Abstract
A model of inducible expansion of the gas exchange area in adult mice would be ideal for the investigation of molecular determinants of airspace regeneration in vivo. Therefore, the post-pneumonectomy (post-PNX) compensatory lung growth in adult C57BL/6 mice was characterised in this study.
Mice underwent left-sided PNX. Right lung volume was assessed on days 1, 3, 5, 7, 10 and 21 after PNX, and total DNA and cellular proliferation of the right lung were determined. Lung histology was studied using immunohistochemistry and quantitatively characterised by detailed stereological investigations. Pulmonary function was assessed using a mouse body-plethysmograph.
Following PNX, right-lung volume rapidly restored the initial volume of left and right lung. Total DNA increased significantly over 21 days and equalled the total DNA amount of both lungs in the control mice. Septal cell proliferation significantly increased after PNX, and included endothelial cells, epithelial cells, smooth muscle cells and fibroblasts. Stereological investigations of left and right control lungs versus right lungs 21 days after PNX indicated complete restoration of body mass-specific alveolar surface area. Pulmonary function testing showed marked alteration at 3 days and normalisation at 21 days post-PNX.
In conclusion, well reproducible reconstitution of alveolar gas-exchange surface based on septal tissue expansion may be provoked by pneumonectomy in adult mice.
This study was partially funded by the German Research Foundation.
Several pulmonary diseases originate from or are deteriorated by the loss of alveolar septae or, alternatively, remodelling processes of the septal walls, which result in severely compromised gas exchange within the alveoli. The principal ability of mammals to completely restore lung function after major losses of lung tissue by compensatory development of additional gas-exchange surface areas provides a rationale for identifying intrinsic regenerative programmes of the lung that may be employed for therapeutic purposes.
The molecular basis of alveolar generation and alveolar remodelling is presently not well understood 1. Furthermore, the cellular components which contribute to repair and remodelling of pulmonary tissue still await ultimate elucidation. Solutions to the problems concerning repair and regeneration of lung tissue for restoration of functional alveoli are at the cutting edge of identifying novel therapeutic options for lung diseases like chronic obstructive pulmonary disease (COPD) and fibrosis.
It has been previously reported that partial resection of the lung results in a rapid compensatory growth process of the remaining lung tissue, restoring normal lung volume, cell mass and organ function, in a variety of mammalian species 2–6. Compensatory lung growth following unilateral pneumonectomy (PNX) has been documented for dogs, rabbits, ferrets, rats and mice. However, it has been most extensively studied in rats, and the time course, volumetric and morphological changes are well characterised in this species 7, 8. Immature dogs show complete restoration of tissue loss after PNX 9, 10, whereas in mature dogs at least 55% of parenchyma has to be resected to evoke post-PNX lung growth 5. Presently, much less detailed information is available for mice, although this species offers a wide spectrum of unique functional approaches in studies where genotypes are manipulated. To assure the suitability of adult mice for studying lung growth and alveolarisation as the basis for further functional investigations, the current authors investigated the time course and quantity of lung growth as well as the morphology after resection of the left lung in C57BL/6 mice.
Methods
Animal surgery
All animal procedures were performed according to the guidelines of good animal experimental practice of the University of Giessen (Giessen, Germany) and were approvedby the local authorities for animal experiments. The 12–16-week-old C57BL/6 mice were anaesthetised by a singlei.p. injection of ketamine (60 mg·kg−1) and xylazine (2 mg·kg−1). The mice were orally intubated with a 21-gauge atraumatic cannula and mechanically ventilated with a mouse ventilator (Hugo Sachs Elektronik, March-Hugstetten, Germany). The left thoracic wall was shaved and disinfected. An incision was performed in the fifth, left intercostal space and extended to the pleural cavity. The left lung was carefully lifted through the opening and the left main bronchus, in combination with the hilar pulmonary vessels, was ligated with 6/0 silk. Afterwards, the lung was resected, and the thorax and skin closed with silk sutures. Mechanical ventilation was terminated at the onset of spontaneous breathing. Post-operative pain was treated with buprenorphin subcutaneously twice daily for 48 h. Chow and water was provided ad libitum.
