Acylamino acid-releasing enzyme, a bifunctional protease with a potential role in aging

Abstract

Acylamino acid-releasing enzyme (AARE) is a highly evolutionary conserved, bifunctional serine protease. In its exopeptidase mode, AARE cleaves N-terminally acetylated or otherwise blocked amino acids from the N-terminus of peptides, and probably even intact proteins. In its endopeptidase mode, AARE cleaves oxidised proteins at internal positions. Although AARE function was discovered 50 years ago and has been biochemically characterized in various organisms, the precise role of this protease in cellular physiology remains elusive. Several other names for AARE do exist in literature, such as acylpeptide hydrolase, acylaminoacyl peptidase, and oxidised protein hydrolase. Recently, the first AARE null mutants have been described in the model moss Physcomitrium patens (Physcomitrella). Comparisons with T-DNA mutants in Arabidopsis thaliana revealed a role for AARE in the timing of the developmental transition from the vegetative to the reproductive state, as well as in the determination of life span. Loss of AARE function was accompanied by a striking increase in oxidised proteins, a hallmark of cellular aging. In mammals, AARE activity is linked to proteasomal function, and dysregulation of AARE function has been observed in different types of cancer and age-related pathologies. Here, we compile the current knowledge on molecular and biological functions of this protease, aiming to derive common roles of AARE in cellular physiology, and potentially in aging, but also highlight differences between species isoforms.

We consolidate information on Acylamino acid-releasing enzyme, a serine protease affecting plant development and aging, and emphasize its conserved features across all kingdoms of life.

Introduction

All living beings are subjected to aging. While we typically link the term ‘aging’ to phenotypical transitions such as from child to adult, and to signs of senescence or physiological deterioration, aging also affects organisms on a molecular level, aside from such obvious phenotypical changes. Moreover, for multicellular organisms, aging also encompasses all developmental steps from the first cell division to organ development and beyond. In a molecular sense, aging may simply be the progression of biochemical reactions within a living cell over time, until an imbalance in the reaction system results in the loss of vital cellular functions. These functions include energy metabolism, catabolism, excitability, reproduction, or division, with imbalance ultimately leading to the destruction of the cell and its contents. This process is largely predetermined by the organism’s genetic code and influenced by environmental factors, such as stressors and nutrient availability. AARE is a bifunctional serine protease that counteracts the accumulation of cytotoxic, oxidatively damaged proteins generated by reactive oxygen species (ROS). In this article, we aim to highlight and discuss a protease that has been relatively understudied in terms of its role in proteostasis and cellular physiology: acylamino acid-releasing enzyme (AARE, EC: 3.4.19.1).

Reactive oxygen species

Throughout their lifetime, all living cells produce reactive oxygen species (ROS) as continuous metabolic by-products or as a consequence of stress (de Almeida et al., 2022). Stress can also lead to increased ROS production in cells. ROS, which are non-radical derivatives or radical forms of molecular oxygen, fulfil indispensable signalling functions when present at physiological levels, but lead to severe molecular damage if present in excess (Sies and Jones, 2020). Due to these contrary functions, cells contain ROS-sensing mechanisms and have several lines of defence to intercept generated ROS to avoid their accumulation to detrimental levels. Plants, as photosynthetically active organisms, have, in general, a higher abundance and broader diversity of antioxidative components than other organisms (Kasote et al., 2015). Regardless of continuous ROS scavenging, oxidative damage within the cell is inevitable and molecules, such as DNA, lipids and proteins, undergo both reversible and irreversible oxidative modifications (Murphy et al., 2022).

Eukaryotic cells contain several major sites of ROS production. In Ophistokonts, such as humans or yeast, the respiratory chain in mitochondria is the main producer, while photosynthetic reaction centres are the major source of ROS in Viridiplantae (Asada et al., 1974; Telfer et al., 1994; Hideg et al., 1998; Balaban et al., 2005; Asada, 2006; Guo et al., 2023). Apart from plastids and mitochondria, eukaryotic cells contain several further production sites, such as the cytosol, and, depending on the nature and the related half-life of a certain molecule species, ROS can diffuse through the cell and across compartments. The diffusion of ROS species within a cell is an integral part of intracellular and inter-organellar signalling. ROS signalling is well known to work via reversible thiol redox switches of solvent-exposed cysteine residues, among other mechanisms, in plants and humans (Mittler et al., 2022; Sies et al., 2024).

ROS production can dramatically increase under conditions of stress or as part of certain pathologies. In plants, ROS production bursts in response to pathogen infections and under high light conditions (Lehtonen et al., 2012; Koselski et al., 2023), but also notably increases during seed aging and at the onset of senescence (Prochazkova et al., 2001; Exposito-Rodriguez et al., 2017; Mao et al., 2018; Yuan et al., 2021). Furthermore, ROS production increased during aging and during developmental transitions from proplastids towards their mature form (Munné-Bosch and Alegre, 2002; Müller-Schüssele et al., 2020; Tripathi et al., 2020). In mammals, accumulation of oxidative damage caused by increasing mitochondrial ROS levels has been proposed as one of the main drivers of aging (Harman, 1956, 1972). This hypothesis has been supported by many studies, although contrary indications exist (Shields et al., 2021). Along with aging, elevated ROS levels in mammals are also associated with several pathologies, such as neurodegenerative and inflammatory diseases (Liu et al., 2017; Checa and Aran, 2020). Effectively, ROS cause oxidative damage on many molecules, among which proteins are the major targets. ROS is an umbrella term for a variety of oxygen derivatives (Sies and Jones, 2020), while protein oxidation summarizes a broad range of modifications to proteins in the presence of ROS (Møller et al., 2007).

Protein modifications and aging

Amino acids, mainly cysteine, methionine, proline, threonine, arginine, lysine, histidine and tryptophan, are affected by ROS-catalysed modifications. While some of these modifications, such as cysteine side chain oxidation to sulphenic and sulphinic acid or the formation of methionine sulphoxide, are enzymatically reversible, the formation of carbonylated side chains (protein carbonylation) is an irreversible reaction. Ultimately, cysteine and methionine can also reach irreversible oxidised states. The protein backbone may represent an additional target for ROS-induced cleavage (Davies, 1996; Reeg and Grune, 2015). Apart from protein modifications, ROS-induced lipid peroxidation in mammals and plants results in the formation of reactive carbonyl species (RCS) such as malondialdehyde (MDA) and 4-hydroxy-2-nonenal (HNE), which in turn can irreversibly modify proteins (Esterbauer et al., 1991; Møller et al., 2007; Ayala et al., 2014; Milkovic et al., 2023). The biological relevance of these RCSs is not yet fully understood, but a role in signalling is repeatedly discussed (Tola et al., 2021; Milkovic et al., 2023).

