SLPI/FBW7/c‐Myc Axis Reduces Gemcitabine Sensitivity in PDAC by Enhancing Glycolysis and Impairing NK Cell Immunity

ABSTRACT

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy with dismal prognosis. Although gemcitabine works by directly inducing tumor cell death, as well as activating NK cell‐mediated antitumor immunity, its efficacy remains rather unsatisfying. Currently, biomarkers precisely ascertaining gemcitabine resistance in PDAC are unavailable; thus, it is of much urgency to identify biomarkers accurately predicting gemcitabine sensitivity and explore innovative targets for developing sensitization strategy. Secretory leukocyte protease inhibitor (SLPI) is overexpressed in several solid tumors, but its role in PDAC remains unclear. This study aims to elucidate the function and mechanism of SLPI in modulating gemcitabine sensitivity in PDAC. In this study, we found that SLPI was markedly upregulated in PDAC cell lines and tissues, inversely correlating with gemcitabine efficacy and survival outcome of patients. In vitro, SLPI overexpression significantly reduced PDAC cell sensitivity to gemcitabine‐induced direct cytotoxicity and NK cell‐mediated lysis through enhancing c‐Myc‐driven glycolysis, whereas SLPI knockdown generated the opposite effects. Mechanistically, SLPI could directly interact with c‐Myc protein, which disturbed the binding of E3 ubiquitin ligase FBW7 to c‐Myc and thereby prevented ubiquitination‐mediated c‐Myc degradation, resultantly increasing c‐Myc stability and expression. In vivo, SLPI knockdown substantially sensitized gemcitabine treatment to suppress PDAC growth, and this effect was at least partly dependent on NK cell‐mediated antitumor immunity. Totally, our findings support that SLPI may serve as a promising biomarker for predicting gemcitabine therapy response, and targeting SLPI may be an effective strategy to improve gemcitabine efficacy in PDAC patients.

All procedures involving animals were conducted in compliance with institutional ethics regulations and approved by the Ethics Committee of the Second Hospital of Lanzhou University (Approval No. D2025‐434).

SLPI reduces the sensitivity of PDAC to gemcitabine by competing with FBW7 for binding to c‐Myc, thereby disrupting FBW7‐dependent ubiquitination and proteasomal degradation of c‐Myc, leading to increased c‐Myc stability and expression. SLPI knockdown could significantly increase gemcitabine sensitivity to gemcitabine‐induced direct cytotoxicity and NK cell‐mediated lysis via the FBW7/c‐Myc/glycolysis axis.

Abbreviations

AUC

Area under the ROC curve

ECAR

extracellular acidification rate

FBW7

F‐box/WD repeat‐containing protein 7

IC 50

half maximal inhibitory concentration

MDSC

myeloid‐derived suppressor cells

MICA/B

major histocompatibility complex class I chain‐related molecules A/B

NK cell

natural killer cell

OS

overall survival

PDAC

pancreatic ductal adenocarcinoma

ROS

reactive oxygen species

SLPI

secretory leukocyte protease inhibitor

ULBP2

UL16 binding protein 2

1. Introduction

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy [1]. Radical resection still represents the only curative therapy for this lethal disease [2, 3]. Nevertheless, a substantial proportion of PDAC patients have missed the opportunity for surgical treatment when diagnosed [4, 5]. Worse still, patients undergoing radical PDAC resection are threatened by tumor recurrence and distant metastasis within 2 years, leading to the 5‐year overall survival (OS) rate of less than 10% [6]. Apart from surgical treatment, gemcitabine, administered alone or combined with other drugs, still represents the first‐line chemotherapy strategy for PDAC [7]. Unfortunately, a large number of patients with PDAC show much poorer sensitivity to gemcitabine‐based chemotherapy regimens, hardly benefiting from this treatment [8]. Currently, clinical biomarkers precisely ascertaining whether PDAC shows resistance to gemcitabine are unavailable; thus, it is of much urgency to identify economic biomarkers accurately predicting the sensitivity of PDAC to gemcitabine and explore innovative targets for developing gemcitabine therapy sensitization strategy.

Cancer cell metabolic dysregulation has been proposed as one hallmark of cancer, and among the abnormal metabolic processes, aerobic glycolysis has attracted growing attention [9, 10, 11, 12]. Particularly, increased glycolysis in cancer cells plays a pivotal role in contributing to gemcitabine resistance in PDAC [13]. Therefore, aerobic glycolysis may represent a great potential target of sensitizing PDAC to gemcitabine, highlighting the necessity of an in‐depth search for molecular mechanisms regulating aerobic glycolysis.

Secretory leukocyte protease inhibitor (SLPI) represents a soluble inhibitor of serine proteases which is primarily found to occupy a central part in epithelial proliferation, tissue regeneration, and anti‐inflammatory responses [14, 15, 16]. Recently, increasing studies showed that SLPI expression is highly expressed across diverse malignancies and has a close relation to unfavorable oncological outcomes [17, 18, 19, 20]. SLPI is also revealed to promote cancer cell proliferation, invasion, metastasis, and chemotherapy resistance [21, 22, 23, 24]. However, it remains undefined whether increased SLPI expression reduces gemcitabine sensitivity by enhancing PDAC cell glycolytic metabolism. Therefore, this study aimed at exploring the expression status, clinicopathological value, and its role in gemcitabine sensitivity in PDAC.

