Role of AMBRA1 in mitophagy regulation: emerging evidence in aging-related diseases

ABSTRACT

Aging is a gradual and irreversible physiological process that significantly increases the risks of developing a variety of pathologies, including neurodegenerative, cardiovascular, metabolic, musculoskeletal, and immune system diseases. Mitochondria are the energy-producing organelles, and their proper functioning is crucial for overall cellular health. Over time, mitochondrial function declines causing an increased release of harmful reactive oxygen species (ROS) and DNA, which leads to oxidative stress, inflammation and cellular damage, common features associated with various age-related pathologies. The impairment of mitophagy, the selective removal of damaged or dysfunctional mitochondria by autophagy, is relevant to the development and progression of age-related diseases. The molecular mechanisms that regulates mitophagy levels in aging remain largely uncharacterized. AMBRA1 is an intrinsically disordered scaffold protein with a unique property of regulating the activity of both proliferation and autophagy core machineries. While the role of AMBRA1 during embryonic development and neoplastic transformation has been extensively investigated, its functions in post-mitotic cells of adult tissues have been limited due to the embryonic lethality caused by AMBRA1 deficiency. Recently, a key role of AMBRA1 in selectively regulating mitophagy in post-mitotic cells has emerged. Here we summarize and discuss these results with the aim of providing a comprehensive view of the mitochondrial roles of AMBRA1, and how defective activity of AMBRA1 has been functionally linked to mitophagy alterations observed in age-related degenerative disorders, including muscular dystrophy/sarcopenia, Parkinson diseases, Alzheimer diseases and age-related macular degeneration.

Abbreviations: AD: Alzheimer disease; AMD: age‐related macular degeneration; AMBRA1: autophagy and beclin 1 regulator 1; APOE4: apolipoprotein E4; ATAD3A: ATPase family AAA domain containing 3A; ATG: autophagy related; BCL2: BCL2 apoptosis regulator; BH3: BCL2-homology-3; BNIP3L/NIX: BCL2 interacting protein 3 like; CDK: cyclin dependent kinase; CHUK/IKKα: component of inhibitor of nuclear factor kappa B kinase complex; CRL2: CUL2-RING ubiquitin ligase; DDB1: damage specific DNA binding protein 1; ER: endoplasmic reticulum; FOXO: forkhead box O; FUNDC1: FUN14 domain containing 1; GBA/β-glucocerebrosidase: glucosylceramidase beta; HUWE1: HECT, UBA and WWE domain containing E3 ubiquitin protein ligase 1; IDR: intrinsically disordered region; LIR: LC3-interacting region; MAP1LC3/LC3: microtubule associated protein 1 light chain 3; MAVS: mitochondrial antiviral signaling protein; MCL1: MCL1 apoptosis regulator, BCL2 family member; MFN2: mitofusin 2; MTOR: mechanistic target of rapamycin kinase; MSA: multiple system atrophy; MYC: MYC proto-oncogene, bHLH transcription factor; NUMA1: nuclear mitotic apparatus protein 1; OMM; mitochondria outer membrane; PD: Parkinson disease; PHB2: prohibitin 2; PINK1: PTEN induced kinase 1; PIK3C3/VPS34: phosphatidylinositol 3-kinase catalytic subunit type 3; PTK2/FAK: protein tyrosine kinase 2; ROS: reactive oxygen species; RPE: retinal pigment epithelium; SAD: sporadic AD; SOCS3: suppressor of cytokine signaling 3; SRC, SRC proto-oncogene, non-receptor tyrosine kinase; STAT3: signal transducer and activator of transcription 3; STING1: stimulator of interferon response cGAMP interactor 1; SQSTM1/p62: sequestosome 1; TBK1: TANK binding kinase 1; TGFB/TGFβ: transforming growth factor beta; TOMM: translocase of outer mitochondrial membrane; TRAF6: TNF receptor associated factor 6; TRIM32: tripartite motif containing 32; ULK1: unc-51 like autophagy activating kinase 1.

Introduction

Aging is defined as the time-related deterioration of the biological functions paving the risks to various aging-related diseases [1]. Aging results from the impact of a variety of dysfunctions on cellular processes over time, including genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, disabled autophagy, deregulated nutrient-sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, altered intercellular communication, chronic inflammation, and dysbiosis [2,3].

Many of those drivers are associated with chronic oxidative stress caused by elevated levels of reactive oxygen species (ROS) mainly generated by mitochondria that, with aging, become highly susceptible to morphological changes and reduced function [4]. Moreover, damaged mitochondria release DNA which induces chronic inflammation and necrotic forms of cell death via cytosolic sensors, such as STING1 (stimulator of interferon response cGAMP interactor 1) and ZBP1 (Z-DNA binding protein 1) [5–7]. For this reason, alterations in mitophagy, a selective form of autophagy that degrades damaged mitochondria, have been recognized as important features of aging [8–10]. In this context, recent evidence shows that impairment of AMBRA1 (autophagy and beclin 1 regulator 1) function inhibits mitophagy in long-lived cells and favors the onset of age-related disorders, such as neurodegenerative and neuromuscular diseases.

AMBRA1 interactome and protein structure

Ambra1 was originally identified using a gene-trap approach in mice searching for genes involved in neuronal development [11]. Mutation of Ambra1 results in exencephaly and spina bifida leading embryo lethality at about E14.5, associated to excessive cell proliferation and cell death, as well as to defective autophagy. The recent characterization of a role of AMBRA1 in cell adhesion/migration, inflammation and transcriptional regulation has further expanded our knowledge of AMBRA1 functions [12–14]. How AMBRA1 controls these cellular activities has been mainly characterized by identifying its interaction partners (Table 1).

Table 1.

AMBRA1-interacting proteins.

