Background
Impaired alveolar development is a leading cause of neonatal morbidity in premature infants weighing less than one kilogram. Deficient alveolar maturation in these children is often characterized by distal airspace enlargement, disruption of elastin fibers and mucus cell hyperplasia. Antenatal exposures of the premature lung may increase susceptibility to inflammation and subsequent postnatal (PN) lung injury. Genes that regulate alveolarization, innate immunity and inflammation are likely to contribute to susceptibility to and outcome in neonatal lung disease.
In a search for downstream targets of glucocorticoid (GC) that regulate lung maturation, we cloned
Lgl1 (late gestation lung 1), a developmentally regulated gene in the lung [
1‐
3].
Lgl1 is a CRISP family (cystine rich secretory protein) protein characterized by a secretory signal and two LCCL (also known as FCH) domains [
4‐
6]. The LCCL domain, an as yet poorly understood module found in over 100 extracellular proteins, has been implicated in directional cell migration and differentiation [
5,
7], extracellular matrix deposition [
5,
8], cell adhesion [
9] and innate host-defense mechanisms [
4,
10,
11]. While
Lgl1 synthesis is almost exclusively restricted to the mesenchyme,
Lgl1 protein is associated with lung epithelial cells from the late canalicular period onward [
2,
3]. We showed previously that
Lgl1 protein stimulates airway branching in lung explant culture [
12]. Maximal fetal expression of
Lgl1 was, however, concordant with the onset of augmented surfactant production in late gestation [
1,
3]. In postnatal rat lung,
Lgl1 concentrated at the tips of budding alveolar septa [
3]. Levels of
Lgl1 were drastically reduced in rat O
2 toxicity models of bronchopulmonary dysplasia (BPD), a chronic lung disease of impaired alveolarization in premature infants, and were restored during recovery in air [
3]. Taken together, these observations suggested that
Lgl1 may regulate both early and late events in lung organogenesis.
We now report on the development of an Lgl1 knockout mouse to investigate the in vivo function of Lgl1 in regulating multiple aspects of lung development. Absence of Lgl1 in homozygous null (Lgl1
-/-) mice was not compatible with life. We describe here the lung phenotype of Lgl1
+/- heterozygous mice. Lungs of developing Lgl1
+/- mice were characterized by disorganized elastin fibers, early expression of inflammatory cytokines and goblet cell hyperplasia. In mature Lgl1
+/- mice, reduced Lgl1 expression was associated with altered lung mechanics.
Methods
Generation of Lgl1 knockout mice
All procedures involving animals were conducted according to criteria established by the Canadian Council for Animal Care and approved by the Animal Care Committee of the McGill University Health Centre. A 13.7 kb EcoR1 fragment of BAC clone 34304 containing the
Lgl1 gene was subcloned into pQZ1BamH1 and a Neo cassette was used to replace exon 2. Not1-Sal1 digestion of this construct produced a 9.6 kb targeting fragment that was electroporated into ES cells. Southern analysis and PCR were used to verify accuracy of targeting. Mouse genotypes were verified by PCR. The primers used were (5'-3'): reverse (wild type) CACTGCTCCGTGTATCAAGCATACAC; reverse (NeoI) GACAATCG GCTGCTCTGATG; or reverse (5' to3') TCGTCGTGACCCATGGCGAT (NeoII) and forward (for all 3 reactions) CAGGTCTGGCTCTGAGGTTCTTGCA. The expected amplification products were: 0.8 kb (wild type), 1 kb (Neo1) and 0.46 kb (Neo2). For details on mouse preparation see Additional file
1.
Isolation of Total Lung RNA
Total RNA was prepared from lungs, brain, heart, kidney, thymus and spleen using Trizol reagent (Invitrogen, Burlington, ON, Canada) according to the manufacturer and was resuspended in 1× RNASecure (Ambion, Austin, TX, USA). RNA was pooled for each litter according to genotype (N ≥ 4).
Quantitative Real-Time RT-PCR
Quantitative real-time RT-PCR was performed on the Mx4000 QPCR system (Stratagene, La Jolla, CA, USA) using the Quantitect One-Step Probe RT-PCR Kit (Qiagen, Mississauga, ON, Canada) as directed by supplier. Gene-specific primers and FAM labeled probes for mouse
LGL1, IL-4, IL-13, Mucin5AC and Tropoelastin were designed using Qiagen's online QuantiProbe Design Software. Quantitect Gene Expression Assay for mouse 18S (Qiagen, Mississauga, ON, Canada) was used to normalize for the input of RNA. The results were analyzed according to the standard curve method. One-step real-time RT-PCR reactions were performed in 25 μL volumes for 40 cycles, using 20 ng of total RNA for
Lgl1, IL-4, IL-13, Mucin5AC and Tropoelastin and 50 pg for 18S. For a list of primers and probes used see Additional file
2.
