In-vivo heparin treatments
Two groups of specific-pathogen-free (SPF) rats were used in this study. The first group, analyzed as preliminary data for the subsequent study, consisted of Fischer rats (n = 9) (Charles River, Wilmington, MA) in 3 groups of 3 animals each: untreated controls, heparin-treated and sacrificed after 24 hours, and heparin-treated and sacrificed after 96 hours.
The second group consisted of Fischer rats (n = 15)(Harlan, Indianapolis, IN) in 5 subgroups of 3 animals each: untreated, saline/vehicle control 24 hours, heparin-treated 24 hours, saline/vehicle control 96 hours, and heparin-treated 96 hours. One untreated control (baseline/reference) was sacrificed at 0, 24, and 96 hours. Vehicle controls and heparin-treated animals were sacrificed at 24 and 96 hours.
Rats weighing between 250 and 285 grams were given intraperitoneally a mixture of 30 mg Ketaset (Fort Dodge, Fort Dodge, IA) and 6 mg X-ject (Burns Veterinary Supply, Inc., Westbury, NY) to induce anesthesia 15 min prior to intratracheal administration of a single aerosolized dose (250 μl) of normal saline or of a 2 mg/ml solution of high molecular weight heparin, 13,500–15,000 MW, (Calbiochem, San Diego, CA) using a MicroSprayer intratracheal aerosolizer (Penn-Century, Philadelphia, PA) to optimize the distribution of heparin throughout the lung. According to the manufacturer, this generates 25–30 μm sized particles, which should be capable of reaching first order (primary) bronchioles. At 24 hours or 96 hours post-treatment, the experimental and control animals were euthanized with an overdose of sodium pentobarbital. The lungs were perfused with saline until white to remove blood and minimize albumin interference in later procedures, then surgically removed from the pulmonary cavity. Pulmonary airways (visible bronchi) were separated by blunt dissection from the parenchyma (small airways and alveolar regions) and the latter flash frozen in liquid nitrogen and stored at -80°C for later analysis. In some cases, the apical lobe of the right lung was tied off and insufflated with 2% paraformaldehyde in 0.1 M phosphate buffered saline for 1 hour. Sections were dehydrated with graded alcohols and infiltrated with xylenes followed by paraffin. Paraffin embedded tissue blocks were cut into 7 μm sections on a standard microtome and prepared for routine histology or immunostaining. Rabbit anti-annexin V (FL-319; sc-8300, Santa Cruz Biotech, Santa Cruz, CA) was applied at 1:100 dilutions overnight and the antigen-antibody complex visualized with a VectaStain Elite Peroxidase kit (Vector Laboratories, Burlingame, CA).
Protein sample preparation
Frozen lung parenchyma was pulverized using a specialized mortar and pestle (Fisher Scientific, Suwanee, GA) under liquid nitrogen. Equal amounts (250 mg) of powdered lung were suspended in a urea-containing buffer containing 20 μg/ml RNase-A, 100 μg/ml DNase-1 inhibitor (Invitrogen, Carlsbad, CA), 1% ampholytes (pH 3–10), and 50 μM dithiothreitol (BioRad, LaJolla, CA). Protein concentrations were measured by chemiluminescence using Protein EZ-Quant (Invitrogen), with 100 μg total lung lysate loaded per gel. All samples were diluted in 6.25 M urea and 2.3 M thiourea buffer to a concentration of 1 mg/ml and stored frozen at -80°C. First-dimensional Readystrips, pH 5–8, 11 cm (BioRad) were used for all samples based on preliminary data that suggested the majority of dynamically regulated proteins focused between pH 5 and pH 8. Samples were applied to strips and equilibrated for 12 hours at 50 V on a Protean X IEF focuser (BioRad). Samples were isoelectrically focused at 250 V for 15 minutes. Voltage was increased by two 500 V ramp increases until achieving 1500 V, held at 1500 V for 2 hours, and then increased in rapid steps to 8000 V until complete isoelectric focus was achieved at 55 kVH.
