LXR deficiency alters lung morphology

To determine the importance of LXR signaling in pulmonary histophysiology, we initially carried out a comparative histological analysis of the lung morphology from WT and Lxr double knockout (Nr1h3−/−/Nr1h2−/−) mice which will be described here as LXR-DKO mice. Macroscopically, we observed a remarkable pale halo bordering the pulmonary lobes (subpleural zone) in LXR-DKO mice, suggestive of lipid accumulation, which is evident at 6 months of animal life (Fig. 1a, lung; brown arrows). A closer histological analysis of lung areas from LXR-DKO mice by hematoxylin and eosin (H&E) staining revealed abundant cellular infiltrates indicative of tissue congestion (Fig. 1a, H&E; blue arrows). To check for possible lipid accumulation in these areas of the lung, we examined consecutive sections by Oil Red-O staining (OR-O). The appearance of OR-O positive areas (Fig. 1a, OR-O; yellow arrows) revealed an extensive accumulation of neutral lipids between the areas of cellular infiltration, showing that LXR deficiency leads to the development of lipidosis in the lung. Simultaneously, we evaluated the presence of immune cells in the subpleural zone by performing enzymatic immunohistochemistry for lymphocyte markers CD45R and CD3 which revealed areas of cellular congestion in the LXR-DKO lungs with increased B and T lymphocytes compared to LXR-WT lungs (Fig. 1A, CD45R/CD3; green arrows).

Fig. 1

Lung histopathology in LXR-deficient mice. a: WT or LXR-DKO lungs from 6-month-old mice were subjected to macroscopic analysis by H&E and oil red O (OR-O) staining, and CD45R/CD3 enzymatic immunohistochemistry. Arrows indicate lipid accumulation (black arrows), tissue congestion (blue arrows), neutral lipid accumulation (yellow arrows), and B- and T-lymphocyte infiltration (green arrows). b: WT or LXR-DKO lungs at 3, 6, 9 and 12 months of age were subjected to OR-O staining. Scale bars of all images are 50 μm. A representative image of 6 mice per genotype is shown

Next, we studied the evolution of morphological changes in the lungs of LXR-DKO mice throughout their life, starting at 3 months of age as the first adult age control group. As shown in Fig. 1b, the appearance of OR-O positive regions and accumulation of immune cellular infiltrates is progressive and increases with age in 3- to 12-month-old LXR-DKO mice compared to LXR-WT. Furthermore, deletion of either LXRα or LXRβ had minimal impact on lung histology in 3- and 9-months-old mice, indicating that both LXR isoforms are required to prevent the development of lung lipidosis and damage (Fig.S1). Taken together, these results demonstrate that loss of both LXR isoforms resulted in altered lung morphology characterized by age-dependent development of lipidosis and infiltration of immunocompetent cells, alterations characteristic of an inflammatory phenotype.

Alveolar macrophage and type II pneumocytes are involved in lung histopathology in LXR DKO mice

Pneumocytes and AMs are the main cells forming the lung parenchyma architecture. To elucidate whether these cells are involved in the histological disorder of LXR-DKO lungs, T2Ps and AMs were analyzed by pro-SP-C (surfactant protein-C) and CD68 expression, respectively. Pro-SP-C immunodetection (Fig. 2a) revealed a similar number of T2Ps between WT and LXR-DKO mice in contrast to AMs (identified by their characteristic morphology and location, purple arrowheads) which appear to be increased in LXR-DKO lungs. Double immunodetection of pro-SP-C and CD68 (Fig. 2a) confirmed the augmented frequency of the AM population in LXR-DKO lungs. In addition, OR-O positive areas were detected in LXR-DKO lungs, which coincided with regions occupied by AMs (Fig. 2a; yellow dashed line). These results suggest that the morphological impairment and lipidosis developed in LXR-DKO lungs were due to a large increase in the number of foamy macrophages in the alveoli.

