Identification and gene expression analysis of circulating tumor cells with high expression of KRAS in pancreatic cancer

The seventh most frequent cancer in the globe and the fourteenth most prevalent cancer overall is pancreatic cancer. It is believed that the incidence of pancreatic cancer is generally higher in wealthy countries than in poorer countries and that the prevalence is essentially the same in men and women.¹ 90% of pancreatic malignancies are pancreatic adenocarcinoma.² The pancreas’ head accounts for around 60%–70% of pancreatic adenocarcinomas, with the remaining 15% occurring in the body and 15% in the tail. Most pancreatic adenocarcinomas are already metastatic when they are diagnosed. Currently, surgical resection is the only effective management of pancreatic cancer, but both the recurrence rate and long-term survival rates are very poor.

CTCs (circulating tumor cells) are released into the peripheral blood by some tumor cells from the primary site or metastasis and can form metastases in other organs of the body, leading to tumor recurrence and metastasis. CTCs have been identified as a potential noninvasive biopsy tool to analyze patients’ primary tumors and their risk of subsequent metastasis.^3–6 At present, CTCs are the main detection substance of liquid biopsy technology and have significant clinical application value in the early diagnosis, efficacy evaluation, and recurrence monitoring of tumor patients. Many studies have found that the expression of epithelial mesenchymal transition related proteins in tumor cells is associated with worse disease prognosis.^7,8 However, there are very few studies on the epithelial mesenchymal transition process of CTCs in pancreatic ductal adenocarcinoma (PDAC).^9,10

KRAS mutations were found in more than 80% of patients irrespective of the PDAC stage.¹¹ KRAS mutations were associated with the development of pancreatic cancer.^12,13 Advances in genomics and next-generation sequencing have accelerated the development of liquid biopsies. NGS and digital PCR have made it possible for doctors to find and identify a significant number of novel nucleic acid biomarkers in only a few short years.¹⁴ 

The most well-established and frequently used technique for CTC collection is magnetic activated cell sorting (MACS). It involves epithelial cell adhesion molecule (EpCAM) magnetic lipid spheres with antibody modifications to isolate CTCs, but this approach still has significant limitations in capturing CTCs with epithelial–mesenchymal transformation.¹⁵ 

In this study, multilocus immunolipid magnetic spheres were used to enrich circulating tumor cells for fluorescence identification of tumor markers. The results showed that the prepared immunolipid magnetic spheres had good CTC separation functions. Therefore, the enrichment and analysis of CTCs from peripheral blood of pancreatic cancer patients are of great clinical value.

Sample collection and treatment: (1) Sample treatment: 7.5 ml of peripheral blood from pancreatic cancer patients was collected using EDTA anticoagulation tubes, stored at 4 °C, and tested within 72 h after thorough mixing. (2) Detection indicators: (1) CTC count (magnetic separation immunofluorescence identification count) was performed in the peripheral blood of patients with pancreatic cancer. (2) Gene mutation detection in circulating tumor cells of pancreatic cancer. (3) CTC gene detection in pancreatic cancer. (3) Patient clinical information collection.

EpCAMs (vimentin) were purchased from Sigma-Aldrich, HSPGs (KRAS monoclonal antibodies) were purchased from Shanghai Sheng Na Industrial Co., DAPI dye (CD45-PE) was purchased from Multisciences (Lianke) Biotech Co. Ltd., and cholesterol (Chol) and methylene chloride were purchased from Sangon Biotech (Shanghai) Co. Ltd.

DOPC, chitosan cetyl quaternary salt, and cholesterol were dissolved in dichloromethane solution. Fe₃O₄-HMN was added and stirred to dissolve and transferred to a pear-shaped flask. A certain amount of PBS solution was added to fix the volume. After incubation, NHS, EDC, HCI, and anti-KRAS solution were added to the pear-shaped flask, and the reaction was stirred at 4 °C for 12 h. Then the solution was separated by a rotary evaporator to remove the residual CH₂Cl₂, and the solution was separated by magnetic separation and washed three times to obtain KRAS immunomagnetic beads. We prepared EpCAM immunomagnetic beads and vimentin immunomagnetic beads in a similar way.

