NCBP2 drives colorectal cancer growth and metastasis through LIPG-mediated lipid droplet accumulation

Abstract

RNA-binding proteins play critical roles in RNA processing and are aberrantly expressed in colorectal cancer (CRC). Through comprehensive analysis of multiple gene expression datasets (GSE20916, GSE18105, GSE21510, TCGA-COAD and TCGA-READ), NCBP2 was identified as a potential tumorigenic gene in CRC. NCBP2 expression was significantly elevated in CRC tissues, correlated with tumour invasion and metastasis, and associated with poor patient survival outcomes. Furthermore, overexpression of NCBP2 in CRC cells was shown to increase cell proliferation, migration, and tumour invasion in both in vitro and in vivo models. Mechanistically, the NCBP2 protein stabilised LIPG mRNA via direct binding to the m7G motif in the 5’-cap structure of LIPG mRNA, thereby increasing LIPG expression. Additionally, NCBP2 promoted lipid droplet accumulation in CRC cells in a LIPG-dependent manner. These findings collectively suggest that the NCBP2-LIPG-lipid droplet axis represents a novel mechanism underlying CRC progression and metastasis, providing a promising therapeutic target for CRC treatment.

Experiments both in vitro and in vivo confirm that the NCBP2-LIPG-lipid droplet axis represents a mechanism underlying the progression and metastasis of colorectal cancer (CRC), providing a promising therapeutic target for the treatment of CRC.

Introduction

Colorectal cancer (CRC) is the third most prevalent solid tumour and the second leading cause of cancer-related death globally^1,2. Two primary drivers of CRC-related mortality are the excessive growth of cancer cells and the occurrence of distant metastasis. CRC cell overgrowth leads to colorectal obstruction and massive haemorrhage, necessitating urgent surgical intervention^3,4, which often results in severe complications and a high risk of fatal outcomes⁵. Beyond the gastrointestinal tract, CRC most commonly metastasises to the liver and lung, with distant metastasis being the most important cause of CRC-related mortality^6,7. Despite remarkable progress in comprehensive treatment modalities, CRC patients with distant metastases still exhibit a dismal 5-year overall survival rate of approximately 15%⁸. Consequently, elucidating the molecular mechanisms underlying CRC growth and metastasis represents a critical imperative for advancing diagnostic and therapeutic strategies for CRC.

RNA-binding proteins (RBPs) bind directly to the RNAs of target genes and subsequently regulate a complex network of biological processes involved in gene expression^9,10. The addition of the cap structure to the 5’-end of pre-mRNAs is essential for efficient gene expression. It stabilises mRNAs and promotes their transcription, splicing and nuclear export^11,12. The addition of the 5’-cap structure to mRNAs occurs in three enzymatic steps. First, the 5’-γ-phosphate group is removed from the first transcribed nucleotide of pre-mRNA; second, a guanosine cap (GpppN) is formed via the transfer of guanine monophosphate nucleotides to the RNA 5’-diphosphate end; finally, GpppN is methylated to produce the 7-methylguanosine cap (m7GpppN, m7G)¹³. The function of the 5’-cap structure in mRNAs is modulated by nuclear cap-binding proteins. Nuclear cap binding protein subunit 2 (NCBP2, also called CBP20) is a subunit of the nuclear cap-binding complex (CBC)¹⁴. NCBP2 coordinates the biogenesis of RNA processes by recruiting the cap structure to newly transcribed mRNA^15,16. Notably, NBCP2 is abnormally expressed in many cancers, including head and neck squamous cell carcinoma and pancreatic cancer^17–19.

Lipid accumulation plays a critical role in accelerating cancer progression and promoting metastasis, as demonstrated in previous studies^20–22. Lipid droplets (LDs), the intracellular form of lipid deposits, are a new hallmark of cancer cells and are strongly correlated with many cellular processes^23–25. LDs consist of a neutral lipid core of triacylglycerol (TG) and cholesterol esters surrounded by a phospholipid monolayer and associated LD surface proteins. LD formation is regulated by two main pathways: lipogenesis and the acquisition of extracellular lipids through lipase-mediated lipolysis²³. It has been suggested that reduced LD formation mediated by ACSS3 suppresses prostate cancer progression²⁶, whereas increased LD accumulation mediated by LPCAT2 supports CRC chemoresistance by impairing caspase cascade activation and ER stress responses in CRC cells²⁷. This evidence suggests that LDs play critical roles in cancer cell proliferation, metastasis and drug resistance.

To date, no study has investigated whether nuclear cap-binding proteins can modulate LD accumulation in CRC cells. In this study, we report that NCBP2, a tumour-promoting gene, contributes to CRC growth and metastasis by promoting LD accumulation in CRC cells. Mechanistically, NCBP2 enhances LD formation in CRC cells by increasing the expression of endothelial lipase (LIPG), a key regulator of LD composition. Notably, NCBP2 interacts with LIPG mRNA through direct binding to m7G in the 5’-cap structure, thereby stabilising LIPG mRNA expression. This study provides critical insights into the role of NCBP2 in CRC patient prognosis and suggests that the NCBP2-LIPG axis represents a promising therapeutic target for CRC treatment.

Results

Bioinformatics analyses reveal that NCBP2 functions as an oncogene in CRC

To identify differentially expressed genes between CRC tissues and normal colorectal tissues, the TCGA database and three colorectal datasets (GSE21510, GSE18105 and GSE20916) were screened on the basis of a fold change (FC) threshold of >1.0 and an adjusted p < 0.01. A total of 394 genes were identified (Fig. 1a). To subsequently identify the subset of 394 candidate genes potentially associated with CRC prognosis, we conducted further analysis of the TCGA-CRC database, which included comprehensive prognostic information. A total of four genes, CTHRC1, THBS2, NCBP2 and SPP1, were significantly associated with CRC prognosis, as shown in Supplementary Fig. 1a–d. Among the four genes examined, three genes, CTHRC1, THBS2 and SPP1, have been well studied in CRC^28–30. NCBP2 attracted our attention because it is an RBP that regulates the biological processes of RNAs¹⁶. In addition, NCBP2 mRNA levels were significantly elevated in CRC tissues than in normal colorectal tissues, as demonstrated by analysis of the TCGA and GSE40967 datasets (Fig. 1b, c). An analysis of data from the TCGA database revealed that high levels of NCBP2 mRNA in CRC tissues were significantly associated with shorter overall survival in CRC patients³¹. Similarly, a high level of NCBP2 protein in cancer indicated a poor prognosis in CRC patients according to data from the Human Protein Atlas (https://www.proteinatlas.org) (Fig. 1d).

Fig. 1. NCBP2 is upregulated in colorectal cancer (CRC) and indicates a poor prognosis.

a Target genes with differential expression between CRC tissues and normal colorectal tissues and with prognostic value in CRC patients were identified from the TCGA, GSE20510, GSE18105 and GSE20916 datasets. NCBP2 mRNA levels were compared between CRC tissues and normal colorectal tissues in the TCGA database (b) and GSE40967 dataset (c). d The Human Protein Atlas was used to analyse the impact of the NCBP2 protein level on overall survival in CRC patients. e NCBP2 mRNA levels were detected in 29 CRC tissues and paired normal colorectal tissues. Representative images of immunohistochemical staining for NCBP2 in CRC tissues and normal tissues are shown in (f); scale bar: 50 μm; NCBP2 protein levels were assessed in 97 CRC tissues and paired normal colorectal tissues in (g). h A Kaplan–Meier curve was generated to assess the effect of NCBP2 protein on overall survival in 97 CRC patients. i NCBP2 mRNA expression was detected in cell lines. P values were determined by two-tailed t-test (b, c, e, g) or one-way ANOVA with Tukey’s multiple comparison test (i). Data are expressed as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001.

