Our cells are constantly engaged in critical life and death decisions. Genes that are responsible for controlling this life or death switch and play a crucial role in safeguarding cells from dying are categorized as cell survival genes. Apoptosis serves as a primary defense against cancer and infection (Wong, 2011; Jorgensen et al., 2017). While apoptosis is a necessary part of our survival, dysregulation or abnormal expression of cell survival genes can lead to pathological conditions, including cancer or neurodegenerative disease, emphasizing the delicate nature of this biological process (Offen et al., 2000; Kalimuthu and Se-Kwon, 2013). In cancer cells, survival genes can contribute to the uncontrolled growth and survival of malignant cells, leading to tumor formation and progression. Survival genes are commonly overexpressed or mutated in cancer cells. This overexpression confers resistance to apoptosis, allowing cancer cells to avoid programmed cell death even when they should be eliminated due to genetic abnormalities or uncontrolled growth (Qian et al., 2022).

In neurons, survival genes are essential for maintaining the health and longevity of these specialized cells. Neurons are post-mitotic cells, which are fully differentiated and non-dividing after neurogenesis, and their survival is critical for the proper functioning of the nervous system (Pfisterer and Khodosevich, 2017). Survival genes play pivotal roles in neurons, preventing apoptosis, protecting against cellular stress, including oxidative stress and DNA damage, promoting neuronal differentiation, and providing neurotrophic support (Morrison et al., 2002; Portt et al., 2011; Hollville et al., 2019).

The study of cell survival genes is of significant interest in CNS degenerative disease or cancer research (Seo and Park, 2020). It is important to accurately understand the role of each survival gene, i.e., whether a certain gene has a dual role and the intra-, inter-, and extracellular conditions of various cell types in which each role is performed. By understanding the mechanisms that control cell survival and apoptosis, researchers can identify targets for the treatment of neurodegenerative diseases or cancer.

In the present review, recent research reports and the findings of our research team on the role of survival genes, including the recently discovered novel gene ‘DX2,’ a splicing variant of AIMP2, lacking the second exon among the total of four exons encoding the full-length AIMP2 (aminoacyl-tRNA synthetase-interacting multi-functional protein 2), in degenerative neuronal tissue cells, cancer cells, and normal healthy cells, will be reviewed. Strategic directions for the clinical application of survival genes in neurodegenerative disease and cancer will also be discussed.

AIMP2 (Aminoacyl tRNA synthase complex-interacting multifunctional protein 2) is one of three auxiliary proteins that form Aminoacyl -tRNA synthetases complex (ARS complex) (Kim et al., 2014). ARSs are enzymes to catalyze the ligation of amino acids to cognate tRNAs in translation (Lee et al., 2004). Among 20 mammalian ARSs, 9 ARSs form the multi-tRNA synthetase complex (MSC) with three non-enzymatic factors for efficient translation. The function of non-enzymatic factors, aminoacyl-tRNA synthetase-interacting multi-functional proteins (AIMPs), is a stabilization of MSC via interaction with their adjacent ARSs, Lysyl-tRNA synthetase (KRS) is the ARS ligating the lysine to tRNALys (Park et al., 2010). Among the AIMPs, aminoacyl-tRNA synthetase-interacting multi-functional protein2 (AIMP2) is the major scaffolding molecule to stabilize the MSC for enhancing the efficiency of translation. KRS shows the strong binding to AIMP2 in MSC (Kim et al., 2022). In mammals, AIMP2 exists in the KRS complex bound form or in a free form in the cytosol (Choi et al., 2011). It has been reported that AIMP2 plays a crucial role in regulating cell fate. It exhibits anti-proliferative activity by augmenting the growth-arresting signal of TGF-β (Kim et al., 2003). Additionally, AIMP2 facilitates cell death through the activation of p53 and the apoptotic signaling pathway of TNF-α (Choi et al., 2009a).

In the similar context, AIMP2 has been considered as a haploinsufficient tumor suppressor (Choi et al., 2009b). Kim et al. demonstrated that mice deficient AIMP2 were neonatal lethal attributed to lung failure stemming from the overproliferation of lung epithelial cells. AIMP2 heterozygous mice with a reduced expression level of AIMP2 exhibited increased susceptibility to tumorigenesis (Kim et al., 2003). Additionally, Kim et al. reported that AIMP2 facilitates the ubiquitin-mediated degradation of FUSE-binding protein (FBP), a transcriptional activator of c-Myc (Kim et al., 2003). Dysregulation or overexpression of c-Myc is common in many cancers and is associated with aggressive tumor behavior.