A nanoparticle size potentiostat (Zetasizer Nano ZSP) was utilized to evaluate the characterization of the immunomagnetic beads, and an atomic force microscope was used to observe the microstructure of the immunomagnetic beads. A specific saturation magnetization analyzer (sigmameter) was employed to investigate the saturation magnetization intensity of immunomagnetic beads, and an ultra-micro spectrophotometer (ND-1000) was used to analyze the density of immunomagnetic beads.

Pancreatic cancer cells used in this study were routinely cultured in a DMEM medium. The cultural environment was 37 °C and 5% CO₂ under humid conditions. The medium dosage is 2 ml medium for 35 mm dish, 3 ml medium for 60 mm dish, and 8 ml medium for 10 cm dish. For freezing and resuscitation of cells, the cells were digested with 1.25% trypsin. When the cells became round and had not yet floated, a proper amount of complete culture medium was added and the cell suspension was formed by repeated blowing. The ratio of cell suspension to glycerol was 900 μl: 100 μl, and the solution was stored overnight at −70 °C in liquid nitrogen. After resuscitation, the cells were immediately placed in a 40 °C water bath, and the freezing tubes were gently shaken to make them all melt rapidly. Then they were transferred to cell culture flasks with the medium and placed in a CO₂ incubator. After 5–8 h, the culture medium was replaced, and conventional subculture was performed.

Pancreatic cancer cells were digested by trypsin and prepared into a single cell suspension, and then a culture medium containing serum was added to neutralize trypsin. After dilution, pancreatic cancer cells were counted and inoculated into 96-well plates with 7000 cells per well. After incubation at 37 °C for 24 h, 100 μl of the medium was added to each well, the original culture medium was removed, and immunomagnetic beads with gradient concentration were added to the cells. The incubation was continued for 48 h. 50 µl of MTT solution with a concentration of 1 mg/ml was added to each well, and the incubation was continued for 2 h. 100 µl of isopropanol was added and shaken for 10 min, and formazan crystals were observed under the microscope after incubation for 3 h.

After removing the medium, 150 µl DMSO solution is added to each well to dissolve the crystallization. The multifunctional Spectramax M5/M5e (Molecular Devices) reader was operated at 490 nm, and the experimental results were analyzed.

Peripheral blood of pancreatic cancer patients was collected using 7.5 ml EDTA anticoagulation tubes, mixed by gently by inverting the blood collection tubes eight times, and then stored at 4 °C and enriched for circulating tumor cells within 48 h. The anti-KRAS immunolipid nanospheres were used to enrich the circulating tumor cells, and the captured tumor cells were fixed and immunostained with keratin cytokeratin antibodies (CK8/18/19, etc.), CD45 antibodies, and finally, the nuclei were identified by 4,6⁃diamidino 2⁃phenylindole (DAPI) staining. Then, the cell suspensions were dropped onto transparent slides and observed under a fluorescence microscope. The CTCs were counted, and the CTCs that met the evaluation criteria are KRAS+/EpCAM+/vimentin+, CK+, DAPI+, and CD45-. The CTCs were counted with the supporting multicolor fluorescence cell counter.

CTCs were separated from the stripping slides. Total DNA was extracted with the TIANamp Genomic DNA Kit, and total RNA was extracted with the TRIzol Plus RNA Purification Kit. The extracted RNA and DNA were quantified by a Nanodrop ND-1000. 120 ng of RNA extracted from CTCs was used for qPCR, and a gene primer reaction was added to detect its Ct value and gene expression. PCR was carried out in a 25 μl reaction tube, 10 μl template DNA and 1 μl primers were added, and the amplification was realized on a T100 thermal cycler (Bio-RAD). PCR products were purified by the QIAquick PCR Purification Kit (Qiagen) and sequenced.