AMBRA1-binding proteins	Function	Binding domains	Ref.
AKAP8, BRG1, ATF2	Recruiting ATF2 to chromatin via CDK9	Undefined	[20]
ALDH1B1	Ubiquitin-mediated inhibition of ALDH1B1 signaling in cancer stem cells	Undefined	[27]
BECN1	Inducing BECN1-PIK3C3/VPS34 activity by promoting protein complex assembly	aa 533–780	[10]
CALPAINS	AMBRA1 degradation during apoptosis and by HPV E7	Undefined	[55]
CASPASES	AMBRA1 cleavage upon apoptosis induction	D482	[55,57]
CCNDs	Proteasomal degradation of CCNDs to inhibit the cell cycle	Undefined	[17–19]
CRL4	Degradative ubiquitination of AMBRA1, non-degradative ubiquitination of BECN1	aa 1-204	[15]
DYNLL1/DLC1-DYNLL2/DLC2	Inhibition AMBRA1 translocation to ER	aa 1104-1106, aa 1116-1118	[14]
ELOB-ELOC	Inhibition of CRL5 mediated degradation of DEPTOR, STAT3 and APOBEC3G	aa 824–1298	[13,15]
ERLIN1	Recruiting AMBRA1 to MAMs for autophagy initiation	aa 533–780, aa 796-1298	[28]
PTK2/FAK-SRC	Regulating cell adhesion and migration in tumor cells	Undefined	[11,31]
HUWE1	Proteasomal degradation of MFN2 and MCL1 to induce mitophagy	Undefined	[25, 26]
CHUK/IKKα	Promoting AMBRA1-LC3 interaction during mitophagy	S1043	[25]
LC3	Promoting engulfment of damaged mitochondria by phagophores in mitophagy	aa 1043-1052	[87]
BCL2	Sequestering AMBRA1 at mitochondria to limit autophagy	aa 1-532, 796-1298	[53]
MTOR	Inhibiting AMBRA1-dependent autophagy	Undefined	[23]
PRKN	Local activation of the BECN1-PIK3C3/VPS34 complex during mitophagy	aa 1-532	[85]
PLK1-NUMA1	Regulation of spindle function and orientation during mitosis	phospho-S1238/S1252	[33]
PINK1-ATAD3A	Promoting PINK1 stability and PRKN activation	Undefined	[86]
PPP2C/PP2A (catalytic subunit)	Inhibiting MYC by dephosphorylation to inhibit proliferation	aa 275-281, aa 1206-1212	[29]
SMAD4	SMAD4 activation in TGFB-driven metastasis	Undefined	[21]
TRAF6	Activation of ULK1 by K63-linked polyubiquitin chains to induce autophagy	aa 618–623, aa 681–686	[23]
ULK1	Activation of AMBRA1 by phosphorylation during autophagy	aa 1-532, aa 796-1298	[54]
WASH-RFN2	Inhibiting autophagy by promoting AMBRA1 degradation by RNF2	Undefined	[22]

At first, yeast two-hybrid screens uncovered the interaction of AMBRA1 with BECN1, groundbreaking the studies of AMBRA1 role in autophagy [11], and with DYNL (dynein light chain), acting as negative autophagy regulators [15]. Afterward, affinity-purification-mass spectrometry (AP-MS) approaches identified AMBRA1 as a substrate receptor of CUL4 (cullin 4)-RING ubiquitin ligase (CRL4) [16].

Recently, the structure of AMBRA1 was determined, providing information on the dynamic characteristics of this protein [17] (Figure 1). AMBRA1 contains three and half WD40 repeats at the N-terminal of the protein (1–204 aa) and three and half at the C-term (853–1044 aa), which join to fold in a β-propeller structure. The 650-residues between these WD40 domains constitute an intrinsically disordered region (IDR), as well as the last 250 aa of the protein. The β-propeller mediates the interaction with the CRL4 adaptor DDB1 (damage specific DNA binding protein 1). AMBRA1 may dynamically adopt both an open conformation, with the N- and C-WD40 repeats not joined, and a close one with the β-propeller formed; with the latter structure stabilized by binding to DDB1. Interestingly, while the β-propeller is sufficient to assemble an active CRL4-DDB1-AMBRA1 ubiquitin ligase, the IDR is required for full ubiquitination activity, suggesting that this region interacts with additional factors regulating its activity.

Figure 1.

Structure of the AMBRA1 protein. (A) Description of structural domains of AMBRA1 proteins. IDR: Intrinsically disorder region. (B) Organization of WD40 domains in AMBRA1 proteins to form a β-barrel structure. (C) Prediction of AMBRA1 structure by alpha fold software. Structural model was obtained from the Alpha fold protein structure database developed by EMBL-EBI and DeepMind (https://alphafold.ebi.ac.uk/entry/Q9C0C7) [142,143]. AlphaFold produces a per-residue confidence score (pLDDT) between 0 and 100. Some regions with low pLDDT may be unstructured in isolation. Color scale Model Confidence: Dark Blue: Very high (pLDDT > 90), Light Blue: Confident (90 > pLDDT > 70), Yellow: Low (70 > pLDDT > 50). Orange: Very low (pLDDT < 50). https://alphafold.ebi.ac.uk/entry/Q9C0C7. The figure was created with BioRender.com.

A breakthrough in understanding the role of AMBRA1 in the regulation of proliferation was the identification of CCND (cyclin D)-type cyclins as targets of CRL4-AMBRA1 ubiquitin ligase [18–20]. ambra1 KO embryos have increased levels of CCNDs, which promotes the uncontrolled cell proliferation originally observed in the developing nervous system [11]. Importantly, the binding to AMBRA1 is reduced by cancer hotspot mutations of CCNDs, resulting in protein stabilization. Of note, in AMBRA1 deficient tumor cells, CCNDs form a complex with CDK2, which reduces their sensitivity to CDK4 and CDK6 inhibitors.

CRL4-AMBRA1 is also involved in the regulation of other CUL ligase activity. ELOC (elongin C), the adaptor protein of CRL2- and CRL5-RING ubiquitin ligases together with ELOB (elongin B), is degraded by CRL4-AMBRA1-mediated ubiquitination [14]. This activity impacts on the anti-inflammatory and the proviral activities of CRL5-SOCS3 (suppressor of cytokine signaling 3) and CRL5-HIV-VIF complexes, respectively. In AMBRA1-deficient cells, hyperactive CRL5-SOCS3 inhibits IL6 signaling by degrading the transcription factor STAT3 (signal transducer and activator of transcription), a master regulator of inflammation. Similarly, the CRL5-VIF ligase, which HIV established in infected cells to degrade the antiviral protein APOBEC3G/A3G, is enhanced by AMBRA1 knockdown. Recently, a role of AMBRA1 in the regulation of STAT3 activity through CRL5-SOCS3 was confirmed in cancer stem cells of group3 medulloblastoma, a childhood malignant brain tumor [21].

Interestingly, CRL4 may also promote self-ubiquitination and degradation of AMBRA1 [16]. In particular, AMBRA1 degradation occurs in prolonged starvation conditions, and it is required for terminating autophagy and avoiding excessive degradation of cellular components.

CRL4-AMBRA1 also mediates non-degradative ubiquitination of target proteins through lysine-63 ubiquitin chain linkage. It ubiquitinates SMAD4 to enhance its transcriptional functions in breast cancer cells to potentiate TGFB/TGFβ (transforming growth factor beta) signaling, promoting epithelial-to-mesenchymal transition/EMT, migration and invasion [22]. Moreover, CRL4-AMBRA1 ubiquitinates the autophagy regulator BECN1 upon nutrient deprivation to enhance its association with the lipid kinase PIK3C3/VPS34 (phosphatidylinositol 3-kinase catalytic subunit type 3) and promote its activity [23].

