InteractiveFly: GeneBrief

ariadne 1: Biological Overview | References

Ariadne-1 (Ari-1) is an E3 ubiquitin-ligase essential for neuronal development, but whose neuronal substrates are yet to be identified. To search for putative Ari-1 substrates, this study used an in vivo ubiquitin biotinylation strategy coupled to quantitative proteomics of Drosophila heads. Sixteen candidates were identified that met the established criteria: a significant change of at least two-fold increase on ubiquitination, with at least two unique peptides identified. Amongst those candidates, Comatose (Comt), the homologue of the N-ethylmaleimide sensitive factor (NSF), which is involved in neurotransmitter release, was identified. Using a pulldown approach that relies on the overexpression and stringent isolation of a GFP-fused construct, Comt/NSF was validated to be an ubiquitination substrate of Ari-1 in fly neurons, resulting in the preferential monoubiquitination of Comt/NSF. The possible functional relevance of this modification was tested using Ari-1 loss of function mutants that displayed a lower rate of spontaneous neurotransmitter release due to failures at the pre-synaptic side. By contrast, evoked release in Ari-1 mutants was enhanced compared to controls in a Ca(2+) dependent manner without modifications in the number of active zones, indicating that the probability of release per synapse is increased in these mutants. This phenotype distinction between spontaneous versus evoked release suggests that NSF activity may discriminate between these two types of vesicle fusion. These results thus provide a mechanism to regulate NSF activity in the synapse through Ari-1-dependent ubiquitination (Ramirez, 2021).

Neurotransmitter release is mediated by a set of protein-protein interactions that include the N-ethylmaleimide sensitive factor (NSF), soluble NSF attachment proteins (SNAPs), and SNAP receptors (SNAREs). These proteins assemble into a tripartite complex in order to elicit synaptic vesicle fusion, which is formed by one synaptic vesicle membrane SNARE protein (v-SNARE), Synaptobrevin, and two plasma membrane SNARE proteins (t-SNAREs), Syntaxin and the 25-kDa synaptosome-associated protein. Following vesicle fusion, the tripartite SNARE complex disassembles by the activities of NSF and SNAPs. Free t-SNAREs from the plasma membrane can then participate in new priming reactions, while the v-SNAREs can be incorporated into recycled synaptic vesicles. These interactions, also routinely used for intracellular vesicle trafficking in all cell types, are conserved across species, including Drosophila (Ramirez, 2021).

Deviations on the rate of neurotransmitter release are at the origin of multiple neural diseases, including Parkinson's disease. Under physiological conditions, the Leucine-rich repeat Serine/Threonine-protein kinase 2 (LRRK2) phosphorylates NSF to enhance its ATPase activity, which facilitates the disassembly of the SNARE complex. However, the most common Parkinson's disease mutation in LRRK2 causes an excess of kinase activity that interferes with the vesicle recycling. Similarly, α-Synuclein, another Parkinson's disease protein, alters neurotransmitter release by preventing the v-SNARE vesicle-associated membrane protein (VAMP)-2, also known as Synaptobrevin-2, from joining the SNARE complex cycle. Correct neural functioning, therefore, requires delicate regulation in vesicle trafficking. This regulation can be achieved by posttranslational modifications, such as ubiquitination. In fact, ubiquitination of certain proteins can affect their activity or life span. At the presynaptic side, for example, increased neurotransmitter release correlates with decreased protein ubiquitination. Similarly, acute pharmacological proteasomal inhibition causes rapid strengthening of neurotransmission (Ramirez, 2021).

Ariadne 1 is an E3 ubiquitin-ligase, first identified in Drosophila, from a conserved gene family defined by two C3HC4 Ring fingers separated by a C6HC in-Between-Rings domain (the RBR motif). Ari-1 had been described to be essential for neuronal development, and its mutants reported to exhibit reduced eye rhabdomere surface and endoplasmic reticulum, as well as aberrant axonal pathfinding. However, despite its importance, no neuronal substrates have been reported so far. Only three Ari-1 substrates have been postulated, either in cultured cells or in vitro, while three Parkin substrates were reported to interact with Ari-1 in COS-1 cells. For this reason, with the aim to identify neuronal Ari-1 substrates, advantage was taken of two methodologies. The first one, the ^bioUb strategy, allows the identification of hundreds of ubiquitinated proteins from neuronal tissues. The system relies on the overexpression of a tagged ubiquitin that bears a 16 amino acid long biotinylatable peptide, which can be biotinylated by the Escherichia coli biotin holoenzyme synthetase enzyme (BirA) in neurons in vivo. Remarkably, this approach can be efficiently applied to identify neuronal E3 ligase substrates. In contrast, the second methodology favors the isolation of GFP-tagged proteins under denaturing conditions to further characterize their ubiquitination pattern under the presence or absence of an E3 ligase (Ramirez, 2021).

This study has combined the ^bioUb strategy with the overexpression of Ari-1 and identified 16 putative neuronal substrates of Ari-1. Among those, focus was placed on Comatose (Comt), the fly NSF orthologue, due to its relevance in normal and pathological function at the synapse. By the isolation of GFP-tagged Comt from Drosophila photoreceptor neurons overexpressing Ari-1, this study confirmed Comt/NSF as an Ari-1 ubiquitin substrate and showed that it is mostly monoubiquitinated. Furthermore, Ari-1 loss-of-function mutants displayed lower rate of spontaneous neurotransmitter release, but enhanced evoked release, due to failures at the presynaptic side. These defects in the mutants are compatible with a deregulation of Com/NSF activity. Altogether, these data show that Ari-1 regulates neurotransmitter release by controlling Comt/NSF activity through ubiquitination (Ramirez, 2021).

