Expression profiling of serum microRNA-101 in HBV-associated chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma

Abstract

MicroRNAs (miRNAs) represent a class of evolutionarily conserved, non-coding small RNAs (18–25 nt) that have emerged as master regulators of several biological processes. Recently, circulating miRNAs have also been reported to be promising biomarkers for various pathological conditions. In the present study, we report the comparative expression profiling of microRNA-101 (miR-101) in serum and tissue samples from chronic hepatitis B (CHB), HBV-associated liver cirrhosis (HBV-LC), and HBV-associated hepatocellular carcinoma (HBV-HCC) patients and healthy controls. The serum miR-101 levels were found to be significantly downregulated in the HBV-HCC patients compared with the HBV-LC patients (P < 0.001), CHB patients (P < 0.001) and healthy controls but were upregulated in the HBV-LC patients compared with the CHB patients (P < 0.001) and healthy controls (P < 0.001). Consistent with the serum data, the expression of miR-101 was also upregulated and downregulated in the HBV-LC and HBV-HCC tissue samples, respectively. A receiver operating characteristic (ROC) analysis of serum miR-101 yielded an area under the ROC curve (AUC) of 0.976 with 95.5% sensitivity and 90.2% specificity when differentiating between HBV-HCC and HBV-LC. Our results suggest that the serum miR-101 level can serve as a potential non-invasive biomarker to differentiate HBV-HCC from HBV-LC.

Introduction

Hepatocellular carcinoma (HCC) is the fifth most common cancer worldwide and ranks as the third leading cause of cancer mortality.¹ Although there are different viral and non-viral causes of HCC, it has been reported that almost 80% of disease cases are associated with hepatitis B virus (HBV) infection. HCC is often diagnosed at an advanced stage when curative resection of the tumor is no longer appropriate because of intrahepatic and extrahepatic metastases. At present, the diagnosis of HCC relies on the identification of a liver mass by radiological imaging tests, including ultrasonography, CT, and/or magnetic resonance imaging (MRI). These imaging techniques are often combined with the use of serum tumor markers, such as α-fetoprotein (AFP), which has been discontinued in several western countries due to its poor sensitivity and specificity, but this protein is still widely used as a serum tumor marker in many Asian countries including China.^2,3

MicroRNAs (miRNAs) are small noncoding RNA (18–24 nucleotides) molecules that have been reported to play important roles in the regulation of several biological processes, such as cell proliferation, differentiation and apoptosis.^4,5 The ability of miRNAs to bind specific sites in the 3′-UTR of target mRNAs leads to translational repression and/or mRNA degradation.⁶ The deregulation of miRNA function can often lead to disorder of the internal environment and eventually to carcinogenesis. Many miRNA genes are located in cancer-associated genomic regions or in fragile sites that are frequently amplified or deleted in human cancers, suggesting that these miRNAs play an important role in malignant transformation.⁷ Apart from their tissue-specific origin and expression, it has recently been shown that miRNAs are also stable and detectable in various body fluids such as serum and plasma.⁸ Similar to the expression patterns in tissue specimens, the expression of circulating miRNAs has been shown to vary between disease cases and healthy controls for several types of cancers and diseases. This observation suggested that the circulating miRNAs could serve as useful, non-invasive biomarkers of disease stage and progression. In the case of HCC, the circulating levels of several miRNAs including miR-21, miR-122, miR-15b, and miR-130b have been reported as potential diagnostic markers.^9,10

Recently, genome-wide miRNA expression profiling has revealed overexpression of miR-101 in HCC tissues compared with normal/healthy controls.¹¹ The aberrant expression and therapeutic potential of miR-101 was also demonstrated in another study showing that miR-101 is downregulated by the HBV X protein and induces epigenetic modifications in HBV-associated HCC (HBV-HCC) tissues.¹² Although aberrant expression of circulating miR-101 has been reported previously in HBV-HCC, parallel expression profiling of miR-101 from the serum and tissues of patients at different stages of HBV-associated liver diseases have not yet been reported.¹³ The development of HCC is a multi-step process, and it is also unknown if serum miR-101 could be used as a marker for differentiating the different stages of HBV-associated liver disease. Therefore, in the present study, we performed serum miR-101 expression profiling of HBV-associated chronic HBV (CHB), HBV-associated liver cirrhosis (HBV-LC), and HBV-HCC patients to determine this marker’s diagnostic potential. Moreover, the hepatic miR-101 expression was also analyzed. For simplicity, the miR-101 that we analyzed in the present study will be referred to as miR-101 throughout the manuscript.

