Charting the N-Terminal Acetylome: A Comprehensive Map of Human NatA Substrates

N-terminal acetylation (Nt-acetylation) catalyzed by conserved N-terminal acetyltransferases or NATs embodies a modification with one of the highest stoichiometries reported for eukaryotic protein modifications to date. Comprising the catalytic N-alpha acetyltransferase (NAA) subunit NAA10 plus the ribosome anchoring regulatory subunit NAA15, NatA represents the major acetyltransferase complex with up to 50% of all mammalian proteins representing potential substrates. Largely in consequence of the essential nature of NatA and its high enzymatic activity, its experimentally confirmed mammalian substrate repertoire remained poorly charted. In this study, human NatA knockdown conditions achieving near complete depletion of NAA10 and NAA15 expression resulted in lowered Nt-acetylation of over 25% out of all putative NatA targets identified, representing an up to 10-fold increase in the reported number of substrate N-termini affected upon human NatA perturbation. Besides pointing to less efficient NatA substrates being prime targets, several putative NatE substrates were shown to be affected upon human NatA knockdown. Intriguingly, next to a lowered expression of ribosomal proteins and proteins constituting the eukaryotic 48S preinitiation complex, steady-state levels of protein N-termini additionally point to NatA Nt-acetylation deficiency directly impacting protein stability of knockdown affected targets.


Introduction
Co-and post-translational protein N-terminal acetylation (Nt-acetylation or NTA) exemplifies an omnipresent and abundant protein modification with one of the highest stoichiometries in mammalian cells reported for protein modifications to date [1,2]. Nonetheless, NTA contributes to an increase in mammalian N-terminal proteoform diversity by the occurrence of partially NTA protein N-termini [3]. NTA is carried out by a family of N-terminal acetyltransferases or NATs which belong to a subfamily of GNATs (GCN5 related N-acetyltransferase fold), with GNATs representing one of the largest protein superfamilies with over 10,000 members reported. To accomplish NTA, the acetyl moiety of acetyl-CoA (Ac-CoA) is being transferred to a protein's free N-terminus by the action of the NAT.
To date, 12 NATs-including 8 distinct eukaryotic NATs NatA to NatH next to 1 archaeal and 3 bacterial NATs-with evolutionary conserved folds and related catalytic mechanisms have been identified [4,5]. These NATs vary greatly in their (subunit) composition and substrate specificities. For the vast majority of identified NATs, structural and biochemical characterizations are at hand, thereby providing molecular insights into their mechanisms of action and modes of regulation besides revealing molecular determinants steering the by and large unique substrate-specific activity potential of NATs (reviewed in [4] and reported recently for NatC [6]). While only a partial redundancy in the specificity profiles of certain NATs could be observed, the targeted protein N-termini per NAT appear by and large unique, with only some exceptions as recently proven for the yeast Lge1 protein representing the first redundant yeast NatA/NatD substrate identified to date [7] studied mutation representative of additional pathogenic NAA10 variants is the Ser37Pro (S37P) mutation in the gene encoding NAA10 (i.e., NAA10 S37P). NAA10 S37P is characterized by reduced NatA catalytic activity and NTA due to impaired NatA complex formation [31][32][33]. Presenting symptoms of this and additional pathogenic NAA10 variants have now collectively been referred to as NAA10-related syndrome [34].
While non-essential in S. cerevisiae, NAA10 catalysed NTA has been reported as essential in a wide variety of higher eukaryotic species (e.g., Caenorhabditis elegans, Drosophila melanogaster, Danio rerio and Arabidopsis thaliana). Notwithstanding, strong effects such as slow growth, mating and decreased survival have been observed under stress conditions in yeast deletion strains [9]. In consideration it was unexpected that Naa10 KO mice were viable. Since functional redundancy was previously demonstrated for the NAA10 homolog NAA11 in mice and humans, intuitively NAA11 was hypothesized to account for the loss of Naa10 whilst not in accordance with its restricted expression profile observed [35]. A recent study by Kweon et al. however univocally demonstrates a compensatory role of the newly discovered and ubiquitously expressed paralogue Naa12 in mice in vivo, as in contrast to Naa10 KO mice, a Naa10 Naa12 double KO in mice turned out to be embryonically lethal [36].
