Stemness CD24 activation promotes hepatocellular carcinoma progression via an immune escape mechanism

Abstract

BACKGROUND

Cluster of differentiation 24 (CD24) serves as a liver cancer stem cell marker, and its upregulation is related to chronic liver disease malignancy. However, the exact relationship between CD24 expression and hepatocarcinogenesis remains unknown.

AIM

To investigate CD24 levels among individuals with chronic liver diseases and confirm the alterations in CD24 and programmed death-ligand 1 (PD-L1) expression in a dynamic model of rat hepatocarcinogenesis.

METHODS

Approved by the ethics committee, CD24 levels were detected in the serum of 129 patients with hepatocellular carcinoma (HCC), 72 patients with chronic hepatitis (CH), 60 patients with liver cirrhosis (LC) and 111 normal control (NC). Receiver operating characteristic curves and clinicopathological characteristics of CD24 were used to evaluate the diagnostic or prognostic value for HCC, and the CD24⁺ T lymphocyte ratio and dynamic alterations in CD24 and PD-L1 expression were confirmed in a rat HCC model.

RESULTS

Compared with those in the CH, LC and NC groups, the average CD24 level in the HCC group was significantly greater (P < 0.01). The CD24⁺ T-cell ratio in the HCC group was greater than that in the CH or NC group. Clinicopathological characteristics of high CD24 in HCC patients included hepatitis B virus infection, single/multicenter status, tumor size, lymph node or extrahepatic metastasis, differentiation degree, tumor-node-metastasis grade, Child-Pugh score, portal vein tumor thrombus and poor prognosis. Mechanistically, CD24 is dynamically upregulated during hepatocarcinogenesis and closely positively correlated with PD-L1 for immune escape, metastasis (CD44), and with respect to HCC markers (Wnt3a, GPC-3 and alpha-fetoprotein).

CONCLUSION

Activated CD24 promoted HCC formation through programmed death-ligand 1 signaling and could be a valuable biomarker for monitoring chronic liver disease malignancy.

Key Words: Clusters of differentiation 24; Hepatocarcinogenesis; Programmed death-ligand 1; Biomarker; Chronic liver disease

Core Tip: Upregulated clusters of differentiation 24 (CD24) expression is associated with hepatocellular carcinoma progression. CD24, a marker of liver cancer stem cells, might exert its oncogenic regulatory effects on the malignancy of chronic liver diseases through the activation of programmed death-ligand 1 signaling. Clinical investigations and hepatocarcinogenesis models have confirmed that upregulated CD24 promotes hepatocyte malignancy via immune escape mechanisms and could be a favorable biomarker for monitoring hepatocarcinogenesis.

INTRODUCTION

Early diagnosis and effective treatment of hepatocellular carcinoma (HCC) are still challenging problems worldwide[1]. HCC is associated with chronic persistent infection with hepatitis B virus (HBV) or hepatitis C virus (HCV)[2], chemical carcinogens, and metabolic dysfunction-associated steatotic liver disease (MASLD)[3], among other pathological factors[4]. The pathogenesis of HCC underlying such an association is complex, especially in terms of liver cancer stem cells (LCSC) activation during chronic liver disease malignancy[5]. Malignant transformation of hepatocytes is induced by the activation of oncogenes, inactivation of anti-oncogenes, and reactivation of certain oncogenes normally expressed only during the embryonic stage [alpha-fetoprotein (AFP)], which leads to the expression of a variety of specific biomarkers. such as clusters of differentiation (CD) that are subsequently secreted into the blood, including LCSCs and signaling molecules[5,6]. Various populations of LCSC with different phenotypical markers (EpCAM, CD133, CD44, CD13, CD90, OV-6, CD47, and side populations) in HCC have been identified or validated in xenotransplantation models with risk factors, prognosis, chemoresistance, and distal metastasis[7]. However, except for those of CD44, their carcinogenic mechanisms have not been substantially explored[8].

