Plasmolipin regulates basolateral-to-apical transcytosis of ICAM-1 and leukocyte adhesion in polarized hepatic epithelial cells

Apical localization of Intercellular Adhesion Receptor (ICAM)-1 regulates the adhesion and guidance of leukocytes across polarized epithelial barriers. Here, we investigate the molecular mechanisms that determine ICAM-1 localization into apical membrane domains of polarized hepatic epithelial cells, and their effect on lymphocyte-hepatic epithelial cell interaction. We had previously shown that segregation of ICAM-1 into apical membrane domains, which form bile canaliculi and bile ducts in hepatic epithelial cells, requires basolateral-to-apical transcytosis. Searching for protein machinery potentially involved in ICAM-1 polarization we found that the SNARE-associated protein plasmolipin (PLLP) is expressed in the subapical compartment of hepatic epithelial cells in vitro and in vivo. BioID analysis of ICAM-1 revealed proximal interaction between this adhesion receptor and PLLP. ICAM-1 colocalized and interacted with PLLP during the transcytosis of the receptor. PLLP gene editing and silencing increased the basolateral localization and reduced the apical confinement of ICAM-1 without affecting apicobasal polarity of hepatic epithelial cells, indicating that ICAM-1 transcytosis is specifically impaired in the absence of PLLP. Importantly, PLLP depletion was sufficient to increase T-cell adhesion to hepatic epithelial cells. Such an increase depended on the epithelial cell polarity and ICAM-1 expression, showing that the epithelial transcytotic machinery regulates the adhesion of lymphocytes to polarized epithelial cells. Our findings strongly suggest that the polarized intracellular transport of adhesion receptors constitutes a new regulatory layer of the epithelial inflammatory response. Supplementary Information The online version contains supplementary material available at 10.1007/s00018-021-04095-z.

The PLLP gene was edited using CRISPR/CAS9, and four PLLP_KO clones were selected and compared to parental HepG2 cells (WT). Clones were analyzed by western blot (a) and confocal immunofluorescence analysis (b) with a specific polyclonal antibody generated to the last 17 residues of human PLLP. (a) Left panels. Western blot of WT cells and PLLP_KO cell clones. Right panels. The clones were pooled and transiently transfected with an expression vector coding for human PLLP. The antibody specifically recognized a 20 KD band in transfected KO cells corresponding to exogenous PLLP expression. (b) Note that the pericanalicular staining with the anti-PLLP antibody disappears in PLLP_KO cells, although some background signal remains. ICAM-1 staining decreased in BC domains (top images) and increased its non-canalicular, basolateral distribution in the KO clones (bottom images) with respect to WT cells. Images were acquired in the confocal microscope with two levels of gain to show changes of ICAM-1 staining in canalicular (low gain) and noncanalicular (high gain) membrane domains (quantifications shown in Figure 4c-d) Scale bars, 10 m. Z-projections of at least 10 confocal planes of 0.6 m thickness are shown. Figure S4. Double immunolocalization of BL-ICAM-1 and PLLP-GFP by transmission electron microscopy. ICAM-1 was basolaterally-labeled as in Figure 1d. Cells were then incubated at 37 o C for 0 min (a) or 90 min (b), and fixed following procedures compatible with immunolocalization by transmission electron microscopy. BL-ICAM-1 was detected with a rabbit anti-mouse antibody followed by protein A conjugated to 15-nm gold particle (purple arrows). Rabbit anti-GFP antibody was followed by incubation with a protein A conjugated to 10-nm gold particles (green arrows). Scale bars 2 m.   Figure 4. (a) Confocal analysis of cellular height, spreading area and perimeter in WT and PLLP_KO HepG2 cells. 75 and 88 Z-stacks, respectively, composed by at least 16 confocal planes, were quantified for each cell type. The triple staining of F-actin, ICAM-1, and ZO-1 was used for quantification. (b) Schematic representation of the sequential incubation with secondary antibodies that bind BL ICAM-1 performed to discriminate between surface and internalized BL-ICAM-1 at 90 min of transcitosis. The incubation with a first secondary antibody at 4 o C before fixation specifically labels surface BL-ICAM-1 that has not been internalized. After fixation and permeabilization, incubation with a second secondary antibody conjugated with a different fluorophore labels the internalized BL-ICAM-1 and part of surface BL-ICAM-1. (c) Comparison between the two secondary-antibody distributions enables discrimination between internalized and surface ICAM-1 populations. Superresolution confocal microscopy (STED) of BL-ICAM-1 after 90 min of transcytosis. The bracket points at regions of transition between surface and internalized ICAM-1. Note that the BC is not labeled by the first secondary antibody added prior fixation and permeabilization. (d) Transcytosis and sequential incubation of secondary antibodies to discriminate surface from internalized BL-ICAM-1 in WT and PLLP_KO cells expressing GFP-Rab11. After image acquisition, surface fluorescence intensity was subtracted from the channel corresponding to the staining performed after permeabilization, so that population of internalized BL-ICAM-1 in the SAC (identified by colocalization with GFP-Rab11) and in other intracellular vesicular domains could be quantified. Note that the representative image shows cells with a slightly greater than average surface distribution of BL-ICAM-1, in order to better illustrate the differences between surface and total labelling of BL-ICAM-1. Scale bars, 10 m. Analysis of macrophage infiltration. Note that infiltrated macrophages were detected in only one tumor generated by WT cells, so statistical analysis could not be performed. Some tumors rapidly collapsed and disaggregated after isolation and could not be processed for IHQ and confocal analysis (N/A).

SUPPLEMENTAL VIDEO LEGENDS
Video S1. Time-lapse confocal microscopy of fluorescent BL-ICAM-1 translocation to the BC in polarized cells expressing PLLP-GFP. Images were acquired at 10 min intervals for 100 min and displayed at 2 frames per second.
Video S2. Time-lapse confocal microscopy of fluorescent BL-ICAM-1 translocation to the BC in which emissions of PLLP-positive tubular structures are detected. Images were acquired at 10 min intervals for 100 min and displayed at 2 frames per second.
Video S3. Time-lapse confocal microscopy of fluorescent BL-ICAM-1 translocation to the BC in polarized WT HepG2 cells. Images were acquired at 10 min intervals for 90 min and displayed at 2 frames per second.
Video S4. Time-lapse confocal microscopy of fluorescent BL-ICAM-1 translocation to the BC in polarized PLLP_KO HepG2 cells. Images were acquired at 10 min intervals for 90 min and displayed at 2 frames per second.