AMBRA1 also associates to and regulates the activity of other ubiquitin ligases, including TRAF6 (TNF receptor associated factor 6), TRIM32 (tripartite motif containing 32), HUWE1 (HECT, UBA and WWE domain containing E3 ubiquitin protein ligase 1) and PRKN [24–28], involved in autophagy signaling through both degradative and non-degradative ubiquitination of substrates, such as ULK1 (unc-51 like autophagy activating kinase 1), MFN2 (mitofusin 2), MCL1 (MCL1 apoptosis regulator, BCL2 family member). The role of these interactions will be discussed in detail in the following dedicated sessions.

- AMBRA1 interacts with ERLIN1, an endoplasmic reticulum lipid raft protein of the prohibitin family [29]. Upon autophagy induction by nutrient starvation, AMBRA1 and ERLIN1 cluster at the raft-like microdomains of ER-mitochondria contact sites to promote the autophagosome formation (Figure 2).

Figure 2.

Role of AMBRA1 in the regulation of autophagy. Description of the role of AMBRA1 in the regulation of ULK1 complex (ULK1, ATG13, ATG101 and RB1CC1) and BECN1 complex (BECN1, ATG14, PIK3C3/VPS34 and PIK3R4/VPS15) activity by regulative ubiquitination (ULK1 dimer not shown). To promote autophagy, AMBRA1 is recruited to lipid rafts present at the ER-mitochondria contact sites by interacting with ERLIN1. AMBRA1 mediates the non-degradative K63-linked polyubiquitination of ULK1 by the interaction with the ubiquitin ligases TRAF6 and TRIM32. Moreover, AMBRA1 is required for BECN1-PIK3C3/VPS34 complex stability via K63-linked polyubiquitination of BECN1, mediated by the interaction of AMBRA1 with the ubiquitin ligase CRL4 and TRAF6. The bottom panel reports different proteins that are responsible for the repressing AMBRA1 autophagic activity in unstressed conditions. PtdIns3P, phosphatidylinositol-3-phosphate. The figure was created with BioRender.com.

- AMBRA1 interacts with the catalytic subunit of the protein phosphatase PP2A to dephosphorylate the proto-oncogene MYC/c-MYC, leading to cell proliferation and tumorigenesis inhibition [30]. Moreover, AMBRA1-PPP2/PP2A control regulatory T-cell (T_reg) differentiation and maintenance, by promoting the stability of the transcriptional activator FOXO3, which, in turn, enhances FOXP3 expression [31]. This activity regulates T_reg induction in mouse models of both tumor growth and multiple sclerosis, highlighting a key role of AMBRA1 in immune homeostasis.

-AMBRA1 binds to the focal-adhesion kinase PTK2/FAK to regulate cell adhesion and migration in cutaneous squamous cell carcinoma [12]. By associating with PTK2, AMBRA1 limits the steady-state levels of active phospho-PTK2 and phospho-SRC (SRC proto-oncogene, non-receptor tyrosine kinase) at focal adhesions. When this binding is perturbed, p-SRC and p-PTK2 levels are elevated at the focal adhesion enhancing cell invasion. This AMBRA1 activity was validated in a mouse model of melanoma, in which AMBRA1 deficiency leads to hyperactive PTK2 signaling leading to enhanced epithelial-to-mesenchymal transition, cell motility and extracellular matrix metalloproteases expression [32].

-AMBRA1 contributes to cell death induced by dsRNA or Semliki Forest virus infection [33]. This proapoptotic activity of AMBRA1 is mediated by its interaction with the mitochondrial antiviral-signaling protein MAVS, favoring MAVS protein stabilization and its activation of CASP8.

-AMBRA1 interacts and is sequentially phosphorylated by CDK1 and PLK1 (polo like kinase 1) at mitosis [34]. These phosphorylation events are critical for proper spindle function and orientation. Phosphorylated AMBRA1 associates with NUMA1 (nuclear mitotic apparatus protein 1) and is responsible for NUMA1 localization at the cell cortex. Loss of AMBRA1 results in PLK1 protein stabilization and increased NUMA1 phosphorylation, leading to spindle orientation defects.

-A role of AMBRA1 in the transcriptional signaling pathway was unveiled by carrying out AP-MS of nuclear fractions [13]. Nuclear AMBRA1 binds and regulates the recruitment to chromatin of a network of chromatin modifiers and transcriptional regulators, including the PRKA/PKA-scaffold AKAP8, the Mediator complex component CDK9, and the transcription factor ATF2. Through this complex, AMBRA1 regulates the expression of genes involved in cancer cell invasion, including ANGPT1 (angiopoietin 1), TGFB2/TGFβ2, TGFB3, and the integrins ITGA8 and ITGB7.

AMBRA1 and regulation of autophagy

Autophagy consists of a sequence of molecular events leading to the formation of the specialized vesicles, called autophagosomes, that enwrap intracellular materials and deliver them to the lysosome for degradation [35,36]. Each step is under the control of specific sets of autophagy proteins (ATGs) whose activity is directly or indirectly regulated by several stress signaling pathways [37,38].

The ULK complex is composed of protein kinases ULK1 or ULK2, the scaffold protein RB1CC1/FIP200 (RB1 inducible coiled-coil 1) and the regulatory subunits ATG13 and ATG101. ULK1 activates autophagy by phosphorylating several ATGs, including members of the BECN1 complex [39,40].

The BECN1-PIK3C3/VPS34 complex is composed of the class III phosphatidylinositol 3-kinase PIK3C3/VPS34, the cofactors BECN1 and ATG14, and the scaffold PIK3R4/VPS15 [41] (Figure 2). This complex mediates the formation of phosphatidylinositol-3-phosphate/PtdIns3P, a membrane lipid that recruits autophagy factors involved in the nucleation of the autophagosome precursor membrane (termed phagophore). BECN1-PIK3C3/VPS34 is inhibited by antiapoptotic members of the BCL2 family, including BCL2, BCL2L1/BCL-xL (BCL2 like 1) and MCL1 [42], whose interaction is displaced by either stress signaling kinases, such as MAPK/JNK and DAPK (death associated protein kinase), or proapoptotic BH3 (BCL2-homology-3)-only proteins [41].

Recent evidence shows that ATG9 positive vesicles, arising from the trans-Golgi network and the endosomal compartment, act as seeds for generating the phagophore [43–46]. ATG9 vesicles both recruit ULK1 and BECN1 complexes and interact with the ER-resident protein ATG2A, which transfers lipids from the ER to the expanding phagophore together with ATG9 itself, acting as lipid scramblases to redistribute lipids to the inner leaflet (Figure 2).

Autophagosome maturation and cargo selection are regulated by the mammalian paralogs of yeast Atg8 (mAtg8s: MAP1LC3/LC3A, MAP1LC3B, MAP1LC3C, GABARAP, GABARAPL1 and GABARAPL2) that are inserted in the phagophore by covalent attachment to the lipid phosphatidylethanolamine [47,48].The membrane insertion site is determined by the presence of phosphatidylinositol-3-phosphate, which is recognized by the WIPI2 (WD repeat domain, phosphoinositide interacting 2) protein that bridges the E3 ligase to the phagophore by interacts with ATG16L1 [49,50]. LC3 interacts with autophagy cargo receptors, e.g., the SQSTM1/p62 (sequestosome 1)-like family proteins, which physically link the targeted structures to the autophagosome membranes, thus conferring selectivity to autophagic degradation [51].

