Introduction

Biliary atresia (BA) is the most common cause of cholestasis in infants under three months. It is characterized by microinflammation and fibrosis of intra- and extrahepatic bile ducts [1]. If therapy is not given, children suffering from BA will have progressive liver fibrosis and cirrhosis and usually will not survive more than two years of age [2].

Liver fibrosis in BA might manifest rapidly after Kasai surgery. Understanding liver fibrosis in BA might benefit patient prognosis, development of biomarkers for diagnosis, and targeted therapy of liver fibrosis in BA patients [3]. A recent study showed that the expression of the collagen gene cluster was strongly associated with liver fibrosis, including BA patients [4,5,6,7]. However, most studies were conducted in developed countries [5,6,7]. Therefore, our study investigated the role of COL6A1, COL6A2, COL6A3, and COL1A1 expressions on liver fibrosis in BA patients in Indonesia.

Material and methods

Patients

Twenty liver tissues of BA patients were acquired during Kasai surgery. The type of BA was determined using the Kasai classification system [8]. Cirrhosis was determined as bridging fibrosis with > 50% of portal tracts encompassed and nodular architecture [9]. At the same time, 18 control liver specimens were obtained from patients who underwent surgeries or biopsies for other diseases, including intrahepatic cholestasis (n = 7), choledochal cyst (5), internal bleeding (1), intraabdominal tumor (n = 2), gastric volvulus (n = 1), liver abscess (1), and Alagille syndrome (n = 1).

Total RNA isolation

Total RNA was isolated from 25 to 30 mg of liver tissue using the Quick-RNA Miniprep Kit (Zymo Research, Irvine, California, US). The RNA concentration was quantified by a NanoDrop 2000 Spectrophotometer (Thermo Scientific, Wilmington, DE, USA) with the OD260/280 ratios ranging from 1.8 to 2.0 to ensure RNA purity. The samples were immediately stored at − 80 °C for future use.

Quantitative RT-PCR

One-step qPCR was performed using SensiFAST SYBR No-ROX Kit (Bioline, Tennessee, USA) and BioRad CFX Real-Time PCR System (California, USA) for collagen gene cluster (COL6A1, COL6A2, COL6A3, and COL1A1). A housekeeping gene, GAPDH, was used as a reference gene. The reaction mix contained 10 μL sensiFAST one-step mix, 0.8 μL of each primer, 0.2 μL Reverse Transcriptase, 0.4 μL RNase Inhibitor, RNAse free water was added to reach 16 μL. Lastly, the mRNA was added, resulting in the final volume of 20 μL. PCR cycling conditions were: 45 °C for 10 min followed by 1 cycle of 95 °C for 2 min and 40 cycles of 95 °C for 5 s (denaturation), 60 °C for 1 min (Annealing/Extension). The primers that were used for collagen gene cluster were as follows: COL6A1 5′—TAAAGGCTACCGAGGCGATG—3′ (forward) and 5′—GCCGTCTTCTCCCCTTTCAC—3′ (reverse); COL6A2: 5′—CTCCTCGGGACCAGGACTTC—3′ (forward) and 5′—CGGTCTTCTCTGGGCAGTTG—3′ (reverse); COL6A3: 5′—TTAGCCAGCACTCGCTATCC—3′ (forward) and 5′—TTACTGGGGCCGATGTTGAG—3′ (reverse); COL1A1: 5′—CAATGCTGCCCTTTCTGCTCCTTT—3′ (forward) and 5′—ATTGCCTTTGATTGCTGGGCAGAC—3′ (reverse) [4]; and GAPDH: 5′—GCACCGTCAAGGCTGAGAAC—3′ (forward) and 5′—TGGTGAAGACGCCAGTGGA—3′ (reverse) [10].

Statistical analysis

The expression of COL6A1, COL6A2, COL6A3, and COL1A1 were determined using the Livak method (2−ΔΔCT). The expression of COL6A1, COL6A2, COL6A3, and COL1A1 was presented as mean ± standard deviation (SD). The normality of the continuous variables was defined by the Kolmogorov–Smirnov test. The differences in the expressions of collagen gene clusters between BA patients and controls were analyzed using an independent t-test with a significance value of p < 0.05. All statistical analyses were performed using the IBM Statistical Package for the Social Sciences (SPSS) version 21 (Chicago, USA).

Results

Baseline characteristics

We ascertained 20 BA patients and 18 controls. Most BA patients were female (60%) and type 3 (60%). The median age of patients who underwent the Kasai procedure was 124 (IQR, 95.5–174.5) days (Table 1).