Fig. 2

Disrupted alveolar physiology in LXR-deficient lungs. a: Consecutive sections of lungs from 6-month-old WT or LXR-DKO mice were subjected to enzymatic immunohistochemistry for the detection of pro-SP-C (left panels), double detection of pro-SP-C and CD-68 (middle panels) or Oil Red O (ORO-O) staining (right panels). Scale bars, 50 μm. Purple arrowheads indicate AMs and areas marked with yellow dashed line indicate coinciding areas with regions occupied by AMs in consecutive sections. A representative image of 5 mice per genotype is shown. b: BAL cells from WT or LXR-DKO 6-month-old mice was subjected to contrast-phase brightfield microscopy (left panels), Bodipy 493/503 fluorescent probe staining (neutral lipids) together with CD68 immunodetection (middle panels), and Bodipy 493/503 together with phospholipid fluorescent probe staining (right panels). Scale bars, 20 μm. Orange arrowheads indicate large hexagonal extracellular crystals. A representative image of 4 mice per genotype is shown. c: Lungs from 6-month-old WT or LXR-DKO mice were subjected to ultrastructural analysis by electron microscopy. Left panels, T2P (8.000 X; Scale bar: 5 μm), areas marked with yellow dashed line indicate cell contours; middle panels, T2P (16.000 X, Scale bar: 1 μm), orange arrowheads indicate mitochondria and red arrowheads indicate lamellar bodies; and right panels, AM (8.000 X, Scale bar: 2 μm). A representative image of 3 mice per genotype is shown

We then investigated whether this alteration could be reflected in the phenotypic characteristics of the BAL cell population from LXR-DKO lungs. BAL cells were analyzed by brightfield and confocal fluorescence microscopy using fluorescent probes. Unexpectedly, we observed through brightfield microscopy the presence of large hexagonal extracellular crystals (orange arrowheads) in the BAL (Fig. 2b) from LXR-DKO mice. These crystals resemble those observed in animal models with pulmonary lipidosis and inflammation, such as in Abcg1 -/- mice and whose biochemical nature could respond to the formation of extracellular complex lipoprotein aggregates [22, 23]. Additionally, the identification of cholesterol droplets by using Bodipy 493/503 fluorescent probe together with immunodetection of AM (CD68+) (Fig. 2b) revealed a large increase in the level of cholesterol accumulated by AMs in BAL from LXR-DKO mice. These data suggest that the origin of these foamy cells could be due to the accumulation of cholesterol in the surfactant material and/or the inability of macrophages to efflux intracellular cholesterol in the LXR-DKO lungs.

Furthermore, taking into account the crucial role of the AM in the homeostasis of the surfactant material and given that approximately 90% of the material is constituted by phospholipids, we performed a first approach to analyze the intracellular phospholipid load of the AM. The simultaneous detection of steatosis and phospholipidosis in vitro (Fig. 2b) revealed that LXR-deficient AMs that had turned into foam cells and accumulated neutral lipids (bodipy, green fluorescence) in the form of vacuoles in their cytoplasm showed weaker labeling of intracellular phospholipids (phospholipidosis, red fluorescence). This suggests that cholesterol accumulation by AMs in the LXR-DKO lung may lead to disruption of their phospholipid metabolism. Moreover, ultrastructural analysis of LXR-DKO lungs by electron microscopy revealed a clear morphological alteration of both T2Ps and AMs. T2Ps appeared hypertrophic with aberrant lamellar bodies (containing the de novo surfactant material) and exhibited increased size and electrodensity compared to LXR-WT lungs (Fig. 2c). While mitochondria (orange arrowheads) and other organelles are similar in size in both genotypes, lamellar bodies (red arrowheads) are the only organelles that show increased size in LXR-DKO T2Ps. In addition, AMs of LXR-DKO mice appeared hypertrophied compared to those in WT lungs (Fig. 2c) and showed large intracytoplasmic inclusions whose thin elongated shape suggests cholesterol accumulation in the form of “cholesterol needles” [24]. Likewise, counting of foamy and non-foamy AMs from BAL showed a progressive increase in the number of both total cells and foamy AMs in BAL from 3- to 9-month-old LXR-DKO mice compared to BAL from WT mice (Fig. 3a, b). OR-O-positive AMs (Fig. 3c, d) from LXR-DKO BAL showed lipid accumulation in the macrophage cytoplasm, confirming that LXR deficiency leads to the development of lipidosis in AM.