AMBRA1 positively regulates autophagy mainly through the interaction and post-translational modification of ULK1 and BECN1-PIK3C3/VPS34 complexes [52] (Figure 2).

AMBRA1 binds to the ubiquitin ligase TRAF6, and promotes autophagosome formation by mediating non-degradative poly-ubiquitination of ULK1, which stimulates self-association, a prerequisite for kinase activation [24]. Moreover, AMBRA1 stimulates ULK1 activity by stabilizing the MTOR (mechanistic target of rapamycin kinase) inhibitor DEPTOR through the inhibition of CRL5 [16].

Interestingly, TRAF6 also ubiquitinates BECN1 within its BH3 domain preventing the inhibitory interaction with BCL2 [53]. Although the involvement of AMBRA1 in TRAF6-mediated ubiquitination of BECN1 remains unassessed, it was reported that CRL4-AMBRA1 ubiquitinates BECN1 at its C-term domain, which enhances BECN1-PIK3C3/VPS34 association [23]. AMBRA1-mediated ubiquitination of BECN1 is negatively regulated by WASH, a factor whose deficiency in mice causes embryonic lethality associated with excessive autophagy [23].

Different mechanisms restrain the autophagic activity of AMBRA1 in unstressed conditions. AMBRA1 activity is suppressed by the MTOR complex MTORC1 phosphorylation on Ser52 [24]. Moreover, different pools of AMBRA1 are sequestered on microtubules through the association with the dynein motor [15], and on mitochondria outer membrane (OMM) by interacting with BCL2 [54]. Upon nutrient starvation, ULK1 phosphorylates AMBRA1 on Ser465/635, freeing it from negative partners and allowing its translocation to the ER to interact with the BECN1 complex [55].

AMBRA1 activity is also tightly regulated to limit the process and avoid excessive cellular self-degradation [16]. CRL4-mediated degradation of AMBRA1 occurs both in nutrient-rich conditions to constitutively restrict AMBRA1 abundance and upon prolonged starvation to terminate the autophagic response. The ring finger ubiquitin ligase RNF2 has also been reported to contribute to limit AMBRA1 abundance [23].

AMBRA1 is also degraded by calpains and caspases upon apoptosis induction to inhibit the pro-survival activity of autophagy [56]. The caspase cleaved AMBRA1 fragment may acquire a proapoptotic function, undergoing a conformational change that increases its binding with antiapoptotic factors of the BCL2 family, through a BH3-like domain, to inhibit their pro-survival activity and establish a cell death amplification loop [57]. Interestingly, the viral oncogene E7 of human papilloma virus binds AMBRA1 and promotes calpain-mediated degradation to limit autophagy [58].

AMBRA1 and regulation of mitophagy

Mitophagy, a selective form of autophagy that targets damaged or unnecessary mitochondria for lysosomal degradation, is essential for development and adult homeostasis [59,60]. Selective mitochondrial degradation is ensured by cargo receptors that mediate the interaction of the outer mitochondria membrane with autophagosome proteins of the mAtg8 family [61]. Interestingly, their binding affinity to mAtg8 is regulated by phosphorylation of residues close to the LC3-interacting region (LIR) mediated by multiple stress signaling pathways.

Different cargo receptors are responsible for mitophagy execution in response to different mitochondrial damaging conditions. In hypoxic conditions, the transmembrane proteins BNIP3 (BCL2 interacting protein 3) and BNIP3L/NIX (BCL2 interacting protein 3 like) accumulate on the OMM upon transcriptional upregulation by HIF1A/HIF1α (hypoxia inducible factor 1 subunit alpha) [62]. The transmembrane protein FUNDC1 (FUN14 domain containing 1) is also involved in hypoxia-induced mitophagy, and its activity is mainly regulated by phosphorylation mediated by CK2, ULK1 and SRC kinases [63]. BNIP3L/NIX also contributes to mitophagy upon ROS accumulation triggered by oxidative phosphorylation and during erythrocyte maturation to trigger mitochondrial clearance [62].

Upon mitochondrial membrane depolarization or accumulation of misfolded mitochondrial proteins, damaged mitochondria are labeled by polyubiquitination of OMM proteins which recruit the cytosolic cargo receptors of the SQSTM1/p62 family [64]. Binding affinity of these receptors to both LC3 and to ubiquitin chains is increased by phosphorylation, mainly mediated by TBK1 (TANK binding kinase 1) [65–67]. An important role in this process is played by PINK1 and PRKN, two genes frequently mutated in familiar early onset forms of Parkinson disease (PD) [68,69]. The protein kinase PINK1 (PTEN induced kinase 1) rapidly accumulates on the OMM and mediates the recruitment and activation of the ubiquitin ligase PRKN through phosphorylation of both ubiquitin chains and PRKN itself within its ubiquitin-like domain [70,71]. Polyubiquitin chains newly synthesized by PRKN are then phosphorylated by PINK1 to recruit additional PRKN molecules, leading to the ubiquitination of a large number of OMM proteins in a feedforward mechanism [72]. Moreover, PRKN activity is required for the exposure of mitochondrial proteins of internal compartments, such as PHB2 (prohibitin 2), NIPSNAP1 and NIPSNAP2, acting also as mitophagy receptors [73].

Ubiquitination of OMM proteins does not exclusively rely on PRKN, as other E3-ubiquitin ligases, such as MUL1, ARIH1, SIAH1, SMURF1 and AMFR/Gp78, cooperate with, or act alternatively to, PRKN downstream of PINK1 [74]. The function of the PINK1-PRKN axis is not limited to prime mitochondria for autophagosome engulfment but it also triggers mitochondrial fission to isolate damaged mitochondria by degrading mitofusins [75], as well as the translocation of the autophagy initiation complexes ULK1 and BECN1 to mitochondria [76,77] (Figure 3).

Figure 3.

Role of AMBRA1 in the regulation of mitophagy. Description of the role of AMBRA1 in the regulation of both PINK1-PRKN-dependent and -independent mitophagy. Upper part of the figure: AMBRA1 is required for the accumulation of PINK1 on the OMM following mitochondrial membrane depolarization by interacting with ATAD3A and the TOMM complex. Moreover, AMBRA1 interacts with PRKN to recruit the BECN1-PIK3C3/VPS34 complex and promote autophagosome formation in proximity to damaged mitochondria. Lower part of the figure: AMBRA1 inhibits mitochondria fusion by mediating the degradative degradation of MFN2 by the ubiquitin ligase HUWE1. Moreover, AMBRA1 acts as a mitophagy cargo receptor by interacting with LC3-II on the OMM. This interaction is stimulated by AMBRA1 phosphorylation mediated by CHUK/IKKα. Mitophagic activity of AMBRA1 is inhibited by BCL2 family members. The figure was created with BioRender.com.

