Plasma levels of sPD-1 and PD-1 genetic variants are associated with hepatitis B infection and related liver disease progression

Programmed cell death-1 (PD-1) variants and circulating levels of soluble PD-1 are associated with susceptibility to malignant and infectious disease. This study aimed to examine the association of PD-1.5 and PD-1.9 variants, and plasma sPD-1 levels with HBV infection and disease progression. The study cohort consists of HBV-infected adults (n = 513) stratied by clinical course, including chronic hepatitis B (CHB, n = 173), liver cirrhosis (LC, n = 134), hepatocellular carcinoma (HCC, n = 206), and matched healthy controls (HC, n = 196). The PD-1.5 (rs2227981 C/T) and PD-1.9 (rs2227982 C/T) genetic variants were genotyped by Sanger sequencing, and then sPD-1 levels were quantied by enzyme immunoassay. levels are associated with liver inammation and progression of liver brosis and the PD-1.5 and PD-1.9 variants are associated with HBV infection and progression of liver disease. assays. HCC patients. Although our current study did not show a signicant association between this polymorphism and susceptibility to HBV infection, we reported for the rst time that the PD-1.5 CT genotype may be a protective factor for HCC predisposition. PD-1 has a soluble form (sPD-1) that can be detected in peripheral blood 14,43 . In our study, the circulating sPD-1 levels of HBV patients were signicantly higher than those of the control group. The results were consistent with studies in HBV infections 17,44,45 , chronic HCV 46 and in acute or chronic inammatory conditions such as pancreatitis 47 , sepsis 48 , autoimmune hepatitis and inammatory bowel disease 49 . These data suggest that increased sPD-Page


Introduction
Hepatitis B virus (HBV) infection is the most common chronic viral infection worldwide. The World Health Organization estimates that 296 million people are chronically infected with HBV in 2019 and 820 000 deaths mostly from cirrhosis and hepatocellular carcinoma 1 . HBV infection can cause a wide spectrum of liver diseases, including acute hepatitis B virus (AHB), chronic hepatitis B (CHB), liver cirrhosis (LC), and hepatocellular carcinoma (HCC). Notably, HBV is accountable for ∼53% of global HCC incidence 2 .
PD-1, encoded by the PDCD1 gene, is a type I transmembrane immunoinhibitory receptor for PD-L1/2 ligands. PD-1 is widely expressed on the surface of T cells, B cells, natural killer cells, natural T killer cells, activated monocytes and cancer cells 13 . Recent ndings have shown that, in addition to the membranebound form, the PD-1 protein also has a soluble form (sPD-1) that regulates immunological responses 14 . Studies on the biological functions of sPD-1 in animal models showed that sPD-1 inhibited the growth of H22 hepatoma cells, enhanced the lysis of tumor cells and ultimately prolonged the overall survival of mice with tumors 15,16 . Previous studies also found that sPD-1 levels are associated with the prognosis of cancer, including HCC 17 . In addition, PD-1 overexpression is associated with T-cell dysfunction and exhaustion in chronic HBV infection and HCC development [18][19][20] . Another study investigated the effects of sPD-1 levels on the long-term dynamics of HBV load and HCC risk showed that elevated sPD-1 levels correlated with a high viral load for greater than four and high sPD-1 levels were associated with an increased risk of HCC 20 .
Genetic variations of the PDCD1 gene in the promoter and transcription binding sites that can alter gene function or expression have been associated with cancers, autoimmune diseases and infectious diseases [21][22][23][24] . PD-1.5 (rs2227981) and PD-1.9 (rs2227982) were selected as candidates because they are located close to exon 5 of PDCD1, which encodes a cytoplasmic domain consisting of two tyrosine motifs associated with inhibitory activities. In addition, PD-1.9 is a non-synonymous substitution (valine to alanine) that modulates the structural and functional properties of PD-1 (www.snpedia.com/index.php/rs2227982). Both these PD-1.5 and PD-1.9 polymorphisms were associated with susceptibility to cancers and other autoimmune diseases [25][26][27] .
There is very little data on the contribution of PD-1.5 and PD-1.9 polymorphisms with HBV infection 28,29 , and there is only a case-control study that investigated the association between these two PD-1 polymorphisms and the progression of HBV-related liver disease, but not in particular on the circulating sPD-1 levels 30 . In this line of research, we conducted a case-control study aiming to examine the association of PD-1.5 and PD-1.9 variants, but also plasma sPD-1 levels with HBV infection and disease progression.

