Amyotrophic lateral sclerosis (ALS) is a multifactorial, multisystem disease caused by progressive loss of both upper and lower motor neurons [1
]. The pathological signature of ALS is the presence of TAR DNA-binding protein 43 (TDP-43) inclusion bodies, the protein from the gene TARDBP
], in the affected regions of brain and spinal cord in over 95% of affected patients with its consequent clearance from the nucleus [3
]. The discovery of causative TARDBP
mutations in 2008 [5
] brought this gene to the forefront of neurodegeneration research, not only as a major pathological marker but also as a causative factor in sporadic and familial ALS. TDP-43 is an heterogeneous nuclear ribonucleoprotein (hnRNP) [7
] able to bind both nucleic acids and proteins, involved in a wide range of RNA processes, including transcription, transport, stability, and splicing [9
]. Together with additional RNA-binding proteins linked to ALS (reviewed in [10
]), TDP-43 gives RNA metabolism a central role in ALS pathogeny. Therefore, among the hundreds of TDP-43′s targets, the determination of a relevant factor for ALS development is crucial.
Increasing evidence shows early morphological and biochemical changes at the neuromuscular junction (NMJ) before complete muscle fiber denervation and functional symptom onset in ALS patients [11
]. In particular, in mutant SOD1G93A
mouse, which replicates several key features of ALS, reduced NMJ complexity at presymptomatic stages has been described, corresponding to an augmented number of disassembled acetylcholine (ACh) receptor clusters on muscle from the mutant SOD1 transgenic mice [13
Structurally, the NMJ is a highly complex cholinergic synapse where the majority of the molecules responsible for its formation are also crucial for its stability and function at the adult stage [14
In this context, the enzyme acetylcholinesterase (AChE) is a master regulator of the signal transduction. In mammals, AChE exists as three distinct variants coming from alternative exon splicing (AChE T, R, and H) [15
]. The main specie, the T (tail), is recruited by specific anchor proteins and targeted to the plasma membrane (if produced by neurons [18
]) or to NMJ basal lamina (if produced by muscles [19
]), where it terminates neurotransmission by hydrolyzing the neurotransmitter ACh [15
]. AChE activity regulates synaptic ACh levels that, in turn, guarantee the functionality and the integrity of the synapse itself [15
In addition, the enzyme has several non-classical roles in neuronal regeneration and development, as demonstrated in dopaminergic neurons [24
]; in rat cultured dorsal root ganglion neurons [26
]; and in amphibian, avian, and mammalian glia and neurons [27
]. NMJ impairments have been demonstrated in a mouse knockout (KO) for AChE [21
], as well as in a mouse model specifically lacking of AChE anchored at the NMJ [29
]. Moreover, transgenic mice overexpressing human AChE display pathological changes at the NMJs [30
]. Altogether, AChE level and activity play an important role in NMJ viability.
In mammals, our understanding of AChE cellular mechanisms is complicated due to the presence of butyrylcholinesterase (BChE), another enzyme able to hydrolyze ACh, although less effectively [32
], and through alternative splicing, which is sensible to AChE inhibitors [33
]. Zebrafish (zf; Danio rerio
) permit the overcoming of these issues since this vertebrate does not appear to express a functional bche and has not been reported to undergo alternative splicing at the respective 5′ or 3′ ends [34
]. Thus, it only presents a T variant, providing the opportunity to specifically investigate the ache function at the plasma membrane and during NMJ development and maintenance during vertebrate embryogenesis in vivo. As previously reported, the inactive form of ache induces progressive motility defects and severe reduction of ach receptor clusters in two independent transgenic zf lines [35
], consistent with a role for ache in synapse stability.
Several studies on pathological tissues [37
] and transgenic mouse models [40
] have proposed a potential role for AChE in ALS. Very recently, this role has been supported by using a spatial transcriptomic analysis approach that showed a reduced expression of AChE transcript in spinal cord regions primarily related to ALS symptoms onset [41
Data summary: In this study, we describe for the first time a functional link between tdp-43 and ache in zebrafish. For this purpose, we employed the zf model previously described by our team [42
], where knockdown of tdp-43 leads to symptoms reminiscent of ALS disease, such as disruption of axonal projections from spinal motor neurons and reduced locomotion. In this model, ache expression is reduced then restored with human TDP-43 overexpression. Importantly, the overexpression of the human AChE-T form partially rescues the motor and NMJ deficits seen in tdp-43 loss of function. Altogether, these results confirm an important role that ACHE through the functional interaction TDP-43 could play in ALS pathogenesis and open new avenues into our understanding of this neuromuscular disorder.
In this study, we extended our understanding of a major pathological factor in ALS, TDP-43, and its potential contribution in this neuromuscular disorder. Here, we established that tdp-43 plays an important action in regulating ache expression and levels in zebrafish. We demonstrated that ache levels that remain membrane-bound are decreased upon tdp-43 knockdown. Importantly, motor phenotype and the NMJ’s defects due to the tdp-43 depletion were rescued upon restoration of ache levels and activity.
