NatB Protects Procaspase-8 from UBR4-Mediated Degradation and Is Required for Full Induction of the Extrinsic Apoptosis Pathway

Abstract

N-terminal acetyltransferase B (NatB) is a major contributor to the N-terminal acetylome and is implicated in several key cellular processes including apoptosis and proteostasis. However, the molecular mechanisms linking NatB-mediated N-terminal acetylation to apoptosis and its relationship with protein homeostasis remain elusive. In this study, we generated mouse embryonic fibroblasts (MEFs) with an inactivated catalytic subunit of NatB (Naa20^-/-) to investigate the impact of NatB deficiency on apoptosis regulation. Through quantitative N-terminomics, label-free quantification, and targeted proteomics, we demonstrated that NatB does not influence the proteostasis of all its substrates. Instead, our focus on putative NatB-dependent apoptotic factors revealed that NatB serves as a protective shield against UBR4 and UBR1 Arg/N-recognin-mediated degradation. Notably, Naa20^-/- MEFs exhibited reduced responsiveness to an extrinsic pro-apoptotic stimulus, a phenotype that was partially reversible upon UBR4 Arg/N-recognin silencing and consequent inhibition of procaspase-8 degradation. Collectively, our results shed light on how the interplay between NatB-mediated acetylation and the Arg/N-degron pathway appears to impact apoptosis regulation, providing new perspectives in the field including in therapeutic interventions.

Introduction

N-terminal protein modifications (NPMs) are among the first and fastest responses to the cellular environment, influencing not only the modified protein, but also cell, tissue, and whole organism phenotypes.^1–4 One major NPM is N-terminal (Nt-) acetylation, an essential multitasking protein modification implicated in many diseases,^5,6 where an acetyl group is transferred from acetyl coenzyme A to the α-amino group of a protein N-terminus catalyzed by Nt-acetyltransferases (NATs).^5,7 The eight currently known eukaryotic NATs have been defined based on their subunits and substrate specificity.^6,8 Nt-acetyltransferase B (NatB) comprises the catalytic subunit NAA20 and the auxiliary subunit NAA25. NAA25 binds to the ribosome and ensures cotranslational activity of NatB,⁹ while the catalytic site of NAA20 anchors and acetylates the N-alpha amino group of the initiator methionine (iMet) of all proteins immediately followed by an asparagine, glutamine, aspartic or glutamic acid residue.^10,11 Together with NatA, which acetylates proteins that have lost their iMet, NatB is the major contributor to Nt-acetylation. NatB substrates cover nearly 21% of human, Caenorhabditis elegans, and Arabidopsis thaliana proteomes and 15% of yeast proteome. In all these organisms, almost all NatB substrates are fully acetylated, unlike the substrates of other NATs (especially NatA), many of which are only partially modified.^10,12,13

For a few well-studied proteins, NatB has been shown to be involved in promoting resistance to aggregation, protein-protein interactions, protein sorting to distinct cellular compartments, and regulating protein half-life.^6,14 However, the overall impact of NatB-mediated Nt-acetylation on proteostasis is still poorly understood. Global protein stability does not seem to be affected by NatB subunit depletion in yeast or human cells,^15–17 suggesting that NatB-mediated Nt-acetylation may not be a dominant factor in proteostasis. Conversely, several studies have highlighted the importance of NATs-mediated Nt-acetylation on the homeostasis of specific proteins in the context of N-degrons, which are N-terminal degradation signals recognized by the N-degron pathways.¹⁸ For instance, the N-terminal acetylated variants of MX-Rgs2, a G-protein regulator, have been shown to be selectively targeted by the Ac/N-end rule pathway when X is either Arg or Gln.¹⁹

Whatever the impact of NatB-mediated Nt-acetylation on proteostasis or protein half-life, the inactivation or downregulation of any NatB subunit influences fundamental cellular processes such as actin cytoskeleton organization,^10,20–23 NAD⁺ homeostasis,²⁴ influenza virus protein PA-X,²⁵ cell proliferation,^22,23,26 and apoptosis.^23,26,27 However, the exact molecular mechanisms linking defective NatB or NatB-mediated Nt-acetylation to specific cellular processes and their interdependence have not yet been elucidated.

Given the prevalence and importance of Nt-acetylation and apoptosis in numerous human diseases,^6,28 we sought to establish the mechanistic role of NatB-dependent Nt-acetylation on apoptosis regulation. Here, we show that NatB inactivation in mouse embryonic fibroblasts (MEFs) has the potential to lead to degradation of proapoptotic proteins, particularly of procaspase-8, mediated by the UBR4 E3 ubiquitin ligase, resulting in a decreased activation of the extrinsic apoptotic pathway. Our findings provide the first evidence implicating NatB-Arg/N-degron pathway in apoptosis regulation.

Results

Complete Naa20 inactivation induces cytoskeletal abnormalities and decreases cell proliferation

Based on our previous studies that showed increased sensitivity of Naa25^-/- MEFs²⁷ and Naa20^-/- HeLa cells²⁶ to the apoptotic inducer MG132, we aimed to further investigate the role of NatB acetylation on apoptosis by inactivating the NatB catalytic subunit in MEFs.

Six days of infection with Ad5CMVCre (Naa20^-/- MEFs) resulted in a 99.6% reduction of Naa20 transcript compared with AdEmpty-infected MEFs (WT) (Figure 1A), and NAA20 protein was no longer detectable in Naa20^-/- MEFs (Figure 1B). Naa20 inactivation was associated with a significant reduction in Naa20^-/- MEF proliferation four days after infection (Figure 1C). Moreover, actin fibers became disorganized in Naa20^-/- MEFs compared to WT (Figure 1D), and, in keeping with this actin disorganization, focal adhesions decreased, a phenotype commonly observed upon NatB complex inactivation or downregulation in other cell types. These results confirmed that Naa20 is inactivated in Naa20^-/- MEF cells and induces phenotypes similar to those previously observed in other NatB-downregulated cellular contexts.^10,20,21,23

Figure 1.

In Naa20^-/- MEFs lacking the catalytic subunit of NatB, Naa20 mRNA and NAA20 protein are not expressed, and cell proliferation and the actin cytoskeleton are perturbed. (A) Naa20 mRNA quantification analysis in MEFs infected with the AdEmpty (WT) and Ad5CMVCre (Naa20^-/-) by RT-PCR. Data are the mean ± SEM of six independent experiments and are normalized to a housekeeping gene (Histone 3) and analyzed with Student’s t test. **P < 0,001, ***P < 0,0001. (B) NAA20 protein levels were detected by Western blotting in WT and Naa20^-/-cells using antibodies targeting NAA20. (C) Cell proliferation was visualized as the number of cells per well along postinfection days. Values represent the average number of WT or Naa20^-/- MEFs from three independent replicates as mean ± SEM (Student’s t test). (D) Confocal microscopy images of actin (phalloidin) and focal adhesions (vinculin) in WT or Naa20^-/- MEFs representative of 500 cells.