In contrast, protein carbonylation and irreversible oxidation of sulphurous amino acids are generally regarded as unwanted and damaging, having not yet been assigned signalling functions. The introduction of side chain carbonyls in proteins results in an altered topology and increased hydrophobicity, which ultimately leads to disruption of protein function and result in protein aggregation (Chao et al., 1997; Tanase et al., 2016). Such dysfunctional proteins need to be cleared to sustain cellular homeostasis and to prevent the formation of cytotoxic aggregates. In eukaryotes, dysfunctional proteins are largely degraded in a ubiquitin-dependent manner via the 26S proteasome (Vierstra, 2009; Bard et al., 2018). Moreover, oxidised proteins can also be degraded in a ubiquitin- and ATP-independent manner in the 20S proteasome in plants (Oracz and Stawska, 2016) and mammals (Raynes et al., 2016). If strongly oxidised or clearance via the proteasome fails, oxidised proteins tend to form cytotoxic aggregates. Accumulation of proteinaceous aggregates can further impair both ubiquitin-dependent and independent proteasomal activity, ultimately triggering cell death (Grune et al., 2004; Thibaudeau et al., 2018). In humans, several age-related pathologies coincide with the deposition of cytotoxic aggregates (Guo et al., 2022). To counteract aggregate deposition, mammals and plants engage chaperons as molecular disaggregation machineries to solubilize existing aggregates, enable refolding, or degradation of the contained proteins (Parsell et al., 1994; Queitsch et al., 2000; Mogk et al., 2018). Finally, if solubilization fails, aggregates can be cleared via autophagy. However, although ROS are likely inducers of autophagy, the autophagic flux declines with aging (Aman et al., 2021; Redza-Dutordoir and Averill-Bates, 2021). Consequently, increasing protein oxidation and aggregate formation, and reduced autophagy capacity are all hallmarks of aging (Sohal and Weindruch 1996; López-Otín et al., 2013; Aman et al., 2021).

Another undesired post-translational modification of proteins is glycation. Glycation is a non-enzymatic reaction of reducing sugars or dicarbonyls with amino acid side chains, which can be spontaneous or in the presence of ROS (Rabbani and Thornalley, 2021; Guo et al., 2022). Successive reactions of glycated proteins lead to the formation of advanced glycation end products (AGEs). Like protein oxidation, glycation ultimately results in unfolding and dysfunction of the modified protein. AGE formation positively correlates with aging and is also associated with neurodegeneration, diabetes, and other pathologies, but is also present in plants (Peppa et al., 2008; Delgado-Andrade, 2016; Rabbani et al., 2020). In plants, various glycation products such as N_ε-carboxyethyl-lysine (CEL), glyoxal-derived hydroimidazolone 1 (G-H1), and N_ε-fructosyl-lysine (FL) form under physiological conditions, undergo dynamic diurnal changes, and increase during leaf aging (Bechtold et al., 2009; Chaplin et al., 2019). Oxidative stress, such as high light, further increases the levels of G-H1 but also of another glycation product, N_ε-carboxymethyl-lysine (CML) (Bechtold et al., 2009). Glycated peptides can act as signalling molecules in mammals, but this function has not yet been well investigated in plants (Delgado-Andrade, 2016; Rabbani et al., 2020). Instead, glycation of Rubisco rendered increased susceptibility for proteolysis (Yamauchi et al., 2002).

Hence, ROS-induced protein oxidation and glycation pose a progressive challenge to cellular homeostasis in all living organisms throughout their lifespan. Nevertheless, evidence is accumulating that protein carbonylation can also fulfil important signalling functions, apart from sole broad level detrimental effects (Tola et al., 2021). In eukaryotes, many aspects and mechanisms of ROS generation and quenching, protein oxidation, and glycation, as well as of the clearance of oxidised proteins, are largely conserved and under extensive research.

AARE—a dual mode protease

AARE is a bifunctional serine (S9) protease with a Ser/Asp/His catalytic triad belonging to the family of prolyl oligopeptidases (Mitta et al., 1998; Polgár, 2002; Rawlings et al., 2018). Its activity is conserved in species covering all kingdoms of life (Tsunasawa et al., 1975; Mori et al 1990; Ishikawa et al., 1998; Yamauchi et al., 2003; Brunialti et al., 2011; Elahi et al., 2019). While AARE isoforms are localized to both the cytosol and nucleus in plants and mammals, their localization to plastids and mitochondria—the primary sources of ROS—seems to be conserved throughout the plant lineage (Shimizu et al., 2003; Nakai et al., 2012; Zeng et al., 2017; Hoernstein et al., 2023). AARE operates in a dual function mode, exhibiting an exopeptidase activity on N^α-acetylated peptides and an endopeptidase activity on oxidised or glycated proteins. Although purified isoforms from various organisms have been characterized biochemically, understanding the precise role of AARE in cellular physiology remains in its early stages. In a recent publication, we characterized AARE knockout mutants in the model moss Physcomitrium patens (Physcomitrella) (Lueth and Reski, 2023) and T-DNA mutant lines in Arabidopsis (Hoernstein et al., 2023). This work revealed the striking contribution of AARE function in the timing of developmental transitions from the vegetative to reproductive state in both species, but also in the determination of their life spans. Mutant phenotypes were accompanied by a substantial increase in the level of oxidised proteins (Nakai et al., 2012; Hoernstein et al., 2023), a well-known hallmark of aging in both plants and mammals. Additionally, this increase is a side effect of the onset of bolting in Arabidopsis (Johansson et al., 2004). In mammalian cell lines, overexpression of AARE protects cells while leading to reduced levels of oxidised proteins under conditions of oxidative stress, and is coordinated with proteasomal activity (Shimizu et al., 2003, 2004; Palmieri et al, 2011). Here, we summarize the extensive knowledge on the biochemical activity of AARE isoforms across various organisms and discuss it in the context of the still limited understanding surrounding the role of AARE in organismal physiology and its potential involvement in the molecular process of aging.

AARE activity was first found by Tsunazawa et al. (1970) in Rattus norvegicus (rat) liver extracts, where AARE cleaved Acetylmethionine-threonine (AcMet-Thr) to AcMet and Thr. A working hypothesis at that time suggested protein biosynthesis at ribosomes would be initiated with acetylamino acid-loaded tRNA (Narita et al., 1968), but it was already known that many proteins were present with free N^α-amino groups in biological samples. Consequently, it was believed that an enzyme with the capacity to liberate N^α-acetylated amino acids from peptides or proteins emerging from the ribosome would exist, and efforts were made to identify the responsible enzyme. Similar activity on N-formyl-Met-Val was observed in Ovis aries (sheep) and Oryctolagus cuniculus (rabbit) erythrocytes (Witheiler and Wilson, 1972; Yoshida and Lin, 1972). In 1975, Tsunasawa and co-workers reported the biochemical characterization of an enzyme purified from rat liver, which they named ‘acylamino acid-releasing enzyme’ (Tsunasawa et al., 1975). In this study, they observed exopeptidase activity on the N-terminus of ovalbumin (AcGly-Ser-Gly) and cytochrome C from Saccharomyces oviformis (AcThr-Glu-Phe). Furthermore, they observed an increased cleavage of the N-terminal acetylated amino acid (AcGly) of ovalbumin after denaturation. In the following years, AARE isoforms were purified from various organisms and tissues, including bovine (Bos taurus), porcine (Sus scrofa domesticus) and rat liver, human erythrocytes, rat brain, and the yeast Rhodotorula glutinis. Their biochemical properties have been partially characterized, primarily using short, synthetic N^α-acetylated peptides (Gade and Brown, 1978; Marks et al., 1983; Tsunasawa et al., 1983; Kobayashi and Smith, 1987; Mori et al., 1990; Jones et al., 1991). Additionally, the primary amino acid sequence of the porcine liver isoform was determined (Mitta et al., 1989), which later facilitated the identification of the catalytic triad Ser/Asp/His (Miyagi et al., 1995; Mitta et al., 1998). The molecular weight of rat and porcine AARE was estimated to be approximately 75 kDa, and were found to assemble as tetramers (Tsunasawa et al., 1975; Mitta et al., 1989).