CD24 is a small, heavily glycosylated and mucin-like cell surface glycoprotein that is highly expressed in stem/progenitor cells, has been linked to LCSCs derived from many types of cancer and has been shown to exhibit oncogenic properties[9,10]. CD24 mRNA overexpression is common in HCC cells in parallel with p53 mutation and tumor differentiation. LCSCs might be involved in distinguishing danger- and pathogen-associated patterns and promoting cell migration, invasion, and proliferation, thereby driving tumorigenesis and progression[11,12]. Recently, clinical data have shown that CD24 or CD44 is significantly associated with HBV infection as a contributing factor to hepatocyte inflammation[13-15]. Additionally, the upregulated expression of programmed death-ligand 1 (PD-L1) in HBV-related HCC is unclear but is associated with poor prognosis[16-18]. However, the mechanisms underlying the activated CD24 and upregulated PD-L1 in HCC remain to be elucidated. This study aimed to investigate CD24 levels in patients with HBV-related chronic liver diseases and establish a model of hepatocarcinogenesis in rats to verify the dynamic expression of CD24 and PD-L1 and corresponding pathological alterations in hepatocytes.

MATERIALS AND METHODS

Patients and data

This study was approved by the Ethics Committees of the Affiliated Hospital of Nantong University (No. 2018-L026), and informed consent was obtained from all patients. This study included 261 consecutive patients with HBV-related chronic liver diseases and 111 healthy persons as normal control (NC) (Table 1) from March 2018 and December 2019. Patients included in the study were diagnosed with HCC based on postoperative pathology and tumor histological type, in accordance with the diagnostic criteria formulated by the Chinese Anti-Cancer Association in 2019, including HBV infection, liver function classified as Child-Pugh grade A or B, no prior anticancer treatment before surgery.

Table 1 Analysis of clinical data of patients with chronic liver diseases.

Group	n	Male/female	Age (years)	HBsAg (+/-)	AFP (ng/mL)			Child-Pugh
					≤ 20	21-399	≥ 400	A	B	C
NC	111	57/56	22-79	0/111	111	0	0	-	-	-
CH	72	55/17	23-75	72/0	55	17	0	-	-	-
LC	60	40/20	32-88	60/0	42	17	1	48	9	3
HCC	129	93/36	37-86	97/32	38	53	38	97	23	9

AFP: Alpha-fetoprotein; NC: Normal control; CH: Chronic hepatitis; HCC: Hepatocellular carcinoma; HBsAg (+/-): Hepatitis B virus surface antigen (positive/negative); LC: Liver cirrhosis.

Blood collection and analysis

Blood samples from patients with HCC, chronic hepatitis (CH), liver cirrhosis (LC) form the Affiliated Hospital of Nantong University or Xinghua People’s Hospital whose clinical records were completed were collected, and the healthy population from physical examination served as the NC group from the Physical Examination Center of the Affiliated Hospital of Nantong University. Hepatitis viral markers (HBsAg, HBcAb, HBV-DNA, and anti-HCV) were negative, with serum alanine aminotransferase (ALT) levels. AFP concentrations were detected with chemiluminescence by Alinityi analyzer (Abbott Shanghai, China) under normal range (< 25 μg/L). Activities of ALT and aspartate aminotransferase (AST) were tested by Kits (Nanjing Built-in Biotechnology Co., Ltd.). HBV surface antigen (HBsAg), HBeAg, and HBV-DNA was analyzed according to the instructions of Kit on the LightCycler480 fluorescence PCR instrument (Roche Shanghai, China), with the requirements of the instructions to detect absorbance value (A) by multi-functional enzyme meter (Bio-Tek Company, United States), and calculate the concentration according to the standard curve.

Dynamic hepatocarcinogenesis model

Male Sprague-Dawley (SD) rats (n = 56), 4 weeks old and weighing 110-120 g, were provided by the Animal Center of Nantong University (S20200318-017) and were randomly divided into a control group (n = 12) and a model group (n = 42). The rats were raised in an environment with a temperature of 22 °C ± 2 °C, a 12-hour light/dark cycle, and 55% humidity. The control group of SD rats was fed ordinary pellet food, and the experimental rats were fed pellet food containing 2-fluoreneacetamido (0.05% 2-FAA, United States). Two control rats and two groups of experimental rats were sacrificed every two weeks, approximately 5 mL of blood was collected from the heart through abdominal anesthesia with ether, and serum was separated and stored at -20 °C. Livers were removed through the abdomen, frozen in liquid nitrogen, cut into slices or tissue homogenates, and fixed in neutral formalin (10%, V/V), after which the paraffin-embedded sections were stained with hematoxylin and eosin (H&E) and immunohistochemical staining, observed or photographed under an OLYMPUS IX71 microscope, and image analysis was performed by Image-Pro Plus 6.0 software.