2. Material and Methods

SLPI expression in PDAC and its associations with clinicopathological parameters and prognosis were analyzed using public databases and tissue microarrays. The impacts of SLPI on gemcitabine cytotoxicity and NK cell‐mediated lysis were explored in vitro using CCK‐8 assays, colony formation, flow cytometry, and lactate dehydrogenase release assays. Glucose consumption, lactate release, and extracellular acidification rate (ECAR) were utilized to evaluate glycolytic activity. The in vivo antitumor effects were validated in xenograft mouse models with gemcitabine treatment and NK cell depletion. The detailed experimental methods are provided in supplementary Doc S1.

3. Results

3.1. SLPI Expression Is Significantly Elevated and Associated With Worse Prognosis in PDAC

A comprehensive analysis across multiple cancer types utilizing data from TCGA and GTEx repositories revealed a pronounced upregulation of SLPI in various gastrointestinal tumors including PDAC (Figure 1A). Transcriptome data from 15 independent PDAC cohorts derived from the GEO and TCGA platforms further supported that SLPI was markedly upregulated in cancerous specimens compared to match adjacent normal tissues (Figure S1). Consistently, by immunohistochemical analysis we also observed an increased SLPI expression in surgically resected PDAC specimens compared to adjacent noncancerous tissues from our center and commercially available tissue microarray. In addition, SLPI expression is observed in both the cytoplasm and the nucleus of tumor cells, with a predominantly cytoplasmic localization (Figure 1B,C). As expected, SLPI expression was found to be upregulated in multiple pancreatic carcinoma cell lines (AsPC‐1, BxPC‐3, MIA PaCa‐2, SW 1990, and PANC‐1) relative to the immortalized nonmalignant pancreatic ductal epithelial line hTERT‐HPNE (Figure 1D,E).

FIGURE 1.

SLPI is overexpressed in PDAC and predicts unfavorable prognosis. (A) Pan‐cancer transcriptomic analysis using TCGA‐GTEx datasets revealed that SLPI is upregulated in several gastrointestinal malignancies, including pancreatic ductal adenocarcinoma (PDAC). Representative image of SLPI histochemistry of house PDAC case (B) and commercially tissue microarray (C). Scale bars, 50 μm; quantitative data (n = 84, p < 0.001). RT‐qPCR (D) and Western blot (E) analysis revealed elevated SLPI mRNA and protein expression in a panel of pancreatic cancer cell lines (AsPC‐1, BxPC‐3, MIA PaCa‐2, SW1990, and PANC‐1) relative to the immortalized nonmalignant pancreatic ductal epithelial line hTERT‐HPNE. (F) Kaplan–Meier survival analysis of TCGA indicated that patients with high SLPI expression exhibited shorter overall survival. (G) The prognostic evaluation on 84 PDAC cases within the tissue microarray showed that increased SLPI levels was markedly correlated with much more unfavorable prognosis (p < 0.05). (H) Kaplan–Meier curve showed that low SLPI expression predicts better survival in PDAC patients receiving postoperative gemcitabine therapy.

Then, we conducted ROC curve analysis to assess the diagnostic utility of SLPI expression in PDAC. Across 15 datasets, the AUC values range from 0.737 to 0.965 (Figure S2). These results indicate that SLPI may be used as a reliable diagnostic biomarker for PDAC. Next, its clinicopathological and prognostic values in PDAC were further explored. Combing Whitney U test and Chi‐squared test, we found that SLPI expression in tumors was positively related to serum CA19‐9 level of PDAC patients (Table 1). Analysis of Kaplan–Meier survival curves revealed a significant correlation between high SLPI expression and short OS time in PDAC patients from both the TCGA and GEO cohorts (Figure 1F and Figure S3). Similarly, the prognostic evaluation on 84 PDAC cases within the tissue microarray also showed that increased SLPI levels was markedly correlated with much more unfavorable prognosis (p < 0.05) (Figure 1G). Meanwhile, a positive association of higher SLPI expression with poorer OS was observed in PDAC patients who received postoperative gemcitabine therapy (p < 0.05) (Figure 1H). Subsequently, by univariate Cox regression analysis, we identified that CA125, CA19‐9, tumor size, T stage, distant metastasis, TNM stage, histological grade, and SLPI expression is closely linked with OS of PDAC patients (Table 2). However, multivariate Cox regression analysis confirmed that only TNM stage (stage IV vs. stage I, HR = 4.299, p = 0.025), histological grade (G3 vs. G1‐2, HR = 3.348, p < 0.001), and high SLPI expression (HR = 3.289, p = 0.001) were significantly associated with OS (Table 2), indicating that TNM stage, histological grade, and high SLPI expression may serve as independent prognostic factors in PDAC. Collectively, our data implies that SLPI is significantly overexpressed in PDAC and has the great potential as a prognostic biomarker and a predictor of gemcitabine treatment sensitivity.

TABLE 1.

Clinical features of pancreatic cancer cases grouped according to SLPI levels determined via IHC analysis on the HPanA180Su10 tissue microarray.

