KIF23 in disease pathogenesis and its therapeutic and diagnostic potential

Abstract

Background

Kinesin family member 23 (KIF23) is a microtubule-dependent motor protein essential for cytokinesis, organelle transport, and signaling pathway regulation. Its dysregulation contributes to both tumorigenesis and non-malignant disorders; however, a comprehensive review integrating recent mechanistic and translational insights is currently lacking.

Methods

A literature search across PubMed, Web of Science, Embase, and public databases (such as TCGA), using keywords including “KIF23,” “MKLP1,” and “cytokinesis” was performed. Published bioinformatic findings, including pan cancer screening and machine learning analyses, linking KIF23 to disease pathogenesis were summarized.

Results

KIF23 is frequently upregulated in various cancers, such as colorectal, gastric, hepatocellular and breast cancer, where it activates key oncogenic pathways including Wnt/β-catenin, PI3K–Akt and NF-κB. It remodels the tumor immune microenvironment and correlates with poor prognosis. In contrast, loss-of-function mutations in KIF23 underlie several non-neoplastic diseases, such as congenital dyserythropoietic anemia type III and primary microcephaly, by causing cytokinesis failure and developmental defects. KIF23 expression is regulated through multilayered networks involving transcriptional, epigenetic and competing endogenous RNA (ceRNA) mechanisms. Preclinical studies underscore its potential as a diagnostic biomarker and a promising therapeutic target.

Conclusions

KIF23 plays a context-dependent, dual role in disease pathogenesis and represents a compelling target for precision medicine. Future research should focus on deciphering the functional heterogeneity of its splice variants, developing tumor-selective inhibitors and validating integrated biomarker panels to advance clinical translation.

Graphical Abstract

Introduction

Microtubule-dependent motor proteins are core “molecular machines” that regulate intracellular cargo transport, organelle positioning and cell division, and are essential for maintaining cellular homeostasis and organismal development [1]. As a member of this superfamily, kinesin family member 23 (KIF23, also known as MKLP1 or CHO1) was first identified in 1995. Its functional repertoire has since expanded from the regulation of cytokinesis to the control of gene expression, signal transduction, tumor immune microenvironment remodeling [2, 3] and the progression of various non-neoplastic diseases [4, 5].

At the cellular level, KIF23 is recruited to the central spindle during late mitosis, where it forms a central spindle complex with CYK4. By mediating antiparallel microtubule bundling and contractile ring assembly, it ensures successful completion of cytokinesis [6]. Loss of KIF23 function results in cytokinesis failure, multinucleated cell formation and genome instability, which represent early events in tumorigenesis [7, 8]. The activity and stability of KIF23 are further fine-tuned by multiple post-translational modifications, including phosphorylation by Cdk1/cyclin B and NDR/LATS kinases, desuccinylation by SIRT7, and K63-linked ubiquitination induced by TRAF6 [9–12]. These modifications are tightly linked to the regulatory mechanisms that control midbody protein stability, as evidenced in [13].

Clinical studies consistently link KIF23 to diverse human diseases. A systematic pan-cancer analysis revealed frequent KIF23 upregulation in at least 15 cancer types [14]. This aberrant expression correlates with aggressive clinicopathological features, including advanced stage and lymph node metastasis, and predicts poor overall survival in multiple cohorts, such as colorectal [2] and gastric cancer [3]. Mechanistically, KIF23 drives oncogenesis through a multi-faceted program: it potently activates the Wnt/β-catenin signaling pathway [2], reprograms the tumor immune microenvironment to foster immunosuppression [3], and induces cell cycle dysregulation by subverting p53-mediated transcriptional control [15, 16]. Loss-of-function mutations in KIF23, such as p.P916R, is well-established pathogenic factors for congenital dyserythropoietic anemia type III and primary microcephaly [4, 5]. In addition, aberrant KIF23 expression is associated with diseases such as asthma, myocardial infarction and perioperative neurocognitive disorders [17–19]. In malignant pleural mesothelioma, synovial sarcoma and other tumors, high KIF23 expression is also closely related to tumor progression and poor prognosis [20, 21].

Despite these advances, critical and unresolved knowledge gaps remain in this field. First, the disease-specific axis linking “upstream triggers–regulatory networks–downstream pathological effects” involving KIF23 has not been systematically elucidated across different disease contexts. Second, the functional heterogeneity and disease-selective roles of KIF23 splice variants, particularly in cancers such as hepatocellular carcinoma, remain poorly characterized and underexplored [22]. Third, the clinical translation of KIF23 as a diagnostic biomarker and therapeutic target is substantially constrained by the lack of large-scale clinical validation and the absence of highly specific and clinically viable KIF23-directed inhibitors.

In the context of precision medicine and targeted drug development, an integrative and critical synthesis of current evidence is urgently needed. To address these gaps, this review summarizes recent mechanistic and clinical evidence retrieved from PubMed, Web of Science, Embase, and public datasets such as TCGA, using keywords including “KIF23,” “MKLP1,” “cytokinesis,” and related terms. Several published integrative bioinformatics studies, including a machine learning analysis of cancer secretory pathways, have independently identified KIF23 as a high priority candidate, suggesting its potential involvement in extracellular communication and microenvironment remodeling [23]. Specifically, this review aims to (i) delineate disease and context specific KIF23 regulatory networks to support biomarker guided patient stratification; (ii) evaluate the druggability and isoform specific therapeutic potential of KIF23; and (iii) identify translational bottlenecks and design principles for the development of KIF23 targeted diagnostics and therapeutics. Finally, building upon the synthesized evidence, prioritized research directions for the next three years are outlined to bridge existing knowledge gaps and advance KIF23 toward clinical application.

Molecular structure and regulatory mechanisms of KIF23

The functional diversity of KIF23 arises from its unique molecular structure and complex regulatory network. Dysregulated expression or abnormal post-translational modification of KIF23 is a key event that disrupts cellular homeostasis and drives disease progression. Understanding its structure–function relationships and regulatory logic is fundamental for the development of targeted strategies. The molecular structure, core functions and multilayered regulatory mechanisms of KIF23 are discussed.

Structural features

The human KIF23 gene is located on chromosome 15q23, a region frequently amplified in non-small cell lung cancer, which constitutes an important genetic basis for KIF23 overexpression [24]. Its transcript is processed to yield a protein of 917 amino acids with a molecular weight of approximately 102 kDa. These domains combine the conserved features of the kinesin family with functional specificity as depicted in Table 1 and Fig. 1.

Table 1.

Core domains of KIF23 protein and their functional features.

Domain type	Key components/sites	Core functions	Consequences of dysfunction	References
N-terminal motor domain	ATP-binding site; microtubule-binding motifs	① Drives conformational changes via ATP hydrolysis to achieve directional movement along microtubules; ② Maintains spindle stability during mitosis	Point mutations such as G194E lead to cytokinesis failure	[6]
Central coiled-coil domain	Dimerization interface mediated by hydrophobic contacts	① Forms homodimers to provide structural basis for motor function; ② Regulates interactions with downstream effector molecules through conformational changes	Dimerization defects disrupt central spindle complex assembly	[6]
C-terminal cargo-binding domain	CYK4/RACGAP1-binding sites	① Forms the central spindle complex and mediates antiparallel microtubule bundling; ② Precisely regulates the timing of cytokinesis onset	Binding site mutations result in multinucleated cell formation	[6, 25]
Post-translational modification sites	Ser716 (phosphorylation); Lys537 (desuccinylation); Lys119/Lys156 (ubiquitination)	① Ser716 phosphorylation regulates the timing of microtubule binding; ② Lys537 desuccinylation enhances protein stability; ③ Lys119/Lys156 ubiquitination promotes autophagic degradation	Aberrant modifications result in cell cycle dysregulation and increased genome instability	[10– [12, 22]

Fig. 1.

