Background
Previous studies have shown that vitamin A deficiency (VAD) in rats is associated with a decrease in gas-exchange surface area, a decrease in the bronchial elastic fiber density, and with an increase in airway responsiveness to cholinergic agents [
1,
2]. Although VAD is uncommon in economically developed countries, it remains an important public health problem in the developing world particularly in children during the first seven years of life, when pulmonary alveolarization occurs [
3]. Vitamin A and its active metabolite retinoic acid influence alveolar development and restoration, however the mechanisms responsible for these effects remain poorly understood [
4,
5]. In our experimental model of VAD, rats do not become deficient until after the period of maximal alveolar formation, which is completed by 3 weeks of age [
2,
6]. During these first 3 weeks of postnatal life there is an increase in the mRNA for tropoelastin, the soluble precursor of cross-linked elastin, which is an important determinant of the mechanical properties of the lung parenchyma and airways [
7]. Once it is cross-linked, elastin normally undergoes very little turnover, although this does occur in pathological conditions such as emphysema [
6,
8].
In order to better identify the mechanisms that are responsible for airway hyperreactivity in VAD rats, with respect to morphological and biochemical characteristics of the pulmonary elastic fiber network, we evaluated the mechanical properties of the lung parenchyma that are most involved in regulating small airway diameter. Airway responsiveness to cholinergic agents is influenced by airway-parenchymal interactions [
9]. The elastic fibers in the walls of alveoli and alveolar ducts, which form a continuous network with elastic fibers in the small and larger airways, are an important structural determinant of these interactions [
10,
11]. The elastic fibers within the airway connect the epithelial basement membrane to the smooth muscle layer [
11]. Fibers in the adventitia that surrounds the airway smooth muscle are connected to parenchymal elastic fibers located in the surrounding alveoli and alveolar ducts. The contractile cells in the alveolar ducts may also influence airway smooth muscle contraction because contractile cells in the two locations are connected through the intervening elastic fiber network [
11]. Physiological measurements of the elastic modulus of the lung are sensitive to alterations in both the airways and the parenchyma [
12]. For an isotropic material, the ability to resist volume and shape distortion, respectively, is described by the bulk modulus (
k, which is proportional to the ability to resist uniform expansion) and the shear modulus (μ, which is proportional to the ability to resist a small isovolume shape distortion). The lung is more constrained in volume expansion than in shape distortion, and
k increases exponentially with volume whereas μ increases arithmetically [
13]. There are three mechanisms whereby the lung resists deformation: (a) altering the spacing between microstructural elements, (b) altering the orientation of the microstructural elements, and (c) stretching of the microstructural elements [
12]. Any or all of these three factors may be affected if there are abnormalities of the elastic fiber network. In pulmonary emphysema there are changes in all three mechanisms. Dilated alveoli and alveolar ducts increase the spacing between elastic fibers, elastic fibers are disarrayed and are abnormally connected, and the remaining alveolar walls and ducts are stretched by dilation. The elastic modulus of the lung parenchyma may also be altered in VAD rats, which have fewer and dilated gas exchange units compared to the lungs of VAS rats [
1]. Because the inhalation of aerosolized cholinergic agents distorts the lung parenchyma producing inter-dispersed regions of localized hyperinflation and atelectasis, one would predict that alterations in the elastic modulus would be accentuated after cholinergic administration [
14]. We hypothesized that because of parenchymal distortion and localized hyperinflation, cholinergic administration would produce a larger increase in the bulk modulus of VAD compared to vitamin A sufficient (VAS) rat lungs. To address this hypothesis we have characterized the effects of VAD on parenchymal mechanics and elastic fiber architecture. We have studied elastic fiber length per unit volume of lung, elastin production, and measured the elastic modulus of the lung parenchyma in VAS and VAD rats before and after the administration of CCh. We further hypothesized that if the elastic fiber network was a major determinant of the bulk and shear moduli, then restoration of the elastic fiber network may restore the elastic moduli to values that are similar to those in VAS rats. Therefore, we administered retinoic acid (RA) to determine whether reversing the tissue effects of VAD would coordinately reverse abnormalities in the elastic fibers and in the bulk and shear moduli. The elastic fiber length per unit volume was decreased in VAD rat lungs and may have contributed to the observed differences in shear modulus. However, other architectural modifications accounted for the observed differences in the bulk modulus in VAD compared to VAS rats.