Results

Expression of miR-101 in liver tissues

First, we measured the expression of miR-101 in CHB, HBV-LC, HBV-HCC, and control tissue samples via in situ hybridization. The patients used for in situ hybridization were independent from those used for serum miRNA profiling, and their characteristics are given in Table S1. The expression levels of miR-101 in the CHB tissue samples were not significantly different from the levels in the normal liver tissue samples (P = 0.999). On the other hand, compared with the normal samples, the expression levels of miR-101 were upregulated in the HBV-LC tissue samples (P = 0.011) yet downregulated in the HBV-HCC tissue samples (P = 0.014) (Fig. 1A–E).

Figure 1. The expression levels of miR-101 in liver tissues. In situ hybridization was performed using locked nucleic acid (LNA)-modified probes for miR-101. Brown yellow staining in the cytoplasm represents positive staining for miR-101. (A) Normal liver tissues were graded positive for miR-101. (B) CHB liver tissues were graded positive for miR-101. (C) HBV-LC tissues were graded strong positive for miR-101. (D) HBV-HCC tissues were graded negative for miR-101. (E) The level of miR-101 expression in HBV-LC patients is higher than in healthy subjects and CHB patients, while it is lower in HBV-HCC patients. (* represents P < 0.05.)

Expression profile of circulating miRNA-101

To investigate whether levels of circulating miR-101 are also altered in patients with CHB, HBV-LC and HBV-HCC, the concentrations of miR-101 were measured in the sera from patients and healthy controls using quantitative real-time PCR. Consistent with our in situ hybridization data, although no significant difference (P = 1.000) was observed in the miR-101 serum levels between the control and CHB samples (Fig. 2), we found that serum miR-101 was significantly upregulated in the HBV-LC group compared with the CHB patients (P < 0.001) and control subjects (P < 0.001) (Fig. 2). We next investigated the levels of serum miR-101 in the HBV-HCC patients compared with the HBV-LC, CHB, and healthy subjects. The results showed that serum miR-101 was significantly downregulated in the HBV-HCC patients compared with the HBV-LC (P < 0.001), CHB (P < 0.001), and healthy subjects (P = 0.003) (Fig. 2). Next, to explore the possible association between miR-101 and the risk of HBV-HCC derived from HBV-LC, binary logistic regression was used controlling for gender, age, alcohol status, and ALT at the same level. A significantly decreased risk of HBV-HCC was found to be associated with upregulated miR-101. The odds ratio of developing HBV-HCC with upregulated miRNA-101 was 0.088 (95% CI = 0.032–0.244, P < 0.001). The correct percentage of the logistic regression model was 92.1%.

Figure 2. Serum miR-101 levels in healthy subjects, patients with CHB, patients with HBV-LC and patients with HBV-HCC. The median level of serum miR-101 was significantly higher in patients with HBV-LC than in patients with CHB or healthy subjects (P < 0.001), whereas the median level was significantly lower in patients with HBV-HCC than in healthy subjects (P = 0.012) or patients with HBV-LC (P < 0.001). The lines represent the range, median and quartiles of relative miRNA expression (ΔCt values) measured by qRT-PCR.

The diagnostic potential of serum miR-101

To evaluate whether serum miR-101 can be used as a diagnostic marker for CHB, HBV-LC, and HBV-HCC, ROC curve analysis was performed. In the case of CHB compared with healthy controls, the AUC was 0.635 (95% CI: 0.548–0.716) with a sensitivity/specificity ratio of 84.8/40.4%, indicating that serum miR-101 may not be an optimal marker for CHB diagnosis (data not shown). This result parallels the lack of a significant difference in the serum miR-101 levels between CHB and healthy controls described above. In the case of HBV-LC compared with healthy controls, the AUC was 0.884 (95% CI: 0.800–0.942), and the sensitivity and specificity calculated at the cut-off value of 10.50 were 80.3% and 80.0%, respectively (Fig. 3A). For HBV-HCC compared with healthy controls, the AUC was 0.788 (95% CI: 0.693–0.865), and at the cut-off value of 8.99, the sensitivity and specificity were 76.1% and 70.0%, respectively (Fig. 3B).

