBI (Pruitt et al., 2007), only one human transcript variant exists, although 2 more isoforms are predicted. In addition to full length Naa15, a N-terminally truncated variant of Naa15 (named tubedown-1), Naa15273–865, has been described (Gendron et al., 2000). However, northern blot analyses of poly(A) mRNA from mouse revealed that full length Naa15 is widely expressed, whereas smaller transcripts were visualized only in heart and testis (Willis et al., 2002).

Similar to the situation with NAA10, a NAA15 gene variant has been identified, NAA16, that originates from an early vertebrate duplication event (Arnesen et al., 2009a). The encoded protein shares 70% sequence identity to hNaa15 and is expressed in a variety of human cell lines, but is generally less abundant as compared to hNaa15 (Arnesen et al., 2009a). Three isoforms of Naa16 are validated so far (NCBI RefSeq). Mouse NAA15 is located on chromosome 2 D and contains 20 exons, whereas mouse NAA16 is located on chromosome 14 D3 and consists of 21 exons.

It has been shown in principle that NatA could assemble from all the isoforms. Naa15 interacts with Naa11, in humans (Arnesen et al., 2006b) and mouse (Pang et al., 2009), and Naa10 interacts with the Naa15 paralogue, Naa16, creating a more complex and flexible system for Nα-terminal acetylation as compared to lower eukaryotes (Arnesen et al., 2009a). Such a system might create the opportunity for functional redundancy or compensation in the event of loss of Naa10 or Naa15, although we are not aware of any studies showing whether NAA11 expression might be upregulated in tissues lacking or having reduced NAA10 expression.

Many studies document the interaction between Naa10 and Naa15 and since a crystal structure of S. pombe Naa10 bound to Naa15 has been published recently, it is safe to conclude that this is a very stable complex. We await such crystal structures for human Naa10 and Naa15. As an enzyme, NatA transiently interacts with a variety of substrates per se. However, some evidence indicates that NatA constitutively interacts with specific proteins to assemble into trimeric or even larger complexes, which might modulate NatA function. One such example is Naa50 (Nα-acetyltransferase 50; NatE; NAT13; Mak3; Nat5, SAN). Naa50 is the catalytic acetyltransferase subunit of NatE, is expressed in several human cell lines, and has been shown to be associated with NatA in yeast (Gautschi et al., 2003), fruit fly (Williams et al., 2003) and humans (Arnesen et al., 2006a). Naa50 has a distinct substrate activity for Met followed by a hydrophobic amino acid in human and yeast (Polevoda et al., 1999, Evjenth et al., 2009, Evjenth et al., 2012) and has been claimed to possess ε-acetyltransferase activity towards K525 in β-tubulin (Chu et al., 2011) and histone 4 (Evjenth et al., 2009). Furthermore, hNaa50 has been shown to harbor autoacetylation activity on internal lysines (K34, K37 and K140) in vitro, modulating Naa50 substrate activity (Evjenth et al., 2009, Evjenth et al., 2012). However, a recent structural study on human Naa50 contradicts these findings. The X-ray crystal structure of human Naa50 revealed, similar to Naa10, a GNAT fold with a specific substrate binding groove that allows acetylation of α-amino substrates but excludes lysine side chains (Liszczak et al., 2011). This seems to be strong evidence against a role for Naa50 in direct acetylation of lysine side chains. Further studies have to sort out these discrepancies.

Because Naa50 has a distinct/different substrate specificity and NAA50Δ cells did not display the NatA phenotype in yeast (Gautschi et al., 2003), Naa50 was considered as an independent NAT and was named NatE (Starheim et al., 2009b). Furthermore, in HeLa cells, more than 80% of endogenous Naa50 is not associated with the NatA complex (Hou et al., 2007). Therefore, future experiments have to examine whether Naa50 has a distinct function independent of NatA and/or if Naa50 works in a cooperative manner with NatA.

Recently, the chaperone like protein HYPK (Huntingtin Interacting Protein K) was shown to interact with Naa10 and 15 and is required for NTA of the known in vivo NatA substrate PCNP (Arnesen et al., 2010). However, it is an open question whether HYPK generally is required for NatA-mediated acetylation of downstream substrates.

