Materials and methods

Animals and ethics statement
Fourteen male castrated SPF cats divided in four different age groups were used in this study: group 1 (c01-c04, 10 weeks), group 2 (c05-c08, 1.5 years), group 3 (c09-c12, 7 years) and group 4 (c13 and c14, 14 years). Cats c13 and c14 from group 4 originated from the same litter; all other individuals were not related to each other. The in vivo experiments were conducted when cats c05-c08 were 3 years of age. All animals were purchased from Liberty Research Inc. (Waverly, NY, USA) and their SPF status was verified as previously described[54]. This study was carried out in strict accordance with regulations of the Swiss law for animal protection (SR 445.1). The Veterinary Office of the Swiss Canton of Zurich officially revised the protocol and approved the study (Permit no. TVB 99/2007 and TVB 100/2007). The animals were housed in groups in an animal-friendly barrier facility under optimal ethological conditions[55]. For blood collections and injections, the cats were sedated with a combination of ketamin and midazolam, and all possible efforts were made to minimize stress and suffering.

Feline PBMC isolation, cell lines, ODNs, cell culture and cell viability assay
Feline PBMCs were isolated from EDTA-anticoagulated whole blood by Ficoll-Hypaque density gradient centrifugation using a standard protocol[56]. Purified cells were counted prior to their utilization in the different experiments using the Sysmex XT 2000iV (Sysmex, Norderstedt, Germany) as described previously[57], and cultured in RPMI 1640 with Glutamax I (Gibco®, Invitrogen, Basel, Switzerland). Adherent CRFK (ATCC no. CCL-94) and FEA cells were maintained in RPMI 1640 with Glutamax I, while adherent fcwf-4 cells (ATCC no. CRL-2787) were cultured in EMEM (ATCC 30–2003). All media were supplemented with 10% heat-inactivated fetal calf serum (Bioconcept, Allschwil, Switzerland), 100 U/mL penicillin and 100 mg/mL streptomycin (Gibco®, Invitrogen).
ODN 2216 and ODN 2243 were obtained from Alexis biochemicals, Enzo Life Sciences AG, Switzerland for in vitro studies and ODN 2216 was synthesized by Microsynth AG, Balgach, Switzerland for the in vivo experiment. Recombinant feline IFNα (rfeIFNα) was obtained from PBL Biomedical, Piscataway, New Jersey, USA. All molecules were solubilized in endotoxin-free PBS. ODN 2243 consists of the same sequence as ODN 2216, with CpG motifs inversed to GpC. For in vitro experiments, both ODNs and rfeIFNα were diluted in RPMI 1640 with Glutamax I supplemented as described above.
Viability of stimulated cells was compared using the trypan blue exclusion test. Briefly, after stimulation for 24 h with increasing concentrations of ODN 2216, ODN 2243 or equivalent volumes of PBS as control, cells were stained with a 0.4% trypan blue solution (Dr Bender and Dr Hobein AG, Zurich, Switzerland) and percentages of viable cells were compared.

Proliferation assay
PBMCs were seeded immediately after isolation at a concentration of 3 × 106 cells/mL in 96-well U-bottom plates. Triplicate cultures for each cat were treated with either 4 μg/mL ODN 2216 or 2243 or an equal volume of endotoxin-free PBS. After an initial incubation of 18 h, the cells were pulsed for 24 h with 3 H-thymidine (Perkin Elmer, Schwerzenbach, Switzerland). Standard liquid scintillation protocols were used for harvesting of the cells and uptake of 3 H was assessed with the Packard Tri-Carb 1600TR liquid scintillation analyzer (Perkin Elmer). Proliferation rates were calculated as the mean counts per minute (c.p.m) of triplicate cultures. The stimulation index (SI) depicted in Figure1 was calculated as follows:Mean c.p.m.of ODN-treated culturesMean c.p.m of PBS-treated cultures
Figure 1 ODN 2216 induce proliferation of primary feline immune cells and enhance their expression of costimulatory surface molecules. (A) The proliferation of feline PBMCs of eight adult cats belonging to groups 2 and 3 after stimulation with ODN 2216 and ODN 2243 was assessed by H3-thymidine incorporation. The dots indicate the stimulation index calculated from triplicate cultures for each cat. (B and C) Expression levels of B7.1 and MHCII surface molecules were assessed 24 h post stimulation with the indicated treatments by flow cytometry. The fluorescence of gated PBMCs of one cat selected as an example is depicted. Isotype control samples are indicated as unstained. (D-G) Mean fluorescence intensity (MFI) for B7.1 (D and E) and MHCII (F and G) of gated lymphocytes and non-lymphocyte subpopulations stimulated with the indicated treatments. Results for eight cats belonging to groups 2 and 3 are shown. *p < 0.05, **p < 0.01.

