A role for growth hormone (GH) in pulmonary function is indicated by the physiological and anatomical changes in the lung in pathophysiological states of GH excess and deficiency. For instance, large lungs (Bar-tlett, 1971), upper airflow obstruction (Trotman-Dickenson et al., 1991), and narrowing of the small airways (Harrison et al., 1978) accompany GH excess, whereas a decrease in muscle strength and a reduction in the maximum inspiratory and expiratory pressure (Merola et al., 1995, 1996) are associated with GH deficiency. The possibility that the lung is a target site for GH action is also indicated by the GH-induced production of superoxide by alveolar macrophages (Edwards et al., 1992), the GH-induced increase in circulating lung neutrophil activation during sepsis, and the accompanying increase in microvascular injury (Liu et al., 2002a). Exogenous GH also induces NFkB activation in the lung (Liu et al., 2002b), increases phosphorylase A activity (Jost et al., 1979), and stimulates the tyrosine phosphorylation of specific proteins in lung epithelial cells (Batchelor et al., 1998). The GH receptor (GHR) gene is also expressed in pulmonary tissues (e.g., Tiong et al., 1989; Garcia-Aragon et al., 1992; Batchelor et al., 1998). Pituitary GH, thus, is likely to be an endocrine regulator of lung growtj and function in juvenile developmen and in adulthood. It is, however, un likely to be a regulator of pulmonar; function in the early fetus.
In many species, lung growth anc differentiation occurs during ontog eny before the appearance of the pi tuitary gland and the differentiation of GH-secreting somatotrophs. Foi instance, lung buds appear at embryonic day (ED) 10 in fetal rats (with a 21-day gestation period; Young et al., 2002) and at ED3.5 (with a 21-day incubation period) in developing chicks (Sakiyama et al., 2000), both of which occur before the presence of pituitary somatotrophs at ED19 in rats (Chatelain et al., 1979; Frawley et al., 1985; Hemming et al., 1986) and at ED16 in chicks (Harvey et al., 1998). The expression of GH receptors in pulmonary tissues also occurs before the appearance of GH in peripheral plasma at ED19 (Strosser and Mialhe, 1975) in rats and at ED17 in chicks (Harvey et al., 1998). Early fetal development, therefore, is thought to be agrowth'Without-GH-syndrome (Geffner, 1996), independent of pituitary GH. The abundant and widespread production of GH in peripheral tissues of early embryonic chicks (Harvey et al., 1998) and mice (Pantaleon et al., 1997) nevertheless suggests GH involvement in early development through autocrine or paracrine mechanisms (Harvey and Hull, 1997). The presence of GH, GHR, and a GH-specific response gene (GH response gene-1) in the lungs of early chick embryos (Harvey et al., 2000, 2001) supports this possibility, although the presence of GH in extrapituitary organs of the mammalian fetus is poorly documented. GH, however, may act as a local growth or differentiation factor in the mammalian lung, because GH mRNA, detected by reverse transcriptase-polymerase chain reaction (RT-PCR), is present in the alveolar macrophages of adults (Allen et al., 2000) and trace amounts of GH immunore-activity are present in whole lung extracts of fetal and adult lungs (Kyle et al., 1981; Costa et al., 1993). The possibility that GH production occurs in the rat lung during perinatal and neonatal periods of lung development, therefore, has been determoned in the present study.
DISCUSSION
These results clearly demonstrate, for the first time, GH gene expression in the lungs of perinatal and neonatal rats. The GH mRNA in the fetal and ¦ neonatal rat lung was identical to that expressed in the adult pituitary gland of these rats. The GH transcript in immune (Rohn and Weigent, 1995) and neural (Baudet et al., 2003) tissues of the rat is also homologous to pituitary GH mRNA. In contrast, tissue-specific GH transcripts are present in central (neural retina) and peripheral (heart) tissues of perinatal chickens that differ from pituitary GH mRNA (Takeuchi et al., 2001). The GH mRNA sequence in the pituitary glands and lungs of our (Sprague-Dawley) rats differed from the published sequence for the Norway rat (Seeburg et al., 1977), but this finding may reflect the presence of polymorphisms in the GH gene (e.g., Malveiro et al. 2001; Sorensen et al., 2002).
GH gene expression has been demonstrated previously in adult human lungs (Allen et al., 2000), although this expression was only in isolated activated alveolar macrophages and was only characterized by RT-PCR. The present study is, therefore, the first to localize GH mRNA within lung epithelia, smooth muscle, mesenchyme, and type I and II cells.
The presence of GH mRNA in the lung provides strong evidence for its synthesis within this tissue during development.
The presence of GH-re-leasing hormone (GHRH) (Shibasaki et al., 1984; Allen et al., 2000) and ghrelin (Gnanapavan et al., 2002; Vol-ante et al., 2002) in the lung may indicate the involvement of these GH secretagogues in the expression of the lung GH gene, as in the pituitary gland.
Within the lung, GH mRNA was present from ED15, before its ontogenetic appearance in the pituitary gland at ED19 (Chatelain et al., 1979: Frawley et al., 1985; Hemming et al., 1986). The widespread presence of GH in the ED 17 lung, thus, is unlikely to be of pituitary origin, especially as the circulating GH concentration is not detectable at this age (Strosser and Miahle, 1975). The abundance and localization of GH immunoreactivity in the mesenchymal, epithelial, and smooth muscle cells of the fetal lung is also unlikely to reflect its sequestration from GH-producing macrophages or monocytes in the alveoli or airways of the developing lung (Allen et al., 2000). Although GH can pass from bronchial airways into the lung inter-stitium and retain its biological activity (Patton et al, 1989), the GH immunoreactivity in the interstitial cells of the perinatal and neonatal lung is far more abundant than in the cells of the bronchial associated lymphoid tissue inside the airway lumen. Indeed, the GH immunoreactivity in pulmonary macrophages reflect its uptake and degradation (Patton et al., 1989), suggesting GH release from lung airway epithelial cells might contribute to the GH found in airway macrophages. Alveolar macrophages, therefore, are unlikely to be the source of the GH immunoreactivity in the perinatal and neonatal lung.