Fig. 3

Accumulation of lipid-laden AMs in LXR-deficient lungs. a: Representative images (n = 5) of BAL cells, at 3- (left panels), 6- (middle panels) and 9-month-old (right panels) WT and LXR-DKO mice, stained with H&E. Scale bars, 20 μm. b: Total cells, non-foamy and foamy AM in the BAL of 3- to 9-month-old WT and LXR-DKO were counted by contrast-phase brightfield microscopy. Data represent the mean and standard error of three experiments (n = 4 per genotype). Unpaired Student’s t test was used for two-group comparisons; **p < 0.01; ***p < 0.001; ****p < 0.0001 compared to WT group; #p < 0.05 compared to 3 months old and P < 0.05 compared to 6 months old mice. c: BAL cells from3- (left panels), 6- (middle panels) and 9-month-old (right panels) WT and LXR-DKO mice were subjected to Oil-Red-O staining and analyzed by contrast-phase brightfield microscopy. Scale bars, 20 μm. d, Representative images of BAL cells from 9-month-old WT and LXR-DKO mice (n = 4 per genotype) stained with Oil-Red-O (100x). Scale bars, 10 μm

To evaluate the origin of lipidosis, we analyzed the whole lung and BALF lipid profiles of LXR-DKO and WT mice by HPTLC. As shown in Table 1, the levels of esterified cholesterol -the biochemical storage form of cholesterol in the cell- were particularly elevated in both the lung and BALF of LXR-DKO mice, whereas triglyceride levels were significantly reduced in the lung of LXR-DKO mice. Besides neutral lipids, lipid analyses also showed a significant reduction in polar lipids, including sphingolipids and phospholipids (overall percentage of 19%). Thus, except phosphatidylcholine, all other phospholipids were significantly reduced in the lungs of LXR-DKO mice, revealing a massive dysregulation of surfactant lipid composition. Comparatively, changes in polar lipids were much less evident in BALF samples (average 4.9%), except phosphatidylglycerol, which was undetectable in BALF of LXR-DKO mice. Taken together, these results indicate that loss of LXR causes an alteration in the homeostasis of cholesterol and phospholipid components of surfactant material in the lung, which appear to affect both the ability of AMs to store cholesterol and of T2Ps to synthesize surfactant material via lamellar bodies.

Table 1 Altered lipid content in the lungs and BALF of LXR-DKO mice. Lipid levels were extracted and determined as described in material and methods. a, Lipid levels in WT or LXR-DKO lungs from 6-month-old mice. Data are expressed as mole percent (mol%) and represent the mean and standard error of mean (SEM). N = 5 (WT), N = 6 (LXR-DKO). Unpaired Student’s t test was used to analyze the statistical differences in lipid content between WT and LXR-DKO lungs. *p < 0.05; **p < 0.01; ***p < 0.001. b, Lipid levels in the BALF of 6-month-old WT or LXR-DKO mice. Data are expressed as mole percent (mol%) of pooled samples from N = 9 mice of each genotypeLXR regulates the expression of genes important for pulmonary surfactant homeostasis

Given the crucial roles of LXRs in lipid and inflammatory homeostasis in several tissues, we next studied the transcriptional activity of LXR in lung tissue. To explore this, we activated LXR in vivo by intraperitoneal injection (10 mg/kg, for 3 days) of the LXR agonist, GW3965, and then analyzed the expression levels of established LXR target genes such as Abca1, Abcg1 and Srebf1 [25]. As shown Fig. S2, the transcript levels of these genes, analyzed by real-time RT-QPCR, were significantly induced in WT mice in response to GW3965, indicating that LXR activity is inducible in the lung by an acute administration of synthetic LXR agonist.