Flow cytometry
PBMCs were treated at a density of 3 × 106 cells/mL with 4 μg/mL ODN 2216 or ODN 2243 or an equivalent volume of endotoxin-free PBS and cultured for 24 h in a 12-well format. During collection of the cells, the adherent cell fraction was removed with 0.05% trypsin-EDTA (Gibco®, Invitrogen). Harvested cells were divided into 3 fractions labeled separately with either anti-feline B7.1 mouse monoclonal IgG (kindly provided by Prof Mary Tompkins, Flow Cytometry and Cell Sorting Laboratory, NC State College of Veterinary Medicine, USA), anti-feline MHCII mouse monoclonal IgG1 (Department of Pathology, Microbiology and Immunology, University of California, Davis, USA) or fluoresceinisothiocyanate (FITC)-conjugated mouse IgG1 as isotype control (BD Bioscience, Allschwil, Swizerland). The fractions were subsequently stained with R-Phycoerythrin (RPE)-conjugated goat anti-mouse IgG1 (BioConcept, Allschwil, Swizerland). Fluorescence data was obtained using the FACSCalibur® instrument (Becton Dickinson, Allschwil, Switzerland) and the CellQuestPro™ software. Gates representing lymphocyte and non-lymphocyte populations were set on the basis of forward versus side scatter, and a total of 50 000 events were acquired in the non-lymphocyte gate. Data was analyzed with the FlowJo software (Tree Star, Olten, Switzerland), whereby an additional gate was set comprising both lymphocyte and non-lymphocyte populations (PBMC gate). MHCII and B7.1 expression levels were determined as mean of fluorescence intensity for each gated cell population. Identical gates were set for all cats in such a way that they comprise the desired cell populations of each animal.

Relative gene expression analysis
PBMCs were stimulated with ODN 2216, ODN 2243 or endotoxin-free PBS at a density of 3 × 106 cells/mL directly after isolation, while CrFK, FEA or fcwf-4 cells were cultured to confluency prior to stimulation. Experiments were carried out in a 96-well format. For stimulation of adherent cells with supernatants (production see below), total cell culture medium was discarded from the wells and the monolayers were further cultured in 100 μL undiluted supernatants for the rest of the experiment. At time points relevant to each experiment, the supernatants were removed and cells were lysed with mRNA lysis buffer (mRNA isolation kit I, Roche Diagnostics, Rotkreuz, Switzerland). mRNA extractions were performed with the mRNA Isolation Kit I and MagNA Pure LC Instrument (Roche Diagnostics) and first strand cDNA was synthesized with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Rotkreuz, Switzerland). Real-time quantitative PCR (qPCR) reactions consisted of 5 μL cDNA in a total volume of 25 μL per reaction using the TaqMan® Fast Universal PCR Master Mix (Applied Biosystems). Thermocycling conditions included an initial denaturation of 20 s at 95°C followed by 45 cycles of amplification by melting at 95°C for 3 s and annealing at 60°C for 45 s. Primers and probes for feline genes have been previously described[58]. mRNA expression factors of selected genes, which correspond to ratios of mRNA levels measured in ODN 2216 or ODN 2243 stimulated versus PBS stimulated cells, were calculated and normalized with GeNorm version 3.5[59], using either both feline β-glucuronidase (GUSB) and tryptophan 5-monooxygenase activation protein zeta polypeptide (YWHAZ) (usually) or Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) alone (when specified) as reference genes, under conditions validated for the feline species[60]. Generally depicted in the graphs are mean expression factors calculated from duplicate experiments carried out simultaneously. Where results of one cat are illustrated, experiments were conducted with cells of at least 3 individual animals and representative data is shown.

