Isocapnic dry gas hyperventilation provokes hyperpnea-induced bronchoconstriction in guinea pigs by releasing tachykinins from airway sensory C-fiber neurons. It is unknown whether dry gas hyperpnea directly stimulates C-fibers to release tachykinins, or whether this physical stimulus initiates a mediator cascade that indirectly stimulates C-fiber tachykinin release. We tested the hypotheses that mucosal hypothermia and/or hyperosmolarity--physical consequences of airway heat and water loss imposed by dry gas hyperpnea--can directly stimulate C-fiber tachykinin release. Neurons isolated from neonatal rat dorsal root ganglia were maintained in primary culture for 1 wk. Cells were then exposed for 30 min at 37 degrees C to graded concentrations of NaCl, mannitol, sucrose, or glycerol (0-600 mOsm) added to isotonic medium, or to isotonic medium at 25 degrees C without or with 462 mOsm mannitol added. Fractional release of substance P (SP) was calculated from supernatant and intracellular SP contents following exposure. Hyperosmolar solutions containing excess NaCl, mannitol, or sucrose all increased fractional SP release equivalently, in an osmolarity-dependent fashion. In marked contrast, hypothermia had no effect on fractional SP release under isotonic or hypertonic conditions. Thus, hyperosmolarity, but not hypothermia, can directly stimulate tachykinin release from cultured rat sensory C-fibers. The lack of effect of glycerol, a solute which quickly crosses cell membranes, suggests that neuronal volume change represents the physical stimulus transduced by C-fibers during hyperosmolar exposure.