Quantitative N-terminomics of Naa20^-/- MEFs reveals reduction in Nt-acetylation of only NatB substrates

We next performed N-terminomics analysis of WT and Naa20^-/- MEFs using the “Stable Isotope Labeling Protein N-terminal Acetylation Quantification” method to identify and quantify N-termini (SILProNAQ²⁹). Analysis of WT and Naa20^-/- MEF cells identified 25,048 protein N-termini corresponding to 1988 non-redundant proteoforms (Supplementary Dataset 1 and Supplementary Table 1). The acetylation yields of 1191 of them (1073 in the WT and 945 in the NatB mutant background) could be quantified. The analysis of these unique proteoforms unraveled that 769 underwent removal of the initial methionine (iMet) in agreement with known N-terminal methionine excision (NME) rules in eukaryotes,³⁰ whereas 258 proteoforms retained their iMet (Supplementary Table 1). The remaining 164 underwent larger cleavages in keeping with leader peptide removal. The quasi-totality of the iMet starting N-termini featured at position two an amino acid with a large lateral chain (Supplementary Dataset 1).

In Naa20^-/- MEFs, there was an overall reduction in Nt-acetylation, primarily in proteins retaining their iMet (Figure 2A and B), with a relative decrease in the number of fully acetylated N-termini to only partially acetylated N-termini (Supplementary Dataset 1 and Supplementary Table 1). By contrast, acetylation of N-termini devoid of iMet or N-termini processed downstream were unaffected (Figure 2C and D). We compared our N-terminomics data to those previously reported for human cells in which NatB was downregulated.¹⁰ Common identified and quantified N-termini of the two datasets revealed mostly decreases of the acetylation yields of NatB substrates (Supplementary Dataset 2). This effect was significantly stronger in Naa20^-/- MEFs (Supplementary Dataset 2). N-termini with the highest decreased Nt-acetylation in Naa20^-/- MEFs (26, Table 1 and Figure 2E and F) corresponded to NatB substrates (23), of which 18 displayed Asp and Glu at position two and, to a minor extent (5), Asn and Gln (Table 1 and Supplementary Dataset 1). In Naa20^-/- MEFs, only three non-NatB substrates also showed strongly decreased Nt-acetylation (Table 1), but for which no information is available for the human counterparts (Supplementary Dataset 2).

Figure 2.

N-terminomic and proteomic analyses of Naa20^-/- MEFs reveal reduced N-terminal (Nt-) acetylation, specifically of NatB substrates, and deregulated protein levels. Accumulated distribution representing the percentage of Nt-acetylated (A) total proteome, (B) peptides with the initial methionine (iMet) present, (C) peptides with Nt-methionine excised, and (D) peptides with processed N-termini identified in normal (WT) and Naa20^-/- MEFs. (E) IceLogo representation of the N-terminal sequences with a minimum 20% decrease in NTA yield to those with < 5% variation, when comparing the Naa20^-/- to the WT MEFs. Volcano plot representing (F) proteins with an Nt-acetylation reduction of 40% or more (green spots). The dashed horizontal line shows the P value cut off, and the eight points highlighted in green indicate the affected proteins with the highest statistical significance. (G) Proteins with a label-free quantification (LFQ protein ratio) either < 0.5 (green spots) or > 2.0 (red spots) in the Naa20^-/- compared with WT samples. The dashed horizontal line shows the P value cutoff, and the points highlighted in green and red represent the downregulated and upregulated proteins, respectively. All plots contain only the N-termini that have been quantified at least once in each condition (WT or Naa20^-/-) from four replicates each. FDR, false discovery rate; N, number of proteins.

Table 1.

Most affected N-terminal acetylated substrates resulting from Naa20 knock-out in MEFs and their correlation with differences in protein expression when label-free quantitation (LFQ) was possible in WT and mutant samples.

Plota	Entry Name	Protein Description	NTA Pos.	N-1 Res.	Nt- Seq.	%NTA (WT)	%NTA (Naa20−/−)	NTA Difference (Naa20−/−-(WT)	NTA Ratio (Naa20−/−)/(WT)	LFQ Protein Ratio (Naa20−/−/WT)
1	RS24_MOUSE	40S ribosomal protein S24	1	–	MNDTVTIRTR	99.4 ± 0.6	33.9 ± 3.4	−65.5	0.341	0.693
2	TM258_MOUSE	Oligosaccharyl transferase subunit TMEM258	1	–	MELEAMSRYT	93.3 ± 4.2	13.4 ± 2.1	−79.9	0.144	–
3	PRUN1_MOUSE	Exopolyphosphatase PRUNE1	1	–	MEDYLQDCRA	99.8 ± 0.3	44.9 ± 1.1	−54.9	0.450	2.075
4	IPO7_MOUSE	Importin-7	1	–	MDPNTIIEAL	99.5 ± 0.8	44.3 ± 3.5	−55.1	0.446	0.516
5	IFM3_MOUSE	Interferon-induced transmembrane protein 3	1	–	MNHTSQAFIT	97.9 ± 0.6	32.8 ± 1.8	−65.2	0.335	–
6	KEAP1_MOUSE	Kelch-like ECH-associated protein 1	1	–	MQPEPKLSGA	99.1 ± 0.7	41.5 ± 1.4	−57.6	0.419	–
7	TEBP_MOUSE	Prostaglandin E synthase 3	1	–	MQPASAKWYD	86.2 ± 7.3	40.7 ± 0.3	−45.5	0.472	0.624
8	AAAS_MOUSE	Aladin	2	M	CSLGLFPPPP	99.9 ± 0.0	5.9 ± 5.5	−94.1	0.059	–
	INT3_MOUSE	Integrator complex subunit 3	1	–	MELQKGKGTV	98.6 ± 0.9	26.1	−72.5	0.265	–
	WDR46_MOUSE	WD repeat-containing protein 46	1	–	METAPKPGRG	100.0	50.3 ± 0.7	−49.7	0.503	–
	PRAF3_MOUSE	PRA1 family protein 3	1	–	MDVNLAPLRA	99.7 ± 0.3	52.8 ± 3.8	−46.9	0.529	–
	BABA2_MOUSE	BRISC and BRCA1-A complex member 2	1	–	MSPEIALNRI	99.2 ± 0.2	57.8	−41.4	0.583	–
	RFA3_MOUSE	Replication protein A 14 kDa subunit	1	–	MEDIMQLPKA	99.6 ± 0.4	58.5 ± 3.1	−41.1	0.588	–
	ILK_MOUSE	Integrin-linked protein kinase	1	–	MDDIFTQCRE	99.9 ± 0.1	63.4 ± 0.9	−36.5	0.635	–
	GTR1_MOUSE	Solute carrier family 2	1	–	MDPSSKKVTG	99.3 ± 0.7	63.3 ± 2.3	−36.0	0.638	0.623
	P4K2A_MOUSE	Phosphatidylinositol 4-kinase type 2-alpha	1	–	MDETSPLVSP	99.0 ± 0.8	66.7	−32.3	0.674	–
	FBX22_MOUSE	F-box only protein 22	1	–	MEPAGGGGGV	99.9 ± 0.0	67.9 ± 2.9	−32.0	0.680	–
	TTL12_MOUSE	Tubulin–tyrosine ligase-like protein 12	1	–	MEIQSGPQPG	99.9 ± 0.1	70.7	−29.2	0.708	0.748
	KT3K_MOUSE	Ketosamine-3-kinase	1	–	METLLKRELG	99.9 ± 0.0	70.9 ± 0.1	−29.1	0.709	–
	BCD1_MOUSE	Box C/D snoRNA protein 1	1	–	MESAAEKEGT	99.5 ± 0.5	71.9 ± 1.8	−27.5	0.723	–
	NDUAC_MOUSE	NADH dehydrogenase 1 alpha subcomplex subunit 12	1	–	MELVEVLKRG	99.9 ± 0.0	73.3 ± 2.0	−26.7	0.733	–
	FARP1_MOUSE	FERM, RhoGEF and pleckstrin domain-containing protein 1	2	M	GEIEQKPTPA	96.9 ± 2.7	70.4 ± 2.6	−26.6	0.726	0.936
	TCPA_MOUSE	T-complex protein 1 subunit alpha	1	–	MEGPLSVFGD	99.6 ± 0.5	74.0 ± 4.3	−25.5	0.743	0.944
	AN32E_MOUSE	Acidic leucine-rich nuclear phosphoprotein 32 family member E	1	–	MEMKKKINME	99.9 ± 0.0	75.8 ± 2.6	−24.2	0.758	0.990
	THIC_MOUSE	Acetyl-CoA acetyltransferase	1	–	MNAGSDPVVI	99.6 ± 0.3	76.9 ± 4.4	−22.8	0.772	0.921
	SYRC_MOUSE	Arginine–tRNA ligase	1	–	MDGLVAQCSA	99.5 ± 0.5	79.5 ± 4.7	−20.0	0.799	1.009
	CASP3_MOUSE	Procaspase-3	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.	0.649
	CASP8_MOUSE	Procaspase-8	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.
	CASP9_MOUSE	Procaspase-9	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.
	MARH6_MOUSE	E3 ubiquitin-protein ligase MARCHF6	1	–	MDTAEEDICR	NTA	NTA	–	–	n.i.
	UBR4_MOUSE	E3 ubiquitin-protein ligase UBR4	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.	n.i.	0.699

