Evaluation of Human Mammaglobin as a Biomarker of Circulating Tumor Cells in Breast Cancer and Clinical Verication

Cytokeratin (CK) is the gold standard marker for the differential diagnosis of epithelial circulating tumor cells (CTC), but the low expression of CK can lead to false negative results.In this study, the specicity and sensitivity of human breast globin (hMAM) as a tumor marker of CTC in peripheral blood of breast cancer patients were analyzed and aim to improve the CTC detection accuracy in breast cancer, enrich the candidate markers for clinical differential diagnosis of breast cancer CTC. EpCAM antibody modied liposome magnetic particles (ELMP) were prepared, and then their physicochemical properties were characterized. ELMP has high physicochemical stability and can eciently enrich the CTC of epithelial breast cancer. We found hMAM is more consistent with clinical and pathological diagnosis results. To be noted, CK19 combined with hMAM method eciently distinguish the separated CTC from breast cancer patients, furthermore, the specicity and sensitivity of CTC detection get promoted.


Introduction
Breast cancer is a type of malignant tumor with a higher risk of metastasis and recurrence [1,2]. Therefore, using sensitive biomarkers to detect metastasis earlier and using reasonable treatment measures are helpful to improve patient survival [3][4][5][6]. Circulating tumor cell (CTC) detection is a viable way for breast cancer diagnosis, staging and transfer risk monitoring [7][8][9]. Due to the heterogeneity of tumors, breast cancer expresses multiple tumor markers at the same time, so accurate biomarkers are urgently needed for patient management and prognosis [10,11]. In recent years, methods to detect and separate CTC from different sources of organs have been developed [12,13]. Current CTC detection techniques include magnetic separation technology based on immune-capture blood CTC, microchip and lter membrane separation method, and improving the convenience and sensitivity of CTC enrichment [14][15][16]. Epithelial-cell adhesion molecule (EpCAM) plays an important role in carcinogenesis and it is usually overexpressed in most tumor cells of epithelial origin, and not in mesoderm-derived blood cells and lymphocytes. The EpCAMbased separation method have been extensively used in clinical for multiple types of cancer [17][18][19].
After CTC is separated, immuno uorescence method is used to identify tumor characteristics. Cytokeratin 19 (CK19) CK19 is a typical marker of epithelial cells and CTCs. However, the level of CK19 expression for those CTCs undergo the Epithelial-mesenchymal transition (EMT) process will present low expression level of CK19 [20,21], inevitably brings false negative results [22].Human mammaglobin (hMAM) are tissue-speci c markers of breast epithelial tissue and their abnormal expression is closely related to micrometastases [23,24], hMAM is expressed only in mammary epithelial cells and overexpressed mainly in breast cancer cells [25,26].
In this study, we prepared EpCAM antibody modi ed liposome magnetic particles (ELMP) in order to separate CTC from breast cancer clinical blood samples. The CTCs counting results by using CK19 alone and the combined use of hMAM and CK19 as breast cancer CTCs markers independently. We compared the correlation between pathological diagnosis of breast cancer patients and the upon two methods for exploring the clinical diagnostic value of hMAM as a speci c tumor marker of CTC in breast cancer patients, aiming to improve the accurate of breast cancer CTCs differential diagnosis.

Cell lines
Human mammary gland cell MCF-10A cells, Cervical cancer Hela cells, breast cancer cell MDR-MB-231, SK-BR-3, MCF-7 and 293T were purchased from ATCC cell bank and kept in the cell lab of Changzheng Hospital; DMEM culture medium, fetal bovine serum and trypsin purchased from Gibco.