Further interaction partners of Naa10 have been identified including Myosin light-chain kinase (Shin et al., 2009), tuberous sclerosis 2 (Kuo et al., 2010) RelA/p65 (Xu et al., 2012) DNA methyltransferase 1 (C.F. Lee et al., 2010), androgen receptor (Wang et al., 2012) and proteasome activator 28β (Min et al., 2013). In high throughput screens cell division cycle 25 homologue (Rual et al., 2005) and Rho guanine nucleotide exchange factor 6 have been shown as interaction partners of NatA (Xiao et al., 2007). Additionally, β-Catenin (Lim et al., 2006, Lim et al., 2008), HIF-1α (Jeong et al., 2002, Arnesen et al., 2005b) and methionine sulfoxide reductase A (Shin et al., 2014) have been suggested to bind to Naa10. In a recent high-throughput study, multiple orthogonal separation techniques were employed to resolve distinct protein complexes. Fractionation of soluble cytoplasmic and nuclear extracts from HeLa S3 and HEK293 cells into 1163 different fractions identified several interaction partners for Naa10 (Naa15, Naa16, Mina, M89BB, TCEA1 and PLCβ3), Naa15 (RT21, ML12A, HYPK and Cap1) and Naa16 (TCEA1, PLCB3, Naa10 and Mina) (Havugimana et al., 2012). However, these interactions seem to be transient in the cell and have not yet been shown to regulate or change NatA function; therefore, we do not list them as part of any putative larger NatA complex. Evidence of a larger stable complex, other than the dimer between Naa10 and Naa15, could come from structural studies, including possibly cryo-electron microscopy.

Mainly from yeast data, it is thought that the auxiliary subunits of NatA as well as other NATs are associated with mono- and polysome fractions and co-translationally acetylate the nascent polypeptide chain as it emerges from the ribosome (Gautschi et al., 2003, Polevoda et al., 2008). In line with this, it has been shown that human Naa10 and Naa15, HYPK (Arnesen et al., 2010), the human paralog of Naa15, Naa16 (Arnesen et al., 2009a) as well as yeast Naa15 (Raue et al., 2007) and rat Naa15 (Yamada and Bradshaw, 1991) are associated with poly- or monosomes. In yeast, NatA binds via the ribosomal proteins, uL23 and uL29 (Polevoda et al., 2008). Further data indicates that NatA preferably associates with translating ribosomes. Particularly, yNatA as well as other ribosome-associated protein biogenesis factors (including the chaperones Ssb1/2 and ribosome-associated complex, signal recognition particle and the aminopeptidases Map1 and Map2) bind with increased apparent affinity to randomly translating ribosomes as compared with non-translating ones (Raue et al., 2007). Hsp70 chaperones may be direct targets of NatA, and NTA by NatA contributes an unanticipated influence on protein biogenesis, both through and independent of Hsp70 activity (Holmes et al., 2014), supporting a role of NatA in protein biogenesis. However, the NatA complex also exists in a ribosome-free context. For instance it has been shown that the majority of hNatA is non-polysomal (Arnesen et al., 2005a) and a minor fraction of cytosolic hNaa10 exists independent of the NatA complex, carrying out post-translational acetylation as mentioned above (Van Damme et al., 2011b). Mammalian Naa10, Naa11 and Naa15 and Naa50 have been reported to be mainly localized in the cytoplasm and to a lesser extent to the nucleus (Fluge et al., 2002, Sugiura et al., 2003, Bilton et al., 2005, Arnesen et al., 2006a, Arnesen et al., 2006b, Chun et al., 2007, Xu et al., 2012, Park et al., 2014, Zeng et al., 2014, Aksnes et al., 2015c). In mouse, an isoform specific localization of Naa10 has been described. mNaa10235 was mainly nuclear in NIH-3T3 and JB6 cells whereas another variant mNaa10225, derived from alternative splicing at a different 3′-splice site, was mainly localized in the cytoplasm (Chun et al., 2007). In humans, Naa10225 is absent and Naa10235 was found to be evenly distributed in both cytoplasm and nucleus of HeLa and HT1080 cells as seen by immunofluorescence, confocal microscopy, and cell fractionation (Chun et al., 2007). Naa10 could be detected in nuclear fractions of doxorubicin treated HEK293 cells whereas a deletion construct lacking amino acids 1–35 could not be detected suggesting that a nuclear localization signal (NLS) resides in the N-terminal part of Naa10 (Park et al., 2012). Sequence analysis had previously identified a putative NLS more C-terminally in Naa10 between residues 78 and 83 (KRSHRR) (Arnesen et al., 2005a). In agreement with that, deletion of this NLS78–83 almost completely abrogated nuclear localization of Naa10, whereas Naa10 wild type was imported to the nuclei of proliferating HeLa and HEK293 cells, especially during S phase (Park et al., 2014). Furthermore, the deletion of NLS78–83 altered the cell cycle and the expression levels of cell cycle regulators and resulted in cell morphology changes and cellular growth impairment, all of which was mostly rescued when the nuclear import of hARD1 was restored by exogenous NLS (Park et al., 2014). Also, Arnesen et al. reported that neither leptomycin B nor actinomycin D significantly changed the localization patterns of Naa10 in HeLa cells, indicating that Naa10 is not actively imported through importin β-dependent mechanisms (Arnesen et al., 2005a).