DX2 was discovered by a research team led by Dr. Sunghoon Kim at Seoul National University. Choi et al. identified ‘DX2’ as a splicing variant of AIMP2, lacking the second exon among the total of four exons that encode the full-length AIMP2 (Fig. 1). This splicing variant has been reported to be overexpressed in various cancer types, including breast cancer, lung cancer and pancreatic cancer. The research team reported that DX2 promotes cancer cell proliferation, inhibits apoptosis, and contributes to cancer progression (Choi et al., 2011, 2012).

Figure 1. A schematic representation of DX2 compared to full-length of AIMP2. AIMP2: Aminoacyl-tRNA Synthetase-Interacting Multifunctional Protein 2, DX2: Exon 2 Deleted Splicing Variant of AIMP2.

Since its discovery, many studies have demonstrated that DX2 supports several key cancer hallmarks, including sustained proliferative signaling, evasion of growth suppressors, and resistance to cell death. DX2 stabilizes KRAS (Kirsten rat sarcoma viral oncogene homolog), a small GTPase transducer protein, by inhibiting its ubiquitin-mediated degradation, thereby enhancing KRAS-induced cellular proliferation and transformation (Kim et al., 2022) . Moreover, studies have shown that DX2 compromises the tumor-suppressive activities of AIMP2 by competing for the binding sites of AIMP2-binding proteins, which normally control various tumor suppressor pathways including p53 and TNF-α (Choi et al., 2009a; Choi et al., 2011). Research has also revealed that DX2 prevents oncogene-induced apoptosis by directly binding to and inhibiting p14/ARF (Oh et al., 2016) or by interacting with heat shock protein 70 (HSP70) (Lim et al., 2020). While direct research on DX2’s impact on the tumor microenvironment and immune evasion is still emerging, its role in stabilizing KRAS and promoting cancer progression likely influences these aspects indirectly. The growing body of evidence highlighting DX2’s importance in cancer has led to efforts to develop targeted therapies, including small molecule inhibitors that disrupt DX2’s interactions with various oncoproteins.

Recent development of the chemical inhibitors specific to oncogenic KRAS mutants revives much interest to control KRAS-driven cancers. It is suggested that DX2, acts as a cancer-specific regulator of KRAS stability, augmenting KRAS-driven tumorigenesis. Kim et al. reported in their recent study that co-expression of DX2 increased the incidence of tumors in mice expressing the KRAS mutant (Freedman et al., 2016), suggesting a potential connection between the two oncogenic proteins in tumor formation and growth (Kim et al., 2022). It was demonstrated that DX2 specifically binds to the hypervariable region and G-domain of KRAS in the cytosol prior to farnesylation, then, DX2 competitively blocks the access of Smurf2 (SMAD Ubiquitination Regulatory Factor 2) to KRAS, thus preventing ubiquitin-mediated degradation. However, the results of the interaction between AIMP2 full length and KRAS, that is, the side-by-side comparison of interaction between AIMP2 to KRAS and DX2 to KRAS, have not been reported in that publication, so further research on this part is justified.

Choi et al. reported that DX2 is often highly expressed in chemo-resistant ovarian cancer and DX2 compromised the tumor necrosis factor alpha-dependent pro-apoptotic activity of AIMP2 via the competitive inhibition of AIMP2 binding to TRAF2 that plays a pivotal role in the regulation of NF-κB. They demonstrated that the direct delivery of siRNA against DX2 into abdominal metastatic tumors of ovarian cancer significantly suppressed the growth rate of tumors and enhanced apoptosis and the decreased TRAF2 level in the treated cancer tissues (Choi et al., 2012).

Research results have revealed that DX2 interacts with target proteins via an AIMP2-independent pathway. Lim et al. introduced an intriguing interaction between HSP70 and DX2. HSP70 identifies not only the amino (N)-terminal flexible region but also the glutathione S-transferase domain of DX2 through its substrate-binding domain. This interaction effectively thwarts the Siah1-dependent ubiquitination of DX2. HSP70 has been shown to enhance DX2-induced cell transformation, thus amplifying cancer progression in vivo. The interplay between these two molecules is supported by the observed positive correlation between HSP70 and DX2 levels in an array of lung cancer cell lines and patient samples (Lim et al., 2020). Additionally, it has been suggested that the elevated level of DX2 across various serotypes of lung cancer is due to DX2 physically binding with P14/ARF, which in turn suppresses oncogene-induced senescence and apoptosis in cancer. As a result, it has been suggested that the increase in DX2 levels can enhance drug resistance in chemotherapy, diminishing its efficacy (Oh et al., 2016).

