The role of neddylation in colorectal cancer and its therapeutic potential

Abstract

Colorectal cancer (CRC) is the third most common malignancy worldwide, and the five-year survival rate for patients with metastatic disease remains below 15% despite advances in current therapeutic approaches. Post-translational modifications (PTMs) play a pivotal role in CRC initiation and progression, among which neddylation—a critical ubiquitin-like modification—is closely associated with tumor cell proliferation, migration, and chemotherapy resistance. This modification covalently attaches neural precursor cell-expressed developmentally downregulated protein 8 (NEDD8) to lysine residues of substrate proteins, thereby regulating protein stability, DNA repair, and immune responses. In CRC, key enzymes in the neddylation pathway, such as NAE1, UBE2M, and DCUN1D1, are frequently aberrantly activated, leading to the stabilization of key oncoproteins and cell-cycle regulators by preventing their ubiquitin-mediated degradation, thereby promoting tumor progression and drug resistance. Although neddylation has been extensively studied in various cancer types, its precise role in CRC has not been fully elucidated. Recent studies have shown that targeting neddylation—particularly with NAE1 inhibitors such as MLN4924—can significantly suppress tumor progression and offer new therapeutic opportunities to overcome chemoresistance. This review systematically summarizes the roles of neddylation in CRC pathogenesis, chemoresistance, and immune microenvironment remodeling, with a focus on the clinical potential of combining neddylation-targeted inhibitors with chemotherapy and immunotherapy, as well as the prospective application of liquid biopsy in precision monitoring, aiming to provide a theoretical basis and future directions for molecular targeted therapy and clinical translation in CRC.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-025-04264-7.

Introduction

Colorectal cancer (CRC) ranks as the third most commonly diagnosed malignancy worldwide, with a mortality rate nearing 50%. The increasing incidence of early-onset CRC, especially in developing countries, has emerged as a major public health concern. Despite advances in conventional treatments—including surgery, chemotherapy, and radiotherapy—the five-year survival rate for patients with metastatic CRC remains below 15%, highlighting the urgent need for more effective therapeutic strategies [1, 2].

In recent years, targeted therapies have been gradually integrated into CRC treatment. Compared to traditional approaches, these therapies have markedly improved the prognosis of patients with advanced malignancies, particularly those with metastatic disease. Among emerging strategies, targeting post-translational modifications (PTMs) of proteins has gained attention as a promising direction to overcome therapeutic resistance [3–5]. PTMs not only regulate fundamental cellular processes and signal transduction pathways, but also contribute to tumor cell proliferation, migration, and drug resistance—thereby offering novel opportunities for therapeutic intervention in CRC [6–8].

Neddylation, a distinct form of PTM, involves the covalent attachment of neural precursor cell-expressed developmentally downregulated protein 8 (NEDD8) to substrate proteins. This modification modulates protein stability, DNA repair, and immune responses [9]. In CRC, aberrant activation of components of the neddylation pathway—such as NAE1, UBE2M, and DCUN1D1—is closely associated with tumor progression and therapeutic resistance [10]. The feasibility of targeting neddylation has been demonstrated in various solid tumors. For instance, the NAE1 inhibitor MLN4924 has shown antitumor activity by inducing DNA damage, cell cycle arrest, and apoptosis, and exerts synergistic effects when combined with platinum-based agents [11].

Although neddylation has been extensively investigated in multiple cancer types, its role in CRC remains underexplored. In particular, the involvement of neddylation-related proteins in regulating tumor proliferation, chemotherapy resistance, and the tumor microenvironment warrants deeper investigation. Elucidating the function of neddylation in CRC is therefore of high clinical relevance—not only for improving patient outcomes but also for advancing precision oncology. This review aims to provide a comprehensive overview of the role of neddylation in CRC pathogenesis, chemoresistance, and immune microenvironment remodeling, while also discussing its translational potential for clinical applications.

Neddylation modification

Molecular mechanisms

The concept of neddylation as a post-translational modification analogous to ubiquitination was first introduced by Tetsu Kamitani in 1997 in the context of cellular responses to external stimuli [12]. NEDD8 (neural precursor cell expressed, developmentally downregulated 8) was initially identified in mouse brain tissue as a gene downregulated during development. Despite its nomenclature, NEDD8 is expressed ubiquitously across various tissues, suggesting its essential role in diverse biological processes [13]. NEDD8 is a relatively small protein, comprising 81 amino acids, with approximately 60% sequence homology to ubiquitin. It is a highly conserved protein predominantly localized in the nucleus [14]. Its C-terminal region contains four glycine residues, among which Gly-75 and Gly-76 are evolutionarily conserved. Gly-76 is critical for forming the covalent linkage between NEDD8 and its substrate proteins, employing a mechanism closely resembling that of ubiquitination [12]. These fundamental structural features provide the molecular basis for the enzymatic cascade illustrated in Fig. 1.

Fig. 1.

The neddylation and deneddylation modification processes. NEDD8 is processed by NEDP1/USP21, activated by the E1 enzyme NAE1, transferred to substrates through E2/E3 ligases, and subsequently removed by CSN or NEDP1 during deneddylation. Created with BioGDP.com [51]

Neddylation is a highly dynamic and reversible multi-enzyme cascade. Initially, the NEDD8 precursor is processed by specific proteases—such as the cysteine protease NEDP1, ubiquitin-specific protease 21 (USP21), and ubiquitin C-terminal hydrolase L3 (UCH-L3)—to expose the C-terminal glycine (Gly-76) [15]. Subsequently, in an ATP- and Mg²⁺-dependent manner, NEDD8 is activated by the E1 activating enzyme complex (NAE1, consisting of APPBP1 and UBA3), forming a high-energy thioester intermediate and releasing pyrophosphate. This intermediate is then transferred to the E2 conjugating enzyme UBE2M (UBC12) or UBE2F via trans-thioester exchange [16]. Finally, as depicted in Fig. 1, an E3 ligase catalyzes the covalent conjugation of NEDD8 to a lysine residue on the target protein, thereby completing one cycle of the modification.

The repertoire of known NEDD8 E3 ligases is relatively limited, comprising approximately 10 identified members, primarily including Cullin family proteins and certain RING domain-containing proteins. Representative E3 ligases include Cullin 1 (ROC1) and its regulator ROC2/SAG [17], mouse double minute 2 (MDM2) [18], casitas B-lineage lymphoma proto-oncogene (c-CBL) [19], SCF FBXO11 [20], ring finger protein 111 (RNF111) [21], inhibitors of apoptosis proteins (IAPs) [22], TFB3, and TRIM40 [23, 24]. After NEDD8 is conjugated to the substrate, the E3 ligase releases the E2 enzyme, enabling subsequent rounds of modification [25–27]. Similar to ubiquitination, neddylation is counterbalanced by a deconjugation system. The principal deneddylating enzymes are the COP9 signalosome (CSN) complex and NEDP1. CSN predominantly removes NEDD8 from Cullin family members, whereas NEDP1 acts on non-Cullin substrates [15, 28]. Together, these activation, conjugation, and deconjugation steps constitute the complete neddylation cycle summarized schematically in Fig. 1.

Importantly, NEDP1 exerts dual enzymatic functions in both processing the NEDD8 precursor and removing NEDD8 from target proteins. Mutations in its catalytic site—for instance, substitution to alanine—completely abolish its deneddylation activity toward non-Cullin substrates [15], emphasizing its pivotal role in maintaining the dynamic equilibrium of the neddylation cycle. As summarized in Table 1, the expanding repertoire of NEDD8 E3 ligases and their substrates underscores the broad involvement of neddylation-mediated post-translational regulation in tumor biology. However, within the specific context of colorectal cancer (CRC), functional evidence to date supports only a subset of E3–substrate axes as having well-defined mechanistic relevance. Accordingly, this review highlights the pathways with the strongest experimental support in CRC, including the Smurf1–p27/PDK1 signaling axis, the β-TrCP1/2–β-catenin cascade, the CRL4–TOP1 regulatory module, and the CRL5–NOXA axis. These constitute the core signaling networks that are systematically analyzed in Sects. 4–6.

Table 1.

Functional implications of E3 Ligase-Mediated modifications on substrate proteins

E3 ligase/E3 complex	Substrate protein	Modification type	Functional consequence	References
	p27	K29-linked ubiquitination	Increases p27 stability, promotes migration	[29]
Smurf1	RRP9(Lys221)	Neddylation	Enhances ribosome biogenesis, promotes proliferation	[30]
	PDK1	Neddylation	Activates PDK1–Akt signaling, supports survival	[31]
	RhoA	Indirect regulation	Increases RhoA expression, promotes migration/invasion	[32]
β-TrCP1 (SCF complex)	β-catenin	Ubiquitination	Proteasomal degradation, Wnt pathway suppression	[33]
β-TrCP2 (SCF complex)	β-catenin	Neddylation	Fast degradation, Wnt pathway suppression	[33]
SIAH-1	β-catenin	Ubiquitination	GSK3β-independent degradation	[34]
MDM2	p53	Neddylation	Inhibits transcriptional activity	[18]
FBXO11 (SCF complex)	p53	Neddylation	Inhibits transcriptional activity	[20]
c-Cbl	EGFR	Neddylation	Alters receptor stability/signaling	[19]
	TGF-β type II receptor	Neddylation	Antagonizes degradation
RNF111	DNA damage substrates	Neddylation	Activates DNA damage-induced ubiquitination	[21]
TRIM40	IKKγ	Neddylation	Promotes NF-κB activation	[23]
Cullin1 CRL complex	IκB	Ubiquitination	Activates NF-κB signaling	[35]
Cullin4–DCAF13 CRL4 complex	TOP1	Ubiquitination	Removes TOP1cc, promotes DNA repair, mediates irinotecan resistance	[36]
Cullin5–UBE2F–RBX2 CRL5 complex	NOXA	Ubiquitination	Degrades pro-apoptotic NOXA, promotes chemoresistance	[37]
UBE2F–ARIH2–CRL5 axis	IL-15RB	Ubiquitination	Impairs NK cell function, promotes immune evasion	[38]

In contrast, other E3 ligases listed in Table 1—such as MDM2, c-CBL, and RNF111—are known to participate in critical oncogenic processes, including the stabilization of oncoproteins and the orchestration of DNA damage responses across various cancer types. Nevertheless, their CRC-specific functional roles remain insufficiently characterized. Therefore, the subsequent sections primarily focus on neddylation-mediated pathways with the most robust evidence in CRC, while the remaining E3–substrate axes are presented as components of broader regulatory networks and as potential directions for future CRC-focused investigations.