To investigate the potential role of NCBP2 in CRC pathogenesis, mRNA and protein expression levels were assessed using RT‒qPCR and immunohistochemical staining. In 29 CRC patients, the NCBP2 mRNA level was significantly increased in CRC tissues than in normal adjacent tissues (Fig. 1e; Supplementary Table 1). In addition, the expression of NCBP2 protein was notably higher in CRC tissues than in adjacent normal tissues from 97 CRC patients (Fig. 1f, g). Elevated NCBP2 protein expression in CRC was significantly associated with increased cancer invasion depth, distant metastasis and advanced TNM stage (Table 1). Furthermore, compared with patients with low NCBP2 expression, CRC patients with high NCBP2 expression had poorer overall survival (Fig. 1h; Supplementary Fig. 2a, b; Table 2). Analyses of five CRC cell lines and NCM460 cells revealed that NCBP2 mRNA expression was higher in CRC cell lines than in NCM460 cells (Fig. 1i). HCT116 and SW480 cells with moderate levels of NCBP2 expression were selected for the cell experiments in this study. Collectively, these findings suggested that a high level of NCBP2 was associated with a poor outcome in CRC patients, supporting its potential as a prognostic biomarker.

Table 1.

The associations between NCBP2 and LIPG expressions and pathological features in CRC

Clinical Variables		NCBP2			LIPG
	n	Low (n = 38)	High (n = 59)	P value	Low (n = 62)	High (n = 35)	P value
Age (years)				0.04			0.27
≥60	51	15	36		30	21
<60	46	23	23		32	14
Gender				0.48			0.80
Male	57	24	33		32	19
Female	30	14	26		30	16
Tumour diameter (cm)		4.03 ± 1.47	4.28 ± 1.53	0.43	3.99 ± 1.32	4.52 ± 1.75	0.10
Tumour location				0.12			0.51
Colon	32	9	23		19	13
Rectum	65	29	36		43	22
Depth of invasion				0.03			0.20
T1/T2	33	18	15		24	9
T3/T4	64	20	44		38	26
Lymph node invasion				0.18			0.001
Negative	38	24	29		32	6
Positive	59	14	30		30	29
Distant metastasis				<0.001			0.30
Negative	80	38	42		53	27
Positive	17	0	17		9	8
TNM stage				<0.001			0.001
I/II	35	24	11		30	5
III/IV	62	14	48		32	30

CRC colorectal cancer, TNM Tumour-Node-Metastasis.

Table 2.

Univariate and multivariate cox regression analyses of the factors associated with 5-year survival in CRC patients

Clinical variables	HR	95% CI	P value
Univariate analysis
Age(≥60 years vs <60 years)	2.07	0.97–4.42	0.061
Gender (female vs male)	1.38	0.67–2.82	0.383
Location (rectum vs colon)	0.78	0.37–1.65	0.520
Tumour diameter (≥4 cm vs <4 cm)	1.84	0.82–4.13	0.140
TNM stage (III&IV vs I&II)	4.99	1.74–14.36	0.003
NCBP2 (high vs low)	2.69	1.15–6.27	0.022
LIPG (high vs low)	3.73	1.79–7.79	<0.001
Multivariate analysis
TNM stage (III&IV vs I&II)	3.84	1.27–11.04	0.017
LIPG (high vs low)	2.78	1.31–5.92	0.008

CRC colorectal cancer, TNM Tumour-Node-Metastasis.

NCBP2 promotes CRC cell proliferation and metastasis

To investigate the biological role of NCBP2 in vitro, HCT116 and SW480 cells were transfected with the NCBP2 overexpression plasmid, and the transfection efficiency was validated using RT‒qPCR and Western blotting (Fig. 2a, b). The biological effects of NCBP2 overexpression on CRC cells were evaluated using CCK-8 and colony formation assays. As shown in Fig. 2c, d, compared with the empty vector (EV), NCBP2 overexpression significantly promoted HCT116 and SW480 cell proliferation and enhanced colony formation. Furthermore, it increased the migratory capacity of HCT116 and SW480 cells (Fig. 2e–g). Notably, NCBP2 overexpression induced epithelial–mesenchymal transition (EMT), as evidenced by increased N-cadherin and vimentin protein expression and decreased E-cadherin protein expression (Fig. 2h, i), and HCT116 and SW480 cells exhibited more pseudopodia according to phalloidin staining (Supplementary Fig. 3a). Additionally, we examined the impact of NCBP2 overexpression on the expression of cancer stem cell-specific markers. The results demonstrated that NCBP2 overexpression enhanced sphere formation (Fig. 2j) and increased the expression of cancer stem cell-specific markers, including CD133, CD44, OCT4 and SOX2 (Fig. 2k, l). In addition, NCBP2 overexpression increased CD133 and OCT4 protein levels in HCT116 and SW480 cells (Supplementary Fig. 3b).

Fig. 2. NCBP2 overexpression promotes CRC growth and metastasis.

HCT116 and SW480 cells were transfected with a NCBP2 overexpression plasmid or an empty vector (EV), and the transfection efficiency was verified via Western blotting (a) and RT‒qPCR (b). CCK-8 (c) and colony formation (d) assays were used to assess the proliferation of NCBP2-overexpressing or NCBP2-EV CRC cells. Wound healing assays (e, f) and transwell migration assays (g) were used to assess the migration of NCBP2-overexpressing or NCBP2-EV CRC cells. Scale bar: 50 μm. h, i Western blotting was used to assess the expression of EMT markers in NCBP2-overexpressing or NCBP2-EV CRC cells. kDa (kilo Dalton), molecular weights of standards. j Sphere formation assays were conducted in NCBP2-overexpressing or NCBP2-EV CRC cells, and representative images were obtained on day 7 after the cells were seeded. Scale bar: 50 μm. k, l The expression of cancer stem cell-specific markers (CD133, CD44, OCT4 and SOX2) was assessed via RT‒qPCR in NCBP2-overexpressing or NCBP2-EV CRC cells. The CRC cells included HCT116 and SW480 cells. EV, empty vector. P values were determined by unpaired two-tailed t-test (a–l). n = 3 (a, b, d–l) and n = 5 (c) biologically independent experiments. Data are expressed as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001.

To investigate the functional role of NCBP2 in regulating tumour growth, subcutaneous engraftment models were established using NCBP2-overexpressing or NCBP2-EV HCT116 cells. Compared with the subcutaneous tumours derived from NCBP2-EV cells, those derived from NCBP2-overexpressing cells exhibited significantly faster tumour growth (Supplementary Fig. 4a–c) and greater cell proliferation, as evidenced by ki-67 staining (Supplementary Fig. 4d, e). Furthermore, NCBP2 overexpression enhanced the pulmonary metastasis of HCT116 cells in a mouse model (Supplementary Fig. 4f, g).