Tissue protein preparation

The protein extracts from 100 mg of liver tissue were homogenized in lysis buffer according to RIPA (Multi Sciences Biotechnology Corporation Limited, WB020) and PMSF (Multi Sciences Biotechnology Corporation Limited) manual step preparation and centrifuged at 12000 rpm for 15 minutes. A bicinchoninic acid (BCA) assay was performed to measure the protein concentrations, and the protein samples were stored in a refrigerator at -80 °C.

Hepatic mRNA transcription

Total RNA was purified from 100 mg of liver tissue in RNA^later (Ambion, TX, United States) according to the instructions of the TRI^zol reagent (Gibco BRL, United States), and the RNA was dissolved in 30 μL of DEPC water. The concentration and purity of the extracted RNAs were determined by a NanoDrop ND1000 spectrophotometer and an Agilent RNA 6000 nanoassay. First-strand cDNAs were generated using reverse transcription T7 oligo as the primer, and then RNase and DNA polymerase were added to synthesize second-strand cDNAs; anti-RNAs were synthesized with the T7 promoter, and aa-UTP was added to form aa-aRNA, after which NHS-CyeDye was added. The amino allyl group undergoes a chemical reaction to convert aa-aRNA into CyeDye-aRNA to complete labeling, followed by a hybridization reaction with Phalanx OneArray TM, and then enters the analysis. The XDR function of the Agilent Microarray Scan (G2505C) was used to scan twice at 100% and 10% PMT at a resolution of 10 μm. Image data were captured using GenePix™ 4 to obtain two primary data files in GPR format, and finally, the signal intensity was converted into five groups of rat AFP and CD44 transcription data.

Enzyme-linked immunosorbent assay

All the serum samples were stored at -80 °C until testing. The serum levels of CD24, CD44, Wnt3a, and GPC-3 were detected by Enzyme-linked immunosorbent assay. All the materials and prepared reagents were equilibrated to room temperature prior to use. In accordance with the manufacturer’s instructions, the samples were diluted, three kinds of wells (blank, standard, and sample) were established, and the absorbance (A) of each well was read on a Bio-Tek microplate reader (ELX808, Gene Company Limited) using 450 nm as the primary wavelength while all operation steps were completed. Finally, the A values of each well were recorded. Six standard values were used to construct the standard curve, and the real concentration of each sample was calculated by the equation of the standard curve.

Immunohistochemistry

Rat liver tissues were embedded in paraffin and sliced. The slices were dewaxed in xylene, dehydrated through graded ethanol solutions, cleaned with phosphate-buffered saline, and then immersed in 3% H₂O₂ for 15 minutes. After that, the tissue sections were incubated with a CD24-APC-conjugated antibody (Shanghai Keman Bi Co., Ltd., China), a PD-L1 antibody and a horseradish peroxidase-labeled goat anti-rabbit IgG antibody (Abbkine, United States). After the sections were stained with DAB and hematoxylin, Image-Pro Plus 6.0 software was used to determine the cumulative integrated optical density value and effective statistical area. The average optical density was calculated, and the average value was used to represent the relative level of CD24 in the liver. The slices were observed and photographed under a microscope and analyzed by Image-Pro Plus v6.0 software to determine the integral optical density.

Flow cytometry for CD24⁺ lymphocytes

Whole blood samples were collected from 20 patients with HCC and 20 patients with CH, with 20 healthy people used as controls. Blood was centrifuged at 2000 rpm for 10 minutes to prepare nucleated cells for flow cytometry (APC method) analysis. Circulating lymphocytes were isolated by using Lymphoprep™ (Fcmacs FMS-FLH100; Nanjing Fcmacs Biotechnology Co., Ltd., China) and surface-stained with antibodies against human CD24 (BioLegend, San Diego, CA, United States). CD24⁺ T cells were acquired using a fluorescence-activated cell sorting Canto II flow cytometer (V961000404. BD, United States), and the data were analyzed using FACSDiva software version.

Statistical analysis

Quantitative data are expressed as the mean ± SD, whereas qualitative data are summarized as n (%). The data were analyzed using one-way ANOVA or Student’s t test. Kaplan-Meier curves and univariate/multivariate Cox regression analyses of CD24 expression were used to explore the risk factors for HCC. Analysis was performed using SPSS version 23.0 statistical software. Image-Pro Plus 6.0, ImageJ, GraphPad Prism 5.0 and Photoshop were used to construct the figures. A P value < 0.05 was considered to indicate statistical significance.