NEDD8-related substrates

Neddylation is a post-translational modification mechanism distinct from ubiquitination. It primarily functions by covalently attaching NEDD8 molecules to lysine residues on target proteins, thereby regulating their conformation, enzymatic activity, stability, and subcellular localization—rather than directly directing them for degradation. This modification is widely involved in critical cellular processes, including cell cycle regulation, signal transduction, and the DNA damage response.

The most extensively studied and representative substrates of neddylation are Cullin-RING ubiquitin ligase complexes (CRLs). CRLs constitute the largest family of E3 ligases in the ubiquitin-proteasome system (UPS), comprising more than 600 genes and accounting for approximately 20% of intracellular protein degradation in humans. The core Cullin family members include CUL1, CUL3, CUL4A, CUL4B, CUL5, CUL7, and CUL9. NEDD8 modification of Cullin proteins is essential for the activation of CRL function [39]. In their unmodified state, Cullins bind to the inhibitory factor CAND1 (Cullin-associated and Neddylation-dissociated 1), which prevents CRL activation. Neddylation induces a conformational change in Cullin proteins, displacing CAND1 and thereby facilitating CRL assembly and enhancing their ubiquitin ligase activity [40].

Moreover, NEDD8 modification directly enhances the enzymatic activity of CRLs without affecting Cullin–RBX1 interactions or subcellular localization. In mammals, UBE2M and RBX1 cooperatively neddylate CUL1 to CUL4, whereas UBE2F and RBX2 function together to modify CUL5, forming distinct enzymatic modules [41, 42]. This process is often referred to as “classical neddylation.” In addition to Cullin proteins, a growing number of non-Cullin NEDD8 substrates have been identified in recent years, including tumor suppressor proteins such as p53 and pVHL (von Hippel–Lindau protein) [43, 44], receptor proteins such as EGFR and the type II TGF-β receptor, as well as transcriptional regulators including HIF1α and HIF2α [19, 45]. These substrates play pivotal roles in regulating cell cycle progression, hypoxic response, oxidative stress, and DNA repair, all of which are critical in tumorigenesis and cancer progression [46, 47].

Accumulating evidence has shown that key enzymes in the neddylation pathway are aberrantly activated in various cancers. In hepatocellular carcinoma, lung cancer, and breast cancer, components such as the NEDD8-activating enzyme E1 (NAE1), E2 conjugating enzymes (UBE2M/UBE2F), and E3 ligases (such as RBX1, RBX2, and DCN1) are frequently overexpressed and are closely associated with poor patient prognosis [48–50]. These findings suggest that hyperactivation of the neddylation pathway may represent a key oncogenic event, functioning as a critical mechanism that supports tumor cell proliferation and survival.

Therefore, a deeper understanding of the neddylation pathway’s mechanistic roles in cancer is of great significance. It not only provides a theoretical foundation for cancer biology but also highlights potential therapeutic targets, particularly for CRC treatment.

The double-edged sword effect of neddylation in tumors

Recent studies have highlighted the dual nature of neddylation as a post-translational modification, exhibiting context-dependent pro-tumorigenic or anti-tumorigenic effects across different tumor types. Its biological function is not fixed but varies depending on factors such as tissue of origin, molecular background, and tumor microenvironment.

In the majority of solid tumors, elevated levels of neddylation are generally associated with tumor-promoting activities. For instance, through the modification of Cullin proteins, neddylation activates CRL-type E3 ubiquitin ligase complexes, which facilitate the degradation of cell cycle regulator p27, thereby accelerating G1/S phase transition and promoting tumor cell proliferation. Moreover, this pathway suppresses cellular surveillance mechanisms by inhibiting tumor suppressors such as p53 and PTEN [52]. In KRAS-mutant pancreatic cancer, UBE2F enhances tumor proliferation by mediating the neddylation of CUL5, which activates CRL5 [53]. Beyond proliferation, neddylation plays critical roles in chemoresistance and tumor microenvironment regulation. In breast cancer, neddylation of PTEN weakens its tumor suppressive function, thereby driving cancer progression and resistance to targeted therapies [54]. Intriguingly, in the immune microenvironment, neddylation promotes degradation of IL-15RB via the UBE2F–ARIH2–CRL5 axis, impairing natural killer (NK) cell function and facilitating immune evasion [38]. Furthermore, activation of the Skp2/Slug pathway inhibits E-cadherin expression, enhancing cell migration and invasion. Under hypoxic conditions, neddylation stabilizes HIF-1α to promote angiogenesis, while Cullin1-mediated degradation of IκB activates the NF-κB pathway, inducing secretion of pro-inflammatory cytokines such as IL-6 and TNF-α, thereby reshaping the tumor microenvironment and promoting immune escape [35, 55].

Despite its prominent tumor-promoting roles, neddylation is also essential for maintaining cellular homeostasis. During DNA double-strand break (DSB) repair, neddylation facilitates the recruitment of Cullin4A to DNA damage sites via the COP9 signalosome, thereby preserving genomic stability [56, 57]. In non-solid tumors such as acute lymphoblastic leukemia (ALL), overexpression of the deneddylating enzyme NEDP1 has been shown to attenuate VP-16-induced apoptosis, whereas pharmacological inhibition of NEDP1 restores drug sensitivity [58]. Evidence from liver fibrosis and hepatocellular carcinoma indicates that excessive inhibition of neddylation can lead to mitochondrial dysfunction and aberrant activation of NF-κB–associated kinases, ultimately resulting in acute liver failure [59].

These findings highlight that the impact of neddylation suppression varies across tissues and tumor types, and that overly reducing neddylation disrupts cellular homeostasis, potentially accelerating disease progression. Collectively, current evidence supports the concept that neddylation functions as a “double-edged sword” in cancer biology: excessive activation promotes malignant progression in many solid tumors, whereas excessive inhibition may impair DNA repair capacity and worsen outcomes in certain hematologic malignancies.

Among solid tumors, colorectal cancer (CRC) displays a particularly strong dependence on neddylation-mediated regulatory mechanisms. This dependency arises because core oncogenic drivers and drug-resistance mechanisms in CRC converge on the neddylation/deneddylation axis. Approximately 90% of CRC tumors harbor mutations in the APC/Wnt axis, which disrupt the canonical β-catenin degradation complex [60]. As a result, tumor cells become increasingly reliant on compensatory mechanisms to maintain β-catenin stability. Studies show that this compensation primarily depends on neddylation balance: β-TrCP2 and CSN5 jointly regulate β-catenin turnover, and disruption of their coordinated function leads to sustained Wnt/β-catenin signaling activation [33]. In addition, KRAS/PI3K-driven signaling enhances PDK1 neddylation via Smurf1, maintaining persistent AKT activation. This axis further interacts with the RRP9/RhoA-linked network involved in ribosome biogenesis and migration, amplifying CRC’s reliance on neddylation-dependent signaling.

Neddylation imbalance is also tightly associated with chemoresistance in CRC. CRL4–DCAF13-mediated clearance of TOP1cc contributes to irinotecan resistance; the UBE2F–CRL5–NOXA pathway reduces apoptotic priming and is reinforced by PRDX1 [36, 61]; and elevated UBE2M/DCN1 expression correlates with Wnt pathway reactivation and resistance to 5-FU and oxaliplatin. Altogether, these observations demonstrate that key oncogenic drivers and drug-resistance modules in CRC are mechanistically linked to neddylation dysregulation. This provides clear molecular rationale for the notion that “CRC is particularly dependent on neddylation modification” and establishes a foundation for prioritizing combination strategies involving NAE1 inhibitors with TOP1 inhibitors, platinum-based chemotherapy, or PD-1/PD-L1 blockade, each of which holds substantial translational promise in CRC.

Advances in research on the function of neddylation in CRC

Research on the relevant functions of Smurf1

Structural characteristics of Smurf1 and traditional ubiquitination functions

Based on structural and functional properties, NEDD8 E3 ligases are generally classified into two main categories: HECT-type and RING-type ligases. HECT-type ligases, such as Smurf1, RSP5, and ITCH, possess catalytically active cysteine residues that form transient thioester intermediates with NEDD8 or ubiquitin, enabling the subsequent transfer of these molecules to substrate proteins. In contrast, RING-type ligases, such as RBX1 and RBX2, function primarily as scaffolding proteins that facilitate the direct transfer of NEDD8 from the E2 conjugating enzyme to the substrate without forming covalent intermediates themselves [62, 63].

Smad ubiquitination regulatory factor 1 (Smurf1), a member of the NEDD4 subfamily of HECT E3 ligases, was initially identified in neural progenitor cells where its expression is downregulated during development. It was originally characterized as a negative regulator of the bone morphogenetic protein (BMP) and transforming growth factor-β (TGF-β) signaling pathways in mammals. Early studies predominantly explored its role in ubiquitin-mediated functions, particularly in physiological contexts such as bone homeostasis and viral autophagy [64].

Functionally, Smurf1 interacts with the E2 ubiquitin-conjugating enzyme UbcH7 to promote K29-linked ubiquitination of the cell cycle inhibitor p27. This modification enhances the stability of p27 without targeting it for proteasomal degradation, thereby promoting increased cellular migratory capacity. Silencing of Smurf1 or UbcH7 has been shown to significantly impair cell migration [29]. Additionally, Smurf1 has been reported to modulate the activity of the Wnt/β-catenin signaling pathway by regulating the function of the scaffold protein Axin [65].

The role of Smurf1 neddylation activity in chemotherapy resistance

Recent research has demonstrated that Smurf1, in addition to its well-established ubiquitin ligase activity, also functions as a thioester-forming NEDD8 ligase. It is capable of mediating neddylation of both itself and its substrates, thereby expanding its regulatory functions in tumor biology. Studies have revealed that Smurf1 expression is significantly upregulated in CRC, accompanied by enhanced neddylation activity, which is closely correlated with poor clinical outcomes.

Previous investigations have shown that knockdown or pharmacological inhibition of Smurf1 expression increases apoptosis induced by gefitinib and cisplatin in CRC cells. Furthermore, suppression of Smurf1 exhibits antitumor effects in both cell line-derived xenograft (CDX) models and patient-derived xenograft (PDX) models treated with either drug alone or in combination [66].

