Cope's gray treefrog, Hyla chrysoscelis, is a freeze-tolerant amphibian that accumulates cryoprotective glycerol during cold acclimation (1,2). Control of intracellular and extracellular water content, and glycerol distribution, promotes extracellular freezing while protecting intracellular structures from ice damage. Regulation of those water and glycerol fluxes is enhanced by aquaporins (AQPs) and aquaglyceroporins (GLPs), integral membrane proteins that facilitate water and, in some cases, glycerol flux across cell membranes (3). Of special interest to the physiology of freeze tolerance is the GLP HC-3, cloned and functionally characterized from H. chrysoscelis (2). HC-3 is expressed in erythrocytes and other tissues (2,4,5), and HC-3 protein abundance in erythrocytes increases during cold acclimation (5). Although previous studies have shown a correlation between enhanced GLP expression, cryoprotection, and freeze tolerance (2, 5-8), the functional role of these proteins in conferring freeze tolerance remains poorly understood. To address this question, an in vitro cell culture model system of HC-3–expressing cells was developed. Suspension cultures of erythrocytes from H. chrysoscelis were chosen because they are nucleated, metabolically active, and therefore capable of regulating gene expression as well as peptide-mediated endocytosis, and can be repetitively harvested as a highly homogenous cell population. Since there are no specific inhibitors for AQPs/GLPs, and tools for gene knockout technology do not exist for H. chrysoscelis, an antisense phosphorodiamidate morpholino (PMO) knockdown approach was pursued. Antisense PMOs act on RNA targets by sterically blocking the translation initiation complex (9). Techniques available for delivery of PMOs to cells in culture include syringe or scrape loading, lipsomes, cationic polymers, free uptake from the media, and electroporation (9). Electroporation has also been used in adult axolotl for studying tissue regeneration (10), whereas microinjection and electroporation are widely used for delivery of PMOs into nonmammalian developmental models including sea urchin, Xenopus, zebrafish, and Drosophila embryos (11). Endo-Porter (GeneTools, Philmoth, OR, USA) is a novel peptide that mediates PMO delivery through nonspecific endocytosis, is effective in a wide variety of cell types, is easy to use, and is nontoxic to cells (12). Endo-Porter has been used in vitro to deliver PMOs to mammalian adherent and nonadherent cells cultured at 37°C (12,13) and ex vivo in tissue explants (14), resulting in a range of 50–80% targeted expression knockdown. One recent study used Endo-Porter–mediated PMO delivery to target endogenous gene knockdown in cultured carp macrophages (15). Endo-Porter has also been used successfully to deliver PMOs in vivo in newts (16). To date, however, Endo-Porter has not been widely used in nonmammalian cell culture systems. The objective of this study was to determine under what conditions and with what efficiency Endo-Porter–mediated PMO delivery could be used to knock down targeted gene expression in nonmammalian cell suspension cultures, specifically in amphibian erythrocytes.

The HC-3 PMO was selected using the criteria defined for custom PMO design available through GeneTools. The HC-3 antisense PMO sequence (5′-CCCATGTTGCTGAGCCTCTAGGTC-3′) was designed based on complementarity to the following region in HC-3 (indicated in brackets on the sense strand, with the start codon in parentheses): ACTACTCAGCCGGCAGCATCACAGCT CTCCCC[GACCTAGAGGCTCAGCAAC(ATG)GG]GCGCCAGAAGGAGGTTCTCA. The commercially available standard control oligo (5′-CCTCTTACCTCAGTTACAATTTATA-3′; Gene Tools) with no known target in H. chrysoscelis was used as a PMO control.

The optimized procedure is as follows: Whole blood from H. chrysoscelis was collected in heparinized capillary tubes, transferred to 15-mL conical tubes containing 5 mL complete cell culture media [CCCM; 250 mOsM, RPMI 1640 medium supplemented with L-glutamine, 100 units/mL penicillin and streptomycin, 0.25 µg/mL amphotericin B (Invitrogen, Carlsbad, CA, USA), and 5% FBS (Fisher Scientific, Hanover Park, IL, USA)], centrifuged at 1000× g for 10 min, and resuspended in CCCM. Approximately 2 × 10⁶ cells were incubated in 200 µL media in round-bottom polypropylene tubes (Fisher Scientific) at 37°C for 2 h while shaking at 190 rpm on a titer plate shaker (Lab-Line Instruments Inc., Melrose Park, IL, USA). The suspension volume was increased to 1 mL followed by addition of one of the following treatments: A, no treatment (control); B, 10 µM Endo-Porter alone (EP); C, 3 µM HC-3 PMO alone (HC-3 M); D, 10 µM Endo-Porter and 3 µM HC-3 PMO (EP + HC-3 M); and E, 10 µM Endo-Porter and 3 µM standard control PMO (EP + control M). Endo-Porter and standard or HC-3 PMO were sequentially added to the erythrocyte cultures. Cultures were maintained at 22–25°C within the range of basal body temperature of live treefrogs for 24 h while shaking. After 24 h, media was replaced with 200 µL fresh CCCM and erythrocytes were again incubated at 37°C for 2 h while shaking. At the end of 2 h, 800 µL CCCM was added to each culture. The relevant cultures were spiked with an additional 3 µM control or HC-3 PMO and/or 10 µM Endo-Porter, and maintained for an additional 24 h at room temperature after which the media was replaced; cells were cultured for a total of 76 h. Initial optimization experiments showed that cell viability at 76 h (~98%) remained unchanged between untreated erythrocytes and erythrocytes cultured in Endo-Porter alone (6–10 µM), HC-3 PMO alone (3–10 µM), or Endo-Porter and HC-3 PMO (each at 10 µM) (data not shown). At 76 h, total cellular proteins were isolated, size-fractionated by SDS-PAGE, and subjected to Western hybridization using a peptide-derived, monospecific rabbit polyclonal antibody raised against 16 C-terminal amino acids of HC-3 (0.44 µg/mL), or mouse anti–β-actin antibody, (diluted 1:5000; Sigma-Aldrich, St. Louis, MO, USA) followed by incubation with HRP-conjugated goat anti-rabbit secondary antibody (1:1000; Santa Cruz Biotech, Santa Cruz, CA, USA) or goat anti-mouse secondary antibody (1:1000; Santa Cruz Biotech) as previously described (5). Relative band intensities were determined by densitometry using Vision Works software on a BioSpectrum Imaging System (UVP, Upland, CA, USA). Averaged HC-3 protein abundance (normalized to β-actin) was expressed as a percentage of the no-treatment control group (A).