RESULTS

Data analysis of patients with chronic liver diseases

A comparative analysis of clinical data of patients with chronic liver diseases (CH, LC and HCC) are shown in Table 1. A total of 261 patients with chronic liver diseases and 111 healthy persons were enrolled in the study, with an age range of 23 years old to 88 years old, 188 male patients and 73 female patients; HBsAg positive were 229 cases, accounting for 75.2% (97/129) in HCC patients. Among HCC patients, 91 cases with higher AFP concentration (> 25 μg/L) accounted for 70.5%. For the Child-Pugh score based on liver function tests, 48 cases were grade A in the LC group, and 12 cases were grade B plus C; 97 cases (75.2%) were grade A in HCC patients, and 32 cases were grade B plus C, accounting for 24.8%.

Upregulated CD24 related to malignancy of chronic liver diseases

A comparative analysis of serum CD24 expression among patients with chronic liver diseases is shown in Figure 1. Compared with those in the NC group, significant differences in CD24 levels were found among the different groups of patients with chronic liver disease (F = 35.571, P < 0.001). When 70 μg/L was used as the upper limit, significant differences were detected among the groups (P < 0.001) and abnormal cases constituted 20.8% of the LC group, 26.7% in the CH group, and 68.2% in the HCC group (Figure 1A), respectively. Also, the AFP concentrations were significantly different among groups (F = 225.410, P < 0.001). When the upper limit of the AFP level was 25 μg/L in the NC group, significant differences were found among the groups (P < 0.001), with the number of abnormal cases accounting for 23.6% in the LC group, 30.0% in the CH group, and 70.5% in the HCC group (Figure 1B). The area under the receiver operating characteristic curve for HCC diagnosis was 0.704 (95% confidence interval: 0.625-0.759; P < 0.01; Figure 1C). A Kaplan-Meier curve of HCC patients revealed that high CD24 expression (Figure 1D) was associated with shorter survival time. The univariate and multivariate Cox regression analysis revealed that CD24 expression could be associated with several predictors such as HBV infection, LC and a poor prognosis for HCC patients (Table 2).

Figure 1 Comparative analysis of serum clusters of differentiation 24 expression among patients with chronic liver diseases. A: Clusters of differentiation (CD) 24 expression in the serum of patients with chronic hepatitis (CH), liver cirrhosis (LC), or hepatocellular carcinoma (HCC) or healthy individuals; B: Alpha-fetoprotein expression in the serum of patients with CH, LC, or HCC or healthy individuals; C: Receiver operating characteristic curve of circulating CD24 for the diagnosis of HCC patients; D: Survival curves of circulating CD24 levels in HCC patients generated by the Kaplan-Meier method. The green line represents the high CD24 group, and the blue line represents the low CD24 group. CH: Chronic hepatitis; LC: Liver cirrhosis; HCC: Hepatocellular carcinoma; AFP: Alpha-fetoprotein; ROC: Receiver operating characteristic; CD: Clusters of differentiation; NC: Normal control.

Table 2 Uni-/multi-variable analysis with several predictors in hepatocellular carcinoma patients.

Variable	Univariate analysis			Multivariable analysis
	HR	P > |z|	95%CI	HR	P > |z|	95%CI
Age (years)
≤ 60 vs > 60	1.482	0.502	0.512-3.661
Gender
Male vs female	0.931	0.828	0.411-2.183
Grade of differentiation
Low vs high	1.316	0.515	0.571-3.545
Tumor number
Single vs multiple	1.105	0.779	0.428-2.806
Liver cirrhosis
Absent vs present	2.805	0.018a	1.161-7.531	2.401	0.013a	1.038-5.187
Hepatitis B virus
Absent vs present	0.422	0.015a	0.156-0.805	0.361	0.008a	0.152-0.796
Periportal embolus
Absent vs present	5.589	0.126	0.791-41.033
AFP (μg/L)
< 400 vs ≥ 400	0.788	0.402	0.516-1.328
TNM staging
I-II vs III-IV	0.221	0.187	0.134-1.485
CD24 expression
Low vs high	12.586	< 0.001a	3.531-52.023	5.58	< 0.001a	2.421-14.928

^aP < 0.05.