Smurf1 regulates ribosome biogenesis and migration through RRP9 and RhoA

With the advancement of research, Smurf1 has been found to play not only a role in drug resistance but also a critical regulatory function in cellular processes. For instance, Du et al. reported that Smurf1 interacts with the ribosome biogenesis-associated protein RRP9 through its E3 ubiquitin ligase activity and mediates NEDD8 modification of RRP9 at lysine residue 221 [30]. RRP9, a core component of the U3 small nucleolar ribonucleoprotein (snoRNP) complex, is essential for pre-rRNA processing and ribosome biosynthesis. Its overexpression is closely linked to the elevated protein synthesis demands of cancer cells [67, 68]. Therefore, by regulating RRP9 neddylation, Smurf1 promotes ribosome biogenesis, which in turn supports the proliferation of colorectal cancer cells.

RhoA (Ras homolog family member A) and epidermal growth factor (EGF) are well-established regulators of tumor cell migration [69, 70]. In breast cancer, studies have shown that activation of transforming growth factor-β1 (TGF-β1) enhances ERK signaling, which subsequently phosphorylates Smurf1. This modification promotes RhoA degradation, leading to cytoskeletal reorganization and epithelial–mesenchymal transition (EMT) [71]. In CRC, however, downregulation of microRNA-145 (miR-145) results in EGF upregulation, which in turn induces Smurf1 expression and its auto-neddylation [72, 73]. Interestingly, unlike in breast cancer, Smurf1 neddylation in CRC does not promote RhoA degradation; rather, RhoA expression is increased.

Further studies have demonstrated that during tumor cell migration, RhoA displays spatially distinct roles within the cell. In the leading edge, its activity is suppressed, whereas in the trailing edge, it is activated. In CRC, elevated RhoA expression may be predominantly associated with its accumulation in the rear of migrating cells, thereby facilitating enhanced cell motility and metastatic potential [32]. These findings suggest that Smurf1-mediated RhoA regulation contributes to CRC metastasis through a mechanism distinct from that observed in other tumor types.

Smurf1 mediates PDK1–Akt axis activation to promote tumor progression

In colorectal cancer, particularly within the KRAS-mutant subtype, Smurf1 is markedly upregulated and strongly associated with poor patient prognosis. Xie [74] et al. demonstrated that Smurf1 forms a thioester intermediate with NEDD8 and its E2 conjugase Ubc12, leading to auto-NEDDylation on multiple lysine residues. This self-NEDDylation enhances Smurf1’s ability to recruit ubiquitin E2 enzymes, thereby significantly promoting its ubiquitin E3 ligase activity and contributing to tumorigenesis. Building upon these findings, recent studies further reveal that activated Smurf1 transfers NEDD8 from Ubc12 to Lys163 of PDK1 through its HECT domain, inducing PDK1 NEDDylation. NEDDylated PDK1 then recruits SETDB1 [31], allowing the formation of a cytoplasmic cCOMPASS complex with Smurf1. This complex facilitates Akt membrane localization and Thr308 phosphorylation, robustly amplifying PI3K–PDK1–Akt signaling and sustaining tumor cell proliferation and survival. Consistently, genetic loss of Smurf1 or pharmacological inhibition of its E3 activity markedly reduces PDK1 NEDDylation and Akt activation, resulting in suppressed tumor formation in colorectal cancer mouse models. These observations establish the Smurf1–PDK1 NEDDylation axis as a critical oncogenic pathway in KRAS-driven colorectal carcinogenesis.

In summary, Smurf1, an atypical E3 ligase with dual ubiquitin and NEDD8 ligase activity, participates in crucial oncogenic processes—including cell migration, signal transduction, and protein biosynthesis—by regulating substrates such as RhoA, PDK1, and RRP9. Its sustained upregulation in CRC not only contributes to tumor progression and chemoresistance but also underscores the pivotal role of the neddylation pathway in shaping the malignant phenotype. Therefore, pharmacological inhibition of Smurf1 and its associated neddylation activity may offer a novel strategy for precision therapy, particularly in aggressive CRC subtypes such as those harboring KRAS mutations, with promising potential for clinical translation (Fig. 2).

Fig. 2.

The multifaceted regulatory roles of Smurf1 in colorectal cancer. Upon NEDD8 conjugation, Smurf1 is activated as a HECT-type E3 ligase that mediates the neddylation of multiple substrates, including PDK1 at Lys163 in the PI3K–PDK1–Akt pathway, RRP9 at Lys221 in ribosome biogenesis, and factors involved in RhoA-dependent cell migration, thereby coordinating proliferative and migratory signaling in colorectal cancer cells. Created with BioGDP.com [51]

Neddylation regulation of the Wnt/β-catenin pathway and its carcinogenic mechanism

The key role of the Wnt/β-catenin pathway in CRC

The Wnt/β-catenin signaling pathway plays a critical role in embryonic development, stem cell maintenance, and tissue homeostasis. Its aberrant activation is closely associated with the pathogenesis of various solid tumors, particularly CRC [75]. Under physiological conditions, β-catenin is continuously degraded, and its stability is tightly regulated by the destruction complex, which includes Axin, adenomatous polyposis coli (APC), glycogen synthase kinase 3 (GSK3), and casein kinase 1 (CK1) [76]. In the absence of Wnt ligands, β-catenin is phosphorylated by CK1 and GSK3 within this complex and subsequently recognized by β-TrCP, leading to its ubiquitination and proteasomal degradation [77].

However, upon Wnt activation or inactivation of GSK3β, β-catenin escapes degradation and accumulates in the cytoplasm. It then translocates into the nucleus, where it forms a transcriptional complex with T-cell factor/lymphoid enhancer-binding factor (TCF/LEF) family transcription factors. This complex activates the transcription of downstream oncogenes such as c-MYC and Cyclin D1, ultimately promoting cell proliferation and tumor progression [78].

β-catenin ubiquitination and neddylation regulatory mechanisms

In CRC, the most common mechanism underlying abnormal activation of the Wnt pathway is the loss of destruction complex function caused by mutations or truncations in APC, Axin, or β-catenin itself. This results in the persistent accumulation of β-catenin and is considered a primary early driver of colorectal adenoma formation [79]. As early as 1998, studies in Drosophila revealed that the Slimb protein mediates β-catenin ubiquitination. In mammals, Slimb has two homologs—β-TrCP1 and β-TrCP2—though most studies have not clearly distinguished between them. Recent work by Wang et al. explored the distinct roles of these homologs and demonstrated that β-TrCP1 primarily mediates β-catenin ubiquitination, whereas β-TrCP2 facilitates its neddylation [33].

Beyond the classical phosphorylation-dependent degradation mechanism, β-catenin can also be targeted for degradation through an alternative pathway that is independent of GSK3β. Notably, the RING-type E3 ligase SIAH-1 (Seven in absentia homolog 1) can directly recognize β-catenin and promote its ubiquitination and subsequent proteasomal degradation, thereby serving as a negative regulator of Wnt signaling [80].

CSN5-mediated deneddylation regulatory mechanism and its dual function

CSN5 (COP9 signalosome subunit 5), a key deneddylating enzyme within the neddylation regulatory system, plays a critical role in modulating Wnt pathway activity. It is highly expressed in several solid tumors, including hepatocellular carcinoma and pancreatic cancer [81, 82], and its mRNA expression is significantly elevated in CRC compared to normal colon tissue [83]. Recent studies have revealed that CSN5 regulates β-catenin stability in CRC through a dual mechanism. First, as deneddylating enzyme, CSN5 removes NEDD8 from Cullin proteins, thereby inhibiting the ubiquitin ligase activity of CRLs, reducing β-catenin degradation, and activating the Wnt pathway to promote CRC progression. Second, CSN5 directly interacts with SIAH-1, promoting its degradation and attenuating its inhibitory effect on β-catenin, which further reduces β-catenin turnover [34]. This dual mechanism cooperatively enhances β-catenin stabilization and amplifies Wnt pathway activation.

In summary, accumulating evidence indicates that CSN5 promotes β-catenin stabilization in CRC by both indirectly suppressing CRL activity and directly degrading SIAH-1. Through these mechanisms, CSN5 facilitates Wnt signaling hyperactivation and drives tumor cell proliferation. Given the central oncogenic role of this pathway in CRC, targeting the CSN5–β-catenin regulatory axis presents a promising therapeutic strategy. Future interventions aimed at inhibiting CSN5 or restoring CRL functionality may provide effective avenues for combating Wnt-dependent CRC progression. (Fig. 3)

Fig. 3.

Regulation of the Wnt/β-catenin signaling pathway by neddylation and its carcinogenic mechanisms. In colorectal cancer, neddylation and ubiquitination cooperatively regulate β-catenin stability: in addition to its classical degradation by the APC–Axin–GSK3–CK1 destruction complex, β-TrCP1-mediated ubiquitination, β-TrCP2-mediated neddylation, and CSN5-dependent control of NEDD8-modified components of the pathway together modulate Wnt target gene transcription (such as c-MYC and Cyclin D1) and thereby promote tumor progression. Created with BioGDP.com [51]

The CAND1–RPL34 axis and ribosome biogenesis and malignant progression of CRC

CAND1 regulates CRL activity and neddylation homeostasis

CAND1 (Cullin-associated and neddylation-dissociated 1) is a critical regulator of Cullin-RING ubiquitin ligase (CRL) activity within the neddylation modification system. It inhibits CRL-mediated substrate degradation by binding to un-neddylated Cullin1 and preventing the assembly of SCF (Skp1–Cullin1–F-box)-like E3 complexes, thereby contributing to the maintenance of protein homeostasis and functional equilibrium [84]. In contrast, DCUN1D1 (defective in cullin neddylation 1 domain-containing protein 1) serves as an E3 cofactor within the CRL complex, promoting Cullin neddylation, enhancing CRL catalytic activity, and facilitating the ubiquitination and degradation of substrate proteins [85].

Together, CAND1 and DCUN1D1 function as antagonistic regulators to maintain the dynamic balance of CRL activity and neddylation homeostasis. Disruption of this equilibrium may lead to dysregulated protein degradation, thereby promoting tumorigenesis and cancer progression.

Molecular mechanisms and target value of the CAND1–RPL34 axis

Ribosomal proteins (RPs), traditionally recognized as core structural components of ribosomes, primarily function in ribosomal assembly and protein translation. However, increasing evidence has revealed that certain RPs possess “extraribosomal functions” in tumor cells, including regulation of the cell cycle, apoptosis, and signal transduction [67, 86]. Among these, ribosomal protein L34 (RPL34), a member of the L34E family, exhibits tumor-specific expression patterns and functional diversity across cancer types. For instance, RPL34 exerts tumor-suppressive effects in cervical cancer by modulating the MDM2–p53 axis [87], but functions as an oncogene in non-small cell lung cancer and glioblastoma, where its overexpression is associated with malignancy and poor clinical outcomes [88].