In the overwhelming majority of ALS cases, TDP-43-mediated neurodegeneration is primarily caused by the loss of normal function of TDP-43 in nuclei without being mutated with a consequent accumulation in the cytosol resulting in pathological aggregation [9
]. However, the shift to a pathological state is still unclear, especially in muscle fibers, where TDP-43 aggregates remain to be unraveled [61
]. In this tissue, cytosolic TDP-43 has been shown to assemble in oligomers, where it recruits specific RNA and proteins directly involved in skeletal muscle formation and regeneration [62
]. The increased assembly or decreased clearance of this pool could be the source of TDP-43-aggregates that commonly occur in neuromuscular disease [62
]. In general, there is a growing acceptance of a pivotal role of TDP-43 in muscle. Indeed, TDP-43 depletion leads to locomotive and NMJs anatomical defects in Drosophila
] as well as in zebrafish [42
]. Our findings strongly support this role and put in evidence an NMJ remodeling partially dependent on the TDP-43/AChE pathway.
The relevance of these findings is evident when taking into consideration the role that AChE displays during neuronal development and NMJ synapse stabilization independently of its activity in cholinergic neurotransmission [65
]. Thus, prior to synapse formation in vertebrates, AChE is expressed in both immature postmitotic neuroblasts that are about to extend the first long tracts and myotomal target cells, reviewed in [66
Despite AChE being considered an “old actor” [67
] in scientific research, the role of this factor in axonal growth and maintenance of the nerve-muscle system is not completely understood. Several in vitro studies have shown that ACh influences neurite extension by inhibiting neurite outgrowth [68
]. Thus, a possible role for embryonic AChE would be to neutralize ACh inhibition and create a corridor for neurite growth. Another explanation comes from AChE sequence homology with Neuroligin 1, a membrane cell-adhesion protein of the family of neuroligins [72
], that binds a specific set of neurexins, another family of cell surface proteins [73
], regulating neuronal axogenesis independently of synaptic activity [74
]. Therefore, it has been demonstrated in vitro that AChE can compete with Neuroligin 1 for the extracellular binding to the common ligand Neurexin, thus regulating neuritogenesis [75
] and synaptogenesis [76
]. Therefore, an AChE–neurexin recognition in vivo for inter-neuronal recognition and axon pathfinding should not be excluded.
In the mature NMJ instead, the AChE classical role involves the hydrolysis of ACh [67
], and has been shown to be critical for NMJ stability [20
]. AChE activity inhibition in fact induces ACh accumulation in the synaptic cleft, which in turn causes (1) the repeated activation of the nicotinic acetylcholine receptors (AChRs) and their consequent desensitization [20
]; (2) the decrease of AChRs number at synapses (as compensatory mechanism) [21
]; and (3) the fragmentation of AChRs clusters, since ACh elicits a negative effect on postsynaptic apparatus stabilization [22
]. Indeed, NMJs are fragmented in mouse model lacking a functional AChE [21
Similarly, the zebrafish model lacking ache catalytic activity presents smaller and fragmented ach receptors clusters, reminiscent of the anatomical abnormalities observed in muscle/NMJ biopsies from ALS patients [53
]. Moreover, the achR
α subunits are increased and hdac4
expression augmented. The increase of these factors is particularly interesting since they are clearly upregulated in ALS-denervated muscle [54
] and are considered markers of muscle atrophy. However, we cannot exclude the presence of re-innervating motorneurons, as NMJs in these embryos are smaller and aberrant branching of motorneurons is observed. Even though γ subunit levels are not altered to properly understand the implication of tdp-43 in denervation/reinnervation mechanisms, the levels of this factor should be measured at further time points of larval development.
In our tdp-43 KD model, NMJs clearly lose their integrity, and ache down-expression can be one of the major causes of NMJ dysfunctions. The discovery that tdp-43 LoF decreases ache expression in zebrafish could provide a novel mechanistic insight for nerve-muscle integrity disruption in ALS pathogenesis. Since TDP-43 downregulation appears to be an early defect in ALS, AChE expression could also be altered at early pre-symptomatic stages of the disease, leading to NMJ remodeling. Indeed, to reinforce these findings, a recent study that performed spatial transcriptomic analysis in ALS pathological tissue observed reduced AChE levels [41
]. Since Neuroligin 1 was also shown to be a target of TDP-43 regulation in mouse brain, these factors could participate in parallel to regulate the neurogenesis and axogenesis at the level of NMJ [81
]. Therefore, future studies are warranted to establish whether TDP-43 binds and regulates the transcription of AChE
mRNA or rather indirectly TDP-43 and AChE can exert a role in NMJ maintenance through trans-acting factors [82
Certain limitations arise from the usage of zebrafish models in this current study. Importantly, compared with other vertebrates, zebrafish embryos appear to be devoid of ache splicing [34
]. For this reason, it will be important to eventually confirm and extend these findings using a transgenic mouse model of ALS, as well as on induced pluripotent stem cell (iPSC)-derived neurons and muscle co-cultures obtained from sporadic and familiar ALS patients. Together with these in vivo results, these approaches could also provide powerful tools to test this hypothesis and eventually to validate compounds that are capable of boosting AChE expression at the NMJ as potential therapeutic avenues in ALS and related neuromuscular disorders. Overall, our work gives novel insight into TDP-43 molecular mechanisms and supports the hypothesis of a direct involvement of AChE in ALS NMJ remodeling.