To demonstrate the effect of NCBP2 inhibition on the growth and metastasis of CRC cells, NCBP2 expression was knocked down using shRNA, and the knockdown efficiency was determined via RT‒qPCR and Western blotting (Supplementary Fig. 5a, b). NCBP2 knockdown significantly reduced the proliferation of HCT116 and SW480 cells (Supplementary Fig. 5c, d). In addition, NCBP2 knockdown suppressed the migration ability of HCT116 and SW480 cells, as evidenced by the results of the scratch test and transwell migration assays (Supplementary Fig. 5e, f). Furthermore, NCBP2 knockdown inhibited EMT, as demonstrated by decreased N-cadherin and vimentin protein levels and increased E-cadherin protein levels (Supplementary Fig. 5g, h) in CRC cells. Taken together, these results suggested that NCBP2 promoted CRC growth and metastasis both in vitro and in vivo.

LIPG is a mediator of NCBP2 function in CRC cells

NCBP2 is an RBP that is involved in the posttranscriptional modification of RNAs¹⁵, and as a m7G reader, it plays an important role in m7G modification^18,19,32. To explore the target mRNAs affected by NCBP2, mRNA sequencing was conducted using NCBP2-overexpressing or NCBP2-EV HCT116 cells (Fig. 3a, b). With a log₂fold change (FC) > 1.0 and P < 0.05, 38 genes were determined to be differentially expressed upon NBCP2 overexpression (Supplementary Table 2). Among these mRNAs, six genes, LIPG, ITH3, PTHCH, ITGB8, ZBED6 and GAL1, were selected for further validation because these genes have been reported to promote cancer progression; however, whether these genes promote CRC progression remains unclear.

Fig. 3. LIPG is a downstream target of NCBP2 and mediates its functions in CRC cells.

Volcano plot (a) and heatmap (b) show the results of mRNA sequencing of NCBP2-overexpressing and NCBP2-EV HCT116 cells. RT–qPCR was used to verify the expression of five genes in NCBP2-overexpressing (c) or NCBP2-knockdown (d) HCT116 cells. RT‒qPCR was used to verify the expression of five genes in NCBP2-overexpressing (e) or NCBP2-knockdown (f) SW480 cells. Western blotting was used to assess the levels of LIPG protein in CRC cells upon NCBP2 overexpression (g) or knockdown (h). kDa (kilo Dalton), molecular weights of standards. i The data obtained for the TCGA-COAD dataset were used to analyse the correlation between NCBP2 and LIPG expression. j Co-IP was performed to detect the efficiency of the antibody against flag-NCBP2 fusion protein in CRC cells. kDa (kilo Dalton), molecular weights of standards. k, l RIP analysis was performed to assess the binding ability of NCBP2 to LIPG mRNA using anti-IgG and anti-flag antibodies in NCBP2-overexpressing or NCBP2-EV CRC cells. m, n NCBP2-overexpressing or NCBP2-EV CRC cells were treated with toyocamycin (2 μg) and collected at the indicated time points for RT‒qPCR in CRC cells. o Co-IP was performed to detect the binding ability of m7G to NCBP2 protein in CRC cells. kDa (kilo Dalton), molecular weights of standards. p, q RIP analysis was performed to assess the binding ability of m7G to LIPG mRNA using IgG and m7G antibody in NCBP2-overexpressing or NCBP2-EV CRC cells. The CRC cells included HCT116 and SW480 cells. EV, empty vector. P values were determined by unpaired two-tailed t-test (c, e, g, k–n, p, q) or one-way ANOVA with Tukey’s multiple comparison test (d, f, h). n = 3 biologically independent experiments. Data are expressed as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001.

We validated whether NCBP2 mediated the expression of these genes in CRC cells. The results revealed that NCBP2 overexpression led to an increase in LIPG mRNA levels and vice versa (Fig. 3c–f). In contrast, the expression of the other five mRNAs did not consistently change in response to NCBP2 overexpression or knockdown in CRC cells (Fig. 3c–f). Furthermore, the protein level of LIPG correspondingly increased or decreased with NCBP2 overexpression or knockdown, respectively, in CRC cells (Fig. 3g, h). Additionally, analysis of TCGA-CRC data revealed a positive correlation between NCBP2 and LIPG expression levels (Fig. 3i). These findings suggested that NCBP2 regulated LIPG mRNA expression and that LIPG was a downstream target of NCBP2.

NCBP2 increases LIPG mRNA stability by interacting with m7G modifications

NCBP2 is a subunit of the nuclear CBC and participates in the regulation of RNA stability by binding to the 5’-end cap structure of newly transcribed mRNAs^14,16,18. To assess whether NCBP2 directly interacted with LIPG mRNA, a RIP experiment was conducted. The results demonstrated that overexpression of NCBP2 increased the amount of LIPG mRNA bound to NCBP2 in HCT116 and SW480 cells (Fig. 3j–l; Supplementary Fig. 6a, b). Furthermore, the effect of NCBP2 on LIPG mRNA stability was assessed in HCT116 and SW480 cells pretreated with toyocamycin, an inhibitor of mRNA synthesis. The findings revealed that NCBP2 overexpression significantly extended the half-life of LIPG mRNA in CRC lines (Fig. 3m, n). Previous studies have reported that NCBP2 influences mRNA stability by modulating m7G modification^18,19,32. Co-IP revealed that NCBP2 interacted with m7G in HCT116 and SW480 cells (Fig. 3o). RIP-qPCR revealed that m7G was detectable in LIPG mRNA and that the m7G level in LIPG mRNA was elevated when NCBP2 was overexpressed in HCT116 and SW480 cells (Fig. 3p, q). Collectively, these results indicated that NCBP2 bound to LIPG mRNA, increased LIPG mRNA stability and upregulated LIPG protein levels in CRC cells by modulating m7G modifications.

LIPG facilitates CRC cell proliferation and metastasis

To address the role of LIPG in the proliferation and metastasis of CRC cells, LIPG was successfully overexpressed in HCT116 and SW480 cell lines (Supplementary Fig. 7a, b). Compared with LIPG-EV, the overexpression of LIPG significantly accelerated CRC cell growth, as evidenced by the results of a CCK-8 assay (Supplementary Fig. 7c) and a colony formation assay (Supplementary Fig. 7d). Similarly, the migratory capacity of CRC cells was enhanced by LIPG overexpression (Supplementary Fig. 7e–g). Furthermore, LIPG overexpression was associated with the upregulation of EMT-specific proteins, including increased N-cadherin and vimentin protein levels, in conjunction with decreased E-cadherin protein levels (Supplementary Fig. 7h, i). Moreover, overexpression of NCBP2 accelerated the formation of cancer cell spheres (Supplementary Fig. 7j) and significantly increased the expression of cancer stem cell-specific markers, including CD133, CD44, OCT4 and SOX2 (Supplementary Fig. 7k–m).

