Nuclei were stained with 50 ng/mL Hoechst 33258. Cell polarization was determined by counting the number

of bile canaliculi (BC) (identified by dense F-actin staining) per 100 cells (identified by fluorescently labeled nuclei) as described.18 Cells were plated on coverslips in the absence or the presence of soluble rGal-1 (7 μM). After 48 hours, each coverslip was mounted in a chamber placed on the stage of a Nikon TE-200 epifluorescence-inverted microscope. Cells were then loaded with 5-chloromethylfluorescein diacetate.19 Hepatocytes capture and metabolize this compound, generating fluorescent glutathione methylfluorescein, which is actively secreted to the canaliculi by MRP2. A total of 5 × 106 HepG2-M (mock-transfected) or HepG2-G2 (Gal-1–overexpressing) cells were injected subcutaneously into the left PLX3397 cell line flank of 6-week-old BALB/c nude mice. Tumor volume was calculated as π/6 × length × width2. Liver, lungs, and tumor-draining lymph nodes were serially sectioned and stained with hematoxylin and eosin. Metastases were

examined in size-matched tumors. All animal care and experimentation was conducted in accordance with the National Academy of Sciences Guide for the Care and Use of Laboratory Animals. To examine the effects of Gal-1 on HCC cell physiology, we first assessed its expression and subcellular distribution Silmitasertib clinical trial in the differentiated HepG2 cell line. We stably transfected HepG2 cells with Gal-1 complementary DNA. Two G418-resistant clones were selected with approximately two-fold (HepG2-G1) and five-fold (HepG2-G2) higher Gal-1 expression compared with nontransfected cells and cells transfected with empty vector (HepG2-M) (Fig. 4-Aminobutyrate aminotransferase 1A). Gal-1 immunostaining of HepG2-G2 cells showed clear cytoplasmic localization, whereas no staining was observed on HepG2 (Fig. 1B) and HepG2-M cells (data not shown). However, immunostaining performed on permeabilized HepG2 cells incubated for 48 hours at 37°C in the presence of exogenously added rGal-1 showed positive cytoplasmic localization. Because real-time polymerase chain reaction analysis showed that rGal-1

does not induce transcription of its own gene (Supporting Information Fig. 1), this result suggests that rGal-1 may be internalized by HepG2 cells. Furthermore, no membrane-associated Gal-1 was detected in nonpermeabilized cells (data not shown). Moreover, examination of Gal-1 secretion revealed a faint immunoreactive band in concentrated serum-free conditioned medium of HepG2 and HepG2-M cultures, an effect that was considerably increased in Gal-1–transfected cells (Fig. 1C). These results indicate that HepG2-G2 cells express and secrete high levels of Gal-1. To study the influence of Gal-1 on hepatocyte function, we performed cell adhesion assays. When cells were incubated for 1 hour on uncoated plates in the presence of rGal-1 (3.