1. The ability of airway epithelial cells to produce insulin-like growth factor I may be important in the pathogenesis of subepithelial fibrosis observed in the airways of patients with asthma. We determined whether human airway epithelial cells are capable of producing polypeptide mediators that could induce fibroblast proliferative activity, in particular insulin-like growth factor I. 2. We examined 12 primary cultures of human airway epithelial cells grown to confluence on collagen gel-coated dishes. Using a colorimetric assay based on the uptake and subsequent release of Methylene Blue, increased proliferation of human fetal lung fibroblasts was detected in conditioned media from airway epithelial cells. The median stimulation of fibroblast proliferation was 49.9% (range 25.6 −113.3%) above control values (observed at 1:2 dilution of media). 3. A neutralizing antiserum to insulin-like growth factor I partly inhibited fibroblast proliferation induced by epithelial cell conditioned media by 52.2% (49.9–109%; n = 5). 4. Radioimmunoassay for insulin-like growth factor I in conditioned media demonstrated a median concentration of 54.1 ng/ml (32.4–96.8 ng/ml). 5. Insulin-like growth factor I mRNA was detected in epithelial cell monolayers by Northern blot analysis using an insulin-like growth factor I cDNA probe. 6. The insulin-like growth factor I gene is expressed in cultured human airway epithelial cells, which also secrete insulin-like growth factor I protein. Insulinlike growth factor I also accounts for the major mitogenic activity for fibroblasts of cultured human epithelial cell conditioned media. Insulin-like growth factor I may function in a paracrine manner to modulate fibroblast behaviour and may be involved in airway processes, such as those occurring in asthma.

1. Airway oedema resulting from increased microvascular permeability is a characteristic pathological finding of asthma. The regional effects of putative mediators involved in asthma on airway microvascular permeability have been studied. 2. The effects of histamine, leukotriene (LT) D 4 and platelet-activating factor (PAF) on microvascular permeability in the nasal mucosa, larynx, trachea, main bronchi and intrapulmonary airways of the guinea pig were assessed by measuring the extravasation of intravenously administered Evans Blue dye. 3. PAF and LTD 4 caused increased microvascular leakage throughout the respiratory tract, although their effects were maximal in different regions. Histamine had no significant effect on intrapulmonary airways. PAF was more potent than LTD 4 and histamine at all airway levels. For example, in the trachea the doses required to cause leakage of 50% of maximal (ED 50 ) were 10.4 nmol/kg, 138 nmol/kg and 11.2 μmol/kg, respectively, for PAF, LTD 4 and histamine. 4. The effect of the three mediators was maximal 5 min after intravenous administration. Histamine, but neither LTD 4 nor PAF, still caused significant leakage 30 min after administration. 5. The increased microvascular leakage induced by the mediators was inhibited by their respective specific receptor antagonists, suggesting that the effect was mediated via specific receptors. 6. Histamine, LTD 4 and PAF have varying potencies in increasing microvascular permeability in the guinea-pig respiratory tract, exert their maximal effect in different regions and have varying durations of action.

1. The bronchial response of 11 normal and ten stable asthmatic subjects to increasing concentrations of methacholine aerosol was assessed by serial measurements of specific airways conductance (scaw) in a body plethysmograph. 2. Cumulative log dose-response curves were constructed. The threshold provocative dose of methacholine needed to cause a 35% fall in starting sCaw (pD35) and the steepest slope of the response were measured from each curve. 3. On separate days subjects were premedicated with 0.9% NaCl solution (control) in duplicate, chlorpheniramine, salbutamol and atropine, the last-named at two different doses, one twice the other. 4. Asthmatic subjects had a lower mean PD35 and a lower mean slope than normal subjects. 5. Pretreatment with salbutamol resulted in a greater increase in sGaw than after atropine but caused a smaller increase in PD35 in both groups. There was a dose-dependent increase in PD3s after the two doses of atropine, but no significant difference in bronchodilatation between doses. Mean steepest slope approximately doubled in these three sets of challenges. 6. Chlorpheniramine caused a small degree of bronchodilatation and there was a non-significant increase in mean PD3s and in mean steepest slope in both normal and asthmatic groups. 7. There was a positive linear correlation between starting sGaw and steepest slope in each group of premedicated challenges, such that when

1. In spontaneously breathing dogs ( n = 8), maintained in the upright position, bronchial luminal diameter was measured from tantalum bronchograms. Changes in total intrathoracic lung volume were measured from radiographs taken at functional residual capacity (FRC R ). 2. With cooling of the cervical vagi to 0–4°C mean bronchial diameter increased to 117 ±15 (sd)% of baseline diameter and FRC R increased to 113 ± 16 of baseline volumes respectively. There was a significant correlation between changes in FRC R and bronchial diameter. 3. After the vagi were rewarmed, pulmonary oedema was induced by rapid intravenous infusion of Hartmann's solution. Bronchi narrowed to a mean of 86 ± 9% of baseline calibre and FRC R decreased (mean = 95 ± 15%). 4. With vagotomy bronchial diameter increased (mean = 102 ± 12% of baseline diameter) but FRC R did not change significantly. 5. Excluding one dog with gross oedema, changes in bronchial diameter due to oedema correlated positively with changes in FRC R ; after vagotomy the relationship between diameter and FRC R was similar to that seen after vagal cooling in the baseline state, though individual values were lower. 6. The shift in the diameter-FRC R relationship with vagotomy demonstrates that the direct bronchoconstrictor effect of the vagus is approximately doubled in the presence of pulmonary oedema. 7. Bronchial and peribronchial oedema does not appear to directly narrow the bronchial lumen.