CREPT serves as a biomarker of poor survival in pancreatic ductal adenocarcinoma

Abstract

Background

Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive human malignancies. Cell-cycle-related and expression-elevated protein in tumor (CREPT) plays an important role in the phosphorylation of RNA Pol II, and has been implicated in the development of several types of cancer. As yet, however, there have been no reports on its role in PDAC. Here, we aimed to explore the value of CREPT as a prognostic biomarker in PDAC.

Methods

CREPT expression was assessed by immunohistochemistry (IHC) on a tissue microarray containing samples from 375 PDAC patients. Kaplan-Meier and Cox regression analyses were performed to explore the independent prognostic value of CREPT expression for the disease-free survival (DFS) and overall survival (OS) of PDAC patients. A Cell Counting Kit-8 (CCK8) assay was used to determine the growth rates and gemcitabine sensitivities of PDAC cells, while a Transwell assay was used to determine the migration and invasion abilities of PDAC cells. Subcutaneous xenografts were used to explore the effect of CREPT expression on tumor growth in vivo.

Results

We found that CREPT is highly expressed in tumor tissues and may serve as an independent prognostic biomarker for DFS and OS of PDAC patients. In vitro assays revealed that CREPT expression promotes the proliferation, migration, invasion and gemcitabine resistance of PDAC cells, and in vivo assays showed that CREPT expression knockdown led to inhibition of PDAC tumor growth.

Conclusions

We conclude that high CREPT expression enhances the proliferation, migration, invasion and gemcitabine resistance of PDAC cells. In addition, we conclude that CREPT may serve as an independent prognostic biomarker and therapeutic target for PDAC patients.

Electronic supplementary material

The online version of this article (10.1007/s13402-020-00569-7) contains supplementary material, which is available to authorized users.

Introduction

Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal human malignancies. According to data from the American Cancer Society, PDAC is 4th most common cause of tumor-related death and has a 5-year survival rate of < 9% [1]. Late diagnosis and chemoresistance are the main causes of a poor prognosis for PDAC patients [2, 3]. In order to improve the prognosis, several biomarkers have been explored to guide current treatment options or to develop new therapeutic targets for PDAC [4, 5]. However, only CA19–9, which has been approved by the US FDA, is currently widely used as diagnostic and prognostic PDAC marker in the clinic. Approximately 5% to 14% of all people, however, lack Lewis antigen glycosyltransferase and cannot synthesize CA19–9, rendering serum CA19–9 undetectable in these patients [6]. Moreover, false positive CA19–9 detection has been found to occur frequently in patients with benign pancreatic disease coupled with obstructive jaundice [7, 8]. Therefore, it is of great clinical significance to continue to search for effective biomarkers and therapeutic targets to improve the prognosis of patients with PDAC.

It has been reported that cell-cycle-related and expression-elevated protein in tumor (CREPT) may be associated with gastrointestinal tumor prognosis. The gene encoding CREPT, also known as RPRD1B (regulation of the nuclear precursor RNA domain containing 1B), is located on chromosome 20. CREPT consists of 326 amino acids, belongs to the RPRD family (a family of cell cycle-dependent kinase inhibitor-associated proteins) and binds to the phosphorylation site of RNA polymerase II (RNA Pol II) [9]. RNA Pol II catalyzes transcription and leads to the synthesis of most mRNA, snRNA and miRNA precursors. RNA Pol II is a large 550-kDa complex containing 12 subunits, one of which is the RPB1 subunit, which has a specific C-terminal domain (CTD) with a highly repetitive Tyr1-Ser2-Pro3-Thr4-Ser5-Pro6-Ser7 sequence. During transcription, phosphorylation of Tyr1, Ser2, Thr4, Ser5 and Ser7 leads the RNA Pol II CTD to produce multiple phenotypes [10]. These different phosphorylation phenotypes and proline isomers of RNA Pol II CTD lead to the recruitment of different transcription factors, which affects the posttranscriptional modification of mRNAs [11, 12]. CREPT, which plays an important role in phosphorylation of the RNA Pol II CTD, has been reported to affect cell cycle regulation [13].