In CRC, RPL34 expression is markedly elevated compared to adjacent normal tissues and negatively correlates with tumor stage and patient survival. Mechanistically, RPL34 has been shown to promote CRC cell proliferation and migration by inducing epithelial–mesenchymal transition (EMT) and activating the JAK2/STAT3 signaling pathway [89]. Moreover, recent studies have identified that CAND1 stabilizes RPL34 protein levels by inhibiting its ubiquitination and degradation, thus forming the “CAND1–RPL34 axis,” which provides a molecular basis for RPL34-driven tumorigenesis in CRC.

These findings suggest that targeting the neddylation-regulatory function of CAND1 and its role in RPL34 stabilization may represent a novel therapeutic avenue for CRC, particularly through intervention in the ribosomal biogenesis and translation pathways.

CRL signal regulation mediated by non-coding RNA

MicroRNAs (miRNAs), a class of non-coding small RNA molecules, play critical post-transcriptional regulatory roles in the development and progression of various tumors [90, 91]. Among them, miR-520b has been characterized as a tumor-suppressive miRNA, shown to inhibit tumor cell proliferation, invasion, and migration in hepatocellular and gastric cancers by targeting oncogenic molecules such as MEKK2, Cyclin D1, and EGFR [92, 93]. In CRC, miR-520b is significantly downregulated and its expression levels negatively correlate with tumor stage and patient prognosis. Notably, DCUN1D1 has been identified as a direct downstream target of miR-520b, and a strong inverse correlation between their expression levels has been observed in CRC tissues [94]. Through downregulation of DCUN1D1, miR-520b can indirectly inhibit Cullin1 neddylation, thereby suppressing CRL activity and the degradation of downstream substrates. These findings suggest that targeting miR-520b or DCUN1D1 may represent a promising strategy for delaying CRC progression.

In addition, emerging studies have highlighted a close association between CAND1 activity and radiosensitivity. Specifically, circAFF2, a circular RNA, is highly expressed in CRC patients who respond well to radiotherapy and is positively correlated with favorable clinical outcomes. Mechanistic investigations have shown that circAFF2 enhances the binding of CAND1 to Cullin1, thereby inhibiting Cullin1 neddylation, reducing CRL activity, and sensitizing tumor cells to radiotherapy [95, 96]. This circAFF2–CAND1–Cullin1 regulatory axis further expands the therapeutic potential of targeting CAND1 in CRC. In addition to these characterized miRNA- and circRNA-mediated mechanisms, emerging studies suggest that other non-coding RNAs, including certain lncRNAs and circRNAs, may also participate in the modulation of neddylation-related components in CRC, although current evidence remains limited and requires further validation.

In summary, CAND1 contributes to CRC development through multiple mechanisms, including the regulation of ribosomal protein stability, oncogenic signaling activation, and modulation of radiosensitivity. Activation of CAND1-related pathway proteins promotes metabolic reprogramming and enhances the migratory capacity of CRC cells, and may also contribute to drug resistance via non-coding RNAs such as circRNAs. Therapeutic interventions targeting key regulators in this pathway—such as miR-520b, DCUN1D1, and circAFF2—may offer novel precision medicine strategies and improve radiotherapy efficacy in CRC.( Fig. 4).

Fig. 4.

The role of the CAND1–RPL34 axis in colorectal cancer and its carcinogenic mechanisms. The schematic shows CAND1 binding un-neddylated Cullin1, DCUN1D1-promoted Cullin1 neddylation and CRL activation, circAFF2-enhanced CAND1–Cullin1 interaction, and RPL34 as a downstream component linked to this axis. Created with BioGDP.com [51]

Cross talk

Recent systematic reviews of NEDDylation modification in colorectal cancer suggest that, although many studies have reported diverse upstream and downstream signaling pathways, only a limited number have been functionally validated, strongly associated with colorectal cancer biological behavior, and shown to exhibit network-level “convergence effects.” These functionally conserved, NEDDylation-dependent pathways primarily include the Smurf1–PDK1–Akt axis, the Wnt/β-catenin pathway, and the CAND1–Cullin regulatory cycle. Driven by highly prevalent driver mutations such as KRAS and APC, these pathways do not operate independently; instead, they converge at an upstream regulatory hub governed by NEDDylation and CRL activity. In KRAS-mutant–dominant subgroups, Smurf1-driven NEDDylation of PDK1 markedly amplifies PI3K–Akt signaling, generating a metabolic and survival dependency on the Smurf1–PDK1–Akt axis. This Akt hyperactivation also indirectly stabilizes β-catenin via the Akt–GSK3β pathway, forming functional coupling and synergistic amplification between Akt signaling and Wnt/β-catenin activity. Conversely, in tumors with APC loss or mutation resulting in constitutive Wnt/β-catenin activation, NEDDylation serves as a key post-translational regulator of CRL-mediated protein homeostasis and DNA damage response. Inhibition of this modification leads to the accumulation of pro-apoptotic and stress-related substrates due to CRL inactivation, producing a strong antitumor effect. At the same time, because CRLs participate in β-catenin turnover, NEDDylation inhibition can further stabilize β-catenin and enhance its transcriptional output, illustrating the characteristic “double-edged sword” nature of the Wnt/β-catenin pathway.

Additionally, CAND1 preferentially binds to unneddylated Cullin, dynamically regulating the assembly, dissociation, and substrate receptor exchange of Cullin–RING ligases. This remodeling of the CRL substrate spectrum affects multiple downstream processes, including DNA damage response, protein quality control, and radiotherapy sensitivity, thereby forming an additional NEDDylation-dependent vulnerability in certain DNA damage repair–dependent or radiotherapy-sensitive colorectal cancer subtypes. Collectively, this integrated framework—shaped by core driver mutations and maintained through both pathway-specific functional dependencies and structural NEDDylation requirements—clarifies the mechanistic connections and hierarchical relationships among the Smurf1–PDK1–Akt, Wnt/β-catenin, and CAND1–Cullin pathways. More importantly, it provides a biological rationale for molecularly stratified therapeutic strategies and outlines a translational direction for the precision development of NEDDylation-targeted interventions in colorectal cancer. To clarify the key pathways described in Sect. 4, the major neddylation-related regulatory axes and their impacts in CRC are summarized in the Table 2 below.

Table 2.

Summary of major neddylation-related pathways and their functional impacts in colorectal cancer

Pathway/regulatory axis	Key neddylation-related components	Core mechanistic insight	Impact in colorectal cancer
Smurf1-mediated regulation	Smurf1, RRP9, RhoA, PDK1–Akt module	Smurf1 modulates ribosome biogenesis, cytoskeletal dynamics, and Akt activation through substrate-specific neddylation	Enhances proliferation, migration, and chemoresistance
Wnt/β-catenin signaling	β-TrCP1/2 (CRLs), SIAH-1, CSN5	Neddylation-dependent CRL activity and CSN5-driven deneddylation jointly regulate β-catenin stability	Sustains stemness, tumor progression, and therapeutic resistance
CRL homeostasis via CAND1–DCUN1D1	CAND1, DCUN1D1, Cullin family	CAND1 and DCUN1D1 antagonistically regulate CRL assembly, disassembly, and cullin neddylation	Drives malignant progression; represents a potential therapeutic vulnerability
Ribosome-biogenesis axis	RPL34, Smurf1/CAND1-regulated CRLs	Neddylation influences ribosomal protein expression and rRNA processing, promoting elevated translational capacity	Supports aggressive proliferation and unfavorable prognosis
lncRNA -mediated modulation of neddylation	miR-520b, lncRNAs targeting CAND1/CRL components	Non-coding RNAs modulate neddylation and CRL signaling by regulating CAND1, DCUN1D1, or cullins	Influences tumor growth, apoptosis, and therapy responsiveness

Neddylation in CRC chemoresistance

Accumulating evidence indicates that the acquisition of drug resistance in colorectal cancer primarily arises from two tightly interconnected biological processes: evasion of apoptosis and remodeling of DNA damage repair (DDR) pathways. Importantly, neddylation functions as an upstream regulatory hub that coordinately governs both processes, thereby exerting a profound influence on tumor cell responses to chemotherapeutic and targeted agents. The suppression of apoptosis and the enhancement of DDR capacity are not independent resistance mechanisms; rather, they represent convergent outcomes driven by neddylation-mediated remodeling of protein stability, signaling dynamics, and stress-response networks. Accordingly, defining neddylation as a major determinant of resistance in colorectal cancer provides a solid mechanistic basis for therapeutic strategies aimed at restoring drug sensitivity.

CRL5–NOXA axis and anti-apoptosis

a first critical step by which neddylation promotes resistance is through attenuation of drug-induced apoptosis via the CRL5–NOXA axis. Neddylation fine-tunes the apoptotic network through multiple downstream effectors, among which the pathway mediated by the CRL5 E3 ligase complex is one of the best characterized mechanisms. As described above, the UBE2F–CRL5 axis facilitates the ubiquitination and subsequent degradation of the pro-apoptotic protein NOXA [37, 61]. Consequently, hyperactivated neddylation accelerates the turnover of key pro-apoptotic mediators, including NOXA, thereby enabling cancer cells to evade cytotoxic insults induced by anticancer agents. Conversely, inhibition of neddylation (for example, with MLN4924) has been shown to restore apoptotic sensitivity by stabilizing NOXA and other pro-apoptotic factors, leading to activation of downstream caspase signaling. Thus, targeting the neddylation–CRL5–NOXA axis represents a central strategy to counteract apoptosis evasion in the context of multidrug resistance (MDR).

CRL4–TOP1 ubiquitin repair pathway

Aberrantly enhanced DNA repair capacity constitutes another major driver of resistance in colorectal cancer, and neddylation orchestrates this process by modulating the activity of specific CRL complexes. A representative example is the CRL4 complex, which is recruited to DNA damage sites through a direct interaction between topoisomerase I (TOP1) and the WD40 domain of its substrate receptor DCAF13 [36]. As noted above, this recruitment facilitates the ubiquitination and clearance of TOP1 cleavage complexes (TOP1cc), which represent the key cytotoxic lesions induced by irinotecan. Elevated neddylation activity allows tumor cells to repair such therapy-induced DNA damage more efficiently, thereby markedly reinforcing resistant phenotypes. Conversely, inhibition of neddylation disrupts CRL-mediated turnover of DDR factors, particularly by preventing effective repair of TOP1cc, and may therefore constitute a promising therapeutic approach to overcoming irinotecan resistance.