Considering the central role of LXR in the maintenance of lipid homeostasis as cholesterol sensors, the expression of those LXR target genes whose dysregulation could be involved in the development of the LXR-DKO phenotype was analyzed, in whole lung tissue and BAL cells by real-time RT-QPCR. As shown in Fig. 4a, loss of LXR resulted in a large decrease in ABCA1 and ABCG1 mRNA levels in lung and ABCA1 in BAL cells, consistent with the cholesterol accumulation observed in LXR-deficient lung cells (AM and T2P) and BALF. Furthermore, the expression of ACAT1 (acetyl-CoA-acetyltransferase 1) and LCAT (lecithin cholesterol-transferase), enzymes involved in cholesterol esterification, was elevated in both lung and BAL cells of the LXR-DKO compared to WT. SREBP-1C-dependent lipogenic pathway was also reduced in lung LXR-DKO cells. In contrast, mRNA levels of glycerol-3-phosphate acyltransferase (GPAT), a limiting enzyme of the de novo pathway of glycerolipid synthesis that plays a key role in regulating triglyceride and phospholipid synthesis, were significantly increased in both lung tissue and BAL cells. Thus, loss of LXR causes the accumulation of cholesterol inside the cells and the derepression of enzymes necessary for the transformation of free cholesterol into esterified cholesterol. Overall, these results are consistent with the altered levels of cholesterol, triglycerides and phospholipid components of the surfactant material found in BALF and lung tissue from LXR-DKO mice.

Fig. 4

Impaired pulmonary surfactant homeostasis gene expression in LXR-deficient mice. Transcript levels of genes involved in lipid (a) and protein (b) metabolism of pulmonary surfactant from BAL cells (BALC) and lung tissue from WT and LXR-DKO 6-month-old mice were analyzed by real-time qPCR. WT gene expression values were normalized to 1 and used as control (dashed line). Data represent the mean and standard error of three experiments (n = 4 mice per genotype). Unpaired Student’s t test with Welch’s correction was used for two-group comparisons; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 compared to WT group

Our results presented above showed that T2Ps displayed aberrant lamellar bodies in LXR deficient mice. Since these subcellular structures are the origin of extracellular surfactant fluid, we next analyzed by RT-qPCR the expression of surfactant component proteins in lung tissue from WT and LXR-DKO mice. Although SP-A and SP-B mRNA levels remain unchanged (Fig. 4b), SP-C and SP-D levels were significantly altered in LXR-DKO lung compared to WT lung, suggesting that SP-C and SP-D protein expression may be under the control of LXR activity and that the absence of LXR in the lung results in altered protein components of surfactant. Furthermore, since SP-C is synthesized mainly by T2Ps, these data suggest that this cell type is likely involved in the development of the LXR-DKO phenotype. Taken together, our results indicate that, in addition to altered expression of several enzymes that regulate cholesterol homeostasis and phospholipid synthesis in the lung, loss of LXR results in abnormal expression of surfactant component proteins, leading to de novo synthesis of defective surfactant material by T2Ps.