Next, loss-of-function experiments for LIPG were conducted. The interference efficiency of LIPG was verified via RT‒qPCR and Western blotting (Supplementary Fig. 8a, b). A CCK-8 assay revealed that LIPG knockdown significantly decreased the proliferation of HCT116 and SW480 cells (Supplementary Fig. 8c). LIPG knockdown also significantly suppressed CRC cell colony formation (Supplementary Fig. 8d). Moreover, LIPG knockdown suppressed cell migration, as shown by the results of the scratch test and transwell migration assays (Supplementary Fig. 8e, f). LIPG knockdown inhibited the expression of EMT-specific markers, which included a decrease in N-cadherin and vimentin protein levels and an increase E-cadherin protein levels (Supplementary Fig. 8g, h). To investigate whether LIPG regulated NCBP2 expression and formed a feedback loop between the two genes, NCBP2 protein levels were assessed in LIPG-overexpressing or LIPG-knockdown CRC cells. The results revealed that LIPG overexpression or knockdown did not affect the NCBP2 protein level in HCT116 and SW480 cells (Supplementary Fig. 7n, 8i). Taken together, these results indicated that LIPG promoted CRC proliferation and metastasis in vitro.

LIPG mediates the biological function of NCBP2 in CRC cells

To investigate whether LIPG functioned as a key mediator of NCBP2 function in CRC cells, LIPG was knocked down in NCBP2-overexpressing HCT116 and SW480 cells (Fig. 4a, b). NCBP2 overexpression alone promoted the proliferation of HCT116 and SW480 cells, whereas LIPG knockdown effectively inhibited NCBP2 overexpression-induced proliferation (Fig. 4c, d). In addition, the LIPG inhibitor GSK256220A significantly inhibited colony formation in NCBP2-overexpressing CRC cells (Supplementary Fig. 9a–d). Compared with NCBP2-EV CRC cells, NCBP2-overexpressing CRC cells exhibited faster wound healing; in contrast, LIPG knockdown inhibited NCBP2 overexpression-induced wound healing (Fig. 4e). Similarly, a transwell migration assay revealed that NCBP2-mediated cell migration was inhibited by LIPG knockdown (Fig. 4f). Furthermore, the LIPG inhibitor GSK256220A inhibited the migration ability of CRC cells overexpressing NCBP2 (Supplementary Fig. 9e, f). In addition, NCBP2 increased the protein levels of N-cadherin and vimentin and decreased the protein level of E-cadherin, and these effects were inhibited by LIPG knockdown (Fig. 4g, h). Collectively, these data suggested that LIPG mediated the functions of NCBP2 in CRC cells.

Fig. 4. NCBP2 promotes CRC growth and metastasis by regulating LIPG expression in CRC cells.

Western blot (a) and RT‒qPCR (b) were performed to assess the expression of NCBP2 and LIPG in CRC cells transfected with NBCP2-OE plasmid alone or combined with the LIPG-knockdown plasmid. kDa (kilo Dalton), molecular weights of standards. CCK-8 (c) and colony formation assays (d) were performed to assess CRC cell growth after NBCP2 overexpression alone or with simultaneous LIPG knockdown. Wound-healing assays (e) and transwell migration assays (f) were used to assess the migration of CRC cells with NBCP2 overexpression alone or with simultaneous LIPG knockdown. Scale bar: 50 μm. g, h Western blotting was used to assess the expression of EMT markers in CRC cells with NBCP2 overexpression alone or with simultaneous LIPG knockdown. kDa (kilo Dalton), molecular weights of standards. The CRC cells included HCT116 and SW480 cells. P values were determined by unpaired two-tailed t-test (a–h). n = 3 (a, b, d–h) and n = 5 (c) biologically independent experiments. Data are expressed as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001; NS indicates no statistical significance.

To confirm whether LIPG mediated the function of NBCP2 in CRC progression and metastasis in vivo, a xenograft mouse model was used. NCBP2 overexpression promoted subcutaneous CRC growth, which was inhibited by LIPG knockdown (Fig. 5a–c). More ki67-positive cancer cells were observed in xenograft tumours derived from NCBP2-overexpressing HCT116 cells than in those derived from NCBP2-overexpressing and LIPG-knockdown HCT116 cells (Fig. 5d–g). More importantly, LIPG knockdown in HCT116 cells significantly attenuated NCBP2-mediated lung metastasis in a mouse model (Fig. 5h, i). Taken together, these data suggested that LIPG played a critical role in the NCBP2-induced growth and lung metastasis of CRC cells both in vitro and in vivo.

Fig. 5. NCBP2 promotes CRC growth and metastasis by regulating LIPG expression in vivo.

a–c A mouse model was used to assess tumour growth after engraftment of HCT116 cells with NCBP2 overexpression alone or with simultaneous LIPG knockdown (n = 6 per group). HE staining (d), Scale bar: 50 μm; and immunohistochemical staining (e–g) were performed to assess tumour formation and the expression of Ki67, NCBP2 and LIPG in tumours (n = 6 per group). h, i A mouse model was used to assess the lung metastasis of HCT116 cells with NBCP2 overexpression alone or with simultaneous LIPG knockdown (n = 4 per group). P values were determined by one-way ANOVA with Tukey’s multiple comparison test (b, c, e–i). Data are expressed as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001; NS indicates no statistical significance.

The NCBP2–LIPG axis promotes LD formation in CRC cells

LIPG facilitates breast cancer progression and metastasis through regulating lipid metabolism and promoting LD formation²¹. Therefore, we hypothesised that NBCP2 promoted CRC progression and lung metastasis via LIPG-mediated LD accumulation. Compared with NCBP2-EV cells, NCBP2-overexpressing HCT116 cells had 925 upregulated (59.5%) and 620 downregulated (41.5%) lipid entities, and compared with NCBP2-overexpressing cells, NCBP2-overexpressing HCT116 cells treated with the LIPG inhibitor GSK264220A had 280 downregulated and 66 upregulated lipid entities (Supplementary Fig. 10a, b). Furthermore, compared with NCBP2-EV cells, NCBP2-overexpressing HCT116 cells had more glycerides (GLs), including diglycerides (DGs) and triglycerides (TGs), whereas compared with NCBP2-EV cells, NCBP2-overexpressing HCT116 cells treated with GSK264220A had significantly lower GL levels (Supplementary Fig. 10a, b). The expression of PLIN1–PLIN5 genes, which encode LD-coated proteins, was measured via RT‒qPCR. NCBP2 overexpression significantly increased PLIN2 expression in HCT116 cells (Fig. 6a) and upregulated the expression of PLIN1, PLIN2 and PLIN3 in SW480 cells (Supplementary Fig. 10c). Similarly, LIPG overexpression increased the expression of PLIN2, PLIN3 and PLIN4 in HCT116 cells (Fig. 6a) and that of all five LD-coated genes in SW480 cells (Supplementary Fig. 10c). Overexpression of NCBP2 or LIPG also increased the TG level in HCT116 and SW480 cells (Fig. 6b; Supplementary Fig. 10d); in contrast, knockdown of NCBP2 or LIPG significantly decreased the TG level in HCT116 and SW480 cells (Fig. 6c; Supplementary Fig. 10e, f). More importantly, LIPG regulated the effect of NCBP2-induced TG in CRC cells (Fig. 6e; Supplementary Fig. 10g). BODIPY 493/503 staining was used to further investigate the effects of NCBP2 and LIPG on LD accumulation in CRC cells. As shown in Fig. 6f–i and Supplementary Fig. 10h, i, HCT116 and SW480 CRC cells overexpressing NCBP2 or LIPG presented increased LD accumulation, whereas interference with NCBP2 or LIPG expression significantly decreased LD accumulation. Moreover, knockdown or inhibition of LIPG expression significantly attenuated the effect of NCBP2 overexpression on the induction of LD accumulation in HCT116 and SW480 cells (Fig. 6j; Supplementary Fig. 10j–n). Collectively, these data suggested that LIPG was a critical mediator of the effects of NCBP2 on lipid deposition and LD formation in CRC cells.