Table 1 Characteristics of BA patients after the Kasai procedure in our institution
Full size table

Expression of the collagen gene cluster in BA patients

There was a significant down-regulated expression of COL6A1 (p = 0.0009), COL6A2 (p = 0.02), COL6A3 (p = 0.024), and COL1A1 (p = 0.0015) in liver BA patients compared to controls with the fold change of 12.53-, 5.58-, 21.35-, and 8.41-times, respectively (Additional file 1: Fig. S1; Table 2).

Table 2 Expression of the collagen gene cluster in BA patients and control liver
Full size table

Association between collagen gene cluster expression and outcomes

Interestingly, the collagen gene cluster expressions were significantly associated with the presence of cirrhosis (p = 0.0085, 0.04, and 0.0283 for COL6A1, COL6A2, and COL6A3, respectively) but not with the survival of BA patients and age at Kasai procedure (Table 3).

Table 3 Association of collagen gene cluster expression and outcomes of BA patients after Kasai procedure
Full size table

Discussion

Our study shows the aberrant expressions of collagen type VI genes (COL6A1, COL6A2, and COL6A3) in BA patients compared with controls. A previous study showed that the collagen gene cluster (COL6A1, COL6A2, COL6A3, and COL1A1) expressions were associated with the hepatic stellate cells (HSCs) activation [4]. Collagen VI (COL6A1, COL6A2, and COL6A3) genes, but not COL1A1, have been associated with liver fibrosis [11,12,13]. The accumulation of type VI collagen might cause the destruction of the liver's structure and function in liver fibrosis [11]. It is compatible with our findings that COL6A1, COL6A2, and COL6A3 expressions, not COL1A1, were strongly associated with the presence of cirrhosis (Table 3). Moreover, collagen type VI has been shown as a potent activating factor for HSCs and an inducing factor for fibrosis-associated gene expression changes, including αSMA and TGF-β [12].

In addition, CO6-MMP, a collagen type VI fragment, was elevated in bile-duct ligation (BDL) and the carbon tetrachloride (CCl4) induced fibrosis rat models. It suggests its role in liver fibrogenesis [4, 11]. However, previous studies investigated the association of collagen type VI genes with liver fibrosis in other diseases, including cholestatic liver diseases and acute liver injury, not BA, as the novelty of our study [4, 11,12,13]. In addition, previous studies showed the aberrant other collagen gene expressions (COL10A1, COL15A1, COL22A1, COL9A2 [5] and COL3A1, COL1A2 [7]), not COL6A1, COL6A2, COL6A3, and COL1A1 genes as our study, in patients with BA. These differences are another novelty of our findings.

Our study also shows the aberrant expressions of COL1A1 in BA patients compared to the control. A recent study demonstrated the involvement of COL1A1 in liver fibrosis through miRNA-29 signaling [14]. Moreover, another study showed a constantly high expression of 12 genes, including the collagen gene cluster, in BA patients one year after a successful Kasai surgery compared to at the time of the Kasai procedure [5]. This suggests that the fibrosis process evolved with time instead of as a one-time process. The genes involved may not be a consequence but a precursor of the fibrotic changes [5]. Moreover, Baiocchini et al. [15] also identified COL1A1 as one of the molecules associated with liver fibrosis in HCV-infected patients.

Interestingly, our findings show downregulated collagen gene cluster expressions in cirrhosis patients compared with non-cirrhosis patients (Table 3). However, a previous study revealed a significantly higher expression of the collagen gene cluster in the fibrosis group than in the inflammation group [16]. Our previous study showed that the expression of COL1A1 and COL6A1 were significantly downregulated in the hypospadias patients with more severe chordee than the milder one (Indonesian) [17]. In contrast, there is no difference in collagen intensity between hypospadias patients and controls (Caucasian) [18]. Moreover, some collagen genes have different molecular evolutionary characteristics among different populations [19].

Conclusions

Our study shows the changes in the collagen gene cluster, particularly collagen type I and VI genes, expressions in patients with BA in a particular developing country. Our findings suggest the role of this collagen gene cluster in the liver fibrogenesis of BA.

Limitations

Our study has several limitations, including a small sample size and one pediatric surgical center that might not reflect the Indonesian population. Moreover, we do not validate the protein levels of the collagen gene cluster in BA, i.e., lack of pathological images and collagen deposition staining of the liver tissues in patients with BA due to limited resources. Notably, the control liver specimens were patients with chronic liver and gallbladder diseases, not subjects with normal liver tissue. These facts should be considered during the interpretations of our findings. In this study, we focus on the effect of the collagen gene cluster expressions in liver fibrogenesis of BA, while the association between the prognostic factors, including histopathological findings, with the survival rate of BA patients after Kasai procedures have been reported in our previous studies [9, 20].