Rapid accumulation of PINK1 on damaged mitochondria results from changes in protein stability [78]. In healthy state, PINK1 is imported into mitochondria through the outer and inner membrane translocases (TOMM and TIMM, respectively) [79]. Here, PINK1 is cleaved by the mitochondrial processing peptidase PMPCA-PMPCB and PINK1-PGAM5-associated rhomboid-like protease PARL and then retro-translocated to the cytosol for proteasomal degradation. Other factors involved in PINK1 import are SAMM50, a channel-forming protein of the sorting and assembly machinery complex, and ATAD3A, an OMM-IMM transmembrane AAA⁺ ATPase, whose depletion results in OMM accumulation of PINK1 even in the absence of mitochondrial depolarization [80,81]. Mitochondrial depolarization prevents PINK1 cleavage and stabilizes the protein on TOMM, facilitating PINK1 dimer formation and intermolecular phosphorylation [71]. Interestingly, when the interaction with TOMM is perturbed, PINK1 is internalized and degraded by a different mitochondrial metalloendopeptidase named OMA1 [82]. Other factors required for PINK1 stability in the OMM are IMMT/Mic60/Mitofilin, a core component of the mitochondrial contact site and cristae organizing system, and PHB2, an IMM-resident membrane scaffold protein required to inhibit PARL activity [83,84]. In addition to mitochondria depolarization, PINK1 stabilization is induced during the mitochondrial unfolded protein response as a consequence of inhibition of the mitochondrial matrix protease LONP1 [85].

Regulation of PINK1-PRKN-dependent mitophagy by AMBRA1

AMBRA1 regulates various types of mitophagy in either PINK1-PRKN-dependent or independent manner (Figure 3). First evidence of a role of AMBRA1 in mitophagy came from a functional proteomic screening which identified AMBRA1 as a PRKN binding protein [86]. PRKN does not target AMBRA1 for ubiquitination, while it promotes the recruitment of AMBRA1 to depolarized mitochondria together with the BECN1-PIK3C3/VPS34 complex, allowing autophagosome formation in close proximity of the damaged organelle.

Recently, a more upstream role of AMBRA1 in PRKN mitophagy has been characterized. AMBRA1 interacts with PINK1 on the mitochondrial outer membrane and contributes to its protein stability upon mitochondrial depolarization [87]. Inhibition of AMBRA1 expression results in reduced amount of PINK1 on OMM, with a consequent impairment of PRKN ubiquitin, phosphorylation and translocation to the OMM. From a mechanism point of view, AMBRA1 interacts with TOMM, favoring PINK1 accumulation on this structure, and with ATAD3A, preventing its ability to promote PINK1 import and degradation mediated by the mitochondrial protease LONP1.

Regulation of PINK1-PRKN-independent mitophagy by AMBRA1

AMBRA1 contributes to two critical steps during PRKN-independent mitophagy: i) inhibition of the fusion process to isolate damaged mitochondria from the organelle network; ii) selective sequestration of damaged mitochondria by the autophagosome membrane. Overexpression of AMBRA1 using a mitochondrial targeting sequence (AMBRA1-ActA) causes a massive induction of mitophagy in PRKN-deficient cells, characterized by depolarization and OMM ubiquitination [88]. This event is accompanied by the proteasomal degradation of the mitochondrial fusion protein MFN2 (mitofusin 2). Interestingly, MFN2 degradation, which is also mediated by PRKN, is achieved by AMBRA1-ActA interacting with a different ubiquitin ligase, called HUWE1 [26]. In particular, AMBRA1 acts as a cofactor for HUWE1 activity favoring its translocation to damaged mitochondria and its interaction with MFN2.

An important role in the regulation of AMBRA1-HUWE1 interaction is played by MCL1, a member of the BCL2-family [27]. In unstressed conditions, MCL1 inhibits HUWE1 recruitment to mitochondria. Upon mitochondria damage, MCL1 is phosphorylated by GSK-3β, which promotes its ubiquitination by HUWE1 and consequent proteasomal degradation.

Once active, mitochondrial AMBRA1 contributes to the phagophore engulfment of ubiquitinated mitochondria through a direct interaction with LC3 through its LIR [88]. The affinity of the mitophagy receptor AMBRA1 to LC3 is enhanced by phosphorylation of a residue just upstream of its LIR mediated by the CHUK/IKKα kinase (component of inhibitor of nuclear factor kappa B kinase complex) [26]. Interestingly, downregulation of HUWE1 expression reduces AMBRA1 phosphorylation upon mitophagy induction, suggesting that mitochondrial ubiquitination mediated by HUWE1 is also a signal to stimulate CHUK/IKKα activity (Figure 3). Importantly, evidence of a physiological role of PRKN-independent mitophagy regulated by AMBRA1 was obtained in a cellular model of ischemia, based on oxygen-glucose deprivation followed by reoxygenation [26,89]. This treatment leads to a rapid phosphorylation of AMBRA1 and mitophagy induction, which is prevented by inhibiting AMBRA1, HUWE1 and CHUK/IKKα activity.

Interestingly, AMBRA1 may also contribute to an alternative process to eliminate damaged mitochondria by promoting secretory autophagy aimed at transferring damaged mitochondria to neighbor cells. In degenerating dopaminergic terminals, mitophagy begins in neurons but degradation occurs in the surrounding astrocytes (transmitophagy) [90]. Axonal fragmentation is accompanied by accumulation of damaged mitochondria in spheroid structures positive to PINK1, PRKN, PINK1 phosphorylated ubiquitin, AMBRA1, BCL2L13 and autophagosome markers. However, these spheroids do not progress toward lysosome degradation but are transferred to astrocytes where they matured to autolysosomes. In this manner, transmitophagy prevents neuroinflammation caused by the extracellular release of mitochondria contents.

Of note, in other cellular systems, AMBRA1-regulated secretion of mitochondria-positive vesicles triggers signals that contribute to pathological progression [91], as described upon myocardial ischemia/reperfusion injury, which induces cardiomyocyte damage associated to fibroblast activation and fibrosis. In ischemic conditions, cardiomyocytes release AMBRA1-positive extracellular vesicles carrying mitochondrial components that are internalized by fibroblasts, where the delivered mitochondrial DNA activates the CGAS-STING1 pathway and promotes fibroblast activation and proliferation. Downregulation of cardiac-specific AMBRA1 expression inhibits the extracellular vesicle release and prevents ischemic myocardial fibrosis. This observation suggests that, at variance with astrocytes, mitochondrial contents is not delivered to cardiac fibroblasts to activate transmitophagy but rather it acts as a signal to induce fibrosis.