Proteins with a significant decrease in NTA according to a FDR < 5%. NTA Pos., NTA position; Nt-Seq., N-terminal sequence; n.i., protein not identified in the experiment; ^amost statistically significant NTA downregulated proteins shown in Figure 2G.

We observed that the functions of many retrieved proteins with reduced Nt-acetylation in Naa20^-/- MEFs were apoptosis-related (Supplementary Dataset 1 and Table 1). Among them, we found TMEM258, a central mediator of ER quality control via apoptosis³¹; importin 7, which is involved via the Hippo pathway in the expression of genes important in apoptosis³²; and IFM3, a member of the interferon-induced transmembrane proteins (IFITMs) family.³³ This intriguing observation led us to note that procaspases-3, -8, -9, BAX, APAF1, and BID, which are major contributors to extrinsic and intrinsic apoptosis (Supplementary Figure 1), systematically displayed Asp or Glu at position 2. These proteins are also inferred as NatB substrates based on sequence similarity between mouse and human (100% conservation of the first amino acids) and a combination of experimental and computational evidence.^10,34–36 Unfortunately, our SilProNAQ approach did not allow us to retrieve and analyze these N-termini due to the distal position of Arg residues required for trypsin cleavage and peptide identification (Table 1).

To further explore the relationship between NatB-dependent Nt-acetylation and proteostasis, we next examined the NatB-dependent proteome. Label-free quantification (LFQ protein ratio) analysis showed that only 36 of 1235 statistically relevant proteins, among which 189 are NatB-substrates, were drastically deregulated in Naa20^-/- MEFs (Supplementary Dataset 1, Table 2, and Figure 2G). Almost all upregulated proteins were NatA substrates and, where Nt-acetylation could be measured, no variation was observed (Table 2). By contrast, the amino terminal sequence of 10 of 21 strongly downregulated proteins indicated that they were predicted as NatB substrates, including ANP32B (Q9EST5), a multifunction protein directly cleaved by caspase-3 and a negative regulator of caspase-3-dependent apoptosis (Supplementary Dataset 1 and Table 2). Although Nt-acetylation of procaspase-3, -8, and -9 was not quantifiable, caspase-3 and the UBR4 E3-ligase were also found to be reduced in Naa20^-/- MEFs to some extent (Supplementary Dataset 1 and Table 1). We obtained information of the Nt-acetylation yield of 12 out of the 36 most drastically deregulated proteins. Of nine downregulated proteins, Nt-acetylation yields were unmodified for five of them (% Nt-acetylation difference <1.3), while four including three NatB substrates—spermidine synthase, caveolae-associated protein 1, and reticulon-4—displayed a parallel reduction in Nt-acetylation and protein accumulation (Table 2).

Table 2.

Differentially accumulated proteins when the NatB catalytic subunit is inactivated in MEFs.

Plota[AQ3}	Entry Name	Protein Description	LFQ Protein Ratio Naa20−/−/ WT	%NTA Difference Naa20−/− – WT
1	TAGL_MOUSE	Transgelin	0.331	−1.3 (Ala2)
2	SERC_MOUSE	Phosphoserine aminotransferase	0.492	−0.4 (iMet)
3	LYOX_MOUSE	Protein-lysine 6-oxidase	0.228	–
4	USO1_MOUSE	General vesicular transport factor p115	0.367	–
5	SPEE_MOUSE	Spermidine synthase	0.444	−3.3 (iMet)
6	AN32B_MOUSE	Acidic leucine-rich nuclear phosphoprotein 32 family member B	0.404	–
7	SET_MOUSE	Protein SET	0.314	–
8	CAVN1_MOUSE	Caveolae-associated protein 1	0.425	−15.9 (iMet)
9	FACR1_MOUSE	Fatty acyl-CoA reductase 1	0.474	–
10	HNRL2_MOUSE	Heterogeneous nuclear ribonucleoprotein U-like protein 2	0.282	–
11	LOXL1_MOUSE	Lysyl oxidase homolog 1	0.423	–
12	RTN4_MOUSE	Reticulon-4	0.462	−4.0 (iMet)
13	RAB18_MOUSE	Ras-related protein Rab-18	0.468	–
14	RS30_MOUSE	40S ribosomal protein S30	0.372	0.0 (Lys1)
15	IMA7_MOUSE	Importin subunit alpha-7	0.416	−0.4 (iMet)
16	EF1B_MOUSE	Elongation factor 1-beta	0.405	–
17	U2AF1_MOUSE	Splicing factor U2AF 35 kDa subunit	0.450	–
18	RLA1_MOUSE	60S acidic ribosomal protein P1	0.406	−0.1 (Ala2)
19	YTHD2_MOUSE	YTH domain-containing family protein 2	0.438	−0.2 (Ser2)
20	QKI_MOUSE	Protein quaking	0.460	–
21	LYAR_MOUSE	Cell growth-regulating nucleolar protein	0.466	–
22	KCD12_MOUSE	BTB/POZ domain-containing protein KCTD12	5.143	–
23	BGLR_MOUSE	Beta-glucuronidase	2.274	–
24	ASAH1_MOUSE	Acid ceramidase	2.084	–
25	NEP_MOUSE	Neprilysin	3.020	–
26	HYEP_MOUSE	Epoxide hydrolase 1	3.681	–
27	HXK1_MOUSE	Hexokinase-1	3.404	–
28	SCOT1_MOUSE	Succinyl-CoA:3-ketoacid coenzyme A transferase 1	2.209	–
29	TPP1_MOUSE	Tripeptidyl-peptidase 1	2.099	–
30	INF2_MOUSE	Inverted formin-2	2.108	−0.2 (Ser2)
31	CSN3_MOUSE	COP9 signalosome complex subunit 3	2.273	0.0 (Ala2)
32	S61A1_MOUSE	Protein transport protein Sec61 subunit alpha isoform 1	2.887	−0.3 (Ala2)
33	H32_MOUSE	Histone H3.2	3.601	–
34	DDX3Y_MOUSE	ATP-dependent RNA helicase DDX3Y	2.001	–
35	HEXB_MOUSE	Beta-hexosaminidase subunit beta	2.052	–
36	BACH_MOUSE	Cytosolic acyl coenzyme A thioester hydrolase	2.098	–

^aMost affected proteins (2 < LFQ protein ratio < 0.5) are presented. For an exhaustive list, see Supplementary Table 1 and Dataset S1. Blue line divides the downregulated proteins from the upregulated proteins. See also Volcano plot, Figure 2G.