This study identified sixteen novel putative substrates of Ari-1 in Drosophila photoreceptor neurons in vivo by means of an unbiased proteomic approach. Remarkably, despite Ari-1 being recently shown to regulate the positioning of the cell nucleus in muscles via a direct interaction with Parkin (Tan, 2018), as well as to interact with some Parkin substrates, there is no overlap between the substrates identified for Ari-1 and those previously identified for Parkin in flies. Taken together, the available data suggest that the 16 targets identified in this study are specifically regulated by Ari-1 in Drosophila photoreceptor neurons and that this E3 ligase has a wide functional repertoire (Ramirez, 2021).

This study focused on Comt, an ATPase required for the maintenance of the neurotransmitter release. Ubiquitination of proteins involved in vesicle trafficking and neurotransmitter release had been previously reported. Similarly, the importance of the ubiquitination machinery for the proper neuronal function has also been demonstrated. The alterations produced on synaptic transmission by ubiquitination are typically attributed to an acute control of synaptic protein turnover. However, many of these presynaptic proteins have been reported to be mainly mono- or di-ubiquitinated, a type of ubiquitin modification that is not usually associated with protein degradation. In line with this, the results showed that Comt/NSF is preferentially monoubiquitinated by Ari-1/ARIH1, suggesting that Ari-1/ARIH1 could be regulating Comt/NSF activity, rather than its life span or expression levels (Ramirez, 2021).

Ari-1 mutations result in abnormal synaptic function at the larval stage, a result consistent with a regulatory function of NSF. All mutant alleles examined exhibit a reduced frequency of spontaneous synaptic release. In addition, ari-1² mutants exhibit a large calcium-dependent evoked release. Analysis of the mechanism for enhanced evoked release in ari-1² suggests that the primary defect consists in an increased probability of vesicle fusion in response to calcium entry in the presynaptic side. First, by comparing the amplitude and time course of spontaneously occurring postsynaptic events in mutant and control animals, the possibility of a postsynaptic modification was ruled out. Since no significant differences were found, it is concluded that the receptor field size and kinetic properties of postsynaptic receptors are normal in the mutant (Ramirez, 2021).

The ari-1 functional defects could result from alterations of synaptic transmission during development; therefore, the number of synaptic contacts was quantified, assuming that most release sites occur within varicosities. No significant difference was observed between mutant and control. Although this study does not include electron microscopy quantification of synaptic vesicles, th confocal microscopy and electrophysiology data point toward an upregulation of release from single synapses (Ramirez, 2021).

The failure analysis from single varicosities represents direct evidence that, at relatively low calcium concentrations, mutant terminals release more quanta than controls in response to an action potential. Whether increased quantal release could be explained on the basis of more release sites being concentrated on mutant terminals was examined. The focal recordings using saturating calcium concentrations argue against this possibility. When mutant terminals are exposed to high calcium, in order to increase the likelihood that all active zones within the bouton will release a quantum, EJP amplitudes in the mutant are indistinguishable from that of controls. These data suggest that the number of release sites in mutant and control terminals is similar and favor the hypothesis that, at physiological calcium concentrations, the probability of vesicle fusion upon calcium entry is increased in the mutant (Ramirez, 2021).

It was found that ari-1 mutants have opposite effects on spontaneous and evoked release. Classically, the two modes of vesicular release have been considered to represent a single exocytotic process that functions at different rates depending on the Ca2+ concentration. However, recent work challenges this idea and supports the alternative model where spontaneous and evoked response might come from different vesicles pools. Several experimental evidences indicate that both forms of release may represent separate fusion pathways. Employing a state-of-the-art optical imaging in larval NMJ, it has been shown that evoked and spontaneous release can be segregated across active zones. Thus, three types of active zones could be defined: those that only release vesicles in response to a rise of intracellular calcium (evoked release), a second population that only participates in spontaneous release, and a third small proportion (around 4%) that participates in both evoked and spontaneous release. This result advocates for a different molecular and spatial segregation of both modes of release (Ramirez, 2021).

Differential content or activity of regulatory SNARE binding proteins could discriminate between spontaneous and evoked release. It has been shown that the presence of the Vamp-7 isoform could participate in this differential release. Vamp-7 preferentially labels vesicles unresponsive to stimulation, and it colocalizes only partially with the endogenous synaptic vesicle glycoprotein Sv2 and the vesicular glutamate transporter Vglut1, suggesting that this vesicle pool does not support evoked transmitter release. Recently, it was shown that the double knockout mouse for Synaptobrevin genes, syb1 and syb2, results in a total block of evoked release, while spontaneous release was increased in both frequency and quantal size without changes in the number of docked vesicles at the active zone, confirming the idea that evoked and spontaneous releases are differentially regulated. Interestingly, Vamp-7 was found by MS to be less ubiquitinated when Ari-1 is overexpressed, suggesting that Ari-1 mutants could be favoring evoked release through NSF and reducing spontaneous release through Vamp7. Thus, Ari-1 could be acting as a repressor and activator of evoked and spontaneous release, respectively. All together, these results evidence a new layer of complexity over the actual fine-tuning of synaptic transmission. A physiological regulatory mechanism for both types of release has been recently demonstrated for inhibitory synapses at the trapezoid body, an important brain area in auditory integration. In this nucleus, activation of metabotropic glutamate receptor mGluR1 differentially modulates both spontaneous and evoked release in both GABAergic and Glycynergic synapses (Ramirez, 2021).