Relationships between CREPT expression levels and clinicopathological features have amply been studied. Wang et al. studied 76 patients with endometrial cancer and found that CREPT expression was significantly higher in cancer tissues than in normal endometrial tissues and that differences in expression were related to the pathological type of the tumor, tumor stage and tumor invasion [14]. Similar results have been reported for gastric cancer and retroperitoneal leiomyosarcoma [9, 15], As yet, however, the role of CREPT in PDAC has remained unclear. Here, we evaluated CREPT protein expression by immunohistochemistry using a PDAC tissue microarray and analyzed its relationship with various clinicopathological features. We also explored the effect of CREPT expression on the biological behavior of PDAC cells in vitro and in vivo. Our data may provide a basis for the feasibility of CREPT as a prognostic biomarker and therapeutic target for PDAC.

Materials and methods

Patient samples

In total 375 PDAC tissue samples were collected from patients who underwent surgery from September 2004 to December 2014 at the Department of General Surgery, Peking Union Medical College Hospital, Beijing, P. R. China. The inclusion criteria were as follows: (1) patients who underwent R0 surgical resection, (2) patients for whom postoperative pathologic diagnosis was clearly pancreatic ductal adenocarcinoma, (3) patients with surgical tissue samples that were suitable for immunohistochemical analysis, (4) patients who did not receive neoadjuvant chemotherapy or radiotherapy before surgery, (5) patients for whom clinical and pathological data were complete, (6) patients who started chemotherapy within 8 weeks after surgery and (7) patients who underwent regular follow-up involving detection of CA19–9 levels, B-ultrasound and/or CT scans. The exclusion criteria were as follows: (1) patients with a history of serious diseases, such as coronary sclerosing heart disease, cerebral diseases, or benign pancreatic, biliary or hepatic disease, (2) patients who experienced serious complications during the perioperative period, such as grade C pancreatic fistula, severe abdominal infection, postoperative severe abdominal bleeding or death and (3) patients without follow-up information. Clinicopathological data were collected from each patient. Tumor stages were classified according to the 7th edition of the American Joint Committee on Cancer TNM classification. Overall survival (OS) was defined as the interval between the date of surgery and the date of death. This research was approved by the Research Medical Ethics Committee of Peking Union Medical College Hospital, and informed consent was obtained from each patient.

Tissue microarray construction and hematoxylin-eosin staining

The tumor tissues and adjacent normal tissues of the enrolled patients were selected and arrayed in sequence. A tissue microarray (TMA) table was designed, and the tissue samples were arranged according to the TMA table. Hematoxylin-eosin (HE) stained sections were evaluated by an experienced pathologist who scored the extent of the diseased tissues. TMAs were prepared using a tissue arrayer (MTA1 Tissue Arrayer) and sectioned using an automated microtome (Leica RM2165 SN. 0903/06–2002). The same TMA was used for HE staining once per 15 sheets. A pathologist conducted quality control of each TMA to determine the pathology type, classification, defect rate and other information.

Immunohistochemical analysis and evaluation

According to a standard immunohistochemistry (IHC) protocol, the TMA samples were prepared using a PV-9000 IHC kit (ZSGB-BIO, Beijing, China) and incubated with an anti-C20orf77 antibody (ab154910; Abcam, Cambridge, UK). Subsequently, DAB from the PV-9000 kit and Harris hematoxylin (REDPHARM, Beijing, China) were employed. Two experienced pathologists independently assessed the results. Staining intensities were graded as 0 (negative), 1 (low), 2 (medium) or 3 (high), while the staining extent was scored from 0 to 100%. Both staining intensity and positive percentage were considered to determine composite expression scores (CESs), and calculated by the following formula: CES = intensity score × percentage score × 100. The average CES (179.9) was chosen to distinguish high CREPT expression from low CREPT expression.