Discussion
Our previous studies of rat lungs
in vivo have shown that the cholinergic-induced increase in total pulmonary elastance (which in this preparation is influenced by both the lung and the chest wall) is greater in VAD rats, and that RA-treatment restores the increase in elastance to a level, which is similar to that observed in VAS rats [
1]. Elastance increases as tissue stiffness increases. In the lung, elastance is increased (a) when lung volumes approach total lung capacity, (b) by atelectasis, or (c) by an increase in rigid structural components (such as collagen) or (d) by an increase in hysteresis, which could result from alterations in alveolar surface tension or disruption of the elastic fibers [
32,
33]. In order to more specifically examine the contribution of the lung parenchyma to the exaggerated CCh-mediated increase in total pulmonary elastance that was observed VAD rats, we ventilated the lung
ex vivo at a small tidal volume. This approach eliminated the contributions of the chest wall and of innervation and avoided the confounding effects of air-trapping that can be induced by large-volume oscillations. We found that the CCh-mediated increase in the elastic bulk modulus was exaggerated in VAD rats. This was manifest as an increase in the pressures required to expand the lung during inflation and a significant increase in hysteresis. In contrast, the CCh-mediated increase in shear modulus was diminished in VAD rats. Administration of RA for up to 21 days did not significantly reverse the effects of vitamin A deficiency on the bulk modulus, but there was a partial normalization of the shear modulus after 21 days of RA-treatment. The VAD-related alterations in the mechanical properties of the lung parenchyma were accompanied by a decrease in the concentration of parenchymal elastic fibers and in lung elastin. The administration of RA for 12 days increased TE mRNA but did not restore the 0.1 M NaOH-resistant lung elastin, although the concentration of parenchymal elastic fibers was increased. Therefore, decreases in lung parenchymal elastic fibers and total pulmonary elastin likely contribute to but do not completely account for to the exaggerated CCh-mediated increase in the elastance and bulk modulus in VAD rats.
Others have shown, using a qualitative pathologic grading system in rats, that VAD is associated with patchy atelectasis as well as emphysema [
33]. Our previous morphometric study, using lungs that were inflated to 20 cm H
2O confirmed the presence of emphysematous areas [
1]. Terminal airway closure that occurs after the administration of aerosolized cholinergic agents results in a non-uniform distribution of atelectatic and hyperexpanded areas of parenchyma, which could exaggerate the pre-existing abnormalities that are associated with VAD [
14]. The data in Table
1 are consistent with this statement, and show that both the SD Lm and SD ATI are increased in VAD relative to VAS lungs, following CCh administration. The exaggerated CCh-mediated increase that we observed in the bulk modulus reflects an increase in the elastance of the lung parenchyma of VAD rats. The deflation volume-pressure characteristics of the excised lung are also consistent with an increase in lung elastance in VAD. This differs from what one would expect in a uniformly emphysematous lung for which elastance would decrease. Furthermore, one might expect that the decrease in elastic fiber concentration (length per mm
3 lung parenchyma) and lung elastin that we observed in VAD rats would be accompanied by a decrease in lung elastance. Therefore, another anatomical abnormality must contribute to the exaggerated increase in parenchymal lung elastance after CCh-administration. It is likely that this abnormality involves localized areas of atelectasis and hyperinflation, which are exaggerated by CCh-administration (see Figure
11 and Table
1). From Figure
9 it is clear that higher pressures are required to expand VAD lungs, compared to VAS lungs, after cholinergic administration. This is particularly obvious at low lung volumes that are similar to those which were used to ventilate the excised lungs during the measurement of the bulk and shear moduli. An increase in surface tension in atelectatic regions is probably the major contributor to this increase in elastance and therefore the bulk modulus. These increased inflationary pressures in cholinergic-exposed VAD lungs resulted in an increase in the hysteresis of VAD compared to cholinergic-exposed VAS lungs (Figure
9).