Taken together, our proteomic investigation suggest that NatB is not likely involved in the proteostasis of all of its substrates, but rather indicates that NatB-mediated Nt-acetylation may still protect a specific set of proteins from degradation, such as components of the apoptotic pathways.

Naa20 inactivation reduces the expression of several components of the apoptotic pathways decreasing activation of procaspases in response to TNF-α plus SMAC mimetic and etoposide

Although a significant number of key components of the apoptotic pathway are NatB substrates (see above and Supplementary Figure 1), our proteomic analysis did not allow us to simultaneously quantify the Nt-acetylation yield and protein accumulation of none of them most likely due to the limitations referred above. To overcome this difficulty we sought to analyze the impact of Naa20 inactivation on the expression levels of intrinsic and extrinsic apoptotic components, including the NatB substrates BID and procaspase-8,-9 and -3, in Naa20^-/- MEFs. However, as it has been shown that Cre recombinase expression could be cytotoxic through unspecific activities,³⁷ we first confirmed that it did not affect the expression levels of intrinsic and extrinsic apoptotic components, including the NatB substrates BID and procaspase-8,-9 and -3, independently of Naa20 inactivation (Supplementary Figure 2). We then performed a targeted time-course assay after Naa20 inactivation to analyze their accumulation. Naa20^-/- MEFs showed no variation in apoptotic NatA substrates such as procaspase-6 and SMAC (or DIABLO) (Figure 3A). However, the levels of procaspases-8, -9, -3 and BID clearly decreased in Naa20^-/- MEFs (Figure 3A and Supplementary Figure 3A), while no differences in procaspase-8, -3 and Bid transcripts were observed six days after infection (Figure 3B). Protein levels of BAX and APAF1, two other NatB substrates, were not decreased; in fact, they accumulated in the absence of NAA20 (Figure 3A). Interestingly, both Bax and Apaf1 mRNA levels were markedly upregulated six days post-infection in the absence of NAA20 (Figure 3B). Furthermore, we observed the presence of cleaved PARP in the absence of NAA20, despite not detecting active caspase-3 (Figure 3A and Supplementary Figure 3A). PARP is typically cleaved and inactivated by caspase-3 during apoptosis. Nonetheless, PARP cleavage can occur independently of caspase-3 activation through different mechanisms,³⁸ such as during senescence. Recent evidence³⁹ demonstrates that inactivation of Naa20 can also induce senescence, which may in turn be implicated in the cleavage of PARP observed under our experimental conditions. Altogether, our data clearly show a strong negative effect of NatB inactivation on procaspase-3, -8, and -9 and BID expression, whereas there is an upregulation of the two pro-apoptotic proteins BAX and APAF1. Unexpectedly, the increase in these two proteins, acting in the intrinsic apoptotic pathway upstream of the proteolytic activation of procaspases-9 and-3, did not result in an evident increase of PARP cleavage most likely due to the decreased levels of cleaved BID and of procaspase-9.

Figure 3.

Inactivation of Naa20 decreases the protein expression of procaspase-8, -9, and -3, and BID, which limits the activation of the extrinsic and intrinsic apoptotic pathways. (A) Representative Western blot images of the NatB substrates procaspase-8, -9, and -3 and of BAX, BID, APAF1 and non-NatB substrate procaspase-6, SMAC and cleaved PARP. MEFs were harvested 4, 5, and 6 days after infection with AdEmpty (WT) or Ad5CMVCre (^-/-) inducing Naa20 inactivation. Procaspase-6 and SMAC, which are not NatB substrates, were used as controls. GAPDH was used a loading control. (B) Relative quantification by RT-PCR of the same NatB substrates as in (A), six days postinfection. Data are the mean ± SEM of six independent experiments and are normalized to a housekeeping gene (Histone 3). Student’s t test was used to evaluate differences between groups with the obtained values indicated when statistical significance was achieved (P < 0.05). (C) Representative Western blot images of procaspase-8, -9, and -3 and of respective cleaved caspases, cleaved PARP and BID in Naa20^-/- MEFs 6 days after infection with AdEmpty (WT) or Ad5CMVCre (^-/-) 0, 8, 12, and 24 h after treatment with 100 ng/mL TNF-α plus 500 nM SMAC mimetic. (D) As in (C) but after treatment with 50 µM etoposide. For the immunoblots, one representative of three independent experiments is shown.

The decrease in procaspases accumulation prompted us to assess their activation in Naa20^-/- MEFs in response to apoptotic inducers like etoposide and TNF-α plus SMAC mimetic. For both stimuli, procaspase-8, -9, and -3 levels slightly decreased in WT cells through proteolytic processing (Figure 3C and D and Supplementary Figure 3B and C). However, procaspases and BID protein levels in Naa20^-/- MEFs—while lower at time 0 compared with WT MEFs—remained constant throughout the experiment and, as expected, were associated with lower levels of cleaved caspases and PARP.

We next questioned whether the observed reduction of procaspase activation upon Naa20 inactivation could be extended to other apoptosis inducers including tunicamycin or MG132, which induce apoptosis by activating caspase-9 and caspase-8, respectively.⁴⁰ Throughout the experiment, the differences in protein levels of cleaved caspase-8, -9, -3, and PARP between Naa20^-/- MEF and WT cells were mostly absent for both treatments, in contrast to the observed decrease when cells were treated with etoposide and TNF-α plus SMAC mimetic (Supplementary Figure 4A and B and Figure 5A and B).

As Naa20 inactivation in MEFs decreased procaspase activation in response to the intrinsic apoptosis inducer etoposide, we addressed whether reduction of apoptosis could be due to the impairment of BAX activation. There were no differences in mitochondrial BAX and no detectable cytosolic cytochrome c (cyt c) and SMAC in WT and Naa20^-/- MEFs under basal conditions (Figure 4A and Supplementary Figure 6A). Curiously, Naa20^-/- and WT MEFs, 12 h after etoposide treatment, showed similar BAX translocation from the cytosol to mitochondria. However, mitochondrial release of cyt c and SMAC was slightly lower in Naa20^-/- than WT MEFs despite increases in cellular cyt c and SMAC after etoposide treatment, as previously described⁴¹ (Figure 4B and Supplementary Figure 6B). In contrast, similar cyt c and SMAC release from mitochondria to the cytosol, and consequently of procaspase-3 activation were observed in WT and Naa20^-/- MEFs 8 h after MG132 treatment (Figure 4C and Supplementary Figure 6C).

Figure 4.