Fig. 6. NCBP2 promotes lipid droplet (LD) formation via the modulation of LIPG expression in HCT116 cells.

a RT‒qPCR was used to assess the expression of the LD-coated genes PLIN1‒PLIN5 in HCT116 cells overexpressing NCBP2 or LIPG. Triacylglycerol (TG) levels were measured in HCT116 cells with NCBP2 or LIPG overexpression (b) or in HCT116 cells with NCBP2 knockdown (c) or LIPG knockdown (d). e TG levels were measured in HCT116 cells with NCBP2 overexpression alone or with simultaneous LIPG knockdown. BODIPY 493/503 staining was performed to assess LD formation in HCT116 cells with NCBP2 overexpression (f) or LIPG overexpression (g). Scale bar: 50 μm. BODIPY 493/503 staining was performed to assess LD formation in HCT116 cells with NCBP2 knockdown (h) or LIPG knockdown (i). Scale bar: 50 μm. j BODIPY 493/503 staining was performed to assess LD formation in HCT116 cells with NCBP2 overexpression alone or with simultaneous LIPG knockdown. Scale bar: 50 μm. P values were determined by unpaired two-tailed t-test (f, g) or one-way ANOVA with Tukey’s multiple comparison test (a–e, h–j). n = 3 (a–d, f–j) and n = 4 (e) biologically independent experiments. Data are expressed as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001; NS indicates no statistical significance.

Targeting LD accumulation inhibits the functions of NCBP2 in CRC cells

DGAT1 inhibition suppresses cancer progression by inhibiting LD synthesis in cancer cells³³. Thus, we investigated whether the suppression of LD formation by a DGAT1 inhibitor could abrogate the functions of NCBP2 in HCT116 and SW480 cells. As shown in Fig. 7a, b, NCBP2 overexpression increased LD formation, and this effect was suppressed by the DGAT1 inhibitor in HCT116 and SW480 cells. Colony formation, scratch and transwell migration assays revealed that the DGAT1 inhibitor significantly inhibited the NCBP2-induced colony formation and migration of HCT116 and SW480 cells (Fig. 7c–h). These data demonstrated that the NCBP2-mediated promotion of CRC cell growth and metastasis was dependent on LD accumulation.

Fig. 7. Inhibition of LD formation inhibits NCBP2-induced growth and metastasis of CRC cells.

a, b BODIPY 493/503 staining was used to assess LD formation in CRC cells with NCBP2 overexpression alone or with simultaneous treatment with DGAT1 inhibitor. Scale bar: 50 μm. c, d Colony formation assays were conducted to assess the growth of CRC cells with NCBP2 overexpression alone or with simultaneous treatment with DGAT1 inhibitor. e–h Wound healing assays and transwell migration assays were used to assess the migration of CRC cells with NCBP2 overexpression alone or with simultaneous treatment with DGAT1 inhibitor. CRC cells included HCT116 and SW480 cells. Scale bar: 50 μm. P values were determined by one-way ANOVA with Tukey’s multiple comparison test (a–h). n = 3 biologically independent experiments. Data are expressed as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001; NS indicates no statistical significance.

High LIPG expression in CRC tissues indicates poor survival in CRC patients

The correlations between LIPG expression and the clinicopathological features of CRC patients were investigated. LIPG mRNA expression in CRC tissues was significantly higher than that in normal colorectal tissues according to the analysis of the TCGA database (Supplementary Fig. 2c, d). LIPG mRNA expression was higher in CRC tissues than in normal colorectal tissues from 29 CRC patients (Fig. 8a), and LIPG mRNA expression was positively correlated with NCBP2 expression in CRC (Fig. 8b). Similarly, compared with that in adjacent normal colorectal tissues, the LIPG protein level in CRC tissues from 97 patients was significantly higher elevated (Fig. 8c, d), and the LIPG protein level was significantly positively correlated with NCBP2 protein expression in CRC tissues (Fig. 8e). As shown in Table 1, high LIPG protein expression was associated with lymph node metastasis and advanced TNM stage. Moreover, high LIPG expression was associated with poor overall survival in CRC patients (Fig. 8f, Supplementary Fig. 2e, f; Table 2). Taken together, these data suggested that LIPG promoted cancer growth and metastasis and was an indicator of poor survival in CRC patients.

Fig. 8. LIPG expression correlates with NCBP2 expression, and a high LIPG level is indicative of poor survival in CRC patients.

a LIPG mRNA levels were assessed in 29 CRC tissues and paired normal colorectal tissues. b The correlation between LIPG mRNA level and NCBP2 mRNA level was analysed in 29 CRC tissues. c, d LIPG protein levels were measured via immunohistochemical staining in 97 CRC tissues and paired normal tissues. Scale bar: 50 μm. e The correlation between the LIPG protein level and NCBP2 protein level was analysed in 97 CRC tissues. Scale bar: 50 μm. f Kaplan–Meier curves were used to assess the effect of LIPG protein level on overall survival in 97 CRC patients. g The diagram of the mechanism shows that NCBP2 promotes the growth and metastasis of CRC cells by regulating LIPG mRNA stability via m7G modification; This image and every element of this image was created entirely by the authors using Adobe Illustrator. No third-party material was used. P values were determined by unpaired two-tailed t-test (a, d). Data are expressed as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001.

Discussion

Previous studies have demonstrated that RBPs play a central role in regulating posttranscriptional RNA modifications, primarily through their influence on transcription, splicing, RNA export, stability and translation^14,34. Some RBPs, such as NELFE and DDX39B, are overexpressed in cancer, leading to the promotion of cancer cell proliferation and distant metastasis, which results in poor survival^35,36. NCBP2, a key player in the nuclear cap-binding complex (NCCB), serves as a potential diagnostic marker for various types of cancers^18,37. Low NCBP2 expression combined with high EIF4E3 expression indicates greater immune activity and a better response to cancer immunotherapy in patients with head and neck squamous cell carcinoma¹⁹. However, whether NCBP2 promotes the progression and distant metastasis of CRC remains unclear. In this study, we elucidated the pivotal role of NCBP2 in promoting CRC progression through increasing fatty acid storage and lipid droplet accumulation mediated by LIPG overexpression. Our data revealed that NCBP2 expression in CRC tissues was significantly greater than that in matched normal colon tissues and that high NCBP2 expression in CRC was strongly correlated with cancer invasion depth, distant metastasis and advanced TNM stage. Moreover, higher NCBP2 expression was associated with poorer overall survival in CRC patients, which is consistent with the findings of a recent study on pancreatic cancer¹⁸. Indeed, NCBP2 expression is higher in pancreatic cancer tissues than in normal pancreatic tissues, and high NCBP2 expression indicates poor overall survival in patients with pancreatic cancer¹⁸. In vitro and in vivo studies demonstrated that NCBP2 overexpression promoted the proliferation and migration of HCT116 and SW480 cells, whereas NCBP2 knockdown decreased the expression of EMT markers in CRC cells. More importantly, overexpression of NCBP2 increased the expression of cancer stem cell markers in HCT116 and SW480 cells. In a murine xenograft model, overexpression of NCBP2 in HCT116 cells significantly promoted tumour growth and enhanced pulmonary metastasis. For the first time, we reported that NCBP2 promoted intracellular LD formation in CRC cells by stabilising LIPG mRNA and promoting LIPG expression through direct binding to LIPG mRNA and interaction with m7G in the cap structure of LIPG mRNA (Fig. 8G).

