Communicated by: Akira Ishihama
Tip60 acetylates six lysines of a specific class in core histones in vitro
Article first published online: 4 JAN 2002
Blackwell Science Ltd, Oxford
Genes to Cells
Volume 3, Issue 12, pages 789–800, December 1998
How to Cite
Kimura, A. and Horikoshi, M. (1998), Tip60 acetylates six lysines of a specific class in core histones in vitro. Genes to Cells, 3: 789–800. doi: 10.1046/j.1365-2443.1998.00229.x
- Issue published online: 3 MAR 2003
- Article first published online: 4 JAN 2002
Background: Tip60, an HIV-1-Tat interactive protein, is a nuclear histone acetyltransferase (HAT) with unique histone substrate specificity. Since the acetylation of core histones at particular lysines mediates distinct effects on chromatin assembly and gene regulation, the identification of lysine site specificity of the HAT activity of Tip60 is an initial step in the analysis of its molecular function.
Results: Tip60 significantly acetylates amino-terminal tail peptides of histones H2A, H3 and H4, but not H2B, consistent with substrate preference on intact histones. Preferred acetylation sites for Tip60 are the Lys-5 of histone H2A, the Lys-14 of histone H3, and the Lys-5, -8, -12, -16 of histone H4, determined by a method which combined matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) measurements and Lys-C endopeptidase digestion, or a method detecting the incorporation of radiolabelled acetate into synthetic peptides.
Conclusion: The lysine site specificity of Tip60 in histone amino-terminal tail peptides in vitro has been characterized by an assay measuring the molecular mass of endopeptidase digested peptides, or a previously described assay. These results agree well with our proposed classification of lysines in core histones. The classification may be useful for an analysis of the relationships between HATs and the substrates of other uncharacterized HATs.
Histone acetyltransferases (HATs) and histone deacetylases (HDACs) are enzymes responsible for the modification of specific lysines in core histones, and are thought to be key modifiers of chromatin structures (for reviews see Grunstein 1997; Wade et al. 1997). After the discovery of histone acetylation (Allfrey et al. 1964), the recent identification of various HAT and HDAC molecules represented a significant breakthrough in our understanding of the functional role of acetylation and deacetylation (Kleff et al. 1995; Bannister & Kouzarides 1996; Brownell et al. 1996; Mizzen et al. 1996; Ogryzko et al. 1996; Rundlett et al. 1996; Taunton et al. 1996; Yang et al. 1996; Chen et al. 1997; Lusser et al. 1997; Spencer et al. 1997; Yamamoto & Horikoshi 1997; Smith et al. 1998). The finding that several transcription factors and coactivators possess HAT activity provides a molecular basis for examining relationships between HAT and transcriptional activation (for example see Brownell et al. 1996). Non-histone proteins are also targets of a number of these acetyltransferases (Gu & Roeder 1997; Imhof et al. 1997). The biological significance of HAT activity was based on the observation that a PCAF mutant lacking HAT activity failed to promote both MyoD-directed transcription and myogenic differentiation in cultured cells (Puri et al. 1997). From studies on Saccharomyces cerevisiae, the acetylation of specific lysines by HATs was shown to function in vivo, based on the following observations: (i) Gcn5p mutants lacking HAT activity abolish both promoter-directed histone acetylation and Gcn5p mediated transcriptional activation (Kuo et al. 1998), and (ii) Gcn5p mutation, together with mutations of specific lysines within histones, results in a synergistic growth defect (Zhang et al. 1998).
Tip60, which was originally isolated as an HIV-1-Tat interactive protein, was shown to modestly activate Tat-dependent transcription (Kamine et al. 1996). Tip60, which was found to localize in the nucleus (Yamamoto & Horikoshi 1997) contains an evolutionarily conserved region named the MYST domain (about 250 amino acids in length; 40–50% identity) which is found in MOZ (Borrow et al. 1996), mof (Hilfiker et al. 1997), YBF2/SAS3 (Reifsnyder et al. 1996), SAS2 (Reifsnyder et al. 1996; Ehrenhofer-Murray et al. 1997), and Tip60. This domain contains a structural motif of 20 amino acids in length which is thought to be an acetyl-CoA binding site (Lu et al. 1996), however, the primary structure of the MYST domain, besides this short structural motif, is not related to known HAT. We tested this evolutionarily conserved MYST domain of Tip60 (Tip60C fragment) for HAT activity and found that it possesses an activity that acetylates histones H2A, H3 and H4, but not histone H2B of core histone mixtures (Yamamoto & Horikoshi 1997). This histone preference had not been observed among previously analysed HATs.