In conclusion, it is well established that Naa10 localizes to the cytoplasm and the nucleus and the possible shuttling between the two compartments might be regulated by an internal NLS. Although the action of cytosolic Naa10 is well described in co-translational acetylation, the significance of its nuclear localization is still not well characterized; however, some studies connect Naa10 to nuclear processes such as transcriptional regulation (see below). Also, the signal pathways that might induce or regulate cytosolic to nuclear translocation are unknown.

Naa15 also harbors a putative NLS between residues 612–628 (KKNAEKEKQQRNQKKKK); however, only Naa10 was found to be localized in the nuclei of HeLa, GaMg, HEK293, MCF-7 and NB4 cells, whereas Naa15 was predominantly localized in the cytoplasm (Arnesen et al., 2005a). In contrast to this, a different study showed that Naa15 localizes to the nucleus where it interacts with the osteocalcin promoter, as shown by cellular fractionation and ChIP experiments in MC3T3E1 calvarial osteoblasts (Willis et al., 2002). Further studies have to resolve these discrepancies and analyze possible cell-type specific differences.

Besides that, it has been shown that Naa10 associates with microtubules in dendrites in cultured neurons (Ohkawa et al., 2008) and Naa15 colocalizes with the actin-binding protein cortactin and the F-actin cytoskeleton in the cytoplasm of IEM mouse and RF/6A rhesus endothelial cells (Paradis et al., 2008).

Section snippets

Naa10 function in mammals

The Nα-acetyltransferase NatA is expressed widely in many tissues and NatA N-termini are overrepresented in eukaryotic proteomes. As pointed out earlier, 80–90% of soluble human proteins are fully or partially acetylated and nearly 40–50% of all proteins are potential NatA substrates according to their sequence in S. cerevisiae, D. melanogaster and humans (Van Damme et al., 2011c, Starheim et al., 2012). Due to its ubiquitous expression in almost all tissues and the broad substrate specificity

NatA in other organisms

Since its first identification in a yeast screen (Whiteway and Szostak, 1985) homologues for Naa10 have been found in a variety of other organisms. In an early review, Polevoda and Sherman used the general BLAST server from the National Center for Biotechnology Information (NCBI) to identify orthologs of yeast NATs. They found candidates for Naa10 in S. pombe, C. elegans, D. melanogaster, A. thaliana, T. brucei, Dictyostelium discoideum, Mus musculus and H. sapiens, and orthologs for Naa50 were

Open questions

NTA is one of the most abundant protein modifications known, and the NAT machinery is very well conserved throughout all Eukarya. Over the past 50 years, the function of NTA has begun to be slowly elucidated, and this includes the modulation of protein–protein interaction, protein-stability, protein function and protein targeting to specific cellular compartments. Many of these functions have been studied in the context of Naa10/NatA; however, we are only starting to really understand the full

Competing interests

G.J.L. serves on the medical advisory board of GenePeeks, Inc. and the scientific advisory board of Omicia, Inc. The study did not involve these companies and did not use products from these companies.

