The role of HLA-G in primary biliary cholangitis and response to therapy

Abstract

Introduction: Primary biliary cholangitis (PBC) is a rare autoimmune liver disease involving bile duct damage and fibrosis. This study explores the role of HLA-G, an immunomodulatory molecule crucial for immune tolerance, in PBC pathogenesis and treatment.

Methods: A cohort of 166 PBC patients from Sardinia was compared to 180 healthy controls and 205 autoimmune hepatitis type 1 (AIH-1) patients. Plasma soluble HLA-G (sHLA-G) levels, HLA-G alleles, and 3'UTR haplotypes were analyzed alongside clinical data, including therapy response to ursodeoxycholic acid.

Results: The UTR-1 haplotype was significantly more frequent in PBC patients than in controls (48.2% vs 34.3%, Pc= 0.0018). The extended haplotype HLA-G*01:01:01:08/UTR-1 was also strongly associated with PBC (23.2% vs 12.5% in controls, Pc = 0.008; 23.2% vs 6.6% in AIH-1, Pc= 2.6×10_-9). PBC patients exhibited lower sHLA-G levels compared to controls and AIH-1 (9.1 U/mL vs 24.03 U/mL and 13.9 U/mL, respectively). Among UTR-1 carriers, sHLA-G levels were particularly reduced in PBC patients. The HLA-G*01:01:01:08/UTR-1 haplotype correlated with the lowest sHLA-G levels and poorer therapy response (60% vs 24.1%, P = 0.0001).

Discussion: These findings suggest HLA-G variants, especially HLA-G*01:01:01:08/UTR-1, as potential biomarkers for PBC prognosis and treatment outcomes.

Keywords: HLA-G; Sardinia (Italy); autoimmune diseases; primary biliary cholangitis; ursodeoxicholic acid.

Figure 1

Patient enrolment workflow diagram. Overview of the study cohort selection according to EASL Clinical Practice Guidelines for the diagnosis and management of primary biliary cholangitis (PBC) (48). Patients with other chronic liver diseases were excluded. PBC, Primary biliary cholangitis; PSC, Primary sclerosing cholangitis; SLD, Steatosic liver disease; ALD, Alcoholic Fatty Liver Disease (50); HFE, Hereditary hemochromatosis; WD, Wilson Disease; HCV, Hepatitis C; HBV, Hepatitis B.

Figure 2

(A) Soluble HLA-G plasma levels (U/mL) were compared between in PBC patients (yellow) and control population (green). The P values reported was computed using the Student’s t test. The boxplot is included in the violin to assess the median and interquartile range. (B) Soluble HLA-G plasma levels (U/mL) were compared between in PBC patients (yellow) and controls who carried the HLA-G UTR-1 haplotype (green). The P values reported was computed using the Student’s t test. The boxplot is included in the violin to assess the median and interquartile range. (C) Soluble HLA-G plasma levels (U/mL) were compared between in PBC patients who carried the HLA-G*01:01:01:08/UTR-1 (light-green), G*01:01:01:01/UTR-1 (yellow) and G*01:01:01:09/UTR-1 (green) alleles combined with UTR-1. The P values reported was computed using the Student’s t test. The boxplot is included in the violin to assess the median and interquartile range.

Figure 3

(A) Soluble HLA-G plasma levels (U/mL) were compared between in PBC (yellow) and AIH patients (light-green). The P values reported was computed using the Student’s t test. The boxplot is included in the violin to assess the median and interquartile range. (B) Soluble HLA-G plasma levels (U/mL) were compared between PBC (yellow) and AIH patients who carried the HLA-G UTR-1 haplotype (light-green). The P values reported was computed using the Student’s t test. The boxplot is included in the violin to assess the median and interquartile range.

Figure 4

(A) Soluble HLA-G plasma levels (U/mL) were compared between Responder (yellow) and Non-responders patients in relation to the therapy (green). The P values reported was computed using the Student’s t test. The boxplot is included in the violin to assess the median and interquartile range. (B) Comparison of soluble HLA-G plasma levels (U/mL) among patients relative to their ALP (IU/L) levels; ALP < 1.5 x ULN (yellow) and ALP > 1.5 x ULN (green). The P values reported was computed using the Student’s t test. The boxplot is included in the violin to assess the median and interquartile range.