	ALL	Low SLPI	High SLPI	p
Age (Y)	65.5 (59.75–71.00)	65.00 (56.00–71.00)	66.00 (60.00–70.50)	0.710
Sex
Female	39 (46.43%)	15 (51.72%)	24 (43.64%)	0.480
Male	45 (53.57%)	14 (48.28%)	31 (56.36%)
CEA (ng/mL)				0.246
≦5	60 (71.43%)	23 (79.31%)	37 (67.27%)
> 5	24 (28.57%)	6 (20.69%)	18 (32.73%)
CA125 (ku/L)	19.80 (10.48–31.90)	15.40 (10.50–23.00)	22.60 (10.80–35.70)	0.111
CA19‐9 (ku/L)	406.65 (34.50–1987.50)	136.90 (22.80–694.30)	700.00 (60.90–3241.15)	0.040
Diabetes				0.555
No	64 (76.19%)	21 (72.41%)	43 (78.18%)
Yes	20 (23.81%)	8 (27.59%)	12 (21.82%)
Max diameter (cm)	3.00 (2.00–4.00)	3.00 (2.00–4.50)	3.00 (2.00–3.55)	0.879
T				0.384
T1–2	60 (71.43%)	19 (65.52%)	41 (74.55%)
T3	24 (28.57%)	10 (34.48%)	14 (25.45%)
N				0.842
0	48 (56.47%)	17 (58.62%)	31 (56.36%)
1	36 (42.86%)	12 (41.38%)	24 (43.64%)
M				0.094
0	79 (94.05)	29 (100.00%)	50 (90.91%)
1	5 (5.95%)	0 (0.00%)	5 (9.09%)
TNM stage				0.241
I	37 (44.05%)	14 (48.28%)	23 (41.82%)
II	42 (50%)	15 (51.72%)	27 (49.09%)
IV	5 (5.95%)	0 (0.00%)	5 (9.09%)
Histological grade				0.192
G1 + G2	69 (82.14%)	26 (89.66%)	43 (78.18%)
G3	15 (17.86%)	3 (10.34%)	12 (21.82%)
TIME	23.00 (8.75–53.50)	50.00 (19.00–67.00)	18.00 (6.00–44.00)	< 0.001
Survival Status				0.002
Live	28 (33.33%)	16 (55.17%)	12 (21.82%)
Dead	56 (66.67%)	13 (44.83%)	43 (78.18%)

Note: Categorical variables were analyzed using the Chi‐squared test, and continuous variables using the Mann–Whitney U test. Given the small number of patients with metastatic disease in our cohort (n = 5), we have employed Fisher's exact test to robustly assess this association.

TABLE 2.

Cox proportional hazards analysis (univariate and multivariate) of clinical and pathological variables associated with overall survival.

	Univariate analysis		Multivariate analysis
	HR (95% CI)	p	HR (95% CI)	p
Age (Y)	1.030 (0.998, 1.062)	0.066
Sex	1.284 (0.756, 2.180)	0.356
CEA	1.359 (0.775, 2.385)	0.285
CA125	1.008 (1.005, 1.012)	0.023	1.007 (1.000, 1.014)	0.056
CA19‐9	2.082 (1.049, 4.133)	0.036	1.000 (1.000, 1.000)	0.241
Diabetes	1.152 (0.629, 2.111)	0.646
Max diameter	1.258 (1.108, 1.427)	< 0.001	1.041 (0.804, 1.348)	0.760
T
T3 VS T1–2	2.134 (1.231, 3.697)	0.007	2.200 (0.859, 5.635)	0.100
N	1.335 (0.789, 2.259)	0.281
M	4.362 (1.674, 11.364)	0.003	—	—
TNM Stage
II vs. I	1.812 (1.031, 3.184)	0.039	1.251 (0.608, 2.574)	0.543
IV vs. I	6.100 (2.196, 16.944)	< 0.001	3.009 (0.930, 9.732)	0.066
Histological grade
G3 vs. G1–2	3.585 (1.925, 6.678)	< 0.001	3.348 (1.727, 6.490)	< 0.001
SLPI staining	2.512 (1.345, 4.693)	0.004	3.289 (1.590, 6.802)	0.001

3.2. SLPI Reduces the Sensitivity of PDAC Cells to Gemcitabine

To explore the influence of SLPI on chemotherapeutic agent sensitivity, we calculated the association between SLPI expression and the sensitivity of PDAC to six common chemotherapy drugs in nine clinical cohorts. As shown in Figure 2A, there was a positive link between SLPI expression and the IC50 values for six drugs including gemcitabine. To specifically analyze the impact of SLPI on the sensitivity of PDAC to gemcitabine, we first divided patients in each PDAC cohort into high‐ and low‐expression subgroups using the optimal cutoff for SLPI mRNA expression, and then compared the IC50 values for gemcitabine. The results showed that the SLPI‐high groups displayed elevated IC50 values compared with the SLPI‐low groups across all nine datasets (Figure 2B), indicating that SLPI may contribute to the reduced gemcitabine sensitivity in PDAC.

FIGURE 2.