The VAD-induced suppression of the CCh-mediated increase in the shear modulus requires an alternate explanation (Figure
4). Although the shear modulus increased as expected after CCh-administration in VAD lungs, the increase was less than in VAS lungs. The shear modulus reflects the ability of the lung parenchyma to resist distortion. As the lung is progressively inflated, the "struts" which surround the airspaces become more distended and rigid [
12]. This leads to a greater resistance to a distorting shear stress. CCh administration increases alveolar distortion resulting in hyperexpanded alveoli, which stiffens the lung and increases the shear modulus [
13]. The cholinergic-induced hyperexpansion and stiffening of the struts appears to occur more uniformly (SD Lm is lower) in VAS lungs, which are not restricted by pre-existing distortion from atelectasis and airspace enlargement, than in VAD lungs. In the areas of VAD lung where alveolar hyperexpansion occurs, there is less elastic tissue to resist shape distortion, which would result in a lower shear modulus. The decrease in elastic tissue likely contributes to the smaller CCh-induced increase in the shear modulus of VAD. The blunted CCh-induced increase in shear modulus in VAD rats likely contributes to their airway hyperresponsiveness, because the shear modulus is thought to be the most important characteristic that mediates airway-parenchymal interdependent opposition to airway contraction [
34].
We observed that VAD, which occurs after the major peak of pulmonary elastin synthesis has occurred, is accompanied by a decrease in lung parenchymal elastin. The loss of elastin in VAD was manifest as a decrease in the alveolar septal elastic fiber concentration (mm length per mm
3 lung parenchyma) and in the quantity of elastin that was resistant to digestion in the presence of 0.1 M NaOH at 98° (Figures
5 and
6, respectively). We also made the novel observation that RA stimulates elastin synthesis and the deposition of elastic fibers, which are important determinants of the mechanical properties of the parenchyma. Northern analysis demonstrated that 12 days of RA-administration increased the steady-state level of tropoelastin mRNA in the lung parenchyma, which is consistent with the restoration of elastic fiber concentration after 12 days of RA-treatment (Figure
7). However, we did not observe an increase in alkali-resistant elastin after 12 days of RA-administration. This may result from one or more of several factors. First, orcein can stain "immature" elastic fibers that contain a larger proportion of microfibrils than thicker fully cross-linked "mature" elastic fibers [
35]. Only the "mature" elastic fibers are resistant to the alkali treatment, which underestimates that quantity of newly formed, incompletely cross-linked elastin. Secondly, our morphometric analysis of elastic fiber concentration did not account for the thickness of the fibers, only the length per unit volume of lung. Therefore, thin newly formed elastic fibers would contain less elastin that could be detected by our biochemical analysis, but the fibers would be detected by our method for determining the concentration of elastic fibers, which is only dependent on the length of the fiber network and not on its thickness.
The airway contraction index after CCh-administration was lower in VAD rats relative to VAS controls. Administration of RA for 12 days was associated with a restoration of the contraction index (after CCh) of airways >0.55 mm diameter to a level that was similar to VAS rats. These data are consistent with our previous study, which demonstrated that 12 days of RA administration restored total (lung plus chest wall) pulmonary elastance and resistance [
1]. The data for the contraction index should be considered in light of our observation that RA-administration does not reverse the exaggerated CCh-mediated alterations in bulk and shear moduli in VAD rats. This consideration suggests that the RA-mediated correction of the increase in total pulmonary elastance that we previously observed in VAD rats with intact chest walls was primarily due to factors within the airways or chest wall rather than the lung parenchyma [
1]. If RA had corrected the lung parenchymal factors, then we would have expected to observe a correction in the VAD-related abnormalities in the parenchymal elastic modulus. These findings suggest that although VAD alters the elastic fiber system, alveolar architecture, and mechanical properties of the lung parenchyma; treatment with RA for 12 days corrects a VAD-related abnormality in the airway, rather than the parenchyma. When considered along with our prior observation that 12 days of RA treatment normalizes the expression of the muscarinic receptor-2, these findings suggest that the salutary effect of administering RA for 12 days is limited to the airways [
1].
Competing interests
The three authors declare that neither has a completing interest that would influence the objectivity of these findings.
Authors' contributions
SEM planned the experiments, performed the physiological measurements, wrote the manuscript, and performed some of the morphometry. AJT performed the studies of elastic fiber concentration, lung elastin contents, airway contraction index and assisted with the preparation of the manuscript. AJH performed the Northern analyses and made substantive contributions to the writing of the manuscript.