Volume determination
The volume of the right lung was determined before PNX (baseline value), and at days 1, 3, 5, 7, 10 and 21 after PNX (n=8 at each time point). The mice were heparinised by an i.p. injection of 500 units of heparin and then sacrificed. The thorax was opened and the trachea and pulmonary artery were cannulated. The lung was then inflated by intratracheal instillation of PBS with a constant hydrostatic pressure of 10 cmH2O and perfusion-fixed with 4% paraformaldehyde viathe pulmonary artery with a pressure of 20 cmH2O for 30 min. Afterwards, the trachea was ligated, and the heart and lung were resected en bloc from the thorax and fixed in fresh fixative overnight at 4°C. The volume of lungs and single lobes was determined by fluid displacement 11.
Bromo-desoxyuridine labelling
To assess proliferation during the time of rapid lung volume increase, mice were fed bromo-desoxyuridine (BrdU) continuously via the drinking water over 10 days, starting on the day of the PNX (0.8 mg BrdU·mL−1, n=3 mice per group). Lungs were cryosectioned and stained with alkaline phosphatase-coupled anti-BrdU antibody developed with nitroblue tetrazolium/5-bromo-4-chloro-3-inodolyl phosphate and counterstained with neutral red. Representative tissue sections of each lung were photographed at 200-fold magnification, BrdU-positive and -negative nuclei were counted and the ratio of BrdU-positive nuclei was calculated.
Immunofluorescent staining
Paraffin sections of 3 µm thickness were sectioned on a rotary microtome (Leica, Wetzlar, Germany). Tissues were deparaffinised and rehydrated, and immunofluorescent staining was performed as described previously 12. Pairs of serial sections were stained for Ki67 and phenotypic cell antigens, respectively, which included: cytokeratin for airway and alveolar epithelial cells; surfactant proprotein B for alveolar epithelial type-2 cells; pecam-1 (CD31) for endothelial cells; α-smooth muscle actin for smooth muscle cells; and vimentin for fibroblasts. For Ki67 staining, antigen retrieval was performed by heating the mixture at 120°C for 10 min in a 10-mM citrate buffer by means of a pressure cooker. The primary antibodies used were against: CD31 (Pharmingen, Hamburg, Germany), pan-cytokeratin, Ki67 (Dako, Hamburg, Germany), proSPB (Chemicon; Temecula, USA), α-smooth muscle actin and vimentin (Sigma, Deisenhofen, Germany). Hoechst 33342 (Sigma) was used to stain nuclei of the Ki67 cells.
Ventricular mass quotient
The hearts of the mice 21 days after PNX (n=5) and the hearts of control mice (n=16) were dissected, left and right ventricles were air dried on glass slides for 48 h at 40°C, then weighed, and right-to-left ventricular weight ratios were calculated.
Preparation of the lungs for stereological studies
Preparation of the lungs for stereological studies was done according to Fehrenbach et al. 13. In brief, after opening the thoracic cavity, fixation was performed by intratracheal instillation of a mixture of 4% paraformaldehyde and 0.1%glutardialdehyde in N-2 hydroxyethylpiperazine-N-2-ethanesulphonic buffer at a hydrostatic pressure of 20 cm for 30 min (n=5 mice per group). After ligation of the trachea, the lungs were excised and stored in cold fixative over night. Immediately before sampling of the tissue slices, the absolute volumes of right and left lungs were determined by fluid displacement according to Scherle 11. Systematic random sampling of the lung tissue was performed according to Michel and Cruz-Orive 14. In brief, the organ was cut into 2-mm slices. After several buffer rinses, slices were post-osmicated for 2 h with 1% osmium tetroxide and stained en bloc over night with half-saturated, aqueous uranyl acetate as described previously 13. The tissue was dehydrated through a graded series of ethanol and embedded in methacrylate resin (Historesin; Leica). Methacrylate embedment was proven to avoid the technical bias by shrinkage (data not shown), which is regularly observed with paraffin-embedded material 15. Semi-thin sections (1 µm) were cut on a Leica RM 2165 rotary microtome (Leica) and stained for 30 min at 60°C with a mixture of methylene blue and Azur II (ratio of 2:1).