Bronchoalveolar Lavage (BAL)
BAL was performed by instilling the lungs four times with 1 ml cold phosphate-buffer saline through a tracheal cannula. Lavage fluid was centrifuged and pellets were resuspended in 0.5 ml cold saline. Total cell numbers were counted with a haemocytometer. For differential cell counts, cytospin slides were prepared (Cytospin 4; Shandon, Pittsburgh, PA) and stained with Diff-Quick; at least 200 cells/slide were counted and percentage of each cell type was calculated.
Lung fixation
Lungs were inflated with 4% paraformaldehyde at a pressure of 20 cm of water. Lungs were gently extracted and fixed in 4% paraformaldehyde overnight. Samples were dehydrated through a series of increasing ethanol washes and embedded in paraffin. 5 μm thick tissue slices were cut through the entire lung.
Antibodies
Lgl1, 1:100 (Covance, Quebec, Canada), β-actin, 1:5000 (Sigma-Aldrich, Oakville, ON, Canada), Anti-Rabbit IgG HRP conjugated, 1:5000 (Amersham, Little Chalfont, Buckinhamshire, UK), Anti-Rabbit IgG AlexaFluor 594 and Anti-Rat IgG AlexaFluor 488 conjugated, 1:200 (Invitrogen, Burlington, ON, Canada), CD34, 1:100 (Abcam, Cambridge, MA, USA).
Immunohistochemistry
Sections of paraffin-embedded lung tissue were stained with hematoxylin and eosin or used for histochemical staining. Ten sections from each of at least 4 and up to 7 animals were assessed for all histochemical experiments and representative images shown in all figures. For immunohistochemistry, slides were rehydrated through a series of decreasing ethanol washes, rinsed with PBS-0.03% Triton and incubated in warm 10 mM sodium citrate for antigen retrieval. Slides were then incubated in H
2O
2 and methanol for 20 minutes to block endogenous peroxidase activity. To block non-specific binding, slides were incubated in PBS-0.03% Triton containing 5% normal goat serum and 1% BSA. Primary antibodies were incubated overnight at 4°C and the following day in corresponding fluorescent conjugated secondary antibody for 30 minutes at room temperature. For CD34 and
Lgl1 co-immunohistochemistry, the slides were then washed with PBS-0.03% Triton and blocked again with 5% normal goat serum and 1% BSA. The second primary antibody was incubated overnight at 4°C and the following day in corresponding fluorescent conjugated secondary antibody for 30 minutes at room temperature. Slides were washed with PBS-0.03% Triton and mounted with pro-long anti-fade media containing DAPI (Invitrogen, Burlington, ON, Canada). For the co-staining, the DAPI is not shown to improve visualization. For
Lgl1 immunohistochemistry, protein levels were quantified using the Northern Eclipse program (Northern Eclipse software, Empix Imaging, Inc. Mississauga, ON, Canada) as per Nadeau
et al. 2006 [
3].
Identification of Goblet cells
Slides were rehydrated through a series of decreasing ethanol washes and stained with Periodic Acid-Schiff kit (Sigma-Aldrich, Oakville, ON, Canada) to visualize goblet cells. For goblet cell staining, sections from at least 6 and up to 10 animals were assessed and representative images shown in all figures.
Elastin Staining
Slides were rehydrated through a series of decreasing ethanol washes and elastin fibers were stained with Fuschin Weigert stain and counterstained with methyl green for better visualization. For elastin staining sections from at least 4 and up to 7 animals were assessed and representative images shown in all figures.
Morphometry
Mouse lungs were fixed under constant distending pressure of 20 cm of fixative. Morphometric measurements were made on hematoxylin and eosin stained lung sections (n > 5 mice). A minimum of 10 representative fields were studied in each lung. A computer generated grid (384 intersections) was superimposed on digital images and grid intersections were examined to determine whether they localized to airspace or tissue. Percent fractional airspace or fractional area of lung parenchyma was quantified using Northern Eclipse software.
Lung Mechanics
At 4 weeks of age, mice were deeply anaesthetized by an i.p. injection of xylazine (8 mg/kg) and pentobarbital (70 mg/kg), tracheotomized and placed on a small animal ventilator (flexiVent, SCIREQ, Canada). Animals were ventilated quasi-sinusoidally (150 breaths/min and tidal volume of 10 ml/kg) and subsequently paralyzed by an i.p. injection of pancuronium bromide (0.8 mg/kg). Maximal resistance and elastance were recorded before and after increasing doses of aerosolized methacholine.
Statistical Analysis
All results are expressed as mean ± standard error of the mean. P ≤ 0.05 was considered to be statistically significant as measured by student t-test or ANOVA as appropriate.