Role of alveolar macrophages in lung histopathology in LXR DKO mice

LXRα-specific transcriptional activity has been reported to be critical for macrophage development in the liver and spleen [20, 26]. Our results presented above have shown that deletion of both LXRα and LXRβ was required to develop an accumulation of foamy AMs in the lungs. To evaluate the contribution of macrophage LXR activity to the development of the lung phenotype of LXR-DKO mice, we performed bone marrow (BM) transplant (BMT) experiments to achieve AM replacement in the lung. First, we confirmed that 12 weeks post-transplant was an adequate time at which nearly all AM in recipient mice had been replaced by the donor BM by using BM progenitors from Csf1r-EGFP (MacGreen) transgenic mice, which contain a bone marrow-derived myeloid lineage expressing the green fluorescent protein (EGFP). Double immunofluorescent detection (Fig. S3a) in lung sections of mice transplanted with Csf1r-EGFP bone marrows revealed GFP signal co-localizing with the CD68 + macrophages, indicating successful replenishment of AMs in recipient mice. Next, we performed reciprocal BMTs using bone marrow from either WT or LXR-DKO mice and transferred into the indicated irradiated recipient mice. After 24 weeks post-transplant, consecutive lung sections were analyzed by histological techniques with hematoxylin-eosin and OR-O. Unexpectedly, transplant of bone marrow from wild-type donors did not ameliorate lipidosis and congestion in the subpleural zone of the lung of LXR-DKO recipient mice. As shown in Fig. 5, LXR-DKO recipient lungs showed infiltrates and oil-red positive cells within the subpleural area (Fig. 5d, h), similar to LXR-DKO control transplanted lungs (Fig. 5b, f). Consistent with this, a reciprocal transplant of BM from LXR-DKO donors resulted in minimal lipid accumulations in the subpleural zone (blue arrows) of the lung of WT mice (Fig. 5c, g). Finally, to rule out the possibility that the inability of WT transplanted macrophages to ameliorate the lipidosis present in LXR-deficient mice was due to unproductive replacement of the pre-existent foamy LXR-DKO AMs present in congested subpleural areas, we performed additional BMT experiments using Csf1r-EGFP donor BM transplanted into LXR-DKO mice. Analysis of GFP expression in consecutive lung sections confirmed the presence of GFP + macrophages in the subpleural area of the lung, thus indicating the overall efficacy of the transplantation. As shown in Fig. S3b, the subpleural area exhibited GFP signal in both WT and LXR-DKO mice. Moreover, this signal was coincident with that of oil red-positive cells (foamy AMs) in LXR-DKO recipient lungs (red dashed line), confirming that myeloid cells have been successfully replaced in the lung tissue. Altogether, these experiments indicated that, although macrophages participate in the pathogenesis of the observed phenotype, the contribution of LXR deficit in AMs cannot be considered exclusively as the main cause of the extensive lipidosis and inflammation observed in LXR-DKO mice.

Fig. 5

Lung histopathology and lipid accumulation in WT and LXR-DKO mice after bone marrow transplant. Morphology (H&E) and lipidosis (Oil-Red-O staining) were assessed in consecutive lung sections from wild-type and LXR-DKO mice 16 weeks after reciprocal bone marrow transplantation (labels: donor→recipient). Scale bars, 50 μm. Images are representative of two independent experiments with five to six mice per group

Deficient regulation of surfactant metabolism by type II pneumocytes in LXR-DKO mice

Our results so far suggested that tissue cells other than myeloid cells, such as surfactant fluid-producing T2Ps, may be responsible for lung histopathology in LXR DKO mice. First, we analyzed LXR activity in a murine cell model of T2Ps, using the MLE-12 pneumocyte cell line (CRL-2110™), to clarify whether LXR activity could be involved in the processing of pulmonary surfactant by T2Ps. Therefore, we examined the expression of LXRα and LXRβ in MLE-12 cells (Fig. S4) using peritoneal macrophages as a cell reference. As shown in Fig. S4a, MLE-12 cells expressed LXRα and LXRβ mRNA, although at lower levels than peritoneal macrophages. In addition, unlike peritoneal macrophages, MLE-12 cells showed similar levels of LXRα and LXRβ mRNA (Fig. S4a) and proteins (Fig. S4c). Subsequent analysis of the expression of ABCA1 in MLE-12 cells showed that its mRNA (Fig. S4b) and protein (Fig. S4c) levels increased in response to GW3965, whereas the mRNA levels decreased dramatically when the cells were co-treated with the antagonist GW233 (Fig. S4b). Analysis of pro-SP-C protein expression (Fig. S4c) by western-blot in MLE-12 cells confirmed expression of the surfactant protein and its levels did not change in response to GW3965. These results indicate that both LXRs are present and active in the MLE-12 pneumocytes, which expresses the pro-SP-C surfactant protein in culture and therefore represents a suitable model for further analysis.