Specific lysines in amino-terminal tails of the four core histones are acetylated in various patterns in vivo. For example, Lys-12 of histone H4 is selectively acetylated in both Drosophila heterochromatin (Turner et al. 1992) and Saccharomyces cerevisiae silent mating type loci (Braunstein et al. 1996), while the Lys-16 of histone H4 is acetylated in the Drosophila hyperactive male X-chromosome (Turner et al. 1992). In accordance with the remarkably specific use of acetylation sites within core histones in vivo, isolated HATs vary in the lysine site specificity in vitro (Kuo et al. 1996; Mizzen et al. 1996; Ogryzko et al. 1996; Parthun et al. 1996; Spencer et al. 1997; Smith et al. 1998). Thus, a mechanistic understanding of functions of Tip60 requires the identification of lysine site specificity of the HAT activity of Tip60.
Taking into consideration that the utilization of specific acetylation sites in the core histones closely correlates with distinct biological processes, native pairs of HATs and substrate lysines in vivo warrant attention. To understand the mechanisms involved in specific site selection with multiple regulation occurring in complex systems, it is useful to manifest the rules governing the potential site specificity of HATs in a minimal in vitro system with the least number of components, such as HAT domains and free histones. We proposed a novel classification of lysine residues based on a comprehensive analysis of all lysines residing in the amino-terminal tail of core histones (Kimura & Horikoshi 1998; see also Table 3). When this classification was compared with the lysine site specificity of catalytic domains of HATs in vitro, all the lysines of a given group were acetylated by the same catalytic domains of HATs. According to this classification, it may be possible to predict the lysine site specificity of other HATs. However, since HATs with characterized site specificity are limited in number, the site specificity of other HATs needs to be analysed in order to investigate the relationships between the classification of lysines and the lysine site specificity of HATs.
An identification of the lysine site specificity of Tip60 will contribute not only to an understanding of functional roles but also to test the applicability of the classification of lysines. We report here an identification of lysine site specificity of Tip60 in histone amino-terminal tail peptides in vitro.
Tip60 preferentially acetylates amino-terminal tail peptides of histones H2A, H3 and H4, but not histone H2B
To characterize the enzymatic activity of Tip60, we examined the lysine site specificity of Tip60 in vitro. When determining the lysine site specificity of HATs, synthetic peptides corresponding to the amino-terminal tails of core histones are mostly used as substrates because all lysines acetylated in vivo reside in these domains (Mizzen et al. 1996; Ogryzko et al. 1996; Parthun et al. 1996; Spencer et al. 1997). To determine if the amino-terminal histone tails themselves are sufficient for specific lysine selection by Tip60, we asked if the unique substrate specificity of Tip60 that has been observed with intact histones would be reproduced when synthetic peptides corresponding to the amino-terminal tails of core histones are used as substrates. Purified recombinant Tip60C (Yamamoto & Horikoshi 1997) was incubated with [3H]-acetyl-CoA and the synthetic peptides. The activity of acetylation was measured as 3H radioactivity incorporated into peptides retained on P81 filter paper. The incorporated activities indicate that acetylation by Tip60C is significant for histones H2A, H3 and H4 amino-terminal peptides, whereas no activity is detected for histone H2B amino-terminal peptide (Fig. 1), which results are consistent on histone substrate specificity for intact histones. Based on these results, the amino-terminal tails of histones are considered to be sufficient to determine the specificity of substrate selection.
Lys-5 of histone H2A is the preferred site for Tip60 acetylation in vitro
To identify specific lysine residues acetylated by Tip60, we utilized matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) analysis (Karas & Bahr 1991) on molecular mass, in combination with the specific enzymatic degradation of peptides. After acetylating the histone amino-terminal peptides with Tip60C, the peptides were digested with Lys-C endopeptidase which selectively cleaves at the carboxy-terminal of unacetylated lysine residues. The acetylation of specific lysine(s) in the peptide not only increases the molecular mass (42-Da mass; -H to -COCH3) but also makes the site(s) resistant to Lys-C endopeptidase cleavage (data not shown). Therefore, the acetylation of specific lysine(s) can be identified from the size of the digested peptide fragments by MALDI-MS.