Using SPSS 21.0 statistical software, the measurement data were expressed as mean ± standard deviation. The t-test was used to test the counted data using X² test, and statistical analysis was performed using the rank sum test. All experiments for cell culture were performed each time in triplicate. The difference was statistically significant at P < 0.05.

The genomic analysis of KRAS-CTC consists of four processes. The flow chart given in Fig. 1 describes the process. The platform for CTC capture first requires the preparation of immunolipid magnetic spheres for the enrichment and capture of CTCs. The preparation process of immunoliposomes is shown in Fig. 1. The antibodies used in the experiment (anti-EpCAM, anti-vimentin, and anti-KRAS) were coupled to GHDC to generate polymers. Then immunolipid magnetic spheres were prepared with DOPC, cholesterol, and Fe₃O₄ raw materials by reversed-phase ultrasound. The modification of GHDC plays an emulsifying and dispersing role in the preparation of immunolipid beads and can help the antibody to bind to the surface of magnetic beads, thereby increasing the amount of antibody carried by immunoliposomes. Cholesterol helps magnetic beads to maintain a stable form during preparation. The oxidation rate of the Fe₃O₄ raw material naked magnetic sphere can be reduced by liposome encapsulation, which lays a foundation for subsequent magnetic separation. There are two crucial functional features of this CTC capturing platform: (1) the Anti-EpCAM antibody is coated with lipid micelles directly arranged on the nanoparticle (NP) for recognition and capture; (2) the soft and protein-like membrane structure can enhance the contact frequency with cell particles.

Clinical samples from patients for CTC capture and identification were analyzed using the prepared system based on immune lipid magnetic spheres and circulating tumor cell capture and separation. The brief process is shown in Fig. 1. Centrifuging the samples of peripheral blood for 10 min at 1000 g gathers the upper-middle layer fluid. The isolated CTCs were stained with immunofluorescence and observed under a fluorescence microscope. The statistical experimental results were analyzed to form a diagnostic report.

The performance of the key KRAS-positive magnetic nanoparticles in the KRAS-CTC capture system would greatly affect the enrichment of KRAS-expressing tumor cells in the blood. Figure 2(a) shows the UV absorption spectrum of KRAS-ML. It can be seen from the figure that the nano-lipid magnetic spheres connected with EpCAM, vimentin, and KRAS antibodies have obvious UV absorption peaks around 280 nm, indicating that the modified antibody exists on the surface of the nano-magnetic beads and the antibody content is controllable. Figure 2(b) shows the magnetic saturation curve of KRAS-ML and other immunomagnetic spheres. The results showed that each magnetic sphere (EpCAM, vimentin, and KRAS) prepared has high saturation magnetization and superparamagnetism. At 300 K, the saturation magnetization of KRAS-ML was 21.5 emu/g while that of Fe₃O₄ magnetic fluid was 42.4 emu/g. Therefore, it can be said that the liposome coating of Fe₃O₄ magnetic fluid reduces the magnetic properties to some extent. In addition, KRAS-ML particle size distribution is shown in Fig. 2(c), where the average particle size is 207.4 ± 5.8 nm, which was not much different from the particle size results shown in atomic force photos. The surface potential of KRAS-ML is 3.8 ± 5.2 mV [Fig. 2(d)], which was neutral and could reduce the non-specific adsorption interference caused by positive charges in previous studies.

Wide-scale application of biological nanoparticles depends on their excellent biocompatibility and lower toxicity. As shown in Fig. 3, the four kinds of IMLs constructed in this study showed a low inhibitory rate on pancreatic cancer cells when the concentration was lower than 100 µg/ml, while the toxicity of immune magnetic beads significantly increased when the concentration was above 200 g/ml. Figure 3 results also showed that the four IMLs had similar inhibitory ability on tumor cells, with no difference between them, except for the concentration of added magnetic spheres. Therefore, the low toxicity of the nano-lipid magnetic spheres to the captured cells under normal use laid a foundation for the culture, surface marker analysis, cell behavior analysis, and gene analysis of the captured CTCs.