Figure 3. ROC curve analysis of serum miR-101 for discriminating; (A) HBV-LC patients and healthy subjects, (B) HBV-HCC patients and healthy subjects.

We next investigated whether serum miR-101 has the potential to serve as a marker of disease progression or to differentiate between CHB, HBV-LC, and HBV-HCC patients. For the discrimination between patients with CHB and either HBV-LC or HBV-HCC, miR-101 had an AUC of 0.861 (95% CI: 0.792–0.913) (Fig. 4A) and 0.777 (95% CI: 0.701–0.842), respectively (Fig. 4B). At the cut-off values of 10.91 (CHB compared with HBV-LC) and 9.70 (CHB compared with HBV-HCC), serum miR-101 had sensitivity/specificity ratios of 75.4/83.5% and 88.1/62.0%, respectively. In the case of HBV-LC compared with HBV-HCC, the AUC was 0.976 (95% CI: 0.931–0.995), and at a cut-off value of 10.08, the sensitivity and specificity of serum miR-101 were 95.5% and 90.2%, respectively (Fig. 4C).

Figure 4. ROC curve analysis of serum miR-101 for discriminating; (A) CHB patients from HBV-LC patients, (B) CHB patients from HBV-HCC patients, (C) HBV-LC patients from HBV-HCC patients, (D) AFP alone for discriminating HBV-LC from HBV-HCC patients. (E) AFP in combination with miR-101 for discriminating HBV-HCC from HBV-LC patients.

Next, the predicted probability of being diagnosed with HBV-HCC from a stepwise logistic regression model based on AFP and miR-101 was used to construct the ROC curve. The AFP alone yielded an AUC of 0.762 (95% CI: 0.674–0.836) with 44.8% sensitivity and 96.6% specificity when differentiating HBV-LC from HBV-HCC (Fig. 4D). The AUC of serum miR-101 combined with AFP was also calculated and found to be 0.973 (95% CI: 0.925–0.994) with a sensitivity/specificity ratio of 96.6/87.9% (Fig. 4E). This result is similar to the ROC analysis of miR-101 alone for differentiating HBV-LC from HBV-HCC (Fig. 4C).

There was no correlation between the serum miR-101 expression levels and the patients’ clinical features, such as age, gender, ALT, T-Bil, PT, viral load, tumor size, Child–Pugh grade, AFP, or TNM classification (data not shown).

Discussion

The aberrant expression of miRNA-101 has been recently reported in several cancers, including ovarian, renal, prostate, gastric, and lung cancer,^14-20 highlighting miR-101 as a bona fide tumor suppressor miRNA. Downregulation of miRNA-101 in association with HCC has also been reported in different studies that have collectively shown that miRNA-101 expression has the potential to inhibit HCC cell proliferation, suppress tumorigenicity, and promote apoptosis by modulating the expression of multiple transcription factors and cell cycle-related genes in HCC.^21-23For instance, Su et al. reported the downregulation of miR-101 in HBV-HCC tissues and cell lines compared with controls. This study suggested that miR-101 downregulation is a frequent event in human HCC tissues that leads to hepatocarcinogenesis through the overexpression of Mcl-1, an antiapoptotic member of the Bcl-2 family, which has a complementary binding site for the miR-101 in its 3′-UTR that is a direct target of miR-101.¹⁹ Similarly, Zhang et al. reported the downregulation of miR-101 and parallel overexpression of SOX9, another direct target of miR-101, in clinical HCC tissues and the correlation of miR-101 downregulation with tumor aggressiveness and poor prognosis.²²