Acknowledgments

This review and the corresponding Gene Wiki article are written as part of the Gene Wiki Review series — a series resulting from a collaboration between the journal GENE and the Gene Wiki Initiative. The Gene Wiki Initiative is supported by the National Institutes of Health (GM089820). Additional support for Gene Wiki Reviews is provided by Elsevier, the publisher of GENE. The laboratory of G.J.L. is supported by funds from the Stanley Institute for Cognitive Genomics at Cold Spring Harbor

References (319)

  • R. Bilton et al.

    Arrest-defective-1 protein, an acetyltransferase, does not alter stability of hypoxia-inducible factor (HIF)-1alpha and is not induced by hypoxia or HIF

    J. Biol. Chem.

    (2005)
  • S. Bonissone et al.

    N-terminal protein processing: a comparative proteogenomic analysis

    Mol. Cell. Proteomics

    (2013)
  • M.C. Brahimi-Horn et al.

    Hypoxia signalling controls metabolic demand

    Curr. Opin. Cell Biol.

    (2007)
  • R. Caesar et al.

    The stress-induced Tfs1p requires NatB-mediated acetylation to inhibit carboxypeptidase Y and to regulate the protein kinase A pathway

    J. Biol. Chem.

    (2004)
  • H.H. Chang et al.

    N-terminal processing of proteins exported by malaria parasites

    Mol. Biochem. Parasitol.

    (2008)
  • E. Charbaut et al.

    N-terminal acetylation of ectopic recombinant proteins in Escherichia coli

    FEBS Lett.

    (2002)
  • X. Chen et al.

    N-acetylation and ubiquitin-independent proteasomal degradation of p21(Cip1)

    Mol. Cell

    (2004)
  • K.H. Chun et al.

    Differential regulation of splicing, localization and stability of mammalian ARD1235 and ARD1225 isoforms

    Biochem. Biophys. Res. Commun.

    (2007)
  • A. Ciechanover et al.

    N-terminal ubiquitination: more protein substrates join in

    Trends Cell Biol.

    (2004)
  • H. Clevers et al.

    Wnt/β-catenin signaling and disease

    Cell

    (2012)
  • A.G. Cumberlidge et al.

    Ribosomal protein modification in Escherichia coli. I. A mutant lacking the N-terminal acetylation of protein S5 exhibits thermosensitivity

    J. Mol. Biol.

    (1979)
  • E.C. de Haan et al.

    Stabilization of peptide guinea pig myelin basic protein 72–85 by N-terminal acetylation-implications for immunological studies

    Mol. Immunol.

    (2004)
  • R. Evjenth et al.

    Human Naa50p (Nat5/San) displays both protein N alpha- and N epsilon-acetyltransferase activity

    J. Biol. Chem.

    (2009)
  • R.H. Evjenth et al.

    Human protein N-terminal acetyltransferase hNaa50p (hNAT5/hSAN) follows ordered sequential catalytic mechanism: combined kinetic and NMR study

    J. Biol. Chem.

    (2012)
  • M. Falb et al.

    Archaeal N-terminal protein maturation commonly involves N-terminal acetylation: a large-scale proteomics survey

    J. Mol. Biol.

    (2006)
  • D.M. Fath et al.

    Histone deacetylase inhibitors repress the transactivation potential of hypoxia-inducible factors independently of direct acetylation of HIF-alpha

    J. Biol. Chem.

    (2006)
  • B. Fauvet et al.

    Characterization of semisynthetic and naturally Nα-acetylated α-synuclein in vitro and in intact cells: implications for aggregation and cellular properties of α-synuclein

    J. Biol. Chem.

    (2012)
  • B. Filner et al.

    TMV coat protein synthesis in vivo: analysis of the N-terminal acetylation

    Virology

    (1974)
  • T.S. Fisher et al.