By making use of CRISPR/Cas9 gene editing, NAA15 haplo-insufficient and deficient induced human pluripotent stem cells (iPSCs) could successfully be obtained. Interestingly, in these iPSCs, a mere partial effect on NatA-dependent NTA acetylation was observed [37]. In case of NAA15 deficiency and since a role for NAA16-NAA10 complexes has been proposed to act as a potential backup system for the NAA15-NAA10 NatA complex [38], akin NAA11 expression, expression of its homologue NAA16 in human cells may also (in part) be responsible for this observation.
Potential redundancy and high NTA efficiency overall have complicated studies aiming at capturing the NatA substrate repertoire in a comprehensive manner further enabling distinguishing efficient from less efficient NAT substrate N-termini.
In the current study, we report on improved human NatA knockdown conditions achieving a significant depletion of NAA10 and NAA15 expression. Using N-terminal proteomics, lowered NTA of at least 25% out of all putative NatA targets identified could be observed. Besides, several previously reported and putative novel NatE substrates were found and a lowered expression of ribosomal proteins and proteins implicated in translation initiation was observed. Finally, a significant impact on protein stability of knockdown affected NatA targets was noticed.
In sum, our study points to the need of a highly significant lowered functional expression of human NatA to comprehensively probe its Nt acetylome and its proteome-wide impact.

N-Terminal Proteomics to Explore the Human Nt-Acetylome upon Pronounced NatA Knockdown
We characterized the in-vivo substrate specificity of the human NatA complex by comparing the Nt-acetylomes of A-431 cells treated with siNAA10/siNAA15 pool (siNatA) versus siControl (siCTR) treated cells ( Figure 1A) by means of N-terminal proteomics. By preforming two consecutive (reverse) siRNA transfections in the presence of pan-caspase inhibition, a very efficient knockdown of more than 99% could be attained, as evidenced by the near absence of NAA10 and NAA15 expressed ( Figures 1B and S1). Since induction of apoptosis was previously reported in human NatA knockdown studies [39], apoptotic cell death was specifically blocked by Z-VAD-fmk.
When performing N-terminal proteomics in combination with differential L-Arg stable isotope-labeling by amino acids in cell culture (SILAC) [40] and in-vitro 13 C2D3-acetylation [41,42], the extent of NTA and relative quantification of N-terminal peptides can be determined [43,44]. In the setup analyzed, the siNatA and siCTR samples were labeled with a light ( 12 C6-L-Arg) and heavy isotopic variant of L-Arg ( 13 C6 15 N4-L-Arg), respectively ( Figure 1). A-431 cells cultivated in 13 C6 15 N4 L-arginine containing SILAC medium were transfected with 10 nM non-targeting siRNA (siCTR) and cells cultivated in 12 C6 L-arginine SILAC medium were transfected with 10 nM siNAA10/siNAA15 pool (siNatA). Knockdown was performed in the presence of the pan-caspase inhibitor Z-VAD-fmk (10 µM). Cells were transfected 24 hours after seeding, re-transfected under identical conditions 48 hours after the initial transfection and harvested 96 hours post initial siRNA transfection. (B) Immunoblots of A-431 cell lysates treated with non-targeting siRNA (siCTR) or siNAA10 and siNAA15 (siNatA). Blots were probed with anti-NAA10 and anti-NAA15 to assess levels of endogenous NAA10 (26 kDa) and NAA15 (101 kDa), thereby confirming highly efficient knockdown. The asterisks indicate non-specific bands with ** serving as loading control.
Following LC-MS/MS analysis, we identified 1498 unique N-termini originating from 1422 human proteins that were in compliance with the rules of NTA and NME [45,46], thereby representing proxies of translation initiation events (Table S1 and Materials and L-arginine containing SILAC medium were transfected with 10 nM non-targeting siRNA (siCTR) and cells cultivated in 12 C 6 L-arginine SILAC medium were transfected with 10 nM siNAA10/siNAA15 pool (siNatA). Knockdown was performed in the presence of the pan-caspase inhibitor Z-VAD-fmk (10 µM). Cells were transfected 24 h after seeding, re-transfected under identical conditions 48 h after the initial transfection and harvested 96 h post initial siRNA transfection. (B) Immunoblots of A-431 cell lysates treated with non-targeting siRNA (siCTR) or siNAA10 and siNAA15 (siNatA). Blots were probed with anti-NAA10 and anti-NAA15 to assess levels of endogenous NAA10 (26 kDa) and NAA15 (101 kDa), thereby confirming highly efficient knockdown. The asterisks indicate non-specific bands with ** serving as loading control.