The numbers of goblet cells in postnatal day 14 lungs displayed a single modal distribution in wild type, but a marked bimodal distribution in heterozygous (Lgl1
+/-) mice. Subgroup analyses of the heterozygous mice revealed the data were not normally distributed; a Mann-Whitney U test showed a significant difference between the subgroups (p < 0.02).
Discussion
The finding that null mutation of the Lgl1 gene in mouse embryos is lethal prior to the onset of lung morphogenesis classifies Lgl1 as an essential early gene in organismal development. The determinants of embryonic lethality in Lgl1
+/- mice remain to be determined and pre-date lung organogenesis. Mutation in Lgl1 has pleiotropic effects. Moreover, it is possible that effects on Lgl1 expression in other organ systems contribute to the postnatal phenotype observed in the lung.
The present study demonstrates that the heterozygous
Lgl1 genotype is sufficient to induce a complex phenotype. In postnatal lung, histologically immature areas with distinctly thickened respiratory interstitium and the appearance of delayed secondary septation alternate with areas of relatively normal appearance. Disorganized elastin fibers, goblet cell hyperplasia and high levels of inflammatory cytokines are present in the absence of detectible differences in
Lgl1 mRNA levels. Our inability to detect changes in
Lgl1 mRNA expression in total RNA isolated from developing lung is likely to reflect a specific requirement for
Lgl1 in a subpopulation of cells in which suppressed expression of
Lgl1 is sufficient to produce the observed phenotype but escapes detection by RT-PCR of total lung RNA. For example, we showed previously that
Lgl1 is maximally expressed in fibroblasts adjacent to the epithelium [
1,
2]. We also demonstrated that
Lgl1 is secreted and taken up by epithelial cells beginning in late gestation and continuing in PN life [
2]. Deficient
Lgl1 expression in a subset of fibroblasts important in mesenchymal-epithelial interactions that regulate alveolarization may contribute to the observed phenotype.
We have backcrossed our Lgl1
+/- mice onto the C57BL/6 background (eight generations). Neonatal Lgl1
+/- mice on this background faithfully reproduce the phenotype of disorganized elastin, goblet cell hyperplasia with elevated levels of MUC5AC and increased IL-4 and IL-13 expression, demonstrating that the phenotype we report is not due to a mixed genetic background.
The association of a distinct respiratory phenotype in the absence of significant reduction in mRNA has been reported previously [
13,
20]. Foxf1 heterozygotes can be divided on the basis of pulmonary levels of Foxf1 mRNA[
13]. Low Foxf1 producers fail to undergo differentiation of terminal airspaces. An
albeit less-severe defect in septation of the lung periphery is
observed in High Foxf1 1 producing heterozygotes mice that express normal or nearly normal (90%) levels of Foxf1 mRNA[
13]. Moreover, these animals display aberrant expression of a number of developmentally important genes in the lung.
There was considerable variability in Lgl1 protein levels in Lgl1
+/- mice. No such variability was observed in wild type animals. We demonstrated regional differences in tissue fraction in postnatal Lgl1
+/- mice. Given that Lgl1 is of mesenchymal origin and that mesenchymal thinning is a prominent feature of lung maturation, variability in Lgl1 protein levels may reflect varying degrees of developmental delay. The observed reduction in Lgl1 protein in the lungs of Lgl1
+/- mice may also reflect effects on RNA stability or protein turnover Absence of Lgl1 in endothelial cells suggests that Lgl1 does not have a direct role in PN angiogenesis.
Genetic modifiers of tissue- and temporal-specific expression of
Lgl1 may also contribute to the observed variation in penetrance of the
Lgl1
+/- phenotype. In this context it is important to recall that low levels of
Lgl1 in pseudoglandular lung are essential to airway branching[
12]. Haplosufficiency for
Lgl1 is clearly sufficient to rescue the branching program. At the same time, effects on tropoelastin expression were observed at E14.5 suggesting that the heterozygous phenotype does impact lung development in the pseudoglandular period.
Reduced
Lgl1 expression in mature
Lgl1
+/- mice was associated with normal baseline lung function. As expected, MCh challenge provoked an increase in airway resistance and elastance in both
Lgl1
+/- and wild type littermates. However, the effects on both these parameters were significantly greater in wild type mice. Moreover, tropoelastin expression in mature
Lgl1
+/- mice was elevated. Elastic interdependence of the lung accounts for orderly elastic recoil of lungs during passive expiration [
21]. Organization of the elastin network is initiated in the pseudoglandular lung and peaks during alveolarization [
16,
17]. At the alveolar level, elastic interdependence is mediated by the correct expression, cross-linking, and orientation of elastin and collagen fibers. Absence of a correctly cross-linked and oriented elastin matrix predisposes to aberrant alveolarization. Deficient alveolar maturation in BPD includes disruption of elastin fibers, distal airspace enlargement, and mucus cell hyperplasia [
14]. All of these properties were present in newborn
Lgl1
+/- mice. Dysregulation of elastin synthesis is also a prominent feature of BPD in murine [
22] and preterm lamb models [
23].
