Different N-glycosylation patterns in mice models of portal hypertension and cirrhosis

INTRODUCTION: The golden standard to asses liver fibrosis is through a liver biopsy, but this technique has several disadvantages. It is invasive and it is associated with several complications such as pain, hospitalization, bleeding, pneumothorax and in rare cases even death. The search for non-invasive alternatives for liver biopsy has not yet resulted in a satisfactory clinical test with high sensitivity and specificity. In 2004, a new promising clinical test was developed by researchers of the Flemish interuniversity institute for biotechnology (VIB). The rationale for this test was that the majority of serum proteins are glycosylated, produced by the liver and in all major liver diseases, changes in this glycosylation occur. Using the DSA-FACE (DNA sequencer-assisted-fluorophore-assisted electrophoresis) technology N-glycan profiles of serum samples were obtained. This glycomic approach resulted in the GlycoCirrhoTest, which is the logarithmic proportion of the peak heights of an agalacto sugar (increased in cirrhosis) and a triantennary sugar (decreased in cirrhosis). AIMS & METHODS: Our study further validates this test by investigating the N-glycan profiles of serum samples obtained from 3 mice models of chronic liver disease. All models were induced in C57BL/6 male mice. A first mice model, common bile duct ligation (CBDL), mimics a human secondary billary cirrhosis. The second mice model of chronic liver disease is partial portal vein ligation (PPVL) and it is a model for a pure portal hypertension without liver damage. Finally, the third mice model is produced by subcutaneous injections with the hepatotoxin CCl4. The pathological picture that is produced in this model is a micronodular cirrhosis with characteristics that resembles a human alcoholic cirrhosis. RESULTS: The CCl4 model shows an increased abundance of multi-antennary glycans (tri- and tetra-antennary glycans). In contrast, the CBDL model is characterized by a clear augmentation of all core-fucosylated glycans. Tetra-antennary glycans also seem to be significantly increased in this model, but the triantennary glycans are mostly significantly lowered. In the PPVL model, no major changes in glycosylation pattern can be seen, but a significant decrease of the most abundant glycan structure in serum is shown. This result is shared with the other two mice models who also have a significant decrease of this most abundant peak (and its derived structures) in the electroferogram. CONCLUSION: The glycosylation patterns of the serum samples obtained from the three mice models are clearly very different. It is therefore advised to analyze the clinical test in patients with different etiologies of chronic liver disease. Although similar features between the mice models can be found, a distinction according to etiological factor is necessary