Stereological studies
One glycolmethacrylate section of every slice of each lung (n=5 mice per group) was used for quantitative histological analysis, and from each lung slice a representative collection of test fields were obtained by meander scanning. This ensured that the structural parameters recorded were representative of the whole organ. Point and intersection counting was performed in a two-level procedure by established stereological methods to determine volumetric parameters as well as the total alveolar surface area 16. At the first level, volume fractions of nonparenchyma (comprising of lumen, wall, and tunica adventitia of airways and vessels) and parenchyma were determined at a primary magnification of ×308. At the second level, volume fractions of alveolar duct airspace, alveolar airspace and alveolar septum were determined at a primary magnification of ×615. Absolute volumes of the respective lung compartments were obtained by multiplication of the total lung volume with the respective volume fraction. In addition, mass-specific volumes (volumes normalised to animal body weight) were calculated to account for structural changes due to physiological postnatal growth processes during the experimental period. Total alveolar surface area was calculated by multiplication of the surface density of the alveolar septum (i.e. surface per unit volume of parenchyma) and the absolute parenchymal volume. The surface density of the alveolar septum was estimated from the sum of intersections of the test lines (35.2 µm of length, at a magnification of ×615) with the alveolar septa and the sum of test points hitting parenchymal structures.
To further characterise the alveolar compartment, the point-sampled intercept method was used, which is an efficient and direct tool for measuring the volume-weighted mean particle volume from single isotropic uniform random sections 17. As described previously 13, the volume-weighted mean volume of alveoli was estimated in a one-step procedure projecting a combined grid of points and intercepts in a random direction on each test field of each section. By measuring the length of the isotropic-test line passing through the sampling point(s) within the alveolar transect, the volume dimensions of the alveoli were determined. Following Massaro and Massaro 18, the alveolus was closed by drawing an imaginary line connecting the ends of the alveolar walls whenever the isotropic test line passed the alveolar opening. This parameter includes information on mean alveolar size and variations of alveolar size. The increase in volume-weighted mean volume of alveoli may, therefore, resemble larger alveoli as well as higher variation in size without change in mean alveolar size.
All stereological analyses were performed by means of a computer-based system (connected to an Olympus BH-2 microscope (Olympus, Planegg, Germany)).
Determination of total lung DNA
Total genomic DNA was isolated from the left and right lungs of nonoperated control (n=5) and day 21 PNX (n=5) mice as described previously 19. The lungs were homogenised and subsequently digested overnight with proteinase K. The DNA was extracted with phenol-chloroform-isoamylalcohol (ratio of 25:24:1), and precipitated in ethanol and ammoniumacetate. The DNA concentration was measured by spectrophotometry (optical density at 260 nm: OD 260). The ratio of OD 260/280 was >1.8 in all accepted measurements.
Lung function testing
For determination of lung function parameters (tidal volume, breathing frequency, inspiration time, expiration time, mid-expiratory flow, peak expiratory flow), two groups of mice (n=4) were pneumonectomised and investigated at day 3 and 21 after the operation. A third group of mice that had not been operated on but were of the same age served as the control (n=4). The lung function testing was done with conscious mice in head-out body plethysmography, as described previously 20.
Statistics
Paired t-tests were employed when not stated differently. Statistical significance was accepted at p<0.05.
Results
Surgical resection of the left mouse lung was followed by a rapid and complete restoration of the initial lung volume, due to compensatory growth of the four right lobes. Volume gain was significant already 1 day after the PNX and increased until day 21. The volume of the right lung after 21 days of compensatory lung growth, corrected for individual body weight (mass-specific volume), equalled the mass-specific total lung volume of left and right lungs of nonoperated mice of the same age (fig. 1⇓). Thereby, the right lung expanded to 145% of the initial volume. In control mice, the left lung had a share of 30.6% of the total lung volume. The remaining lobes of the right lung after PNX increased disproportionately in volume, with the cardiac lobe displaying the strongest relative volume gain over 21 days (fig. 1⇓ and table 1⇓).