On the other hand, Dr. Levens D’s lab presented a slightly different story. They demonstrated that AIMP2 dissociates from the ARS complex, translocates to the nucleus, associates with the far upstream element-binding protein (FBP), and co-activates the transcription of a new FBP target: ubiquitin-specific peptidase 29, thereby stabilizing p53 in response to oxidative stress. The accumulated p53 quickly induces apoptosis. Thus, FBP and AIMP2 help coordinate the molecular and cellular response to oxidative stress. In this report, they showed that DX2 interacted with endogenous FBP similarly to AIMP2. Additionally, both DX2 and AIMP2 were ubiquitinated to a similar extent, indicating that both AIMP2 and DX2 retained the molecular features necessary to interact with FBP and modify its function (Liu et al., 2011).

Dr. Kim’s research team has recently developed a couple of DX2 inhibitors that targets lung cancer. They discovered a potent DX2 inhibitor that is most efficacious in H460 and A549 cells, utilizing a ligand-based drug design strategy (Lee et al., 2022). They reported the synthesis and discovery of the first potent proteolysis-targeting chimera (PROTAC) degrader of DX2 as a lung cancer drug, combining DX2 inhibitors with selective E3-ligase ligands with optimized linkers (Lee et al., 2023). However, it’s important to note that the cancer treatment effect achieved by inhibiting DX2 does not imply that DX2 is an oncogene. This understanding aligns with the context of β-catenin or FAK-related studies (Yoon et al., 2015; Shang et al., 2017).

Neurons are fully differentiated cells and do not undergo cell division after neurogenesis. Neurodegenerative diseases are a group of disorders characterized by the progressive degeneration or death of neurons in the nervous system. The dysregulation or dysfunction of cell survival genes can contribute to the development and progression of neurodegenerative diseases.

Parkinson’s disease (PD) is a movement disorder presenting primarily with a combination of bradykinesia, rigidity and tremor and non-motor symptoms such as autonomic, cognitive, and psychiatric disturbances. Destruction of dopaminergic neurons in the substantia nigra pars compacta with a consequent reduction of dopamine actions in the corpus striatum, parts of the basal ganglia system that are involved in motor control has been thought to be a commonly pathogenesis of sporadic and familial forms of PD (Poewe et al., 2017). Mutations in the genes such as parkin or α-synuclein are thought to be associated with familial forms of PD. In addition, chronic neuroinflammation, sustained levels proinflammatory cytokine including TNF-α have been associated with neurodegeneration in the substantia nigra (Aloe and Fiore, 1997; Ferrari et al., 2006; De Lella Ezcurra et al., 2010; Pott Godoy et al., 2010; Chertoff et al., 2011).

Amyotrophic Lateral Sclerosis (ALS) is a neurodegenerative disease linked to the selective motor neuron death in the spinal cord, brain stem and motor cortex (Al-Chalabi and Hardiman, 2013). Typical ALS patients suffer from a gradual loss of motor function caused by muscle atrophy and degeneration (van Es et al., 2017). To date, the main cause of ALS has not been fully elucidated while various pathological hallmarks such as misfolded protein aggregation, glutamate excitotoxicity and neuroinflammation seems to be related to the development ALS symptom (Pasinelli and Brown, 2006; Yang et al., 2010; Bucchia et al., 2015; Palomo and Manfredi, 2015). In Alzheimer’s disease, abnormal processing and accumulation of proteins, such as amyloid-beta and tau, lead to the formation of plaques and tangles in the brain. These abnormal protein aggregates can interfere with cell survival pathways, leading to neuron dysfunction and cell death (d’Errico and Meyer-Luehmann, 2020).