In line with previous reports [21,27,37], and as evidenced from protein expression levels when probing for the NatA subunits by means of Western blotting, exclusive knockdown of NAA10 also results in a significant destabilization of NAA15 ( Figure S1).
When performing N-terminal proteomics in combination with differential L-Arg stable isotope-labeling by amino acids in cell culture (SILAC) [40] and in-vitro 13 C 2 D3acetylation [41,42], the extent of NTA and relative quantification of N-terminal peptides can be determined [43,44]. In the setup analyzed, the siNatA and siCTR samples were labeled with a light ( 12 C 6 -L-Arg) and heavy isotopic variant of L-Arg ( 13 C 6 15 N 4 -L-Arg), respectively ( Figure 1).
Following LC-MS/MS analysis, we identified 1498 unique N-termini originating from 1422 human proteins that were in compliance with the rules of NTA and NME [45,46], thereby representing proxies of translation initiation events (Table S1 and Materials and Methods). Of these, 1482 N-termini started at position 1 or 2, thus matching database annotated N-termini, whereas 16 started beyond position 2, the latter indicative of database non-annotated N-termini generated upon alternative translation initiation or translation of alternative (unknown) transcripts [47].
Of the 1306 identified N-termini carrying this modification (as inferred from MS or MS 2 data (Table S1)) and representing 87% of all identified N-termini (a percentage in line with previous reports [2,32]), 182 N-termini displayed a reduction in the degree of NTA upon hNatA knockdown ( Figure 2 and Table S2). Methods). Of these, 1482 N-termini started at position 1 or 2, thus matching database annotated N-termini, whereas 16 started beyond position 2, the latter indicative of database non-annotated N-termini generated upon alternative translation initiation or translation of alternative (unknown) transcripts [47]. Of the 1306 identified N-termini carrying this modification (as inferred from MS or MS 2 data (Table S1)) and representing 87% of all identified N-termini (a percentage in line with previous reports [2,32]), 182 N-termini displayed a reduction in the degree of NTA upon hNatA knockdown ( Figure 2 and Table S2). The effect of siNatA knockdown on the Nt-acetylome. The scatterplot displays the correlation of the determined degrees of NTA when comparing the siCTR (X-axis) and siNatA (Y-axis) N-terminome datasets (n=1404). The enlarged dots (siNatA affected = 1) represent N-termini displaying a significant reduction in the degree of NTA upon siNatA knockdown. The circles are colour coded according to their indicated NAT substrate specificity profiles (see also NAT type column in Tables S1,S2). siNatA knockdown causes a decrease in the levels of NTA as compared to the siCTR setup especially in case of NatA type N-termini (purple dots).
Expectedly, of these, the grant majority -173 or 95% -of affected N-termini displayed the NatA substrate specificity ( Figure 3, Tables 1,2 and Tables 1,2), meaning that more than 25% out of the putative NatA targets in the siCTR setup (i.e., 173 of 687 NTA NatA type N-termini) were susceptible to siNatA knockdown (Tables (S)1 and 2). Notably, in our original study only 7% of all putative hNatA substrates were affected by the reduced level of hNatA (i.e., 16 of 242) and this despite a significant, but somewhat lower, NatA knockdown efficiency [2]. Overall, this demonstrates the need for excessive NatA depletion or NatA perturbation to comprehensively probe for human NatA substrates in a cellular context.
MS spectra of representative siNatA affected NatA Ala-substrate N-termini are shown in Figure 4 and Figure S2 and among others, include the translational activator GCN1 ( 2 AADTQVSETLKR 13 ) as a representative and previously confirmed NatA substrate [2,32,37], besides two newly identified mitochondrial protein substrates; namely, alpha-ketoglutarate-dependent dioxygenase alkB homolog 7 and import inner membrane translocase subunit Tim8 B ( Figure 4, panels A, B and D). 74% of all siCTR NTA NatA type Figure 2. The effect of siNatA knockdown on the Nt-acetylome. The scatterplot displays the correlation of the determined degrees of NTA when comparing the siCTR (X-axis) and siNatA (Y-axis) N-terminome datasets (n = 1404). The enlarged dots (siNatA affected = 1) represent N-termini displaying a significant reduction in the degree of NTA upon siNatA knockdown. The circles are colour coded according to their indicated NAT substrate specificity profiles (see also NAT type column in Tables S1 and S2). siNatA knockdown causes a decrease in the levels of NTA as compared to the siCTR setup especially in case of NatA type N-termini (purple dots).