Materials And Methods
Methods used in this study were in accordance with the relevant guidelines and regulations and were approved by the institutional review board and an independent Ethics Committee of the Institute of Clinical Medical and Pharmaceutical Sciences, Hanoi, Vietnam.

Ethics statement
This study was conducted in accordance with the Declaration of Helsinki. The study was approved by the institutional review board of the 108 Institute of Clinical Medical and Pharmaceutical Sciences, Hanoi, Vietnam. Informed written consent was obtained from all participants after explanation of the study at the time of sampling.

Study participants
This case-control study recruited 513 adult HBV-infected patients, who were referred for clinical management at the 108 None of HBV-infected patients had evidence of chronic comorbidities such as: autoimmune diseases, alcoholic liver disease, type 2 diabetes, addiction to smoking and alcohol or treatment with immune inhibitors. In addition, 196 individuals visited 108 hospital for routine medical check-up showed the normality of hematological and biochemical testing were collected as heathy control (HC) group. All patients and HCs were con rmed negative for anti-HCV and anti-HIV by ELISA assays. We collected 5 ml of peripheral blood from each participant. Plasma was immediately separated then frozen at -80°C until use.

PD1 genotyping
Genomic DNA was isolated from 200 µl of peripheral blood using a DNA isolation kit (Qiagen, Hilden, Germany), following manufacturer's instructions. The procedure of PD-1 genotyping was followed the study protocol as described 30 . Brie y, the amplicon containing the variants PD-1.5 and PD-1.9 was ampli ed by PCR using the speci c primer pairs PD-1.5/9_F: 5′-GCA AGA ATG CCA GGG ACA TTT CAG AG-3′ and PD-1.5/9_R: 5′-TGC CTG GTG CAG GTG CAG-3′. PCR products were puri ed using the Exo-SAP-IT PCR product cleanup reagent (Afymetrix Santa Clara, USA) 5 µl of puri ed PCR products were used as sequencing templates. Direct sanger sequencing was performed using the BigDye terminator v.1.1 cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) on an ABI 3130XL DNA sequencer according to the manufacturer's instructions.

Statistical analysis
All statistical analyses were performed using R version 3.1.2 (http://www.r-project.org). Genotype and allele frequencies were determined by simple gene counting. Deviations from Hardy-Weinberg equilibrium were calculated for each group. Chi-square, Kruskal-Wallis and Mann-Whitney-Wilcoxon tests were used to compare differences between groups for qualitative or quantitative variables where appropriate. For genetic analyses, we used multivariate logistic regression models adjusted for age and gender to test for associations between variants and HBV-related liver diseases, applying different genetic models (codominant, dominant, recessive models) in which, adjusted odds ratios (OR) with 95% con dence intervals (CI) were calculated. Correlations between sPD-1 and other laboratory parameters were assessed using the Spearman's rank correlation test. The level of signi cance was set at a two-sided p-value of < 0.05.

Results
Baseline characteristics Table 1 describes the baseline pro les of patient subgroups and 196 healthy controls. Most HBV patients were male (85%) compared to 49% in healthy group. Among the patient subgroups, the AST, ALT, total bilirubin and direct bilirubin levels were higher in CHB patients than in LC and HCC groups (P < 0.0001). Albumin, prothrombin and platelet counts were signi cant lower in patents with LC compared to CHB and HCC group (P < 0.0001). AFP levels were signi cantly higher among the HCC compared to LC and CHB group (P < 0.0001).