Domain architecture of KIF23 and representative disease-/regulation-relevant sites. KIF23 comprises an N-terminal motor domain with an ATP-binding site, a central coiled-coil region mediating dimerization, and a C-terminal cargo-binding domain containing the CYK4 (RACGAP1) interaction interface [6, 25]. Representative regulatory sites include Ser716 (phosphorylation) [12], Lys537 (SIRT7-associated desuccinylation) [10], and Lys119/Lys156 (TRAF6-associated ubiquitination) [11]. Disease-associated alterations highlighted here include G194E within the motor region (associate to impaired ATPase activity and cytokinesis failure) and p.P916R in the C-terminal region (linked to midbody instability and congenital dyserythropoietic anemia type III), whereas KIF23 deficiency has been implicated in neurodevelopmental phenotypes such as microcephaly [4–6].

The N-terminal motor domain contains the ATP-binding site and microtubule-binding motifs, and drives directional movement along microtubules via ATP hydrolysis. Point mutations such as G194E can abolish its ATPase activity and microtubule binding, ultimately leading to cytokinesis failure [6]. The central coiled-coil domain mediates homodimer formation through hydrophobic interactions, providing structural support for motor function, and its dynamic conformational changes modulate interactions with downstream effector molecules [6]. The C-terminal cargo-binding domain specifically binds CYK4 and RACGAP1 to form the central spindle complex, thereby regulating the timing of cytokinesis onset [6, 25].

Recent studies have revealed the functional significance of KIF23 splice variants, exemplified by the contrasting prognostic associations of nuclear V1 and cytoplasmic V2 isoforms in hepatocellular carcinoma. Specifically, in hepatocellular carcinoma, KIF23 exists as two major splice variants: V1 (nuclear localization) and V2 (cytoplasmic localization). The V2 isoform enhances the stability of β-catenin in the cytoplasm and promotes its nuclear translocation, thereby activating the Wnt pathway and driving tumor proliferation; patients with high V2 expression have a median survival that is 14 months shorter than that of those with low V2 expression. In contrast, the V1 isoform localizes to the nucleus and binds p53 to inhibit its transcriptional activity, and is inversely correlated with tumor progression [22, 26]. The differences in subcellular localization and functional heterogeneity between these splice variants provide new molecular markers for precise subtyping and targeted therapy in liver cancer. In addition, the network of post-translational modifications continues to expand: in anaplastic thyroid cancer, SIRT7-mediated desuccinylation of Lys537 enhances KIF23 protein stability [10], whereas TRAF6-mediated K63-linked ubiquitination regulates its protein homeostasis through autophagy [11]. Together with phosphorylation events, these modifications constitute a dynamic regulatory code.

Core biological functions

The multifaceted roles of KIF23 in cell division, signal transduction, and organelle dynamics underlie its contributions to physiological homeostasis and pathological processes, as detailed in the following discussion.

Regulation of cell division

KIF23 is a key regulator in cell division. During metaphase and anaphase, KIF23 forms a complex with CYK4 at the central spindle, where it mediates the bundling of antiparallel microtubules and maintains spindle stability. The interaction with CYK4 induces conformational rearrangements in KIF23, thereby enhancing microtubule bundling efficiency. As mitosis proceeds, the complex accumulates at the spindle midzone and collaborates with the RhoA signaling pathway to promote contractile ring assembly, providing the mechanical force required for cytokinesis [6, 27]. Loss of KIF23 function results in multinucleation and genome instability. For example, v-Src activation disrupts Aurora B-mediated proper localization of KIF23, thereby inhibiting cytokinesis [8, 28].

Recent studies further demonstrate that KIF23 cooperates with KIF18A to regulate central spindle assembly [25]. Aurora B-mediated phosphorylation of KIF23 delays the breakage of chromatin bridges [8] in colorectal cancer, missense mutations in KIF23 and CENPE synergistically reduce chromosome segregation fidelity [29] and dynamic balance between dephosphorylation by the PP1β-MYPT1 phosphatase and phosphorylation by Aurora B kinase ensures precise timing of cell division [27].

Modulation of signaling pathways

The Wnt/β‑catenin signaling pathway is closely associated with the regulation of genes involved in cell proliferation, differentiation, and tumor progression, and emerging evidence suggests that KIF23 (Kinesin Family Member 23) may function as a downstream target or functional mediator of this pathway. KIF23 can directly bind the ARM domain of β-catenin, promote its nuclear translocation, and activate the transcription of downstream target genes such as cyclin D1 and c-Myc, thereby driving cell proliferation [2, 30]. In colorectal cancer, N-acetyltransferase 10 (NAT10) introduces N4-acetylcytidine (ac4C) modifications on KIF23 mRNA, markedly enhancing its stability and translational efficiency and providing a molecular basis for its overexpression [31]. In triple-negative breast cancer, Forkhead Box M1 (FOXM1) drives KIF23 transcription by increasing histone H3 lysine 4 trimethylation (H3K4me3) at its promoter region [32]. In esophageal squamous cell carcinoma and clear cell renal cell carcinoma, KIF23 likewise promotes tumor progression by activating this pathway [33, 34]. In gastric cancer, KIF23 competitively binds Amer1 and releases its negative control over the Wnt pathway [35].

KIF23 expression is positively correlated with the activity of the PI3K–Akt/mTOR pathway. In nasopharyngeal carcinoma, the curcumin derivative CVB-D downregulates KIF23 and inhibits Akt/mTOR signaling [36]. In myocardial infarction models, KIF23 drives cardiac fibroblast proliferation via RhoA and suppresses Ces1d expression and function through the RhoA/ROCK1 axis [18]. In chronic myeloid leukemia, KIF23 participates in imatinib resistance through regulation of the PI3K–Akt/mTOR pathway [37]. In idiopathic pulmonary arterial hypertension, KIF23 modulates the PI3K/AKT and MAPK pathways to affect pulmonary artery smooth muscle cell function [38]. In cervical cancer, KIF23 promotes tumor progression by inhibiting NLRP3 inflammasome-mediated pyroptosis [39].

Organelle transport and immune homeostasis

KIF23 binds specific receptors on the surface of the Golgi apparatus and peroxisomes to participate in directed organelle transport. Its transport rate for peroxisomes (~ 0.8 μm/s) is significantly higher than that for the Golgi (~ 0.3 μm/s) [1]. In immune regulation, high KIF23 expression in gastric cancer is associated with increased infiltration of regulatory T cells and M1-type macrophages as well as upregulation of immune checkpoint molecules such as programmed death-ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated protein 4 (CTLA4). By activating the NF-κB pathway, KIF23 promotes PD-L1 transcription and suppresses the anti-tumor function of cytotoxic T cells [3]. In head and neck squamous cell carcinoma, KIF23 interacts with SHC binding protein 1 (SHCBP1) to activate pro-survival signaling pathways and regulate immune checkpoint expression [40]. In lung adenocarcinoma, KIF23 acts as a phagocytosis regulator influencing tumor-associated macrophage function [41]. Notably, KIF23 is not alone in this regard; other secretory or membrane‑associated proteins have also been implicated in tumor immune modulation. For example, ZG16B, a secretory lectin‑like glycoprotein, has been shown to shape the tumor immune microenvironment by enhancing the immunosuppressive functions of myeloid‑derived suppressor cells and M2 macrophages, while also promoting dendritic cell maturation [42].

In patients with systemic lupus erythematosus complicated by diffuse large B-cell lymphoma, Epstein–Barr virus (EBV) infection modulates hub genes including KIF23, thereby mediating immune cell infiltration and immune dysregulation [43]. In line with this, a recent transcriptomic study in rectal cancer identified KIF23 as a core stemness‑associated gene. The KIF23‑containing module was significantly correlated with G2/M checkpoint activation and alterations in the tumor immune microenvironment, further supporting its role in immune modulation [44].