Second-dimension electrophoresis and staining
Bis-Acrylamide gels (12.5%) with stackers (4%) were hand-cast in a 16 cm × 18 cm gel slab Multicasting System (Hoefer Scientific, Cambridge, MA). Gels were allowed to polymerize overnight with n-butanol overlays. First-dimension electrofocused strips were alkylated in 0.25% iodoacetamide (BioRad) for 20 minutes on an orbital shaker, rinsed with 1× Tris-Glycine-SDS running buffer, and then carefully laid onto second-dimension gels. A layer of 0.5% agarose containing bromophenol blue was poured to keep strips in place. Multiple gels were run simultaneously (to minimize group run variation) at 200 V for 6 hours, rinsed for 1 minute in 250 ml of ultra-pure water twice, and fixed overnight in 10% methanol, 5% glacial acetic acid on a horizontal shaker. Gels were incubated in Sypro-Ruby stain (Molecular Probes, Carlsbad, CA) in the dark for 16 hours.
Image capture and analysis
Gel images were captured with an Epi Chemi UVP scanner supported by a 16 bit CCD camera for enhanced high resolution images (BioImaging Systems, Upland, CA). Aperture width and exposure time were adjusted so that only the most abundant spots on the gels reached saturation (pixel intensity was determined by the software). In addition, gels were normalized based on total protein fluorescence per gel as well as reference spot fluorescence comparison. The same settings were used for all gels. Acquired gel images were analyzed using PDQuest software (BioRad), transformed to adsorption images (dark spots against light background), cropped, and size-adjusted across all images. The effects of occasional streaking were minimized by during the analysis with the application of Guassian equations to the raw images with the PDQuest software. Spot detection parameters were optimized on a single gel representative of each individual sample, then normalized among the sub-group, and finally normalized across groups. Areas of interest were assigned to the program for further magnification. Images in each set were matched and a representative map was generated for each individual and subsequently for each sub-group per time point. Correlation coefficients of variance were calculated for each spot within sub-groups and compared with the mean correlation coefficient of variance in the software program, which then assigned significance values to each spot. Only spots with optical densities that were significantly changed by heparin (based on the Mann-Whitney statistical technique) were further analyzed for the purposes of this study.
Spot selection, digestion, and protein analysis
After selection of the most dynamically-regulated spots, gel plugs were manually excised from replicate gels (n = 3), placed in a 96-well plate, trypsin-digested overnight on an automated digester, and placed on zip-plates (Millipore Bioscience, Bedford, MA) to remove any salts that could adversely affect results. Bovine albumin was added to the Maldi plate as a positive control for MS detection. Replicate samples and control proteins were resuspended in 5 μl of 30% acetonitrile/0.1% trifluoroacetic acid. One μl of each sample was spotted twice, allowing the spot to air dry completely each time. Matrix (alpha-cyano-4-hydroxycinnamic acid, Agilent Technologies, Palo Alto, CA) (0.5 μl) was overlaid on each spot. An ABI-4700 Proteomics Analyzer MALDI-time of flight/time of flight (TOF/TOF) mass spectrometer (Applied Biosystems, Foster City, CA) was used for the analysis. The instrument was set to acquire an initial MS spectrum followed by up to 15 MS/MS spectra for each spot. The instrument was internally calibrated using a mixture of 6 peptides. A default "plate" calibration was performed based upon 6 calibration spots distributed on the plate, resulting in accuracies better than 100 ppm. For the analysis of sample spots, the instrument attempted to perform an internal calibration for the individual spots using the tryptic autodigest fragments, which resulted in accuracies better than 20 ppm. If the instrument was unable to perform the internal calibration, the default "plate" calibration was applied to the sample spot. The MS and MS/MS data were processed by GPS Explorer, version 2.0 (Applied Biosystems), and submitted to our in-house MASCOT search engine (Matrix Science). The NCBI non-redundant database was searched with tolerances set to 150 ppm for the MS spectra and 0.5 kDa for the MS/MS spectra. GPS Explorer then processed the search results from MASCOT.