As the association between miR-101 and HCC has been strengthened by the available data, it has been suggested that miR-101 could serve as a potential prognostic marker and therapeutic target for HCC.^21-23 We were interested in determining the differences in miR-101 expression at different stages of HBV-associated liver diseases (CHB, HBV-LC, and HBV-HCC) and, moreover, whether miR-101 could serve as a non-invasive diagnostic biomarker. For this purpose, we performed expression analysis of miR-101 in the liver tissue and serum samples from CHB, HBV-LC, HBV-HCC, and control groups by in situ hybridization and real-time PCR, respectively. Two independent study groups were used for tissue and serum miRNA-101 analysis, and collectively, the results from each group were in agreement. We showed that the miRNA-101 levels in the sera and liver tissues were down- regulated in HBV-HCC patients but upregulated in HBV-LC patients compared with CHB patients and healthy controls. To evaluate the diagnostic potential of serum miR-101, ROC curves were generated. Our data showed that serum miR-101 is superior to AFP for diagnosing HBV-HCC derived from HBV-LC, as combining serum miR-101 with AFP conferred no advantage over serum miR-101 alone for detecting HBV-HCC derived from HBV-LC. Combining the data from the tissue and serum miRNA levels, ROC curves and binary logistic analysis, we propose that serum miRNA-101 could serve as a non-invasive biomarker to monitor HBV-HCC that is derived from HBV-LC and also the progression of CHB infection to HBV-LC.

The progression of HBV-associated liver disease to HCC consists of several stages. Additionally, the expression of several miRNAs is stage-dependent in many diseases; some miRNAs show early expression, whereas some are expressed only during the advanced stages of disease. For instance, the liver-specific miRNA-122 has been reported to be upregulated in the tissue and serum samples of CHB patients, yet downregulated in the tissue and serum samples of HBV-HCC patients.²⁴ Other examples include miR-25 and let-7f, which were altered significantly in HBV-HCC patients but unchanged in CHB patients compared with controls.²⁵ Based on our current analysis, we also suggest that miR-101 is a potential disease progression marker; while we observed no significant difference in the expression of miR-101 in tissue and serum samples from CHB patients compared with healthy controls, a significant expression difference emerged when comparing these samples to those from HBV-LC patients (upregulation) and HBV-HCC patients (downregulation). In our opinion, however, the lack of a significant difference between the CHB and control samples in the present study can be attributed to the following factors. First, the group used for serum analysis in this study consisted of CHB patients with low staging and grading scores, which therefore did not have high hepatic activity index (HAI) scores. However, we also suspect that the inclusion of patients with higher inflammation and fibrosis scores in the CHB group may bias the data toward a false-positive result, as patients with liver cirrhosis usually have higher HAI scores and the inclusion of these patients in the CHB group may lead to an increase in the total level of miR-101 expression in this group. Another possible explanation is that the transcription factors associated with liver inflammation may not be targets of miR-101; the previously reported targets of miR-101 that are liver-associated transcription factors have mostly been those involved in hepatocarcinogenesis and advanced-stage liver diseases.^21,22 However, these speculations warrant further investigations with a larger sample size. Similarly, the inclusion of matched liver samples before and after HCC development could lead to a better understanding of the dynamics of miR-101 expression in HCC progression as miRNA-101 has been linked to tumor migration and recurrence.^16,26,27 Finally, whether miR-101 has the same function in other liver disease and HCC etiologies such as HCV and alcoholic cirrhosis remains unclear.

Taken together, the findings from our present study demonstrate that miR-101 levels are elevated in the sera and liver tissues of HBV-LC patients and decreased in HBV-HCC patients, which suggests that serum miR-101 is a potential biomarker for monitoring the development of HBV-HCC from HBV-LC and the development of HBV-LC from CHB.

Materials and Methods

Study subjects and clinical characteristics

The study protocol was approved by the Ethics Committee of Beijing YouAn Hospital, Capital Medical University and adhered to the tenets of the Declaration of Helsinki. Written informed consent was obtained from the participants for the use of their blood, tissue samples and clinical records in this study. The study population consisted of 67 HBV-HCC patients, 61 HBV-LC patients, 79 CHB patients and 30 healthy subjects. The patients’ characteristics are summarized in Table 1. All patients were positive for HBsAg and did not have any other liver diseases, such as CHC, alcoholic liver disease, autoimmune liver disease, or metabolic liver disease, based on clinical reports. The degree of inflammation of all of the patients was between G₁ and G₂. The fibrosis staging of the CHB patients was between S0 and S2. The fibrosis staging of the HBV-LC and HBV-HCC patients was between S3 and S4. The blood samples from the HBV-HCC patients were obtained before the surgical resections were performed. The data on all of the subjects were obtained from medical records, pathology reports, and personal interviews with the subjects. The data collected included age, gender, serum albumin (ALB) level, total bilirubin level (T-Bil), alanine aminotransferase (ALT) level, prothrombin time (PT), HBV DNA viral load, AFP, tumor number and size, tumor differentiation, tumor stage, relapse time, and time of death. The clinical stage of HBV-HCC was evaluated based on the TNM classification system. Child–Pugh scoring was performed to categorize the LC and HCC patients at Child–Pugh grades A, B, or C.

