The in situ piglet preparation
We have previously described an ex-vivo perfusing piglet preparation, which permits perfusion restricted to the lung and the liver using autologous blood . Piglets were anesthetized, a tracheostomy was performed, and cannulas were placed in a carotid artery and a jugular vein. After heparinization, the animals were ex-sanguinated and the blood was used to fill the blood reservoir and the perfusion circuit. The lungs were perfused at a constant flow rate of 40 ml/kg/minute (up to 400 ml/min); half of the blood in the efferent circuit transverses the liver before perfusing the lung. In all the preparations, lungs were ventilated under the same conditions with a piston ventilator (Harvard Apparatus, Dover, MA, USA) with room air and 5% CO2. The minute ventilation was adjusted to maintain perfusate pH between 7.35 and 7.45. Throughout the experiments, pressures in the pulmonary artery, left atrium and portal vein and flow in and out of both organs was monitored continuously and the data were stored in a computer for later analysis.
The study adhered to National Institute of Health guidelines on the use of experimental animals and was approved by the institutional animal care and use committee of Emory University. Animals were maintained in the Emory University Division of Animal Resources, an AAALAC approved facility.
Experimental protocols and data collection
Six groups of experiments were performed with four to six animals in each group. The groups were designated as: 1) endotoxin alone, 2) endotoxin + freshly isolated autologous bone marrow nuclear cells, 3) endotoxin + non-hematopoietic bone marrow autologous cells CD45(-), 4) endotoxin + hematopoietic bone marrow cells CD45(+), 5) endotoxin + buffy coat and 6) endotoxin + in vitro expanded swine CD45 negative adherent allogeneic bone marrow cells (cultured CD45neg). For all the groups, 5 μg/kg body weight of endotoxin was used (Serotype O55:B5 Difco, Detroit, MI, USA), directly into the blood reservoir fifteen minutes after endotoxin cells were added to the blood reservoir. A total of 100 x106 whole bone marrow cells were used. In the case of the subpopulations, CD45 (-), CD45(+), buffy coat and cultured CD45neg, 50 x106 purified cells were perfused in each preparation. Blood samples were collected before and after endotoxin administration. Blood gas analysis was performed every 15 minutes for the duration of the experiment using a portable blood gas analyzer (Rapidlab 248, Diamond Diagnostics, East Walpole, MA, USA). Serum samples were separated and stored at -80o for measurement of TNF-α, IL-6 and IL-1β. Automated cell count and differential counts were performed every 15 minutes using a device which is specifically designed for use in laboratory animals (Hemevet, CDC Technologies, Oxford, CT, USA). At each time point, we recorded pressures and flows in and out of the lungs and flow to the liver. Pulmonary vascular resistance (PVR) was calculated as in-flow pressure (pulmonary artery pressure) minus out-flow pressures (left atrial pressure) divided by total blood flow through the lungs and expressed in cm H2O/ml/min. Temperature was measured continuously and kept between 38°C and 40°C by adjusting the temperature of the water in the water-jacket surrounding the reservoir.
Lung and liver tissue specimens were obtained at three time points (30 minutes before endotoxin and 45 minutes and 135 minutes after endotoxin) for the determination of histology, cell engraftment and gene expression in a microarray system. The study was terminated two and a half hours after endotoxin administration and both lungs were taken for determination of wet-to-dry weight ratio.
Generation and administration of cells
Briefly, fresh swine bone marrow derived cells were isolated by mechanically extracting bone marrow from both femurs of the animal while the preparation was being set up. The harvested bone marrow was cut in small fragments and cells detached by continuous vortex for 30 seconds followed by 1X trypsin-ethylenediaminetetraacetic acid (EDTA) for 2 minutes. Cell suspensions were centrifuged for 2 minutes at 700 RPM to separate insoluble material. Supernatant was collected and cells transferred to a 50 ml conical tube where cells were washed twice with PBS containing penicillin-streptomycin. Red blood cells were lysed by re-suspending the cell pellet with 3 ml of sterile water. After 15 seconds lysis was stopped by adding 27 ml of 10X PBS for a final concentration of 1X PBS. Cells were washed twice with sterile PBS and counted in a hemocytometer. To isolate CD45 (+) and CD45 (-) cells, 300 × 106 bone marrow cells were transferred into a 15 ml conical tube and incubated with 35 μg of fluorescein isothiocyanate (FITC) conjugated mouse anti-porcine CD45 (ABD Serotec, Oxford, UK) for 10 minutes at 4#176;C. Cells were washed twice with cold MACS buffer (PBS +0.5% BSA+2mM EDTA). After washing, cells were re-suspended in 1.5 ml MACS buffer and incubated with 300 μl of anti-FITC microbeads (Myltenyi Biotec, Auburn, CA, USA) for 15 minutes at 4#176;C. CD45 (-) cells were sorted using LD columns (Myltenyi Biotec, Auburn, CA, USA). Cell sorting was repeated once to ensure the purity of the cells. Cells that were retained in the column were used as CD45 (+) cells. After the final elution, cells were washed twice with sterile saline solution. Before infusion, cells were counted in a hemocytometer with Trypan blue to determine cell numbers and viability.
Autologous buffy coat cells were obtained by separating leukocytes from fresh heparinized peripheral blood by osmotic gradient using HISTOPAQUE® 1083 (Sigma, St Louis, MO, USA). Twenty ml of fresh blood was centrifuged and red blood cells lysed. Cells were re-suspended in 20 ml of PBS and placed on top of 20 ml of Histopaque 1083. The opaque interface, containing the mononuclear cell band, was collected using a Pasteur pipet and transferred into a 50 ml conical centrifuge tube. Cells were washed twice with PBS. After the last wash, cells were counted to determine viability and cell numbers.
