METHODS

Fragments of C5f were purified from yeast-activated rabbit serum, the terminal arginine residue removed by incubation with carboxypeptidase as previously described (11). New Zealand white rabbits weighing 2.5 to 3.5 kg were anesthetized intramuscularly with xylazine (2 to 3 mg/kg) and ketamine (30 to 35 mg/kg). Using an Olympus BF-7 pediatric bronchoscope (Olympus Corp. of America, New Hyde Park, NY), a polyethylene catheter (PE 50) was placed under direct vision into the right cranial lobar bronchus, and 150 µl of C5f (vide infra) dissolved in 850 µl phosphatebuffered saline (PBS) were instilled into the dorsal branch. The rabbit was immediately placed in the right decubitus position for 10 min to limit the distribution of the C5f to the dorsal portion of the right cranial lobe. Prior studies using indocyanine green and radiopaque dye have shown that this is feasible and leads to an inflammatory response that extends throughout the dorsal portion of the right cranial lobe (11, 17). Four rabbits were killed at specified time points (1, 30, 60, 120, and 240 min) using pentobarbital overdose (20 animals in total). The thorax was opened by a median sternotomy, the heart was tied off to prevent redistribution of blood, and the heart and lungs were removed en bloc. The right (experimental) and left (control) cranial lobes were selectively lavaged with 10 ml of PBS containing 10 U of heparin per ml. Lavage fluid (10 ml per lavage) was reinfused into the lobe an additional three times to harvest the maximal number of airway and airspace inflammatory cells. Subsequently, the volume of lavage recovered was measured using a volumetric pipette (usually> 80% of the volume infused), total cell count was estimated by a Neubauer hemocytometer, and a differential count was completed using a cytospin and Diff-Quik stain (American Scientific Products, McGaw Park, IL). At least 200 cells were counted for each sample. After lavage, lungs were fixed by the intratracheal instillation of a glutaraldehyde (1%)-paraformaldehyde (1%) fixative in 0.2 M cacodylate buffer (400 mosmol) at 25 cm of fixative pressure. After 8 to 12 h, whole lung and individual lobe volumes were estimated by weight displacement (18). The right and left cranial lung lobes were used for selection of specific airway generations using an airway dissection technique (19). Five bronchi at generations 7, 10, and 13, their subadjacent bronchioles and surrounding parenchyma were selected from the right cranial lung lobe (figure 1). Corresponding bronchi were selected from the left cranial lobe and served as an internal control. Each airway and its surrounding parenchyma were cut to a block face of 5 by 10 mm, postfixed in osmium tetroxide, dehydrated in a graded series of ethanol and propylene oxide, and embedded in Aralite 502. After 8h at 60 C, blocks were cut 1 µm thick using a Sorvall JB4 microtome (Sorvall, Newtown, CT), and the sections were mounted on glass slides and stained with hematoxylin-eosin and azure 2. Stratified random sampling was used to select 10 parenchymal fields per slide at x400 (20). For parenchymal tissue and neutrophil volume per surface area of interalveolar septa (VSPMN, IAS) was estimated using point count and linear intersection methods and a 100 point lattice grid (21) as follows: