Reply:

Abstract

We thank Dr. Ramadori and Dr. Mansuroglu for their comments on our recent publication1 reporting the classification of fetal liver mesenchymal cells into three populations based on the location and immunostaining: hepatic stellate cells (HSCs) in the liver parenchyma, perivascular mesenchymal cells around the veins, and submesothelial cells beneath the basal lamina near the liver surface. As they pointed out, we found that there are rare desmin-positive alpha smooth muscle actin (αSMA)–positive mesenchymal cells near the veins in the developing mouse liver. We assume that this minor population represents heterogeneity of fetal HSCs as previously reported in the human liver2 or a transitional cell type between HSCs and perivascular mesenchymal cells. It is also possible that this minor population may become the cells within the arteriole wall in the adult liver. A cell lineage analysis of this minor perivascular cell population is important, but this requires identification of new and definitive molecular markers. We identified activated leukocyte cell adhesion molecule (ALCAM) as a marker for submesothelial cells and mesothelial cells in the developing mouse liver. Unfortunately, fetal HSCs and perivascular mesenchymal cells do not express ALCAM and therefore cannot be separated with fluorescence activated cell sorting. We are looking for new distinct cell surface markers to achieve this goal. Using anti-ALCAM antibodies, we separated an ALCAMhigh population containing submesothelial cells and mesothelial cells.1 The cultured ALCAMhigh population rapidly acquired a myofibroblastic phenotype on a plastic dish; the expression of αSMA messenger RNA (mRNA) showed a nearly 400-fold increase already at day 1, and this was maintained on day 8. When the cells were cultured in collagen gel, αSMA mRNA was increased 25- and 157-fold on days 1 and 8, respectively; this suggests that the myofibroblastic conversion is delayed in collagen gel (unpublished data, 2009). We are currently examining gene expression of these cells upon retinol and palmitic acid treatment. Although we did not test ALCAM immunostaining on the cultured ALCAMhigh cells in collagen gel, quantitative reverse-transcription polymerase chain reaction revealed reduced mRNA expression in a collagen gel culture (unpublished data, 2009). Kinji Asahina Ph.D.*, Hidekazu Tsukamoto D.V.M., Ph.D.*, * Southern California Research Center for Alcoholic Liver and Pancreatic Disease and Cirrhosis, Department of Pathology Keck School of Medicine, University of Southern California, Los Angeles, CA.