We used laser confocal microscopy to examine cells that were attached to the microscope slide (Fig. 4). IMLs adhere to the cell surface. Immunoliposomes gradually accumulated around the CTCs. The IMLs eventually surround the cell surface to the point where they were no longer attached. The immunofluorescence of KRAS immunoliposome expression overlaps well with the red fluorescent cell membrane dye Dil, indicating that the immunoliposomes were uniformly dispersed on the CTC surface. At least three months after preparation, the immunoliposomes still exhibit a high degree of stability and can be applied to capture tumor cells. The results of Prussian blue staining showed a blue fluorescence on the surface of CTCs, demonstrating that KRAS-ML effectively binds to the membrane surface of pancreatic cancer cells. The fluorescence around the cells elevated with time. The binding of immunomagnetic beads to KRAS receptors on the surface of pancreatic cancer cells reached a peak at 20 min.

The cell suspension of pancreatic cancer cells was prepared and added to the 7.5 ml PBS buffer and anticoagulant blood at different concentrations of cells after counting to mimic CTC suspension. The prepared ML, K-ML, Ep-ML, and Vi-ML magnetic beads were used for capture. After capture, it was stained with DAPI and fluorescent antibodies and observed under a fluorescence microscope. In this study, the suspension was divided into ten groups—Ep-ML group, Vi-ML group, K-ML group, Ep-LMB/Vi-LMB/K-ML group, Ep-ML + Vi-ML + K-ML group, Vi-ML + Ep-ML + K-ML group, K-ML + Ep-LMB + Vi-ML group, K-ML + Vi-ML + Ep-ML group, Vi-ML + K-ML + Ep-ML, and Vi-ML + Ep-ML + Vi-ML group—to evaluate the efficiency of tumor cell enrichment under various combinations of magnetic beads. We found that compared with ML, EP-ML, and VI-ML, K-ML can capture pancreatic cancer cells more stably at different concentrations in PBS and blood systems [Figs. 5(a) and 5(b)]. The enrichment efficiency of combining the three beads was superior to that of using individual beads. Furthermore, the order for sequential binding had no significant difference in the capture efficiency of tumor cells. Binding had no significant difference in the capture rate of the cancer cells [Figs. 5(c) and 5(d)].

The pancreatic cancer CTC detection system constructed by EpCAM/vimentin/KRAS magnetic spheres was used to detect the peripheral blood of 62 patients with pancreatic cancer, as shown in Fig. 6(a). The cells captured by magnetic spheres were subsequently identified by immunofluorescence. CTCs of patients with pancreatic cancer showed obvious cell morphology in bright fields. CK19-FITC showed strong positive green fluorescence, and DAPI showed strong positive blue fluorescence. The two kinds of fluorescence were superimposed and overlapped, CD45 staining was negative, and the enrichment cells in line with the above-mentioned description were identified as pancreatic cancer CTCs. In the peripheral blood of 62 pancreatic cancer patients enrolled, the results of CTC count analysis are shown in Fig. 6(b). When KRAS-ML, Ep-ML, and Vi-ML were added sequentially, the average number of enriched cells was 5.21/7.5, 4.73/7.5, and 4.54/7.5 ml, respectively, which showed no significant difference in the number of CTCs enriched by KRAS-ML, Ep-ML, and Vi-ML [Figs. 6(b) and 6(c)].

We counted the KRAS mutation status in 62 blood samples and 21 tissue samples from pancreatic cancer patients. The results showed that KRAS mutations were detected in the CTCs of 32 patients and the tissues of eight patients. Overall, KRAS magnetic beads detected CTCs of 93.2% for KRAS mutations in patients with pancreatic cancer [Fig. 6(d)]. To further understand the relationship between the KRAS mutation and clinicopathological characteristics of pancreatic cancer patients, a statistical analysis was conducted on the clinicopathological features of 62 patients with pancreatic cancer, as well as the correlation analysis of age, tumor size, lymph node metastasis, selection of distant metastasis, and TNM stage, as shown in Table I. The findings demonstrated that KRAS mutation was not related to clinicopathological characteristics.