Cell lines and culture conditions

Five human PDAC-derived cell lines, PANC-1 (RRID: CLCV_0480), AsPC-1 (RRID: CLCV_0152, MIA PaCa-2 (RRID: CLCV_0428), SU.86.86 (RRID: CLCV_3881), BxPC-3 (RRID: CLCV_0186) and T3M4 (RRID: CLCV_4056), were provided by Professor Freiss from the University of Heidelberg, Germany. All cell lines were authenticated using STR (or SNP) profiling within the last three years. All experiments were performed with mycoplasma-free cells. The cells were cultured in RPMI-1640 medium supplemented with antibiotics and 10% fetal bovine serum at 37 °C in a humidified atmosphere containing 5% CO₂. The culture medium and supplements were purchased from HyClone (Northbrook, IL, USA).

Quantitative RT-PCR

Total RNA was extracted from PDAC-derived cell lines by TRIzol (Invitrogen, Carlsbad, CA, USA) and processed for reverse transcription and quantitative PCR using a Reverse Transcription System (Promega, Madison, WI, USA) and a One Step SYBR® PrimeScript™ RT-PCR Kit (Takara, Tokyo, Japan) according to the manufacturer’s instructions.

Western blot analysis

Protein samples were obtained from PDAC-derived cell lines, separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membranes. Next, the membranes were incubated with primary antibodies (1:200–1:1000) followed by a secondary antibody (1:5000–1:10000). Proteins were visualized using chemiluminescence (ECL) reagents and a FluoChem SP ECL gel imager (Alpha Innotech, USA). The images were analyzed using Image-Pro Plus 6.0.

Plasmid construction for CREPT knockdown

The gene sequence encoding CREPT was retreived by searching the National Center of Biotechnology Information (NCBI) database. Three shRNAs were designed based on the CREPT gene sequence, CREPT shRNA01: 5’-GCAACAGAAGAGAAGAAATCTCGAAAGATTTCTTC TCTTCTGTTGC-3′; CREPT shRNA02: 5’-GGGCAAGTCATGGTTGGAAATCGAAATTTCCA ACCATGACTTGCCC-3′ and CREPT shRNA03: 5’-GCACGAAGATTAGGTGCATTTCGAAAA ATGCACCTAATCTTCGTGC-3′. The three shRNAs were inserted into the lentiviral vector plasmid “pLV-U6-BsaI-EF1α-EGFP-2A-Puro” at the BsaI site to construct the respective lentiviral gene silencing plasmids. The lentiviruses were packaged, and the titers were determined. Stably expressing cell lines were selected and identified as previously described [16].

Cell Counting Kit-8 assay

The Cell Counting Kit-8 (CCK8) assay is widely used to evaluate cell proliferation and cytotoxicity. CCK8 reagent (DOJINDO, Kyushu, Japan) was added to PDAC cells treated for 0, 24, 48, 72 or 96 h and incubated for 2.5 h. Next, optical densities at 450 nm (OD450) were measured using a microplate reader (Spectra Max 190, Molecular Devices, USA) and used to draw cell growth curves. To assess the sensitivity of PDAC cells to gemcitabine, a gradient concentration of gemcitabine was added, after which the OD was used to calculate the inhibition rate (inhibiting rate = 1-(treated group OD-blank group OD) / (control group OD-blank group OD)).

Migration and invasion assays

For the migration assays, a Transwell system (Corning, USA) was used according to the manufacturer’s instructions. Briefly, 6 × 10⁴ cells in serum-free culture medium were seeded into the upper chamber, whereas the lower chamber was filled with culture medium containing 10% FBS (HyClone, Northbrook, IL, USA). For the invasion assays, polycarbonate membranes with 8-μm pores were coated with 0.5 mg/ml Matrigel (Sigma-Aldrich, Saint Louis, MI, USA) for 6 h. After the membranes were rehydrated, cancer cells in serum-free culture medium were seeded into the upper chamber of the Transwell unit, whereas the lower chamber was filled with culture medium containing 10% FBS (HyClone, Northbrook, IL, USA). After 24 h, the migrated or invaded cells were fixed by 4% paraformaldehyde and stained with hematoxylin and eosin. Next, after the membranes were dried, the number of invading cells was counted in five selected (above, below, left, right and middle) microscopic fields. Mean values were determined in triplicate assays.