The 5’-cap structure containing m7G is critical for the transcription and stabilisation of many cellular mRNAs because it protects them against exonucleases¹³. Together with NCBP1, NCBP2, an important member of the nuclear cap-binding protein family, forms the CBC that orchestrates downstream RNA biogenesis^13,14,38. In CBC, NCBP1 stabilises NCBP2 and NCBP2 directly binds to mRNA by interacting with the mRNA cap structure through cap-binding pockets consisting of two aromatic amino acids³⁸. NCBP2 mediates efficient gene expression by increasing mRNA stability and has been reported to be involved in cancer progression^17–19,39. For example, NCBP2 promotes pancreatic cancer progression by regulating c-JUN mRNA stability through interactions with m7G in the 5’-cap structure¹⁸. Consistent with these findings, we observed that NCBP2 promoted CRC progression and metastasis and was closely correlated with CRC patient survival. More importantly, through mRNA sequencing of NBCP2-overexpressing or control HCT116 cells, LIPG was identified as a downstream regulator that mediated the effects of NCBP2 on CRC cells.

LDs, as cytoplasmic lipid-enriched organelles⁴⁰, are involved in various tumour-related processes, including cancer initiation, progression and metastasis^23,40. Preclinical studies have demonstrated that targeting LD accumulation is a promising strategy for inhibiting cancer progression and metastasis^33,41,42. LIPG, identified as a downstream target of NCBP2 in the present study, releases free fatty acids and lysophosphatidylcholine by cleaving phosphatidylcholine (PC) from high-density lipoprotein through its phospholipase A1 activity and thus promotes intracellular lipid accumulation and LD formation⁴³. FoxA-mediated LIPG expression is upregulated in breast cancer tissues and promotes breast cancer cell proliferation via the uptake of extracellular lipids to increase intracellular LD accumulation, which protects breast cancer cells from lipotoxicity ^21,44,45. Accordingly, we hypothesised that NCBP2 mediated lipid metabolism by increasing LIPG expression in CRC cells. Indeed, NCBP2 significantly promoted LD accumulation in CRC cells, as reflected by elevated TG levels, increased LD-coated PLIN2 expression, and increased LD staining in CRC cells. In addition, LD accumulation in CRC cells was critical for the NCBP2 overexpression-induced promotion of cancer cell proliferation and metastasis since suppressing LD formation via treatment with a DGAT1 inhibitor reversed the effects of NBCP2 on CRC cells. These data suggested that NCBP2 could be targeted to inhibit LD formation for the treatment of CRC. In addition, targeting LD formation might improve the efficacy of anticancer drugs in CRC patients with high NCBP2 expression.

These findings revealed conclusively that the expression of NCBP2 was significantly greater in CRC tissues than in normal colorectal tissues and that its increased expression was associated with an unfavourable outcome in CRC patients. Furthermore, our data demonstrated that NCBP2 promoted CRC cell proliferation and metastasis in both in vitro and in vivo models. Mechanistically, NCBP2 enhanced CRC progression by facilitating LD accumulation, a process achieved through stabilising LIPG mRNA stability and upregulating its expression (see Fig. 8g). These findings are particularly significant because they not only suggest a potential biomarker for CRC prognosis but also highlight the therapeutic promise of targeting NCBP2 in CRC treatment.

Methods

Bioinformatics analysis

Clinical data and related gene expression data for CRC patients in TCGA-COAD and TCGA-READ datasets were downloaded from The Cancer Genome Atlas (TCGA) data portal (https://portal.gdc.cancer.gov). The genes that were differentially expressed between CRC patients and healthy controls were identified as part of the analysis using the R package ‘limma’. The GSE20916⁴⁶, GSE21510⁴⁷, GSE18105⁴⁸ and GSE40967⁴⁹ datasets for CRC patients were downloaded from the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo). The cut-off values of log₂(FC) > 1 and adjusted P < 0.01 were used to screen the genes that were differentially expressed between CRC patients and normal subjects in the TCGA-COAD and TCGA-READ datasets, GSE20916, GSE21510 and GSE18105. To determine which genes were potentially associated with CRC prognosis, we identified the differentially expressed genes that overlapped with those in the TCGA-COAD and TCGA-READ datasets for the prognosis of CRC patients with hazard ratios (HRs) > 1 and adjusted P < 0.05.

We used the Human Protein Atlas (HPA) (https://www.proteinatlas. org), which maps human proteins in cells, tissues and organs, to analyse the impact of the NCBP2 protein on CRC prognosis.

Cell lines and cell culture

The cell lines used in this study (Caco-2, HCT116, SW480, SW620, HCT-8, NCM460 and 293T) were purchased from the Cell Bank of the Chinese Academy of Science (Shanghai, China). The cells were cultured in 10 cm culture dishes (Cat. 12311; Labselect, China) in Dulbecco’s modified Eagle’s medium (DMEM) (Transgene Biotech, China) supplemented with 10% foetal bovine serum (FSP500; ExCell Bio, Australia) and 1% penicillin at 37 °C in a humidified incubator containing 5% CO2. HCT116 and SW480 cells were recently authenticated using standard short tandem-repeat-based DNA profiling. The presence of mycoplasma was tested in all cell lines every month to rule out mycoplasma infection.

Plasmid construction and lentiviral transduction

Using cDNA derived from the mRNA of HCT116 cells as the template, the CDS of NCBP2 or LIPG was amplified by PCR. After enzymatic digestion, these sequences were ligated into the PiggyBac transposon vector containing a Flag tag, and successful construction of the plasmid was verified by DNA sequencing technology. The shRNA plasmid was constructed using the pLKO.1-TRC vector (#10878, Addgene plasmid, China). A specific shRNA sequence for NCBP2 or LIPG was inserted between the AgeI and EcoRI sites. The 293 T cells were transfected with target gene plasmids (3 μg) or the helper plasmids plp1, plp2 and VSVG (1 μg each) using Lipofectamine 2000 (8 μg; Yeasen Biotechnology, China). The lentivirus was harvested 2 days later. When the CRC cell density reached 60%, the lentivirus and culture mixture were added at a ratio of 1:1 to 6-well culture dishes (Cat.11110, Labselect, China), and the selection of stably transduced cells was conducted with puromycin (1:1000). The primers used to amplify the CDSs of NCBP2 and LIPG are listed in Supplementary Table 3. The sequences of the shRNAs used for NCBP2 and LIPG knockdown are listed in Supplementary Table 4.