Our finding of GH immunoreactivity in the lung is in agreement with earlier studies that measured radio-immunoassayable GH in whole-lung extracts in the adult lungs (Kyle et al., 1981) and in the fetal lungs (Costa et al., 1993) of humans. In the latter study, the concentration of immunore-active GH in the fetal lung was >fivefold higher than in adults, suggesting its presence was ontogenetically regulated and possibly linked to lung growth and development.
It is now well established that the lung is a target site for GH action. Indeed, radioligand binding sites for GH have been demonstrated in adult and fetal rabbit lungs (Amit et al., 1987; Labbe et al., 1992). Northern blotting .has also demonstrated the presence of GHR mRNA in the lungs of fetal rats (Walker et al., 1992) and rabbits *(Tiong et al., 1989). GHR mRNA is also expressed in the human lung from the first trimester of gestation (Zogopoulos et al., 1996). Immu-nocytochemistry and in situ hybridization (ISH) have also localized GHR and GHR mRNA within fetal rat lung epithelial cells (Garcia-Aragon et al., 1992; Edmondson et al, 1995) and GHRs have also been identified by Western blotting in fetal (Walker et al., 1992) and adult (Frick et al., 1998) rat lungs. Batchelor et al. (1998) also found that the GHR mRNA was present in the rat lung from ED 16-ED21, corresponding to part of the pseudoglandular, the canalicular, and part of the saccular stages of rat lung development, and found that it was 50% more abundant than in the liver, which is a recognized GH target site (Baumbach et al., 1989). Moreover, these authors showed that GH stimulation of the receptor-induced tyrosine kinase activity, indicating that the fetal rat lung GHR is functional. Shoba et al. (1999) similarly correlated GHR abundance in the rat lung with the activity of proteins involved in GHR signaling (STAT-1, -3, -5, and JAK 2), and Lu et al. (2001) found that a GH-regulated gene (Grtpl) was also present in the rat lung. The presence of GH and GHR in the developing lungs of perinatal and neonatal rats, therefore, suggests GH actions in lung development or in pulmonary function.
Actions of exogenous GH in promoting lung development have been shown in neonatal rats (Dubreuil and Morisset, 1986) and hypophysecto-mized mice (Sondergaard et al., 2003). Pathological pituitary GH excess in acromegaly is similarly correlated with lung hypertrophy (Trotman-Dickenson et al., 1991), particularly in alveolar size and alveolar surface area (Donnelly et al., 1995), whereas a decrease in lung size is a characteristic of pituitary GH deficiency (De Troyer et al., 1980). Exogenous GH also increases respiratory muscle strength (Felbinger et al., 1999) in GH-defi-cient patients. Other actions of GH within the rat lung include an increase in the activity of pulmonary guanylate cyclase (Vesely, 1981), increased activities of the antioxidant glutathione (Youn et al., 1998), and a decrease in lung lipid peroxidation (Weigent et al., 1992). Whereas roles for lung GH in lung development are currently unknown, it may participate in its vascularization, as GH stimulates angiogenesis in other tissues (Struman et al., 1999; Corbacho et al, 2002). It may also participate in cellular differentiation (Sanders and Harvey, 2004) or regulate immune function within the lung (Batchelor et al., 1998; Waters et al., 1999; Allen et al., 2000).
The abundance and widespread distribution of GH and GH mRNA in the perinatal and neonatal lung suggests GH action during development is of physiological significance. Moreover, as GH expression occurs in the lung before its appearance in the pituitary gland at ED19 (Chatelain et al., 1979; Frawley et al., 1985; Hemming et al.. 1986) and the presence of GH in systemic circulation at ED 19 (Strosser and Mialhe, 1975), actions of GH in the developing lung are likely to result from local autocrine or paracrine actions, at least in the late pseudoglandular stage. Local actions of GH within immune (van Garderen et al., 1997), mammary (Zhang et al., 1997), and orthodontic tissues (Mertani et al., 2001), and GH-expressing cell lines (e.g., Kaulsay et al., 1999) are now well established, and a similar local mechanism may be operative in the lung during development. Such actions, however, may be indirect and mediated by an array of growth factor mediators (Sanders and Harvey, 2004), including insulin-like growth factors (IGFs).
The possibility that actions of GH within the developing lung might be mediated by an IGF-I-dependent mechanism is supported by the distribution of IGF-I mRNA in the fetal rat lung, because it is comparable to our findings of GH and GH mRNA localization. IGF-I mRNA is abundantly localized within fetal mesenchymal lung cells, especially those surrounding airway epithelium (Retsch-Bogart et al., 1996), although it is also present in airway epithelial cells (Wallen and Han, 1994). As the expression of IGF-I in the developing lung was correlated with cellular proliferation and the maturation of connective tissue (Lallelmand et al., 1995), it is possible that these actions reflect the upstream expression of GH in the same tissues and cells. The increased production of IGF-I in the lungs of GH-treated rats (D'Ercole et al., 1984) supports this possibility.
In summary, these results demonstrate the expression of GH mRNA, and GH-immunoreactivity, within the lungs of perinatal and neonatal rats and suggest autocrine/paracrine actions of pulmonary GH are involved in the pseudoglandular through the alveolarization stages of development of this tissue.