Subsequently, we modeled in vitro the environmental conditions of lipidosis in which LXR-DKO lung pneumocytes are found by adding bronchoalveolar fluid to cultured cells. Initially, the effect of BALF obtained from WT and LXR-DKO lungs on WT peritoneal macrophage cultures was analyzed. The percentage of foamy cells was determined by using Bodipy 493/503, after 24, 48 and 72 h incubation of macrophages with fractions of BALF. As shown Fig. 6a, the presence of WT BALF resulted in low lipid accumulation in macrophages at 72 h. Importantly, the magnitude and timing of foamy cell formation was robustly observed earlier when macrophages were cultured in the presence of LXR-DKO BALF (Fig. 6a), emphasizing the lipid accumulation inside the macrophages observed from 24 h to 72 h of incubation with BALF LXR-DKO. These results indicate that bronchoalveolar fluid from LXR-DKO animals is capable of inducing macrophage-lipidosis in culture. We then hypothesized that we could translate this model to drive lipid handling in vitro in T2Ps, cells with the capacity to uptake and recycle surfactant material. Thus, we cultured MLE-12 cells with LXR-DKO BALF in the presence of GW3965 agonist or G233 GW233 antagonist and measured lipid uptake by analyzing bodipy 493/503 signal. As shown in Fig. 6b, pneumocytes cultured with BALF experienced a slight increase in lipid accumulation in their cytoplasms, as did peritoneal macrophages, confirming the validity of this model with MLE-12 cells. Interestingly, lipid uptake was strongly enhanced in the presence of GW3965 agonist, an effect that was completely reversed when MLE-12 cells were co-treated with GW233 antagonist (Fig. 6b). These data indicate that LXR activation is important for the uptake and handling of surfactant material by T2Ps expressing LXRs. Data also suggests that diminished LXR activity could lead to defective surfactant handling by T2Ps and, consequently, accumulation of this material in the alveolar space.

Fig. 6

Modulation of LXR activity affects the handling of surfactant material by MLE12 cells. In vitro lipidosis model: Peritoneal macrophages from WT mice were cultured and exposed to BALF from WT and LXR-DKO mice. Subsequently, (a) fluorescence microscopy analysis and percentage of foam cells were determined using Bodipy 493/503 after 24, 48 and 72 h incubation of peritoneal macrophages with BALF. A representative image is shown (left panel) and the data are presented as the mean and standard error (right panel), n = 4 mice per genotype. Kruskal-Wallis with a post hoc Dunn’s test was used for multiple comparisons; ****p < 0.0001 and **p < 0.001 compared to control group; #P < 0.05 and ##P < 0.001 compared to BALF WT groups. (b) MLE-12 cells were pre-incubated with GW3965 agonist or GW233 antagonist for 24 h and then exposed to LXR-DKO BALF for additional 24 h. Lipid uptake was measured by bodipy 493/503 signal (A-G) or bright field (B-H) using confocal microscopy. Cells not exposed to BALF were used as control. A representative image of 3 independent experiments is shown. Scale bars of all images are 50 μm

The results obtained with MLE-12 lung epithelial cells led us to study the role of LXRs in the regulation of surfactant material by primary T2Ps. These cells were isolated from mice using enzymatic digestion, immunomagnetic purification and cultured on a matrigel-coated surface. Monitoring of pro-SPC expression by immunocytochemistry revealed an enrichment of the T2Ps population by 80–95% (Fig. 7a) during routine isolation of these cells from murine lungs. Furthermore, RT-qPCR analysis of the expression of LXR target genes (Fig. 7b, upper panel) such as Abca1, Abcg1 and Srebf1 in primary pneumocytes isolated from LXR-DKO mice showed significantly reduced mRNA levels compared to WT mice, confirming that LXR is relevant for the expression of these genes in primary cells. Next, we analyzed other genes involved in lipid and protein synthesis related to the production and recycling of surfactant material by T2Ps. While ACAT1 and LCAT mRNA levels did not change, importantly, ABCA3 (membrane transporter of different classes of lipids associated with lamellar bodies) and LPCAT1/3 (lysophosphatidyl acyltransferases involved in the recycling or remodeling pathway, Land’s cycle, of phospholipids) mRNAs were drastically diminished in T2Ps from LXR-DKO mice compared to the WT (Fig. 7b, upper and middle panel). Furthermore, the expression levels of the surfactant component proteins, SP-A, but especially SP-B and SP-C, were significantly decreased in the LXR-deficient T2P (Fig. 7b, lower panel). Taken together, these results demonstrate that LXR is present in primary T2PS and its activity is crucial for the correct synthesis and recycling of surfactant material, and suggest that defective LXR transcriptional activity by the T2P could be the main cause of the lung histopathology found in LXR-DKO mice.