Early functional studies of NSF employing the fly thermo-sensitive mutant allele comt^ts17, have reported a reversible reduction of synaptic transmission. Consistent with a role of NSF on SNARE dissociation, this inhibition parallels an increase in the number of synaptic vesicles at the presynaptic terminal. At this point, it can only be speculated how specifically Ari-1/ARIH1 regulates Comt/NSF activity within the presynaptic terminal. Opposite to the role of NSF mutant comt^ts17, which impairs SNARE complex disassembly, a change that enhances NSF functionality due to the lack of its ubiquitination would favor the dissociation of the so-called trans-SNARE complex. Further, this would build up the number of SNARE complexes assembled per vesicle, thus increasing the efficiency of fusion machinery in a Ca2+-dependent manner. Consistent with this interpretation, it has been shown that fast release of a synaptic vesicle requires at least three SNARE complexes, whereas slower release may occur with fewer complexes (Ramirez, 2021).

Interestingly, some of the additional putative substrates identified are also related to synapse physiology and neurotransmitter release. PPO1 is an enzyme with L-DOPA monooxygenase activity, hence, may be involved in the metabolism of dopamine neurotransmitter. Similarly, GstO3 is involved in glutathione metabolism, another type of neurotransmitter. Vha44 and Vha68-1 are components of the vacuolar proton-pump ATPase, whose mutations have been reported to impair neurotransmitter release. Vha44 has also been described as an enhancer of Tau-induced neurotoxicity, and CG15117, orthologue of human GUSB, has been associated with neuropathological abnormalities. The long recovery time from paralysis observed in comt⁶ ari2/comt6 females could result from the role of Ari-1 in the ubiquitination of these additional substrates, in addition to the role of Comt in tissues other than the nervous system. The data reported in this study may be relevant in the context of Parkinson's disease. It should be noted that most Parkinson's-related genes encode proteins involved in vesicle recycling and neurotransmitter release at the synapse. Thus, the kinase LRRK2 phosphorylates NSF to enhance its ATPase activity upon the SNARE complex and facilitate its disassembly. Pathological mutations in this protein, such as G2019S, cause an excess of kinase activity that interferes with vesicle recycling. Deregulated synaptic aggregates of α-Synuclein may target VAMP-2 hampering the formation of the SNARE complex. Parkin is a structural relative of Ari-1, based on their common Cysteine rich C₃HC₄ motif, which is also at the origin of some forms of Parkinson's disease. All these genes and their corresponding mechanisms of activity sustain the scenario in which several types of Parkinson's disease seem to result from a defective activity of the synapse. In this context, the role of Ari-1/ARIH1 emerges as a mechanism to regulate a key component of the SNARE complex, Comt/NSF. Conceivably, Ari-1/ARIH1 may become a suitable target for either diagnosis or pharmacological treatment of Parkinson's and related diseases (Ramirez, 2021).

Ari-1 regulates myonuclear organization together with Parkin and is associated with aortic aneurysms

Nuclei are actively positioned and anchored to the cytoskeleton via the LINC (Linker of Nucleoskeleton and Cytoskeleton) complex. This study identified mutations in the Parkin-like E3 ubiquitin ligase Ariadne-1 (Ari-1) that affect the localization and distribution of LINC complex members in Drosophila. ari-1 mutants exhibit nuclear clustering and morphology defects in larval muscles. Ari-1 mono-ubiquitinates the core LINC complex member Koi. Surprisingly, functional redundancy between Parkin and Ari-1 was discovered: increasing Parkin expression rescues ari-1 mutant phenotypes and vice versa. It was further shown that rare variants in the human homolog of ari-1 (ARIH1) are associated with thoracic aortic aneurysms and dissections, conditions resulting from smooth muscle cell (SMC) dysfunction. Human ARIH1 rescues fly ari-1 mutant phenotypes, whereas human variants found in patients fail to do so. In addition, smooth muscle cells (SMCs) obtained from patients display aberrant nuclear morphology. Hence, ARIH1 is critical in anchoring myonuclei to the cytoskeleton (Tan, 2018).

The localization of a nucleus in a specific subcellular location is a highly regulated and dynamic process. Depending on the cell type, nuclear localization can affect cell division and differentiation, or cellular function. In humans, loss of proper nuclear positioning has been implicated in muscular dystrophy, lissencephaly, cerebellar ataxia, and deafness. The LINC complex is required for proper nuclear localization and consists of various proteins (Lamins, SUN [Sad1 and UNc84 homology proteins] and KASH [Klarsicht, ANC-1, and Syne Homology] proteins) that cross the nuclear envelope and project into the cytoplasm to connect the nucleoskeleton to the cytoskeletal network. Lamins, the main component of the nuclear lamina, interact with SUN proteins that cross the inner nuclear membrane into the perinuclear space. Within the perinuclear space, the SUN domain interacts with the C terminus of KASH proteins that in turn span the outer nuclear membrane and interact with various members of the cytoskeleton. Apart from placing a nucleus in its appropriate position in the cell, the LINC complex also endows a cell with the ability to respond to changes in its micro-environment, such as changes in cell shape or migration. This process was recently dubbed nuclear mechanosensing. External stimuli that alter cell shape introduce mechanical tension on the plasma membrane, which is transferred through the cytoskeleton and the LINC complex to alter nuclear morphology and/or transcription. This suggests that the LINC complex is required for mechanosensing, an important feature for cells, in particular muscle cells that have to respond to constant changes in mechanical forces that are applied to them (Tan, 2018).