Mouse xenograft model

We used 4–6-week-old female nude BALB/c mice (Charles River, Beijing, China). The breeding environment was specific pathogen-free (SPF) grade, and the number of mice per cage was less than 5. T3M4 cells stably expressing shCREPT were used as the shCREP group, whereas the control group was labeled as shCongroup. After the skin was disinfected with alcohol, the cells were inoculated into the nude mice at the armpits (8 × 10⁶ cells/mouse). The lengths and widths of the tumors were measured twice a week, and the tumor volumes were calculated as [V (volume) = π / 6 (W1 × W2 × W2)], where W1 represents the maximum diameter of the tumor and W2 represents the smallest diameter of the tumor. All the animal experiments were approved by the Ethics Committee of Peking Union Medical College Hospital.

Statistical analysis

The data were analyzed using Student’s t test or analysis of variance (ANOVA). The Chi-square test was used for counting data. Fisher’s exact test was used for variables with an expected value < 5. Kaplan-Meier single factor analysis was used to evaluate differences in survival times between groups by the Log-rank method. Multivariate analysis was used to assess relative risks by the Cox proportional hazards regression model. SPSS 24.0 was used for statistics, and p < 0.05 was considered significant. GraphPad Prism 8.2 was used to generate graphs.

Results

Clinicopathological feature of the patients

In total 375 patients were enrolled in our study including 213 males (56.8%) and 162 females (43.2%) with a mean age of 60.3 years (95% CI 59.2–61.3). The tumors were located in the head of the pancreas in 157 patients (41.9%), and these patients underwent pancreaticoduodenectomy, while 218 patients (58.1%) had tumors located in the tail of the pancreas and underwent distal pancreatectomy. According to the inclusion criteria, the enrolled patients did not have serious postoperative complications. All the patients were diagnosed with pancreatic ductal adenocarcinoma based on postoperative pathology. A total of 13 patients (3.5%) was in TNM stage I, 330 (88.0%) in stage II, 7 (1.9%) in stage III, and 25 (6.7%) in stage IV. Clinical follow-up showed that the disease-free survival (DFS) time was 17.4 months (95% CI 15.6–19.2) and ranged from 1 to 129 months. The overall survival (OS) time was 21.0 months (95% CI 19.2–22.7) and ranged from 2 to 129 months. Detailed clinicopathological information is provided in Table S1.

CREPT expression is upregulated in PDAC tissues

To assess the expression pattern of CREPT, IHC staining was performed in both tumor and adjacent normal tissues from 375 PDAC patients. We found that the CREPT protein was mainly localized in the nuclei of the tumor cells (Fig. 1a). Subsequent quantitative analyses revealed that CREPT protein expression was enhanced in tumor tissues compared to normal tissues (179.9 (95% CI 168.7–191.2) vs 54.9 (95% CI 46.6–63.3), p < 0.0001) (Fig. 1b). In addition, we found that the average CREPT staining score of the tumor tissues was 179.9. Therefore, 180 was used as the cut-off value for tumor CREPT expression. Next, the patients were divided into a high CREPT expression group (CES ≥ 180) and a low CREPT expression group (CES < 180), after which we evaluated correlations between CREPT levels and clinicopathologic features. No feature was found to be significantly correlated with the level of CREPT expression (Table S2).

Fig. 1.

IHC detection of CREPT expression in primary PDAC tissues. a CREPT expression is obvious in PDAC tissues, mainly in the nucleus, while no obvious CREPT expression is observed in normal control tissues. b Box-whisker plot of CREPT expression data in cancer tissues and adjacent normal tissues (mean value 179.9, 95% CI 168.7–191.1 vs 54.9, 95% CI 46.6–63.3, **** p < 0.0001). 5x: 5x magnification; 20x: 20x magnification