SLPI reduces PDAC cells sensitivity to gemcitabine. (A) Heatmap depicting the correlation between SLPI mRNA expression and the predicted IC50 values for six common chemotherapy drugs across nine clinical cohorts. (B) Violin plots illustrating the consistently elevated IC50 values for gemcitabine in the SLPI‐high group relative to the SLPI‐low group. CCK‐8 assays demonstrated that SLPI knockdown significantly decreased the IC50 value of gemcitabine in SW 1990 cells (C), while SLPI overexpression in MIA PaCa‐2 cells resulted in a notable elevation (D). (E) SLPI knockdown significantly enhanced gemcitabine‐mediated inhibition on colony formation of SW 1990 cells, whereas SLPI overexpression generated the inverse effect. (F) Flow cytometry was used to evaluate apoptosis in SLPI‐knockdown SW 1990 cells and SLPI‐overexpressing MIA PaCa‐2 cells, with or without gemcitabine treatment.

To validate this result, we knocked down SLPI expression in SW 1990 cells and overexpressed it in MIA PaCa‐2 cells (Figure S4A,B), and then these cells exposed to gemcitabine or not were subjected to CCK‐8, colony formation, and flow cytometry assays. CCK‐8 assays showed that SLPI knockdown significantly increased gemcitabine sensitivity in SW 1990 cells, as evidenced by a sharply decreased IC50. In contrast, enforced SLPI expression in MIA PaCa‐2 cells resulted in a notable elevation in IC50 value (Figure 2C,D). As illustrated in Figure 2E, SLPI knockdown significantly enhanced gemcitabine‐mediated inhibition on colony formation of SW 1990 cells, whereas SLPI overexpression generated the inverse effect. Flow cytometry analysis showed that SLPI knockdown alone exerted little influence on apoptosis but markedly boosted gemcitabine‐induced apoptotic cell death in SW 1990 cells. Conversely, SLPI overexpression dramatically counteracted gemcitabine‐mediated apoptosis of MIA PaCa‐2 cells (Figure 2F). Collectively, these findings indicate that SLPI plays a pivotal role in reducing the sensitivity of PDAC cells to gemcitabine.

3.3. Glycolysis Mediates the Impact of SLPI on the Sensitivity of PDAC Cells to Gemcitabine

Evidence indicates that glycolysis is a key mechanism underlying chemoresistance in various cancer cells [25, 26]. To investigate the potential involvement of SLPI in regulating glycolytic processes, GSEA was performed using transcriptomic profiles from PDAC samples in the TCGA database. The result revealed a significant association between high SLPI expression and the enrichment of glycolysis‐related gene signatures (Figure 3A). Therefore, we further asked whether SLPI decreased the sensitivity of PDAC cells to gemcitabine by modulating glycolysis. To this end, we first examined the effect of SLPI on glycolytic activity of PDAC cells. As shown in Figure 3B,C, SLPI knockdown in SW 1990 cells significantly decreased glucose uptake and lactate secretion. Seahorse XF analysis further demonstrated that SLPI knockdown markedly reduced the ECAR, reflecting impaired basal glycolysis and glycolytic reserve (Figure 3D,E). Inversely, overexpression of SLPI in MIA PaCa‐2 cells led to increased glucose utilization, lactate output, and elevated ECAR levels (Figure 3F–I). These results demonstrated that SLPI significantly boosted glycolysis in PDAC cells. Next, MIA PaCa‐2 cells with enforced SLPI expression were exposed to the glycolysis inhibitor 2‐DG in combination with gemcitabine, and subsequently we subjected these cells to colony formation and flow cytometry assays. It was observed that SLPI overexpression could alleviate gemcitabine‐mediated suppression on PDAC cell colony formation, whereas 2‐DG treatment markedly reversed this effect (Figure 3J,K). The flow cytometry showed that SLPI overexpression significantly reduced gemcitabine‐induced apoptosis, but this effect was effectively restored by 2‐DG treatment (Figure 3L,M). Totally, these findings indicate that SLPI can augment glycolytic activity in PDAC cells, and thereby diminish the sensitivity of PDAC cells to gemcitabine.

FIGURE 3.

Glycolysis mediates the impact of SLPI on the sensitivity of PDAC cells to gemcitabine. (A) GSEA analysis revealed that glycolysis and Myc targets pathways were significantly enriched in PDAC tissues with high SLPI expression. (B–E) SLPI knockdown in SW 1990 cells markedly reduced glucose uptake (B), lactate secretion (C), and the extracellular acidification rate (ECAR) (D, E). (F–I) SLPI overexpression in MIA PaCa‐2 cells enhanced glucose uptake (F), lactate release (G), ECAR (H), and glycolytic capacity (I). (J, K) SLPI overexpression could alleviate gemcitabine‐mediated suppression on MIA PaCa‐2 cell colony formation, whereas 2‐deoxyglucose (2‐DG) treatment markedly reversed this effect. (L, M) The flow cytometry showed that SLPI overexpression significantly reduced gemcitabine‐induced apoptosis, but this effect was effectively restored by 2‐DG treatment.

