Targeting the Transnitrosylation Cascade Provides a Novel Therapeutic Strategy for Alzheimer’s Disease

As a major public health problem in modern society, Alzheimer’s disease (AD) is currently the most widely prevalent neurodegenerative disease, which will indubitably increase dramatically in the coming years, unless drugs that can prevent or cure the disease become available. Because AD is characterized histopathologically by extracellular deposition of amyloid beta (Aβ) surrounded by dystrophic neuritis forming senile plaques (SPs), and intracellular neurofibrillary tangles (NFTs) of hyperphosphorylated tau, and clinically by chronic and progressive cognitive impairments due to synapse loss, therapeutic approaches for AD treatment mainly focus on the tau pathologies, Aβ toxicity, and synaptic damage. Actually, many drug discovery and development strategies based on tau and Aβ have not been successful, and thus have been declared as failures. More and more therapeutic strategies are now being shifted to disease-modifying therapies (DMTs) which temporarily slow the worsening of dementia symptoms in patients with AD and other dementias [1, 2]. The 2020 Alzheimer’s disease drug development pipeline revealed that synaptic plasticity/neuroprotective agents have increased to up to 23.5% of DMTs in Phase 3 clinical trials, and 27.3% of DMTs in Phase 2 clinical trials [3], implying that correction of altered synaptic functions in patients with AD and other dementias might be a promising strategy in the management of these diseases. Synapses are regarded as the initial and key target for the different molecular assaults that lead to the development and progression of AD, and synaptic dysfunction also correlates with the degree of cognitive decline in AD patients and transgenic mice with Aβ toxicity [4]. However, the specific mechanism of Aβ-induced synaptic dysfunction has not been fully elucidated. Gong et al. reported that ubiquitin hydrolase Uch-L1 rescues Aβ-induced decreases in synaptic function and contextual memory via the restoration of normal levels of the PKA-regulatory subunit IIa, PKA activity, and CREB phosphorylation [5], implying that down-regulation of Uch-1 is associated with synapse loss in AD.

Post-translational modifications of neuronal proteins have been implicated in the pathogenesis of AD. Besides common modifications like phosphorylation, ubiquitination, acetylation, and SUMOylation on tau, APP, and BACE1 [6], more and more attention is being given to the implication of S-nitrosylation in the development and/or progression of AD. Protein S-nitrosylation is a covalent reaction of a nitric oxide (NO) group with the reactive thiol group of a cysteine to form S-nitrosothiol (SNO), which transfer NO-mediated signals. NO can in turn S-nitrosylate a particular set of protein thiols by nitrosylating enzymes to form protein-protein transnitrosylation cascades [7]. For example, Cdk5 is S-nitrosylated at the thiol groups of cysteine residues 83 and 157 to form SNO-Cdk5, which is significantly higher in the brains of AD than in controls. SNO-Cdk5 S-nitrosylates dynamin-related protein 1 (Drp1) to form SNO-Drp1 via transnitrosylation, a process that might account for the Aβ-induced dendritic spine loss [8]. Recently, Lipton’s group at the Scripps Research Institute, La Jolla, coupled and extended ubiquitin hydrolase Uch-L1 and the Cdk5/Drp1 transnitrosylation pathway with synapse loss in AD [9]. Using the biotin-switch assay to analyze S-nitrosylated proteins, they identified an upstream and a downstream transnitrosylation reaction between the above proteins and clarified that the aberrant transnitrosylation cascade from SNO-Uch-L1 to Cdk5 to Drp1 contributes to the synaptic pathology observed in AD (see the middle panel of Fig. 1).

Fig. 1.

Schematic of Aβ-induced synaptic impairments in AD. Left panel, an increase in NO as a result of toxic Aβ and/or an oxidized milieu leads to SNO-Uch-L1 formation and inhibition of Uch-L1 activity. This altered Uch-L1 activity promotes Aβ production/oligomerization as well as downregulation of the PKA/CREB pathway, resulting in synaptic impairments. Middle panel, Aβ oligomers in turn induce NO generation which then activates the Uch-L1/Cdk5/Drp1 transnitrosylation cascade, leading to mitochondrial dysfunction and synaptic impairments. Right panel, the resulting mitochondrial dysfunction promotes further ROS production and an oxidizing environment, where increased NO generation triggers the transnitrosylation cascade and further disrupts mitochondrial function. Together, this form a vicious cycle where Aβ triggers increased NO production resulting in the Uch-L1/Cdk5/Drp1 transnitrosylation cascade leading to mitochondrial and synaptic impairments, and at the same time increasing Aβ and ROS production, so the cycle continues.