Quantity and kinetics of post-pneumonectomy (post-PNX) lung growth in mice. a) The volume of the right lung (RL) increased rapidly after left-sided pneumonectomy. The time course of lung growth leading to complete compensation of volume loss is shown. Lung volume was corrected for individual body weights. b) The four right lobes did not increase equally in volume over 21 days. The cardiac lobe showed the largest volume gain relative to its initial volume and the upper lobe the smallest (n=8 in all experiments). LL: left lung; xd PNX: x days after PNX. Data are presented as mean±sem. *: significant (p<0.05) volume increase compared with control or §: with control and preceding time point.
Volumetric data
To elucidate whether this complete regeneration of organ volume was paralleled by an increase in lung tissue and, consequently, in the growth of alveolar walls and formation of alveoli rather than by the mere distension of the airspaces, total lung DNA was determined and stereological studies were performed. Total DNA content of the right lung 21 days after left-sided PNX was significantly increased in comparison with the total DNA of the right control lungs, and equalled total DNA content of left and right lungs of nonoperated control mice (fig. 2a⇓). Proliferation in alveolar septa during the first 10 days of compensatory lung growth was investigated by continuous feeding of BrdU over this time and subsequent BrdU-staining of the lungs. The proliferation index of septal cells was significantly increased in compensatory grown lungs and locally accentuated in subpleural areas (fig. 2b–d⇓). Comparable results were obtained by staining for the proliferation antigen Ki67 in control lungs and lungs 7days after PNX, showing dramatically increased proliferation activity at this time point (fig. 3⇓). To further clarify which cell types proliferate at this time point, pairs of serial sections were stained for phenotypic cell antigens and Ki67, respectively. Specific staining was done for airway and alveolar epithelial cells (cytokeratin), for alveolar epithelial type-2 cells (proSPB), for endothelial cells (CD31), for smooth muscle cells (α-smooth muscle actin) and for fibroblasts (vimentin). It was demonstrated that all these cell compartments proliferate and thereby contribute to the post-PNX lung growth (fig. 4⇓).
Measurement of total DNA mass (a) and cellular proliferation (b–d). a) The total DNA content of the right lung before pneumonectomy (RL control), right lung 21 days after left-sided pneumonectomy (RL 21d PNX), and right and left lung of nonoperated mice (RL+LL control, n=5). The DNA mass of right lungs 21 days after pneumonectomy was significantly increased compared with right lungs from nonoperated mice, and equalled DNA mass of left and right lungs of nonoperated mice. b) Proliferation of septal cells over 10 days compensatory lung growth. Proliferating cells were labelled by continuous feeding of bromo-desoxyuridine (BrdU), and visualised by specific antibody staining in nonoperated control mice (c) and pneumonectomised mice (d). Blue nuclear stain: BrdU positive; brown/red nuclear stain: BrdU negative, n=3. Data are presented as mean±sem (*: p<0.05). Scale bar=50 µm.
Proliferation of pulmonary cells 7 days after pneumonectomy (PNX). a, c) Nuclear staining for the proliferation antigen Ki67 (red), and b, d) the overlay of Ki67-staining and the nuclear dye Hoechst 33342 (blue). Massive increase of proliferation can be detected in the lung 7 days after PNX (c, d) when compared with lung tissue of nonoperated mice of the same age (a, b). Scale bar=100 µm.
Phenotypical characterisation of proliferating cells in a lung 7 days after pneumonectomy. Serial sections stained red for the phenotypic markers cytokeratin (a), pro-SPB (c), pecam-1 (f), α -smooth muscle actin (h), vimentin (k) and Ki67 (b, d, e, g, i, j, l ). Nuclear stain with Hoechst 3342 (blue) and tissue autofluorescence (green) are shown for better orientation where needed. Proliferation of bronchial epithelial cells (arrowheads) and alveolar epithelial cells (arrows) are shown in a) and b). Further specification for proliferating type II cells was done in c–e) (arrowheads). The figures f) and g) show a proliferating endothelial cell in a pulmonary artery (arrowheads). Proliferating smooth muscle cells in the media of a small pulmonary artery are visible in h–j) (arrowheads). Proliferating alveolar fibroblasts are shown in k) and l) (arrowheads). Scale bars=100 µm.
