Introduction

Endocytosis, a hallmark of all eukaryotic cells, results in the intracellular uptake of membrane proteins, lipids, and a variety of extracellular ligands. Endocytosis is involved in cellular processes such as nutrient uptake, morphogenesis, synaptic molecule recycling, and regulation of the cell surface expression of signaling receptors, transporters, and channels (1). The first endocytic event, the internalization step, has been extensively studied in both lower and higher eukaryotic cells. Studies in mammalian cells have revealed the existence of multiple endocytic pathways, including macro-pinocytosis and clathrin-mediated endocytosis, as well as internalization via caveolae, caveolae-like structures, and lipid rafts (2,3,4). Multiple studies have focused on uptake through clathrin-coated pits because about 50% of all ligands are internalized by this route. Therefore clathrin-mediated endocytosis is often considered as a first tentative internalization pathway for a given molecule (2).

One major obstacle in identifying a particular endocytic pathway as unique for the uptake of a ligand resides in developing an experimental strategy exclusively suppressing that specific internalization route without affecting other internalization pathways. For example, cholesterol extracting or sequestrating drugs such as cyclodex-trins, filipin, or nystatin disrupt not only caveolae/lipid rafts, but destabilize clathrin-coated pits as well (5,6). This lack of specific pharmacological inhibitors frequently hampers the study of internalization processes. Other experimental approaches such as dominant negative overexpression, gene targeting, selective inactivation, and RNA interference (RNAi) have been exploited in the study of clathrin-mediated endocytosis (7,8,9,10,11,12). A frequently used tool to disrupt internalization via coated pits is dominant negative overexpression of a K44A mutant of dynamin; next to dynamin, amphiphysin and EPS15 have been subjected to dominant negative overexpression, however mutant proteins have their limitations as well (13,14,15). Therefore alternative approaches like gene targeting, selective inactivation, and RNAi have been exploited.

Here, a method was developed to suppress clathrin-dependent uptake very specifically. To this end, the RNAi technique was used to silence expression of a core component of the clathrin-mediated endocytosis in a sequence-specific manner. Introduction of small interfering RNA (siRNA) in mammalian cells induces strong and specific suppression of the gene of interest. However, this effect is transient due to the short lifespan of synthetic RNAs and the absence of RNA-dependent RNA polymerase in mammalian cells, thus limiting its applications (16,17,18). To overcome these limitations, a vector-based production of short hairpin RNA (shRNA) was used, offering perspectives to construct a cell line that permanently restrains translation of messenger RNA (mRNA) encoding for the target protein. Based on recent insights at the molecular level of clathrin-mediated endocytosis, we have chosen epsin as the most appropriate factor to target shRNA.

According to Ford et al. (19) epsin is an essential accessory protein implicated in the clathrin-mediated uptake. It is widely expressed in many mammalian cell types (20,21,22,23,24,25) and a homologous protein has also been found in Drosophila melanogaster (24). Epsin is composed of different domains including an epsin N-terminal homology (ENTH) domain, clathrin binding motifs, ubiquitin interacting consensus motifs, and C-terminal Eps15 homology (EH) domain binding sequences (25). The ENTH domain has been proposed to represent a separately folded protein module (20) displaying a super helix of seven α-helices with a supplementary phosphatidylinositide (PIP₂)-induced α₀ helix misaligned with the super helical axis (26,27,28). Insertion of the α₀ helix in the inner plasma membrane leaflet is crucial for induction of membrane curvature and therefore one of the first events in coat formation (27). The ENTH domain is highly conserved (29,30) and unique for clathrin-mediated endocytosis at the plasma membrane (20)(31,32,33,34,35,36), rendering it suited to select stretches of 20 nucleotides to target shRNA (22,31,37,38). All these features make epsin an ideal target protein to suppress in order to study involvement of clathrin-mediated endocytosis in the uptake of bioligands.

Materials and Methods

Cell Culture

Human intestinal colon carcinoma cells, Caco-2^wt (passages 5 to 34) and Caco-2^eps- cells (passages 1 to 12) were cultured in modified Eagle's medium (MEM) supplemented with 10% heat-inactivated fetal bovine serum, nonessential amino acids, 1 mg/L amphotericin B, 50 mg/L gentamicin, 50 mg/L penicillin/streptomycin, and 55 mg/L sodium pyruvate (all from Gibco®; Invitrogen, Carlsbad, CA, USA). Cell medium was changed twice a week, and cells were passed weekly. After transfection, cell medium was supplemented with 2 g/mL selection antibiotic Zeocin® (Invitrogen).