Dysregulation of autophagy and impaired lysosomal function have been linked to the accumulation of toxic protein aggregates in neurodegenerative diseases (Seranova et al., 2017). Dysfunction of DNA repair mechanisms results in the accumulation of DNA lesions, leading to neuronal degeneration (Maynard et al., 2015). Dysfunctional mitochondria can also lead to increased oxidative stress and impaired energy production, contributing to neurodegenerative diseases (Norat et al., 2020). Cell survival genes play a role in modulating autophagy, and their dysfunction can disrupt protein clearance pathways, contributing to the pathology of neurodegenerative diseases (Guo et al., 2018). Cell survival genes are also involved in the proper functioning of DNA repair pathways and in maintaining mitochondrial integrity, thus preventing oxidative stress and energy problems in neurodegeneration (McKinnon, 2009; Cowan et al., 2019; Volkert and Crowley, 2020). Understanding the involvement of cell survival genes in neurodegenerative diseases has prompted the exploration of potential therapeutic interventions. Researchers are investigating various strategies to modulate cell survival pathways and enhance neuronal survival, including gene therapy, small molecule inhibitors, and other targeted approaches.

Our research team has revealed that DX2 plays multifaceted survival roles when it is overexpressed in neuronal cells that are exposed to cell death- inducing stressful environments, including neuroinflammation or ROS.

PARP1 dependent pathway

A previous study reported that AIMP2 expression level is augmented in postmortem brains of PD patients as well as parkin knock-out mice (Ko et al., 2005). AIMP2 leads to aberrant poly (ADP-ribose) polymerase-1 (PARP1) activation dependent neuronal cell death in Parkinson’s disease (Lee et al., 2013). PARP-1 has a critical role to recover damaged DNA(de Murcia et al., 1997). On the other hand, overactivation of PARP-1 causes depletion of cytosolic NAD+ forming poly-ADP-ribose (PAR), releasing mitochondrial apoptosis-inducing factor (AIF), nuclear translocation of AIF, and causing cell death called parthanatos (Hong et al., 2004). AIMP2 accumulation results in PARP1 overactivation and neuronal cell death even without DNA damage, via direct association of these proteins in the nucleus. Inhibition of PARP1 overactivation could protect neuronal cell death in neurodegenerative disease. PARP1 mediated DX2 MOA in the treatment of neurodegenerative disease is represented in Fig. 2A. Lee et al. reported that DX2 competes with AIMP2, binds to PARP1 more strongly than AIMP2, but doesn’t induce its overactivation, resulting in neuronal cell survival (Lee et al., 2013).

Figure 2. (A) PARP1-mediated neuronal survival pathway. DX2 competes with AIMP2, binds to PARP1 more strongly than AIMP2, but doesn’t induce its overactivation, resulting in neuronal cell survival. (B) TRAF-mediated neuronal survival pathway. DX2 competes with AIMP2, thus inhibiting AIMP2 stimulated apoptosis in TNF-α-activated conditions. DX2 interacts much stronger than AIMP2 with TRAF2, which inhibits c-IAP1 binding to TRAF2 leading to stabilization of TRAF2, even upon TNF-α signals, thereby allowing cell survival. (C) p53 mediated cell survival pathway. DX2 competes with AIMP2, inhibiting the apoptotic activity of AIMP2 by interrupting interaction with p53. While AIMP2 blocks the interaction between p53 and MDM2, DX2 does not block MDM2 binding to p53. AIMP2: Aminoacyl-tRNA Synthetase-Interacting Multifunctional Protein 2, DX2: Exon 2 Deleted Splicing Variant of AIMP2, NDD: Neuronal Degenerative Disease, PARP-1: Poly (ADP-ribose) polymerase-1, PAR: Poly-ADP-ribose, TRAF2: TNF Receptor Associated Factor 2, C-IAP1: Cellular Inhibitor of Apoptosis Protein 1, Mdm2: Mouse double minute 2 homolog.

TRAF dependent pathway

It’s been reported that AIMP2 promotes TNF-α-dependent cell death via TRAF2 pathway (Choi et al., 2009a). In this pathway, free form of AIMP2 binds to TNF receptor-associated factor 2 (TRAF2), a key player in the TNF-α signaling pathway, followed by recruiting the cellular inhibitor of apoptosis protein 1 (c-IAP1). c-IAP1 plays a role as E3 ubiquitin ligase to facilitate the degradation process of a targeted protein. In this process, TRAF2 can be degraded, which finally leads to cell death. It is proposed that the DX2 competes with AIMP2, thus inhibiting AIMP2 stimulated apoptosis in TNF-α-activated conditions.