Excessive degradation of the elastin matrix underlies the loss of elastic interdependence and alveolar degeneration associated with chronic lung disease in adults. Indeed, it is believed that individuals with even limited elastin insufficiency may suffer damage sufficient to preclude alveolar repair from a less severe injury than would be required to have this effect in normal subjects. Survivors of preterm birth are at particular risk to develop premature COPD [
21,
24]. The disrupted elastin architecture in
Lgl1
+/- mice appears to resolve at maturity. Yet these animals have elevated tropoelastin levels and impaired lung function at maturity. It is tempting to speculate that latent effects on elastin integrity may increase vulnerability of
Lgl1
+/- mice to respiratory insult at maturity and that
Lgl1 haploinsufficiency may modify risk to respiratory insult in later life.
Given that the disruption of elastin expression and/or organization has been associated with lung injury in the newborn period, it was of interest to establish whether
Lgl1
+/- mice displayed any other characteristic features associated with neonatal lung injury. Inflammation is known to interfere with lung development in model systems and is present chronically in the lungs of preterm infants who develop BPD [
25]. Goblet cell hyperplasia is associated with multiple respiratory disorders including asthma, BPD and emphysema. It has been suggested that exposure of the developing lung to inflammation may be central to the development of BPD[
25]. Analysis of goblet cell number in PN14 lungs of
Lgl1
+/- mice identified a bimodal distribution, with one subgroup of animals displaying significantly elevated numbers of goblet cells in trachea and bronchi associated with elevated mucin (MUC5AC) production. A second group of animals showed no increase in goblet cell number. The presence of these 2 distinct groups is likely to reflect the incomplete penetrance of the
Lgl1
+/- phenotype and may be attributed to the genetic contribution from C57BL/6 and 129/J strains. Indeed, we have found that on the C57BL6 background, goblet cell hyperplasia in
Lgl1
+/- mice is much less variable.
The differentiation of epithelial cells to goblet cells is induced by inflammatory cytokines [
19]. We found dramatically elevated levels of the T
H2 cytokines IL-4 and IL-13 in the postnatal
Lgl1
+/- mouse lung, concordant with evidence of goblet cell hyperplasia. Both IL-4 and IL-13 have been shown to induce goblet cell hyperplasia and mucus hypersecretion in mouse airways [
26‐
28]. It is of interest that despite the pronounced rise in IL-4 and IL-13 observed at PN7, we did not detect an inflammatory infiltrate in BAL. Both Il-4 and IL-13 are cytokines associated with the inflammatory process in diseases such as asthma, but neither of these cytokines has chemokine properties. Indeed, Wills-Karp et al. [
28] have shown that these cytokines can induce pathophysiological features of asthma via mechanisms independent of eosinophil recruitment. Thus the inflammation observed in postnatal
Lgl1
+/- mice that are free of allergen and infection is un likely to involve recruitment of cellular infiltrates. The combined findings of goblet cell hyperplasia and induction of inflammatory cytokines are nevertheless consistent with a role for
Lgl1 in innate immunity suggesting that
Lgl1 may function to protect the lung from external insults.
To date, the domain structure of
Lgl1 has offered little to our understanding of its molecular function. The LCCL domain family now includes more than 100 proteins. Orthologues of
Lgl1 are typified by having two LCCL modules. Multiple functions have been attributed to the LCCL module (also known as FCH[
5]). These include roles in cell differentiation, motility and migration [
5,
7]; cell adhesion and matrix deposition [
8,
9]; and host-defense and innate immunity [
4,
10,
11]. The results of our studies provide the first evidence for a potential role of
Lgl1 in innate immunity. Mutagenesis of the LCCL modules in
Lgl1 will be necessary to explore such effects
in vitro.
There are a number of limitations to this study. While haploinsufficiency for Lgl1 is the only reasonable explanation for the pulmonary phenotype in Lgl1
+/- heterozygotes, we were not able to identify the precise time and localization of deficient Lgl1 that triggers the postnatal phenotype. There was considerable variability in several outcome measures among Lgl1
+/- heterozygotes. The development of a mouse model with conditionally regulatable expression of Lgl1 will allow a more definitive analysis of the role of Lgl1 in early lung development.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JL generated the knockout mouse. LR and IM characterized Lgl1 mRNA and protein, elastin and tropoelastin mRNA, and goblet cells and mucin mRNA in Lgl1
+/- heterozygotes and wild type controls. TB participated in H and E histochemistry and protein quantitation. KN and SC carried out lung function studies. NBS contributed to the design of the study and the preparation of the manuscript. FK conceived and participated in the design of the study and had a primary role in preparation of the manuscript.