Inactivation of Naa20 has no impact on BAX localization in response to etoposide, TNF-α + SMAC mimetic, and MG132 treatments but decreases the release of cytochrome c and SMAC in response to etoposide. (A) Representative Western blot images of BAX, cytochrome c, COX IV, and SMAC in total extracts, the cytosolic fraction, and mitochondrial fraction of Naa20^-/- MEFs six days after infection with AdEmpty (WT) or Ad5CMVCre (^-/-), under basal conditions. (B) As in (A) 12 h after 50 µM etoposide treatment. (C) As in (A) 8 h after 5 µM MG132 treatment. WT and Naa20^-/- MEF cells were fractionated 6 days after adenovirus inoculation. Cytosolic GAPDH and mitochondrial COX IV were used as loading controls of cytosolic and mitochondrial fractions, respectively. For the immunoblots, one representative of two independent experiments is shown.

The Arg/N-degron pathway is involved in the reduction of procaspase-8, -9 and BID levels in Naa20^-/- MEFs

As BID and procaspase-8, -9, and -3 protein levels decreased in the absence of Naa20 (Figure 3A), we next assessed if the possible lack of the N-terminal acetyl group in NatB substrates was sensed as an Arg/N-degron and consequently targeted for degradation. To this end, we used siRNAs to silence E3 ubiquitin ligases of the Arg/N-degron pathway (Ubr1, Ubr2 and Ubr4). As it has been reported a functional complementarity between Arg/N-degron and Ac/N-degron pathways, we also used siRNAs to silence the Ac/N-recognins Cnot4 and March6 in Naa20^-/- and WT MEFs, and evaluated procaspase and BID protein levels (Figure 5A and Supplementary Figure 7A). Although we faced challenges in obtaining specific antibodies for the direct confirmation of N-recognins downregulation, we successfully validated their decreased expression through reverse transcription quantitative polymerase chain reaction (RT-qPCR) (Supplementary Table 2).

Figure 5.

Silencing the ubiquitin ligases Ubr4 and Ubr1 in Naa20^-/- MEFs rescues the decrease in procaspase-8 and -9 protein expression, respectively, as well as in BID protein expression and promotes the activation of procaspase-8 and procaspase-3 in response to TNF-α plus SMAC mimetic. (A) Representative Western blot images of procaspase-8, -9, and -3 and BID protein levels after silencing March6, Ubr4, Cnot4, Ubr1, and Ubr2 ubiquitin ligases in Naa20^-/- MEFs six days after infection with AdEmpty (WT) or Ad5CMVCre (Naa20^-/-). A specific siRNA (siControl) was used as control. (B) Representative Western blot images of the procaspases and cleaved caspase-8, -9, and -3 protein levels after silencing the Ubr4 ubiquitin ligase in NAA20 MEFs six days after infection with AdEmpty (WT) or Ad5CMVCre (^-/-), before (basal conditions) and 12 h after treatment with 100 ng/mL TNF-α plus 500 mM SMAC mimetic. PBS plus DMSO was used as a negative control of TNF-α plus SMAC mimetic. A nonspecific siRNA (siControl) was used for control. For the immunoblots, one representative of two independent experiments is shown.

While procaspase-3 levels were not positively affected by silencing any of the tested E3 ubiquitin ligases, Ubr4 silencing increased procaspase-8 levels almost comparable to the level in WT MEFs, while Ubr1 silencing slightly increased procaspase-9 in Naa20^-/- MEFs. Moreover, silencing Ubr4 or Ubr1 also increased BID protein levels. Interestingly, downregulation of Ubr2 negatively affected procaspase-8 and procaspase-3 expression in WT MEFs. Furthermore, silencing of March6 and Cnot4 did not affect the expression levels of procaspases and BID.

Ubr4 silencing in Naa20^-/- MEFs partially reverts the decrease of procaspase-8 expression and its activation in response to TNF-α plus SMAC mimetic

Our data suggest that the procaspase-8, -9 and BID N-termini may be recognized as degradation signals when Naa20 is inactivated. We then wanted to explore the physiological relevance of these NAA20-dependent degradation signals on the induction of the two major apoptotic pathways (e.g., extrinsic and intrinsic ones). Therefore, we investigated whether the procaspase-8 and BID increase observed after silencing of Ubr4, and the increase in procaspase-9 and BID after silencing of Ubr1, could rescue procaspase activation in Naa20^-/- MEFs treated with TNF-α plus SMAC mimetic or etoposide, respectively. In addition to restoring procaspase-8 accumulation in Naa20^-/- MEFs, Ubr4 silencing also activated caspase-8 12 h after TNF-α plus SMAC treatment, as it recovered and increased caspase-8 p43 polypeptide production and caspase-8 p18 polypeptide accumulation, respectively (Figure 5B and Supplementary Figure 7B). Ubr4 downregulation also promoted slight caspase-3 activation in Naa20^-/- MEFs. Nonetheless, procaspase-3 activation was less pronounced than procaspase-8 activation. Therefore, the decrease in procaspase and caspase-8 in response to TNF-α plus SMAC mimetic caused by Naa20 inactivation can be partially reverted by silencing the Ubr4 E3 ubiquitin protein ligase. Conversely, although Ubr1 silencing restored procaspase-9 accumulation in Naa20^-/- MEFs (Figure 5A), etoposide treatment barely induced procaspase-9 activation in the double mutant (Supplementary Figure 8A). Of note, no caspase-9 activation was observed when silencing Ubr4 in response to etoposide (Supplementary Figure 8B).

Discussion

By targeting around 21% of the proteins, NatB is the second major contributor to the Nt-acetylome and its inactivation or downregulation affects several cellular processes such as actin cytoskeleton function,^10,20,21,42 stress responses,^13,43 proliferation,^22,23,26 and apoptosis.^26,27 Although NatB is known to regulate apoptosis in MEFs and tumor cells, the association between NatB and apoptosis regulation and its interdependence with N-degron pathways have not previously been investigated. Curiously, several main components of intrinsic and extrinsic apoptosis pathways are NatB substrates based on their second amino acid identity (Supplementary Figure 1). Accordingly, BAX and procaspase-3 are Nt-acetylated by NatB in human cells.¹⁶ In addition, we recently showed that human BAX acetylation by NatB is essential to maintain BAX in an inactive conformation in the cytosol of MEFs.²⁷ Thus, to further reveal how apoptosis is modulated by NatB-mediated Nt-acetylation, we explored the effect of the catalytic subunit Naa20 inactivation on protein and gene expression of several apoptotic pathway components. We observed a strong and selective decrease in the NatB substrates procaspase-8, -9, and -3 and BID, but not procaspase-6 and SMAC, both of which are NatA substrates. Nevertheless, when Naa20 was inactivated, Apaf1 and to lesser extent Bax mRNAs were upregulated while procaspase-8 and -3 and Bid were not affected. In D. melanogaster, Naa20 knockdown decreased protein expression of NatB substrate Drk, while accumulation of its mRNA was unaffected.⁴⁴ In sum, decreases in procaspase-8 and -3 and BID proteins in Naa20^-/- MEFs is not caused by transcriptional repression of their genes or through destabilized mRNA, suggesting a mechanism of selective translational repression or instability of these proteins.