Regulation of cell cycle and genome stability

KIF23 is a key regulator of the G2/M cell cycle transition. In normal cells, p53 downregulates KIF23 expression via the p21–DREAM–CDE/CHR pathway to maintain cell cycle arrest and genome stability; when p53 function is lost, dysregulated KIF23 expression leads to aberrant cell division [15, 16]. In colorectal cancer, missense variants in KIF23 and CENPE synergistically reduce chromosome segregation efficiency [29]. In pre-malignant stages of ovarian cancer, KIF23 overexpression triggers replication stress and cytokinesis failure [45]. In lung adenocarcinoma, KIF23 maintains cancer stemness by regulating the cell cycle [46]. In malignant peripheral nerve sheath tumors, tumor cells are dependent on KIF23 for survival [47].

Taken together, KIF23 plays a dual role: it safeguards genome integrity and intracellular homeostasis under physiological conditions, but its dysregulation converts it into a key driver of genomic instability, aberrant signaling, and immune evasion.

Upstream regulatory mechanisms

Transcriptional regulation

Transcriptional control of KIF23 is a highly refined and context-dependent process involving the coordinated actions of multiple transcription factors, epigenetic mechanisms and non-coding RNA networks as depicted in Table 2.

Table 2.

Key transcriptional regulators of KIF23 and their disease-specific roles.

Transcription factor	Regulatory mechanism	Disease type	Biological effect	References
FOXM1	Recruits RNA polymerase II and increases H3K27ac at the promoter	Hepatocellular carcinoma	Upregulates KIF23, enhances cell proliferation and sorafenib resistance	[48]
AR	Binds promoter AREs to drive transcription	Nasopharyngeal carcinoma	Thr645-phosphorylated AR enhances interaction with Akt, amplifying KIF23-mediated effects and activating Wnt/β-catenin	[49]
ETV5	Binds EBS in the promoter to inhibit transcription	Perioperative neurocognitive disorders	Reduced ETV5 expression relieves repression of KIF23, aggravating caspase-3/GSDME-mediated neuronal pyroptosis	[50]
TAZ–TEAD complex	Directly binds the KIF23 promoter region	Hepatocellular carcinoma	Drives KIF23 transcription, promoting cytokinesis and proliferation	[51]
WDR5	Catalyzes H3K4me3 and regulates FOXM1 transcriptional activity	Triple-negative breast cancer	Indirectly upregulates KIF23, activates Wnt/β-catenin, promotes epithelial–mesenchymal transition	[32]
TCF-4	Directly binds the KIF23 promoter	Glioblastoma	Transcriptionally activates KIF23, promoting tumor cell proliferation	[53]

In liver cancer, FOXM1 recruits RNA polymerase II and increases H3K27ac at the KIF23 promoter region, thereby upregulating its expression [48]. In nasopharyngeal carcinoma, the androgen receptor (AR) directly binds androgen response elements (AREs) in the KIF23 promoter to drive transcription, and Thr645 phosphorylation enhances its interaction with Akt [49]. In models of perioperative neurocognitive disorders, ETS variant transcription factor 5 (ETV5) binds ETV5 binding sites (EBS) in the KIF23 promoter to repress transcription, and decreased ETV5 expression relieves this repression [50]. In addition, the TAZ–TEAD complex directly binds the KIF23 promoter to promote its transcription in liver cancer [51] WD repeat domain 5 (WDR5) regulates FOXM1 transcriptional activity via catalyzing H3K4me3 in triple-negative breast cancer, thus indirectly driving KIF23 expression [32]. T-cell factor 4 (TCF-4) directly binds the KIF23 promoter and activates its transcription in glioblastoma [52].

Epigenetic regulation

At the DNA methylation level, hypomethylation of the KIF23 promoter in triple-negative breast cancer releases transcriptional repression [54]. In ovarian cancer, hypermethylation of the miR-424/503 cluster promoter leads to its silencing and indirectly results in KIF23 upregulation [55]. In diffuse large B-cell lymphoma, hypomethylation of the KIF23 promoter is a major mechanism underlying its overexpression [56]. In vitiligo, DNA hypomethylation of KIF23 in perilesional skin precedes changes in transcript levels [57]. In esophageal squamous cell carcinoma, lactate regulates the expression of hub genes including KIF23 and drives lactylation of epithelial cells [58].

Regarding histone and RNA modifications, FOXM1 recruits histone acetyltransferase p300 in liver cancer, increasing H3K27ac at the KIF23 promoter region [48]. In colorectal cancer, NAT10-mediated ac4C modification of KIF23 mRNA significantly enhances its translational efficiency [31].

Regulation by non-coding RNAs

CeRNA networks represent a major pathway regulating KIF23 expression, wherein various non-coding RNAs modulate its levels through specific pairing interactions as shown in Table 3.

Table 3.

Key ceRNA networks regulating KIF23 and their disease associations.

Non-coding RNA type	Regulatory molecule	Interaction mode	Cancer type	Biological effect	References
lncRNA	UCA1	Acts as a ceRNA to sponge miR-135a/143, relieving repression of KIF23	Colorectal cancer	Upregulates KIF23, promoting cell migration and chemoresistance	[59]
circRNA	circ_0067934	Acts as a ceRNA to sponge miR-1301-3p, relieving repression of KIF23	Gastric cancer	Upregulates KIF23, enhancing proliferation, migration and invasion	[60]
lncRNA	PVT1	Acts as a ceRNA to sponge miR-15a-5p, relieving repression of KIF23	Prostate cancer	Upregulates KIF23, suppresses apoptosis, and promotes tumor growth and metastasis	[61]
lncRNA	KCNQ1OT1	Acts as a ceRNA to sponge miR-339-3p, relieving repression of KIF23	Retinoblastoma	Upregulates KIF23, promoting proliferation and migration while inhibiting apoptosis	[62]
lncRNA	LINC00467	Acts as a ceRNA to sponge miR-107, relieving repression of KIF23	Cervical cancer	Upregulates KIF23, driving epithelial–mesenchymal transition and tumor progression	[63]
miRNA	miR-424-5p	Directly targets the 3′UTR of KIF23 mRNA to inhibit its expression	Hepatocellular carcinoma	Downregulates KIF23 and increases tumor cell sensitivity to sorafenib	[68]
miRNA	miR-195-5p	Directly targets the 3′UTR of KIF23 mRNA to inhibit its expression	Triple-negative breast cancer	Downregulates KIF23, suppressing epithelial–mesenchymal transition and metastasis	[66]

Long non- coding RNA (lncRNA) UCA1 sponges miR-135a/143 [59], circularRNA (circRNA) circ_0067934 sponges miR-1301-3p [60] and lncRNA PVT1 sponges miR-15a-5p [61], all of which relieve miRNA-mediated repression of KIF23, leading to its upregulation. In retinoblastoma, lncRNA KCNQ1OT1 upregulates KIF23 by sequestering miR-339-3p [62]. In cervical cancer, LINC00467 functions as a ceRNA for miR-107, similarly lifting repression of KIF23 [63]. In lung adenocarcinoma, lncRNA LINC00337 acts as a “molecular sponge” to sequester miRNAs and relieve their inhibition of KIF23 [64]. In schizophrenia, lncRNA-C2orf48A sponges hsa-miR-20b-5p and hsa-miR-17-5p, thereby releasing their suppression of KIF23 [65].