Some of the LINC complex components have been shown to be the subject of few posttranslational modifications. However, a role for ubiquitination in the regulation of the LINC complex has not been documented so far. The human genome encodes ∼700 E3 ligases, 14 of which are RING-between-RING (RBR) ligases. The best-known member of the RBR family in Drosophila is encoded by the parkin locus, because mutations in human PARK2 (parkin) cause a recessive form of early-onset Parkinson's disease. Parkin plays a role in clearing dysfunctional mitochondria based on fly mutants and mammalian cell culture experiments, but numerous mouse knockouts of Park2 have revealed little about the protein's in vivo function in mice. Much less is known about the other members of the RBR family of E3 ligases. The human homolog of Ari-1 (ARIH1) was recently shown to function together with Cullins of the RING family of E3 ligases to regulate several cell cycle regulators (Scott, 2016). In this setting, ARIH1 transfers the first ubiquitin molecule to the Cullin substrate, upon which Cullin lengthens the poly-ubiquitin chain. In flies, Ari-1 was shown to be an essential protein that can dimerize and can regulate the insect hormonal ecdysone signaling pathway (Aguilera, 2000, Gradilla, 2011). Interestingly, although Parkin and Ari-1 are thought to have unique biological functions, the two proteins are similar at the structural level and Ari-1 interacts with a subset of established Parkin substrates (Parelkar, 2012). In addition, both proteins are activated by the same E2 conjugating enzyme, UbcD10 (fly)/UBCH7 (human) (Tan, 2018).

This study identified lethal mutations in ari-1 from an X chromosome mutagenesis screen in Drosophila. Phenotypic characterization of the mutants revealed a role for Ari-1 in positioning nuclei in muscles by regulating the LINC complex. This study shows that Ari-1 mono-ubiquitinates Koi. In the absence of Ari-1, Parkin can partially compensate for the loss of Ari-1 and vice versa. In contrast, ari-1 missense mutations act in a dominant negative fashion and the mutant forms of Ari-1 affect Parkin, inhibiting its function in nuclear positioning. Furthermore, three variants of ARIH1 were identified in patients with aortic or cerebrovascular aneurysms. Aortic tissues from these patients reveal that vascular smooth muscle cell (vSMC) nuclear morphology is severely affected. Further testing of the identified ARIH1 variants in Drosophila show that the mutant forms of ARIH1 are less functional or non-functional. In sum, this study has identified an evolutionarily conserved role of Ari-1/ARIH1 with Parkin in regulating myonucleo-cytoskeletal anchorage and shape, and variants in ARIH1 are associated with developing aortic aneurysms in humans(Tan, 2018).

This study reveals that the RBR E3 ubiquitin ligase Ari-1 regulates the subcellular localization and morphology of muscle nuclei in flies. Loss of ari-1 affects the protein levels and/or distribution of members of the LINC complex. Interestingly, Ari-1 and Parkin, a second RBR E3 ligase family member, exhibit redundant functions. Indeed, overexpressing parkin in ari-1 mutants suppresses the observed phenotypes and vice versa. Koi, a subunit of the LINC complex, can be ubiquitinated by both Ari-1 and Parkin. Finally, rare dominant variants were identified in the human homolog of ari-1, ARIH1. These variants are associated with cerebrovascular and aortic aneurysms and this study documented that patient samples have altered nuclear morphology (Tan, 2018).

Loss of ari-1 leads to clustering of nuclei in Drosophila striatal skeletal muscles. The protein distribution of several LINC complex members is affected in ari-1 mutants. The data show that Koi is the target of Ari-1. First, Koi protein levels are upregulated in ari-1 mutants, whereas Koi mRNA levels remain unchanged. Second, in human vSMCs, the protein level of the Koi homolog, SUN2, is elevated upon ARIH1 knockdown. Third, overexpressing Koi phenocopies ari-1 mutants. Fourth, Koi is the only LINC complex member that physically interacts with Ari-1 when tested in cell culture. Fifth, Ari-1 mono-ubiquitinates Koi in in vitro experiments. It is proposed that Ari-1's mono-ubiquitination of Koi primes Koi for subsequent poly-ubiquitination and degradation by another E3 ligase. This other E3 ligase could be Cullin-1, as, in human cells, the Cullin-1 ubiquitination complex regulates the turnover of SUN2. Furthermore, a cooperative ubiquitin-tagging activity has been shown for vertebrate ARIH1 and Cullin-1 (Tan, 2018).

How do elevated levels of Koi impair the LINC complex? One possibility is that excessive amounts of Koi proteins disrupt the stoichiometry of the functional LINC complex, resulting in imbalanced mechanical tensions applied across the nuclear envelope. Also, excessive Koi at the nuclear envelope may redistribute Koi to other compartments continuous to the nuclear envelope, including the endoplasmic reticulum (ER). In fact, overexpressed SUN1 was found to enter the ER-Golgi secretory pathway, and retention of SUN in the Golgi is associated with the pathology of laminopathies, supporting observations that elevated Koi/SUN levels can lead to cellular and physiological defects (Tan, 2018).

The observation that a missense mutation induces a stronger phenotype than its corresponding null mutation was recently dubbed 'recessive antimorph'. This genetic phenomenon reveals the presence of a second, functionally redundant protein that is inhibited by a missense mutant but can execute its function in the absence of the protein. The current data suggest that the missense mutations function as recessive antimorphs because (1) the mutations cause recessive lethality with a dominant muscle nuclear phenotype; (2) mutant phenotypes are rescued by introducing wild-type copies of the gene; (3) crossing ari-1 missense mutant alleles to a deficiency diminishes the nuclear positioning phenotype compared with homozygous/hemizygous missense mutants; and (4) the missense mutant alleles can be rescued in a dose-dependent fashion. These findings strongly support the view that the three identified missense mutations behave as dominant negative alleles (Tan, 2018).