Our proteomic analysis is consistent with NatB not being involved in the overall control on the stability of its substrates, with the accumulation of most of these proteins being not affected by Naa20 inactivation. Instead, NatB-dependent NTA of specific proteins—particularly of the pro-apoptotic factors procaspase-8, -9, -3 and BID—prevents these proteins from degradation. Therefore, we addressed whether their reduced levels affected Naa20^-/- MEFs susceptibility to different apoptotic stimuli. Loss of apical initiators caspase-8 and -9 is known to block extrinsic and intrinsic apoptosis, respectively.^45,46 Our results revealed a decrease in procaspase-8, -9, and -3 activation in response to etoposide- and TNF-α plus SMAC mimetic-induced apoptosis in Naa20^-/- but not in WT MEFs, reflecting a negative impact of Naa20 inactivation on both pathways. Intriguingly, Naa20^-/- MEFs display basal levels of cleaved PARP (Asp214), which suggests the activation of caspase-3, even though the cleaved caspase-3 peptide remains undetectable. However, this particular state does not promote the induction of apoptosis upon cell exposure to etoposide and TNF-α plus SMAC mimetic treatment suggesting the dysfunction or inhibition of apoptotic signaling components. Interestingly, Naa20^-/- MEFs present a senescent state³⁹ that is consistent with the inactivation of PARP and the reduction of apoptosis induction observed in this work.

To understand whether this apoptosis reduction was related to impaired BAX activation, we assessed its subcellular localization and cyt c and SMAC mitochondrial release. In Naa20^-/- MEFs, 12 h after etoposide treatment, BAX mitochondrial translocation was similar to WT MEFs; however, cyt c and SMAC release was attenuated, indicating a potential inability to activate procaspase-9 in response to etoposide. As etoposide is a typical caspase-9-dependent drug,^45,46 the reduction in native and cleaved procaspase-9, as well as of cytosolic cyt c and SMAC mediated by Naa20 inactivation, likely affect etoposide’s ability to elicit apoptosis. Conversely, TNF-α, a pleiotropic ligand of TNFR1 and 2, promotes cell survival by activating NF-κB or cell death by activating procaspase-8⁴⁴. The observed decrease in procaspase-8 protein levels in the absence of NatB decreases caspase-8 activation which ultimately hampers the apoptotic cascade.⁴⁷

An early study investigating the biological function of the catalytic subunit NAA20 in human revealed increased susceptibility of HeLa cells to MG132 after Naa20 knockdown.²⁶ Consistently, our previous study with Naa25^-/- MEFs also revealed increased susceptibility to MG132²⁷. Moreover, another study performed in caspase-9^-/- MEFs revealed efficient MG132-induced but weak etoposide-induced apoptosis, which was restored by synergizing with active cytosolic SMAC.⁴⁸ Interestingly, the reduced initial protein levels of procaspase-8, -9 and -3 and their active forms in Naa20^-/- compared with WT MEFs did not promote a clear differential procaspase activation and PARP1 cleavage in response to MG132 or tunicamycin, neither affected BAX activation nor cyt c and SMAC release when the proteasome was inhibited. It is possible that inhibition of the proteasome by MG132 could counteract the degradation of the pro-apoptotic factors in the absence of NatB-mediated acetylation. Furthermore, proteasome inhibition and ER stress induction are two biological processes that can promote apoptosis through diverse molecular pathways. Importantly, these mechanisms can operate independently of apoptosome formation and can utilize either the extrinsic or intrinsic pathway. Together, these data indicate that the negative effect of Naa20 inactivation on apoptosis induction, likely due to reductions in procaspase-8, -9 and -3 and BID protein levels, is stimulus-dependent.

Since Nt-acetylation prevents protein degradation and plays a role in maintaining partial proteome stability in MEFs, we sought to identify the N-recognins involved in Naa20-mediated degradation of apoptosis effectors. Partial reversion of the decrease in procaspase-8, -9 and BID protein expression levels through Ubr1 or Ubr4 downregulation confirmed that their unacetylated N-terminus may be sensed as an N-degron and targeted for degradation. Accordingly, it has been documented that NatC activity serves to shield proteins from degradation by UBR4 and, to a lesser extent, by UBR1/UBR2.⁴⁹ Consequently, these findings collectively support the notion that the Arg/N-degron pathway primarily governs the stability of unacetylated N-termini through the UBR4 E3 ligase, while NatB and NatC activity act to safeguard their substrates from degradation.

Interestingly, Ubr4 silencing in Naa20^-/- MEFs mostly restored procaspase-8 expression and activation after TNF-α plus SMAC mimetic treatment. However procaspases-3 and -9 expression were just slightly recovered when the Ubr4 N-recognin was downregulated in Naa20^-/- MEFs and consequently caspase-3 and -9 activation was less pronouncedly restored in response to these stimuli. The partial recovery of phenotypes as a result of Ubr4 silencing can be due to the fact there is not a complete absence of the protein, possible involvement of other N-recognins and/or reflect that cleavage of BID and procaspase-3 by caspase-8 was insufficient to activate properly the initiator and effector caspases of the intrinsic pathway. Similarly, we show that UBR1 recognizes procaspase-9 as an N-degron, as its decreased protein level in Naa20^-/- MEFs was also partially restored after Ubr1 silencing. However, Ubr1 downregulation in Naa20^-/- MEFs did not promote procaspase-9 activation in response to etoposide. This implies that the observed partial recovery of procaspase-9 levels was inadequate to reinstate the activation of the intrinsic pathway. However, we cannot rule out the possibility that unidentified molecular mechanisms may also be involved in regulating the stability of NatB substrates when NatB is inactivated.

Taken together, our data reveal that NatB activity protects several apoptotic factors from degradation probably by specific UBR E3 ligases-mediated action. This further highlights the impact of NatB and Arg/N-degron pathway interdependence on apoptosis. Most interestingly, our findings indicate that the absence of NatB activity may lead to the exposure of unacetylated N-termini of procaspase-8, making them susceptible to UBR4-mediated degradation. This process could consequently decrease caspase-8 activation in response to extrinsic stimuli (Figure 6). Our study supports the relevance of NatB-dependent NTA for cellular proteostasis of key pro-apoptotic factors and consequent apoptosis, paving the way for new therapeutic strategies.

Figure 6.

Model depicting the role of NatB-dependent Nt-acetylation in the control of extrinsic apoptosis induction. Procaspase-8 and procaspase-3 are NatB substrates Nt-acetylated in normal cells. TNF-α binding to its receptor in normal cells promotes procaspase-8 activation, which activates procaspase-3 and promotes apoptosis. Inactivation of the NatB catalytic subunit (Naa20 KO) reduces procaspase-8 Nt-acetylation, and therefore E3-ubiquitin ligase UBR4 recognizes procaspase-8 N-terminus as an N-degron, marks, and sends procaspase-8 for proteasomal degradation. Procaspase-3 is also less stable when Naa20 is inactivated, but the E3-ubiquitin ligase has not been identified. Therefore, binding of TNF-α to its receptor when NatB is inactivated limits apoptosis induction as a consequence of reduced procaspase-8 and -3 in these cells. Image created with Biorender.