For microRNAs (miRNAs), multiple species directly target the 3′UTR of KIF23 to inhibit its expression. For example, downregulation of miR-195-5p in triple-negative breast cancer [66], miR-424/503 cluster in ovarian cancer [55], and miR-424 in glioma [67] contributes to aberrant KIF23 overexpression. In liver cancer, miR-424-5p targets KIF23 and increases tumor cell sensitivity to sorafenib [68]. In brain metastases of lung adenocarcinoma, decreased expression of miR-195-5p and miR-195-3p releases KIF23 from repression [69].

The expression of KIF23 is thus precisely controlled by a multilayered regulatory network, and its dysregulation under pathological conditions determines its aberrant expression patterns as depicted in Fig. 2.

Fig. 2.

Multilayered upstream regulatory network controlling KIF23 expression. KIF23 transcription is regulated by transcription factors (such as FOXM1, AR, and ETV5) [48–50] and epigenetic mechanisms including H3K4me3 enrichment [32] and promoter hypomethylation [54]. Post-transcriptionally, ceRNA circuits (such as lncRNA UCA1, circ_0067934, and lncRNA PVT1) promote KIF23 expression by sponging specific miRNAs, thereby shaping disease- and context-dependent KIF23 expression programs [59–61]. Arrows denote activating or inhibitory regulation across modules

The multilayered regulatory network of KIF23 exhibits marked disease specificity: in tumors, KIF23 is often overexpressed due to FOXM1/AR-mediated transcriptional activation, promoter hypomethylation or ceRNA network activation (the specific ceRNA molecules, target miRNAs and disease associations are summarized in Table 3), whereas in non-neoplastic diseases, transcriptional repression by ETV5 or genetic mutations often result in loss of function [50, 54, 56]. These regulatory abnormalities drive disease-specific pathological phenotypes via pathways such as Wnt/β-catenin and PI3K–Akt, forming a core link between molecular regulation of KIF23 and disease mechanisms.

Roles of KIF23 in tumorigenesis

KIF23 is highly expressed in multiple tumor types and plays a central driving role in tumor initiation, progression and drug resistance via activation of oncogenic signaling and remodeling of the tumor microenvironment, thereby presenting therapeutic potential. Below, we discuss its tumor-specific mechanisms in major cancer types in combination with clinical evidence as depicted in Table 4.

Table 4.

Roles and clinical associations of KIF23 in major tumor types.

Tumor type	Core regulatory pathways/molecular interactions	Key biological effects	Clinical associations (prognosis/therapy)	References
Colorectal cancer	(1) NAT10-mediated ac4C modification stabilizes KIF23 mRNA; (2) Drives β-catenin nuclear translocation and activates Wnt; (3) Cooperates with CENPE mutations to promote genome instability	(1) Enhances proliferation, migration and invasion; (2) Impairs chromosome segregation	(1) High expression correlates with advanced TNM stage, lymph node metastasis and reduced 5-year survival; (2) Remodelin reduces liver metastases by 50% in PDX models; (3) Associated with 5-FU resistance	[2, 29, 31]
Hepatocellular carcinoma	(1) Transcriptionally activated by TAZ–TEAD and FOXM1; (2) DEPDC1B–KIF23 axis suppresses p53; (3) Regulated by miR-424-5p	(1) Promotes cytokinesis and proliferation; (2) Contributes to sorafenib resistance	(1) High expression shortens median survival by 14 months and correlates with microvascular invasion (OR = 2.15); (2) FOXM1 inhibitor FDI-6 reverses resistance; (3) V2 splice variant overexpression predicts poor prognosis	[26, 48, 51, 68]
Triple-negative breast cancer	(1) Downregulation of miR-195-5p releases repression; (2) Transcriptionally activated via WDR5/FOXM1/H3K4me3 axis; (3) KIF23–PRC1 interaction suppresses p53	(1) Promotes EMT and metastasis; (2) Activates Wnt/β-catenin; (3) Enhances pro-survival autophagy	(1) Promoter hypomethylation (~ 40% reduction) serves as early warning biomarker; (2) High expression correlates with metastatic potential and poor prognosis; (3) FOXM1 inhibitors suppress tumor progression	[32, 66, 80]
Gastric cancer	(1) Competitively binds Amer1, impairing Wnt inhibition; (2) Activates NF-κB and upregulates PD-L1; (3) Regulated by circ_0067934/miR-1301-3p axis	(1) Drives G2/M progression and proliferation; (2) Recruits Tregs and M1 macrophages, forming immunosuppressive microenvironment	(1) High diagnostic AUC (training: 0.958; validation: 0.867); (2) High expression associated with intraperitoneal paclitaxel resistance (1.8-fold higher in resistant group); (3) Positively correlated with H. pylori infection	[3, 35, 60]
Nasopharyngeal carcinoma	(1) AR directly binds promoter AREs; (2) Activates Wnt/β-catenin; (3) CVB-D downregulates KIF23 and inhibits Akt/mTOR	(1) Thr645-phosphorylated AR enhances KIF23 effects; (2) Promotes proliferation, migration and invasion; (3) Induces apoptosis	(1) High expression correlates with lymph node metastasis and poor prognosis; (2) CVB-D achieves a 52% tumor growth inhibition rate in vivo; (3) Associated with EBV-mediated pathway dysregulation	[36, 49, 88]

Colorectal cancer

In colorectal cancer, high KIF23 expression is a recognized indicator of poor prognosis [2]. Its overexpression is driven by unique post-transcriptional regulation: NAT10-mediated ac4C modification enhances KIF23 mRNA stability [31]. Functionally, KIF23 directly binds β-catenin and promotes its nuclear translocation, representing a critical step in sustained activation of the Wnt pathway [2], and cooperates with CENPE mutations to exacerbate genome instability [29]. Overexpression of KIF23 also promotes stemness in cancer stem cells; extracellular vesicles secreted by colon cancer cells that carry KIF23 enhance the invasive capacity of recipient cells [70, 71].

From a translational perspective, the small-molecule inhibitor Remodelin reduces liver metastatic nodules by 50% in colorectal cancer xenograft models [31]. lncRNA UCA1 regulates KIF23 via a ceRNA mechanism, forming a potential therapeutic pathway [59]. Additionally, KIF23 overexpression is associated with resistance to fluoropyrimidine-based chemotherapy [31].

Hepatocellular carcinoma

High KIF23 expression is an independent adverse prognostic factor in hepatocellular carcinoma, and is closely associated with tumor size, vascular invasion and poor overall survival [72]. At the molecular level, the TAZ–TEAD transcriptional complex directly binds the KIF23 promoter and drives its transcription [51]. FOXM1 upregulates KIF23 expression by recruiting RNA polymerase II and increasing H3K27ac at its promoter [48]. In the DEPDC1B–KIF23–p53 axis, DEPDC1B positively regulates KIF23, which in turn suppresses the p53 signaling pathway [26]. In HBV-related hepatocellular carcinoma, RACGAP1 modulates interactions involving KIF23 and other molecules, promoting early tumor recurrence [73].

Therapeutically, the FOXM1 inhibitor FDI-6 reduces KIF23 expression by 45% and increases the tumor growth-inhibitory effect of sorafenib by 2.3-fold when used in combination [48]. miR-424-5p targets KIF23 and enhances tumor cell sensitivity to sorafenib [68], circRNA hsa_circ_0070934 and certain lncRNAs upregulate KIF23 via ceRNA mechanisms, constituting additional potential intervention targets [74, 75].

Breast cancer

KIF23 is aberrantly expressed in different breast cancer subtypes, with distinct underlying mechanisms. In triple-negative breast cancer, its expression is significantly higher than that observed in other subtypes and is associated with invasive/metastatic potential and poor prognosis [66]. In the luminal B subtype, high KIF23 expression affects the cell cycle and correlates with unfavorable outcomes [76]. Loss of p120 expression causes mislocalization of KIF23, resulting in cytokinesis failure, multinucleation and chromosomal instability [77].