AMBRA1-regulated mitophagy in aging-related neuronal disorders

AMBRA1 and Parkinson disease

An increasing number of studies reported evidence of alteration of AMBRA1-regulated mitophagy in neurodegenerative diseases. PD is one of the most common neurodegenerative motor disorders affecting 1% of the population over age 50 [92]. Most cases of PD are sporadic, caused by a complex interplay of genetic and environmental factors [93]. About 15% of cases of PD are inherited, caused by mutations in genes mostly involved in mitochondrial functions. Pathologic features of PD are the selective loss of dopamine neurons in the ventral substantia nigra associated to ubiquitinated protein deposits (Lewy bodies), having SNCA/α-synuclein as a primary structural component [92]. What triggers cell death in PD remains incompletely understood. Most evidence suggests the important contribution of altered mitochondria-related processes, including impaired mitochondrial complex I function, intracellular redox state, dysfunction of proteasome- and autophagy-degradative processes, and activation of proapoptotic factors [94]. Aging remains the biggest risk factor for developing sporadic PD, and it is still largely unknown which specific age-related factors predispose individuals to develop this neurodegenerative disease [95].

A direct link between PD and mitophagy emerged when PINK1 and PRKN, two genes mutated in the familial form of this disease, were characterized as mitophagy regulators [68,69]. Absence of neuronal phenotypes in pink1 and prkn knockout mice apparently failed to confirm the relevance of these genes in protecting neurons from premature cell death, although they showed increased sensitivity to neurodegeneration when exposed to additional stresses and insults, such as SNCA/α-synuclein or MPTP induced toxicity, as well as intestinal bacterial pathogens [96]. Recently, this apparent contradiction was in part clarified by using monkey models. In fact, PINK1 inactivation in both monkey embryonic and adult brain results in neurodegeneration and motor deficits [97,98], suggesting that rodent brains are more resistant to Pink1 deficiency when compared to the humans and primates. The authors proposed that this specificity could rely on a higher basal expression of PINK1 in human brain compared to rodents, as well as to a cytoplasmatic localization of the protein in addition to mitochondrial one [97,98].

Recent evidence suggests that inhibition of the mitophagic activity of AMBRA1 are associated with PD pathogenesis (Figure 4). A first example is represented by the reduction of AMBRA1 levels in patients with mutations in the GBA (glucosylceramidase beta) gene [99], which codes for a lysosomal enzyme that degrades glucosylceramide to ceramide and glucose. While homozygous GBA mutations cause a lysosomal disorder, named Gaucher disease, heterozygous GBA mutations are the most common genetic risk factor for PD, found in 7–20% of all PD cases [100]. Heterozygous GBA mutations are associated to an earlier age at onset, greater cognitive decline, and a faster rate of disease progression [101]. An impairment of mitophagy was observed in vitro in cultured hippocampal neurons, and in vivo in mouse brain expressing the GBA^L444P mutant protein, as well as in postmortem anterior cingulate cortical tissue from PD patients with heterozygous GBA mutations [99]. In cells expressing the GBA^L444P mutant protein, mitophagy is affected both at the early and late stages [99]. Decreased levels of wild type GBA impair lysosomal function, limiting their ability to degrade autophagosome-delivered materials. Mice expressing the mutant GBA protein show defective mitophagy with reduced recruitment of PRKN, NBR1, and BNIP3L to mitochondria. This is paralleled by a decrease in the total protein levels of AMBRA1 and BECN1, observed both in vitro using neuronal cells overexpressing the mutant GBA^L444P and in vivo analyzing brain tissues from GBA-PD patients and GBA^L444P heterozygous mice.

Figure 4.

Role of AMBRA1 in the regulation of mitophagy in aging-related degenerative disorders. Description of the role of AMBRA1 in the regulation of mitophagy in age-relate degenerative diseases. Upper panel: decreased AMBRA1 levels in neuronal cells overexpressing GBA^L444P mutant and brain tissues from GBA-PD patients and GBA^L444P heterozygous mice. MIR103A-3p, inhibits AMBRA1 expression levels, as well as PINK1 and PRKN expression, in pharmacologically-induced cellular and mouse models of PD. AMBRA1 associates with SNCA/α-synuclein and this interaction is increased by the phosphorylation of SNCA/α-synuclein observed in Lewy bodies. Right panel: AMBRA1 expression was observed to decrease in cellular and mouse models of Alzheimer disease expressing pathogenic APP mutant or APOE4. Decreased AMBRA1 is associated to a reduction of mitophagy. Left panel: defects in the retinal pigment epithelium (RPE) were observed in heterozygous ambra1 mice showing a reduction in mitochondrial membrane potential, accumulation of protein aggregates, increased neuroinflammation and an impaired antioxidant response, triggering age-related vision loss. Lower panel: AMBRA1 is an important factor regulating both muscle homeostasis and muscle atrophy. AMBRA1 interacts with the ubiquitin ligase TRIM32, promoting the regulative K63-linked polyubiquitination of ULK1 and the induction of mitophagy in muscle cells. The figure was created with BioRender.com.

The mechanisms for the reduction in AMBRA1 and BECN1 levels by mutant GBA remain to be elucidated, but it is likely to depend on different regulative events, from gene expression to protein stability. For example, the ER retention of mutant GBA protein causes an unfolded protein response that may activate calpains and caspases [102], which are known to target AMBRA1 and BECN1 for degradation [56,103].

Of note, reduction of AMBRA1 levels, as well as of PINK1 and PRKN, were also observed in pharmacologically-induced cellular and mouse models of PD [104]. A role in this downregulation is played by MIR103A-3p, a miRNA whose levels are increased in human brain or plasma samples from PD patients, as well as in MPTP-treated cells and mice [104]. At the molecular level, MIR103A-3p targets the 3’ UTR of PRKN, while the effect on AMBRA1 mRNA was not investigated in detail.

AMBRA1 is also involved in multiple system atrophy (MSA), an adult-onset, sporadic, rapidly progressing, fatal neurodegeneration combining symptoms of cerebellar ataxia, poorly L-dopa-responsive parkinsonism and dysautonomia [105]. MSA is pathologically characterized by glial cytoplasmic inclusions/GCIs, and, to a lesser extent, neuronal cytoplasmic inclusions/NCIs, neuronal nuclear inclusions/NNIs and glial nuclear inclusions/GNIs [105]. Brain specimen from MSA patients shows an intense AMBRA1 staining in glial cytoplasmic inclusions, neuronal cytoplasmic inclusions and threads, which are also positive for SNCA/α-synuclein, while neuronal nuclear inclusions and glial nuclear inclusions were negative for AMBRA1 [106]. Notably, AMBRA1 associates with SNCA/α-synuclein and this interaction is increased by the phosphorylation of SNCA/α-synuclein observed in Lewy bodies. Modulation of AMBRA1 levels impacts on SNCA/α -synuclein levels [106]. Ectopic expression of AMBRA1 in mouse primary cultured neurons increases autophagy flux and decreases phospho- SNCA/α-synuclein levels, while AMBRA1 silencing significantly increases SNCA/α-synuclein dot-like structures in the cytoplasm [106], suggesting that AMBRA1 is critical to clear SNCA/α-synuclein inclusions through autophagy degradation.