Ari-1 and Parkin are members of the same E3 ligase subfamily and are evolutionarily related. An evolutionary tree of the RBR domains of all Ariadne and Parkin RBR proteins reveals that ariadne genes are the ancestral genes as they, in contrast to parkin, are present in all unicellular eukaryotic organisms tested. Parkin has an established role in mitochondrial quality control. However, loss of PARK2 homologs have been associated with significant phenotypic discrepancies across species and even among cell types in the same organism. These differences have challenged the ability to unravel the in vivo role of Parkin and have questioned the cellular requirements of Parkin. The data argue that there is some functional redundancy between Ari-1 and Parkin. First, the double homozygous mutants are first instar lethal, whereas the single null mutants are viable (parkin) or late pupal lethal with escapers (ari-1). Second, double heterozygous mutants display a much stronger myonuclear clustering phenotype than either heterozygous mutant. Third, overexpressing Parkin partially rescues ari-1 mutant phenotypes. Fourth, overexpressing ARIH1 in parkin mutants ameliorates phenotypes associated with the loss of parkin. Moreover, Ari-1 and Parkin can form homodimers or homomultimers. In addition, Ari-1 can be found in a complex with Parkin. These data are therefore most consistent with the following model: it is proposed that there are three complexes (Ari-1/Ari-1, Parkin/Parkin, Ari-1/Parkin), each of which can perform some functions related to nuclear and mitochondrial phenotypes, but they differ in efficiency. Under normal conditions, Ari-1 mainly regulates nuclear positioning and morphology but is aided by Parkin. In the absence of Ari-1, Parkin can still regulate nuclear positioning but with lower efficiency leading to a mild nuclear clustering phenotype. Although the requirement of Ari-1 is clearly more prominent for nuclear phenotypes than the requirement of Parkin, the reverse is true for mitochondrial quality control. In the latter case, Parkin plays the major role and no mitochondrial morphology phenotype is observed in Ari-1 mutants, but Ari-1 can clearly suppress some of the mitochondrial phenotypes associated with loss of parkin. Finally, in an ari-1 missense mutant background, mutant Ari-1 sequesters Parkin in a non-functional complex, causing a stronger nuclear positioning phenotype (Tan, 2018).

Is this functional redundancy observed among all RBR family members? Although the data show that Ari-1, Parkin, and Ari-2 can form pairwise complexes such as Ari-1/Ari-2, Parkin/Ari-1, and Parkin/Ari-2, overexpressing Ari-2 does not rescue nuclear clustering phenotypes in ari-1 mutants, suggesting no redundancy exists between these two Ari members for this function. It is, however, possible that, in different cellular contexts, one or more RBR protein(s) may be able to compensate for the loss of another RBR protein, and different species may have evolved to express these proteins at different levels in various tissues. Hence, the observed phenotypic discrepancies between parkin mutants in different species and cells may be attributable to redundancy/dependency issues between the Ari-1 and Parkin isoforms (Tan, 2018).

This study has identified three rare variants in ARIH1 in patients with aortic and cerebrovascular aneurysms. All three variants likely represent loss-of-function alleles because they, in contrast with wild-type ARIH1, fail to rescue nuclear clustering, eclosion, and longevity of ari-1 mutant flies. Similarly, the nuclei of patients' vSMCs display aberrant nuclear morphology, and reducing ARIH1 expression in vSMCs induces nuclear morphological changes, as observed in the patients (Tan, 2018).

How could loss of ARIH1 lead to aneurysms? To date, the genes associated with aortic aneurysms can be divided into several groups based on the cellular process that is affected, one of which is cellular mechanotransduction. Mechanotransduction is important for muscles to adapt and respond to dynamic mechanical forces in their environment. In vascular smooth muscle cells (vSMCs), proper contraction and mechanosensing depends on functional contractile units and their connections to the extracellular matrix through focal adhesions. Genes that encode major structural components of the contractile unit, ACTA2 (encodes SMC-specific α-actin), MYH11 (SMC-specific myosin heavy chain), MYLK (myosin light-chain kinase), PRKG1 (type I cyclic glutathione S-transferase [GMP]-dependent protein kinase), and the extracellular matrix component FBN1 (encodes fibrillin-1) have been shown to be mutated in aortic aneurysm patients. It has been proposed that, in these cases, mechanotransduction between vSMCs and the extracellular matrix is impaired, affecting mechanosensing of the vSMCs and thereby altering molecular pathways leading to aortic disease. Given that contractile units in vSMCs are connected to both the focal adhesions and the nuclear membrane, and that proper mechanosensing is important in preventing aneurysms, the observation that ARIH1 disease-associated variants disrupt the nuclear envelope's structure in vSMCs implicates a defect in mechanosensing that may be responsible for the aortic and cerebrovascular diseases (Tan, 2018).

Interestingly, a growing body of evidence implicates the LINC complex in mechanosensing. Integrin adhesion complexes relay extracellular signals via the cytoskeleton and the LINC complex onto the nuclear envelope to permit mechano-regulation of gene expression. Mutations in the LINC complex members have been associated with striatal and cardiac muscle diseases as well as vascular diseases. For example, mutations in Lamin (LMNA), and SUN genes (SUN1 and 2) are associated with muscular dystrophies. Interestingly, missense mutations in SYNE1 (the human Msp-300 homolog) are associated with dominant muscular dystrophies, while mutations that encode truncated SYNE1 cause recessive cerebellar ataxias with no overt muscular phenotype. However, all of the above mutations affect nuclear positioning or nuclear envelope integrity (Tan, 2018).