Expectedly, of these, the grant majority-173 or 95%-of affected N-termini displayed the NatA substrate specificity (Figure 3, Tables 1 and 2), meaning that more than 25% out of the putative NatA targets in the siCTR setup (i.e., 173 of 687 NTA NatA type N-termini) were susceptible to siNatA knockdown (Tables 1 and 2). Notably, in our original study only 7% of all putative hNatA substrates were affected by the reduced level of hNatA (i.e., 16 of 242) and this despite a significant, but somewhat lower, NatA knockdown efficiency [2]. Overall, this demonstrates the need for excessive NatA depletion or NatA perturbation to comprehensively probe for human NatA substrates in a cellular context.   Table 2 and Tables S1,S2). Mean values are shown as '+'. Table 2. Overview of the Nt-acetylation status of all identified protein N-termini. N-termini are categorized according to their NAT substrate class (with totals of substrate classes italicized) and the setup analyzed. Notable differences in NTA levels between the siCTR and siNatA setups were enriched in the NatA category of N-termini (See also Table 1 and Tables  Table 2 and Tables S1 and S2). Mean values are shown as '+'. Table 1. The effect of siNatA knockdown on the Nt-acetylation states of NatA type N-termini. All types of N-termini identified in this study matching previously reported NatA substrate specificities ((M)A-, C-, G-, S-, T-, V-) [2,15] are listed. In all cases, the numbers of uniquely identified N-termini and the number of N-termini for which the degree of NTA could be univocally calculated/determined in both setups is indicated with their corresponding number and proportion of complete and partial NTA. Whenever a reduced degree of NTA could be observed upon siNatA knockdown, the corresponding number and proportion of affected N-termini are indicated. 1 The (X)P-rule that indicates the prevention of NTA was considered for defining putative NatA substrate N-termini (i.e., N-termini carrying a proline residue at P2' were not considered) [48].   Table 1 and Tables S1  and S2 for the complete list of all (affected) N-termini identified). Only N-termini of which the degree of NTA could be determined and which were compliant with the rules NTA and NME [46,49] were used for the overall calculation of NTA. MS spectra of representative siNatA affected NatA Ala-substrate N-termini are shown in Figure 4 and Figure S2 and among others, include the translational activator GCN1 ( 2 AADTQVSETLKR 13 ) as a representative and previously confirmed NatA substrate [2,32,37], besides two newly identified mitochondrial protein substrates; namely, alpha-ketoglutarate-dependent dioxygenase alkB homolog 7 and import inner membrane translocase subunit Tim8 B (Figure 4, panels A, B and D). 74% of all siCTR NTA NatA type N-termini (e.g., 2 ASELEPEVQAIDR 14 of the E3 ubiquitin-protein ligase UBR2 ( Figure 4C)) however remained unaffected (Table 1).  Besides affected NatA type N-termini, 9 non-NatA type N-termini displayed a reduced degree of NTA (Figure 2, Tables S1,S2). More specifically, the latter category of Besides affected NatA type N-termini, 9 non-NatA type N-termini displayed a reduced degree of NTA (Figure 2, Tables S1 and S2). More specifically, the latter category of affected N-termini belongs either to iMet-retaining NatA type N-termini (7 N-termini); five MVand two MT-N-termini (NAT category "other"), besides two ML-N-termini (NAT category "NatC"). Moreover, this set includes two N-termini of which the degree of NTA was previously found to be reduced in patient derived fibroblasts and/or B-cells harboring the Ogden syndrome mutation in the gene encoding NAA10 (i.e., NAA10 S37P) [32] (Tables S1  and S2). Since besides a reduced catalytic capacity of NatA and the decreased NTA of a subset of NatA and NatE substrates, an impaired interaction between human NAA10 S37P and NAA50 (NatE) was demonstrated in Ogden cells, our data hints to these 9 non NatA N-termini as being highly likely NatE substrates ( Figure 2, Table S2).