3.4. SLPI Promotes Glycolytic Activity by Upregulating c‐Myc Expression in PDAC Cells

c‐Myc, as a key intracellular transcription factor, has been shown to exert oncogenic effects by regulating glycolysis [27]. Recent whole‐genome RNA sequencing and ChIP‐seq studies have further elucidated that c‐Myc upregulates nearly all genes associated with glycolysis [28]. Notably, activation of c‐Myc has been shown to confer PDAC resistance to gemcitabine through a glycolysis‐dependent mechanism [29, 30]. Given this, identifying and exploring the key molecules that regulate c‐Myc‐driven glycolysis could provide significant opportunities for sensitizing PDAC to gemcitabine. Thus, we next investigated whether SLPI promotes glycolysis in PDAC through regulating the c‐Myc signaling pathway. GSEA of TCGA datasets revealed that the c‐Myc target gene signature was significantly enriched in PDAC tissues with high SLPI expression (Figure 3A). Based on this result, we further explored the regulatory effect of SLPI on c‐Myc expression in PDAC cells. Western blotting assay demonstrated that the c‐Myc protein level was markedly decreased upon SLPI knockdown, whereas it was substantially increased by SLPI overexpression (Figure 4A). Notably, our RT‐qPCR analysis showed that SLPI modulation had little impact on c‐Myc mRNA expression (Figure 4B). The correlation analysis based on TCGA data also demonstrated that there was no significant association between SLPI and c‐Myc transcript levels (Figure 4C). These findings indicate that SLPI regulates c‐Myc expression post‐transcriptionally. Hence, we further explored the contribution of c‐Myc to SLPI‐mediated glycolysis. As shown in (Figure 4D–G), c‐Myc knockdown significantly abolished the effects of SLPI overexpression in promoting glucose uptake, lactate release, and ECAR of MIA PaCa‐2 cells. Conversely, c‐Myc overexpression dramatically reversed the roles of SLPI knockdown in reducing glucose uptake, lactate release, and ECAR of SW 1990 cells (Figure 4H–K). It has been revealed that c‐Myc can directly transcriptionally activate the expressions of several glycolysis‐related key enzymes such as HK2, LDHA, and GLUT1, consequently firing glycolysis in PDAC cells [31]. Consistently, we demonstrated that SLPI knockdown significantly inhibited the expressions of HK2, LDHA, and GLUT1 in PDAC cells, whereas c‐Myc overexpression markedly reversed this effect (Figure 4L,M). Collectively, these results demonstrated that c‐Myc represents a key mediator of SLPI‐mediated glycolytic reprogramming in PDAC cells.

FIGURE 4.

SLPI promotes glycolytic activity by upregulating c‐Myc protein expression. (A) Western blot analysis of c‐Myc protein levels in SLPI‐knockdown SW 1990 cells and SLPI‐overexpressing MIA PaCa‐2 cells. (B) RT‐qPCR analysis showed that SLPI modulation had little impact on c‐Myc mRNA expression. (C) Correlation between SLPI and c‐Myc mRNA expression based on TCGA datasets. (D–G) Measurements of glucose consumption (D), lactate production (E), ECAR (F), and glycolytic reserve (G) in MIA PaCa‐2 cells overexpressing SLPI, with or without c‐Myc knockdown. (H–K) Evaluation of glucose uptake (H), lactate secretion (I), ECAR (J), and glycolytic capacity (K) in SLPI‐knockdown SW 1990 cells, with or without c‐Myc overexpression. (L, M) RT‐qPCR analysis of HK2, LDHA, and GLUT1 expression in cells with enforced c‐Myc expression (L) or c‐Myc knockdown (M). Data are expressed as mean ± SD.

3.5. SLPI Suppresses FBW7‐Dependent Ubiquitination and Proteasomal Degradation of c‐Myc

Ubiquitination represents a key post‐translational modification facilitating proteasomal‐mediated degradation of proteins to modulate protein stability and expression [32]. FBW7 (F‐box/WD repeat‐containing protein 7), a tumor suppressor in diverse malignancies including PDAC, has been considered an essential E3 ubiquitin ligase that promotes the ubiquitination and proteasome degradation of many oncogenes including c‐Myc [33]. Interestingly, a study by Taggart et al. reported that SLPI could participate in regulating the proteasomal degradation of protein [34]. Moreover, in silico molecular docking analysis revealed that both SLPI and FBW7 could bind to c‐Myc, with partially overlapping binding regions located within the central domain of c‐Myc (Figure 5A). Therefore, we performed a series of experiments to explore whether SLPI competes with FBW7 for binding to c‐Myc to disturb FBW7‐dependent ubiquitination and proteasomal degradation of c‐Myc, resultantly upregulating c‐Myc expression. The endogenous Co‐IP assays confirmed that SLPI directly interacted with c‐Myc (Figure 5B), and SLPI knockdown significantly enhanced the binding of FBW7 to c‐Myc (Figure 5C). However, SLPI had no impact on FBW7 expression in PDAC cells (Figure 5D). Accordingly, we continued to explore the effect of SLPI on c‐Myc ubiquitination in PDAC cells. As shown in (Figure 5E), SLPI knockdown was found to increase the ubiquitination level of c‐Myc, but this effect was markedly reversed by simultaneously silencing FBW7. Conversely, SLPI overexpression in PDAC cells significantly reduced c‐Myc ubiquitination, which was abrogated by co‐overexpression of FBW7 (Figure 5F). As expected, CHX chase experiments further demonstrated that SLPI knockdown significantly reduced c‐Myc protein stability in SW 1990 cells, whereas simultaneously silencing FBW7 dramatically mitigated this effect (Figure 5G). In contrast, overexpression of SLPI in MIA PaCa‐2 cells markedly elevated c‐Myc protein stability, but co‐overexpression of FBW7 could effectively reverse this effect (Figure 5H). More importantly, it was observed that the proteasome inhibitor MG132 treatment substantially neutralized the effect of SLPI knockdown in downregulating c‐Myc protein expression (Figure 5I). Overall, these findings support that SLPI stabilizes c‐Myc protein by directly binding to c‐Myc and competitively inhibiting FBW7‐mediated ubiquitination in PDAC cells.