Table 1 lists all the possible Lys-C digested fragments and their calculated m/z of acetylated amino-terminal peptide isoforms of histone H2A. Figure 2 shows the mass spectra of the Lys-C digestion mixture of the peptide acetylated by Tip60C, measured using two different spectrometers. A peak appearing at m/z= 959 is the only significant peak which corresponds to the possible fragment listed in Table 1. This significant peak represents an acetylated Lys-C fragment Ac-SGRGK(-Ac)QGGK, suggesting that the Lys-5 of histone H2A is the preferred acetylation site for Tip60 in vitro. Three small peaks appeared at m/z= 687, 857 and 1428 which could be signals corresponding to the sizes of the acetylated fragments (see Table 1), while the other peaks did not agree with size of the possible Lys-C digested fragments. The small peaks at m/z= 857 and 1427 corresponding to the fragment size, suggesting that acetylation occurs at Lys-9. The peak at m/z= 687 appears nonspecifically when peptides corresponding to histone H3, or histone H4 amino-terminal tails are used as substrates, thus indicating that this peak is a background signal rather than a peak indicating the acetylation of Lys-13 of histone H2A (ARAK(-Ac)AK) (see Table 1). These background signals also appear when the samples without peptides and/or enzymes are analysed (data not shown), indicating that these signals are not caused by unexpected peptide cleavage but by chemical noise in the reaction buffer. The small peaks at m/z= 857 and 1427 corresponding to the fragment size suggest that acetylation occurs at Lys-9.
Mass spectrometric analysis of acetylation site(s) in histone H2A revealed that Lys-5 is the major acetylation site for Tip60, while Lys-9 can be marginally acetylated. Since this was the first time a mass spectrometric approach had been used to analyse the acetylation of histone tails, a more conventional filter binding assay was also performed. Preacetylation at the Lys-5 of the histone H2A tail peptide greatly (≈ 10-fold) reduced the incorporation of radioactivity, and Lys-5/-9 preacetylation reduced the incorporation completely to background levels in the filter binding assay (data not shown). This indicated that the results of the mass spectrometric assay are consistent with those of the conventional assay.
Since the Lys-5 of histone H2A belongs to group A in our classification, it is predicted that the Lys-5 and Lys-12 of histone H4 (see Table 3), the other sites of group A, are also preferred acetylation sites for Tip60.
Lys-14 of histone H3 is the preferred site for Tip60 acetylation in vitro
Lysines in the amino-terminal tail of histone H3 belong to groups B, D, E and F in the classification (see Table 3). An identification of acetylation site(s) preferred by Tip60 in histone H3 will reveal whether the lysine specificity of Tip60 is restricted to one of the four groups. Table 2 lists all the possible Lys-C digested fragments and their calculated m/z of acetylated amino-terminal peptide isoforms of histone H3. Figure 3 shows the mass spectra of the Lys-C digestion mixture of histone H3 amino-terminal peptide acetylated by Tip60C, as measured using two different spectrometers. As the significant peak at m/z= 943 represents an acetylated Lys-C fragment STGGK (-Ac)APRK it is suggested that the Lys-14 of histone H3 is the preferred acetylation site for Tip60 in vitro. The small peak at m/z= 1101 suggests that acetylation might occur at Lys-4 (Fig. 3A). Since the signal was weak and could not be detected when a different spectrometer was used (Fig. 3B), we considered that Lys-4 is not a preferred site for acetylation by Tip60.
An analysis of acetylation site(s) in histone H3 revealed that Lys-14 is the preferred acetylation site for Tip60. Since this site belongs to group B in the classification, other sites of group B, the Lys-8 and Lys-16 of histone H4, can be predicted to be acetylated by Tip60 (see Table 3). This finding also indicates that Tip60 does not prefer the lysines of groups D, E and F. This means that Tip60 seems to prefer specific lysines of class I, consisting of groups A and B, but not lysines of class II (groups C and D) or class III (groups E and F).
Tip60 acetylates Lys-5, -8, -12 and -16 of histone H4
The preferred acetylation sites for Tip60 in histones H2A and H3 belong to groups A and B in our classification. In histone H4, four sites (Lys-5, -8, -12, and -16) belong to groups A and B, and are predicted to be acetylated by Tip60. When these four lysines are acetylated independently, 16 isoforms with different acetylation patterns appear. In this case, the population of each peptide may be too sparse to be detected using the same assay conditions employed for histones H2A and H3. In actual fact, there are no significant peaks for either acetylated or unacetylated fragments in the mass spectra of Lys-C digestion mixture of the peptides acetylated by Tip60C (data not shown).