NatA
In contrast to the general lower NTA levels of previously reported NatA substrates here identified in our siNatA setup [2,32,37], it is noteworthy that presumed NatE substrates were in general (somewhat) less affected in their NTA levels (e.g., in the case of p53 and DNA damage-regulated protein 1 (ML-) and Peptidyl-prolyl cis-trans isomerase A (MV-)), or even unaffected (when determined) in case of Kinesin-like protein KIF21A (ML-), ARL14 effector protein (MM-) and SUMO-activating enzyme subunit 1 (MV-). These results signify that the impaired interaction between NAA50 and NatA in the context of Ogden patients likely has a more prominent negative impact on NatE NTA as compared to the destabilization of NAA50-and thus NatE activity-in the NatA knockdown conditions employed in this study.

hNatA Affects Translation, Lipid Metabolism and Regulation of Cyclin-Dependent Protein Kinase Activity
When performing a comparative statistical analysis of N-terminal peptide MS-signal intensities in the siCTR versus siNatA setup (see Materials and Methods), 76 out of 1415 unique N-termini with a determined expression ratio displayed significantly altered steady state levels (p ≤ 0.01). Among the regulated N-termini, the Nt-free N-terminal peptide of NAA15 ( 2 PAVSLPPKENALFKR 16 ) was found as the N-terminus of which the expression which was most significantly reduced (over 7-fold) upon siNatA knockdown (Tables S1 and S3). More specifically, the levels of 27 and 49 protein N-termini were respectively lowered and elevated upon siNatA knockdown (Table S3). Interestingly, among the regulated proteins, and out of the 24 regulated NatA type N-termini with determined degrees of NTA (29 regulated NatA N-termini in total), 11 N-termini (46%) were here identified as NatA substrate N-termini. Overall, this indicates a~2-fold enrichment of NatA substrates in the category of regulated over non-regulated NatA type N-termini (i.e., 171 NatA substrate N-termini in case of 716 non-regulated NatA N-termini (24%) with determined degrees of NTA). This observation was statistically significant as determined by a chi-square test of independency (p < 0.02), indicating that deficiency of NatA NTA may thus directly impact protein stability and concomitantly the steady-state expression levels of certain target protein.
Out of the 11 regulated NatA substrate N-termini, 3 and 8 respectively displayed a significantly decreased or increased levels upon NatA depletion. A representative example of a NatA substrate N-terminus originating from the mitochondrial alpha-ketoglutaratedependent dioxygenase alkB homolog 7 ( 2 AGTGLLALR 10 ), classified as being fully NTA in the control setup while being only partially NTA (80%) in the siNatA knockdown sample, was found significantly upregulated (p ≤ 0.01) upon siNatA knockdown (ratio 2.6) (Figure 4, panel D).
As inferred from a ranked GO gene enrichment analysis using GOrilla [50] and considering the highest ranking instance per gene, in line with a recent proteomics study performed in NAA15 haplo-insufficient and deficient induced pluripotent stem cells (iP-SCs) [37], a general and significant downregulation of ribosomal proteins clustering near the ribosomal NatA docking site and head of the small ribosomal subunit could be observed besides a lowered expression of proteins constituting the eukaryotic 48S preinitiation complex ( Figure 5 and Tables S3 and S4). Overall, these altered N-terminal steady-state levels appear to affect translation (initiation). Moreover, cytoskeleton protein and integral membrane components next to lipid metabolic and cyclin-dependent protein kinase activity regulation processes were significantly upregulated in the siNatA setup (Table S4).
x FOR PEER REVIEW 10 of 17 appear to affect translation (initiation). Moreover, cytoskeleton protein and integral membrane components next to lipid metabolic and cyclin-dependent protein kinase activity regulation processes were significantly upregulated in the siNatA setup (Table S4).  [50] and p values were corrected for multiple hypotheses testing using the Benjamini and Hochberg false discovery rate (FDR) ( Table  S4). Only non-redundant terms with at least 5 members and corrected p values < 0.2 were considered. The bubble size corresponds to the number of regulated proteins with a given GO term and the colour code represents the p-value scale.

Discussion
In line with previous reports [1,2,32] 87% of the proteome was here shown to carry an acetyl moiety at their N-terminus.