The biotin-switch assay is a key technique for characterizing protein-protein transnitrosylation, in which the level of S-nitrosylation of a protein of interest is measured by western blot after substituting SNO with a more stable biotin group via ascorbate-mediated chemical reduction and hence removal of the NO group [10]. The researchers first explained why they investigated Uch-L1 in this non-canonical transnitrosylation cascade as follows. Uch-L1, as an ubiquitin hydrolase, promotes cyclin-dependent kinase activity via an indistinct molecular mechanism. This means that its ubiquitin hydrolase activity does not participate in this reaction mechanism. Therefore, the authors studied the second, non-canonical function of Uch-L1 besides its ubiquitin hydrolase activity. They found that Uch-L1 is a potential substrate for SNO in HEK-nNOS cells, and that Uch-L1 S-nitrosylation is also markedly increased in both hAPP-J20 and Tg2576 AD mouse models which produce Aβ oligomers. Mutation assay and mass spectrometry identify Cys152 as the predominant site of Ucl-L1 S-nitrosylation. The authors then tested the effect of formation of SNO-Uch-L1 on its ubiquitin hydrolase activity, and found that exposure to SNOC decreases Uch-L1 activity and denitrosylation of Uch-L1 increases its ubiquitinylation activity. Together with the previously reported decreased Uch-L1 activity in the AD brain [5], their findings imply that SNO-Uch-L1 formation in the AD brain might account for Uch-L1 inhibition, which also induces synaptic impairments via down-regulation of the PKA/CREB pathway (see the left panel of Fig. 1).

Because over-expression of Uch-L1 decreases Aβ production and the very high abundance of Uch-L1 favors its own S-nitrosylation, the authors next evaluated the relationship between the pathological effects on synapses following Aβ oligomer treatment and Uch-L1 overexpression. They reported that overexpression of Uch-L1 does not greatly affect Aβ-induced synaptic loss. However, C152S mutant Uch-L1, mimicking non-S-nitrosylated Uch-L1, blocks Aβ-induced synaptic loss, strongly supporting the idea that SNO-Uch-L1 might at least in part contribute to synaptic impairments in AD. The authors had previously demonstrated that the transition from SNO-Cdk5 to SNO-Drp1 transnitrosylation contributes to Aβ-induced dendritic spine loss. To enrich and clarify the transnitrosylation pathway involvement in Aβ-related pathology, they performed kinetic, cell-based S-nitrosylation assays to monitor the possible transfer of an NO group from SNO-Uch-L1 to Cdk5 and then to Drp1. The biotin-switch assays showed that both knockdown and mutation of Uch-L1 induce decreases in NO-Cdk5 and SNO-Drp1 generation. Meanwhile, Cdk5 depletion by immunoprecipitation decreases the level of SNO-Drp1. In cell lysates expressing either Cdk5 or Uch-L1, they found that SNO-Uch-L1 S-nitrosylates Cdk5 to form SNO-Cdk5. Conversely, SNO-Cdk5 cannot S-nitrosylate Uch-L1. These findings fully support the conclusion that, under Aβ oligomer treatment, NO is transferred from SNO-Uch-L1 to SNO-Cdk5 and then to SNO-Drp1 to form a transnitrosylation cascade. This is strongly supported by a quantitative method based on the Nernst equation to characterize these transnitrosylation reactions thermodynamically. Finally, the authors showed that the level of SNO-UchL1 is significantly higher in the brains of late AD patients than in controls. Some pathological alterations of AD like Aβ toxicity and mitochondrial dysfunction can each trigger NO generation, and this probably initially ignites S-nitrosylation of Uch-L1 and evokes this transnitrosylation cascade (Uch-L1/Cdk5/Drp1), resulting in mitochondrial deficiency and synaptic impairments. Meanwhile, mitochondrial deficiency further promotes NO generation, to form a vicious cycle (see the right panel of Fig. 1), which will irreversibly worsen AD progress without effective drugs. In addition, increased production of free radicals due to mitochondrial impairment could increase oxidative stress, which might lead to intracellular calcium influx and the activation of cascades like calpain-CDK5-STAT3-BACE1 and thus increasing Aβ production [11], which joins the cycle and aggravates the situation. By mass spectrometry analysis, the authors found an acidic (Glu)-basic (Arg) region as a candidate motif for S-nitrosylation surrounding the critical Cys residue of Uch-L1. Synthesized small-molecule compounds docking this motif may block S-nitrosylation of Uch-L1 and in sequence prevent or at least attenuate the transnitrosylation reactions to stop the vicious cycle-induced synaptic loss in AD. Therefore, targeting the transnitrosylation cascade might provide a novel therapeutic strategy for AD. Meanwhile, their findings show that unrelated enzymes aberrantly work together like aberrant transnitrosylation reactions and may be involved in brain dysfunction. Given the many proteins that can be S-nitrosylated, it would be of interest to determine whether such transnitrosylation exists among other proteins and how common it may be; this could pose challenges to the existing pathway analysis tools. Certainly, further identification of S-nitrosylated or transnitrosylated proteins in the AD brain will shed more light on the molecular mechanisms and pharmaceutical research on the disease. An immediate next question posed by the current study is to further investigate whether Uch-L1 acts as the initial S-nitrosylated protein in the present transnitrosylation cascade. Moreover, it is worth exploring whether SNO-Uch-L1 formation in the AD brain inhibits the primary ubiquitination function of Uch-L1, which will better explain the mechanism of Aβ oligomer-induced synaptic loss in AD.

Conflict of interest

The authors declare that they have no conflict of interest.