Given that Ari-1 regulates the LINC complex in flies, it is highly likely that it assumes a similar role in human vSMCs. Loss of ARIH1 in aortic vSMCs would thereby affect mechanosensing, weaken the aortic wall muscles to induce aneurysms and dissections. It is intriguing to note that patients with ARIH1 mutations have no reported skeletal muscle abnormalities, even though mutations in members of the LINC complex are associated with muscular dystrophies. The current data suggest that functional redundancies of RBR proteins in skeletal muscles may underlie the lack of phenotype. Interestingly, patients with Parkinson's disease are about two times more likely to develop heart failure, suggesting that parkin may be a risk factor for heart disease (Tan, 2018).

In summary, Ari-1 regulates nuclear morphology in Drosophila skeletal muscles and human vSMCs. Loss of ARIH1 is associated with aortic and cerebrovascular aneurysms, likely due to the disruption of mechanosensing in vSMCs. This study also found that Ari-1 and Parkin are to some extent functionally redundant as either gene can partially suppress some of the phenotypes associated with the loss of the other gene. Hence, Ari-1 and Parkin regulate nuclear positioning and mitochondrial quality control respectively but have each retained the ability to perform both functions to some extent (Tan, 2018).

Isoform-specific regulation of a steroid hormone nuclear receptor by an E3 ubiquitin ligase in Drosophila melanogaster

The steroid hormone 20-hydroxyecdysone (20E) regulates gene transcription through the heterodimeric nuclear receptor composed of ecdysone receptor (EcR) and Ultraspiracle (USP). The EcR gene encodes three protein isoforms -- A, B1, and B2 -- with variant N-terminal domains that mediate tissue and developmental stage-specific responses to 20E. Ariadne-1a is a conserved member of the RING finger family of ubiquitin ligases first identified in Drosophila melanogaster. Loss-of-function mutations at key cysteines in either of the two RING finger motifs, as well as general overexpression of this enzyme, cause lethality in pupae, which suggests a requirement in metamorphosis. This study shows that Ariadne-1a binds specifically the isoform A of EcR and ubiquitylates it. Co-immunoprecipitation experiments indicate that the full sequence of EcRA is required for this binding. Protein levels of EcRA and USP change in opposite directions when those of ARI-1a are genetically altered. This is an isoform-specific, E3-dependent regulatory mechanism for a steroid nuclear receptor. Further, qRT-PCR experiments show that the ARI-1a levels lead to the transcriptional regulation of Eip78C, Eip74EF, Eip75B, and Br-C, as well as that of EcR and usp genes. Thus, the activity of this enzyme results in the regulation of dimerizing receptors at the protein and gene transcription levels. This fine-tuned orchestration by a conserved ubiquitin ligase is required during insect metamorphosis and, likely, in other steroid hormone-controlled processes across species (Gradilla, 2011).

Ariadne-1: a vital Drosophila gene is required in development and defines a new conserved family of ring-finger proteins

The identification and functional characterization of ariadne-1 (ari-1), a novel and vital Drosophila gene required for the correct differentiation of most cell types in the adult organism is reported. A sequence-related gene, ari-2, and the corresponding mouse and human homologs of both genes, are described. All these sequences define a new protein family by the Acid-rich, RING finger, B-box, RING finger, coiled-coil (ARBRCC) motif string. In Drosophila, ari-1 is expressed throughout development in all tissues. The mutant phenotypes are most noticeable in cells that undergo a large and rapid membrane deposition, such as rewiring neurons during metamorphosis, large tubular muscles during adult myogenesis, and photoreceptors. Occasional survivors of null alleles exhibit reduced life span, motor impairments, and short and thin bristles. Single substitutions at key cysteines in each RING finger cause lethality with no survivors and a drastic reduction of rough endoplasmic reticulum that can be observed in the photoreceptors of mosaic eyes. In yeast two-hybrid assays, the protein ARI-1 interacts with a novel ubiquitin-conjugating enzyme, UbcD10, whose sequence is also reported in this study. The N-terminal RING-finger motif is necessary and sufficient to mediate this interaction. Mouse and fly homologs of both ARI proteins and the Ubc can substitute for each other in the yeast two-hybrid assay, indicating that ARI represents a conserved novel mechanism in development. In addition to ARI homologs, the RBR signature is also found in the Parkinson-disease-related protein Parkin, adjacent to an ubiquitin-like domain, suggesting that the study of this mechanism could be relevant for human pathology (Aguilera, 2000).

This study reports on ARIADNE-1, the first member of a conserved new family of proteins whose multiple structural domains include a RING finger that interacts with an E2 enzyme of the ubiquitin system. In Drosophila, ari-1 is expressed in all tissues throughout development although the mutant phenotypes indicate differential requirements in tissues and developmental stages (Aguilera, 2000).

A new protein family: The primary sequence of ARI-1 serves to define a novel protein family characterized by the string of motifs ARBRCC. Drosophila ARI-1 and ARI-2 are the first two members of this family and conserved homologs in yeast, worms, mice, and humans have been identified as well. On the basis of structural criteria, ARI proteins are related to the RBCC family that is composed mostly by signal transducers and transcription factors. In these proteins, the RBCC signature appears in conjuction with other motifs like bromodomain, RFP, and PHD fingers. Within the string of motifs that the ARI family defines, RBR is proposed as a new signature. As in the case of RBCC proteins, RBR is also found associated to other domains, i.e., an ubiquitin-like domain in Parkin or an ill-defined N-terminal domain in ARA54. In addition, the RBR signature is found in many other functionally uncharacterized proteins across eukaryotic species. Most of these sequences are derived from genome projects and their number is increasing rapidly, suggesting that RBR is an ancient and widespread modular element of protein structure (Aguilera, 2000).