Cell proliferation

The CRC cells were inoculated into 96-well plates (cat. 11510, Labselect, China) at a density of 2 × 10³ cells per well and incubated for 72 h. At each designated time point, the culture medium was discarded, and fresh culture solution (100 μL) was added to each well. Subsequently, 10 μL of CCK-8 solution (cat. C0038; Beyotime, China) was added for further incubation. The optical density (OD) values at 450 nm were subsequently measured using an automatic enzyme marker (SpectraMax M5; Molecular Devices, USA).

Colony formation

CRC cells were inoculated into a 6-well plate (cat.11110; Labselect, China) at 3000 cells per well and cultured for 2 weeks. The cell colonies were fixed with 4% paraformaldehyde, dyed with 1 mL of 0.5% crystal violet for 20 min, slowly washed with PBS 3 times, and counted with ImageJ software (v1.52; National Institutes of Health Freeware, USA) after imaging.

Wound healing assay

When the cell density reached 90%, a cell wound scratch assay was performed using a 10 μL pipette tip. A 6-well plate (cat.11110; Labselect, China) was placed under a microscope to select an appropriate area for monitoring and image acquisition, and corresponding marks were made. Cell wound healing was monitored after 24 h of culture, and the area of the healed region was calculated using ImageJ software (v1.52; National Institutes of Health Freeware, USA).

Cell migration assay

Cell migration was assessed in a transwell culture system placed in a 24-well plate (cat. 14341, Labselect, China). A total of 8 × 10⁴ CRC cells in each well were suspended in serum-free DMEM (100 μL) and added to the upper chamber, and 600 μL of fresh culture medium supplemented with 10% foetal bovine serum was added to the lower chamber. After incubation for 24 h, the cells on the upper side of the membrane were gently removed with a cotton swab, and the migrated cells that had migrated to the lower side of the membrane were fixed with 4% paraformaldehyde. The cells were stained with 0.5% crystal violet for 20 min, after which the migrated cells were counted using ImageJ software (v1.52; National Institutes of Health Freeware, USA).

Sphere formation assay

CRC cells with NCBP2 overexpression, LIPG overexpression, or empty vector were seeded onto 24-well Ultra-Low Attachment plates (Cat. 3473; Corning, China) at a density of 10,000 cells per well in tumour sphere medium containing DMEM (Gibco, USA) supplemented with 20 ng/mL human recombinant bFGF, 20 ng/mL EGF and 5% B27 (Gibco, USA) for 7 days. Subsequently, spheres >50 μm in size were photographed and counted with a microscope.

Animal experiments

All animal experiments were approved by the Animal Ethics Committee of the First Affiliated Hospital, University of Science and Technology of China (USTC) (No. 2021-N(A)-134) and were performed in accordance with the guidelines of the Animal Ethics Committee of the First Affiliated Hospital, USTC. Six-week-old male BALB/c nude mice were obtained from Ziyuan Experimental Animal Technology Co., Ltd. (Hangzhou, China). There were two sets of animal experiments. First, to determine whether NCBP2 promoted CRC growth, a total of 5 × 10⁶ HCT116-empty vector (NCBP2-EV) cells or NCBP2-overexpressing (NCBP2-OE) HCT116 cells were injected subcutaneously on the right side of the hind leg to monitor tumour growth (n = 6 per group). For the lung metastasis experiments, a total of 3 × 10⁶ HCT116-EV or NCBP2-OE HCT116 cells were suspended in 100 μL of saline and injected via the tail vein (n = 4 per group). Second, to determine whether LIPG mediated the functions of NCBP2 in CRC in vivo, NCBP2-EV, NCBP2-OE&LIPG-EV HCT116, or NCBP2-OE&LIPG sh#1 HCT116 cells were used as a subcutaneous xenograft model (n = 6 per group). For the lung metastasis experiments, CT116 cells were suspended in 100 μL of saline and injected via the tail vein (n = 4 per group). The long (L) and short (W) diameters of the subcutaneous tumours were measured weekly, and the tumour volume was calculated as follows: tumour volume (mm³) = length × width²/2. Both subcutaneous tumours and lung tissues were fixed with 4% paraformaldehyde and embedded in paraffin wax. Haematoxylin and eosin (HE) staining of lung tissues was conducted to analyse the metastatic capacity by counting the number of lung metastatic nodules. In addition, subcutaneous tumour tissues were stained with antibodies against Ki67 (ab15580; Abcam), NCBP2 (11950-1-AP; Proteintech, China) and LIPG (PA5-144247; Invitrogen). All the mice were kept under specific pathogen-free (SPF) conditions with a 12-h light/12-h dark cycle and were allowed to eat and drink freely.

Real-time fluorescence quantitative PCR (RT‒qPCR)

Twenty-nine paired adjacent normal and CRC tissues from CRC patients were subjected to RT‒qPCR. The baseline data of the 29 CRC patients are displayed in Supplementary Table 1. RNA isolation was conducted using a total RNA extraction kit (Yeasen Biotechnology, China), and reverse transcription to cDNA was performed using the cDNA Synthesis SuperMix for qPCR kit (Yeasen Biotechnology, China). Quantitative RT‒PCR was performed using Universal SYBR qPCR Master Mix (Biosharp, China). Relative expression was calculated via the 2-DCq method⁵⁰ and normalised to β-actin expression. The primers used are listed in Supplementary Table 5. This study was approved by the Ethics Committee of the First Affiliated Hospital, USTC (No. 2022-RE-362).

Western blotting

In brief, the cells were lysed with precooled RIPA buffer (P0013B; Beyotime Biotechnology, China) supplemented with protease inhibitors. The protein concentration was determined using a bicinchoninic acid (BCA) assay. The samples were separated on a 10% SDS‒PAGE gel and electrically transferred to PVDF membranes, which were blocked with 5% skim milk for 4 h and then incubated with the indicated primary antibodies at 4 °C overnight. The antibodies used included NCBP2 (11950-1-AP, Proteintech, China; 1:1000), NCBP2 (ab91556, Abcam; 1:1000) for IP, LIPG (PA1-16793, Invitrogen; 1:1000), N-cadherin (ab76011, Abcam; 1:1000), E-cadherin (ab314036, Abcam; 1:1000), vimentin (60330-1-Ig, Proteintech, China; 1:1000), CD133 (18470-1-AP, Proteintech, China; 1:1000), OCT4 (11263-1-AP, Proteintech, China; 1:1000), m7G (RN017M, MBL, Japan; 1:1000), Flag (F1804, Sigma; 1:1000), β-actin (380624, ZENBIO, China; 1:1000) and β-tubulin (66219-1-Ig, ZENBIO, China; 1:1000). The membrane was subsequently incubated with secondary antibody at room temperature for 2 h. The immunoreactive bands were visualised via exposure to a gel imaging system (Tanon 5200multi, Shanghai).

Nontargeted lipidomics

HCT116 cells transfected with empty vector, NCBP2-overexpressing lentivirus (NCBP2-OE), or NCBP2-OE plus the LIPG inhibitor GSK264220A were mixed with quenching agent at a 1:5 volume ratio. The cells were subsequently centrifuged at 1000 × g for 5 min, placed in liquid nitrogen for 30 s and stored at 80 °C. Cell samples (n = 8 per group) were sent to BIOTREE Biotechnology (China) for nontargeted lipidomic detection.

mRNA sequence

HCT116-vector or NCBP2-overexpressing cells were collected using TRIzol (Invitrogen), and mRNA sequencing was conducted by Shanghai Majorbio Technology Co., Ltd. (China). Each group had two replicates. The differentially expressed mRNAs were analysed on the basis of a log₂fold change (FC) > 1.0 and a p < 0.05. The sequencing data are available at the GEO repository (GSE270402).