Mechanistically, downregulation of miR-195-5p in triple-negative breast cancer relieves repression of KIF23 and promotes tumor cell proliferation and migration [66]. In hormone receptor-positive breast cancer, high KIF23 expression increases phosphorylation of GSK-3β, suppresses β-catenin degradation and activates downstream target genes [30]. The MMB–FOXM1 complex directly binds the KIF23 promoter and drives its transcription [78], and in triple-negative breast cancer, a WDR5/FOXM1/KIF23/Wnt–β-catenin positive feedback loop is present [32]. Furthermore, decreased miR-548 F-3p expression attenuates its inhibition of KIF23 and FOXM1 [79]; KIF23 cooperates with PRC1 to suppress the p53 pathway and enhance pro-survival autophagy [80]; and miR-30a-3p targets KIF23 to inhibit tumor proliferation and metastasis [81]. Clinically, KIF23 can serve as a prognostic biomarker in breast cancer [82], and promoter hypomethylation of KIF23 may act as an early warning indicator for triple-negative breast cancer [54].

Gastric cancer

KIF23 expression is significantly higher in gastric cancer tissues than in adjacent normal tissues, and its upregulation correlates with advanced pTNM stage, lymph node metastasis, and poor prognosis by activating the Wnt/β-catenin signaling pathway [83]. Receiver operating characteristic (ROC) analysis shows excellent diagnostic performance of KIF23 mRNA/protein expression in gastric cancer (training cohort AUC = 0.958; validation cohort AUC = 0.867) [3]. Mechanistically, KIF23 competitively binds Amer1, disrupts its normal complex formation with APC, and weakens its negative regulation of the Wnt/β-catenin pathway [35]. In terms of immune regulation, high KIF23 expression is associated with increased infiltration of regulatory T cells and M1-type macrophages, as well as upregulation of immune checkpoint molecules such as PD-L1 and CTLA4 [3].

For therapy response prediction, high KIF23 expression is associated with poor response to intraperitoneal paclitaxel in gastric cancer patients [84]. circ_0067934 promotes tumor proliferation and metastasis by sponging miR-1301-3p and upregulating KIF23 [60]. Genistein downregulates KIF23 and other kinesins, inducing G2/M arrest in gastric cancer cells [85]. Several circRNAs also relieve suppression of KIF23 and other hub genes by absorbing miRNAs [86].

Nasopharyngeal carcinoma

KIF23 expression is significantly higher in nasopharyngeal carcinoma tissues than in normal nasopharyngeal mucosa, and is associated with tumor stage, lymph node metastasis and poor prognosis [49]. Overexpression of eight hub genes including KIF23 drives cell cycle progression and inhibits apoptosis, promoting nasopharyngeal carcinoma cell proliferation [87]. The core mechanism involves AR-mediated transcriptional regulation: AR binds AREs within the KIF23 promoter to activate transcription, and high KIF23 expression enhances proliferation, migration and invasion of cancer cells via activation of Wnt/β-catenin signaling [49]. The phytochemical CVB-D downregulates KIF23, inhibits Akt/mTOR signaling and induces tumor cell apoptosis [36]. KIF23 also participates in the regulation of p53, PI3K–Akt and IL-17 pathways [88].

KIF23 drives tumor progression through coordinated activation of multiple pathways, and its core mechanisms must be understood by integrating upstream drivers, downstream effectors and clinical intervention targets as depicted in Fig. 3.

Fig. 3.

Integrated mechanistic landscape of KIF23-driven tumor progression and representative therapeutic opportunities. KIF23 overexpression can arise from upstream genomic and epigenetic events (such as 15q23 locus amplification [24] and promoter hypomethylation [54]) and from activation of ceRNA networks [59–61]. Elevated KIF23 is associated with activation of oncogenic pathways and phenotypes including cell-cycle dysregulation (G2/M), PI3K-Akt signaling, immune evasion (such as PD-L1–linked immunosuppression), and enhanced invasion/metastasis, collectively contributing to unfavorable clinical outcomes [2, 3], 36– [38]. Representative interventions targeting KIF23-centered axes are indicated, including FOXM1 inhibition by FDI-6, NAT10 inhibition by Remodelin (ac4C modulation), and CVB-D–associated downregulation of KIF23 in relevant preclinical settings [31, 36, 48]

Pan-cancer roles and clinical associations

KIF23 exhibits broad oncogenic functions across cancers. Its upregulation is closely linked to cell cycle dysregulation, activation of key signaling pathways, and poor prognosis. The core mechanisms, biological effects, and clinical associations of KIF23 in major cancer types (such as colorectal, hepatocellular, breast, gastric, and nasopharyngeal carcinomas) are systematically summarized in Table 4.

In bladder cancer, functional studies have demonstrated that KIF23 promotes cancer cell proliferation, migration, and invasion in both in vitro and in vivo models, strongly supporting its role as a bona fide oncoprotein and a potential driver of tumor progression in urological malignancies [89].

Beyond these, KIF23 also plays indispensable roles in a variety of other malignancies (such as cervical cancer, pancreatic cancer, glioma), where its functions exhibit marked tissue-type specificity. Although the specific upstream regulators and downstream effector pathways vary among these cancers, KIF23 consistently promotes tumor proliferation, invasion, and therapy resistance through common mechanisms, such as regulating cell cycle progression and conferring resistance to cell death. The specific mechanisms, clinical implications, and supporting references for KIF23 in these tumors are compiled in Table 5.

Table 5.

Mechanisms and clinical associations of KIF23 in other tumor types.

Tumor type	Core regulatory pathways/mechanisms	Key effects	Clinical associations (prognosis/therapy)	References
Cervical cancer	Inhibition of NLRP3-mediated pyroptosis; Wnt/β-catenin activation	Suppresses pyroptosis, promotes proliferation and migration; LINC00467/miR-107/KIF23 axis drives EMT	High expression associated with poor prognosis; NLRP3 agonist nigericin reverses pyroptosis inhibition	[39, 63, 90, 91]
Endometrial cancer	Activation of ERK/AKT/PI3K; HIF-1α-regulated lactate metabolism	Downregulates BAX/caspase-3, upregulates BCL-2; hypoxia-induced KIF23 promotes tumor growth	High expression reduces cisplatin/radiotherapy sensitivity; independent adverse prognostic factor	[92–94]
Pancreatic cancer	Cell cycle driving; regulation of macrophage phenotype switching	Promotes G2/M progression and recurrence; SLC16A1-mediated lactylation upregulates KIF23	High expression associated with early recurrence; KIF23 targeting inhibits proliferation	[95–98]
Ovarian cancer	Cell cycle dysregulation; p53 pathway suppression	Silencing of miR-424/503 leads to KIF23 upregulation; pre-malignant overexpression causes genome instability	High expression predicts resistance to MAPK inhibitors and correlates with poor prognosis	[45, 99]
Papillary thyroid carcinoma	Wnt/β-catenin activation; mitophagy suppression	Activates downstream target genes and suppresses mitophagy, promoting proliferation and migration	High expression associated with lymph node metastasis; potential therapeutic target	[10, 100]
Glioma	Copy number gains; miR-424 downregulation	TCF-4-mediated transcriptional activation of KIF23; promotes EMT and metastasis	Copy number gain predicts poor prognosis; miR-424 mimics inhibit progression	[53, 67, 101, 102]

Roles of KIF23 in non-neoplastic diseases

Dysfunction of KIF23 represents a core pathogenic factor in multiple non-neoplastic diseases, yet its mechanisms differ significantly from those observed in cancers. In developmental disorders such as congenital anemia and microcephaly, KIF23 primarily exhibits loss of function, directly disrupting fundamental biological processes including cell division. In contrast, in acquired conditions such as asthma and fibrotic diseases, its pathogenicity often stems from aberrant upregulation of expression or activity, driving disease progression by promoting inflammation, pyroptosis, or tissue remodeling. The following sections outline these core modes of action, while detailed mechanisms, key effects, and clinical implications are systematically summarized in Table 6.