FIGURE 5.

SLPI enhances c‐Myc stability by suppressing FBW7‐mediated ubiquitination and degradation. (A) In silico docking analysis illustrating potential binding interfaces among c‐Myc, FBW7, and SLPI. (B) Co‐IP confirms direct interaction between SLPI and c‐Myc in SW 1990 cells. (C) SLPI knockdown markedly enhanced the binding of FBW7 to c‐Myc in SW 1990 cells. (D) Western blot showed that SLPI manipulation does not affect FBW7 protein abundance. (E) Analysis of c‐Myc ubiquitination by immunoprecipitation in SLPI knockdown SW 1990 cells under conditions of FBW7 presence or silence, following MG132 exposure. (F) Ubiquitination analysis of c‐Myc in MIA PaCa‐2 cells with SLPI overexpression, with or without FBW7 coexpression, under MG132 exposure. (G) Cycloheximide (CHX) chase assay assessing c‐Myc stability in SW 1990 cells following SLPI knockdown, with or without FBW7 knockdown, at 0, 4, 8, and 16 h. (H) CHX chase in MIA PaCa‐2 cells overexpressing SLPI, with or without FBW7 coexpression, to evaluate c‐Myc degradation kinetics. (I) Western blot detection of c‐Myc protein levels in SW 1990 cells following MG132 treatment (10 μM, 4 h), comparing control and SLPI knockdown conditions.

3.6. SLPI Inhibits NK Cell Cytotoxicity Partly Through Promoting c‐Myc‐Mediated Glycolysis

Apart from direct cytotoxic activity, gemcitabine has also been revealed to augment NK cell antitumor immunity, which makes a great contribution to therapeutic efficacy as well. Of note, solid evidence shows that lactic acid, a key glycolytic product, can cause a direct inhibition on NK cell cytotoxicity via intricate mechanisms. Therefore, we were interested in whether SLPI in PDAC cells could repress NK cell cytotoxicity in a c‐Myc‐mediated glycolysis‐dependent manner. As shown in (Figure 6A–C), SLPI knockdown significantly enhanced NK cell‐mediated lysis of SW 1990 cells, but this effect was dramatically mitigated by the concurrent upregulation of c‐Myc. Meanwhile, SLPI overexpression prominently reduced the sensitivity of MIA PaCa‐2 cells to NK cell‐mediated killing, whereas c‐Myc knockdown significantly restored the susceptibility of these cells to NK cell‐mediated lysis (Figure 6D–F). Consistent with this result, we further demonstrated that glycolysis inhibitor 2‐DG also attenuated the effect of SLPI overexpression in impairing NK cell cytotoxicity. Interestingly, 2‐DG could only partially reverse SLPI overexpression‐mediated inhibitory effect on NK cell cytotoxicity, and its effect in this regard was much weaker than c‐Myc knockdown. Collectively, our data implies that SLPI inhibits NK cell cytotoxicity at least partly through promoting c‐Myc–mediated glycolysis.

FIGURE 6.

SLPI inhibits NK cell cytotoxicity partly through promoting c‐Myc‐mediated glycolysis. (A) Cytotoxic activity of NK‐92 cells against SW 1990 cells was determined by LDH release assay at effector‐to‐target (E:T) ratios of 0:1, 1:1, and 5:1. (B) The viability of SW 1990 cells following NK‐92 cell treatment was further assessed by Calcein AM fluorescence staining, and representative fluorescence images are shown. (C) Quantitative analysis of the Calcein AM staining assay results is presented as a bar graph. (D) Cytotoxic activity of NK‐92 cells against MIA PaCa‐2 was assessed by a LDH release assay. (E) Fluorescence images of MIA PaCa‐2 cells stained with Calcein AM. (F) Quantification of the Calcein AM assay is presented as a bar graph. Data were analyzed by one‐way ANOVA.

3.7. SLPI Knockdown Sensitizes Gemcitabine to Inhibit PDAC in Vivo

To further confirm the effect of SLPI on gemcitabine efficacy in PDAC and its corresponding mechanisms, we established a subcutaneous xenograft model using SW 1990 cells with stable SLPI knockdown or not in BALB/c nude mice, and these mice were randomly allocated into five groups with different treatments (Figure 7A). It was observed that SLPI knockdown significantly enhanced gemcitabine to reduce the volume and weight of PDAC in mice (Figure 7B–D). Nevertheless, NK cell depletion with antiasialo‐GM1 antibody substantially abrogated this effect. These findings confirmed that SLPI knockdown could improve the sensitivity of PDAC to gemcitabine, and NK cell‐mediated antitumor immunity is involved in this phenomenon. Besides, our immunohistochemistry validated that SLPI knockdown and gemcitabine cooperatively decreased Ki‐67 expression and increased cleaved caspase‐3 expression in PDAC tissues (Figure 7E), which is attributed to the direct cytotoxicity of gemcitabine to tumor cells in terms of proliferation and apoptosis. Collectively, these findings demonstrate that SLPI knockdown could increase the sensitivity of PDAC cells to gemcitabine‐induced direct cytotoxicity and NK cell‐mediated lysis via FBW7/c‐Myc/glycolysis axis, consequently improving gemcitabine efficacy (Figure 8).