To identify specific acetylation sites for Tip60 in histone H4, we used a different method, as described in previous studies (Mizzen et al. 1996; Ogryzko et al. 1996; Spencer et al. 1997). We prepared synthetic peptides corresponding to 16 isoforms of histone H4 amino-terminal peptides with four lysines acetylated in various patterns. The peptide sequence and positions of acetate groups incorporated during synthesis are shown at the bottom of Fig. 4. When we investigated whether Tip60 would incorporate radiolabelled acetate groups into the unmodified lysines of the peptides, we found that the incorporation was independent of positions of preacetylated lysine(s), thereby indicating that all four lysines can be acetylated in vitro. The preacetylation of any single lysine in the peptides decreases the incorporation of radioactivity. For example, when the incorporation activities examined for the Lys-5 preacetylated peptide and the Lys-5/-8 di-acetylated peptide are compared, the incorporation of radioactivity to the di-acetylated peptide is decreased according to the preacetylation of Lys-8, which suggests that Lys-8 is a potential acetylation site for Tip60.
Therefore, it seems obvious that the Lys-5, -8, -12, -16 of histone H4 can be acetylated almost equally by Tip60. These lysines belong to group A or group B in the classification.
We identified the lysine site specificity of Tip60 in vitro using histone amino-terminal tails as substrates. For this we used a method which utilizes Lys-C digestion and MALDI-MS. The specific acetylation sites of Tip60 are the Lys-5 of histone H2A, the Lys-14 of histone H3, and the Lys-5, -8, -12, -16 of histone H4, based on our analysis. As shown in Table 3, all of the preferred acetylation sites identified in this study fall into a specific class of lysines which we classified in our earlier study (Kimura & Horikoshi 1998).
Identification of acetylation sites using mass spectrometry
We applied the MALDI-MS technique to identify acetylation sites in histone tails. Two methods have been used in previous studies to determine the lysine site specificity of HATs. One involves the detection of radiolabelled acetyl-groups at positions of acetylated residues released during N-terminal peptide sequencing. This approach was used to determine the lysine site specificity of yGcn5p (Kuo et al. 1996), yHat1p (Parthun et al. 1996) and yEsa1p (Smith et al. 1998). With this method, intact histones can easily be handled as substrates, but radioactive isotope labelling and deblocking of the amino-terminal are required. The other method involves the detection of radiolabelled acetyl-CoA incorporation into various synthetic peptides with specific lysines premodified in various patterns during peptide synthesis. This method was used to determine the lysine site specificity of p300/CBP (Ogryzko et al. 1996), dTAF230 (Mizzen et al. 1996) and SRC-1 (Spencer et al. 1997). While it is convenient, this method requires various isoforms of synthetic peptides.
MALDI-MS is a sensitive technique for measuring the molecular mass of peptides and proteins, and for detecting qualitative changes in molecules from modifications such as phosphorylation (Karas & Bahr 1991; Yip & Hutchens 1992). The method we used to locate acetylation sites takes advantage of the speed and mixture analysis capability of MALDI-MS. The change that is due to the acetylation of specific lysine(s) can be detected accurately by measuring the molecular mass of the Lys-C digested peptide fragments using MALDI-MS. The major advantages in this method, compared to the previous two methods, are the absence of requirements for: (i) radioactive isotope labelling, (ii) deblocking of amino-terminal, or (iii) preparing various synthetic acetylated peptides. However, if many acetylation sites reside in one peptide, as is the case for
histone H4, this approach is not appropriate because it may be difficult to detect a low concentration of each single isoform. We used the second method described above to analyse the site usage of Tip60 for histone H4.
Histone amino-terminal tails as substrates
To identify the lysine site specificity of HATs, free intact histones or amino-terminal peptides were used in previous studies (for example see Kuo et al. 1996; Ogryzko et al. 1996). The histone substrate specificity of Tip60 that has been observed with the intact histones mixture is also observed when synthetic peptides corresponding to amino-terminal histone tails are used as substrates (Fig. 1). This coincidence of substrate preference in the two forms of substrates was observed for other HATs such as dTAF230, hGCN5 (Mizzen et al. 1996) and hSRC-1 (Spencer et al. 1997), suggesting that the specificity of substrate selection by HATs is primarily determined by the amino-terminal histone tails alone, without taking carboxy-terminal core domains into consideration. Thus, we consider that the lysine site specificity of HATs does not change when peptides corresponding to amino-terminal histone tails or intact free histones served as substrates. Therefore amino-terminal tail peptides seem to serve as appropriate substrates for identifying the potential lysine site specificity of HATs in free histones.