Materials and Methods

Establishment of mouse embryonic fibroblast strains

MEFs were obtained from Naa20^{tm1a(KOMP)Wtsi} transgenic mice generated at the transgenesis unit at the CNB-CSIC using embryonic stem (ES) cells provided by the Knockout Mouse Program (KOMP). Embryos isolated from the uteruses of pregnant mice were transferred to petri dishes containing sterile PBS. The head and internal organs were removed from embryos. Each embryo was transferred to a well of a six-well plate with trypsin/EDTA, and plates were incubated for 30 min at 37 °C. Then, embryos were pulled through an 18-gauge needle to disaggregate them to isolate fibroblasts. Cells were grown in a 100 mm petri dish with Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (P/S). When the cells reached confluency, they were replated in 60 mm petri dishes and immortalized the next day by retroviral infection with a recombinant virus expressing the SV40 large T antigen and zeocin resistance gene. The infection process was repeated after 24 h to increase efficiency. At confluency, cells were harvested and plated in 100 mm petri dishes. Forty-eight hours later, zeocin (200 µg/mL) was added to the plates for selection of T antigen-expressing cells. Cells were incubated in medium with antibiotic replaced every two days to remove dead cells and refresh the antibiotic. When cell clones were clearly observed, they were harvested with a micropipette and plated in 96-well plates to be grown separately. Each clone was amplified and, once established, they were used in downstream experiments.

Inactivation of Naa20 in MEFs

Naa20 was inactivated with a recombinant adenovirus expressing CRE recombinase (Ad5CMVCre, MOI 2000). An empty adenovirus was used as a negative control (AdEmpty, MOI 2000). Recombinant adenovirus was inoculated in DMEM supplemented with 2% FBS and 1% penicillin/streptomycin and, 24 h later, regular DMEM was added. Cells were trypsinized and replated, two or five days post-infection.

Real-time PCR

Total RNAs were extracted with the Maxwell^® RSC simplyRNA kit (Promega, Madison, WI). Reverse transcription was performed as previously reported.²⁶ Real-time PCR was performed with the iQ SYBR Green Supermix (Bio-Rad, Hercules, CA) in a CFX96 Real-Time System (Bio-Rad) using specific primers for each gene: Naa20 WT F 5′-CCTTCACCTGCGACGACCTGTT R 5′-GGAATTCAGGGGCGACAGAG; Histone F 5′-AAAGCCGCTCGCAAGAGTGCG R 5′-ACTTGCCTCCTGCAAAGCAC; Caspase-3 F 5′TACATGGGAGCAAGTCAGTGG R 5′CACATCCGTACCAGAGCGAG; Caspase-8 F 5′TCAGAAGAAGTGAGCGAGTTGG R 5′ATCCTCGATCTTCCCCAGCA; Caspase-9 F 5′GGGAAGATCAGGGGACATGC R 5′TCTTGGCAGTCAGGTCGTTC; Apaf1 F 5′CTCCTTGGACGACAGCCATT R 5′AAACACGCGTGGTAAACAGC; Bax F 5′ACCAAGAAGCTGAGCGAGTG R 5′ATGGTTCTGATCAGCTCGGG; and Bid F 5′-GGCGTCTGCGTGGTGATTC R 5′-CCAGTAAGCTTGCACAGGCA. Transcript was quantified using the formula: 2^{ct(β−Histone)−ct(gene)}, with ct being the point at which the fluorescence rises substantially above background fluorescence.

Immunofluorescence and confocal microscopy

For immunofluorescence experiments, cells were seeded in six-well plates containing glass coverslips. Six days postinfection, cells were fixed with 4% paraformaldehyde (PFA, 16% formaldehyde solution; Thermo Fisher Scientific, Waltham, MA) for 15 min at room temperature (RT). After rinsing with 1× PBS, cells were permeabilized with 0.1% Triton X-100 (Sigma-Aldrich, St Louis, MO) over 15 min at RT and washed. Following this process, cells were incubated with anti-α-vinculin (V9131, Sigma-Aldrich) antibodies and Alexa Fluor 488-phalloidin (Invitrogen, Waltham, MA) in TBS-Tween 20 (TBST 1×)-3% BSA for 30 min, 37 °C. Subsequently, cells were washed and incubated with anti-mouse IgG Cy3 conjugate developed in sheep (Sigma-Aldrich) or anti-rabbit IgG Alexa488 conjugate developed in donkey (Molecular Probes, Eugene, OR) over 30 min at 37 °C. After mounting the coverslips in Vectashield mounting medium with DAPI (Vector Laboratories, Newark, CA), samples were maintained at 4 °C until visualization. Images were acquired with an Axiovert 200 M confocal LSM 510 META Zeiss microscope using a 40× objective.

Protein extraction

Cells were harvested, and 2× Laemmli Sample Buffer (Bio-Rad) plus β-mercaptoethanol with sodium dodecyl sulfate (SDS) 2% (Bio-Rad) supplemented with Tris-HCl 0.5 M pH 7.4 and protease inhibitors (1 mM PMSF, 0.001 mg/mL aprotinin, 1 mM sodium orthovanadate, and 1 mM sodium pyrophosphate (Roche, Basel, Switzerland)) was used to collect cell lysates. The soluble protein concentration was normalized using the Revert 700 Total Protein Stain (LI-COR) and Image Studio Lite Software (LI-COR, Lincoln, NE).

Western blotting

Protein samples were separated by SDS polyacrylamide gel electrophoresis and transferred onto nitrocellulose membranes (Bio-Rad). Next, to avoid nonspecific interactions, membranes were blocked in 5% BSA or 5% non-fat milk in 1× TBST for 30 min at RT with agitation. Membranes were then incubated 1 h at RT with primary antibodies, namely mouse monoclonal anti-GAPDH (AbD Serotec, Bio-Rad), mouse monoclonal anti-α-tubulin (Sigma-Aldrich), and rabbit polyclonal anti-Nat5 (ProteinTech). Rabbit polyclonal anti-caspase-3, anti-cleaved caspase-3, anti-cleaved caspase-9, anti-cleaved caspase 8, anti-BAX; rabbit monoclonal anti-caspase-8, anti-Cyt c, anti-APAF1, anti-COX IV, anti-caspase-6, anti-SMAC; and mouse monoclonal anti-caspase-9 and anti-cleaved PARP (Asp214) were all purchased from Cell Signaling Technology (Danvers, MA). Rat monoclonal anti-BID (truncated and full length) was purchased from R&D systems. Subsequently, membranes were incubated with secondary anti-mouse IgG or anti-rabbit IgG antibodies (Cell Signaling Technology). Chemiluminescence detection was performed using the ECL Ultra detection system (Lumigen, Southfiled, MI) and an Odyssey^® Fc Imaging System (LI-COR).

Cell proliferation assay

Cells were seeded in six-well plates and infected the day after. Then, cells were trypsinized and replated two and five days after infection, as described above. Cell counting was performed at days 2, 3, 4, 5, and 6 postinfection, and four wells per condition/per day were used. Cells were counted using a TC20^™ Automated Cell Counter (Bio-Rad).

MEF fractionation

Cells were seeded in six-well plates, infected, and, six days postinfection, cells were harvested (basal conditions) or treated with 50 µM etoposide and 5 µM MG132 for 12 h and 8 h, respectively. Then, 6.6 × 10⁶ cells per condition were harvested and cell fractionation was performed using a Cell Fractionation kit (Abcam, Cambridge, UK; AB109719) following the manufacturer’s guidelines. 1 × 10⁶ cells per condition were used for total extracts. 4× Laemmli Sample Buffer (Bio-Rad) plus β-mercaptoethanol with SDS 2% (Bio-Rad) supplemented with Tris-HCl 0.5 M pH 7.4 and protease inhibitors was used to collect cellular fractions. GAPDH and COX IV were used as cytosolic and mitochondrial controls, respectively. Protein bands were quantified Image Studio Lite Software System (LI-COR).