In conclusion, the neddylation pathway modulates CRC chemosensitivity by regulating two critical ubiquitination axes: CRL5–NOXA and CRL4–TOP1. Targeted inhibition of key components in these pathways may enhance the therapeutic efficacy of existing agents and offer new avenues for the development of precision therapies in CRC.( Fig. 5).

Fig. 5.

Neddylation activates the CRL5–NOXA and CRL4–TOP1 ubiquitin ligase axes in colorectal cancer cells, promoting NOXA degradation, apoptosis evasion, enhanced TOP1cc repair, and reduced chemosensitivity. Inhibition of neddylation or its downstream CRL components may restore apoptosis, impair TOP1cc repair, and resensitize tumor cells to irinotecan and other chemotherapeutic agents. Created with BioGDP.com [51]

Neddylation regulates the immune microenvironment and therapeutic response

The immunosuppressive tumor microenvironment (TME) is widely recognized as a major barrier to effective antitumor immunity and a key determinant of resistance to immunotherapy [97]. Emerging evidence indicates that neddylation regulates not only the phagocytic activity of innate immune cells, such as macrophages, but also the metabolic fitness and effector function of adaptive immune cells, particularly CD8⁺ T cells. These immunoregulatory processes do not operate in isolation; instead, they form an interconnected neddylation-driven network that promotes immune evasion. In this section, we focus on tumor-associated macrophages and CD8⁺ T cells to delineate how neddylation coordinately shapes an overall immunosuppressive TME in colorectal cancer (CRC).

CD47/SIRPα axis and SHP2 deneddylation

The CD47/SIRPα axis is a prototypical innate immune checkpoint that enables tumor cells to evade macrophage-mediated phagocytosis. CD47 expressed on tumor cells engages signal regulatory protein-α (SIRPα) on macrophages and activates immunoreceptor tyrosine-based inhibitory motifs (ITIMs), leading to the recruitment of the phosphatase SHP2 and subsequent suppression of phagocytic activity [98]. In multiple malignancies, including CRC, CD47 and SIRPα are frequently upregulated, thereby establishing a “don’t eat me” signal barrier that markedly impairs innate immune clearance of tumor cells [99].

Recent studies have revealed that SHP2 activation is tightly regulated by neddylation. Under physiological conditions, SHP2 maintains an autoinhibited conformation through covalent NEDD8 modification of lysine residues K358 and K364, which prevents aberrant phosphatase activation. Upon CD47–SIRPα engagement, tumor cells selectively recruit the deneddylase NEDP1 to remove NEDD8 from SHP2, inducing a conformational switch and activating its phosphatase function. This neddylation-dependent activation of SHP2 ultimately inhibits macrophage phagocytosis [98]. These findings identify neddylation as a molecular switch governing SHP2 activity and suggest that modulation of SHP2 neddylation may serve as a promising strategy to restore macrophage function in CRC. A dual-targeting approach combining CD47/SIRPα blockade with inhibition of SHP2 deneddylation may synergistically enhance the antitumor activity of tumor-associated macrophages (TAMs), thereby improving the generally limited immunotherapeutic responses observed in CRC.

Clinically, microsatellite instability–high (MSI-H) CRC is characterized by a high tumor mutational burden and abundant neoantigen generation, and patients with this subtype derive substantial benefit from PD-1/PD-L1 immune checkpoint inhibitors [100]. By contrast, approximately 75% of CRC cases are microsatellite-stable (MSS), exhibit poor immune cell infiltration, and show minimal responses to current immunotherapies [101]. Recent work has shown that, in MSS CRC, combined PD-L1 blockade and SHP2 inhibition significantly augment antitumor immune responses, further underscoring SHP2 and its neddylation-dependent regulation as attractive immunotherapeutic sensitization targets.

The immunoregulatory function of neddylation also appears to be shaped by the molecular subtype of CRC. In MSI-H tumors, the highly inflamed TME—with increased immune infiltration and abundant neoantigens—renders neddylation particularly important in the regulation of antigen presentation, PD-L1 stability, SHP2 activation, and macrophage–T-cell crosstalk. In this context, neddylation inhibition may not only relieve SHP2-driven suppression of macrophages but also potentiate T-cell–mediated tumor recognition and killing, thereby generating strong synergy with PD-1/PD-L1 blockade. In contrast, MSS CRC, which represents the majority of “immune-cold” tumors, is characterized by limited immune infiltration, inefficient antigen presentation, and constrained T-cell activation. In these tumors, the neddylation pathway more critically supports tumor growth and survival through CRL-mediated control of cell cycle progression, DNA damage repair, and metabolic adaptation. Pharmacological blockade of neddylation in MSS CRC can induce DNA damage accumulation, impair the turnover of substrates such as TOP1 and NOXA, reinstate apoptosis, and enhance chemosensitivity. Thus, in MSS disease, neddylation inhibitors are more likely to exert benefit as sensitizers when combined with chemotherapy, radiotherapy, or immunotherapy.

Collectively, these data indicate that neddylation fulfills distinct functional priorities in MSI-H and MSS CRC: immune regulation and checkpoint control are more prominent in MSI-H tumors, whereas DNA repair and cell-cycle regulation predominate in MSS tumors. This subtype-specific view provides a mechanistic basis for rationally tailoring neddylation-targeted strategies according to the molecular context of CRC.

NAE1 regulation dependent on CD8⁺ T cell metabolism

In addition to macrophage-mediated suppression, neddylation further reinforces tumor-induced immune dysfunction by reshaping the metabolic and functional state of CD8⁺ T cells. As central effector cells of the adaptive immune system, CD8⁺ T cells depend on T-cell receptor (TCR)–driven transcriptional and metabolic reprogramming to exert cytotoxic activity against tumors. Recent studies have identified neddylation as a critical post-translational modification that governs the functional state of CD8⁺ T cells [102]. TCR stimulation activates nuclear factor of activated T cells 1 (NFATc1), which in turn upregulates expression of the neddylation-activating enzyme NAE1. CD8⁺ T cells lacking NAE1 exhibit markedly impaired effector function, including reduced expression of activation markers such as CD44 and granzyme B, increased apoptosis, diminished intratumoral infiltration, and substantially weakened antitumor capacity [103].

Mechanistically, elevated NAE1 expression enhances neddylation-mediated stabilization of multiple proteins involved in energy metabolism and T-cell activation, including components of the mitochondrial import complexes TOMM40/TIM23, tricarboxylic acid (TCA) cycle enzymes, and effector cytokines such as IFN-γ and TNF-α. These factors are indispensable for maintaining the metabolic fitness and sustained cytotoxic activity of CD8⁺ T cells. In murine models, CD8⁺ T cell–specific deletion of NAE1 accelerates tumor growth, whereas NAE1 overexpression enhances antigen-specific CD8⁺ T-cell responses and suppresses CRC progression.

Analyses of clinical CRC specimens further corroborate these experimental findings, revealing a positive correlation between NAE1 expression and CD8⁺ T-cell infiltration. Patients with high NAE1 expression display more favorable clinical outcomes, underscoring the biological significance of NAE1 within the CRC immune microenvironment [104]. These observations suggest that therapeutic strategies designed to preserve or augment NAE1-dependent neddylation in CD8⁺ T cells may strengthen antitumor immunity and help overcome immune tolerance in CRC.

In summary, neddylation shapes a multilayered and self-reinforcing immune escape network by concurrently regulating SHP2-dependent macrophage phagocytic activity and NAE1-dependent metabolic and effector functions of CD8⁺ T cells, thereby establishing two synergistic immunosuppressive axes within the tumor microenvironment. Precision targeting of neddylation-related nodes according to specific cell types and molecular subtypes holds promise for globally reprogramming the tumor microenvironment and enhancing the responsiveness of colorectal cancer to immunotherapy (Fig. 6).

Fig. 6.

Neddylation regulates the immune microenvironment and treatment response in colorectal cancer. The schematic illustrates how neddylation modulates the CD47/SIRPα–SHP2 axis in macrophages and NAE1-dependent neddylation in CD8⁺ T cells, thereby linking NEDD8 signaling to phagocytosis control and T cell metabolic and cytotoxic functions. Created with BioGDP.com [51]

Neddylation-targeted therapeutic strategies

Progress in research on NAE1 inhibitor MLN4924

Antitumor mechanisms and clinical trial progress

MLN4924 (also known as Pevonedistat) was initially developed by Millennium Pharmaceuticals in 2009 through high-throughput screening (HTS) combined with structure-based optimization. It functions as a mimic of the NEDD8-adenylate intermediate and forms a stable complex with the ATP-binding pocket of NAE1, thereby blocking the transfer of NEDD8 to its substrate. This inhibits the activity of Cullin-RING ligases (CRLs), resulting in substrate accumulation and subsequent downstream pathway disruption [105]. Through this mechanism, MLN4924 exerts potent antitumor effects in CRC. However, its global inhibition of CRL activity also suggests a potentially elevated risk of target-related toxicity in normal tissues.

In CRC cells, MLN4924 has been shown to significantly enhance radiation-induced G2/M phase arrest and apoptosis by promoting p27 accumulation, demonstrating notable radiosensitizing properties [106]. Additionally, it induces nuclear translocation of RPL11 and prevents MDM2-mediated degradation of p53, thereby stabilizing p53 activity and activating apoptotic signaling [107]. Notably, studies suggest that MLN4924 demonstrates greater anticancer efficacy in CRC models with wild-type p53, indicating that its therapeutic effects may be, at least in part, p53-dependent.

Mechanistically, MLN4924 triggers both intrinsic and extrinsic apoptotic pathways. It activates the mitochondrial pathway by promoting Bax/Bak-mediated mitochondrial outer membrane permeability (MOMP), while simultaneously engaging the ATF4–CHOP axis to upregulate death receptor TRAIL-R2, thus initiating extrinsic apoptosis [108]. Beyond CRC, MLN4924 exhibits similar antiproliferative effects in oral squamous cell carcinoma, lung cancer, and renal cell carcinoma, reflecting its broad-spectrum anticancer potential [107, 109, 110].

In addition to MLN4924, several NAE1-targeted inhibitors have been developed. TAS4464, discovered by Taiho Pharmaceutical via virtual screening, shows robust and sustained inhibition of the NAE1 complex and has demonstrated efficacy in both solid tumors and hematologic malignancies [111]. Compound X-10, reported by Wang et al., selectively inhibits CUL1/CUL3-mediated neddylation and induces reactive oxygen species (ROS) production to trigger apoptosis [112]. Another compound, C11, acts as a dual inhibitor targeting both the microtubule and neddylation pathways, exhibiting strong antiproliferative activity [113]. Collectively, these agents represent a growing arsenal of NAE1-targeted therapies, underscoring the therapeutic potential of inhibiting the Neddylation pathway in cancer.