Production of supernatants
For each cat, PBMCs were resuspended in supplemented RPMI 1640 with Glutamax I (Gibco®, Invitrogen) at a concentration of 106 cells/mL in 6-well plates and stimulated immediately after isolation with 4 μg/mL ODN 2216, 4 μg/mL ODN 2243 or an equivalent volume of endotoxin-free PBS. After 24 h incubation, supernatants were harvested by centrifugation of the cultures twice at 2000 × g for 10 min, aliquoted and stored at −20°C. Large supernatant quantities were produced with PBMCs purified after one blood collection, enabling the utilization of the same supernatants for virus inhibition experiments concerning VSV, FCV, FCoV, FHV and FPV. New supernatant batches were produced for use in FeLV inhibition assays. Supernatants derived from PBMCs stimulated with ODN 2216, ODN 2243 and endotoxin-free PBS are referred to as Sup2216, Sup 2243 and Sup Neg in the text and in the figures.

Western blot
CrFK and fcwf-4 grown to confluency in 12-well plates were stimulated with 600 μL of PBMC supernatants produced as explained above. At the time points indicated, cells were harvested and counted. 106 cells of either cell line were resuspended in 30 μL sample buffer (0.5 M Tris(hydroxymethyl)aminomethane, 5% SDS, 10% β-mercaptoethanol, 40% glycerol, and 0.05% bromphenol blue) and boiled at 95°C for 5 min. SDS-PAGE separation and submersed immunoblotting procedures were carried out as previously described[61]. The Spectra Multicolor Broad Range Protein Ladder (Fermentas GmbH, St. Leon-Rot, Germany) served as molecular weight standard marker for each blot. For protein visualization, membranes were first cut immediately below the 80kB marker band. The top and bottom membrane fractions were incubated with murine anti-human Mx MAb M143 (generously provided by Dr J. Pavlovic, Institute for Virology, University of Zürich, Switzerland) and murine anti-β-actin monoclonal antibody as a loading control (Sigma Aldrich GMbH, Buchs, Switzerland) respectively. Both fractions were subsequently incubated with a peroxidase-labelled goat anti-mouse IgG (Jackson Immunoresearch, Newmarket, Suffolk, UK). Bands were digitalized using the Chemigenius 2 BioImaging System (Syngene, Cambridge, UK).

Viruses and viral inhibition assays
VSV Indiana strain (Institute of Virology, Vetsuisse Faculty, University of Zurich, Switzerland), FCoV Wellcome strain (a generous gift from Prof. A. Kipar, University of Liverpool, Great Britain), FPV (kindly provided by Prof. U. Truyen, University of Leipzig, Germany), FHV ZH5-04 strain and FCV F9 strain (kindly provided by Veterinaria AG, Zurich, Switzerland) were titrated on both CrFK and fcwf-4 cells. Viral stock dilutions inducing 95% cytopathic effect (CPE) after 24 h (72 h for FCoV and FPV) were selected for inhibition experiments in order to ensure proper measurement of inhibitory effects. Monolayers of CrFK and fcwf-4 cells in 96-well plates were incubated for 24 h with 100 μL of the supernatants produced with PBMCs from cats of groups 1, 2 and 3. With the exception of assays carried out with FPV, viral inhibition experiments were conducted simultaneously and with supernatants thawed an equal number of times. The treated cells were then inoculated with virus (VSV, FCV, FHV, FCoV) or trypsinized with 0.05% trypsin-EDTA (Gibco®, Invitrogen) and allowed to settle in viral suspension (FPV), and inhibition assays were carried out after 24 h (72 h for FCoV and FPV) according to the procedure described previously[62]. Briefly, supernatants were discarded and cell debris was removed from the wells by 3 cycles of washing with Hank’s balanced salt solution (HBSS) (Gibco, Invitrogen) and shaking on an orbital shaker for 15 s. Remaining cells were fixed with 5% formalin and stained with a crystal violet solution. For spectrophotometric measurements, 100% methanol was added to the dried out wells and absorbance was read at 595 nm on a SpectraMax Plus 384 microtiter plate reader (Molecular Devices, Bucher Biotec AG, Basel, Switzerland). Viral inhibition rates were calculated with the following formula:Mean optical densityODvalues of duplicate wells treated with SupernatantMean OD values of quadruplicate wells treated with medium alone
FeLV-A Glasgow-1 strain (a generous gift from Prof. M. Hosie and O. Jarret, University of Glasgow, Great Britain) was titrated on FEA cells, and the lowest stock dilution leading to productive infection of the cells after 48 h was used for inhibition assays. Experiments were carried out in 96-well plates and cells were treated with 100 μL of supernatants or relevant controls immediately prior to inoculation. Every second day thereafter, 50 μL culture medium was replaced by the same volume of fresh supernatant. At appropriate time points, cells and supernatants were harvested and total nucleic acid was extracted from both the cells and supernatants using the MagNA Pure LC DNA Isolation Kit I and MagNA Pure LC Instrument (Roche Diagnostics). Viral replication in supernatants and proviral loads in cells were measured by real-time RT-PCR and real-time PCR respectively, with assays previously described[63]. The time course experiments were conducted with supernatants derived from PBMCs of three cats and the measurements on day 8 post inoculation were carried out with material derived from two additional cats. In order to facilitate interpretation of the figures illustrating measurements of viral RNA loads, 45 cycles-cycle threshold (Ct) values were calculated and means of duplicate wells are depicted.