TRAF2 mediated DX2 MOA in the treatment of neurodegenerative disease is represented in Fig. 2B. Kook et al. (2023) reported that the DX2 transgene encoded by AAV2-DX2, competes with AIMP2, thus inhibiting AIMP2 stimulated apoptosis in TNF-α-activated conditions. The free form of AIMP2 binds to TRAF2 protein to promote ubiquitination, thus inducing cell death. However, ubiquitination of TRAF2 by AIMP2 is significantly reduced in the presence of DX2, which compromises the TNF-α-dependent pro-apoptotic activity of AIMP2 via the competitive inhibition of AIMP2. DX2 interacts much stronger than AIMP2 with TRAF2, which inhibits c-IAP1 binding to TRAF2 leading to stabilization of TRAF2, even upon TNF-α signals, thereby allowing cell survival.

p53 dependent pathway

p53 is a key regulator to direct cell death. In the neurodegenerative disease, p53 is appreciated a causative gene to sustain neuronal death. Choi et al. demonstrated that AIMP2 binds to the N-terminal of p53, a binding domain for MDM2, thereby preventing MDM2-mediated ubiquitination and degradation of p53, which induces the stability of p53 and pro-apoptotic activity (Choi et al., 2011). DX2 competes with AIMP2, inhibiting the apoptotic activity of AIMP2 by interrupting interaction with p53. While AIMP2 blocks the interaction between p53 and MDM2, DX2 does not block MDM2 binding to p53. P53 mediated DX2 MOA in the treatment of neurodegenerative disease is represented in Fig. 2C.

DX2 cell survival MOA: AIMP2-dependent or AIMP2-independent pathway

DX2 appears to play its cell survival role through either an AIMP2-dependent or AIMP-independent pathway. When DX2 acts as a competitive antagonist of AIMP2, a known tumor suppressor, it becomes logically plausible to consider DX2’s anti-apoptotic role in cancer cells. However, the role of DX2 in normal healthy tissue cells or degenerative cells (ex. neurons in neurodegenerative disease), where the expression or cellular function of AIMP2 and DX2 is distinct from in cancer cells, suggests a different perspective. The level of AIMP2 might be elevated in neurons in neurodegenerative diseases and/or there could be distinct expression or translocation patterns in cancer or in neurodegenerative disease or other disease. Furthermore, Kook et al. demonstrated the effects of DX2 on non-neuronal cells in the central nervous system through the neuroinflammatory TRAF2 signaling pathway. In the central nervous system, microglia secrete pro-inflammatory cytokines such as TNF-α and astrocytes support activation of microglia in neurodegenerative diseases. These processes exacerbate disease progression and induce motor neuron death in ALS. Kook et al. observed increased glial cell activation in the spinal cords of mutant SOD1 mice. Notably, DX2 reduced the number of microglia and astrocytes in the spinal cord compared to mutant SOD1 mice without DX2 expression (Kook et al., 2023).

Taken together, the role of DX2 in apoptosis/ cell survival seems to differ depending on the cell types and conditions. In particular, when DX2 acts via AIMP2-independent manner, DX2 tumorigenic potential becomes slimmer. Our laboratory performed the following experiments to test the tumorigenic potential of DX2 in normal tissue cells or neuronal cells.

Recent research highlights several key advances in the role of neural input and regulated cell death in cancer control. Cancer cells not only invade nerves but also manipulate the nervous system through processes like perineural invasion (PNI) and neurogenesis. PNI allows cancer cells to migrate along nerve fibers, which facilitates metastasis. Meanwhile, neoneurogenesis refers to the formation of new nerves within tumors, driven by neurotrophic factors secreted by cancer cells, such as NGF and netrin-1. These processes reshape the tumor microenvironment, promoting growth and resistance to therapy by integrating neural components into tumor biology (Krishna and Hervey-Jumper, 2022; Prillaman, 2024) .

These insights suggest that targeting neural interactions and exploring novel mechanisms of regulated cell death could open new avenues for cancer treatment. Researchers are actively developing therapies based on these discoveries to improve patient outcomes, particularly for tumors resistant to conventional treatments. Although there is currently no direct evidence linking DX2 to neural inputs in cancer, further investigation is needed to assess whether DX2 inhibition could have therapeutic potential in suppressing cancer progression and metastasis in cancer patients.

DX2 does not induce any histopathological abnormalities or neoplastic pattern in whole body organs of mice

To evaluate the tumorigenicity potential of DX2, our laboratory conducted a toxicity test using a mouse model. DX2 overexpression AAV vector was administered to 7- week- old C57BL/6J mouse (male and female, n=100, each) via intrathecal administration. Whole body histopathological analysis, as well as clinical observations and clinical chemistry were conducted at week 3, 13 and week 26 (Fig. 3).