Beppu et al. (1994) investigated human erythrocytes under oxidative stress and found the presence of membrane-bound proteases that specifically cleave oxidised proteins. Accordingly, Fujino et al. (1998) purified a protease of around 80 kDa, which exhibited strong activity on oxidised and on glycated bovine serum albumin (BSA). This protease, called oxidised protein hydrolase (OPH), was present in the cytosol of erythrocytes, but was also associated with membranes, especially under oxidative conditions. Partial amino acid sequence determination and immunoblotting revealed close homology between OPH and AARE (Fujino et al., 2000a). Further reciprocal experiments with recombinant AARE and purified OPH revealed that AARE and OPH are in fact the same protease (Fujino et al., 2000a). Incubation of both enzymes generated similar degradation patterns of glycated or oxidised BSA on SDS-PAGE, indicating intrachain cleavages on the target. In summary, AARE/OPH is a bifunctional protease that cleaves N^α-acetylated amino acids via an exopeptidase mode, whereas oxidised or glycated proteins are processed via an endopeptidase mode.

Several different names for AARE were introduced, such as acylpeptide hydrolase (APEH/ACPH), acylaminoacyl peptidase (AAP) and oxidised protein hydrolase (OPH). Here, we will continue using the abbreviation AARE, which stands for acylamino acid-releasing enzyme, as originally referred to in the first publication of this protease (Tsunasawa et al., 1975).

Conservation of AARE function and structure

Until now, AARE function has been found in organisms from all kingdoms of life. Typically, presence of AARE function is proven using its exopeptidase activity towards artificial N^α-acetylated test substrates, such as AcAla-pNA and AcLeu-pNA (Fig. 1). The endopeptidase function is typically analysed by visualization of AARE cleavage products of artificially oxidised or glycated test proteins, or via detection of protein oxidation through derivatization with 2,4-dinitrophenylhydrazine (Fig. 1, DNPH) (Fujino et al., 2000a; Yamauchi et al., 2003; Nakai et al., 2012; Hoernstein et al., 2023). Using the analysis of the exopeptidase activity, AARE isoforms were identified in the archaea Aeropyrum pernix, Pyrococcus horikoshii, and Sulfolobus solfataricus (Ishikawa et al., 1998; Kiss et al., 2007; Gogliettino et al., 2012), in the bacteria Sporosarcina psychrophila and Geobacillus stearothermophilus (Brunialti et al., 2011; Chandravanshi et al., 2024), in yeast Rhodotorula glutinis (Mori et al., 1990), in rats, humans, and the silkworm Bombyx mori (Kobayashi and Smith, 1987; Scaloni et al., 1994; Fu et al., 2016), in the fish species Chionodraco hamatus, Dicentrachus labrax and Trematomus bernacchii (Gogliettino et al., 2014; Riccio et al., 2015), and in the plant species Arabidopsis thaliana, Physcomitrium patens, and Cucumis sativus (Yamauchi et al., 2003; Nakai et al., 2012; Hoernstein et al., 2023). In protists, the situation is not fully clear, but AARE function seems to be required for metabolic functions after host cell infection in Plasmodium falciparum (Elahi et al., 2019). Intriguingly, it seems that this parasite internalizes a proteolytically active AARE fragment of the human host cell.

Fig. 1.

Schematic representation of acylamino acid-releasing enzyme’s (AARE) dual operation mode. AARE cleaves N-terminally acetylated amino acids (AA) from polypeptides via its exopeptidase function. The structure of the human AARE monomer (P13798), predicted with AlphaFold (Jumper et al., 2021; Varadi et al., 2024), is depicted at the top. Exopeptidase activity is assayed by monitoring the release of para-nitroanilide (pNA) from acetylated test amino acids (here: acetyl-alanine (AcAla). Via the endopeptidase mode, oxidised or glycated proteins are cleaved. Protein oxidation is represented by the addition of double-bonded oxygen atoms (carbonyl groups). Structures were drawn in ChemSketch (version 2024.1.13; Advanced Chemistry Development, Inc. (ACD/Labs), Toronto, ON, Canada, www.acdlabs.com). Images were taken from BioRender (Created in BioRender. Huesgen Lab (2025) https://BioRender.com/g41v537).

Nevertheless, the question arises as to which function of AARE dual operation modes (endopeptidase/exopeptidase) is truly conserved or crucial across all kingdoms of life. N-terminal acetylation is a highly abundant and conserved protein modification in eukaryotes and affects approximately 75-90% of the soluble, cytoplasmic proteome (Bienvenut et al., 2012; Linster and Wirtz, 2018; Hoernstein et al., 2024). N-terminal acetylation also occurs in archaea (Halobacterium salinarum, Natronomonas pharaonis), albeit present only up to 30% (Falb et al., 2006; Aivaliotis et al., 2007; Soppa 2010). In bacteria (Escherichia coli), the degree of N-terminal acetylation is even lower, reaching only up to 10% (Bienvenut et al., 2015), whereas approximately 5% retain their N^α-formylation. AARE activity towards formylated N-terminal amino acids has also been demonstrated repeatedly, although with slower reaction rates compared to the acetylated counterpart (Tsunasawa et al., 1975; Gade and Brown, 1978; Kobayashi and Smith, 1987; Ishikawa et al., 1998). Remarkably, AARE function seems to be absent in E. coli (Witheiler and Wilson, 1972). Hence, AARE’s exopeptidase function may not be only specific for acetylation, but also for other blocked N-termini.

Furthermore, protein oxidation—targeted by AARE’s endopeptidase activity—is a common occurrence in all living cells. However, its prevalence is likely much lower compared to N-terminal modifications, such as acetylation. Due to the diverse and complex nature of oxidative protein modifications, accurately estimating their overall abundance within a given proteome remains inherently challenging. It is likely that oxidative modifications, particularly those induced by stress, affect only a subset of a given protein population, as proteins exhibit varying susceptibilities to oxidation. Moreover, the specific types of oxidative modifications can differ significantly across organisms. Apart from the oxidation of cysteine and methionine side chains, metal-catalysed protein oxidation (MCO) by ROS via the Fenton reaction, which leads to the formation of primary carbonyls on lysine, arginine, threonine, and proline residues (Stadtman and Levine 2003; Estévez et al., 2022), may represent the most widely shared mechanism across lifeforms. In contrast to primary carbonyls, secondary carbonyls on proteins are derived from highly reactive dicarbonyls such as HNE or MDA. Both HNE or MDA are present in plants and humans (Møller et al., 2007; Ayala et al., 2014; Milkovic et al., 2023), but the general composition of the dicarbonyl variety may differ between organisms. The same holds true for reducing sugars which cause unwanted glycation leading to the formation of AGEs.