While in our original study merely 7% hNatA substrates were found to be affected by the reduced level of hNatA despite a significant knockdown efficiency (i.e. 30% and 5% of normal NAA15 and NAA10 level, respectively) [2], in the current study, pronounced NatA knockdown -as evidenced by the nearly undetectable levels of NAA10 and NAA15 -pointed to 25% out of the putative NatA targets with lowered NTA.  [50] and p values were corrected for multiple hypotheses testing using the Benjamini and Hochberg false discovery rate (FDR) ( Table S4). Only non-redundant terms with at least 5 members and corrected p values < 0.2 were considered. The bubble size corresponds to the number of regulated proteins with a given GO term and the colour code represents the p-value scale.

Discussion
In line with previous reports [1,2,32] 87% of the proteome was here shown to carry an acetyl moiety at their N-terminus.
While in our original study merely 7% hNatA substrates were found to be affected by the reduced level of hNatA despite a significant knockdown efficiency (i.e., 30% and 5% of normal NAA15 and NAA10 level, respectively) [2], in the current study, pronounced NatA knockdown-as evidenced by the nearly undetectable levels of NAA10 and NAA15pointed to 25% out of the putative NatA targets with lowered NTA.
More so, the number of human NatA substrate N-termini identified in our study vastly exceeds the numbers reported in related N-terminomics efforts aiming at identifying human NatA targets as no less than 182 unique N-termini were here proven affected. This number thus greatly contrasts the mere 16 and 34 affected N-termini reported in a previous NatA knockdown study in Hela cells [2] and in null and/or haplo-sufficient NAA15 iPSC cells, respectively. The latter number in iPSCs could however be increased to 42 when considering the NTA rules applied here and in [2,43] to assign NAT substrates [37]. All in all, in the current effort to chart for NatA targets in human cells, an over 4 to 11-fold increase in the number of affected N-termini by NatA could be identified.
This increase in NAT targets identified are in line with some of our previous study results reporting on affected substrate N-termini in HeLa siNatB knockdown proteomes. Here, a more efficient knockdown resulted in an over 4-fold increase of NatB-type substrates displaying reduced NTA (i.e., 9% vs. 39% of the potential NatB-type substrates displayed a reduced Nt-acetylation upon pronounced knockdown), indicating a strong correlation between NAT and NTA levels [2,37,43,44]. Moreover, these reports firmly strengthen that suboptimal NAT substrates are primarily affected upon NAT perturbation.
That levels of NatA can be lowered drastically before any NTA effect becomes apparent is quite striking and might possibly be explained by the low levels of ribosome-associated and complexed NatA required to efficiently carry out NTA. Moreover, associations between NAA15/NAA10/NAA50 (NatA/NatE) and even associated HYPK (NatE/HYPK) have been shown to have a mutually (functionally) stabilizing effect on top of influencing the catalytic activities of both NAA10 and NAA50 ( [14,21,27,28,30,37] and this study). These stabilizing effects mean that complexed NatA/NatE is more efficient and stable, and thus potentially lowered less drastically in concentration upon knockdown as compared to the concentration of its composing subunits present in monomeric form at early timepoints post knockdown.
Interestingly, while 95% affected N-termini displayed the NatA substrate specificity, 7 iMet-retaining NatA type N-termini were affected. This substrate specificity is indicative of them being putative NatE substrates likely affected by an impaired interaction between NatA and NAA50 (NatE) in case of NatA perturbation [32]. Besides, coordinated (and thus influenced) NTA actions by both NatE catalytic subunits have been demonstrated [22].
The observation that abolishing NAA15 expression in NAA15 null iPSCs affects the NatA Nt-acetylome less pronounced than the NatA knockdown model assessed in this study might indicate a compensatory role for the NAA15 paralogue NAA16 in human cells when considering cotranslational NatA Nt-acetylation [38]. While nonetheless NAA16 expression was low in iPSCs and generally less abundant as compared to abundant NAA15 [38], our previous and current knockdown studies demonstrates that limited levels of NAA10 and NAA15-and thus possibly also by extension low NAA16 levels-might still be sufficient to maintain the essentially unaltered NTA levels of NatA substrates reported for NAA15 null iPSCs.