AMBRA1 and Alzheimer disease

A decrease in AMBRA1 expression was reported in samples from Alzheimer patients and in in vitro and in vivo models of Alzheimer disease (AD). AD is the most common type of dementia affecting thought, memory and language in 1 out of 10–15 people over the age of 65 [107]. The pathological cause of AD is considered to be the senile plaque formed by amyloid beta (Aβ) and neurofibrillary tangles/NFTs composed of phosphorylated MAPT/tau protein, mainly in cerebral cortex and hippocampus [108]. AD is mainly late onset and sporadic disease/SAD, although early-onset and familial cases are also present/FAD [109]. Familial AD is mainly associated with mutations in the Aβ precursor APP (amyloid beta precursor protein) and presenilin genes PSEN1 and PSEN2, which are the catalytic subunit of the γ-secretase complex that processes APP [110]. SAD has a complex etiology, involving genetic, environmental, metabolic and viral factors. The APOE (apolipoprotein E) variant 4 (APOE4) is the strongest genetic risk factor for SAD [111]. The higher affinity of APOE4 for Aβ with respect to other APOE isoforms is thought to interfere with the processing of Aβ [112].

Alzheimer is a disease whose symptoms worsen over the years and the disease process is thought to begin several years before the first symptoms appear [113]. Although, many processes involved in AD pathogenesis have been identified, the molecular mechanisms responsible for AD onset and progression are still elusive.

AMBRA1 expression was observed to decrease in a mouse model of AD expressing a pathogenic APP mutant that results in elevated levels of Aβ and ultimately amyloid plaques [114]. In detail, in neocortex AMBRA1 levels are significantly lower in 3-month old transgenic animals than in WT animals, and remain low at 6 months of age when these mice show in vivo memory impairment. Interestingly, BECN1 and LC3 levels are unchanged at the onset of AD pathology. Lower AMBRA1 levels are also observed in the hippocampus of transgenic mice at 6 months of age. Notably, AMBRA1 decrease was recently confirmed in iPSC-derived neuronal cultures expressing either a different APP mutant or APOE4 [74]. AMBRA1 decrease is associated to mitophagy inhibition and a reduction of other mitophagy proteins, including FUNDC1, BCL2L13, MUL1, as well as the phosphorylated active forms of TBK1 and ULK1. Importantly, two potent neuronal mitophagy-inducing agents, urolithin A, and actinonin, are able to rescue mitophagy defects by potentiating the expression of AMBRA1, PINK1, PRKN, BECN1, BCL2L13 and phospho-ULK1 [74] (Figure 4).

AMBRA1 and age‐related macular degeneration

A role of AMBRA1 in age‐related macular degeneration (AMD) has been characterized in last years.

The retina is an organ that has one of the highest metabolic rates body-wide and is constantly exposed to photooxidative damage and external stressors. Of note, autophagy decreases with aging in the retina, and autophagy stimulation has neuroprotective activity [115,116]. Although age-associated autophagy downregulation has been described in the retina, organ specific increases in mitophagy have also been reported as a counteracting mechanism to limit inflammation triggered by cytosolic mitochondrial DNA release during aging [117].

AMD is the leading cause of vision loss in the elderly population. The progression of AMD is initially characterized by atrophic alterations in the retinal pigment epithelium (RPE) and formation of lysosomal lipofuscin and extracellular deposits [118]. Located immediately outside the neuroretina, the RPE provides nutrients and trophic support to photoreceptors, and mediates recycling of outer segment photoreceptor, thereby eliminating toxic intermediates generated in the phototransduction process [119]. The RPE of AMD patients contains mitochondria with abnormal cristae and reduced mitochondrial respiration and glycolytic function, causing an energetic imbalance that ultimately impacts the adjacent neuroretina [120,121]. Regarding mitophagy, decreased levels of PINK1 have been described both in vivo [122] and in primary RPE cells from AMD donors in response to mitochondrial uncoupling [123].

An indication of the involvement of AMBRA1 in AMD has emerged from a transcriptomic analysis of samples from AMD patients showing that AMBRA1 expression is upregulated in retina of the macular region of AMD patients, suggesting an active contribution to the autophagic response to retinal damage [124]. The functional contribution of AMBRA1 to AMD has been in-depth investigated using heterozygous ambra1 mice as a viable organism with a moderate reduction in autophagic flux [124]. Defects in the RPE were observed in young mutant animals showing a reduction in mitochondrial membrane potential and in number of autophagosomes, and accumulation of protein aggregates. Aged ambra1 mutant mice show increased mitochondrial mass in the absence of mitochondrial biogenesis suggesting a defect in mitophagy. This is accompanied by accumulation of lipofuscin, increased neuroinflammation and an impaired antioxidant response, which eventually causes exacerbated age-related vision loss [124] (Figure 4). RPE and photoreceptors of ambra1 heterozygous mutant mice are also more sensitive to sodium iodate, a pharmacological model of AMD-associated atrophy, and show reduced survival of retinal ganglion cells after optic nerve injury in old animals. This is associated to impaired oxidative stress response and defective mitophagy as indicated by reduced increase in BNIP3L and BNIP3 levels [125].

Altogether, these results suggest that the defective mitophagy caused by AMBRA1 deficiency may play a key role in aging-related neurodegeneration, boosting this process could constitute a valid therapeutic strategy to limit disease progression.

AMBRA1-regulated mitophagy and muscle aging

Mitochondria are critical for muscle health ensuring the execution of key processes for muscle activity, such as energy production, Ca²⁺ homeostasis and programmed cell death [124]. Mitophagy contributes to mitochondrial homeostasis by rapidly removing dysfunctional mitochondria which could cause muscle damage by generating excessive ROS [126].

Muscle atrophy is a highly regulated process aimed at reducing muscle mass during prolonged inactivity, preserving cell survival and allowing tissue regrowth when activity is resumed [127]. Muscle cells are not damaged by atrophy unless aberrantly induced by pathological stimuli, muscle gene mutations, chronic inflammation or metabolic dysfunction [128]. Excessive/chronic muscle atrophy is a major health problem in elderly people, with no treatment available to counteract the progressive decline in skeletal muscle mass and strength, a process called sarcopenia [129]. Both proteasome and autophagy degradation contribute to the dismantling of muscle fibers [130]. In particular, autophagy mainly targets supernumerary organelles, such as mitochondria, preventing the accumulation of harmful metabolic products, such as ROS [127]. Notably, accumulation of dysfunctional mitochondria contributes to the development of sarcopenia, which is parallel by decreased mitophagy activity [131].