AARE function was proven on both artificially oxidised and glycated proteins (Fujino et al., 2000a; Yamauchi et al., 2003; Gogliettino et al., 2014; Riccio et al., 2015) which raises the question whether there is a certain specificity or whether AARE exhibits promiscuous activity via its endopeptidase function. In 2007, Møller and colleagues suggested that, due to the wide variety of oxidative modifications, it is unlikely that specific proteases are solely responsible for removing oxidatively modified proteins. Instead, they proposed that responsible proteases are more likely to target proteins with a more open conformation (Møller et al., 2007). However, knowledge regarding the cleavage specificity of the endoprotease activity of AARE isoforms is limited to only a few studies (Chongcharoen and Sharma, 1998; Fujino et al., 2000b; Yamin et al., 2009; Sandomenico et al., 2021). For instance, a proteolytically active 55 kDa fragment of bovine lens AARE was shown to cleave the insulin alpha chain near oxidised cysteine residues (Chongcharoen and Sharma, 1998), whereas full-length recombinant human AARE (APEH) cleaved the amyloid-beta peptide C-terminally after oxidised histidine or between oxidised phenylalanine residues (Yamin et al., 2009). In contrast, human AARE purified from red blood cells cleaved synthetic peptides of GDF11 (Growth Differentiation Factor 11) comparably distant from two oxidised methionines (Sandomenico et al., 2021). Cleavage sites in oxidised BSA by AARE have been identified (Fujino et al., 2000b). However, no data are available on oxidatively modified amino acids in close proximity to these cleavage sites.

These findings may support the hypothesis of Møller et al. (2007) but further research employing state-of-the-art mass spectrometry is required to fully address this question. Notably, if AARE shows a preference for unstructured and open conformations, cleavage of non-oxidised, unfolded proteins would be significant. This is further supported by the early findings of Tsunasawa et al. (1975), who observed enhanced cleavage of AcGly from ovalbumin under denaturing conditions. Regardless of that, even if AARE exhibits broader substrate specificity, one would expect differences between species and kingdoms which is already reflected via AARE exopeptidase mode. Whereas most investigated AARE isoforms from archaeal species prefer AcLeu over AcAla, plant and fish AARE enzymes prefer AcAla and AcMet (Table 1).

Table 1.

Summary of purified and biochemically characterized acylamino acid-releasing enzyme (AARE) enzymes and their specificities based on test substrates. pH and temperature optimum values were determined in the referenced studies. *Substrate specificity of Physcomitrella (Physcomitrium patens) PpAARE1 is based on observations from KO mutants and not based on assays with purified isoforms.

Organism	Protein	pH	Temperature [°C]	Substrate specificity	Multimer	Publication
Aeropyrum pernix	apAPH	7.0	90	AcLeu/AcPhe	Dimer	Bartlam et al. (2004), Kiss et al. (2007)
Pyrococcus horikoshii	PhAAP1	7.0	90	AcLeu	Dimer	Ishikawa et al. (1998), Szeltner et al. (2009)
	PhAAP2			AcLeu/AcPhe	Hexamer
Sulfolobus solfataricus	APEH-3Ss	7.5	80	AcAla/AcLeu/AcPhe	Trimer	Palmieri et al. (2010), Gogliettino et al. (2012)
	APEHSs		90		Dimer
Sporosarcina psychrophila	SpAAP	-	40	AcLeu/AcPhe	Dimer	Brunialti et al. (2011), Brocca et al. (2016)
Geobacillus stearothermophilus	S9gs	9.0	45	AcAla/AcLeu	Tetramer	Chandravanshi et al. (2024)
Cucumis sativus	CsAARE	7.0	37	AcAla/AcMet/AcGly	Tetramer	Yamauchi et al. (2003)
Arabidopsis thaliana	AtAARE	7.0	37	AcAla/AcMet/AcGly	Tetramer
Physcomitrium patens	PpAARE1*	-	-	AcAla/AcLeu*	-	Hoernstein et al. (2023)
	PpAARE2	-	-	-	-
	PpAARE3	-	-	-	-
Bombyx mori	BmAPH	7.5	50	AcAla-		Fu et al. (2016)
Trematomus bernacchii	APEH-1Tb	8.0	60	AcAla/AcMet/AcLeu	Tetramer	Gogliettino et al. (2014)
	APEH-2Tb		50	AcAla/AcMet
Chionodraco hamatus	APEH-1Ch	7.5	50	AcAla/AcMet/AcLeu	Tetramer	Riccio et al. (2015)
	APEH-2Ch		40	AcAla/AcMet/AcLeu/AcPhe
Dicentrachus labrax	APEHDl	7.5 & 8.9	50	AcAla/AcMet	Tetramer	Riccio et al. (2015)
Rattus norvegicus	RnAARE	7.2-7.6	-	AcMet/AcAla/AcSer/AcLeu	Tetramer	Tsunasawa et al. (1975)
Homo sapiens	HsAPEH	-	-	AcMet/AcAla/AcSer/AcGly	Tetramer	Scaloni et al. (1994), Sandomenico et al. (2021)
Bos taurus	BtAPEH	8.2	-	AcAla/AcMet/AcGly/AcSer	Tetramer	Gade and Brown (1978)

Despite low sequence similarity across isoforms from different kingdoms, AARE enzymes display highly conserved tertiary and quaternary structures in addition to their shared functional characteristics. For instance, archaeal and human AARE share only about 27% sequence similarity, while human and porcine AARE isoforms are approximately 90% identical (Bartlam et al., 2004). The first available crystal structure of the archaeal isoform, apAPH, confirmed the predicted structure for AARE isoforms consists of an N-terminal seven bladed β-propeller, which is involved in protein/protein interactions and multimerization, and a C-terminal α/β-fold hydrolase domain harbouring the catalytic centre (Fig. 2; Bartlam et al., 2004). Although successful crystallizations of human and porcine AARE isoforms were reported already several years ago (Freese et al., 1993; Wright et al., 2005), the cryo-EM structure of the porcine AARE complex was published recently (Fig. 2; Kiss-Szemán et al., 2022). The crystal structure of a bacterial AARE isoform from Geobacillus stearothermophilus is also available (Fig. 2; Chandravanshi et al., 2024), confirming the conserved seven bladed β-propeller and the α/β-fold hydrolase domain with the catalytic centre. Substrate access to the catalytic centre of AARE isoforms occurs through a central tunnel within the β-propeller structure (Bartlam et al., 2004), as well as through an additional lateral opening (Kiss-Szemán et al., 2022; Chandravanshi et al., 2024). Additionally, AAREs form homomultimers with varying subunit numbers, ranging from dimers to hexamers (Table 1). Typically, plant and animal AAREs assemble as tetramers, while archaeal and bacterial isoforms exhibit a broader range of multimeric configurations. The interactions between distinct monomers within the multimer, as well as substrate binding, appear to induce structural rearrangements that regulate the active or inactive states of AARE complexes. These observations suggest that the degree of multimerization may modulate different functional modes (endopeptidase/exopeptidase) and influences substrate specificity (Kiss-Szemán et al., 2022; Chandravanshi et al., 2024). Notably, several organisms, including Physcomitrella, rice (Oryza sativa), Funaria hygrometrica, as well as certain fish and archaeal species, encode multiple AARE isoforms (Ishikawa et al., 1998; Szeltner et al., 2009; Palmieri et al., 2010; Gogliettino et al., 2012, 2014; Riccio et al., 2015; Hoernstein et al., 2023). This diversity suggests the potential for forming heteromeric complexes with varying subunit stoichiometries. In Physcomitrella, we recently identified specific interactions among the three isoforms (PpAARE1-3) using co-immunoprecipitation (Co-IP) followed by mass spectrometry (MS) analysis (Hoernstein et al., 2023). If these interactions indeed form heteromers, they may provide additional mechanisms for fine-tuning functional modes or substrate specificities. Finally, proteolytic processing also generates functional AARE fragments which are proteolytically active (Sharma and Ortwerth 1993; Chongcharoen and Sharma, 1998). Such a process may also lead to a modulation of AARE function regarding endo- or exopeptidase mode.