CREPT expression relates to PDAC prognosis

Next, we analyzed the relationship between the level of CREPT expression and the prognosis of PDAC patients; 25 patients in stage IV were excluded. Kaplan-Meier analysis revealed that the DFS time of patients with a low CREPT expression was significantly longer than that of patients with a high CREPT expression (18 months vs 13 months, p = 0.036, Fig. 2a). Similarly, we found that the OS time of patients with a low CREPT expression was significantly longer than that of patients with a high CREPT expression (25 months vs 20 months, p = 0.005, Fig. 2b). Univariate analysis showed that DFS time and OS time were negatively correlated with the tumor M stage, TNM stage and CREPT expression level (Table 1). Cox regression was used for multivariate factor analysis of history of diabetes, M stage, TNM stage, capsule invasion, preoperative CA19–9 value and CREPT expression. We found that high CREPT expression in tumor tissues served as an independent risk factor for a shorter DFS (HR = 1.42, 95% CI 1.06–1.90, p = 0.020) and OS (HR = 1.53, 95% CI 1.11–2.09, p = 0.009) time in PDAC patients (Table 2).

Fig. 2.

DFS and OS of PDAC patients. a The DFS time of patients with a low CREPT expression was longer than that of patients with a high CREPT expression (18 months vs 13 months, p = 0.036). b The OS time of patients with a low CREPT expression was longer than that of patients with a high CREPT expression (25 months vs 20 months, p = 0.005)

Table 1.

Univariate analysis between clinicopathological varaibles and DFS/OS

Variables	DFS (Month)	95% CI	P value	OS (Month)	95% CI	P value
Gender
Male	28.0 ± 3.9	20.3–35.6	0.76	39.1 ± 4.4	30.5–47.8	0.72
Female	30.6 ± 3.6	23.6–37.6		37.0 ± 4.2	28.8–45.2
Age
<65	27.9 ± 3.4	21.1–34.5	0.16	37.3 ± 4.0	29.5–45.1	0.23
≥65	30.9 ± 3.9	23.3–38.5		38.0 ± 4.1	30.0–46.0
CA19–9 (u/ml)
≥37	27.5 ± 2.7	22.1–32.8	0.076	34.6 ± 3.3	40.1–78.7	0.076
<37	46.6 ± 8.7	29.6–63.7		59.4 ± 9.8	28.2–41.0
Tumor location
head and neck	27.7 ± 3.8	20.1–35.2	0.87	34.4 ± 4.6	25.3–43.5	0.40
body and tail	32.5 ± 3.7	25.3–39.7		41.3 ± 4.1	33.4–49.3
Tumor differentiation
Well	29.0 ± 4.8	19.5–38.4	0.91	36.6 ± 5.5	25.9–47.3	0.73
Mediated or low	29.5 ± 2.8	24.0–35.0		38.1 ± 3.6	31.2–45.1
Capsule invasion
Yes	12.6 ± 2.3	8.0–17.2	0.053	22.3 ± 3.5	15.5–29.2	0.52
No	30.8 ± 3.0	24.8–36.8		39.4 ± 3.4	32.7–46.0
Vessel invasion
Yes	31.3 ± 3.0	25.4–37.1	0.32	38.3 ± 3.4	31.7–44.9	0.25
No	38.9 ± 9.9	19.5–58.2		44.4 ± 12.5	19.9–68.8
Nerve invasion
Yes	27.5 ± 3.0	21.7–33.3	0.92	33.7 ± 3.2	27.5–39.9	0.76
No	31.2 ± 4.2	23.0–39.4		41.3 ± 4.8	31.9–50.8
T stage
T1/2	32.6 ± 9.5	14.0–51.2	0.65	45.4 ± 11.1	23.6–67.2	0.25
T3/4	31.5 ± 3.1	25.5–37.5		37.9 ± 3.5	31.1–44.7
N stage
N0	30.1 ± 3.4	23.4–36.9	0.28	39.4 ± 4.4	30.9–47.9	0.45
N1/2	30.7 ± 4.0	22.9–38.5		37.5 ± 4.3	29.1–45.9
M stage
M0	30.7 ± 3.0	24.8–36.7	0.031	39.7 ± 3.4	33.1–46.4	0.026
M1	11.8 ± 2.2	7.6–16.1		16.2 ± 1.9	12.6–19.9
TNM stage
I/II	31.0 ± 3.1	25.0–37.0	0.016	40.2 ± 3.4	33.5–46.9	0.002
III/IV	12.4 ± 1.9	8.6–16.1		15.4 ± 1.7	12.1–18.8
CREPT expression
High group	28.7 ± 4.0	21.0–36.5	0.039	33.5 ± 4.7	24.2–42.8	0.007
Low group	31.2 ± 3.4	24.6–37.7		41.6 ± 4.2	33.4–49.8

DFS, Disease-free survival time; OS, Overall survival time; CI, Confidence interval

P < 0.05 indicates statistical significance of differences

Table 2.