AFP: Alpha-fetoprotein; HCC: Hepatocellular carcinoma; TNM: Tumor-node-metastasis; HR: Hazard ratio; CD: Clusters of differentiation; 95%CI: 95% confidence interval.

The diagnostic values of CD24 or AFP for HCC were 68.2% or 70.5% sensitivity, 60.7% or 84.9% specificity, 63.3% or 7.8% accuracy, 45.4% or 69.0% positive predictive value, and 80.5% or 60.7% negative predictive value, respectively. Combining these two markers increased the diagnostic sensitivity to 90.7%, representing a substantial improvement, especially in AFP-negative HCC.

Clinicopathological features of CD24 expression in HCC

Clinicopathological features of serum CD24 in HCC patients are shown in Table 3. No significant difference in patient’s age, sex, or AFP concentration was found by pairwise comparison of different groups. However, circulating CD24 levels in sera of HCC patients were remarkedly associated with to HBV infection (P = 0.006), tumor size (P < 0.001), unifocal/multifocal (P < 0.001), periportal embolus (P = 0.020), lymph node metastasis (P = 0.008), differentiation degree (P < 0.001), extra-hepatic metastasis (P = 0.007), tumor-node-metastasis (TNM) grade (P = 0.008), and Child-Pugh classification (P < 0.001). Serum CD24 level in the HCC group (128.62 ± 29.10 μg/L) was significantly higher (P < 0.001) than those in the CH group (36.64 ± 27.43 μg/L), LC group (45.41 ± 23.74 μg/L) or NC group (23.72 ± 14.63 μg/L), respectively. Significant upregulated CD24 levels were a promising biomarker for chronic liver disease malignancy or HCC patients’ prognosis.

Table 3 Clinicopathological characteristics of serum clusters of differentiation 24 levels in hepatocellular carcinoma (mean ± SD).

Group	n	CD24 (μg/L)	t value	P value
Age (years)
≤ 60	66	99.7 ± 27.5	0.467	0.641
> 60	63	97.3 ± 30.8
Gender
Male	93	100.8 ± 30.5	1.443	0.502
Female	36	92.6 ± 24.4
AFP (ng/mL)
< 400	91	98.9 ± 29.1	0.763	0.447
≥ 400	38	94.6 ± 29.4
HBV infection
Absent	32	88.0 ± 18.3	2.773	0.006
Present	97	103.7 ± 30.2
Tumor size (cm)
≤ 5	73	88.5 ± 21.4	4.795	< 0.001
> 5	56	111.4 ± 32.7
Gross classification
Unifocal	85	88.8 ± 19.9	6.096	< 0.001
Multifocal	44	117.4 ± 34.9
Periportal embolus
Absent	119	96.8 ± 28.9	2.358	0.02
Present	10	119.2 ± 28.3
Lymph node metastasis
Without	59	91.4 ± 20.4	2.675	0.008
With	70	104.8 ± 33.6
Differentiation
Poor	35	118.9 ± 35.5	5.364	< 0.001
Moderate & well	94	90.9 ± 22.1
Extra-hepatic metastasis
Without	97	94.6 ± 25.6	2.743	0.007
With	32	110.5 ± 35.8
TNM stage
I-II	71	91.8 ± 27.8	2.713	0.008
III-IV	58	106.8 ± 35.0
Child-Pugh classification
A	97	83.2 ± 25.1	5.592	< 0.001
B & C	32	114.7 ± 34.3

AFP: Alpha-fetoprotein; HBV: Hepatitis B virus; CD: Clusters of differentiation; TNM: Tumor-node-metastasis.

High CD24⁺ T lymphocyte ratio in HCC patients

The distributions of the subpopulations of nucleated cells and CD24⁺ T lymphocytes in the peripheral blood of HC, CH and HCC patients are shown in Figure 2. The percentage of CD24⁺ T lymphocytes in peripheral blood was 86.2 ± 16.3% in the HCC group (n = 20; Figure 2A), 56.8 ± 11.3% in the CH group (n = 20; Figure 2B), and 21.6 ± 9.6% in the NC group (n = 20; Figure 2C), and the ratio between the groups was significant (F = 7.685; P < 0.001); the percentage of CD24⁺ T lymphocytes in the CH group was significantly greater than that in the HC group (q = 4.658; P < 0.001). In addition, the number of CD44⁺ T lymphocytes in the HCC group was significantly greater than that in the CH group (q = 2.387, P = 0.04).