Fig. 2.

Tertiary structures of archaeal, mammalian and bacterial acylamino acid-releasing enzyme (AARE) isoforms showing the conserved N-terminal β-propeller (cyan) and the C-terminal α/β-fold hydrolase domain (gold) harbouring the catalytic centre. Isoform names are according to the corresponding publication reporting the structure. Structures were downloaded from the Protein Data Bank (https://www.rcsb.org/structure/1VE6 (Bartlam et al., 2004); https://www.rcsb.org/structure/7QUN (Kiss-Szemán et al., 2022); https://www.rcsb.org/structure/8WT1 (Chandravanshi et al., 2024).

AARE function in physiology and disease

A limited number of studies have investigated AARE function in relation to organismal physiology, largely due to the lack of viable mutants. So far, this has been reported only in plants with an RNA interference (RNAi) as well as T-DNA mutant lines in Arabidopsis and knockout (KO) mutants in Physcomitrella (Nakai et al., 2012; Hoernstein et al., 2023). Transient RNAi suppression of AARE function was reported in human bronchial epithelial cells (Palmieri et al, 2011) and in adenocarcinomic human alveolar basal epithelial (A549) cells (Zeng et al., 2017).

Apart from genetic mutants, a range of AARE inhibitors have been reported (Scaloni et al., 1992a; Yamin et al., 2007; Adibekian et al., 2011; Palmieri et al., 2011; Bergamo et al., 2013; Bachovchin et al., 2014). These comprise a broad range of chemical compounds which inhibit AARE function reversibly, such as AcMet-OH and AcAla-OH (Scaloni et al., 1992a; Yamin et al., 2007), but also irreversible inhibitors, such as Acetyl-leucyl-chloromethylketone (ALCK) and 1,2,3-triazole urea (Scaloni et al., 1992a; Yamin et al., 2007; Adibekian et al., 2011). Moreover, naturally occurring molecules, such as conjugated linoleic acid (CLA) isomers, were identified as efficient inhibitors of AARE function (Palmieri et al., 2011; Bergamo et al., 2013).

In the moss Physcomitrella, single KOs of the three distinct AARE isoforms did not reveal any striking phenotypic deviation between wild type (WT) and mutant lines regarding plant size and development, suggesting a certain degree of functional compensation between the different isoforms (Hoernstein et al., 2023). This finding is remarkable since exo- and endopeptidase function seem to be split across two of the three isoforms (PpAARE1, PpAARE2). Double knockout (doKO) mutants, which lack both functions, displayed an accelerated transition from the filamentous protonema (2D growth stage) to the adult leafy gametophore (3D stage). Similarly, the KO of the conserved cysteine protease Defective Kernel 1(DEK1) caused strongly enhanced bud formation in the protonema (Demko et al., 2024) but these buds did not develop further into mature gametophores. In contrast, the doKO of PpAARE1/2 did give rise to gametophores, although they remained dwarfed and exhibited a strikingly reduced life span (Hoernstein et al., 2023). The apparent distinct phenotypes were only observable in the dominant generation, the leafy gametophore, whereas growth of the filamentous protonema appeared to be unaffected. In those haploid tissues, PpAARE1 is typically the most highly expressed, whereas PpAARE2 expression is most highly expressed in the diploid spores. In gametophore leaflets, PpAARE3 is the most highly expressed. However, the additional KO of PpAARE3 did not further enhance the observed effects, leaving the function of this isoform unclear (Hoernstein et al., 2023).

In contrast to these marked differences, the phenotype of T-DNA mutants in Arabidopsis, which encodes only a single AARE gene (AtAARE), was less pronounced (Hoernstein et al., 2023). While overall plant size remained comparable, the mutants displayed an accelerated transition from the vegetative to the reproductive phase, as evidenced by premature bolting. In line with the analysis of an AtAARE RNAi line (Nakai et al., 2012), an increase of oxidised proteins was observed in the AtAARE T-DNA mutant line (Hoernstein et al., 2023). Unfortunately, phenotypes of the RNAi line compared to WT were not described (Nakai et al., 2012). An increase of protein oxidation was also observed in the Physcomitrella KO of PpAARE2, but the accelerated transition from 2D to 3D growth was not observed in this single KO (Hoernstein et al., 2023). Conversely, overexpression of AtAARE in tobacco did not affect the level of oxidised proteins (Nakai et al., 2012). Here, accumulation of oxidized proteins during leave aging, analysed via OxyBlot, did not reveal remarkable differences between the overexpression line and the corresponding WT (Nakai et al., 2012).

In mammalian cell line COS-7, overexpression of AARE enhanced cell survival and reduced levels of oxidised proteins under oxidative stress (Shimizu et al., 2003), while AARE function appeared to be linked to proteasomal activity (Shimizu et al., 2004). Effectively, AARE inhibition led to an increase of oxidised proteins, along with down-regulation of proteasomal activity (Shimizu et al., 2004). This effect is observed in different mammalian cell types, such as COS-7 cells and human bronchial epithelial cells, as well as a variety of human cancer cell lines treated with AARE inhibitors, such as N-Acetyl-leucine-chloromethyl-ketone (ALCK) and conjugated linoleic acid variants (CLA) (Shimizu et al., 2004; Palmieri et al., 2011; Bergamo et al., 2013; Palumbo et al., 2016). AARE inhibition was also associated with pro-apoptotic activity and reduced cell viability after CLA treatment, in line with an imbalance of the cellular redox system (Bergamo et al., 2013). Since AARE activity appeared unaffected after exogenous hydrogen peroxide (H₂O₂) application (Shimizu et al., 2003), this may indicate that AARE is not directly inhibited by H₂O₂ but its activity is linked to the cellular redox system. In fish, AARE isoforms have distinct biochemical properties and their differential gene expression strengthen a potential role of AARE in the antioxidative defence system (Gogliettino et al., 2014; Riccio et al., 2015). This is further evidenced by the fact that AARE supports the X-Ray Repair Cross-Complementing protein XRCC1 in nuclear DNA single strand break repair under oxidative stress (Zeng et al., 2017).