Stereological investigation of semi-thin sections (fig. 5a–f⇓) revealed that the mass-specific parameters (normalised to individual body weight) for lung volume, alveolar surface area, parenchyma, nonparenchyma and airspaces were not significantly different in right lungs 21 days after PNX compared with left and right control lungs of nonoperated mice (table 2⇓). The only significant difference observed was in the mass-specific volume of alveolar septa. Focal areas of thickened septa were observed in subpleural regions. Masson-Goldner staining showed increased cellularity, rather than increased matrix deposition, in those thickened septa (fig. 5g, h⇓). Notably, the volume-weighted mean alveolar volume wassignificantly larger (p<0.001) in the right lungs 21 daysafter PNX (89.5±4.3×103 µm−3) in comparison with control lungs (51.4±9×103 µm−3), an effect that was independent of body weight (data not shown). Although the absolute total airspace volume of alveoli was significantly larger (p=0.009) in right lungs 21 days after PNX (453±19 mm3) in comparison with left and right lung of nonoperated animals (370±15 mm3), the concomitant increase in the volume-weighted mean alveolar volume indicates that this was at least only partially achieved by the formation of new alveoli.
Light micrographs of nonoperated control lung (a, d) and of right lung 21 days after left lung pneumonectomy (PNX) (b, c, e, f, g, h). PNX- and age-matched control lungs exhibited similar microscopic appearance of terminal bronchioles, alveolar ducts and alveoli, particularly in central regions. Subpleural regions exhibited focal areas (c outlined by - - - - , and g) with thickened alveolar septa and increased cellularity, while closely adjacent regions (e, h) exhibited similar architecture as control lungs (d). Masson-Goldner staining for nuclei (red) and extracellular matrix (green) shows higher cellular density in thickened alveolar septa (arrowheads) rather than increased deposition of matrix material. A: arterial vessel; PL: pleura; V: venous vessel. Scale bars: a–c)=100 µm; d–f)=33.3 µm; and g–h)=25 µm.
Stereological data
The ventricular mass quotient (right/left ventricular mass) was 0.27±0.01 for control animals (n=16) and 0.273±0.02 for mice 21 days after PNX (n=5).
To assess parameters of in vivo pulmonary function, head-out body plethysmography was performed on spontaneously breathing mice. The breathing frequency and tidal volumes of control mice and mice 21 days after PNX were similar, whereas breathing frequency was increased and tidal volume was decreased at 3 days after PNX. Furthermore, mid-expiratory airflow, a parameter for lung volume and bronchial resistance, was decreased 3 days after PNX but reached the control value 21 days after PNX (fig. 6⇓).
Lung function testing of spontaneously breathing mice before (control), and 3 and 21 days after pneumonectomy (3d and 21d PNX, respectively). At 3 days post-PNX, all parameters indicated an alteration of lung function due to tissue loss with increased breathing frequency, decreased tidal volume, airflow parameters, and inspiratory and expiratory times. At 21 days following pneumonectomy, the parameters returned to the initial values. Data are presented as mean±sem (n=4). For statistical significance, analysis of variance was performed; #: 0.023; ¶: 0.055; +: 0.375; §: 0.356; ƒ: 0.004; ##: 0.002.
Discussion
The compensatory lung growth of mice was investigated to evaluate if it can be used as a reliable model of inducible lung growth and alveolar wall formation in this species. Functional studies on compensatory lung growth in mice have already been carried out 19, 21, 22, but detailed analysis of the time frame and kinetics of lung growth, as well as stereological investigations for the quantification and proof of alveolar septal wall formation are still lacking. In contrast, detailed studies have been performed in rats and dogs 10, 23–25.
Compensatory lung growth has not been observed in adult humans. Children may show partial compensation of pulmonary parenchymal loss 26. The transplantation of adult lung lobes into children results in a distension of the transplant during body growth, but not in compensatory growth of alveolar septa in the transplanted lung 27. This might compare to compensatory lung growth following PNX in canines, which is restricted in adult dogs in that it only occurs after resection of >50% of lung tissue 5. In contrast, immature dogs seem to have the capacity of complete compensatory lung growth due to vigorous tissue gain and alveolarisation 10. An age dependency of post-PNX lung growth has also been described for rabbits 28, but could not be demonstrated in rats 7.