Figure 3. DX2 overexpression AAV2 vector were administered to 7-week-old C57BL/6J mouse (male and female, n=100, each) via intrathecal administration. Whole body histopathological analysis, as well as clinical observations and clinical chemistry were conducted at week 3, 13 and week 26.

There was no death or moribundity occurred in any group during the study period. No changes were noted in the clinical observations, body weights, clinical chemistry. No gross pathological lesions were observed in any group. Specifically, no histopathological lesions were observed at the injection site (lumbar), including the dorsal root ganglion in any group.

DX2 does not induce any histopathological abnormalities or neoplastic pattern in whole body organs of rabbits

Next, our laboratory performed a toxicity test using a rabbit model to evaluate the tumorgenicity potential of DX2. DX2 overexpression AAV vector were administered to 13- week-old male Chinchilla rabbits (male, n=4-6) via intravitreal or subretinal administration. Whole-body histopathological analysis was conducted at week 4. Experiment data confirmed that DX2 overexpression AAV vector administered animals did not show any histopathological abnormalities or neoplastic pattern in any of the tissues evaluated.

DX2 demonstrates no cytotoxic effect and does not impact cell growth in both MEF and neuron cells under normal conditions

Additionally, the effects of overexpression of DX2, on cellular growth has been evaluated by Choi et al. (2011). Ectopic expression of DX2 did not hinder the growth of or induce any morphological abnormalities of neuronal or other cell types. DX2 efficiently suppressed Adriamycin-induced cell death. DX2 has no cytotoxic effect or no effect on cell growth or morphology of MEFs under normal conditions. Lee et al. also conducted similar experiments using primary neuronal cells and SH-SY5Y cells (Lee et al., 2024a). DX2 efficiently inhibits oxidative-induced neuronal cell death in a dose dependent manner; however, DX2 has no cytotoxic effect / no effect on cell growth of neuronal cells under normal conditions.

DX2 does not induce oncogenicity

Lee et al. conducted an in vitro ‘anchorage independent colony forming assay’ to assess the oncogenic potential of DX2 (Lee et al., 2024a). In this study, DX2 expressing plasmid with or without c-Myc and K-Ras expressing plasmids were transfected into primary neuronal cells, MEF (mouse embryonic fibroblast) cells or HEK293 cells and the anchorage independent growth was analyzed (Lee et al., 2024b). DX2 expression did not affect colony formation, indicating that DX2 does not induce oncogenic cell growth. Consistent to the results obtained in neuronal cells, DX2 expression did not affect colony formation MEF or HEK293 cells as well, indicating that DX2 does not induce oncogenicity in the neuronal cells as well as other cells in normal condition.

Previously developed DX2 transgenic mice were used to test the potential tumorigenicity of DX2. A construct was made that expressed DX2 under the control of a CMV promoter, allowing expression of DX2 in mouse whole body. DX2 TG animals did not show any neoplastic pattern in the kidney, liver, spleen, brain, lymph node, lung and etc. The survival was also same between wild type and TG group. No difference in survival rate was observed between WT and DX2 TG mice (Lee et al., 2024b).

Gene therapy stands as a significant and emerging strategy in the treatment of neurodegenerative disorders. Table 1 shows representative promising therapeutic targets for gene therapy of neurodegenerative diseases (Chen et al., 2020b). It is noteworthy that AKT and X-box binding protein 1 (XBP1) are now being utilized as curative transgenes for targeting neurodegenerative diseases. AKT and XBP1 were previously classified solely as oncogenes, however, they have recently been employed in gene therapy for neurodegenerative disorders, promoting neuronal survival.

Table 1 Representative promising therapeutic targets for gene therapy of neurodegenerative disorders (Chen et al., 2020b)