Accordingly, disruption of the antioxidative system was observed in an AtAARE RNAi line of Arabidopsis thaliana (Nakai et al., 2012). Several proteins associated with the cytosolic redox system were up-regulated, including cICDH (ISOCITRATE DEHYDROGENASE, AT1G65930.1), ATGSTF9 (GLUTATHIONE S-TRANSFERASE, AT2G30860.1), ATGSTF10 (GLUTATHIONE S-TRANSFERASE, AT2G30870.1), and APX1 (ASCORBATE PEROXIDASE 1, AT1G07890.1). However, the oxidised sites of these proteins were not identified. The identification of these proteins was based on MS analysis following OxyBlot detection, leading to the conclusion that they are likely oxidised targets of AtAARE (Nakai et al., 2012). Notably, there is a partial overlap between the proteins identified in Arabidopsis (Nakai et al., 2012) and homologous proteins in rice, such as APX2 (Q0D3B8) and ICDH (Q7XMA0), which have been shown to undergo carbonylation (Zhang et al., 2016). These findings suggest that AtAARE plays a crucial role in maintaining the antioxidative system by clearing oxidised proteins also in plants (Nakai et al., 2012).

AARE is involved in both age-related development and as part of the antioxidative system in plants, whereas the latter also holds true for mammals. The involvement of AARE in the developmental processes of mammals is less clear. However, changes in AARE activity and expression have been associated with certain types of cancer, age-related pathologies, such as cataract formation, and Alzheimer’s disease, but also lung inflammation (Scaloni et al., 1992b; McGoldrick et al., 2014; Santhoshkumar et al., 2014; Komatsu et al., 2016; Palmieri et al., 2017; Tangri et al., 2021). As such, cancer can be regarded as abnormal growth or development.

In small cell lung carcinoma cell lines, AARE function is apparently absent (Scaloni et al., 1992b), whereas the expression of the N-terminal acetyltransferase D (NatD) which acetylates specifically N-termini of histones is up-regulated during epithelial to mesenchymal transition, promoting cancer progression (Ju et al., 2017). It is not known whether down-regulation of AARE function is causative or a side effect of cancer progression, but histone acetylation seems to be a key factor in the progression of these types of cancer. Hence, absence of AARE function could also indicate a potential exopeptidase activity of AARE on the N-terminal acetylation state of histones. In contrast, AARE activity is elevated in prostate cancer cell lines (McGoldrick et al., 2014). While up-regulation of the N-alpha-acetyltransferase 10 protein (Naa10p), a subunit of N-alpha-acetyltransferase complex A (NatA), is also observed, cancer progression may be more closely linked to the promiscuous activity of this protein which also acetylates lysine side chains (DePaolo et al., 2016) and are not substrates for AARE. Simultaneously, ROS levels are markedly increased in prostate cancer cells, driving aggressive progression (Kumar et al., 2008; Ruscica et al., 2018). Thus, elevated AARE activity here may be a secondary response to mitigate the accumulation of oxidised proteins independent of protein acetylation.

AARE expression is also elevated in brain tissues of Alzheimer’s disease (AD) patients with a high load of amyloid plaques (Yamin et al., 2009), which could have two causes simultaneously. Firstly, AARE was shown to degrade the amyloid-β peptide, which is the main source of amyloid plaque formation in AD (Yamin et al., 2007, 2009). Secondly, progression of AD is also accompanied by a strong increase in oxidative stress caused by amyloid-β oligomer accumulation (Butterfield and Boyd-Kimball, 2017). Hence, increased gene expression of AARE would be a response to sustain cellular homeostasis. In contrast, AARE activity, as well as proteasomal activity, are significantly down-regulated in erythrocytes of AD patients (Palmieri et al., 2017). This may be the result of either tissue-specific differences of AARE activity, or increased gene expression in the affected brain tissues that is not correlated with AARE activity, but rather a struggle to compensate for a loss of AARE activity.

It seems that AARE of plants and mammals is responsive to cell redox state, and its function is tightly linked to the redox state of a cell. Whether the absence or presence of AARE function is causative in mammalian pathogenesis, or instead a concomitant feature, is not yet clear. Effectively, AARE function differs between cell types and across tissues in plants and mammals (Tsunasawa et al., 1975; Yamauchi et al., 2003; Bergamo et al., 2013; Komatsu et al., 2016; Covarrubias et al., 2022) and may also depend on the oligomerization state. AARE function is decreased in age-related diseases with increased oxidative stress, such as AD and Type 2 diabetes (Palmieri et al., 2017; Marshall et al., 2019). Accordingly, AARE function declines with leaf age in Arabidopsis (Yamauchi et al., 2003), which is in line with increasing ROS in plastids and AGE levels during plant aging (Munné-Bosch and Alegre, 2002; Chaplin et al., 2019). AARE activity in mammals seemed less affected by exogenous H₂O₂ application, but is strongly impaired by endogenous lipid peroxides caused by e-cigarette vapour (Shimizu et al., 2003; Tyler et al., 2021). Hence one may speculate that an increase of intracellular lipid peroxides inhibits AARE function, thereby promoting age-related molecular deterioration of the cell.

Plant and animal AAREs exhibit several conserved features, including multimerization, substrate specificity (Table 1), and subcellular localization (Fig. 3). However, plant AARE isoforms show additional targeting to plastids and mitochondria (Fig. 3), which is distinct from animals. In animals, AARE has not yet been observed to localize to mitochondria (Shimizu et al., 2003; Zeng et al., 2017) nor does it harbour a predicted mitochondrial presequence (Hoernstein et al., 2023). Another conserved aspect appears to be the linkage between AARE function and the cellular redox state.

Fig. 3.

Summary of the current knowledge of acylamino acid-releasing enzyme (AARE) localization and function in plants and animals. AtAARE (green) is depicted as plant AARE and human AARE (blue) as animal AARE. Structures of both isoforms (AF-Q84LM4-F1, AF-P13798-F1; https://alphafold.ebi.ac.uk/) were predicted with AlphaFold (Jumper et al., 2021; Varadi et al., 2024). Middle shows commonalities between plants and animals: Localization of AARE to the nucleus and to the cytosol has been shown in Arabidopsis and human cell lines (Shimizu et al., 2003; Nakai et al., 2012; Zeng et al., 2017; Hoernstein et al., 2023). AARE functions as part of the antioxidant defence system in plants and animals (Shimizu et al., 2003; Nakai et al., 2012; Gogliettino et al., 2014; Riccio et al., 2015; Zeng et al., 2017). Left (only plants): AARE localizes to both plastids and mitochondria in plants (Hoernstein et al., 2023). AARE function influences the transition from vegetative to reproductive phase in Physcomitrella and Arabidopsis, and influences life span and plant size (Hoernstein et al., 2023). Right (only animals): AARE function is necessary for cell viability and linked to proteasomal function (Shimizu et al., 2004; Palmieri et al., 2011, 2017; Bergamo et al., 2013; Palumbo et al., 2016; Gogliettino et al., 2021). AARE interacts with X-ray repair cross-complementing protein (XRCC1) to support DNA single strand break repair under oxidative stress (Zeng et al., 2017). Plant image parts have been taken from BioRender (Huesgen Lab (2025) https://BioRender.com/g41v537) and from Falz and Müller-Schüssele (2019).