The fact that reconstitution of lung parenchyma and gas exchange surface area can be triggered in different mammalian species, in both youth and adulthood, suggests the existence of intrinsic programmes that may regulate pulmonary maintenance, alveolar septal growth and alveolarisation.A review on this matter is given by Brown et al. 1. If these programmes could be identified, new therapeutic approaches to a variety of pulmonary diseases would be conceivable. In the era of in vivo genetic manipulation, the mouse is the ideal species for the investigation of such regenerative programmes of the lung.
In this study evidence is presented for rapid and complete restoration of lung volume in mice following PNX (fig. 1⇑). Stereological investigations revealed that the volumes of airspace (of both alveolar ducts and alveoli) as well as the total gas exchange surface area did not significantly differ between controls and pneumonectomised animals 21 days after PNX. In this study, a significant difference for tissue volume of alveolar septa that was increased in the pneumonectomised mice was observed (table 2⇑). This can be explained by the observation of areas with thickened and putatively not fully matured septa 21 days after PNX (fig. 5⇑). These septa were comprised of densely packed cells rather than of extensive extracellular matrix deposition, which is in line with the assumption that these septa, after rapid proliferation of septal cells, still await further maturation and thinning. Interestingly, the volume-weighted mean alveolar volume was also significantly increased in the PNX group. This may represent an increase in alveolar size or a higher variation in alveolar sizewithout increase of mean alveolar size. The classical stereological parameters determined in this study do not allow for a definite conclusion about the relative contribution of newly formed alveoli to the compensation of alveolar surface area, and, therefore, for the generation of new alveolar septa versus the growth of already existing septa. This will require the direct determination of the absolute number of alveoli by design-based stereological methods that have been developed in parallel to this study 29.
Lung functional parameters of tidal volume, breathing frequency and mid-expiratory airflow, measured in vivo, were markedly altered at 3 days after PNX and normalised by compensatory lung growth at 21 days after PNX, which indicates not only histological but also functional restoration.
A massive cellular proliferation must underlie the rapid generation of tissue in this model. Only very low basal levels of proliferation can be observed in nonoperated control lungs, whereas lung tissue 7 days post-PNX shows dramatically increased proliferation (fig. 3⇑). Phenotypical characterisation of proliferating cells showed that all major cellular compartments participated in the proliferative process (fig. 4⇑). Continuous labelling of proliferating cells over 10 days with BrdU revealed areas of BrdU-positive thickened septa in subpleural areas, which is in line with findings and with theoretical considerations suggesting subpleural and peribronchial regions for preferred lung growth 30. A significantly higher rate of proliferation in subpleural regions compared with central areas of the lung were shown very recently in newborn dogs 23. This poses the question “what is the main driving force for the post-PNX lung growth programme?”. Although it was not the main focus of this study, the findings of preferred subpleural lung growth and preferred growth of the cardiac lobe directly neighbouring the newly formed cavity favour a major contribution of sensing of mechanical forces and mechanotransduction pathways as an underlying mechanism. This issue should be addressed in further studies. The ratio of right/left ventricular mass did not change after PNX, indicating absence of pulmonary hypertension.
In conclusion, this study shows that unilateral pneumonectomy of adult mice is followed by a rapid and complete compensatory growth of the remaining lung, which includes cellular proliferation and leads to restoration of an adequately sized alveolar surface area and respiratory function parameters. The lung volume and alveolar surface area reconstitution iscompleted at 21 days after pneumonectomy, although focalareas of the lung still show signs of septal immaturity. Proportional neo-alveolarisation could not be definitely proven in this study. Interestingly, the extent of growth response differs among the different lobes of the right lung, being largest in the cardiac lobe. This study in wildtype C57BL/6 mice provides a basis for future studies in genetically altered mice, which will allow questions about the intrinsic regulative pathways of lung growth to be addressed, which, in turn, could help to develop new therapeutic strategies leadingto a successful induction of alveolarisation in adult mammals.
Acknowledgments
The authors wish to gratefully acknowledge the expert technical assistance of T. Rausch, N. Weisel and S. Lay.
- Received January 14, 2004.
- Accepted June 5, 2004.
- © ERS Journals Ltd
References