Apoptosis assays

Cells were seeded in six-well plates, infected, and six days postinfection treated with 50 µM etoposide, 100 ng/mL TNF-α plus 500 nM SMAC, 5 µM of MG132, or 5 ng/µL of tunicamycin for 0, 8, 12, and 24 h. At each time point, cells were harvested and collected for protein extraction for further analysis by Western blotting.

N-terminal acetylation assays

Eight biological replicates of WT or Naa20^-/- MEFs (10⁶ cells) were harvested and six days after adenovirus infection with AdEmpty and Ad5CMVCre, respectively, were centrifuged at 6000 rpm for 5 min at 4 °C, and pellets were washed with PBS 1× and centrifuged again. Next, cell pellets were resuspended in 300 µL of protein extraction buffer (50 mM HEPES/NaOH pH 7.2; 1.5 mM MgCl₂; 1 mM EGTA; 10% glycerol; 1% Triton X-100; 150 mM NaCl; 2 mM PMSF; 1 protease inhibitor cocktail tablet (EDTA+) in 50 mL), subjected to lysis by sonication, centrifuged at 17,000 rpm for 20 min at 4 °C, and the supernatant recovered for proteomic analysis.

For each sample, 1 mg of protein was used following the SILProNAQ protocol previously described.²⁹ Briefly, N-acetoxy-[²H₃] succinimide was used to label protein N-termini and lysine ε-amino groups before being digested by trypsin. Peptides were then fractionated by chromatography using a strong cation exchange (SCX) column to discriminate acetylated and unacetylated internal peptides. Individual fractions 2-11 were analyzed by LC-MS using 55 min methods on the LTQ-Orbitrap Velos mass spectrometer (Thermo Fisher Scientific, Waltham, MA), and three pools of fractions (2-5, 6-8, and 9-11) were analyzed using 100 min methods on the TIMS-ToF (Bruker, Billerica, MA). Data were searched against the Mus musculus SwissProt database through Mascot Distiller then parsed using the in-house eNcounter script to validate the detected N-termini and obtain the Nt-acetylation yield.⁵⁰

Label-free proteome quantitation

In parallel with the N-terminal acetylation assay, the same MEF samples were used for full proteomic analysis. Thirty micrograms of protein extract were loaded on an SDS-PAGE gel and digested following an in-gel digestion protocol.⁵¹ The resulting peptides were extracted, dried, and then analyzed on the TIMS-ToF mass spectrometer using the same 100 min method as before. Data were processed using MaxQuant 2.3.9.⁵² to identify and quantify. Statistical analysis were achieved by processing the exported data with Perseus 2.0.⁵³ Only the proteins with a minimum of two valid quantitation values in at least one of the conditions were retained. Then the P values were calculated through the use of the two-sample Student’s t test function, with an S0 value maintained at 0, a minimum number of valid values set to 1, and a P value threshold of 0.05. The resulting matrix was exported and finalized with Excel.

siRNA silencing of E3 ubiquitin ligases

WT or Naa20^-/- MEF cells were seeded in six-well plates and infected the day after with the corresponding adenovirus as described above. Two days later, cells were trypsinized and replated as described previously and, 4 days postinfection, cells were replated and transfected with the corresponding siRNAs. 150 000 cells/mL were transfected in suspension with the siRNAs and the Lipofectamine^® RNAiMAX Transfection Reagent (Invitrogen, Thermo Fisher Scientific 13778) according to manufacturer’s instructions. Transfected cells were collected 48 h posttransfection for Western blot and real-time PCR analysis.

The following siRNAs were used: siRNA Silencer^™ Select Negative Control No. 1 (4390843), Silencer^® Select Ubr1 siRNA (4390771-s75706), Silencer^® Select Ubr2 siRNA (4390771-s104948), Silencer^® Select Ubr4 siRNA (4390771-s87462), Silencer^® Select C4 siRNA (4390771-s79257), and Silencer^® Select March6 siRNA (4390771-s104524), all from Thermo Fisher Scientific.

Statistical analysis

Statistical analyses were performed using Prism 8 (GraphPad Software, La Jolla, CA). Normality of groups was assessed with the Shapiro–Wilk test. The unpaired t test was used to compare parametric data the Mann–Whitney test for nonparametric data. Two-way ANOVA was used to analyze proliferation data. The significance of enriched NatB substrates in different MS quantifications was calculated using the built-in two-tailed t test function for comparison of two samples with equivalent variances in Microsoft Excel for the N-terminome analysis, while a two sample Student t test, with a threshold P value of 0.05, was used in Perseus for the label-free proteome analysis.

Funding Statement

This work was supported by the KatNat (ERA-NET, ANR-17-CAPS-0001-01) and CanMore (France-Germany PRCI, ANR-20 CE92-0040) grants funded by the French National Research Agency (ANR) to C.G. to support JB.B, by Foundation ARC (ARCPJA32020060002137) grants to T.M., by ISCIII Consolidation Program Grant and Ministerio Español de Economía y Competitividad Torres Quevedo Program (PTQ-13-06466) to support RA, by Departamento de Desarrollo Económico del Gobierno de Navarra (0011-1383-2018-000011) grant to RA, from the facilities and expertise of the I2BC proteomic platform (Proteomic-Gif, SICaPS) supported by IBiSA, Ile de France Region, Plan Cancer, CNRS and Paris-Saclay University, and from ProteoCure COST (European Cooperation in Science and Technology) action CA20113. The proteomic experiments were partially supported by Agence Nationale de la Recherche under projects ProFI (Proteomics French Infrastructure, ANR-10-INBS-08) and GRAL, a program from the Chemistry Biology Health (CBH) Graduate School of University Grenoble Alpes (ANR-17-EURE-0003). Joana P. Guedes acknowledges the PhD fellowship SFRH/BD/132070/2017 funded by FCT.

Data Availability

Mass-spectrometry based proteomics data are deposited in ProteomeXchange Consortium (https://proteomecentral.protemeexhange.org) via the PRIDE repository (https://www.ebi.ac.uk/pride/) with the data set identifier PXD029641, username: reviewer_pxd029641@ebi.ac.uk, Password: O8uvsKWw.

Author Contributions

Conceptualization, R.A, C.G. and M.C.R.; methodology design, C.G., R.A., M.C.R., T.M., JB.B., J.P.G., J.E., and B.C.; formal analysis, C.G., R.A., M.C.R., T.M., JB.B., J.P.G., J.E., B.C., V.R. and L.S.; investigation, C.G., R.A., M.C.R., T.M., JB.B., J.P.G., J.E., B.C., V.R. and L.S.; writing original draft, R.A, C.G. M.C.R. and J.P.G.; writing-review and editing, C.G., R.A., M.C.R., J.P.G., T.M., and JB.B.; visualization, C.G., R.A., M.C.R., J.P.G., T.M., and JB.B.; supervision, R.A, C.G. and M.C.R; project administration, R.A, C.G. and M.C.R; funding acquisition, R.A, C.G., and T.M.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Materials and Correspondence

Correspondence and material requests should be addressed to Manuela Côrte-Real, Carmela Giglione, and Rafael Aldabe.