Conversely, certain aspects of AARE function, such as their roles in developmental transitions (Hoernstein et al., 2023) or their connection to proteasomal activity, appear more kingdom-specific (Shimizu et al., 2004; Palmieri et al., 2011, 2017; Bergamo et al., 2013; Palumbo et al., 2016) and cannot yet be easily generalized. These differences raise the question of whether these characteristics represent unique adaptations within kingdoms or species, or if they reflect variations in broader, conserved mechanisms. Further research is essential to elucidate these possibilities.

Open questions and critical points

More than 60 publications have investigated AARE’s biochemical and physiological function. Despite this extensive body of research, many questions remain unanswered, and some findings may appear counterintuitive.

One important unresolved question is how the dual operational mode is regulated among AARE isoforms. Although assays to evaluate AARE function typically focus on exopeptidase activity, distinctions between endo- and exopeptidase activities may vary across isoforms or differ in specific cell types (Covarrubias et al., 2022; Hoernstein et al., 2023). Therefore, a reduction in AARE exopeptidase activity does not necessarily indicate a corresponding decrease in endopeptidase function. In Arabidopsis, AARE transcript and protein levels are positively correlated (Hoernstein et al., 2023). However, this relationship does not consistently apply to exopeptidase activity and AARE protein levels in mammalian systems (Marshall et al., 2019). Consequently, AARE transcript or protein levels alone may not provide a comprehensive picture of AARE functionality, underscoring the need to assess both exo- and endopeptidase activities for an accurate evaluation.

Another intriguing question is why the absence of AARE promotes cell proliferation and development in certain cell types, such as small cell lung carcinoma and moss protonema, while in other types, such as osteosarcoma, AARE inhibition triggers cell death (Jones et al., 1991; Scaloni et al., 1992b; Palmieri et al., 2011; Gogliettino et al., 2021; Hoernstein et al., 2023). One hypothesis suggests a link between AARE function and the stabilization of N-terminal acetylated proteins. Consistent with this, AARE inhibition in T-cells leads to increased levels of N-terminal acetylated proteins and promotes cell proliferation (Adibekian et al., 2011). N-terminal acetylation especially of initiator methionines is known to stabilize proteins, extend their half-life, and regulate various cellular pathways (Varland et al., 2023). Arfin and Bradshaw (1988) proposed that AARE might cleave these N-terminally acetylated residues, thereby generating substrates for the Arg/N-degron pathway (Varshavsky, 2019). Although AARE exhibits relatively low activity toward substrates such as AcMet-Glu or AcMet-Asp in short peptide assays, this activity has been independently observed in AARE isoforms from both plants and mammals (Krishna and Wold, 1992; Yamauchi et al., 2003). Supporting this hypothetical function, findings in Physcomitrella suggest AARE functions on full-length proteins (Hoernstein et al., 2016). More precisely, MS analysis of PpAARE1 itself revealed N-terminal arginylation of the asparagine residue after the intitatior methionine(Asp²). An exposure of an Asp² is unexpected since its preceding Met is usually not targeted by methionine amino peptidases (Dissmeyer, 2019; Fuentes-Terrón et al., 2025). Moreover, Met¹ with subsequent Asp² is typically acetylated by the N-terminal acetyl transferase complex B (NatB; Giglione and Meinnel, 2021). Both at the N-terminus, the arginylated Asp² and the acetylated Met¹ were unambiguously identified (Fig. 4), whereas an N-terminus with a non-acetylated Met¹ was not identified in Hoernstein et al. (2016). This dual modification state indicates the cleavage of the preceding acetylated Met¹  in vivo, but it should also be considered that internal cleavage of a longer PpAARE1 isoform harbouring an N-terminal extension (Hoernstein et al., 2023) could expose this Asp residue instead.

Fig. 4.

Higher-energy collisional dissociation (HCD) fragment spectra and tables showing the dual modification state of the N-terminus of PpAARE1. Database search results in Hoernstein et al. (2016) were loaded in Scaffold5^TM software (V5.3.3; www.proteomesoftware.com) to export the depicted spectra. A) HCD fragment spectrum and fragment table of the acetylated peptide MDQVTSGKAR of Physcomitrella PpAARE1 (Pp3c2_30720V3.1). The methionine is N-terminally acetylated (+42.01 Da) and oxidized (+15.99 Da). Free amino groups (here: K) have been dimethylated (+34.06 Da) B) HCD fragment spectrum and fragment table of the post-translationally arginylated peptide DQVTSGKAR of Physcomitrella PpAARE1 (Pp3c2_30720V3.1). Free amino groups were dimethylated (+34.06 Da) causing a mass shift of +190.16 Da for a dimethylated N-terminal arginine. Data are available via ProteomeXchange with identifier PXD003232. Tables depict expected masses of b and y fragment ions of the peptides after HCD fragmentation. Coloring indicates a match between an expected and an experimentally observed mass. b/y: incremental number of a b or y fragment ion. b/y Ions: MH⁺ masses of expected b or y fragment ions. b/b+2H: doubly protonated b or y fragment ion. b/y-NH₃: b or y fragment ion with neutral loss of NH₃. b/y-H₂O: B or Y fragment ion with neutral loss of H₂O.

This hypothesis links directly to another open question, namely whether there is AARE activity on N-terminal acetylated proteins. Although shown in the initial publication of AARE (Tsunasawa et al., 1975), several publications describe a decline of AARE activity towards the N-terminal acetylated amino acid in correlation with increasing peptide length (Gade and Brown, 1978; Krishna and Wold, 1992). Despite 25 potential targets (Adibekian et al., 2011), no physiological targets of AARE have been unambiguously identified. Using an N-terminomics approach, Adibekian et al. (2011) identified an N-terminus whose abundance increased upon selective inhibition of AARE, while remaining unchanged during inhibition of the acetyl-hydrolase PAFAH2 (PLATELET-ACTIVATING FACTOR ACETYLHYDROLASE-2). Several of these proteins are known to undergo N-terminal acetylation. For five of them, synthetic peptides mimicking the N-terminally acetylated sequence were generated, and, in all cases, AARE activity on these peptides was demonstrated (Adibekian et al., 2011). Taken together, these data suggest that these proteins are in vivo targets of AARE. The authors also stated that AARE function would likely preferentially alter the N-terminal modification state, instead of impacting the overall stability of proteins, since candidate protein levels did not change. As there is clear AARE activity on intact proteins via the endopeptidase mode, it can be assumed that there is also activity on intact proteins via the exopeptidase mode.

It has to be further investigated if AARE truly has a role in delaying aging as proposed by Hoernstein et al. (2023). Is it possible to transfer the knowledge of AARE function in Physcomitrella and Arabidopsis development to a general function across kingdoms? This question cannot be answered now, but several key points should be considered: 1) Independent of the organism, aging is always influenced by stress linked to ROS formation along with detrimental protein oxidation. 2) Protein oxidation and glycation increase during aging in plants and mammals. 3) Oxidised and glycated proteins are targeted by AARE endopeptidase mode. 4) Accumulation of oxidised and glycated proteins is a hallmark of aging and detrimental to the cell leading to perturbed proteostasis and cell death. 5) AARE function promotes cell viability.