Site specificity of Tip60 and the proposed classification
Acetylation-site utilization analysis revealed that the Lys-5 of histone H2A, the Lys-14 of histone H3 and the Lys-5, -8, -12 and -16 of histone H4 are highly preferred acetylation sites for Tip60 in vitro. Lys-9 of histone H2A, another residue that can be classified into group B based on its flanking sequence (Kimura & Horikoshi 1998), is also shown in our results to be weakly acetylated. Although this site has not been reported to be acetylated in human cells (Marvin et al. 1990; Thorne et al. 1990), the Lys-9 of histone H2A may perhaps be acetylated in human cells under certain conditions, as well as in other organisms (Doenecke & Gallwitz 1982).
All of the preferred acetylation sites of Tip60 belong to class I (groups A and B) in our classification (Table 3). In addition, all lysines which belong to class I are acetylated by Tip60. Furthermore, none of the lysines which belong to class II (groups C and D) and class III (groups E and F) are preferred by Tip60. This indicates that Tip60 discriminates the lysines of groups A and B (class I) from those of groups C, D, E and F (classes II and III), and our hypothesis that HATs can distinguish among lysines of the proposed groups is thus given support.
Tip60 is the first member, among a group of proteins containing the MYST domain, that has been shown to possess HAT activity (Yamamoto & Horikoshi 1997). Saccharomyces cerevisiae Esa1p, another protein containing the MYST domain, also possesses HAT activity (Smith et al. 1998). The lysine site specificity of Esa1p was determined by the N-terminal peptide sequencing method using intact histones as substrates. This lysine site specificity is similar to that of Tip60 determined in our study by the different methods using peptides as substrates. This result is the first evidence that human Tip60 and Saccharomyces cerevisiae Esa1p are functional homologues with respect not only to their evolutionarily conserved primary structures but also to their enzymatic properties. Factors containing MYST domains may have HAT activities with similar types of lysine site specificity as Tip60 and Esa1p.
Classification of HATs based on lysine site specificity
Tip60 is a HAT with ‘class-specific’ site specificity which acetylates lysines of two groups (A and B) belonging to class I. Saccharomyces cerevisiae Esa1p (Smith et al. 1998) can be considered another ‘class I-specific’ HAT, based on our classification. The presence of ‘class I-specific’ HATs suggests the existence of class II- or class III-specific HATs. Hat1p, which acetylates lysines of group A, or Gcn5p which acetylates lysines of group B, can be defined as ‘group-specific’ HATs that acetylate lysines of a single group. p300/CBP can acetylate histone H2B (Ogryzko et al. 1996) which contains lysines of class II or class III. This indicates that p300/CBP can acetylate lysines not only of class I but also of class II and/or class III. From this broader spectrum of acetylation site specificity, we can categorise p300/CBP as HATs with ‘interclass’ type specificity. Based on these observations, the lysine site specificity of HATs may be categorized to ‘group-specific’, ‘class-specific’ and ‘interclass’ types.
There are no apparent ‘consensus sequences’ for substrate recognition by class I-specific HAT-like Tip60. Since a characteristic of the ‘class I’ lysines is the location of G/A residues on the amino-terminal side of the lysines, these G/A residues are probably essential for recognition by ‘class I-specific’ HATs. However, this is insufficient for recognition because there are many -(G/A)K-sequences which are not acetylated by ‘class I-specific’ HATs. Taking into consideration that ‘class I’ lysines consist solely of ‘group A’ lysines and ‘group B’ lysines, the substrate recognition surfaces of ‘class I-specific’ HATs may consists of two distinct surfaces recognizing the consensus sequence of group A and that of group B, or one surface that can recognize both sequences. This speculation, based on the proposed classification, can explain the specific lysine selection by HATs without an apparent tight target consensus sequence.
In this report, we analysed the lysine site specificity of Tip60 in minimal sets of enzymes and substrates (i.e. HAT domains and amino-terminal histone tails), which will be an initial step for the elucidation of the molecular basis of mechanisms involved in specific site selection in more complex states. Lysine site selection in vivo is thought to be influenced by factors other than minimal catalytic domains of HATs and histone amino-terminal tails. The formation of nucleosome structure has inhibitory effects on the acetylation of histones by HATs (for example see Grant et al. 1997). Other nonhistone chromatin components may also affect the accessibility of HATs to substrate lysines. Some HATs have been shown to form multiprotein complexes (Parthun et al. 1996; Grant et al. 1997, 1998; Ogryzko et al. 1998) and modulate the site specificity of the catalytic domains of HATs (Parthun et al. 1996; Grant et al. 1997). For better a understanding of the specificity between the HATs and lysines, it will be necessary to: (i) identify the HATs with novel lysine specificity, (ii) analyse the lysine site specificity in a more complex system (e.g. HAT complexes and nucleosomes as enzymes and substrates) and (iii) determine the tertiary structure of histone tails and substrate recognition surface of HATs.