Multivariate analysis between clinicopathological variables and DFS/OS

Variables	HR	95% CI	P value	HR	95% CI	P value
History of diabetes
	1.12	0.73–1.71	0.602	1.16	0.73–1.83	0.537
M stage
	1.09	0.37–3.20	0.871	0.75	0.25–2.23	0.601
TNM stage
	1.43	0.58–3.54	0.438	2.34	0.94–5.80	0.067
Capsule invasion
	0.64	0.30–1.38	0.258	0.69	0.31–1.58	0.383
CA19–9 (u/ml)
	1.44	0.98–2.12	0.062	1.53	0.99–2.35	0.055
CREPT expression
	1.42	1.06–1.90	0.020	1.53	1.11–2.09	0.009

HR, Hazard ratio; CI, Confidence interval

P < 0.05 indicates statistical significance of differences

CREPT affects PDAC tumor cell features in vitro and in vivo

Using qRT-PCR we found that the mRNA expression levels of CREPT in Panc-1, AsPC-1, MIA 3PaCa-2, SU.86.86, BxPC-3 and T3M4 PDAC cells differed significantly. The expression level was highest in T3M4 cells (Fig. S1). The CREPT protein expression levels were consistent with the mRNA expression levels, as determined by Western blotting (Fig. S1). Next, we used shRNA to knockdown the expression of CREPT in T3M4 cells. A subsequent CCK8 assay revealed that, compared to control cells (CREPT KDNC), the growth rate of CREPT knockdown (CREPT KD) cells was significantly decreased (p < 0.05) (Fig. 3a), and that the sensitivity of CREPT KD cells to gemcitabine was significantly increased (p < 0.05) (Fig. 3b). Next, we used Transwell assays to evaluate the migration and invasion abilities of CREPT KD and CREPT KDNC cells. We found that compared with those of CREPT KDNC cells, the migration and invasion abilities of CREPT KD cells were significantly lower (p < 0.05) (Fig. 3c-d).

Fig. 3.

In vitro effects of CREPT expression in PDAC cells. a Regulatory effect of CREPT on PDAC cell proliferation. The proliferation rate of CREPT KD cells was significantly decreased compared to that of CREPT KDNC cells (*p < 0.05). b Regulatory effect of CREPT expression on PDAC cell sensitivity to gemcitabine. The sensitivity of CREPT KD cells to gemcitabine was significantly increased compared to that of CREPT KDNC cells (*p < 0.05). c Regulatory effect of CREPT on PDAC cell invasion. The invasion ability of CREPT KD cells was significantly inhibited compared to that of CREPT KDNC cells (*p < 0.05). d Regulatory effect of CREPT on PDAC cell migration. The migration ability of CREPT KD cells was significantly inhibited compared to that of CREPT KDNC cells (*p < 0.05)

A shRNA expressing lentivirus system was subsequently used to inhibit CREPT expression (Lenti-shCREPT) or not (control lentivirus, Lenti-shCon) in T3M4 cells. After subcutaneous injection of Lenti-shCREPT- and Lenti-shCon-infected T3M4 cells into nude mice, we found that the growth rate of the implanted tumors in the Lenti-shCREPT group was significantly slower than that in the Lenti-shCon group. Also, the tumor weights were lower in the Lenti-shCREPT group than those in the control group (p < 0.05, Fig. 4a-c), indicating that CREPT expression knockdown can inhibit the growth of PDAC tumors in vivo.

Fig. 4.