Toxic side effects and mechanisms of drug resistance

Although MLN4924 has demonstrated promising antitumor efficacy in preclinical studies, it is not without limitations, including dose-related adverse effects and potential tumorigenic risks. Studies have shown that at low concentrations (0–0.1 µM), MLN4924 can upregulate c-Myc expression, promote cancer stem cell (CSC) self-renewal and proliferation, and increase PD-L1 expression—factors that may contribute to immune evasion and tumor recurrence [114, 115]. These effects are largely dependent on the activation of the MEK/ERK and JNK signaling pathways. Notably, MLN4924-induced PD-L1 upregulation can be reversed by MEK/JNK inhibitors or by knockdown of ERK1/2 and c-JUN, suggesting a modifiable signaling axis with therapeutic potential [116]. In contrast, at higher concentrations (>1 µM), its primary antitumor mechanism is mediated through inhibition of the neddylation pathway. These findings underscore the necessity for precise dosage control and rational combination strategies in clinical applications.

MLN4924 has advanced into Phase I, II, and III clinical trials across multiple cancer types and is currently being evaluated for safety, efficacy, and synergistic potential in combination therapy settings [117]. However, the number of CRC-specific clinical trials remains limited, and the overall objective response rate across solid tumors is only moderate. This suggests that MLN4924 is unlikely to function as a universal monotherapy for all patients; instead, its more realistic role lies in being incorporated into biomarker-guided combination treatment regimens. However, the emergence of intrinsic or acquired resistance has impeded broader clinical adoption. One identified resistance mechanism involves mutations in the ATP-binding pocket of UBA3, a catalytic subunit of the NAE1 complex, which disrupt the formation of the MLN4924–NEDD8 intermediate complex and abrogate drug activity [118]. This observation suggests a feedback protection mechanism within the neddylation pathway and highlights the need for structural optimization of existing compounds or development of alternative targets to overcome resistance. Beyond UBA3 target mutations, context-dependent reprogramming of signaling networks—including DNA damage response (DDR) pathways, WNT/β-catenin signaling, and immune evasion mechanisms—has also been shown to attenuate the sensitivity of CRC models to MLN4924. These findings further highlight the necessity of embedding mechanistic investigations within ongoing and future clinical trials.

In summary, MLN4924, as the first clinically validated inhibitor of the neddylation pathway, represents a significant advancement in molecular targeted therapy for CRC. However, its dose-dependent toxicity and emerging resistance mechanisms necessitate further optimization of dosing regimens, development of structurally refined analogs, and strategic combination therapies to improve clinical translation efficiency and ensure long-term therapeutic efficacy. Therefore, the development of such agents should be accompanied by prospective stratification of CRC patients—based, for example, on MSI status, RAS/RAF mutational profile, baseline DNA repair capacity, and the expression levels of key ubiquitin-like modification components such as NAE1, UBE2M, UBE2F, and DCN1—in order to enrich for tumors that are truly dependent on this pathway. This strategy is expected to maximize the proportion of patients who can derive meaningful benefit from ubiquitin-like modification–targeted therapies.

Development of downstream E2 enzyme targets

Although NAE1 represents the initial and most critical enzymatic step in the neddylation cascade, its central role has raised concerns regarding the consequences of direct inhibition. Emerging evidence suggests that targeting NAE1 may result in irreversible disruption of downstream protein function and pathway integrity, along with clinically relevant adverse effects—such as poor therapeutic response in acute myeloid leukemia, hepatotoxicity, drug resistance, and aberrant glycolysis [117]. These drawbacks present significant challenges to its broader clinical application.

Consequently, increasing attention has turned toward downstream targets, particularly the E2 conjugating enzymes UBE2M and UBE2F, which may provide improved specificity and a more favorable safety profile. Targeting these enzymes could potentially circumvent the limitations associated with upstream NAE1 inhibition, including drug resistance to agents such as MLN4924 [119]. Notably, both UBE2M and UBE2F are frequently overexpressed in multiple human cancers—including CRC—and are strongly correlated with poor prognosis, further supporting their potential as therapeutic targets. Compared with NAE1 inhibitors, E2-directed strategies confer a narrower, yet more precisely controllable, spectrum of regulation over specific CRL subpopulations. This pharmacologic profile renders E2 targeting particularly suitable for incorporation into rationally designed combination regimens or maintenance therapy settings in patients with CRC.

UBE2M-targeted drugs and DCN1 inhibitors

UBE2M, a pivotal E2 enzyme in the neddylation modification pathway, plays a crucial role in the neddylation of CUL1 through CUL4. In CRC, UBE2M is markedly upregulated, and its elevated expression is not only correlated with poor prognosis but also contributes to chemotherapy resistance—particularly to 5-fluorouracil (5-FU) and oxaliplatin—by activating the Wnt/β-catenin signaling pathway [120]. As such, therapeutic strategies targeting UBE2M are emerging as promising alternatives to NAE1 inhibition.

Functionally, UBE2M relies on its interaction with the cofactor DCN1, which, despite lacking a canonical RING domain, facilitates the stable binding of the UBE2M–RBX1 complex within the E3 ligase, thereby promoting the neddylation of Cullin family proteins [121]. Disruption of this UBE2M–DCN1 interaction has thus been identified as a key regulatory approach to modulate CRL1–CRL4 activity, potentially leading to the accumulation of substrates such as NRF2, p21, and p27—proteins known to exert tumor-suppressive effects.

In recent years, several small-molecule and peptide-mimetic inhibitors have been developed to disrupt the UBE2M–DCN1 interaction. For example, Zhou et al. designed the peptide mimetics DI-591 and DI-404 based on the N-terminal 12 amino acids of UBE2M, which effectively block UBE2M–DCN1 binding [122]. Guy et al. identified small molecules such as NAcM-HIT via high-throughput screening of more than 600,000 compounds. Additionally, Zhao et al. developed DC-1 and DC-2—DCN1 inhibitors constructed on a pyrimidine scaffold [123]. These compounds demonstrate high selectivity for inhibiting CUL1- or CUL3-mediated neddylation, while exerting minimal effects on other Cullin family members, indicating excellent target specificity and therapeutic promise. Although still in preclinical development, their favorable pharmacodynamic and pharmacokinetic profiles have laid a solid foundation for future clinical translation. More importantly, the tumor-biased overexpression of UBE2M and DCN1 in CRC and a variety of other solid malignancies suggests that these downstream components may function not only as therapeutic targets but also as predictive biomarkers for identifying patient subgroups most likely to benefit from UBE2M/DCN1-directed interventions. Such an approach holds promise for achieving more tumor-selective inhibition of ubiquitin-like conjugation than that afforded by upstream NAE1 blockade.

The UBE2F-CRL5-NOXA axis and radiosensitization

Compared to UBE2M, research on UBE2F remains relatively limited. However, as another critical E2 enzyme in the neddylation pathway, UBE2F plays an essential role in regulating apoptosis by facilitating the neddylation of CUL5 in conjunction with RBX2. Zhou et al. were the first to report that UBE2F promotes the ubiquitination and degradation of the pro-apoptotic protein NOXA via activation of the CRL5 E3 ligase complex, thereby supporting the survival of lung cancer cells [37]. In CRC, subsequent studies have further clarified this mechanism, demonstrating that PRDX1 interacts with UBE2F and CUL5 to form a PRDX1–UBE2F–CUL5 ternary complex, which enhances CRL5 activity and accelerates NOXA degradation, ultimately inhibiting apoptosis [61].

Notably, this UBE2F-centered regulatory axis is closely linked to radioresistance. Its expression can be upregulated by radiation and other stress stimuli, which increase reactive oxygen species (ROS) levels and promote NOXA degradation, allowing cancer cells to evade apoptosis. In contrast, inhibiting UBE2F prevents NOXA degradation, thereby enhancing the apoptotic response to radiotherapy and suggesting its potential as a radiosensitizer [96]. From a translational standpoint, the activation status of the UBE2F–CRL5–NOXA axis in CRC tumors may serve as a functional biomarker for identifying patient subgroups most likely to benefit from UBE2F inhibition or from intensified radiotherapy and chemotherapy regimens.

Beyond its influence on NOXA, PRDX1 also plays additional regulatory roles. For instance, PRDX1 can function as a molecular chaperone that binds to CUL3, suppressing its ability to ubiquitinate and degrade nuclear factor erythroid 2–related factor 2 (NRF2). This stabilization facilitates NRF2 nuclear translocation and activation, enhancing the cellular antioxidant response, inhibiting ferroptosis, and promoting CRC progression [124].

Given UBE2F’s role in regulating NOXA stability within the CRL5 complex and its involvement in radioresistance, it represents a promising therapeutic target. Inhibition of PRDX1 or UBE2F has been shown to restore apoptosis, impair CRC progression, and significantly enhance sensitivity to platinum-based chemotherapeutics, thereby offering novel avenues for targeted intervention.

In recent years, HA-9104 has emerged as a novel small-molecule inhibitor specifically targeting UBE2F. It binds directly to UBE2F, reduces its protein levels, and disrupts the CUL5-mediated neddylation cascade, thereby stabilizing NOXA and inducing tumor cell death [125]. Although still in the preclinical stage, HA-9104’s selective inhibition of the CRL5–NOXA axis highlights its potential as a candidate for anticancer therapy and chemoradiotherapy sensitization.

In conclusion, UBE2M and UBE2F, as key E2 enzymes within the neddylation machinery, exhibit distinct but complementary roles in the tumorigenesis and therapy resistance of CRC. Targeting their respective signaling axes—such as UBE2M–DCN1 and UBE2F–CRL5—holds significant therapeutic promise. Continued investigation into their structural biology, functional mechanisms, and druggability is expected to accelerate the transition of related inhibitors from preclinical development to clinical application. Therefore, downstream E2-targeting strategies are more likely to function as synergistic and complementary approaches to upstream NAE1 inhibition rather than as simple replacements. They hold considerable potential for incorporation into multi-layered combination regimens within biomarker-stratified therapeutic strategies for colorectal cancer.