Association of PD-1 variants with HBV infection and liver disease progression
The distribution of genotype and allele frequencies of PD-1.5 and PD-1.9 variants and their association with hepatitis B and related liver diseases are summarized in the Table 2 and Table 3. The variants were in Hardy-Weinberg equilibrium for both cases and controls (P > 0.05). No signi cant association was found with susceptibility to HBV infection and the PD-1.5 variant, but only in chronically infected HBV-infected patients, where the frequency of the PD-1.5 CT genotype was signi cantly lower in HCC patients compared to non-HCC patients (CHB + LC). Among subgroups of HBV patients, the PD-1.5 CT genotype was found to be a protective factor for HCC (HCC vs. non-HCC: OR = 0.6, 95%CI: 0.4-0.9, Padj = 0.031). For other PD-1.5 genotypes, no signi cant associations with HBV-related liver progression were found in various between-group comparisons ( Table 3).  For PD1.9 polymorphism, the frequency of the genotype TT was signi cantly lower in HBV-infected patients compared to healthy controls (HBV patients vs HC: OR = 0.53, 95%CI:0.31-0.9, P adj =0.026). In addition, the TT genotype was more frequent in the LC, HCC and HCC + LC patient groups than in the CHB patient group (LC vs. CHB: OR = 4.3, 95%CI: 1.7-11.3, P adi =0.0012; HCC vs. CHB: OR = 4.1, 95%CI: 1.6-10.3, P adi =0.0012; HCC + LC vs. CHB: 4.2, 95%CI = 1.8-9.9, P adi =0.00022, respectively). The result also showed that PD-1.9 TT genotype was a risk factor for HCC (HCC vs. non-HCC: OR = 2.0, 95%CI: 1.1-3.7, P adj =0.017).
No signi cant association was observed between PD-1.5 or PD-1.9 variants and any clinical laboratory parameters (data not presented).
Association of sPD-1 levels with PD-1 variants We analyzed whether PD-1.5 and PD-1.9 variants could in uence the expression of sPD-1 in HBV infected patients. For SNP PD-1.5, patients carrying the CC and CT genotype had the higher sPD-1 levels than TT genotype (P = 0093 and 0036, respectively). Meanwhile, there were no signi cant differences in sPD-1 levels between PD-1.9 genotypes (Fig. 2).

Correlation between sPD-1 and liver enzyme parameters
The correlation between sPD-1 levels and biochemical parameters in 191 HBV infected patients is shown in Fig. 3. A positive correlation between sPD-1 levels and AST, ALT enzyme activity, total bilirubin, and direct bilirubin levels was observed (rho = 0.57; 0.57; 0.27; 0.33; respectively) ( Fig. 3). We did not nd the signi cant correlation between sPD-L1 and other parameters including albumin, platelet counts, prothrombin and AFP levels, and HBV viral loads (data not presented).

Association of plasma sPD-1 levels with brosis score
Analysis of the association of plasma sPD-1 with HCC stages based on BCLC classi cation (BCLC-A/B/C/D) and liver brosis progression (FIB-4 and APRI scores) found no signi cant differences between plasma sPD-1 concentrations and HCC stages (data not shown).
The patients with an APRI score > 1 had signi cantly higher plasma sPD-1 levels compared to those with APRI score ≤ 1 (P < 0.0001). Analysis of correlation between sPD-1 levels and APRI score also revealed a positive correlation (rho = 0.53, P < 0.0001). Nevertheless, the sPD-1 levels were not correlated with Fib-4 score (Fig. 4).