Table 6.

Mechanisms and effects of KIF23 in non-neoplastic diseases.

Disease type	Core mechanism	Key pathological effects	Clinical implications (Diagnosis/Therapy)	References
Congenital dyserythropoietic anemia type III	Loss-of-function mutations (such as p.P916R) in KIF23.	Failure of cytokinesis in erythroblasts, leading to multinucleated cells; ineffective erythropoiesis and resultant anemia.	Genetic mutation testing can confirm diagnosis; targeting KIF23 function repair is a potential therapeutic strategy.	[4, 103], 109– [111]
Primary microcephaly	KIF23 deficiency leads to disordered spindle orientation and impaired cytokinesis.	Premature cell cycle exit of NSPCs; precocious neurogenesis and neuronal apoptosis, resulting in cortical maldevelopment.	Serves as a molecular marker for neurodevelopmental disorders.	[5]
Perioperative neurocognitive disorders	KIF23 upregulation activates the caspase-3/GSDME-mediated pyroptosis pathway.	Increased neuronal pyroptosis; elevated ROS generation and inflammatory factor release, leading to cognitive impairment.	The ETV5/KIF23 ratio may serve as an early diagnostic indicator.	[19, 50]
Asthma	KIF23 activates the p53 pathway and promotes inflammation/oxidative stress.	Exacerbates airway inflammation and apoptosis; induces lung tissue injury and airway hyperresponsiveness.	Knockdown of KIF23 alleviates symptoms; represents a potential therapeutic target.	[17]
Idiopathic pulmonary arterial hypertension	KIF23 modulates the PI3K/AKT/MAPK signaling pathway.	Imbalance between proliferation and pyroptosis in pulmonary arterial smooth muscle cells; increased pulmonary arterial pressure and right ventricular hypertrophy.	High expression indicates disease progression; targeted inhibition can alleviate vascular remodeling.	[38, 104]
Myocardial infarction	KIF23 inhibits Ces1d function via the RhoA signaling pathway.	Exacerbates cardiac fibrosis.	Targeting KIF23 may improve fatty acid β-oxidation and reduce fibrosis.	[18]
Liver fibrosis	KIF23 acts as a pro-fibrotic protein.	Promotes hepatic stellate cell activation; collagen deposition and destruction of liver architecture.	Luteolin alleviates fibrosis by downregulating KIF23.	[105]
Cerebral ischemia-reperfusion injury	Functions as an autophagy-related protein, upregulated by lncRNA NEAT1.	Exacerbates neuronal injury.	Knockdown of lncRNA NEAT1 downregulates KIF23 and exerts neuroprotective effects.	[106]
Vitiligo	Altered DNA methylation of KIF23.	Involvement in disease pathogenesis.	Methylation abnormalities precede transcriptional changes, holding diagnostic potential.	[57, 107]
Polycystic ovary syndrome	Dysregulation of KIF23 and other cell cycle-related genes.	Impedes endometrial decidualization, affecting embryo implantation.	Associated with patient fertility.	[108]

Congenital dyserythropoietic anemia

Mutations in the KIF23 gene are a primary etiology of Congenital Dyserythropoietic Anemia type III (CDA III). Loss of KIF23 function (such as the p.P916R missense mutation) leads to failure of cytokinesis in erythroid precursors, resulting in the formation of characteristic multinucleated erythroblasts and ultimately ineffective erythropoiesis and anemia [4, 103].

Microcephaly

During cerebral cortex development, KIF23 is crucial for maintaining normal mitosis in neural stem and progenitor cells (NSPCs). Its deficiency causes misorientation of the mitotic spindle and impaired cytokinesis, subsequently triggering premature neurogenesis and neuronal apoptosis, thereby disrupting normal cortical layering [5].

Perioperative neurocognitive disorders

In models of perioperative neurocognitive disorders, KIF23 expression is significantly upregulated in the hippocampal tissue. It exacerbates neuronal damage and cognitive decline by positively regulating the caspase-3/GSDME-dependent pyroptosis pathway [19, 50].

Other non-neoplastic diseases

Dysregulated KIF23 expression is involved in the pathological processes of various other diseases. For instance, in asthma, it exacerbates airway pathology by activating the p53 pathway and promoting inflammatory responses [17]; in myocardial infarction models, it aggravates cardiac fibrosis via the RhoA signaling pathway [18]; in idiopathic pulmonary arterial hypertension, its high expression in pulmonary arterial smooth muscle cells dysregulates the PI3K/AKT/MAPK pathway, driving an imbalance between cell proliferation and pyroptosis [38, 104]; in liver fibrosis, it acts as a pro-fibrotic protein involved in hepatic stellate cell activation [105]; and in cerebral ischemia-reperfusion injury, KIF23 functions as an autophagy-related protein whose expression is regulated by lncRNA NEAT1, exacerbating neuronal damage [106]. Furthermore, epigenomic studies suggest that altered DNA methylation of KIF23 contributes to the pathogenesis of vitiligo, and its dysregulation is also associated with endometrial dysfunction in polycystic ovary syndrome [57, 107]. In PCOS, disrupted expression of cell cycle genes including KIF23 impedes endometrial decidualization, affecting embryo implantation [108]. The functions of KIF23 in disease show marked heterogeneity, and a clinical translational framework has been constructed based on these features as depicted in Fig. 4.

Fig. 4.

Disease- and context-dependent roles of KIF23 and a translational roadmap. KIF23 shows gain-of-function characteristics in cancers, where upstream activation drives KIF23 upregulation and pro-tumor signaling (such as Wnt/β-catenin signaling in colorectal cancer), associating with unfavorable outcomes and enabling biomarker-guided stratification [2, 31]. KIF23-centered regulatory axes (including FOXM1-linked control in hepatocellular carcinoma) provide actionable targets for precision therapy [48, 51]. In non-neoplastic disorders, KIF23 more often reflects loss-of-function, exemplified by congenital dyserythropoietic anemia type III and neurodevelopmental phenotypes such as microcephaly [4, 5]. Translational applications include diagnostic performance based on KIF23-related signatures (such as gastric cancer AUC = 0.958) [3] and representative interventions targeting the KIF23 axis, including FDI-6 (~ 2.3-fold enhanced inhibition) [48], CVB-D (tumor growth inhibition rate = 52%) [36], and PVT1-siRNA–based downregulation of KIF23-associated circuits [61]

Clinical translational applications of KIF23

Based on its roles in disease, a translational framework of “diagnostic biomarker – therapeutic target – prognostic assessment” has been established for KIF23, providing new strategies for precision medicine.

Disease diagnostic markers

Abnormal expression, epigenetic modification and regulatory network alterations of KIF23 can serve as specific markers for early diagnosis and subtype discrimination in multiple diseases as shown in Table 7.

Table 7.

Clinical evidence of KIF23 as a diagnostic biomarker.

