Abstract Background Liver sinusoidal endothelial cells (LSECs) have transcellular pores, called fenestrations, participating in the bidirectional transport between the vascular system and liver parenchyma. Fenestrated LSECs indicate a healthy phenotype of liver while loss of fenestrations (defenestration) in LSECs is associated with liver pathologies. Methods We introduce a unique model of systemic inflammation triggered by the deletion of Mcpip1 in myeloid leukocytes (Mcpip1fl/flLysMCre) characterised by progressive alterations in LSEC phenotype. We implement multiparametric characterisation of LSECs by using novel real-time atomic force microscopy supported with scanning electron microscopy and quantitative fluorescence microscopy. In addition, we provide genetic profiling, searching for characteristic genes encoding proteins that might be connected with the structure of fenestrations. Results We demonstrate that LSECs in Mcpip1fl/flLysMCre display two phases of defenestration: the early phase, with modest defenestration that was fully reversible using cytochalasin B and the late phase, with severe defenestration that is mostly irreversible. By thorough analysis of LSEC porosity, elastic modulus and actin abundance in Mcpip1fl/flLysMCre and in response to cytochalasin B, we demonstrate that proteins other than actin must be additionally responsible for inducing open fenestrations. We highlight several genes that were severely affected in the late but not in the early phase of LSEC defenestration shedding a light on complex structure of individual fenestrations. Conclusions The presented model of LSEC derived from Mcpip1fl/flLysMCre provides a valuable reference for developing novel strategies for LSEC refenestration in the early and late phases of liver pathology.