Functional role: ARI-1 becomes an additional case lending credence to the role that the ubiquitin system seems to play in neural connectivity. The first precedent was illustrated by the bendless mutant. This gene encodes a ubiquitin-conjugating enzyme whose depletion leads to neural defects. The UbcD10 identified as interacting with ARI belongs to the same subfamily as UbcD3 encoded in ben. Mutations in ben have the most extreme effects in neural connectivity, particularly in the optic ganglia. Defective medulla rotation, the abnormal course of axon bundles, and reduced rhabdomeres, mostly R7, are reported as consistent features of ben eyes along with the original phenotype of incomplete bending of the cervical giant fiber axon. Adult muscles are also affected, albeit the tubular TDT to a greater extent than the indirect flight muscles. Finally, ben null mutants are pupal lethals with a few short-lived and motor-impaired adult escapers. Mutant ari-1 traits reported in this study coincide with those of ben; however, ari-1 phenotypes are more extreme in every tissue analyzed. Preliminary tests (double heterozygotes and gene dosage) were performed, searching for genetic evidence of interactions between ari-1 and the nine Ubc encoding genes including ben, but no positive indication has been found so far. Thus, it seems that UbcD3 on the one hand and UbcD10 and ARI-1 on the other are components of two mechanisms within the ubiquitin system (Aguilera, 2000).

Considering the general expression of ari-1, it is likely that all cell types require its activity. However, at the light microscope level and under the procedures used, the mutant phenotypes become most evident in adult tissues. The failure to detect larval phenotypes cannot be justified, at least solely, on the basis of a large maternal deposit of the normal product. The germline mosaics of null mutations yield regular escapers identical in frequency and phenotype to those produced from heterozygous ovaries. It is more probable that the phenotypes in the larval tissues are subtler than in the adult, requiring a level of observation beyond that used in this study. It is worth noting that the cell types in which the mutant phenotypes are most apparent are those that require a massive and rapid membrane deposition, namely, macrobristles, photoreceptors, large tubular muscles, and rewiring neurons at metamorphosis. These processes represent a high demand on the physiology of the cell. In particular, proteins and membranous cisternae must be supplied efficiently (Aguilera, 2000).

A substantial reduction of rough endoplasmic reticulum was detected in mutant photoreceptors. This trait is more difficult to evaluate in other cell types because their ultrastructure is not as well characterized as that of the photoreceptors. However, this membrane system plays such a central role in the biology of the cell that it is plausible that it could account for the variety of phenotypes observed. The endoplasmic reticulum is the port of entry of most compartmentalized, membrane-anchored, and secretory proteins. It is also the site of folding and modification of nascent protein chains and assembly of multisubunit complexes. The ubiquitin pathway plays a central role in the 'quality control' machinery to remove those proteins that fail to fold properly or to oligomerize, as has been demonstrated for the cystic fibrosis transmembrane conductance regulator and the T-cell receptor. Interestingly, a RING-finger protein, Der3p/Hrd1p, is an integral component of a retrograde transport system in yeast endoplasmic reticulum. The system interacts with the 26S proteasome and includes a Ubc enzyme, Ubc7p. Also, mutations in the Drosophila ninaA cyclophilin homologue, involved in the protein folding and secretory pathway of rhodopsin 1, lead to an accumulation of endoplasmic reticulum in photoreceptors. The reduction of endoplasmic reticulum in ari-1 mutant photoreceptors could indicate a role of ARI proteins in ubiquitin-mediated mechanisms. In turn, a reduced endoplasmic reticulum is a likely origin for the small rhabdomeres and short bristles that all alleles show (Aguilera, 2000).

Regarding the cellular compartment in which ARI-1 performs its function, the protein appears to localize mainly in the cytoplasm according to Western blots of cytoplasm vs. nuclear enriched fractions. Nevertheless, a transient nuclear localization cannot be formally excluded since a putative nuclear localization signal, KKWIKK, can be found between amino acids 268 and 273. In conyrast, the N-terminal acid cluster shows the requirements for transcription activation domains and, in a yeast reporter assay, it was found that this domain is necessary for the autonomous transcription activity of the ARI-1-GAL4BD fusion. Thus, considering the multiple domains present in ARI-1, it is plausible that the protein might perform distinct interactions in several cellular compartments (Aguilera, 2000).

Interaction mechanisms: Cysteines 150 and 309 are key residues for the activity of each RING finger of ARI-1 since alleles ari-1² and ari-1³ are full lethals. In addition, the C150Y version of the protein fails to sustain one of the identified interactions of ARI-1, that with UbcD10. These in vivo and in vitro observations coincide with findings in the breast/ovarian cancer susceptibility gene product BRCA-1 where the RING motif of this human protein also mediates an interaction with a ubiquitin enzyme, the ubiquitin hydrolase BAP1. The implication of a RING-finger protein in the ubiquitin pathway raises the possibility that ARI-1 might be a substrate for UbcD10-mediated ubiquitination and targeted for degradation. ARI-1 Western blots show a single 59-kD band throughout development without any smear or ladder-like signals that could result from ubiquitin-mediated degradation. Although ARI-1 ubiquitination cannot be excluded at this point, a more likely role for ARI proteins might be the modulation of Ubc activity upon other substrates. Recent studies have shown common structural features among ubiquitination complexes such as SCF (Skp1-Cdc53/CUL1-F-box), APC (anaphase promoting complex), or VCB (Von Hippel-Landau-ElonginC-ElonginB). Apart from proteins with E2 and E3 activities, SCF, APC, and VCB also contain adaptor proteins with a variety of protein-protein interaction modules and a conserved domain (F-box or SOCS-box). None of these domains are present in ARI proteins; thus, ARI might represent a new class of adaptor proteins in ubiquitination complexes. In fact, the recent identification of Rbx1 as a common component of SCF and VCB complexes already has introduced the RING-finger motif as a keystone in the combinatorial control of ubiquitination. Rbx1 is a RING-finger protein that interacts with E2 and E3 enzymes as well as with some adaptor proteins. These interactions are necessary to stimulate the catalytic activity of the fully assembled complex, thereby limiting its substrate specificity. This speculative proposal of ARI-1 as an adaptor in a multimeric complex need not be restricted to a ubiquitin tagging system. For example, the SUMO-1 or Rub-1 tagging systems share components with the ubiquitination complexes. In this context, PML, a RBCC protein involved in acute promyelocytic leukemia, is covalently modified by SUMO-1. This modification requires the interaction between Ubc9, a nuclear E2 enzyme, and the R motif in PML (Duprez et al. 1999) (Aguilera, 2000).