Swine CD45 negative adherent allogeneic bone marrow cells (cultured CD45neg) were isolated by negative selection of CD45(+) cells, using an anti-CD45-FITC antibody, followed by anti-FITC-conjugated magnetic beads (Miltenyi Biotec). Cultured CD45neg cells were expanded in vitro in 175 cm2 flasks in Iscove's modified Dulbeccos's medium (IMDM) containing 9% fetal bovine serum (FBS, Atlanta, Biologicals, Norcross, GA, USA), 9% horse serum, and 1% penicillin-streptomycin. Cells were harvested at 70% confluence using 0.25% trypsin. For in vivo infusion, cells were detached using 0.25% trypsin at 37°C for 5 minutes. Trypsin was neutralized by adding IMDM with 10% serum. The cell suspension was centrifuged and suspended in sterile PBS (without Ca and Mg). After two additional washes with PBS, cells were counted in a hemocytometer with trypan blue and were re-suspended in PBS at 5 × 106 cells per ml final concentration.
To track the cells in the swine preparation, BMCs were surface-labeled using PKH (Sigma. St Louis MO, USA). Prior to infusion, 10 × 107 cells were re-suspended in 5 ml of staining buffer and 50 μl of PKH solution were added. After incubating for 5 minutes at room temperature, the reaction was stopped by adding 10 ml of complete FBS. Cells were washed twice with 50 ml of sterile saline solution. Before infusion, cells were counted to determine viability and cell numbers.
Blood samples obtained from preparations were collected at different time points. Red blood cells were lysed by hypotonic shock and samples analyzed for PKH-26 staining on a fluorescence-assisted cell sorting (FACS) Excalibur (Becton Dickinson, San Jose, CA, USA), using CellQuest software (Becton Dickinson, San Jose, CA, USA) and analyzed using FlowJo software (Tree Star, San Carlos, CA, USA).
Pig lungs were placed in 2% paraformaldehyde and processed for paraffin embedding. Sections (5 mm) were cut, mounted on the slides, and subjected to antigen retrieval in a decloaking chamber (BioCare Medical, Gaithersburg, MD, USA). Endogenous peroxidase activity was quenched with 3% peroxide for 5 minutes. Slides were incubated with anti-myeloperoxidase antibody (MPO) (Abcam Cambridge, MA, USA) overnight at 4#176;C. Immune complexes were visualized with biotinylated secondary antibody and 3,3'-diaminobenzidine tetrahydrochloride using the streptavidin-biotin complex method.
Plasma concentrations of TNF-α, IL-6 and IL-1β were measured separately using Quantikine® ELISA kits (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions. Briefly, samples were dispensed into 96-well micro-titer plates that are precoated with porcine monoclonal or polyclonal antibodies specific to the above cytokines. After washing away any unbound substances, an enzyme (horseradish peroxidase)-linked monoclonal (TNF-α) or polyclonal (IL-6 and IL-1β) antibody specific to the above cytokines was added to the wells. Following a wash to remove any unbound antibody-enzyme reagent, a substrate solution (hydrogen peroxide/tetramethylbenzidine) was added to the wells. The reaction was terminated with a stop buffer and absorbance read at 450 nm using an MRX Revelation (Dynex Technologies, Chantilly, VA, USA) multi-well plate reader. Values of IL-6, TNF-α and IL-1β were determined by reference to a standard curve constructed using the porcine proteins and computer software capable of generating a four parameter logistic curve-fit.
Measurements of lung water
At the completion of each experiment, both lungs were harvested and homogenized. The homogenate was weighed (wet weight), then dried to constant weight in a microwave oven and reweighed (dry weight). The ratio of wet/dry lung weight was calculated as a measure of the amount of lung water (edema) that was present.
Liver RNA isolation and Affymetrix GeneChip array
We used a swine specific microarray chip from Affymetrix with 27,000 genes represented. RNA was isolated from liver tissue samples taken 30 minutes after endotoxin in a liver-lung perfused preparation or after 30 minutes of perfusion without addition of endotoxin as a control. We performed microarray analyses of these samples and calculated the ratio of expression in the sample from the endotoxin-treated preparation to the control sample. Liver tissue samples from both endotoxin and endotoxin-bone marrow cell groups were homogenized in TRIzol reagent (Gibco BRL, Gaithersburg, MD, USA), and RNA was extracted according to the instructions of the manufacturer. RNA precipitates were briefly air-dried then subjected to an RNeasy protect mini kit and DNAse I treatment (Qiagen, Miami, FL, USA). Total RNA concentration was measured spectrophotometrically at 260 and 280 nm and dissolved in RNAse-free water to 1 µg/µL and stored at -80°C. For control of concentration and integrity, 1 µg of each RNA sample was subjected to formaldehyde-agarose electrophoresis.
Total RNA (5 µg) was used for microarray analysis on swine Affymetrix chips. Preparation of complementary RNA, hybridization, scanning and image analysis of the microarrays were each performed according to the manufacturer's protocol (Affymetrix, Santa Clara, CA, USA) at the Emory Bio-Core facility with a final ArrayExpress accession: E-MTAB-1436
Statistically significant differences were determined by using analysis of variance (ANOVA) and Tukey-Kramer for post hoc test. P <0.05 was considered significant. In most cases the physiological data are presented as change from baseline, with the baseline calculated as the mean of values at thirty minutes prior and immediately prior to administering endotoxin.