Table 1. Summary of clinical characteristics of CHB, HBV-LC, HBV-HCC and healthy subjects.

		Group I: CHB (n = 79)	Group II: HBV-L (n = 61)	Group III: HBV-HCC (n = 67)	Group IV: Control (n = 30)	P values
Age (y)	Mean ± SD	32.38 ± 11.93	51.19 ± 9.57	51.69 ± 10.43	37.26 ± 10.79	<0.001
Gender n (%)	Male	54 (68.4)	43 (70.5)	57 (85.1)	21 (70)	0.089
	Female	25 (31.6)	18 (29.5)	10 (14.9)	9 (30)
Alcohol n (%)	Yes	14 (17.7)	14 (23.0)	12 (17.9)		0.506
	No	55 (69.6)	44 (72.1)	46 (68.7)
	Unknown	10 (12.7)	3 (4.9)	9 (13.4)
ALT (IU/L) n (%)	Elevated	25 (31.7)	28 (45.9)	34 (50.7)		0.175
	Normal	52 (65.8)	32 (52.5)	31 (46.3)
	Unknown	2 (2.5)	1 (1.6)	2 (3.0)
Total bilirubin (mg/dL); n (%)	Elevated	18 (22.8)	48 (78.7)	26 (38.8)		<0.001
	Normal	58 (73.4)	12 (19.7)	39 (58.2)
	Unknown	3 (3.8)	1 (1.6)	2 (3.0)
Viral load (log)	Median (range)	7.02 (0–7.9)	2.77 (0–7.62)	0 (0–7.26)		0.422
AFP	Median (range)	2.12 (0.66–34.1)	5.02 (0.77–222.3)	31.79 (0.72–17342)		<0.001
Grade n (%)	1–2	29 (36.7)		67 (100)
	3–4	0		0
	Unknown	50 (63.3)		0
Stage n (%)	0	15 (19.0)		0
	1–2	14 (17.7)		0
	3–4	0		67 (100)
	Unknown	50 (63.3)		0
Tumor size (cm); n (%)	<5			46 (68.7)
	≥5			21 (31.3)
Tumor differentiation n (%)	Well			8 (11.9)
	Moderate			40 (59.7)
	Poor			19 (28.4)
Tumor stage n (%)	Stage I			42 (62.7)
	Stage II			6 (9.0)
	Stage III–IV			19 (28.3)

Serum preparation

Ten milliliters of peripheral blood was collected from each individual directly into serum tubes at the time of liver biopsy or before surgery. The tubes were initially centrifuged at 1500 g for 10 min. The serum was then aliquoted and additionally centrifuged at 13 000 g for 15 min at 4 °C to completely remove the cell debris and any remaining cells. The supernatant was then transferred to new 2.0-mL tubes without disturbing the pelleted debris. Any samples with signs of hemolysis were excluded from the study. The serum samples were then stored at −80 °C until further use.

RNA extraction and reverse transcription

Total RNA was extracted from 200 μL of serum using a QuantoBio Total RNA Isolation Kit (QuantoBio Co.) following the manufacturer’s protocol. Briefly, 200 μL of lysis buffer and 40 μL of lysis enhancer solution were added to each sample and vortexed for 15 s. For the normalization of sample-to-sample variation during RNA isolation and as an internal control, 25 fmol of synthetic Caenorhabditis elegans miRNA-39 (cel-miR-39-3p) was added to each lysed sample as described previously.⁸ This step was followed by the addition of 440 μL of acidified phenol:chloroform and vigorous shaking for 30 s. The samples were centrifuged at 16 000 g in a microcentrifuge at 4 °C for 10 min to separate the aqueous phase (upper phase) and the organic phase (lower phase). The aqueous phase was then transferred to new RNase-free microcentrifuge DNA LoBind tubes (Eppendorf). Ethanol was added to the aqueous phase to a final concentration of 30%, and this mixture was vigorously shaken for 15 s. Then, the 650 μL mixtures were centrifuged at 10 000 g at room temperature for 30 s to pellet the RNA. After the washing steps, the total RNA including miRNAs was resuspended in 50 μL RNase-free water and stored at −80 °C until further use. The concentration and purity of the total RNA was determined using a NanoDrop 8000 (Thermo Fisher Scientific).