Figure 2 Blood mononuclear cell population and clusters of differentiation 24⁺ T lymphocyte ratio (%). A: Mononuclear or clusters of differentiation (CD) 24⁺ T cells in the normal control group; B: Mononuclear or CD4⁺ T cells in the chronic hepatitis group; C: Mononuclear or CD24⁺ T cells in the hepatocellular carcinoma group. CD: Clusters of differentiation.

Dynamic upregulated CD24 during hepatocarcinogenesis

The relationships between hepatocyte malignancy and dynamic activation of CD24 or PD-L1 during hepatocarcinogenesis in rats are shown in Figure 3. After H&E staining was performed to determine the liver pathology of the rats, microscopic observation revealed that the hepatocytes of the control rats (n = 12; a1 in Figure 3A) were arranged neatly, the chromatin was sparse, and there was no inflammatory cell infiltration in the hepatic lobules or portal area. After the rats were fed a 2-FAA diet, the hepatocytes were disorganized, swollen, and inflammatory at the early stage (n = 17; a2 in Figure 3A), with inflammatory cell infiltration observed in the hepatic lobules and portal area at the middle stage (n = 15; a3 in Figure 3A), and substantial cell necrosis and the formation of cancer nodules at the late stage (n = 10; a4 in Figure 3A). Immunohistochemical analysis of CD24 expression in rat livers revealed weak staining in the control group (b1 in Figure 3B) and weak staining after the 2-FAA diet at the early stage (b2 in Figure 3B), significant staining at the middle stage (b3 in Figure 3B), and overexpression in cancer nest nodules at the late stage (b4 in Figure 3B). Similar PD-L1 expression was weakly detected (c1 in Figure 3C) in the control livers and weakly detected at the early stage (c2 in Figure 3C), significantly elevated at the middle stage (c3 in Figure 3C), and highly expressed in the cancer nest nodules at the late stage (c4 in Figure 3C). Similarly, both of liver CD24 (Figure 3D) or PD-L1 (Figure 3E) at protein or mRNA level were dynamically upregulated during rat hepatocarcinogenesis.

Figure 3 Verification of dynamic clusters of differentiation 24 activation in rat hepatocarcinogenesis. A: Hematoxylin and eosin staining of control livers (a1); a2: Liver pathology at an early stage; a3: Liver pathology at the middle stage; a4: Liver pathology at a late stage; B: Clusters of differentiation (CD) 24 expression in control livers (b1); b2: CD24 expression at the early stage; b3: CD24 expression at the middle stage; b4: CD24 expression at the late stage; C: Programmed death-ligand 1 (PD-L1) expression in control rat livers (c1); c2: PD-L1 expression at the early stage; c3: PD-L1 expression at the middle stage; c4: PD-L1 expression at the late stage; D: Dynamic CD24 mRNA expression during hepatocarcinogenesis; E: Dynamic PD-L1 mRNA expression during hepatocarcinogenesis. ^aP < 0.01. NC: Normal control; CD: Clusters of differentiation; H&E: Hematoxylin and eosin; PD-L1: Programmed death-ligand 1.

CD24 promoting hepatocarcinogenesis via an immune escape mechanism

The results of the quantitative analysis of the dynamic upregulation of CD24, CD44, Wnt3a, GPC-3 and AFP expression during hepatocarcinogenesis in rats are shown in Table 4. The dynamic upregulation of the expression of these genes was extremely significantly high in cases of rat hepatocyte malignancy. Compared with those in the NC group, the serum ALT and AST activities in the different experimental groups significantly increased, indicating a 4- to 8-fold increase in liver cell damage. Based on above data, dynamically progressive up-regulated CD24 was closely positive correlated (P < 0.001) with promoting HCC formation biomarkers (AFP, Wnt3a and GPC-3) or tumor metastasis related-marker (CD44).

Table 4 Serum clusters of differentiation 24 and hepatocellular carcinoma markers in hepatocarcinogenesis (mean ± SD).