PDAC tumor formation in nude mice. a Growth curves of shCREPT- and shCon-derived tumors. CREPT knockdown leads to decreased tumor volumes (*p < 0.05). b Weights of shCREPT- and shCon-derived tumors. CREPT knockdown leads to decreased tumor weights (*p < 0.05). c Pictures of dissected tumors from the shCREPT (left) and shCon (right) groups

Discussion

It is commonly believed that dysregulated oncogenes and tumor suppressor genes play important roles in cancer development and cell cycle deregulation [17]. CREPT, which is highly homologous to RPRD1A, can bind to the phosphorylation site of RNA Pol II and affect transcription to regulate cell cycle progression [18]. At the initiation of transcription, Ser5 and Ser7 are phosphorylated by transcription factor II H (TFIIH) and cyclin-dependent kinase 7 (CDK7) [19]. Ser7 phosphorylation (S7P), in turn, recruits CDK9 and phosphorylates Ser2 under the synergistic action of CDK12, CDK13 and BRD4 [20–23]. During elongation, S5P is gradually dephosphorylated by Rtr1 (RPAP2 in humans) [24, 25]. At the end of transcription, the remaining S5P and S7P residues are dephosphorylated by Ssu72, and S2P becomes the major phenotype of RNA Pol II CTD [26–29]. Programmed phosphorylation at specific sites of Pol II CTD effectively regulates the cell cycle. At initiation, S5P provides a binding site for mRNA-capping enzymes to modify the 5′ end of the newly synthesized mRNA [29, 30]. During transcription, when S5P is dephosphorylated and Ser2 is phosphorylated, RNA Pol II starts to leave the promoter region [31, 32]. At the end of transcription, S2P binds to Rtt103 (p15RS in humans) to end transcription [33]. CREPT and p15RS show a high homology in both their N-terminal and C-terminal domains [13, 33, 34]. However, they have different effects on cell proliferation. p15INK4b, which is encoded by p15RS, inhibits cell proliferation [35], whereas CREPT is highly expressed in many types of cancer [9, 14]. Although the specific reasons for these phenomena are unclear, some researchers believe that the junction domains of the N-terminal and C-terminal regions differ, leading to distinct functions [34]. Other researchers have suggested that CREPT may bind to the promoter of cyclinD1 and, at the same time, its 3′ end, while p15RS and Rtt103 cannot, and that this unique interaction between CREPT and its target gene explains the difference in function [13].

Lu et al. explored the relationship between CREPT and cell cycle-related proteins [9] and found that CREPT could promote the proliferation and tumorigenesis of HepG2, HeLa and NIH3T3 cells, both in vitro and in vivo. The authors also showed that CREPT can shorten the cell cycle and accelerate progression from the G1 to S phase, and that CREPT synchronizes better with cyclinD1 than with other proteins. Since the authors also found that after binding to RNA Pol II, CREPT can bind both the promoter and the 3′ poly(A) region of cyclinD1, they hypothesized that the CREPT and RNA Pol II complex forms a hairpin structure for cyclinD1. In the absence of CREPT, RNA Pol II passes through the cyclinD1 gene and reaches the 3′ poly(A) region to terminate transcription. CREPT, on the other hand, allows RNA Pol II to be concentrated at the cyclinD1 gene promoter and upregulate the expression of cyclinD1 to shorten the cell cycle.

By exploring the molecular pathways related to CREPT, cyclinD1 was found to play an important role in multiple pathways, including the Wnt/β-catenin, STAT3 and NF-κB signaling pathways [36–39]. Multiple transcription factors can interact with the cyclinD1 promoter region to affect its transcription. TCF4 is an important component of the Wnt/β-catenin signaling pathway that can bind to transcriptional repressors, such as Groucho and HDAC. Activated Wnt signaling leads to the replacement of transcriptional inhibitors with β-catenin and the binding of β-catenin to TCF4 to form a TCF4/β-catenin complex. This complex recruits transcriptional co-stimulatory factors (such as BRG1, CBP/p300, Bcl9 and Pygopus) and stimulates downstream target genes (such as cyclinD1 and c-MYC) to activate transcriptional processes [40–42]. The stability of the TCF4/β-catenin complex during this process is critical for the entire signaling pathway to remain active. Lu et al. further explored the relationship between CREPT and the Wnt/β-catenin signaling pathway and found that CREPT can bind to TCF4/β-catenin in the nucleus to stabilize the complex and enhance the activity of the Wnt/β-catenin signaling pathway, which was also demonstrated in vitro. Thus, downstream target genes, such as cyclinD1 and c-MYC can be upregulated and, thereby, affect cell proliferation.