Relevance to human pathologies: The interaction between ARI-1 and UbcD10 detected in the yeast two hybrid assay is a conserved mechanism since mouse and fly homologues substitute for each other. Although a genetic test that could validate this interaction in vivo cannot be carried out at this point, the interspecific functional substitution is a strong argument to support its biological relevance. This functional substitution has a potential clinical interest. The sequence of human UbcH7 is 100% identical to UbcM4. In turn, UbcH7 is known to interact with E6-AP, a ubiquitin ligase that uses the tumor suppressor p53 as a substrate. In addition, a defective E6-AP is implicated in Angelman's syndrome. As a logical derivative from the data in Drosophila, the possible interaction between human ARI-1 and the doublet UbcH7/E6-AP should be considered (Aguilera, 2000).

Several human pathologies are caused by functional deficits in RING-finger proteins. Aside from PML and SIAH-1 discussed above, mutations that delete the RING motif of PEX10 cause defects in peroxisome biogenesis leading to Zellweger's syndrome. However, the most direct link with ARI proteins is found in parkin, a gene associated with juvenile parkinsonism. During the search for sequence homologies, the Parkin protein was identified because of a RING-finger motif. Upon closer inspection, however, an RBR signature was evident. In addition, this motif seems to be directly linked to the pathology as the disease can be caused by a point mutation, T240R, which corresponds to the R1 motif in ARI-1. Since this protein also exhibits a ubiquitin-like domain, it has been proposed that Parkin might cause alterations in the ubiquitin system or defects in its own functional maturation eventually leading to selective neurodegeneration. Although defective ubiquitination is a common feature of cellular inclusions such as the Lewy bodies, a hallmark of Parkinson disease, the juvenile form of parkinsonism does not exhibit Lewy's bodies. Also, the early onset of the disease points toward a developmental etiology. As an alternative mechanism to ubiquitination, Parkin could be involved in protein tagging of the SUMO-1 or Rub-1 type. In any event, two important questions arise from the results reported in this study: (1) Are human ARI proteins involved in the biology of Parkin? and (2) Does Parkin interact with a Ubc enzyme? (Aguilera, 2000).

Search PubMed for articles about Drosophila Ariadne-1

Aguilera, M., Oliveros, M., Martinez-Padron, M., Barbas, J. A. and Ferrus, A. (2000). Ariadne-1: a vital Drosophila gene is required in development and defines a new conserved family of ring-finger proteins. Genetics 155(3): 1231-1244. PubMed ID: 10880484

Gradilla, A. C., Mansilla, A. and Ferrus, A. (2011). Isoform-specific regulation of a steroid hormone nuclear receptor by an E3 ubiquitin ligase in Drosophila melanogaster. Genetics 189(3): 871-883. PubMed ID: 21900267

Parelkar, S. S., Cadena, J. G., Kim, C., Wang, Z., Sugal, R., Bentley, B., Moral, L., Ardley, H. C. and Schwartz, L. M. (2012). The parkin-like human homolog of Drosophila ariadne-1 (HHARI) can induce aggresome formation in mammalian cells and is immunologically detectable in Lewy bodies. J Mol Neurosci 46(1): 109-121. PubMed ID: 21590270

Ramirez, J., Morales, M., Osinalde, N., Martinez-Padron, I., Mayor, U. and Ferrus, A. (2021). The ubiquitin ligase Ariadne-1 regulates neurotransmitter release via ubiquitination of NSF. J Biol Chem: 100408. PubMed ID: 33581113

Scott, D. C., Rhee, D. Y., Duda, D. M., Kelsall, I. R., Olszewski, J. L., Paulo, J. A., de Jong, A., Ovaa, H., Alpi, A. F., Harper, J. W. and Schulman, B. A. (2016). Two distinct types of E3 ligases sork in unison to regulate substrate ubiquitylation. Cell 166(5): 1198-1214 e1124. PubMed ID: 27565346

Tan, K. L., Haelterman, N. A., Kwartler, C. S., Regalado, E. S., Lee, P. T., Nagarkar-Jaiswal, S., Guo, D. C., Duraine, L., Wangler, M. F., University of Washington Center for Mendelian Genomics, Bamshad, M. J., Nickerson, D. A., Lin, G., Milewicz, D. M. and Bellen, H. J. (2018). Ari-1 regulates myonuclear organization together with Parkin and is associated with aortic aneurysms. Dev Cell 45(2): 226-244 e228. PubMed ID: 29689197

date revised: 4 October 2021

Home page: The Interactive Fly © 2011 Thomas Brody, Ph.D.