However, it is noteworthy that all 3 NatE constituents are listed "core essentialome" components in case of the human haploid HAP1 and the near-haploid KBM7 cell lines [51]. In accordance, we were unable to successfully obtain single NAA15, NAA10 or NAA50 haploid HAP1 CRISPR knockout cell lines (unpublished data), indicating a cell-type dependent rather than a general compensatory role for NAA16 correlating with NAA16 expression levels [38].
Alternatively, in NAA15 null cells, NAA10 could also carry out post-translational Nt-acetylation independent of NAA15. This scenario is however very unlikely for at least 3 reasons. First, a different substrate specificity profile of non-complexed NAA10 was reported [21,24]. Second, a clear overlap of the substrate repertoire identified in the NAA15 iPSCs [37] and a previous as well as current NatA knockdown study was notable, and third, ribosome interaction has been proven a crucial aspect for NatA activity in yeast [20].
Remarkably, in contrast to previously reported proteome-wide findings, our results indicate that deficiency of NatA NTA may directly impact protein stability and thus the expression levels of NatA targets. Especially, the presumed less efficient or suboptimal NatA-and thus NatA knockdown affected-substrates appear prime NatA targets being affected in their steady-state levels.
This apparent discrepancy with previous studies might likely be attributed to the increased statistical power provided by the higher number of NatA substrates here found to be affected upon NatA knockdown [2,32,37]. Comprehensive shotgun analysis is however deemed necessary to extend this observation form the N-terminome to the proteome, and to further confirm this observation.
Nonetheless, in a previous attempt to study the effect of NTA on protein stabilityand while for most of the substrates analyzed no such effect was discernible-we already showed that the steady-state levels of the-here also identified and confirmed-NatA substrate THOC7 (THO complex subunit 7 homolog) was strongly reduced and destabilized [32].
Despite the relative large numbers of differentially expressed proteins here inferred from their N-terminal peptide abundances and from previous differential shotgun proteomics analyses of NAA15 haplo-insufficient and deficient iPSCs, only very few, if any, significant differences upon NatA perturbation were apparent at the mRNA level [37]. Further, and notwithstanding the significant lowered transcription and translation of NAA15 and NAA10 observed, we found no notable if any changes at the level of the translatome by means of ribosomal profiling (unpublished data). Thus, it can be concluded that transcriptome and translatome findings by and large rule out a possible role for NatA in steering transcription and translation efficiency apart from the NTA deficiency and NatA NTA steered protein stability effects observed.
Comprehensive NatA substrate discovery and discrimination between inefficient and inefficient substrates overall provided a better understanding of NTA by NatA. Such data is of much need to be able to further unravel the wide-ranging developmental defects observed in humans carrying mutations in NAA10 and NAA15 [34]. For L-Arg SILAC (stable isotope-labeling by amino acids in cell culture) labelling [40] of A-431 cells, cells were grown in DMEM SILAC medium (Invitrogen, cat n • 61870-010) containing either natural ( 12 C 6 ) or 13 C 6 15 N 4 -labelled L-Arg (Cambridge Isotope Laboratories, Andover, MA, USA) at a concentration of 80 µM (i.e., 16.8 mg/L or 20% of the suggested concentration present in DMEM medium at which L-Arg to Pro conversion was not detectable for A-431 cells) and supplemented with 10% dialyzed fetal bovine serum (Invitrogen, cat n • 26400-044). Cells were cultured for at least six population doublings for complete incorporation of the labelled L-Arg.

Cell Culture
Cells were cultured at 37 • C in a humidified incubator at 37 • C and 5% CO2 and passaged every 3-4 days.

N-Terminal Proteomics
Cells were detached using cell dissociation buffer (Invitrogen, cat n • 13151014). The obtained cell pellets (~9 × 10 6 cells) were lysed in 1 mL 50 mm sodium phosphate buffer pH 7.5 and 50 mm NaCl supplemented with a complete protease inhibitor mixture tablet (Roche Diagnostics GmbH, Mannheim, Germany) using multiple cycles of freeze-thawing (3×). After lysate clearance by centrifugation for 15 min at 16,000× g and 4 • C, protein concentrations were adjusted to 2 mg/mL as measured using the DC Protein Assay Kit (Bio-Rad, Hercules, CA, USA)). Aliquots (50 µg) were analyzed in parallel by SDS/PAGE and Western blotting to confirm efficient knockdown of NAA10 and NAA15 (Figure 1).