Clinical samples collection
This clinical study was approved by Zibo Central Hospital. All the patient and healthy volunteer consents were written informed consents, and that this was conducted in accordance with the Declaration of Helsinki. Peripheral blood was collected from patient and healthy volunteer, each person was collected two tubes of venous blood, 8 ml per tube at Zibo Central Hospital, including 20 normal blood samples from healthy volunteers, 100 blood samples from breast cancer patient. All breast cancer patients were con rmed by breast ultrasonography and MRI, 60 of the 100 breast cancer patients have complete clinical case information and follow-up records. All blood samples were performed CTC separation and identi cation within 48 h after the blood collected in anticoagulant vessel collection. ELMP preparation ELMP was prepared refer to the reverse evaporation method [27]. First of all, DSPE-PEG 2K -MAL 0.5 mg and DSPE-PEG 2K 2.0 mg was dissolved in 2.5 mL ddH 2 O, DOPC 3.0 mg, raw magnetic beads 1.0 mg and Chol 2.0 mg was dissolved in 1.5 mL chloroform, the water phase and the organic phase was mixed, followed by ultrasonic treatment to obtain emulsion. Then liposome solution was obtained under reduced pressure by means of rotary evaporator.
Secondly, ELMP was prepared by the method in reference [28,29], the reaction steps are as follows, 5.0 mg liposome solution was added into 4 mL HEPES buffer (50 mM, pH 6.5), then 40 µg EpCAM Antibody solution was added and stirred for 12 h at room temperature. The reaction mixture was dialyzed in a 3,500 Da dialysis bag for 12 h at 4 ℃ to obtain ELMP.

Characterization of ELMP
The surface potential and particle size of ELMP was measured by particle size potential analyzer (Zetasizer Nano ZS 90). The morphology of ELMP was observed by transmission electron microscopy (TEM), ELMP was dropped onto the copper mesh and observed after drying. Determination of saturation magnetization of ELMP, after the machines were warmed up, instructions and software prompts were followed to calibrate the machine rst, and then the right amount of dry sample of magnetic ball powder was taken to be installed and to start measurements.
Intrinsic coercivity HCJ value can be drawn by a magnetic regression line that can come from the test. The saturation magnetization value of the sample can be obtained by its mass values.

hMAM-QDs was reparented by hMAM antibody labeled with quantum dot
The rabbit anti-human hMAM antibody was labeled with a quantum dot labeling kit (QDream Antibody Labeling Kit), the brief steps are as follows, 10 × reaction solution and DEionized water were used to dilute the antibody concentration to 2 mg/ml, and the buffer system was 1 × reaction solution; according to the molar ratio of activating reagent: antibody = 20:1, the activating reagent was added to the above diluted antibody solution, mixed completely, and incubated for 30 minutes at room temperature; according to the molar ratio of antibody: quantum dots = 2:1, the quantum dot solution was added to the above reaction mixture, fully mixed, and incubated at room temperature for 30 minutes, labeled step was end. Flow cytometry was used to detect the immunostaining of breast cancer cells by CK19 and hMAM at the same time.
MDA-MB-231 cells were plated at a density of 80,000/well in a 12-well plate, and CK19-FITC and hMAM-were added after overnight adhesion. QDs were used for 1 h, washed twice with PBS, digested with trypsin, centrifuged at 1000 rpm for 5 min, collected cells, resuspended with 300 µL of PBS containing 1% FBS, and analyzed by ow cytometry.

Process of isolation and identi cation breast cancer CTC
The steps for CTC separation and identi cation of breast cancer were as follows: (1)  Clinical study on isolation and identi cation of circulating tumor cells ELMP was used to separate CTC from 120 peripheral blood samples (20 normal and 100 patient samples), Hoechst 33342, anti-CK19-FITC, and PE-labeled CD45 antibody were used for the separated CTC identi cation in single tumor marker method. Hoechst 33342, anti-CK19-FITC, and hMAM-QDs were used for the separated CTC identi cation in double tumor marker method. The CTC positive determination criteria are greater than 5 in 7.5 mL blood.

Statistical Analyses
Data were analyzed using descriptive statistics, for multiple comparisons, a one-way ANOVA test was performed, and Data are expressed as mean ± standard deviation (S.D.) derived averaged across three to at least three independent measurements. Comparisons among different groups were performed by Chi-square tes. Differences were considered as signi cant and the null hypothesis of no difference among treatment groups was rejected for P values < 0.05.