How mitophagy is regulated in normal or atrophic muscle cells remains partly characterized. Interestingly, recent works identifies AMBRA1 as an important factor regulating both muscle homeostasis and muscle atrophy in vivo. Generation of skeletal muscle-specific ambra1 knock-out mice showed that ambra1 deletion results in decreased myofiber sizes in both young and adult mice [132]. In addition, AMBRA1 deficiency displayed a switch from oxidative to glycolytic myofibers. These alterations were confirmed by global gene expression analysis of tibialis anterior muscles, which show altered expression of genes coding for proteins involved in sarcomere function as well as for mitochondria-related proteins. In line with these results, Ambra1-deficient muscles accumulate myofibers with swollen mitochondria with disorganized cristae associated to impaired regulation of mitochondrial membrane potential, while the rate of mitochondria biogenesis is not significantly affected. Notably, mitophagy is severely impaired in Ambra1-deficient mice, while the bulk autophagic flux is not appreciably affected. Mitophagy defects are characterized by a reduced mitochondrial recruitment of both DNM1L/DRP1 (dynamin 1 like), a regulator of mitochondria fission, and PRKN. Moreover, an impaired mitophagy flux was also demonstrated by analyzing the recruitment of lipidated LC3 levels in isolated mitochondria-enriched fractions from muscles upon starvation, which is severely reduced in Ambra1-deficient mice.

Role of AMBRA1 in muscle atrophy

Evidence that AMBRA1 plays a role in regulating autophagy during muscle atrophy were initially suggested by the interaction of AMBRA1 with the ubiquitin ligase TRIM32 [25,133,134]. Although expressed in various cell types [135–137], TRIM32 has a prevalent role in muscle cells as indicated by the fact that TRIM32 gene mutations are causative of limb-girdle muscular dystrophy 2 H (LGMD2H), recently renamed as limb-girdle muscular dystrophy recessive 8 (LGMDR8) [138,139]. TRIM32 regulates muscle atrophy as observed in trim32 knockout mice that show impaired muscle fibers regrow after arm immobilization [140]. In this context, TRIM32 is required for both preventing premature senescence of satellite cells by degrading the E3 SUMO ligase PIAS4 [140] and autophagy induction in differentiated muscle cells [25]. Autophagy inhibition in TRIM32-deficient myotubes causes the exacerbation of muscle damage in response to atrophic stimuli, as shown by high levels of mitochondrial ROS and increased induction of the atrophic gene TRIM63/MURF-1, which suggest an impairment of the mitophagic activity of AMBRA1. These defects were confirmed in Trim32 knockout mice and in muscle cells carrying LGDM2H mutants. The proautophagic activity of TRIM32 relies on its ability to bind and activate ULK1, through unchained ubiquitination, to phosphorylate BECN1 complex components and promote autophagosome formation. Importantly, AMBRA1 plays a key role in TRIM32 signaling by promoting TRIM32-ULK1 interaction and ULK1 ubiquitination (Figure 4).

Conclusion and Future perspectives

It is now evident that AMBRA1, in post mitotic cells of neuronal and muscle tissues, plays a key role in the regulation of mitochondrial homeostasis, guaranteeing effective mitophagy both in basal and stressed conditions. Notably, a defective mitophagic activity of AMBRA1 is associated to some relevant age-related degenerative disease, suggesting that AMBRA1 may represent an important therapeutic target to prevent disease onset or reduced its progression.

Thus the development of competitive inhibitors specific for the binding of AMBRA1 with the CUL4-DDB1 complex and/or caspases and calpains, responsible for regulating the degradation of AMBRA1 might represent valid tools for pharmacologically intervening in the development of these diseases. The concern in this approach is represented by the inhibition of regulatory complexes that are involved in multiple pathways at the basis of cellular proteostasis and thus their inhibition will not be specific. The more feasible possibility is to identify small molecules or specific peptides able to block the interface of interaction between AMBRA1 and Cullin–RING ubiquitin ligases (CRLs). This approach is under evaluation, although the disordered 3D AMBRA1 structure might represent a major problem in designing specific inhibitors.

Numerous studies have highlighted how the reduced expression of AMBRA1 is correlated to the development of degenerative diseases, as a consequence of the ability of AMBRA1 to regulate the function of numerous E3 ligases and kinases (PRKN, PINK, HUWE1 and TRIM32). In fact, these enzymes are involved in the autophagic degradation of damaged mitochondria, a cellular homeostatic mechanism whose alteration is among the factors underlying the development of numerous degenerative diseases. Therefore, the in-depth study of the molecular mechanisms through which AMBRA1 is able to positively regulate the activity of various E3 ligases and their regulated pathways is of fundamental importance to identify possible pharmacological strategies to compensate for its reduced expression in certain pathologies.

Another possible approach is the block of AMBRA1 gene expression, however, there is a lack of studies on the AMBRA1 promoter that would be useful to identify transcription factors or microRNAs involved in the regulation of AMBRA1 expression. Indeed, the specific modulation of AMBRA1 expression by miRNAs such as MIR200B and MIR103A-3p in neurological setting including the Parkinson disease has been recently reported [104,141]. In keeping with this, to pharmacologically modulate the expression of AMBRA1, experimental studies aimed at identifying the transcription factors on the promoter or mRNA epigenetic regulation such as the m6A methylation that block or stabilize the mRNA of AMBRA1 in those specific pathologies would be of interest.

In conclusion, although the loss or reduced expression of AMBRA1 has been correlated to the development of numerous degenerative diseases linked to aging, the molecular mechanisms underlying the defect in expression and/or function of AMBRA1 in these pathologies have not been completely clarified, studies aimed at identifying specific factors capable of inhibiting the expression and/or function of AMBRA1 in a specific pathological context, appear to be fundamental importance. Furthermore, given the presence of a genetic component in the development of some of these degenerative diseases, it would be interesting to investigate whether the presence of mutations in the AMBRA1 gene is associated to the development of degenerative diseases.

Funding Statement

This study was supported by grants from Ministry of Health: Ricerca Corrente 2024 LINEA 3 – PROGETTO 2 and Ricerca Finalizzata PNRR-MAD-2022-12375755 to GMF. We acknowledge co-funding from Next Generation EU through the Italian Ministry of University and Research, in the context of the National Recovery and Resilience Plan, Investment PE8 – Project Age-It: “Ageing Well in an Ageing Society” to Gian Maria Fimia, co-financed by the Next Generation EU (DM 1557 11.10.2022) CUP B53C22004090006, and PE6 “Heal Italia” to Mauro Piacentini CUP E83C22004670001. We also acknowledge AIRC (IG26394 to GMF). We also acknowledge funds from 5 × 1000 tax donation to MDR.

Disclosure statement

No potential conflict of interest was reported by the author(s).