cDNA was reverse transcribed from the total RNA via the polyadenylation method. Briefly, 1 ng of the total RNA template was mixed with 2 μL 10× buffer, 2 μL dATP (10 mM), 0.5 μL poly(A) polymerase (NEB), 0.5 μL RNase inhibitor (Promega), and RNase-free water (Promega) to a final volume of 20 μL and incubated at 37 °C for 1 h. Then, 1 μL of 0.5 μg/μL RT primer was added, and the reactions were incubated at 70 °C for 5 min followed by immediate incubation on ice for at least 2 min to disrupt the secondary structures of the RNA and the primer. In the final step, the 20 μL reaction mixture from the above step was mixed with 4 μL 5× buffer, 1 μL dNTP (10 mM) (Sigma), 0.5 μL M-MLV reverse transcriptase (Promega), 0.5 μL RNase inhibitor (Promega), 10 μL A-Plus reaction mix, and 4 μL RNase free water (Promega), and incubated at 42 °C for 1 h. The undiluted cDNA samples were then stored at −20 °C in 0.2 mL DNA LoBind microcentrifuge tubes (Eppendorf) until further use.

Real-time quantitative PCR (qPCR)

The real-time PCR was performed in a total reaction volume of 20 μL, which contained 1 μL of diluted cDNA (1:100), 10.4 μL 2× qPCR mix with ROX, 2 μL 10× universal reverse primers, 2 μL 10× microRNA-specific forward primers (QuantoBio), and 4.6 μL PCR grade RNase-free water. The PCR reactions were run on an ABI 7900 HT real-time PCR system (Applied Biosystems) with the following cycle conditions: 95 °C for 5 min, followed by 40 cycles of 95 °C for 30 s, 60 °C for 1 min and a dissociation stage of 95 °C for 15 s, and finished at 60 °C for 1 min. All reactions were performed in duplicate. Ct values over 35 were defined as undetectable. The relative expression levels of the miRNAs were calculated using the ΔCt method. The miRNA levels in the serum samples were normalized against the spiked-in cel-miR-39 because the circulating levels of the potential microRNA reference genes are also increased in the serum of cancer patients, and normalization to a spiked-in synthetic microRNA controls for variable RNA isolation efficiency across the samples.²⁸

In situ hybridization

A locked nucleic acid (LNA)-modified probe for miR-101 (mercury-LNA detection probe, Exiqon) was used. The probe sequence was 5′-AGCATCAGCA CTGTGATAAC TG-3′. In situ hybridization was performed using an in situ hybridization kit (Bohengkechuang Co.). The in situ hybridization steps were performed as previously described.²⁹ The staining was evaluated by two investigators who were unaware of the study objectives and graded as follows: negative (−), positive (+), and strong positive (++). Paraffin-embedded tissue samples from HCC, LC and CHB patients and healthy individuals were obtained from the Alenabio Company. Details of the clinical data are provided in Table S1.

Statistical analysis

Statistical analysis was performed using the SPSS 19.0 software (SPSS, Inc.) and MedCalc (11.4.2.0) software. The differences between groups were evaluated using a Pearson Chi-square, Kruskal–Wallis, or ANOVA test with post-hoc testing. The association between miRNA-101 and the risk of developing HBV-HCC from HBV-LC was analyzed using binary logistic regression, controlling for age, gender, alcohol status, and ALT. Hosmer and Lemeshow tests were used to calculate the goodness-of-fit of the model. A Spearman non-parametric rank test was used to test the correlations between the expression of serum miR-101 and the clinical parameters. Receiver operating characteristic (ROC) curves were generated to classify patients in different groups, as well as for the evaluation of the diagnostic potential of serum miR-101 via calculation of the area under the ROC curve (AUC), sensitivity and specificity according to standard formulas. P values were two-sided, and values less than 0.05 were considered to be statistically significant.

Glossary

Abbreviations:

CHB

chronic hepatitis B

LC

liver cirrhosis

HCC

hepatocellular carcinoma

ROC

receiver operating characteristic

AFP

α-fetoprotein