Marker	NC (n = 12)	Model rats in hepatocarcinogenesis
		Early (n = 17)	Middel (n = 15)	Late (n = 10)
CD24 (μg/L)	3.68 ± 0.63	7.28 ± 3.17a	10.61 ± 4.52b	14.33 ± 7.14b
ALT (IU/L)	14.75 ± 3.02	62.19 ± 29.79b	55.55 ± 26.80b	46.08 ± 13.39b
AST (IU/L)	15.88 ± 3.29	44.46 ± 9.25c	61.99 ± 18.21c	47.91 ± 21.22c
CD44 (μg/L)	5.32 ± 1.66	12.36 ± 8.18b	21.69 ± 5.38b	28.6 ± 9.14b
AFP (μg/L)	0.81 ± 0.16	1.25 ± 0.32	1.58 ± 0.51a	1.96 ± 0.75b
Wnt3a (μg/L)	0.96 ± 0.47	2.25 ± 2.24	6.28 ± 2.12b	9.98 ± 5.62c
GPC-3 (μg/L)	0.02 ± 0.08	1.01 ± 0.27	1.62 ± 0.42a	5.83 ± 1.54b

^aP < 0.05 vs the normal control (NC) group.

^bP < 0.01 vs the NC group.

^cP < 0.001 vs the NC group.

AFP: Alpha-fetoprotein; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GPC-3: Glypican-3; CD: Cluster of differentiation.

On the basis of the above clinical investigations and model analysis, a possible novel mechanism through which CD24 activation promotes hepatocarcinogenesis is shown in Figure 4. In this study, CD24 and other HCC biomarkers were involved in HCC progression. First, stronger CD24 expression and a higher CD24⁺ T-cell ratio are associated with chronic liver disease. Second, the dynamic hepatocarcinogenesis confirmed the upregulation of CD24 expression at the protein or mRNA level. Third, immune escape is related to dynamic alterations in PD-L1 expression at the protein or mRNA level. These data provide a concise overview of the molecular pathogenesis of the HCC sequence, indicating that CD24 might be an important hepatic progenitor in the immune escape microenvironment for hepatocyte malignancy.

Figure 4 Mechanism through which clusters of differentiation 24 promotes hepatocarcinogenesis. Chronic liver diseases involve a small subpopulation of cells called liver cancer stem cells (LCSCs), which are key drivers of hepatocellular carcinoma (HCC) formation or progression. Clusters of differentiation (CD) 24, like other HCC biomarkers, is involved in HCC progression, with abnormal expression at the protein or mRNA level, and provides a concise overview of the molecular pathogenesis of the HCC sequence. CD24 might be an important progenitor in hepatocarcinogenesis in the immune escape microenvironment. HBV: Hepatitis B virus; NF-κB: Nuclear factor kappa-B; TGF-β1: Transforming growth factor β1; PD-1: Programmed cell death 1; PD-L1: Programmed cell death ligand 1; AST: Aspartate aminotransferase; ALT: Alanine aminotransferase; ALP: Alkaline phosphatase; GGT: Gamma-glutamyl transpeptidase; CD: Clusters of differentiation; HCC: Hepatocellular carcinoma; AFP: Alpha-fetoprotein.

DISCUSSION

HCC is a common malignant tumor, and monitoring hepatocyte malignancy in high-risk groups is helpful for early diagnosis and early treatment of HCC[2]. HBV infection is the main causative agent of high-incidence HCC[3,19]. HCC has no obvious symptoms in the early stage and is prone to missed diagnosis; thus, the development of simple, noninvasive diagnostic or sensitive molecular monitoring technologies is urgent[20]. The prognosis of patients with HCC is poor, and these patients are prone to recurrence, which may be related to the activation of LCSCs[21]. LCSCs have self-renewal and resistance abilities, such as LCSC activation or abnormal differentiation of hepatocytes, which are key factors leading to poor prognosis in HCC patients[22,23]. However, the mechanisms underlying the activation of LCSC early in the development of HCC remain to be elucidated. In this study, the CD24 levels and clinicopathological features of patients with HBV-related chronic liver diseases were investigated, and dynamic CD24 and PD-L1 expression was confirmed through a rat model of hepatocarcinogenesis.