The five-year overall survival rate for pancreatic ductal adenocarcinoma (PDAC) has risen from 3% in the 1970s to a meager 9% in 2020.¹⁶ Currently, most patients are diagnosed only after their tumors have metastasized and therefore are not suitable for surgery.¹⁷ Even those patients at the early stages of disease face additional challenges as there is currently no way to stratify a patient’s risk of metastasis to help guide neoadjuvant therapy and adjuvant therapy.

CTCs are disseminated tumor cells that have broken off from the primary lesion and are therefore considered to be the basis of distal metastasis. With CTCs, it is possible to understand cancer progression and track response to treatment in real time.¹⁸ In addition, longitudinal monitoring of changes in CTCs after chemotherapy can provide insight into treatment outcomes in patients with PDCA.¹⁹ A feasible strategy for improving the prognosis of pancreatic cancer was a better understanding of the pancreatic cancer metastasis process, and according to the prognosis of pancreatic cancer and metastasis biomarkers, cancer cells for the treatment of patients²⁰ with layers usually lose some characteristics of epithelial cells and gain a more mesenchymal phenotype, which was known as “epithelial–mesenchymal transition.”²¹ Epithelial–mesenchymal transformation increased mobility and invasiveness and was thought to contribute to metastasis.²² However, the expression of EpCAM in CTCs varies greatly. The interstitial transformation of CTC results in poor CTC capture effect of EpCAM magnetic spheres for some cancer types. Our study showed that direct screening of CTCs with KRAS mutation in pancreatic cancer patients can guide patients to use precise medication and more precisely test which patients are appropriate for further therapy. Approximately 80% of pancreatic cancer patients carried KRAS mutations, and these patients had a worse prognosis than those with KRAS wild type.^23–25 The protein encoded by the KRAS is a GTPase that plays an important role in intercellular signaling pathway transduction. Through the binding of GTP to the protein encoding KRAS, the signaling pathway is activated, thereby promoting cell proliferation.²⁶ CellSearch was the only FDA-approved platform for CTC capture, which used immunomagnetic beads to enrich CTCs expressing EpCAM.²⁷ However, Bidard et al. used the CellSearch system and detected CTCs in only 11% of the 79 pancreatic cancer patients.²⁸ We speculate that this was mainly due to the inability of the CellSearch system to detect mesenchymal CTCs. In this study, we used KRAS-ML to capture pancreatic cancer CTCs and significantly increased the number of CTCs carrying KRAS mutations. Notably, the results suggest that we can stratify pancreatic cancer patients prognostically by the CTC enrichment capture system, and the number of CTCs is closely related to the prognosis of patients. In the future, longitudinal monitoring of CTCs can also be included in the experiments to dynamically assess the correlation between changes in the number of CTCs and the prognosis and treatment outcome of patients in real time. In conclusion, the immunomagnetic beads modified by KRAS antibody constructed in this study based on pancreatic cancer is more suitable for CTC detection than EpCAM immunomagnetic beads in pancreatic cancer patients. In addition, CTCs can screen for KRAS mutations in pancreatic cancer, thus realizing the possibility of early diagnosis and prognosis assessment in pancreatic cancer patients.

The supplementary material contains the results of CTC testing for 62 pancreatic cancer patients using magnetic spheres, and the table within exhibits the amount of captured tumor cells in the peripheral blood of the same patients. Both Kras-ML and Ep-ML were used, as well as Vi-ML.

This study was supported by the Startup Fund for Scientific Research of Fujian Medical University (Grant No. 2020QH1119).

The authors have no conflicts to disclose.

Zhiqiang Zheng: Conceptualization (lead); Writing – original draft (equal); Writing – review & editing (equal). Zhichao Chen: Data curation (lead); Formal analysis (lead); Writing – original draft (equal); Writing – review & editing (equal). Yonghua Lin: Methodology (lead); Writing – original draft (equal); Writing – review & editing (equal). Jianfeng Wei: Project administration (lead); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available within the article.