Results And Discussion
Preparation and characterization of ELMP ELMP was prepared by indirect method (Fig. 1), the antibodies were modi ed on the surface after the preparation of the liposome magnetic particles, raw Fe 3 O 4 beads were embedded inside of the liposome, and EpCAM antibodies were modi ed outside the liposome.
As shown in Fig. 2A, the average diameter of ELMP was 90.1 ± 0.84 nm (in ddH 2 O and at 25℃), with a polydispersity index of 0.08. Surface Zeta potential of ELMP was 16.7 ± 0.32 mV( Figure 2B). Transmission electron microscope results was shown in Fig. 2C, the diameter of ELMP was around 30 ~ 90 nm in dehydration state. The magnetizing curve under 300 K temperature is shown in Fig. 2D, the result showed that maximum saturated magnetization of the Fe 3 O 4 raw beads was 43.5 Am 2 /kg, and the saturated magnetization of ELMP was 31.3 Am 2 /kg, accounting for 72.0% of the saturated magnetization of the Fe 3 O 4 raw beads, this suggesting that liposome wrapped the surface of the Fe 3 O 4 raw beads. The solid content of ELMP was 2.8 ± 0.15 mg/mL, which fully complied with the test requirements, so the antibody and lipid components were distributed around the magnetic NPs, thereby reducing the saturated magnetization.

hMAM expression analysis in breast cancer tissue
The expression of hMAM in benign and malignant breast cancer tissues was analyzed by immunohistochemistry. As shown in Fig. 3A, the expression of hMAM in low-grade benign breast cancer tissues was very low, and the expression of hMAM in high-grade benign tumors increased slightly, however the expression of hMAM in malignant breast cancer tissues was high. Figure 3B shows the pathological results of four common malignant tumors, we can clearly nd that threr was almost no hMAM expression in other tumor tissues. At the same time, it can be seen from

Isolation and identi cation of CTC in peripheral blood
By magnetic separation, CTC separation of breast cancer clinical blood samples was carried out by means of ELMP. Immuno uorescence staining was performed on the separated cells by two methods, and uorescent micrographs were taken for statistical analysis of CTC. Observed by uorescence microscope, the CTC results identi ed by the two methods are shown in Fig. 7. Figure 7A shows that CK19-FITC stains green, cell nucleus stains blue, and CD45 staining is negative to determine breast cancer CTC. Figure 7B shows that CK19-FITC stained green, nuclei stained blue, and hMAM-QDs stained cells showing red uorescence were judged as breast cancer cells. The CTCs that were positive in both determination methods showed large cells, a high proportion of nuclei, and irregular cell morphology, which was consistent with the CTC morphological consensus.

Comparative analysis of clinical results
CTC was isolated from the blood samples of 20 healthy volunteers and 100 breast cancer patients. Two methods were used for CTC identi cation. The results are shown in Table 1 after statistics. The CTC statistical results of 20 healthy volunteers showed that 1 case of CTC was misdiagnosed as positive with the single label method, and the positive rate was 5%. There was no CTC misdiagnosis with the single label method, and the positive rate was 0%.
There was no signi cant difference in statistical results (p = 0.311). The CTC test results of clinical blood samples of 100 breast cancer patients showed that the CTC positive rate using the single tumor marker method was 82%, and the CTC positive rate using the double tumor marker method was 92%. The statistical results of the two methods were signi cantly different (p = 0.036). The ROC curve was drawn based on the CTC statistical results of 120 clinical blood samples, and the speci city and sensitivity of the two methods were compared. As shown in Fig. 8 and Table 2, the distribution of the area under the curve of the single tumor marker method at the 95% con dence interval was 0.819 ~ 0.039, the area under the curve is 0.889, the distribution of the area under the double tumor marker method in the 95% con dence interval is 0.908 to 0.987, and the area under the curve is 0.96. The above results indicate that the use of the double standard method can reduce the false positives of CTC identi cation results. And the probability of false negatives, improve detection accuracy.  Table 3, which showed that there was no correlation between CTC detection results and age and pausimenia, but close correlation with blood metastasis, Lymphatic metastasis, P < 0.05.There was a signi cant difference between the positive CTC and TNM stage by the double-standard method (p = 0.019 < 0.05).However, there was no signi cant difference between the positive results of CTC and TNM staging by single tumor marker method (p = 0.205 > 0.05).
TNM stage is an important indicator of tumor. The higher the stage, the higher the risk of metastasis in theory and the corresponding CTC test result should be positive. Therefore, the double tumor marker method can improve the accuracy of CTC test for breast cancer, with higher consistency with the pathological test results.   The preparation method of EpCAM antibody modi ed liposome magnetic particles (ELMP). EpCAM antibodies were modi ed on the surface after the preparation of the liposome magnetic particles, raw Fe3O4 beads were embedded inside of the liposome.