In vivo experiment
Cats c05 and c06 were administered ODN 2216 once subcutaneously at a dose of 0.2 mg/kg body weight while cats c07 and c08 received endotoxin-free PBS as control. The treatments were distributed in equal portions bilaterally in axillary and inguinal regions. Heparinized blood was collected at time points 0 h (just before injection of the treatment), 8 h, 24 h, 48 h, 96 h and 192 h. The cats were clinically examined and body temperature was measured at each time point. Whole blood samples were used in part for the assessment of hematological paramaters with the Sysmex XT 2000iV (Sysmex, Norderstedt, Germany) as previously described[57]. The rest of the blood was mixed with mRNA lysis buffer (Roche Diagnostics) immediately upon blood collection and processed as recommended by the manufacturer. Plasma was obtained by centrifugation of whole blood samples for 10 min at 1500 × g and conserved at −80°C until use for in vitro experiments. Mx mRNA expression was measured by qPCR and normalized to the expression of two feline housekeeping genes as described above. The expression factors depicted in the figures represent the ratio of mRNA levels from a specific time point to mRNA levels from time point 0 h. For viral inhibition assays, the plasma collected from each cat at each time point was incubated for 24 h with fcwf-4 cells prior to their inoculation with FCV. The assays were carried out as described above and viral inhibition factors were calculated as follows:Mean optical densityODvalues of duplicate wells treated with plasma from timepoint0hMean OD values of duplicate wells treated with plasma from timepoint Xh

Statistical analysis
All statistical analyzes were performed using GraphPad Prism for Windows, version 3.0 (GraphPad Software, San Diego California USA). Due to the limited number of cats integrated in the study, we refrained from using a parametric approach in the statistical tests. As such tests require a larger sample size than n = 4, it was not possible to calculate p-values for the induction in expression of each gene shown in Figure2. Differences between treatment groups in proliferation of PBMCs were analyzed with a Mann–Whitney test and expression of co-stimulatory molecules on the surface of these cells were analyzed with a Wilcoxon signed rank test, where values for each cat were paired. Relative Mx mRNA expression, OD values from viral inhibition assays, ratios for FeLV provirus and Ct values for FeLV viral RNA loads were analyzed with a Wilcoxon signed rank test with pairing of values for each cat when treatment with different supernatants were compared, or a Mann–Whitney test when incubation with supernatant was opposed to treatment with medium alone or rfeIFNα. Normalized and relative TLR9 mRNA expression ratios between different animal age groups were also compared with a Mann–Whitney test. Longitudinal effects on FeLV viral and proviral loads were compared with each other using a Mann–Whitney test carried out with Area Under the Curve (AUC) values. Correlations were assessed using the Spearman test. p-values < 0.05 were considered statistically significant.
Figure 2 ODN 2216 induces an antiviral gene expression profile in PBMCs of adult cats. (A-D) mRNA expression factors of the indicated genes were measured in PBMCs isolated from fourteen cats belonging to the indicated age groups and stimulated with ODN 2216 for 24 h. (E) mRNA expression factors of the indicated IFNα subtype genes were measured in PBMCs isolated from four adult cats (group 2) and stimulated with ODN 2216 for 24 h. (F) mRNA levels of TLR9 were measured by real-time qPCR in unstimulated PBMCs of all four groups of cats and normalized to the expression of a feline housekeeping gene (GAPDH). (G) Relative TLR9 mRNA expression factors were measured in ODN 2216 stimulated PBMCs from cats of all four groups. *p < 0.05, Perf = perforin, GranB = granzyme B.