Why babies don't drown at birth?

The paper of Aslan et al. (1) in this issue, demonstrating a link between certain β-adrenoceptor polymorphisms and the development of transient tachypnoea of the newborn, is the latest to address the question of the physiological mechanisms controlling liquid movement across the respiratory epithelium at birth – a question that has long exercised respiratory physiologists (see reference 2 for a recent review of the developmental regulation of lung fluid transport). The abrupt removal of the liquid present within the lung at birth is perhaps the most dramatic demonstration of the importance of lung transport mechanisms that operate throughout life, and it is consequently not surprising that paediatricians (including paediatric pathologists) and obstetricians have been at the forefront of attempts to understand the key physiological mechanisms underlying these processes. Knowledge of the fluid absorptive capacity of the lung is more than 150 years old (3) but until the late 1970s it was believed that the removal of lung liquid at birth was a passive process due to the compression of the chest during its passage down the birth canal. However, there are several lines of evidence which argue against this simple explanation (see 2), not the least of which is the observation that occlusion of the trachea in experimental animals does not impede lung liquid absorption. It had already been established that lung liquid is secreted by a process involving ‘active’ chloride transport (see 2) when Enhorning and co-workers (4) and Walters and Olver (5) demonstrated the key role of β-adrenoceptor stimulation in the rapid switch in the direction of net liquid movement across the lung epithelium at birth. Enhorning et al. (4) noted the dehydrating effect of the synthetic β-adrenoceptor agonist, isoxuprine, on the newborn lung in a study principally focussed on the control of the release of surfactant, while Walters and Olver (5) went several steps further by directly measuring the rate of absorption of liquid from the lumen of the foetal lung and demonstrated that the effect was mediated via β-adrenoceptors by comparing the relative potencies of naturally occurring catecholamines and the inhibitory action of propranolol. They observed that very rapid removal of foetal lung liquid could be accomplished by modest adrenaline dosage but that the sensitivity of the lung to β-adrenoceptor agonists was critically dependent on gestational age. Studies performed at about the same time by Lagercrantz and Bistoletti (6) on human foetuses undergoing labour showed that there was a marked rise in blood catecholamines as labour proceeded and that the blood concentrations achieved were higher than during any other physiological event, and even higher than in a patient with a catecholamine secreting tumour. They further demonstrated high cord blood catecholamine concentrations in babies requiring instrumental delivery, both emergency caesarean section (usually performed after a period of labour or foetal stress) and elective caesarean section. However, since the mother is usually not in labour prior to an elective caesarean section, the foetus is exposed to these high catecholamine concentrations for only a brief period of time. These findings help explain the higher risk of TTN after elective caesarean section and are consistent with the studies of Bland and co-workers (7) which showed that extravascular lung water was lowest in newborn rabbit pups born vaginally, or abdominally after a period of labour, compared to abdominal delivery without labour. The relevance of the foetal studies described above to the changes in lung liquid movement at birth was underscored by Brown et al. (see 2) who showed, using dose response curves to adrenaline, that all the foetal lung liquid absorption that takes place during labour and birth in sheep could be accounted for by the natural rise in catecholamines which occurs in the foetus and newborn during the birth process. That β-adrenoceptor stimulation leads to activation of amiloride-sensitive Na+ channels and the consequent absorption of sodium (and chloride) ions and thus of liquid has been well documented (e.g. 8,9). Although the identity of the Na+ channel mediating the Na+ absorptive response is a matter of dispute (2), Hummler et al. (10) have demonstrated the critical importance of ENaC to respiratory adaption at birth by showing that knockout of the α-subunit (but not the β- or γ-subunits) leads to death due to respiratory distress within 48 h of birth. Notwithstanding the findings of Aslan et al. (1), it should be noted that the development of the foetal lung liquid absorptive mechanism in the latter part of gestation is not limited by the expression of β-adrenoceptors. An analogue of cyclic AMP which mimics β-adrenoceptor stimulation follows the same developmental curve of increasing effect during gestation as does adrenaline (11), indicating that the rate limiting step in development is downstream of the generation of intracellular cAMP. As shown by Barker et al. (e.g. 12) and others (2,13), the developmental appearance of the absorptive mechanism requires the presence of cortisol and thyroid hormone in vivo and a number of studies (see 2) have shown that these hormones are a prerequisite for the development of a normal physiological phenotype in cells grown in culture, an effect which is very likely mediated by ENaC expression or trafficking. The paper by Aslan et al. (1) on the association of certain polymorphisms of the β-adrenoreceptor with TTN adds a further level of detail and complexity to the story. As far as we are aware there are no naturally occurring lethal conditions or diseases in which there is a defect in the foetal lung liquid absorptive mechanisms, which suggests that the process of foetal lung liquid clearance has been the subject of intense evolutionary pressure (it should be remembered that all mammals, all birds and many amphibian species have to take a ‘first breath’ and clear their lungs of liquid). This may explain why nature has developed a ‘Belt and Braces’ (i.e. built in redundancy) approach to a situation which is critical for survival, evolving additional physiological mechanisms which allow or promote lung liquid absorption in parallel to the β-adrenoceptor-mediated system described. These include the transient increase in pulmonary epithelial permeability at the onset of breathing demonstrated by Egan et al. (2); a change in the hydrostatic gradient across the epithelium favouring liquid absorption once an air–liquid interface has been established (e.g. Misserocchi, 14); the stimulatory effect of AVP on lung liquid absorption (15); the regulation of conductive Na+ transport by arachidonic acid and G proteins (16); a direct effect of oxygen on epithelial permeability (17), ENaC (2,13,18,19) and on Na/K ATPase (20) and the mysterious ‘stretch factor’ described by Nelson and Perks (2) which produces an amiloride insensitive (i.e. non-ENaC) absorption. There is evidence that β-adrenoceptor stimulation induces increased lung liquid absorption in the postnatal lung when excess liquid is present as in experimental pulmonary oedema (21,22). How polymorphisms such as those described by Aslan et al. (1) may affect the β-adrenoceptor-dependent system in the postnatal lung is yet to be determined.