Disease type	Biomarker type	Sample/technique	Diagnostic performance (AUC/sensitivity/specificity)	References
Gastric cancer	KIF23 mRNA/protein expression	Tissue / IHC / qRT-PCR	Training cohort AUC = 0.958; validation cohort AUC = 0.867	[3]
Non-small cell lung cancer	KIF23 gene amplification + protein expression	Tissue / FISH / IHC	AUC = 0.82 for lung adenocarcinoma; no diagnostic value in squamous carcinoma	[24, 112]
Triple-negative breast cancer	KIF23 promoter methylation	Tissue / methylation-specific PCR (MSP)	40% decrease in methylation; AUC = 0.89	[54]
Perioperative neurocognitive disorders	ETV5/KIF23 ratio	CSF / qRT-PCR	AUC = 0.79; sensitivity 72%; specificity 68%	[50]
Vitiligo	KIF23 DNA methylation	Perilesional skin / 850 K array	Methylation changes precede transcriptional changes; AUC = 0.81	[57]
CDA type III	KIF23 gene mutations	Peripheral blood / whole-exome sequencing (WES)	Mutation detection rate 92%; p.P916R accounts for 65% of cases	[4]

As a diagnostic biomarker, KIF23 shows remarkable potential across cancers. Its diagnostic performance is particularly outstanding in gastric cancer (AUC up to 0.958) [3]. In non-small cell lung cancer, KIF23 overexpression is driven by amplification at chromosome 15q23, and a diagnostic model based on its expression level demonstrates discriminative power for the lung adenocarcinoma subtype, with an area under the curve (AUC) of 0.82 [24, 112]. In triple-negative breast cancer, promoter hypomethylation of KIF23 serves as an early warning signal [54]. In diffuse large B-cell lymphoma, promoter hypomethylation correlates with high KIF23 expression [56]. In ovarian cancer, high KIF23 expression helps facilitate subtype diagnosis and disease assessment [113].

In non-neoplastic diseases, KIF23 expression in the hippocampus of patients with perioperative neurocognitive disorders is inversely correlated with ETV5, and the ETV5/KIF23 ratio yields an AUC of 0.79 [50]. In vitiligo, DNA hypomethylation of KIF23 in perilesional skin precedes pigmentary alterations [57]. In CDA III, detection of pathogenic KIF23 mutations such as p.P916R provides a definitive diagnosis [4]. In IPAH, high KIF23 expression in PASMCs may serve as a potential diagnostic indicator [38].

Therapeutic target

Therapeutic strategies targeting KIF23 have exhibited promising effects in preclinical studies as shown in Table 8. From an intervention standpoint, targeting KIF23 or its upstream regulators yields potent anti-tumor effects in preclinical models. In hepatocellular carcinoma, inhibition of FOXM1 significantly downregulates KIF23 and reverses sorafenib resistance [48]. In prostate cancer, siRNA-mediated knockdown of the upstream lncRNA PVT1 reduces KIF23 levels, leading to dramatic decreases in tumor volume and metastasis [61]. In cervical cancer, the NLRP3 agonist nigericin blocks KIF23-mediated suppression of pyroptosis and increases tumor cell apoptosis by 35% [39]. Across cancers, the NAT10 inhibitor Remodelin reduces KIF23 ac4C modification and translation, decreasing liver metastases by 50% in colorectal cancer xenograft models [31]. In nasopharyngeal carcinoma, CVB-D directly downregulates KIF23 and inhibits Akt/mTOR signaling [36]. In triple-negative breast cancer, FOXM1 inhibitors attenuate tumor progression by reducing KIF23 expression [32].

Table 8.

KIF23-targeted interventions and their preclinical efficacy.

Intervention type	Target/mechanism	Disease type	Preclinical efficacy (in vitro / in vivo)	References
Small-molecule inhibitor	FOXM1 (indirect inhibition of KIF23)	Hepatocellular carcinoma	in vitro: FDI-6 reduces KIF23 by 45% and increases sorafenib sensitivity by 2.1-fold; in vivo: tumor growth inhibition rate increased by 2.3-fold	[48]
siRNA	lncRNA PVT1 (indirect downregulation of KIF23)	Prostate cancer	in vitro: apoptosis increased by 40%; in vivo: tumor volume reduced by 60% and metastasis by 55%	[61]
Inflammasome agonist	NLRP3 (reverses KIF23-mediated suppression of pyroptosis)	Cervical cancer	in vitro: apoptosis increased by 35%; in vivo: tumor weight decreased by 42%	[39]
RNA modification inhibitor	NAT10 (reduces ac4C modification of KIF23 mRNA)	Colorectal cancer	in vitro: KIF23 translational efficiency decreased by 50%; in vivo: liver metastatic nodules reduced by 50%	[31]
Phytochemical	CVB-D (directly downregulates KIF23)	Nasopharyngeal carcinoma	in vitro: p-Akt/p-mTOR decreased by 45%, apoptosis increased by 38%; in vivo: tumor growth inhibition rate 52%	[36]
miRNA mimics	miR-424-5p (targets KIF23)	Hepatocellular carcinoma	in vitro: reduces KIF23 and enhances sorafenib sensitivity; in vivo: strengthens tumor growth inhibition	[68]

In non-neoplastic diseases, lentivirus-mediated KIF23 knockdown reduces PASMC proliferation by 40% in IPAH models [38]. In asthma, siRNA-mediated KIF23 silencing decreases pulmonary IL-1β levels by 55% in mice [17]. In myocardial infarction, KIF23-targeted interventions improve fatty acid β-oxidation and reduce cardiac fibrosis [18]. In liver fibrosis, luteolin alleviates fibrosis by downregulating KIF23 [105].

Prognostic assessment

KIF23 expression is clearly associated with disease progression and therapeutic response, making it an important indicator for prognostic evaluation. For tumor risk stratification, high KIF23 expression is an independent adverse prognostic factor in diffuse large B-cell lymphoma. Univariate analysis revealed that it was associated with progressively worse survival at 3 years (HR = 1.28), 5 years (HR = 1.41), and 10 years (HR = 1.45), with all P values below 0.05 [56]. In ovarian cancer, high KIF23 expression predicts resistance to MAPK inhibitors [113]. In hepatocellular carcinoma, patients with high KIF23 expression have a median survival 14 months shorter than controls [7]. In colorectal cancer, high KIF23 expression correlates with advanced TNM stage, lymph node metastasis and lower 5-year survival [2]. In breast cancer, high KIF23 expression is closely linked to invasive/metastatic potential and poor prognosis [66].

Regarding prediction of therapeutic efficacy, high KIF23 expression is associated with intraperitoneal paclitaxel resistance in gastric cancer [84]. In hepatocellular carcinoma, KIF23 levels negatively correlate with response to sorafenib (r = − 0.62) [68]. In non-small cell lung cancer, high KIF23 expression is associated with radioresistance [114]. In chronic myeloid leukemia, KIF23 overexpression contributes to imatinib resistance [37]. This notion is reinforced by a recent systematic review, which confirms that KIF23 enhances sorafenib and cisplatin resistance in hepatocellular carcinoma, underscoring its broad involvement in chemoresistance across different cancer types [115]. At the mechanistic level, KIF23 has recently been demonstrated to promote primary cisplatin resistance in cervical cancer by activating the MYH9/MCM2/PCNA axis, providing the first detailed molecular pathway that directly links KIF23 to chemoresistance [116].

Challenges and future strategies for clinical translation of KIF23-targeted therapies

Despite encouraging preclinical findings, several obstacles must be overcome before KIF23‑targeted therapies can be successfully translated into clinical practice. First, because KIF23 is essential for cytokinesis in normal dividing cells, systemic inhibition may cause on‑target toxicity in rapidly proliferating tissues, such as bone marrow and intestinal epithelium. To improve tumor selectivity, the following strategies could be considered: (i) exploiting synthetic lethal interactions, particularly in the context of TP53 inactivation, which may render cancer cells more dependent on KIF23 [15, 16, 117, 118]; (ii) developing tumor‑specific delivery systems, including proteolysis‑targeting chimeras (PROTACs) and antibody–drug conjugates (ADCs) [119, 120]; and (iii) targeting tumor‑enriched KIF23 splice variants (such as the V2 isoform in hepatocellular carcinoma) to achieve isoform‑selective inhibition.

Second, off‑target effects and safety profiles warrant thorough evaluation in preclinical models, with particular emphasis on bone marrow suppression, gastrointestinal toxicity, and the identification of reliable biomarkers for early detection of adverse events.