Disorder	Gene - delivery system	Target pathway
Traumatic optic nerve injury	AAV2-XBP-1	ER stress and UPR
Parkinson’s disease	AAV2-XBP-1
Parkinson’s disease	AAV5-Bip
Amyotrophic lateral sclerosis	AAV6-SIL1
Huntington’s disease	AAV2-XBP-1
Optic nerve injury	AAV2-AKT	mTOR signaling
Optic nerve injury	AAV2-S6K1
Optic nerve injury	AAV2-PTEN
Parkinson’s disease	AAV1-AKT
Alzheimer’s disease and Parkinson’s disease	AAV1-AKT
Huntington’s disease	AAV1-caRheb
Parkinson’s disease	AAV2-HSP70	Mitochondrial function
Alzheimer’s disease	AAV2-PINK1
Alzheimer’s disease	AAV2-PSD95-6ZF-VP64	Epigenetic regulation
Alzheimer’s disease	AAV2-PINK1	Autophagy
Parkinson’s disease	AAV6-Lamp2a
Parkinson’s disease	AAV2-TFEB
Amyotrophic lateral sclerosis	AAV9-snapin
Alzheimer’s disease	AAV2/8-sTREM2	Microglial and astrocyte function
Alzheimer’s disease	Lentivirus-PGRN

The PI3K/AKT pathway stands as one of the most frequently over-activated intracellular signaling pathways in several human cancers. Several proteins serve as targets for AKT, including the FOXO family and mTOR, most of which function as transcription factors and can induce alterations in the function and metabolism of cancer cells if mutated (Tapia et al., 2014). In light of the emerging alterations in PI3K/AKT pathway genes have been extensively reported in recent cancer research, the inhibitors of PI3K/AKT pathway have brought a new era for targeted therapy of cancer. More PI3K inhibitors have been introduced, since the first approval of idelalisib (CAL-101) validated the druggability of the PI3K pathway (Brown and Banerji, 2017; Huck and Mochalkin, 2017).

XBP1 is a multitasking transcription factor and functions as a key mediator for the endoplasmic reticulum stress (ERS) response. Upon disturbance of homeostasis in the endoplasmic reticulum (ER) and activation of the unfolded protein response (UPR), the mRNA of XBP1 gene is processed to an active form by an unconventional splicing mechanism (Bommiasamy et al., 2009; Park et al., 2021). XBP1 enhances the transcription of genes encoding molecular chaperones and ER-associated protein degradation (ERAD) components, thereby restoring ER homeostasis (Bommiasamy et al., 2009; Shoulders et al., 2013; Dewal et al., 2015; Park et al., 2021). Cancer cells often experience ERS due to the aggregation of misfolded proteins within the ER. It exhibits high expression levels in various cancers, playing a significant role in biological processes of tumor cells. XBP1s is closely associated with invasion and metastasis in tumorigenesis, while inhibition of XBP1 expression reduces the viability and drug resistance of tumour cells (Chen et al., 2020a).

As shown in Table 1, the role of PI3K-AKT has been viewed from various perspectives, notably its implication in cell survival within neurodegenerative diseases. Its’ been shown PI3K-AKT mediated signaling pathway protect dopaminergic neurons, hippocampal neurons, and cortical neurons, and suppress the activation of microglia , which may help with the prevention and treatment of like Parkinson’s disease (PD) and Alzheimer’s disease (AD) (Long et al., 2021). The involvement of the PI3K-AKT signaling pathway has been noted in neuroprotection through the reduction of inflammation, oxidative stress, and apoptosis across various trials, including clinical studies. Findings from a study on Alzheimer’s disease (AD) patients revealed that modulation of the p-AKT/PTEN pathway affected key inflammatory and oxidative stress markers implicated in AD pathology, ultimately mitigating the AD pathological cascade and cognitive decline (Mohamed et al., 2019). These studies indicate the clinical relevance of the PI3K-AKT pathway and suggest that this pathway plays a significant role in mediating neuroprotective actions in neurodegenerative disorders such as AD and PD (Goyal et al., 2023).

Recent research has highlighted the importance of the UPR, including XBP1, in the context of neurodegenerative diseases treatment (van Ziel and Scheper, 2020). Many neurodegenerative diseases are associated with the accumulation of misfolded proteins in neurons (Sweeney et al., 2017). Adaptation to ERS is mediated by the activation of the UPR, an integrated signal transduction pathway that attenuates protein folding stress by controlling the expression of distinct transcription factors including XBP1 (Hetz et al., 2011). The UPR, including the activation of XBP1, is an attempt by neurons to manage and eliminate these protein aggregates (Vidal et al., 2021). However, in chronic and severe cases, the UPR can become overwhelmed, leading to neuronal dysfunction and cell death (Lindholm et al., 2006). This process contributes to the progressive nature of neurodegenerative diseases. As shown in Table 1 clinical trials, targeted expression of XBP1 on a viral-mediated vector suggest a potential use of gene therapy strategies to modulate the UPR in the context of neurodegenerative disease such as Parkinson’s disease, Huntington’s disease or traumatic optic nerve injury etc.