Discussion
Increased PD-1 expression on T lymphocytes, including CD4 and CD8 T lymphocytes, is believed to lead to T cell exhaustion and is an important mechanism for tumor cell immune defense. In this case-control study, we investigated PD-1.5 and PD-1.9 variants and soluble PD-1 levels in HBV-infected patients and controls. We observed that there is a potential clinical signi cance of PD-1.5 / PD-1.9 polymorphisms and plasma sPD-1 levels in HBV infection.
Several genetic association studies have investigated the association of polymorphisms of the PDCD1 gene with infectious diseases. In particular, ve single nucleotide polymorphisms of PD1: PD-1.1 (-538G/A), PD-1.3 (+ 7146G/A), PD-1.5 (+ 7785T/C), PD-1.6 (+ 8669G/A) and PD-1.9 (+ 7625C/T) have been associated with human malignancies and are highly expressed in several cancers 22,33−36 . Furthermore, few studies have investigated the association of these PD-1 variants and susceptibility to HBV infection and liver disease progression 28,37−40 . In our current study, the PD-1.9 variant but not PD-1.5, is associated with risk of HBV infection and clinical outcome.
PD-1.9 is a non-synonymous variant that results in the amino acid substitution from valine to alanine at codon 215, which may lead to a change in structure or function of PD-1. Our study has shown that PD-1.9 TT genotype may be a risk factor for HBV infection as well as for disease progression, which is consistent with our previous ndings 30 .Two of our earlier studies which utilized different HBV patient cohorts con rm the study conducted by Fang Li et al.
showing that PD-1.9 may be involved in hepatocarcinogenesis in HBV infection 29 . However, no signi cant association between PD-1.9 variant and overall cancer susceptibility has been found in meta-analyses 22,25,41 .
The PD-1.5 SNP is located in the exon 5 region, the transition from C to T does not alter the amino acid sequence of PD-1. The signi cant associations between PD-1.5 and cancer are likely due to a linkage disequilibrium of the PD-1.5 variation with other PD-1 gene polymorphisms, which can lead to altered PD-1 expression levels 21 . Several meta-analyses have shown that the PD-1.5 TT genotype, as well as the T allele reduce the risk of cancers 22,41,42 . However, all studies did not include HCC patients. Although our current study did not show a signi cant association between this polymorphism and susceptibility to HBV infection, we reported for the rst time that the PD-1.5 CT genotype may be a protective factor for HCC predisposition.
PD-1 has a soluble form (sPD-1) that can be detected in peripheral blood 14,43 . In our study, the circulating sPD-1 levels of HBV patients were signi cantly higher than those of the control group. The results were consistent with studies in HBV infections 17,44,45 , chronic HCV 46 and in acute or chronic in ammatory conditions such as pancreatitis 47 , sepsis 48 , autoimmune hepatitis and in ammatory bowel disease 49 . These data suggest that increased sPD-1 levels are associated with in ammation and higher in ammation tends to lead to higher plasma sPD-1 levels. Additional data from our current study was the positive correlation between plasma sPD-1 levels and indicators of liver damage such as AST and ALT. Furthermore, previous studies have shown that sPD-1 levels are elevated in HBV patients in a manner that corresponds with in ammatory factors such as AST, ALT, IL-10, IL-17, TNF-α and IFN-γ 17,44,50 . The above evidence could explain the signi cantly higher levels of plasma sPD-1 in CHB patients in this study cohort.
FIB-4 and APRI indices are considered good predictors of liver brosis in chronic hepatitis C 51 and CHB 52 . In the current study, we found for the rst time a positive correlation between plasma sPD-1 levels and APRI scores, suggesting that sPD-1 could be a complementary marker for the diagnosis of brosis in patients with chronic hepatitis B infection. In addition, this marker has many advantages, such as being less invasive and monitoring the kinetics of brosis progression. This shall be a useful indicator in the prognosis of HBV related diseases and in the treatment of patients. Therefore, sPD-1 could be a useful marker for monitoring the progression of HBV-related liver disease. However, studies in a larger number of patients are needed to con rm and extend the results of this study.
The current study has limitations as the results would have been more meaningful if we could have determined the expression of PD-1 membrane-bound proteins in PBMC (peripheral blood mononuclear cells) as well as protein expression in liver tissue by immunohistochemical assays. However, in this study, neither liver tissue samples nor blood samples were available for isolation of PBMC.
In summary, this study showed that plasma sPD-1 levels correlate with liver in ammation and are associated with liver brosis in chronically HBV-infected patients. The PD-1.5 and PD1.9 polymorphisms are associated with HBV infection and progression of HBV-related liver disease.
Declarations Figure 2 Association of sPD-L1 with PD-1 polymorphisms in HBV patients. (A) PD-1.5 (B) PD-1.9 polymorphism Box-plots illustrate median values with range (minmax) and outliers; NS: not signi cant, p-values were calculated by Wilcoxon tests (A) or Kruskal Wallis test (B).

Figure 3
Correlation between sPD-1 levels with laboratory parameters in HBV patients. Correlation between sPD-1 levels with AST (A); ALT (B); total bilirubin (C); direct bilirubin (D). The correlation coe cient between sPD-1 levels with laboratory parameters: was calculated by using Spearman's rank correlation coe cient.
Spearman's rho (rho) and P value are given.

Figure 4
Association and correlation of sPD-1 and brosis scores in HBV patients Box-plots illustrate median values with range (min-max) and outliers with P-values were calculated by Wilcoxon tests and Kruskal Wallis test. The correlation coe cients between sPD-1 levels with APRI (A) and FIB-4 (C) scores were calculated by using Spearman's rank correlation coe cient. Spearman's rho (rho) and P value are given.