FIGURE 7.

SLPI knockdown sensitizes gemcitabine to inhibit PDAC in vivo. (A) Schematic diagram outlining the in vivo experimental design. (B) Visual presentation of subcutaneous tumor xenografts obtained from various experimental treatment cohorts. (C) Tumor growth curves over time. (D) Final tumor weights measured on day 32. (E) Representative images of H&E staining and immunohistochemical staining for Ki67 and cleaved caspase‐3. Scale bars, 200 μm.

FIGURE 8.

Graphical summary illustrating the mechanism by which SLPI reduces the sensitivity of PDAC to gemcitabine. SLPI competes with FBW7 for binding to c‐Myc, thereby disrupting FBW7‐dependent ubiquitination and proteasomal degradation of c‐Myc, leading to increased c‐Myc stability and expression. SLPI knockdown could significantly increase PDAC cell sensitivity to gemcitabine‐induced direct cytotoxicity and NK cell‐mediated lysis via FBW7/c‐Myc/glycolysis axis, consequently augmenting gemcitabine efficacy.

4. Discussion

In this study, SLPI was found to be markedly overexpressed in PDAC and closely associated with poor prognosis in patients receiving postoperative gemcitabine treatment. Notably, we observed a compelling link between SLPI and metastatic progression. SLPI expression was most pronounced in SW 1990 cells (derived from splenic metastasis) and AsPC‐1 cells (derived from ascites of metastatic PDAC). Consistently, clinical data from the tissue microarray analysis demonstrated that all five patients with metastasis (M1) exhibited high SLPI expression, although the difference was not statistically significant, possibly due to the limited number of metastatic events in the cohort. This biological trend aligns with a prior report in a breast cancer mouse model, where SLPI promoted metastasis by enhancing intravasation and exerting anticoagulant function [23]. While this underscores the plausible biological role of SLPI in metastasis, its specific mechanisms and contribution to metastatic progression in PDAC warrant dedicated future investigation.

Moreover, our data demonstrated that SLPI could reduce the sensitivity of PDAC to gemcitabine by enhancing glycolysis in cancer cells. The contribution of glycolysis to gemcitabine resistance in PDAC involves multiple mechanisms. It has been disclosed that gemcitabine can impede DNA replication to induce PDAC cell tumor growth arrest [35]. Conversely, enhanced glycolysis in cancer cells augments pyrimidine biosynthesis to increase the intrinsic levels of deoxycytidine triphosphate, which can competitively suppress gemcitabine activity and resultantly elicit PDAC resistance to gemcitabine therapy [36]. In addition to DNA replication inhibition, gemcitabine can also cause PDAC cell death through triggering oxidative stress to induce DNA damage [37]. Of interest, glycolysis in several kinds of cancer cells has been demonstrated to promote DNA damage repair and thereby cause chemotherapy resistance [38]. These pieces of evidence suggest that glycolysis‐meditated DNA damage repair may also make a contribution to the resistance of PDAC to gemcitabine, which deserves in‐depth research in future. Apart from drug‐meditated direct cytotoxic activity, gemcitabine can also enhance NK cell antitumor immunity by killing myeloid‐derived suppressor cells (MDSCs), as well as upregulating the expressions of NK cell activating ligands such as MICA/B and ULBPs [39]. Notably, lactate, a key glycolytic product, can directly inhibit NK cell cytotoxicity through disturbing nicotinamide adenine dinucleotide metabolism, impairing mitochondria, and downregulating nuclear factor of activated T cells (NFAT) to limit IFN‐γ production in NK cells [40]. Consistent with this knowledge, we found that SLPI overexpression significantly inhibited NK cell‐mediated killing of PDAC cells in vitro, whereas glycolysis inhibitor 2‐DG could attenuate this effect. Per contra, SLPI knockdown in PDAC cells dramatically augmented NK cell activity. Interestingly, glycolysis inhibition by 2‐DG only partially reversed SLPI overexpression‐meditated inhibitory effect on NK cell cytotoxicity, implying that the mechanisms independent of glycolysis also involve the diminished NK cell cytotoxicity caused by SLPI upregulation in PDAC cells. Importantly, our in vivo experiments demonstrated that SLPI knockdown could sensitize gemcitabine to retard PDAC growth, whereas NK cell depletion significantly abrogated this therapeutic benefit. Overall, our findings suggest that aberrant SLPI upregulation restricts gemcitabine therapy efficacy in PDAC through strengthening glycolysis to promote PDAC cell resistance to direct killing by gemcitabine and NK cells, and amplifying NK cell cytotoxicity to PDAC cells independent of glycolysis.