To address the relationships between HAT activities and the biological processes governed by Tip60, it is important to identify associating regulators which influence the site preference of Tip60, or associating DNA/RNA binding proteins which recruit or dissociate HATs to specific loci. HIV-1-Tat, an RNA binding protein, may be a good candidate for such proteins (Kamine et al. 1996). However, since Tat is a viral protein, other cellular proteins that recruit or dissociate Tip60 to specific loci may be involved. The present study of the lysine site specificity of Tip60 represents an initial stage of investigation of the functional roles of Tip60.
Synthesis of histone amino-terminal peptides
Histone amino-terminal peptides were synthesized, both with and without acetyl groups, on the ε-amino groups of specific lysines by Sawady Technology Co., Ltd. The α-amino group of the amino-terminal serine residues of peptides corresponding to histones H2A and H4 were also acetylated during synthesis.
Histone acetyltransferase assay
Filter binding assays for HAT activity were done as described (Yamamoto & Horikoshi 1997), except that synthetic peptides (200 ng) were substituted for core histones.
The acetylation of peptides was done essentially as described above. Peptides (1 μg) were incubated for 20 min at 30 °C with Tip60C (200 ng), acetyl CoA (50 pmol, sodium salt, Sigma) in buffer containing 50 mm Tris-HCl (pH 8.0 at 25 °C), 1 mm dithiothreitol, 0.1 mm EDTA, 10% (v/v) glycerol in 25 μL. After the acetylation of the peptides, the pH was adjusted to 8.5 by adding 3 μL of 0.24 m Tris (final concentration to 60 mm), and 2 μL of 25 ng/μL Lys-C endopeptidase (Boehringer Mannheim). The mixture was incubated at 37 °C for 18 h. MALDI spectra were obtained on Voyager DE-STR (Perseptive) or REFLEX III (Bruker) time-of-flight mass spectrometers, equipped with nitrogen lasers (337 nm, 3 ns for Voyager DE-STR, 5 ns for REFLEX III). The accelerating voltages in the ion source were 20 kV for the Voyager DE-STR, and 19 kV for REFLEX III. Data were acquired with 0.5 ns resolution. α-cyano-4-hydroxycinnamic acid (Aldrich for Voyager DE-STR, Sigma for REFLEX III) was used as the MALDI matrix. The matrix was dissolved in a solution (acetonitrile [Nacalai Tesque: 0.1% aq trifluoroacetic acid [Nacalai Tesque] = 1:1 for Voyager DE-STR, acetonitrile [Wako]: 0.1% aq trifluoroacetic acid [Merck] = 1:2 for REFLEX III] to give 10 mg/mL. To prepare the sample, the solution containing the peptide sample was mixed with four volumes of matrix solution, and 1 μL of the mixture was applied to a sample plate. The mixture was then allowed to air-dry before being introduced into the mass spectrometer. Time-to-mass conversion was achieved by external calibration using a peak-for-matrix peak (α-cyano-4-hydroxycinnamic acid, [2M+H]+ at m/z= 379.09). For Voyager DE-STR, a peak for angiotensin I ([M+H]+ at m/z= 1269.69) was also used for external calibration. For REFLEX III, peaks for angiotensin II ([M+H]+ at m/z= 1046.54), substance P ([M+H]+ at m/z= 1347.74), and bombesin ([M+H]+ at m/z= 1619.82) were also used for external calibration.
We thank all the members of our laboratory for valuable discussion, especially Drs T. Yamamoto and T. Suzuki. We also thank M. Ohara for helpful comments. This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan, and the Exploratory Research for Advanced Technology (ERATO) of the Japan Science and Technology Corporation (JST), in addition to grants from the Asahi Glass Foundation, the Toray Science Foundation, and the New Energy and Industrial Technology Development Organization (NEDO).
- 131997) mof, a putative acetyl transferase gene related to the Tip60 and MOZ human genes and to the SAS genes of yeast, is required for dosage compensation in Drosophila. EMBO J. 16, 2054–2060.DOI: 10.1093/emboj/16.8.2054, , , (