Fig. 7

LXR deficiency alters primary type 2 pneumocytes gene expression. a, T2Ps were isolated from WT and LXR-DKO 6-month-old mice by enzymatic digestion and immunomagnetic purification, and subsequently cultured on a Matrigel-coated surface. Enrichment of T2P through the isolation method was monitored by pro-SPC expression by immunocytochemistry in the total lung homogenate (upper panel), after recovery of non-adherent cells (middle panel) and at the end point of isolation after the positive selection column (low panel). A representative image (n = 4) is shown. Scale bars of all images are 50 μm. b, Transcript levels of genes involved in lipid (upper and middle panels) and protein (low panel) metabolism of pulmonary surfactant from T2P, isolated from WT and LXR-DKO 6-month-old mice, were analyzed by real-time qPCR. WT gene expression values were normalized to 1 and used as control (dashed line). Data represent the mean and standard error of three experiments (n = 4 mice per genotype). Unpaired Student’s t test with Welch’s correction was used for two-group comparisons; **p < 0.01; ***p < 0.001; ****p < 0.0001 compared to the WT T2P group

Inflammation and impaired pulmonary function in LXR DKO lungs

The alteration of lung morphology in LXR-deficient mice, characterized by the progressive development of lipidosis and infiltration of immunocompetent cells, are pathological events associated with an inflammatory process that could affect lung function. The development of the inflammatory process was confirmed by the subsequent detection of elevated levels of immunoglobulins IgA, IgG and IgM in perfused LXR-DKO lung tissue (Fig. 8a). These data indicate that the numerous leukocyte infiltrates, previously observed in the lung tissue of LXR-DKOs, may be producing these immunoglobulins and contributing to the proinflammatory environment. Furthermore, the expression of genes whose overexpression is related to the development of an inflammatory process: Spp1 (SPP-1, secreted phosphoprotein-1), Mip-1β (MIP-1β, macrophage inflammatory protein-1β), Mmp-8 and Mmp-12 (MMP-8 and MMP-12, matrix metalloproteinase-8 and − 12), was analyzed in WT and LXR-DKO lungs. As shown in Fig. 8b, MMP-12, SPP-1 and MIP-1β expression levels were increased, in LXR-DKO lungs. These metalloproteinases have been shown to promote the recruitment of macrophages and other immune cells to sites of inflammation and fibrosis [27, 28].

Fig. 8

LXR inactivation leads to a chronic inflammatory process in the lung. a, Whole cell lysates were prepared from homogenized lung tissue cells from perfused WT and LXR-DKO 6-month-old mice and IgA, IgM or IgG proteins were analyzed by Western blot. Membranes were stripped and reprobed with β−actin antibody as loading control. A representative western blot (n = 3 mice) is shown. b, Transcript levels of inflammatory related-genes, from homogenized lung tissue cells from WT and LXR-DKO 6-month-old mice, were analyzed by real-time qPCR. WT gene expression values were normalized to 1 and used as control (dashed line). Data represent the mean and standard error of three experiments (n = 3 mice per genotype). Unpaired Student’s t test with Welch’s correction was used for two-group comparisons; *p < 0.05; **p < 0.01; ***p < 0.001 compared to WT group. c, Masson’s trichrome (MT) staining (left panels) and periodic acid-Schiff (PAS) staining (middle panels) were evaluated in consecutive lung sections from 6-month-old WT and LXR-DKO mice. Pretreatment of tissues with PAS-diastase (α-salivary amylase) (PAS-D) (right panels) was assessed to confirm PAS-positive material. A representative image of 4 mice per genotype is shown. Scale bars of all images are 20 μm