In this context it should be noted that, although increased levels of oxidized proteins due to AARE deficiency have been independently demonstrated (Nakai et al., 2012; Tyler et al., 2021; Hoernstein et al., 2023), the fate of these proteins remains unclear. It is plausible that their accumulation contributes to the formation of cytotoxic aggregates, a hallmark of molecular aging. However, aggregate formation in the context of AARE inhibition has not yet been investigated. Given the different cell lines, organisms, and experimental setups which provide insights into AARE function in physiology, it is challenging to integrate these key points into the phenotypes caused by AARE loss-of-function observed in Physcomitrella and Arabidopsis (Hoernstein et al., 2023). This raises the question of how AARE function is linked to development at the molecular level. Recent studies in mice exposed to e-cigarette vapour have shown increased levels of oxidised proteins alongside reduced AARE activity (Tyler et al., 2021), suggesting that AARE inhibition in vivo may disturb cellular redox balance, potentially accelerating disease development and progression. In this sense, AARE inhibition, due to shifts in redox balance, may stabilize a signal promoting the transition from the vegetative to the reproductive phase in plants. In Arabidopsis, protein oxidation increases prior to bolting, then decreases after bud emergence (Johansson et al., 2004), suggesting finely controlled modulation or temporary inactivation of AARE activity. This transition is also marked by reduced ascorbate peroxidase activity and a rise in lipid peroxides, particularly in leaves (Ye et al., 2000). Lipid peroxides are strong AARE inhibitors (Tyler et al., 2021), potentially providing a regulated mechanism for suppression of AARE activity during bolting. While it remains speculative whether increased protein oxidation directly drives developmental transitions, acetylation through conserved NatA and NatB complexes is another key mechanism modulating flowering time in Arabidopsis (Kapos et al., 2015).

In order to understand the potential link between accelerated developmental transitions and aging in Arabidopsis and Physcomitrella (Hoernstein et al., 2023), it is essential to identify in vivo targets of both endopeptidase and exopeptidase activities. At the same time, investigating cellular proteolytic systems, such as proteasome function and autophagic flux—key processes in maintaining cellular homeostasis and longevity—may clarify whether the observed effects in Physcomitrella and Arabidopsis are driven by cellular deterioration, or if oxidative modifications and/or acetylation of specific AARE-targeted proteins may serve signalling roles.

A major critical point in future AARE research is the lack of mutants in other organisms apart from Arabidopsis and Physcomitrella. The lack of identified mutants in any mammalian organism is intriguing, particularly given the biochemical studies on purified AARE isoforms. A reason for this might be lethality caused by AARE loss-of-function. In this context it appears puzzling how Arabidopsis AARE T-DNA mutants exhibit an accelerated transition to bolting despite any other phenotypic deviation such as reduced growth or premature senescence. This may indicate either functional compensation by another protease or process, or may be due to the nature of the T-DNA mutants available which are not null mutants. The latter could be investigated by generating null mutants, e.g. with CRISPR-Cas. Given the phenotype of Physcomitrella PpAARE1/2 double mutants and studies showing a link between AARE and proteasome function (Shimizu et al., 2004; Palmieri et al., 2011, 2017; Bergamo et al., 2013; Palumbo et al., 2016; Hoernstein et al., 2023), it is likely that AARE has a crucial role in cellular homeostasis and that a complete shutdown of its function has a striking impact on cellular physiology. A similar pattern is observed with ARGINYL-TRNA-PROTEIN TRANSFERASE 1 (ATE), a central component of the arginylation branch of the N-degron pathway. Null mutants of ATE are embryo-lethal in mice and Drosophila melanogaster, while knockouts in Physcomitrella and Arabidopsis remain viable (Spradling, 1999; Kwon et al., 2002; Yoshida et al., 2002; Graciet et al., 2009; Lee et al., 2012; Schüssele et al., 2016). In Physcomitrella, however, ATE knockouts exhibit significant developmental abnormalities, whereas the phenotypic effects in Arabidopsis T-DNA mutants are noticeable but considerably milder (Graciet et al., 2009; Schüssele et al., 2016). Besides the lack of viable mutants in other plant and non-plant species, another critical aspect would be the analysis of AARE function in perennial plants. It is of particular interest to investigate AARE function over time and in relation to annual senescence. This would clarify how AARE function is linked to annual accumulation of oxidized proteins during senescence and their clearance during bolting (Ciacka et al., 2020).

Lastly, we cannot exclude the possibility that some organisms have lost AARE function and employ alternative strategies to cope with protein oxidation and N-terminally blocked proteins and/or peptides. One prominent example is E. coli, which apparently lacks at least AARE exopeptidase function (Witheiler and Wilson, 1972). Moreover, we do not find any significant homology to bacterial or human AARE in the E. coli genome.

Conclusions

Proposing AARE as part of a unified concept of aging may be bold, especially as only two mutant studies have been conducted specifically in plants (Nakai et al., 2012; Hoernstein et al., 2023). Nonetheless, several conserved features provide the basis for exploring its physiological role within the framework of aging. This requires a definition of aging itself. We understand aging as the progression of biochemical processes over time, regulated by the genetic program and influenced by environmental factors. Aging encompasses these processes from the initial cell division through development to the death of the cell or the entire organism.

In this context, AARE function delays aging of cells, as loss-of-function accelerates developmental progression in two model plant species (Hoernstein et al., 2023). Effectively, the molecular consequence of a cell suffering from a loss of AARE is the accumulation of oxidised proteins, impling a connection to the redox state of a cell. Plants seem to be more tolerant to the accumulation of oxidised proteins since null mutants are viable, whereas cell death as a consequence of AARE inhibition is triggered in mammalian cell lines (Palumbo et al., 2016; Gogliettino et al., 2021). Nevertheless, the physiological consequences of accumulating oxidised proteins, acetylated peptides, or proteins apart from oxidation may vary between tissues and metabolic states and, hence, can have different physiological consequences in the short term. However, the long-term consequence is molecular deterioration.

Glossary

Abbreviations:

AAP

acylaminoacyl peptidase

AARE

acylamino acid-releasing enzyme

Ac

acetyl

AcAla-pNA

N-Acetyl-alanine para-nitroanilide

AGE

advanced glycation end product

BSA

bovine serum albumin

HNE

4-hydroxy-2-nonenal

MCO

metal catalysed oxidation

MDA

alondialdehyde

OPH

oxidised protein hydrolase

RCS

reactive carbonyl species

ROS

reactive oxygen species

Author contributions

SNWH compiled the literature and wrote the manuscript together with RR. AAM provided images and data and helped writing the manuscript. RR acquired funding.

Conflict of interest

The authors declare no conflict of interest.

Funding

Funding by Germany’s Excellence Strategy (CIBSS – EXC-2189 – Project ID 390939984) is gratefully acknowledged.