In this research, we for the first time found that SLPI augmented PDAC cell glycolysis by upregulating c‐Myc expression. Particularly, we demonstrated that SLPI overexpression upregulated c‐Myc protein level without altering c‐Myc mRNA expression, indicating the post‐transcriptional regulation of SLPI on c‐Myc expression. The ubiquitination is one of post‐transcriptional protein modification, which plays a pivotal role in facilitating the proteasomal degradation of proteins. Concerning this point, we explored whether SLPI increased c‐Myc protein expression by modulating its ubiquitination. Expectedly, our data showed that SLPI overexpression prominently reduced the ubiquitination level of c‐Myc but elevated its stability, whereas SLPI knockdown generated inverse effects. Moreover, we further verified that the proteasome inhibitor MG132 could effectively reverse the effect of SLPI knockdown in diminishing c‐Myc expression. Obviously, these findings indicate that SLPI upregulates c‐Myc expression by repressing the ubiquitination and proteasomal degradation of c‐Myc protein. FBW7, an E3 ubiquitin ligase, has been recognized as a tumor suppressor in various cancers including PDAC [41, 42]. Remarkably, previous studies have confirmed that FBW7 can boost proteasome‐mediated c‐Myc degradation by increasing the ubiquitination of c‐Myc in PDAC cells [43, 44]. Therefore, we asked whether SLPI increased c‐Myc expression through interfering with FBW7‐mediated ubiquitination and proteasomal degradation of c‐Myc. As our data showed, SLPI did not alter FBW7 protein levels but markedly blocked the binding of FBW7 to c‐Myc. FBW7 overexpression significantly mitigated SLPI overexpression‐mediated effects in reducing the ubiquitination levels, enhancing the stability, and increasing the expression of c‐Myc protein. By coimmunoprecipitation assays, we confirmed a direct interaction between SLPI and c‐Myc. Moreover, in silico docking analysis showed that there was an overlap between the SLPI‐interacting domain and the FBW7‐interacting domain of c‐Myc. Totally, these results suggest that SLPI inhibits the ubiquitination and proteasomal degradation of c‐Myc through competing with FBW7 for binding to c‐Myc, consequently increasing c‐Myc expression to drive glycolysis in PDAC. Apart from the enhancement of glycolysis, c‐Myc was also revealed to suppress NK cell cytotoxicity through downregulating the expressions of natural killer group 2 member D (NKG2D) ligands including MICA, MICB, and ULBP1 on tumor cell surface [45]. Therefore, it will be of much interest to determine whether the c‐Myc/NKG2D ligands axis involves in the ability of SLPI to mitigate NK cell cytotoxicity to PDAC cells in our future research.

This study has several limitations that should be considered. First, our in vivo findings rely on the use of antiasialo‐GM1 antibody for NK cell depletion in BALB/c nude mice. While this model addresses concerns about concomitant T‐cell depletion, a known off‐target effect of the antibody in immunocompetent hosts, the simplified immune landscape of these mice may not fully capture the complex interactions within an intact tumor microenvironment. Therefore, while our model isolates the role of NK cells, future validation in immunocompetent models is essential to better assess the translational relevance of our findings. Second, although we establish SLPI as a robust tissue‐based prognostic biomarker, our study does not provide experimental data on its detectability in serum or other liquid biopsies. Given that SLPI is a secreted protein, there is a strong rationale for investigating its potential as a noninvasive circulating biomarker. However, claims regarding its utility in this context remain speculative and warrant further investigation in future translational studies.

In summary, SLPI was highly expressed in PDAC tissues, and its high expression predicted frustrating survival prognosis in patients receiving postresection gemcitabine chemotherapy for PDAC. Moreover, SLPI knockdown could significantly increase PDAC cell sensitivity to gemcitabine‐induced direct cytotoxicity and NK cell‐mediated lysis via the FBW7/c‐Myc/glycolysis axis, consequently augmenting gemcitabine efficacy. Therefore, SLPI may serve as a promising biomarker for predicting gemcitabine therapy response, and targeting SLPI may be an effective strategy to improve gemcitabine efficacy in PDAC patients.

Author Contributions

Haofei Chen: conceptualization, investigation, data curation, formal analysis, visualization, writing – original draft. Xin Li: methodology, investigation, writing – original draft. Weixiong Zhu: validation, writing – original draft. Yanxi Mu: validation, writing – original draft. Guoqing Zhang: data curation, investigation, writing – original draft. Youheng Zhang: writing – original draft, investigation, data curation. Yusheng Cheng: conceptualization, supervision, validation, writing – review and editing. Wence Zhou: funding acquisition, project administration, conceptualization, writing – review and editing.

Funding

Cuiying Science and Technology Innovation Project of the Second Hospital of Lanzhou University, CY2024‐CQ‐01. National Natural Science Foundation of China, 82260555.

Ethics Statement

All procedures involving human tissues and animals were conducted in compliance with institutional ethics regulations and approved by the Ethics Committee of the Second Hospital of Lanzhou University (Approval No. 2025A‐705).

Animal Studies: All procedures involving animals were conducted in compliance with institutional ethics regulations and approved by the Ethics Committee of the Second Hospital of Lanzhou University (Approval No. D2025‐434).

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Transcriptome analysis revealed a significant upregulation of SLPI in cancerous tissues compared to noncancerous tissues across 15 independent PDAC cohorts.

Figure S2: ROC curve analysis of SLPI as a PDAC biomarker.

Figure S3: Survival curves revealed a significant correlation between high SLPI expression and short overall survival time in PDAC patients.

Figure S4: RT‐qPCR analysis verified successful genetic manipulation of SLPI.

Data S1: Supporting Information and Methods.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.