Combined therapy with neddylation inhibitors

As the diverse and pivotal roles of the neddylation pathway in the initiation, progression, and drug resistance of CRC are increasingly elucidated, related inhibitors—particularly MLN4924—have shown notable antitumor activity. However, the efficacy of monotherapy is often constrained by tumor heterogeneity and adaptive resistance mechanisms, underscoring the need for combination treatment strategies. Combining neddylation inhibitors with chemotherapeutic agents or immune checkpoint inhibitors is emerging as a promising approach to enhance therapeutic responses and overcome drug resistance in CRC. However, current combination regimens also face substantial challenges, including the optimization of treatment timing, sequencing, and dosing. As noted above, although the combination of MLN4924 with conventional chemotherapy has shown promising efficacy in preclinical models, toxicity observed in clinical settings remains non-negligible. Similarly, combining neddylation inhibitors with immune checkpoint blockade requires careful consideration of their impact on the tumor immune microenvironment, as well as the risk of immune-related adverse events.

Accordingly, future drug development should prioritize the refinement of combination strategies and the design of personalized therapeutic regimens guided by tumor molecular characteristics. Clinical trials ought to incorporate patient stratification based on predictive biomarkers—such as MSI status, RAS mutational profile, and the expression of ubiquitin-like ligases—in order to identify those patient subgroups most likely to derive meaningful benefit.

Combination chemotherapy: cisplatin, oxaliplatin, irinotecan

Platinum-based chemotherapeutic agents, including cisplatin, carboplatin, and oxaliplatin, exert their antitumor effects primarily through the induction of DNA cross-linking and damage. These agents remain central components of standard chemotherapy regimens for CRC. However, their clinical efficacy is often compromised by the ability of cancer cells to develop resistance mechanisms, particularly through enhanced DNA repair capacity and evasion of apoptosis.

Nucleotide excision repair (NER) is the primary pathway responsible for repairing cisplatin-induced DNA lesions, with DNA damage-binding protein 2 (DDB2) acting as a key sensor and regulator of cisplatin sensitivity. Recent studies have demonstrated that MLN4924 inhibits the transcription of DDB2 by suppressing E2F1 activity via CUL4A inactivation. Given that DDB2 is a central determinant of cisplatin sensitivity, these findings suggest that MLN4924 enhances the cytotoxicity of cisplatin by downregulating DDB2 expression, thereby exerting synergistic antitumor effects [126–128].

Moreover, the combination of MLN4924 with cisplatin has shown potent synergistic efficacy in various solid tumors, including head and neck squamous cell carcinoma, pancreatic cancer, prostate cancer, and pleural mesothelioma, supporting its potential application in CRC [128–131]. Phase I clinical trials have also indicated that MLN4924 combined with carboplatin is well tolerated in patients with advanced solid tumors [132].

Progress has also been made in combination strategies involving oxaliplatin. In a CRC model, Zheng et al. found that MLN4924 potentiates oxaliplatin-induced cell cycle arrest and apoptosis by activating the DNA damage response (DDR) and upregulating p-CHK2 expression, thereby sensitizing tumor cells to oxaliplatin [133].

Furthermore, irinotecan exerts antitumor activity by inhibiting topoisomerase I (TOP1), leading to the accumulation of the TOP1–DNA cleavage complex (TOP1cc). The CRL4 complex facilitates the degradation of TOP1cc via ubiquitination, contributing to irinotecan resistance. MLN4924 has been shown to inhibit CRL4 activity, thereby preventing TOP1cc degradation, increasing DNA damage accumulation, and enhancing apoptosis induced by irinotecan. These synergistic effects have been validated in both xenograft and metastasis mouse models [36].

In summary, combining MLN4924 with platinum-based agents or topoisomerase inhibitors represents a promising strategy to overcome drug resistance and enhance the therapeutic efficacy of current treatments for advanced CRC.

Combined immunotherapy: PD-L1/PD-1 inhibitors

Immunotherapy has become a key strategy in the precision treatment of CRC, particularly for patients with mismatch repair deficiency (dMMR) or microsatellite instability-high (MSI-H), where PD-1/PD-L1 checkpoint inhibitors have demonstrated notable therapeutic efficacy. However, more than half of MSI-H patients and the vast majority of microsatellite stable (MSS) patients remain largely unresponsive to immunotherapy [134].

Recent studies have revealed that MLN4924 can induce the accumulation of misfolded proteins and trigger the unfolded protein response (UPR), thereby initiating immunogenic cell death and enhancing tumor immunogenicity. In an MSI-type endometrial cancer model, the combination of MLN4924 with a PD-L1 antibody exhibited significant antitumor effects [100].

Paradoxically, MLN4924 can also upregulate PD-L1 expression by activating the MEK/JNK signaling axis, which enhances PD-L1 transcription and reduces T cell-mediated immune clearance [116, 135]. Therefore, the combination of MLN4924 with PD-1/PD-L1 inhibitors not only enhances tumor immunogenicity but also counteracts immunosuppressive signals, achieving a dual regulatory effect. This mechanism has also been validated in MSS CRC models (such as CT26), significantly expanding the potential applicability of immunotherapy and offering a promising breakthrough for patients who are typically resistant to immune checkpoint blockade [116].

In summary, neddylation inhibitors—especially MLN4924—exhibit broad therapeutic potential in CRC through synergistic applications with both chemotherapy and immunotherapy. Future clinical studies are warranted to further elucidate their mechanisms of action in combination regimens, identify predictive biomarkers, and address safety concerns to facilitate their translation into clinical practice for precision oncology.( Fig. 7).

Fig. 7.

Mechanism of action of NAE1 inhibitors (MLN4924), downstream E2-binding enzyme-related inhibitors, and their combination with platinum-based drugs or the immune checkpoint inhibitor PD-L1. The schematic summarizes the effects of the NAE1 inhibitor MLN4924 on CRL activity, UBE2M/UBE2F-dependent neddylation pathways, and its combination with platinum-based chemotherapeutic agents, topoisomerase I inhibitors, and PD-L1 immune checkpoint blockade. Created with BioGDP.com [51]

NAE1 inhibition and T cell function: resolving the therapeutic paradox

Following the recognition that the NAE1-mediated NEDDylation pathway is essential for maintaining metabolic plasticity and antitumor activity in effector CD8⁺ T cells, the application of MLN4924—a potent NAE1 inhibitor—in immunotherapy appears paradoxical. To address this contradiction, current evidence indicates that MLN4924 forms a reversible NEDD8-adenylate–like adduct at the active site of NAE1, leading to transient accumulation of oncogenic proteins such as Cdt1 and p27, and resulting in DNA damage, cell-cycle arrest, and apoptosis. Importantly, this adduct dissociates after drug withdrawal, allowing restoration of NAE1 activity and explaining the reversible, time- and dose-dependent pharmacological properties of MLN4924 [105]. Although direct evidence is lacking, compensatory activation of enzymes such as NEDP1—a known “eraser” of NEDDylation—is consistent with general biological principles in which cells respond to perturbations of post-translational modification pathways by upregulating antagonistic enzymes [15, 136]. Thus, short-term NAE1 inhibition may transiently impair T-cell cytotoxicity, whereas prolonged or high-intensity inhibition is more likely to disrupt CD8⁺ T-cell metabolism and effector functions.

In parallel, MLN4924 induces robust immunomodulatory effects within tumor cells. It triggers substantial DNA damage and immunogenic cell death (ICD), enhances antigen presentation, increases chemotactic factor expression, and activates the cGAS–STING pathway [137, 138], collectively improving the tumor immune microenvironment and promoting CD8⁺ T-cell infiltration and activation. Although MLN4924 also upregulates PD-L1 expression—potentially limiting monotherapy efficacy—this adaptive response creates a mechanistic window for synergistic combination with PD-1/PD-L1 blockade. In vivo studies have shown that MLN4924 combined with immune checkpoint inhibitors significantly enhances antitumor immunity, with ICD-induced immunogenicity complementing checkpoint blockade–mediated relief of T-cell suppression [116].

Taken together, these findings indicate that NAE1 inhibition exerts a “double-edged effect” rather than a mutually exclusive contradiction with T-cell function. A rational strategy may involve short-course or intermittent NAE1 inhibition to minimize prolonged suppression of CD8⁺ T-cell metabolic and effector activity, combined with a sequential therapeutic approach in which MLN4924 first induces ICD, antigen-presentation enhancement, and PD-L1 upregulation in tumor cells, followed by administration of immune checkpoint inhibitors at the peak of immunogenicity and PD-L1 expression. Moreover, future drug-development efforts targeting selective NEDDylation substrates or specific CRL subtypes may further maximize antitumor efficacy while preserving T-cell function and immune memory. Under optimized timing, dosage, and combination conditions, MLN4924 has the potential not to weaken but to synergistically enhance antitumor immunity, providing a mechanistically grounded rationale for combining NAE1-targeted therapy with immunotherapy.

Challenges and future directions

Feedback resistance driven by UBA3 mutations

UBA3, as the catalytic subunit of the NEDD8-activating enzyme (NAE1), plays a pivotal role in initiating the neddylation cascade. Functional abnormalities in UBA3 directly influence the activity of Cullin-RING ubiquitin ligases (CRLs). MLN4924 inhibits the neddylation pathway by targeting NAE1 and preventing NEDD8 activation. However, mutations in UBA3—particularly within its ATP-binding pocket—can hinder the formation of the MLN4924–NEDD8 complex, resulting in drug resistance [139, 140].

Recent studies have demonstrated that UBA3 is overexpressed in various malignancies and is strongly associated with poor patient prognosis. In lung cancer, UBA3 overexpression promotes the phosphorylation and degradation of activated nuclear factor κB (NF-κB), leading to continuous activation of NF-κB signaling. This suppresses tumor suppressor pathways such as p53 and p21, contributing to chemoresistance [141]. In acute myeloid leukemia (AML) models, deletion of UBA3 reduces the neddylation of Cullin proteins and upregulates the expression of p53 and p21, thereby enhancing the efficacy of chemotherapeutic agents [142]. These findings suggest that UBA3 mutations or sustained activation may establish a feedback-driven resistance network by stabilizing oncogenic proteins like MDM2 and perpetuating survival signaling pathways such as NF-κB.

In addition, UBA3 is significantly upregulated in intrahepatic cholangiocarcinoma (ICC), where it facilitates tumor cell proliferation, invasion, and migration through activation of the MAPK signaling pathway [143]. From a therapeutic perspective, UBA3 represents a promising target for overcoming resistance. Designing next-generation selective inhibitors that recognize mutant or aberrantly active forms of UBA3 may help interrupt resistance mechanisms at the molecular level. Moreover, combining UBA3-targeted agents with immune checkpoint inhibitors, such as PD-L1 antibodies, could counteract immune escape caused by UBA3 overactivation and improve antitumor immunity.