One combined sample (Figure 1, 1/1 (siNatA/siCTR) protein mix) was subjected to Nterminal COFRADIC analysis as described in [1,44]. More specifically, N-hydroxysuccinimide ester of (stable isotopic encoded) acetate labelling of free amines at the protein level (i.e., NHS esters of 13 C 2 D 3 acetate) was used (i.e., in vitro 13 C 2 D 3 -acetylation) in combination with differential L-Arg SILAC (stable isotope-labeling by amino acids in cell culture) labeling, which allows for the calculation of the extent of NTA, as well as the relative quantification of N-terminal peptides [1,52,53].

LC-MS/MS Analysis and Data Analysis
45 fractionated samples were introduced into the Ultimate 3000 (Dionex, Amsterdam, The Netherlands) in-line connected to an LTQ Orbitrap XL mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) and liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis was performed as described previously [44,54]. Spectra were searched against the Swiss-Prot database (taxonomy Homo sapiens) with Mascot using the Mascot Daemon interface (version 2.2.0, Matrix Science). Lys heavy ( 13 C 2 D 3 − ) acetylation (+47.036 Da), Cys carbamidomethylation (+57.021 Da) and Met oxidation (+15.995 Da) were set as fixed modifications. Variable modifications included light/heavy ( 13 C 2 D 3 − ) NTA (+42.011 Da or 47.036 Da) and pyro-Glu formation from Gln (−17.026 Da). Mass tolerance on precursor ions was set to 10 ppm (with Mascot's C13 option set to 1) and on fragment ions to 0.5 Da. Endoproteinase semi-Arg-C/P (Arg-C specificity with Arg-Pro cleavage allowed) was set as enzyme with no missed cleavages. The peptide charge was set to 1+, 2+, 3+ and instrument setting was put to ESI-TRAP. Only N-terminal peptides that were ranked one and scored above the threshold score, set at 99% confidence, have a minimum amino acid length of six, and were compliant with the rules of NTA or initiator methionine (iMet) processing [45,46] were withheld. More specifically, iMet processing was considered in case of iMet-starting N-termini followed by any of the following amino acids; Ala, Cys, Gly, Pro, Ser, Thr, Met, or Val, or exceptionally Glu and Asp in case of actin representative of NatH type substrate N-termini [55]. Quantification of the degree of NTA was performed as described previously [1]. Considerable variations in the degree of NTA (defined as a minimum difference of 10% in the degree of NTA, or 5% in case of the clear presence of two isotopic envelopes in only one of the setups, while a unique envelope signifying an Nt-free or fully NTA N-terminus in the siNatA or siCTR setup, respectively. When additionally considering the latter category of affected N-termini, 68 N-termini (out of 182 affected N-termini in total) were further classified as being affected upon siNatA knockdown (Supplemental Table S2).
SILAC quantifications ( 12 C 6 L-Arg versus 13 C 6 15 N 4 L-Arg) were carried out using the Mascot Distiller Quantitation Tool (version 2.2.1). Ratios for all peptides were calculated by comparing the XIC peak areas of all matched light versus heavy peptides and the calculated ratios originally reported as FALSE were all verified by visual inspection of the highest scoring MS spectra [44]. A distribution of all individual peptide ratios was determined followed robust statistics as described in [56] (Median: 1.038 and Huber scale: 0.381). Peptides displaying a ratio that was significantly up-or down-regulated (99% confidence interval) were considered affected by the NAA10/NAA15 knockdown (Table S3).
For the GO gene enrichment analysis using GOrilla a ranking of all determined Nterminal peptide-based expression ratios were used as input, keeping only the highest ranking instance of each gene. More specifically, of the 1498 unique N-termini identified in total (Table S1), 1415 (94%) had determined values. Of these, the system recognized 1413 genes and 66 duplicate genes were removed (keeping the highest-ranking instance of each gene) leaving a total of 1347 genes. Of these, 1335 were associated with a GO term and thus used for the enrichment analysis. Only enriched (non-redundant) GO terms with minimally 5 members were considered.
Data management was done in ms_lims [57], KNIME, Excel and RStudio (R version 4.1.0). For data visualization GraphPad Prism version 9.2.0 was used.