LCSC is a subpopulation in undifferentiated tumor cells, and the survival of LCSC is usually limited to specific microenvironments in hypoxic hepatic tissue[5], it is closely related to tumor initiation, metastasis, recurrence, and drug resistance[6]. Although LCSC as a minority of HCC cells have the ability to self-renew and form tumors, which is a key factor leading to poor prognosis[24-26]. The univariate and multivariate Cox regression analysis revealed that CD24 expression could be associated with several predictors such as HBV infection, LC and a poor prognosis for HCC patients. In the progression of chronic liver disease, circulating CD44 as a marker of LCSC was aberrantly expressed, especially in patients with HBV infection[14,27,28], which was related to HBV replication in vivo. CD24 concentration in sera of patients with chronic liver diseases was significantly higher more than that in the healthy population. Also, the proportion of CD24⁺ T lymphocytes in the peripheral blood was remarked increased in patients with chronic liver diseases. Circulating CD24 level in HCC was significantly associated with HBV infection, tumor size, single or multiple nodules, lymph node metastasis, differentiation degree, TNM stage Child score, portal vein tumor thrombus, AFP and distal metastasis, etc. HCC data from the TCGA database show that PD-L1 was substantially correlated with various immune cells[29]. In cases with HBV-associated HCC, high PD-L1 and CD24 levels were related to poor overall survival (OS) and progression-free survival (PFS)[30]. It suggested that CD24 levels was positively correlated with AFP concentrations, which was a useful monitoring biomarker for hepatocyte malignancy or monitor the occurrence of HCC[31].

According to immunophenotype functional characteristics, nearly 20 cell surface markers are used to separate and identify LCSC, with surface proteins such as CD133, CD90, CD44, aldehyde dehydrogenase, EpCAM, OV-6, CD13, CD24, DLK1, α2δ1, CD47 and so on[32,33]. CD24 is costimulatory molecule for T cell response and as an auto-immune regulatory factor promotes HCC progression together with PD-L1[34,35]. HCC cells interact with immune cells and participate in immunosuppressive formation, escaping from immune surveillance[36,37]. Clinically, under the condition persistent chronic infection of hepatitis virus, CD8⁺ T cells in HCC patients gradually exhausted or dysfunctional, losing proliferative capacity and effector function[38]. Based on clinical data, circulating CD24 levels were upregulated in cases with HBV-related chronic liver disease, with higher ratio of CD24⁺ T cells. CD24 and PD-L1 were significantly associated with the staging of HBV-associated HCC. Those data revealed that hepatic PD-L1 level was closely positively correlated with CD24 level, and activated stemness CD24 promoted hepatocyte malignancy via mechanisms related to aggravate inflammation and immune escape[39,40].

HCC is a heterogeneous cancer with diverse inter- and intra-heterogeneity, also in terms of histology, prognosis, and molecular profiles. Reported stemness CD24 can play important roles in regulating self-renewal, tumorigenesis, metastasis, and radio-chemotherapy resistance of tumor cells[41,42]. However, its exact rules in hepato-carcinogenesis are unclear. Mechanically, a dynamic model of hepatocyte malignancy had been made for systematically observed dynamic expression of liver CD24. Both of liver and blood CD24 were dynamically progressive expression with hepatocyte malignancy, and hepatic CD24 at mRNA or protein level was upregulated expression in parallel with PD-L1 in rat hepatocarcinogenesis. Interestingly, dynamic increasing CD24 level was closely positive correlated with any of specific HCC biomarkers such as CD44[43], Wnt3a[44], GPC-3[45], and clinical routine AFP[46-48] concentration. These data provided new evidences for the occurrence and development of HCC, and suggested that stemness CD24 could be a functional liver tumor-initiating cell marker for monitoring hepatocyte malignancy[49,50].

CONCLUSION

In conclusion, this is the first report to investigate the relationship between CD24 activation and hepatocarcinogenesis. These novel findings are promising, and the initial evidence confirmed that hepatic CD24 was one of the early molecules involved in chronic liver disease malignancy. It is speculated that the stemness of CD24 is related to lymphocyte homing, CD24⁺ T cells and PD-L1 expression and may also reduce drug apoptosis or increase immune escape and other mechanisms to promote hepatocyte malignancy. The progressive increase in CD24 combined with AFP is of specific help for monitoring hepatocarcinogenesis. However, this study preliminarily reported the abnormal elevation of stem cell marker CD24 in the process of benign liver disease transforming into HCC, but it only a single-center study. The need for multi-center collaborative studies, as well as its exact biological mechanism or whether it could be used as a molecular therapeutic target for HCC in the future.

ACKNOWLEDGEMENTS

The authors would like to thank the staff of the Research Center of Medical Research, Affiliated Hospital of Nantong University, China.