LIPG mRNA stability assay

After cell adhesion was observed, toyocamycin (MCE, 2 μg) was added to each well to inhibit RNA synthesis. After incubation for the indicated times (0, 4, 12 and 24 h), total RNA was extracted and reverse transcribed to cDNA for RT‒qPCR.

Co-IP assay

In brief, 20 μL of anti-flag magnetic beads (B26101; Selleck, China) was incubated with 10% BSA overnight and then washed with precooled PBS 5 times. NCBP2-OE CRC cells were placed on ice and washed 3 times with precooled PBS containing protease inhibitors, after which 600 μL of IP lysis solution was added. The cells were shaken at 4 °C for 15 min, scraped off with a knife, rotated on a shaking table at 4 °C for 15 min, and then transferred to a centrifuge at 12,000 rpm for 15 min at 4 °C. After 120 μL of the sample was removed for use as the input sample, the remaining supernatant was incubated with magnetic beads overnight, after which Western blotting was performed on the prepared protein samples.

RNA immunoprecipitation (RIP) assay

The RIP experiment was conducted using a Pierce^TMMagnetic RNA‒Protein Pull-Down Kit (20164; Thermo Scientific) following the manufacturer’s instructions. In brief, 1 mL of cell lysis buffer was added to the cells. After lysis, the cell lysate was transferred to a 1.5-ml EP tube and vortexed. The EP tube was then placed on ice for 30 min and centrifuged at 12,000 rpm for 5 min, and the supernatant was collected. Fifty microlitres of magnetic protein A/G beads were labelled with IgG or target antibodies (Flag-NCBP2, NCBP2 (Cat. ab91556, Abcam) or m7G antibody) and rinsed with RIP wash buffer 3 times. Afterwards, 300 μL of RIP wash buffer was used to resuspend the magnetic beads. Five microlitres of target antibody-labelled or IgG antibody-labelled protein A/G magnetic beads was added to 300–500 μL of the supernatant and incubated at 4 °C overnight. The precipitated mRNA was eluted with elution buffer and incubated with protease K buffer at 55 °C for 30 min. TRIzol was added to each tube, and the resulting RNA was purified with chloroform and ethanol and precipitated at −80 °C. Quantitative RT‒qPCR was subsequently performed. The sequences of the primers used for RIP‒qPCR are shown in Supplementary Table 6.

Triacylglycerol (TG) measurement

The TG level in the CRC cells was measured using a triglyceride assay kit (ab65336; Abcam) following the manufacturer’s instructions. In brief, the triglyceride concentration of the cell extract was measured using a microplate reader at an OD of 570 nm. The total protein concentration of the cell extract mixture was quantified using a BCA protein concentration assay kit (Beyotime Biotechnology, Shanghai, China), and the triglyceride level was calculated as the protein concentration per mg.

BODIPY 493/503 staining

In brief, 1 × 10⁴ CRC cells were cultured in 96-well plates (cat. 11510; Labselect, China) for 24 h. The CRC cells were subsequently starved of serum-free DMEM for 6 h and cultured with fresh culture solution supplemented with 10% FBS for 1 h. The cells were fixed with 4% PFA and stained with BODIPY 493/503 (Beyotime Biotechnology, Shanghai) and Hoechst stain (Merck, Germany) for 10 min. The LDs in the CRC cells were photographed under a fluorescence microscope and counted using ImageJ software (v1.52; National Institutes of Health, USA). The fluorescence intensity of the LDs was normalised to the total CRC cell count for each group.

Immunohistochemical (IHC) staining

CRC specimens were obtained from 97 CRC patients who signed informed consent at the First Affiliated Hospital, USTC. Patients with CRC were diagnosed by pathological examination and underwent curative surgery. In addition, patients with other malignant tumours were excluded. IHC staining was conducted according to previously described methods²⁰. The primary antibodies used included an anti-NCBP2 antibody (11950-1-AP; Proteintech, China; 1:200) and an anti-LIPG antibody (PA5-144247; Invitrogen; 1:200). Two pathologists independently reviewed the slides without knowledge of the patient information. IHC staining was quantified using a scoring system ranging from 0 to 9. The staining intensity was classified as negative (score 0), weak (score 1), moderate (score 2) or strong (score 3); the percentage of positively stained CRC cells was scored as 0 (no staining), 1 (≤10% staining), 2 (10–25% staining) or 3 (>25% staining). The final staining score was calculated by multiplying the intensity score by the percentage staining score. High expression was defined as a final staining score greater than 4.

Statistical and reproducibility

All data are reported as the mean ± standard deviation (SD), and all results of the cell culture experiments were obtained from at least three independent replicates. IBM SPSS statistics 22 was used for all the statistical analyses, and GraphPad Prism 8 was used to visualise the data. All the data were compared for normality and variance, and differences between two groups or three groups were analysed by two-sided Student’s t test or one-way ANOVA with Tukey’s multiple comparison test, respectively. The association between NCBP2 expression and clinicopathological characteristics was analysed by the Pearson chi-square (χ²) test. The log-rank test and Kaplan–Meier curves were used to assess differences in the survival of CRC patients. Univariate and multivariate Cox regression analyses were used to identify predictors of 5-year survival. *P < 0.05, ^#P < 0.05; **P < 0.01, ^##P < 0.01; ***P < 0.001, ^###P < 0.001; NS, no statistical significance. No statistical methods were used to predetermine the sample size. A blinding strategy for allocation during the experiments and outcome assessment was not used, and a randomisation strategy for allocation was not adopted in this study.

Ethics approval and consent to participate

All animal experiments were approved by the Animal Care and Use Committee of the Animal Ethics Committee of the First Affiliated Hospital, USTC (ref. No. 2021-N(A)-134). We have complied with all relevant ethical regulations for animal use. The Animal Ethics Committee of our hospital permitted a maximum subcutaneous tumour volume of 2 cm³, and none of the experiments exceeded this limit. Studies involving human samples were approved by the Ethics Committee of the First Affiliated Hospital, University of Science and Technology of China (USTC) (ref. No. 2022-RE-362), and these studies were conducted in accordance with the Declaration of Helsinki. All human participants provided informed consent, and all ethical regulations relevant to human research participants were followed in this study.

Author contributions

Liu Liu, X.X. Wang and Lin Liu designed the study; Liu Liu, W. Lu, S.Y. Miao, Y.Yu and X.B. Zhang conducted the experiments and analysed the data; Liu Liu, W. Lu, Y. Yu and S.Y. Miao prepared the figures and drafted the manuscript. S.D. Ye and Lin Liu reviewed and revised the manuscript. All the authors have read and approved the final version of the manuscript.

Peer review

Peer review information

Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Georgios Giamas and Johannes Stortz. [A peer review file is available].

Data availability

All uncropped blots are provided in Supplementary information. The source data behind the graphs in the manuscript can be found in Supplementary Data 1. Data related to mRNA sequence analysis have been deposited in the GEO, and the accession number is GSE270402. All other data supporting the findings of this study are available from the corresponding author upon reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-026-09903-5.