Third, combination therapies may enhance efficacy and help overcome resistance. Based on current evidence, KIF23 inhibitors could be combined with sorafenib in hepatocellular carcinoma [48, 68], with taxanes or PARP inhibitors (taking advantage of cell‑cycle synergy), or with immune checkpoint inhibitors, given the established role of KIF23 in regulating PD‑L1 expression [13].

Finally, patient stratification will be critical for clinical success. Candidate biomarkers include KIF23 expression levels, splice variant ratios, and the activation status of upstream regulators such as FOXM1 and AR. Prospective validation of these biomarkers in well‑designed clinical trials is urgently needed to guide patient selection and maximize therapeutic benefit.

Future prospects and research directions

Compelling evidence for KIF23 involvement in disease underscores strong translational potential; however, several critical challenges remain in bridging mechanistic insights to clinical application. A structured research agenda, organized around four interconnected strategic pillars, is proposed to advance KIF23 from a promising biological target toward tangible clinical implementation.

Decoding the functional heterogeneity of splice variants

Although the distinct subcellular localization and functional roles of major KIF23 splice variants, such as the nuclear V1 and cytoplasmic V2 isoforms in hepatocellular carcinoma are established, a comprehensive understanding of their expression patterns, interaction networks, and disease-specific regulatory functions across different cancers and non-neoplastic conditions remains lacking [22, 26]. Moving forward, integrating isoform-specific sequencing, gene-editing tools, and super-resolution imaging will be crucial to systematically map the expression landscapes, protein-protein interactions, and pathological associations of these variants [121]. This work should clarify how individual variants uniquely contribute to tumor progression, therapy resistance, or physiological homeostasis. Ultimately, establishing a functional classification framework for KIF23 splice variants will help assess their clinical potential as precision diagnostic markers or subtype-specific therapeutic targets, thereby informing the development of individualized treatment strategies.

Advancing tumor-selective therapeutic strategies

The essential role of KIF23 in normal cell division raises concerns that its direct inhibition could lead to systemic toxicity. While preclinical studies of small-molecule inhibitors like Remodelin and CVB-D are promising, enhancing their tumor specificity is critical for clinical translation. Future efforts should therefore focus on developing safer, more selective approaches. First, identifying synthetic lethal interactions with KIF23 dysregulation, for instance, in the context of TP53 inactivation [117, 118, 122] could help delineate patient subgroups most likely to benefit from KIF23-targeted therapies [15, 16]. Second, engineering tumor-microenvironment-responsive prodrugs or antibody-drug conjugates may enable precise delivery to malignant cells, minimizing off-target effects [119, 120]. Third, rational combination therapies that co-target upstream regulators (such as FOXM1, AR, NAT10) or downstream effector pathways (such as Wnt/β-catenin) should be optimized and rigorously tested in preclinical models for synergistic efficacy and safety [31, 48, 49].

Clinical translation of integrated biomarker panels

Relying solely on KIF23 expression as a diagnostic or prognostic marker has inherent limitations. A more robust approach will require the integration of multi-dimensional data, encompassing KIF23 transcript levels, epigenetic alterations (such as promoter hypomethylation [54, 57], ceRNA network activity, and downstream effector molecules [68]. Illustrative examples include the high diagnostic accuracy of KIF23 mRNA in gastric cancer (AUC up to 0.958) [3], the early-warning potential of its promoter hypomethylation in triple-negative breast cancer [54], and its strong association with reduced long-term survival in diffuse large B-cell lymphoma (10-year survival decreased by 42%) [56]. For example, integrating molecular biomarkers with pathological microenvironment characteristics, such as the density of lymphatic vessels within the tumor capsule, has recently been demonstrated to improve the accuracy of postoperative recurrence risk prediction in patients with early-stage hepatocellular carcinoma [123].

Beyond this, a similar paradigm is emerging for other secretory proteins. ZG16B, for example, has been evaluated as a diagnostic and prognostic biomarker across multiple malignancies, and several ZG16B‑specific therapeutic strategies such as monoclonal antibodies, RNA aptamers, and trans‑splicing ribozymes have shown preclinical efficacy, with a humanized ZG16B antibody currently in clinical trials for advanced pancreatic cancer [42].

To advance this paradigm, future studies should leverage multi-omics resources and data from multi-center prospective cohorts to build and validate integrated biomarker panels. Priority should be given to evaluating their utility in concrete clinical scenarios, such as the early detection of gastric cancer and predicting responses to therapies like sorafenib in hepatocellular carcinoma [124–126].

Leveraging single-cell and spatial omics to decipher context-specific mechanisms

Current understanding of KIF23 is primarily derived from tissue-level analyses, which limits insight into its cell-type-specific functions within complex disease microenvironments. For instance, in gastric cancer, KIF23 can modulate immune cell infiltration and PD-L1 expression [3]. To resolve these mechanisms at a cellular resolution, future work should employ single-cell multi-omics (such as scRNA-seq, scATAC-seq) and spatial transcriptomic/proteomic platforms [132–135]. These approaches will enable precise mapping of KIF23 expression across distinct cell subpopulations, such as malignant cells, tumor-associated macrophages, or neural stem cells within pathological niches. Furthermore, they can reveal how KIF23 shapes intercellular crosstalk and microenvironment remodeling [127–130]. Uncovering these cell-type-specific regulatory circuits will open new avenues for identifying actionable upstream targets tailored to particular cellular contexts [3, 41, 131].

By pursuing the four strategic directions outlined above in a coordinated manner, foundational discoveries about KIF23 can be systematically translated into clinically useful precision diagnostics, safer targeted therapies, and refined prognostic frameworks. This integrated effort promises to establish a new paradigm for precision medicine across both malignant and non-malignant diseases.

Conclusions and perspectives

This review integrates evidence across molecular, cellular, and clinical domains, positioning KIF23 as a pivotal cellular regulator whose context-dependent dysregulation contributes to a broad range of human diseases. In cancer, KIF23 acts as a potent oncogenic driver, promoting tumor progression through the activation of multiple signaling pathways, induction of genomic instability, and remodeling of the immunosuppressive tumor microenvironment. Conversely, in non-neoplastic settings, loss of KIF23 function disrupts essential processes such as cytokinesis, underlying congenital anemias and neurodevelopmental disorders. This “functional duality” highlights its significant therapeutic potential while also revealing the key challenge of achieving selective targeting in the clinic.

Translating this mechanistic understanding into clinical applications remains a substantial endeavor. As outlined in our proposed research agenda, progress will depend on focused efforts in four critical areas: clarifying the distinct roles of KIF23 splice variants; developing tumor-selective therapeutic strategies to improve safety; advancing the clinical validation of integrated, multi-parameter biomarker panels; and employing single-cell and spatial omics to unravel cell-type-specific regulatory networks in diseased tissues.

By addressing these priorities through coordinated and innovative research, KIF23 can evolve from a compelling molecular target into a practical cornerstone of precision medicine. This trajectory holds the promise of delivering novel diagnostics and therapies, ultimately improving outcomes for patients with both malignant and non-malignant diseases.

Author contributions

Conceptualization was carried out by FFL and CML. The literature search and data curation were performed by DWY and QQM, while formal analysis and interpretation were conducted by MZQ. The original draft of the manuscript was written by CML, and the review and editing were completed by FFL. Visualization, including Figs. 1, 2, 3 and 4 and the graphical abstract, was prepared by FXQ. Supervision of the study was provided by MZQ and FFL. All authors read, critically revised, and approved the final manuscript.

Funding

This research received no external funding.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable. This review article is based on the analysis of previously published literature and does not involve any original studies with human participants, animal subjects, or personal data requiring ethical approval or informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.