In summary, UBA3 plays a multifaceted role in tumor progression and treatment resistance by modulating the tumor microenvironment and multiple signaling pathways. Future therapeutic strategies should prioritize integrated, multi-targeted approaches that focus on UBA3 as a central regulatory node to overcome resistance and enhance clinical outcomes in cancer therapy. Accordingly, it may be useful for future work not only to optimize UBA3-directed inhibitors, but also to explore approaches for detecting UBA3 alterations at an early stage, which could help to identify emerging resistance to neddylation-targeted therapy in CRC patients.

NEDP1-regulated immunity and chemotherapy tolerance

NEDP1, a critical deNeddylation enzyme, is frequently overexpressed in relapsed acute lymphoblastic leukemia (ALL) and has been implicated in promoting resistance to VP-16 (etoposide) by counteracting the effects of neddylation inhibitors. This resistance mechanism primarily involves the stabilization of MDM2, leading to suppression of p53 transcriptional activity and attenuation of chemotherapy-induced apoptotic responses. Furthermore, NEDP1 has been shown to impair antitumor immune responses by modulating the deNeddylation of key immune regulatory molecules, including components of the T cell receptor (TCR)/NF-κB signaling axis and immune checkpoint proteins such as PD-L1 [58, 142]. These findings highlight the critical role of NEDP1 in facilitating both chemoresistance and immune evasion in tumor cells, underscoring its potential as a therapeutic target for simultaneously modulating treatment resistance and antitumor immunity. These observations suggest that future therapeutic designs might consider combining NEDP1 modulation with strategies that reshape the immune microenvironment, and exploring biomarkers that have the potential to capture chemotherapy tolerance and immune escape in a more dynamic manner.

Prospects for joint strategies

UBA3 and NEDP1 mediate resistance to neddylation inhibitors through distinct mechanisms involving enzyme activity-dependent and non-enzymatic functions, respectively. Combination therapeutic strategies targeting both pathways are currently under active investigation. On one hand, recent efforts have focused on designing specific inhibitors that disrupt the interaction between UBA3 and NAE1, aiming to enhance therapeutic selectivity in tumors that are highly dependent on CRL activity. On the other hand, inhibition of NEDP1 has shown promise not only in reversing VP-16 resistance but also in enhancing antitumor immune responses. This is achieved by targeting PD-L1 and augmenting the cytotoxic function of CD8⁺ T cells. Additionally, optimized dosing strategies for MLN4924 have demonstrated that low-dose, continuous administration can effectively suppress compensatory autophagy, improve therapeutic efficacy, and reduce adverse effects, thereby meeting clinical treatment requirements. Future research should prioritize the exploration of how UBA3 mutations affect the structure and activity of the NAE complex, as well as elucidate the cell-type-specific mechanisms of NEDP1 within different immune subsets. These insights will provide new molecular targets and therapeutic strategies to overcome drug resistance and immune evasion in CRC and other malignancies.

Monitoring potential of liquid biopsy in Neddylation-targeted therapy

After elucidating the molecular mechanisms that underlie resistance to neddylation-targeted therapies—such as those driven by UBA3 and NEDP1—enhancing the therapeutic efficacy of this pathway, particularly overcoming resistance to MLN4924, requires early detection and dynamic monitoring of resistance-associated molecular alterations. Achieving this capability represents a critical step toward advancing personalized and precision oncology.

Liquid biopsy technologies, especially the comprehensive molecular profiling of circulating tumor DNA (ctDNA), offer a promising avenue for real-time monitoring. ctDNA, released into the bloodstream by tumor cells, can be obtained through minimally invasive sampling. With continuous advances in sequencing technologies, ctDNA has become widely used in oncology—for example, to track molecular responses and predict clinical outcomes in lung cancer [144]. In colorectal cancer (CRC), ctDNA has demonstrated high sensitivity and specificity for monitoring tumor burden and detecting resistance-associated mutations such as EGFR T790M and BRAF V600E [145]. However, ctDNA predominantly captures mutational events in classical driver genes and does not reflect the dynamic biochemical activity of the neddylation pathway. Therefore, improving the translational value of neddylation-targeted therapies requires liquid biopsy strategies capable of directly or indirectly indicating pathway activity.

Accumulating evidence suggests that exosome-derived neddylation pathway proteins may serve as promising pharmacodynamic biomarkers. Exosomes are abundant in bodily fluids and stably transport proteins and nucleic acids that mirror tumor cell physiology. Their proteomic signatures have been widely applied across cancer types for diagnosis, prognostic evaluation, and treatment monitoring [146]. Although clinical studies investigating exosomal NAE1 or UBE2M in CRC are still lacking, their essential regulatory functions support a reasonable hypothesis: during treatment with MLN4924 (pevonedistat) or emerging UBE2M–DCN1 inhibitors, a sustained decrease in exosomal NAE1 or UBE2M protein/mRNA abundance may reflect effective pathway suppression, whereas a subsequent rebound may indicate reactivation of neddylation signaling and the onset of therapeutic resistance [6, 147].

In addition, neddylation-associated microRNAs represent another class of potential liquid biopsy biomarkers. miR-520b is markedly downregulated in CRC and correlates strongly with tumor progression and unfavorable prognosis. Functional studies demonstrate that miR-520b directly targets DCUN1D1 to inhibit cullin neddylation and malignant phenotypes. Thus, circulating or exosomal levels of miR-520b and DCUN1D1 may provide a measurable readout of the miR-520b–DCUN1D1–neddylation regulatory axis, enabling dynamic assessment of the intensity and durability of neddylation inhibition during therapy.

By integrating ctDNA-based genomic information with exosomal protein/transcript levels of key enzymes (e.g., NAE1, UBE2M) and regulatory non-coding RNA signatures such as the miR-520b–DCUN1D1 axis, it may be possible to establish a multi-layered neddylation-pathway monitoring system. This integrated biomarker framework would capture mutation status, pathway activity, and regulatory network alterations, thereby providing a more comprehensive understanding of the biological effects of neddylation inhibition and the evolution of drug resistance. Such a strategy could offer robust molecular support for patient stratification and treatment response prediction in future clinical studies.

Summary and outlook

In colorectal cancer (CRC), neddylation has been demonstrated to participate in multiple critical aspects of tumorigenesis and progression, including cell proliferation, metabolic reprogramming, DNA damage response, and remodeling of the immune microenvironment. This review focuses on several representative signaling axes and regulatory pathways. For instance, signaling cascades such as Smurf1–RRP9/PDK1 and CAND1–RPL34 are closely associated with sustained activation of tumor cell protein synthesis and KRAS-related signaling. The CRL5–NOXA- and CRL4–TOP1-related pseudo-ubiquitination pathways are strongly linked to the efficacy of platinum-based drugs and topoisomerase I inhibitors and have a substantial impact on chemotherapy sensitivity. In addition, SHP2-mediated CD47/SIRPα signaling and NAE1-dependent regulation of CD8⁺ T-cell function contribute to immune evasion and further remodeling of the immune microenvironment. Collectively, these findings demonstrate that neddylation does not operate as a simple linear pathway, but rather as a pivotal regulatory hub integrating multiple signaling networks and exerting broad effects on CRC progression and therapeutic response.

Drug development targeting neddylation-related mechanisms has achieved notable progress, including NAE1 inhibitors represented by MLN4924 and studies focusing on downstream molecules such as UBE2M/UBE2F, Smurf1, and DCUN1D1. However, multiple challenges persist in clinical translation. The biological consequences of neddylation inhibition are highly dependent on tissue specificity and variations in the tumor microenvironment. While inhibition can weaken the DNA repair capacity and survival of tumor cells, it also affects normal tissues and immune cells, exhibiting a classic double-edged sword effect. Broad inhibition of NAE1 may further encounter resistance during treatment (e.g., UBA3 mutations that impair inhibitor binding), thereby limiting the dosage and duration of therapy. Moreover, molecular subtypes of CRC that reflect dependence on the neddylation pathway have not yet been established, and stable, reliable neddylation-associated biomarkers remain lacking. This hampers accurate clinical identification of patients most likely to benefit from targeted interventions. Existing evidence is derived predominantly from in vitro cellular experiments or animal models and therefore provides only a limited representation of the complexity of the human tumor microenvironment and its immune interactions.

To address these challenges, future research should prioritize comprehensive evaluation of neddylation-related pathways and signaling axes. By integrating multidimensional data from transcriptomics, proteomics, and spatial omics, it may be possible to achieve finer-grained stratification of CRC patients based on characteristic patterns involving Smurf1–RRP9/PDK1, CAND1–RPL34, CRL5–NOXA, CRL4–TOP1, SHP2/CD47–SIRPα, and NAE1/CD8⁺ T-cell pathways. Such stratification would provide a biological foundation for the design and selection of rational combination therapies. Therapeutic strategies should progressively shift from reliance solely on upstream NAE1 inhibition toward combining this approach with precise modulation of key downstream signaling axes. For example, selectively targeting Smurf1, DCUN1D1, or the UBE2F–CRL5–NOXA pathway within tumor cells, while preserving or enhancing beneficial neddylation signaling in CD8⁺ T cells and optimizing dosing regimens, may mitigate immunosuppressive feedback and treatment-related toxicity, thereby enhancing antitumor efficacy while reducing adverse effects.

Dynamic monitoring represents another critical direction for the development of neddylation-targeted strategies. Liquid biopsy technologies based on circulating tumor DNA (ctDNA), exosomes, and non-coding RNAs hold promise for establishing composite biomarker systems. Such systems can track changes in molecules such as UBA3, NAE1, and UBE2M, detect the expression of pathway-related proteins in exosomes, and assess the activity of regulatory axes such as miR-520b–DCUN1D1. If validated in clinical studies, these indicators could signal emerging resistance to MLN4924 or other pathway inhibitors, inform adjustments to combination regimens and dosing strategies, and support earlier, more personalized interventions. In summary, building a research framework that integrates molecular profiling, rational target combination, and dynamic monitoring, based on the molecular mechanisms outlined in this review, may represent a crucial direction for advancing neddylation from fundamental mechanistic research toward precision treatment for colorectal cancer.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

CM: writing—original draft preparation. FC: writing—review and editing. SYW: writing—review and editing. YYD: supervision. YZL: visualization. HTY: project administration and funding acquisition. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Nature Science Foundation of Gansu Province (Grant No.24JRRA3221), the “Cuiying Science and Technology Innovation” program (Grant No. CY2024-MS-B04), and the Scientific Research Project of Traditional Chinese Medicine of Gansu Province (Grant No. GZKZ-2024-26).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.