In the present review, recent research reports and our findings on the role of survival genes, including DX2 gene, in degenerative neuronal tissue cells and cancer cells, were discussed. A strategic direction for clinical application of survival genes in neurogenerative disease and cancer is needed (Fig. 4, 5).

Figure 4. Strategic application of DX2 in cancer. AIMP2: Aminoacyl-tRNA Synthetase-Interacting Multifunctional Protein 2, DX2: Exon 2 Deleted Splicing Variant of AIMP2, PARP-1: Poly (ADP-ribose) polymerase-1, TRAF2: TNF Receptor Associated Factor 2, FBP: FUSE-binding protein, K-RAS: Kirsten rat sarcoma viral oncogene homolog.

Figure 5. Strategic application of DX2 in degenerating neuron. AIMP2: Aminoacyl-tRNA Synthetase-Interacting Multifunctional Protein 2, DX2: Exon 2 Deleted Splicing Variant of AIMP2, PARP-1: Poly (ADP-ribose) polymerase-1, TRAF2: TNF Receptor Associated Factor 2, PARKIN: 465-amino acid residue E3 ubiquitin ligase.

DX2 overexpression is observed in various cancer types and it promotes cancer cell proliferation and cancer progression. However, in degenerating neurons that are exposed to conditions such as neuroinflammation or ROS, etc., DX2 overexpression plays multifaceted survival roles. The survival effect of DX2, as an antagonistic competitor of AIMP2, through PARP1, TRAF, and p53 mediated pathway, has been demonstrated in degenerating neurons (Han et al., 2008; Choi et al., 2009a; Lee et al., 2013). Although the initial findings of DX2 have generated considerable interest as a potential anti-cancer therapeutic target, further research and validation has been justified to determine the efficacy, safety, and clinical potential of DX2-targeted therapies, particularly in neurodegenerative disease, given its dual role.

Our study results suggest that whole body high expression of DX2 is not toxic or tumorigenic in normal healthy subjects or neurodegenerative patients whose endogenous DX2 levels are not elevated. The expression of DX2 in normal, healthy cells in whole-body mouse organs is observed to be very insignificant and it seems to be mainly expressed in only in spleen and thymus. Notably, endogenous expression of DX2 is rarely observed in the brain (Lee et al., 2024a) . In neurons where the DX2 gene is absent, its inhibition would have negligible effects on cellular function. Instead, the focus should be on elucidating mechanisms to inhibit accumulated AIMP2’s actions in degenerating neurons. In-depth analysis is warranted for the pattern of DX2 expression across whole human tissue and the difference in DX2 expression pattern in various diseases.

Although study reports on the cell death mechanisms of accumulated AIMP2 and the pathogenesis of neurodegenerative diseases have begun to emerge, substantial amount of clinical data on the free form AIMP2 accumulation and neuronal death across various neurodegenerative diseases needs to be mounted. As more data accumulates on the precise accumulation levels and timing at which accumulated AIMP2 induces cell death, it will be possible to establish more accurate and detailed treatment strategies for gene therapy aimed at treating neurodegenerative diseases by delivering DX2 (Yun et al., 2017; Ham et al., 2020; Kook et al., 2023; Lee et al., 2024a).

Furthermore, it is necessary to set the direction of clinical applications: suppressing DX2 expression or activity when DX2 is highly secreted in cancer cells, and stimulating neuronal cell survival through DX2 overexpression in neurodegenerative diseases caused by excessive nerve cell death.

In addition, depending on the characteristics of the disease and the treatment purpose, it is necessary to find the optimal approach for delivering survival gene including DX2 inhibitors and overexpression vectors, to the lesion through target-site-specific delivery. Several studies have demonstrated an inverse association between cancer and neurodegenerative disease, including dementia, supporting our argument (Attner et al., 2010; Bajaj et al., 2010; Driver et al., 2012; Musicco et al., 2013; Ganguli, 2015; Freedman and Pfeiffer, 2016; Frain et al., 2017).

I gratefully thank Hyorin Hwang and Minhak Lee for manuscript editing.

This work was supported by the National Research Foundation of Korea (NRF) Grant (NRF-2019R1A2C1006752) and Academic Research Support Program of Gangneung-Wonju National University (2022100153).

We have no conflicts of interest to disclose.