    Analysis of ARD1 function in hypoxia response using retroviral RNA interference

    J. Biol. Chem.

    (2005)
  • H. Foyn et al.

    Protein N-terminal acetyltransferases act as N-terminal propionyltransferases in vitro and in vivo

    Mol. Cell. Proteomics

    (2013)
  • H. Geng et al.

    HIF1α protein stability is increased by acetylation at lysine 709

    J. Biol. Chem.

    (2012)
  • Y. Gordiyenko et al.

    Acetylation of L12 increases interactions in the Escherichia coli ribosomal stalk complex

    J. Mol. Biol.

    (2008)
  • M.G. Grütter et al.

    Crystals of the complex between recombinant N-acetyleglin c and subtilisin. A preliminary characterization

    J. Biol. Chem.

    (1985)
  • V. Adam et al.

    Synergistic and selective cancer cell killing mediated by the oncolytic adenoviral mutant AdΔΔ and dietary phytochemicals in prostate cancer models

    Hum. Gene Ther.

    (2012)
  • M. Aivaliotis et al.

    Large-scale identification of N-terminal peptides in the halophilic archaea Halobacterium salinarum and Natronomonas pharaonis

    J. Proteome Res.

    (2007)
  • D. Akopian et al.

    Signal recognition particle: an essential protein-targeting machine

    Annu. Rev. Biochem.

    (2013)
  • H. Aksnes et al.

    N-terminal acetylation by NatC is not a general determinant for substrate subcellular localization in Saccharomyces cerevisiae

    PLoS One

    (2013)
  • T.R. Alderson et al.

    Biophysical characterization of α-synuclein and its controversial structure

    Intrinsically Disord. Proteins

    (2013)
  • C. Ansong et al.

    Top–down proteomics reveals a unique protein S-thiolation switch in Salmonella Typhimurium in response to infection-like conditions

    Proc. Natl. Acad. Sci. U. S. A.

    (2013)
  • A.D. Aragon et al.

    Characterization of differentiated quiescent and nonquiescent cells in yeast stationary-phase cultures

    Mol. Biol. Cell

    (2008)
  • C.S. Arendt et al.

    Eukaryotic 20S proteasome catalytic subunit propeptides prevent active site inactivation by N-terminal acetylation and promote particle assembly

    EMBO J.

    (1999)
  • K.J. Armache et al.

    Structural basis of silencing: Sir3 BAH domain in complex with a nucleosome at 3.0 Å resolution

    Science

    (2011)
  • N. Arnaudo et al.

    The N-terminal acetylation of Sir3 stabilizes its binding to the nucleosome core particle

    Nat. Struct. Mol. Biol.

    (2013)
  • T. Arnesen

    Towards a functional understanding of protein N-terminal acetylation

    PLoS Biol.

    (2011)
  • T. Arnesen et al.

    Identification and characterization of the human ARD1-NATH protein acetyltransferase complex

    Biochem. J.

    (2005)
  • T. Arnesen et al.

    Characterization of hARD2, a processed hARD1 gene duplicate, encoding a human protein N-alpha-acetyltransferase

    BMC Biochem.

    (2006)
  • T. Arnesen et al.

    Induction of apoptosis in human cells by RNAi-mediated knockdown of hARD1 and NATH, components of the protein N-alpha-acetyltransferase complex

    Oncogene

    (2006)
  • T. Arnesen et al.

    A novel human NatA Nalpha-terminal acetyltransferase complex: hNaa16p–hNaa10p (hNat2–hArd1)

    BMC Biochem.

    (2009)
  • T. Arnesen et al.

    Proteomics analyses reveal the evolutionary conservation and divergence of N-terminal acetyltransferases from yeast and humans

    Proc. Natl. Acad. Sci. U. S. A.

    (2009)
  • T. Arnesen et al.

    The chaperone-like protein HYPK acts together with NatA in cotranslational N-terminal acetylation and prevention of Huntingtin aggregation

    Mol. Cell. Biol.

    (2010)
  • Cited by (0)

    View full text