In total 375 patients were enrolled in the current study. According to the IHC results, the expression of CREPT in PDAC tumor tissues was significantly higher than that in adjacent normal tissues, which is in line with reports on other tumors [9, 43]. The DFS and OS times of patients with high CREPT expression were significantly shorter than those of patients with low CREPT expression. Therefore, CREPT is an independent risk factor for a shorter DFS and OS time in PDAC. We also found that M stage and TNM stage were negatively correlated with the DFS and OS time, which is consistent with other reports [44]. Type of resectability or tumor location showed no significant correlation with the prognosis of PDAC patients in our study. Others found that patients with resectable pancreatic head cancer (PHC) exhibited a worse survival than those with resectable pancreatic body or tail cancer (PBTC) because PHC has a more aggressive tumor biology than PBTC [45–47]. Others, however, reported that PBTCs had a worse overall survival than PHCs because PBTCs lack early symptoms, such as obstructive jaundice and, therefore, are usually more advanced at diagnosis than PHC [48, 49]. Further studies are required to solve the underlying causes of these disparities.

We also explored the effects of CREPT expression on PDAC cells and found that CREPT knockdown inhibits the proliferation of PDAC cells and the growth of PDAC tumors. We also found that CREPT can increase gemcitabine resistance in PDAC cells. Together, these results provide a basis for the use of CREPT as a prognostic PDAC biomarker and therapeutic target. In future studies, the possibility of using CREPT as prognostic biomarker through its detection in peripheral blood or tissues by fine needle aspiration, or by other less invasive methods, may be explored to expand its putative prognostic value. In addition, CREPT may be employed as a putative therapeutic target to design CREPT-targeted inhibitors, siRNAs or other drugs.

This study has some shortcomings. We used strict inclusion and exclusion criteria to ensure a consistent baseline of recruited patients. For example, patients with benign pancreatic, biliary or hepatic diseases or without follow-up data were excluded. These criteria may cause selection bias. Additionally, we did not investigate whether CREPT can upregulate cyclinD1 gene expression in PDAC cells and act through the Wnt/β-catenin signaling pathway Furthermore, whether CREPT can affect the transcription of other genes or influence other signaling pathways was not explored. In future studies, the effect of CREPT on downstream genes to increase gemcitabine resistance should be clarified.

Abbreviations

CCK8

Cell Counting Kit-8

CES

composite expression score

CI

confidence interval

CREPT

cell-cycle-related and expression-elevated protein in tumor

CTD

C-terminal domain

DFS

disease-free survival

HE

Hematoxylin-eosin

IHC

immunohistochemistry

OS

overall survival

PDAC

pancreatic ductal adenocarcinoma

RPRD1B

regulation of the nuclear precursor RNA domain containing 1B

SDS-PAGE

SDS-polyacrylamide gel electrophoresis

TMA

tissue microarray

Authors’ contributions

All authors contributed equally to the writing and editing of this manuscript. All authors read and approved the final manuscript.

Funding

This study was supported by grants from the National Natural Science Foundation of China (No. 81772639, No. 81802475, No. 81972258, No. 81974376), the Natural Science Foundation of Beijing (No. 7192157), the CAMS Innovation Fund for Medical Sciences (CIFMS) (No.2016-I2M-1-001) and the Non-profit Central Research Institute Fund of the Chinese Academy of Medical Sciences (No. 2018PT32014).

Compliance with ethical standards

Ethics approval and consent to participate

This research was approved by the Research Medical Ethics Committee of Peking Union Medical College Hospital. Informed consent was obtained from each patient.

Consent for publication

All authors agreed to publish the article.

Competing interests

The authors declare no potential conflicts of interest.