To evaluate if this inflammation also results in fibrosis in the lungs of LXR-DKO, we performed Masson’s trichrome staining of lung sections. As shown in Fig. 8c, collagen fibers (blue color), in WT lungs, were present forming part of a thin connective tissue layer accompanied by laxly arranged fibroblasts (nuclei, purple color), which constitute the alveolar septa, separating and supporting the alveolar sacs. In contrast, the stroma constituting the alveolar septa appeared congested (Fig. 8c, double green arrows) in LXR-DKO lungs, constituted by increased cellularity and thick collagen fibers that patently occupy almost the entirety of the alveolar septa. Moreover, the alveolar space was enlarged in LXR-DKO mice due to a nearly complete occupied space by numerous AMs (nuclei, purple color, cytoplasm, red color), which were foamy and hypertrophied. Next, we analyzed by periodic staining with Schiff acid (PAS) the possible accumulation of glycoprotein components of the surfactant material in the alveolar space because of an alteration in its synthesis and recycling in LXR-DKO lungs. As shown in Fig. 8c, staining detected PAS-positive material filling the alveolar spaces of LXR-DKO lungs (pink color) compared to WT, corresponding to accumulated surfactant material. In addition, this staining revealed PAS-positive material in the cytoplasm of AMs (Fig. 8c, green arrowhead), coincident with the OR-O-positive foamy macrophages observed previously. Pretreatment of tissues with PAS-D diastase (α-salivary amylase) confirmed that it is indeed a PAS-positive material (Fig. 8c). Taken together, these results suggest that inactivation of LXR leads to a chronic inflammatory process in the lung as a consequence of mismanagement of surfactant biosynthesis, ultimately resulting in progressive alveolar proteinosis aggravated with age.

Finally, we used a mouse model of house dust mite (HDM)-induced allergic airway inflammation to analyze whether the baseline lipid and immune defects developed by LXR-DKO mice could affect HDM-challenged lung functions [29]. To this purpose, WT and LXR-DKO mice were sensitized through intranasal instillations of an HDM extract for 10 days and then the airway reactivity to a methacholine challenge and the lymphocytic infiltration in the lungs were analyzed. HDM-exposed LXR-DKO mice presented more pronounced airway resistance in response to methacholine challenge (Fig. 9a) and increased infiltration in perivascular and peribronchiolar areas (Fig. 9b, c), compared with HDM-exposed controls. Next, WT mice were stimulated with an LXR agonist, before HDM exposure and during the entire course of sensitization. Pretreatment with GW3965 markedly improved the lung response to methacholine challenge in WT mice exposed to HDM (Fig. 9d). These results indicate that LXR pharmacological activation might be useful as a therapeutic alternative in pulmonary allergic inflammatory diseases. Taken together, our results demonstrate that LXR is involved in the control of surfactant synthesis and recycling by T2Ps and AMs, and that LXR deficiency promotes a chronic inflammatory process that impairs lung function.

Fig. 9

LXR pharmacological activation ameliorates lung reactivity in mice exposed to HDM. WT or LXR-DKO 3-month-old mice were administered with HDM extract daily intranasally at a dose of 25 µg/mouse for 10 consecutive days. a, Lung resistance (RL) to increasing doses of methacholine was assessed 24 h after the last HDM exposure. Data represent the mean and standard error of n = 5 (WT) or n = 4 (LXR-DKO) mice. The effect of genotype was analyzed using one-way ANOVA with a post hoc Bonferroni test; * p < 0.05. b, Histological analysis of the number (left panel) of perivascular and/or peribronchiolar inflammatory infiltrates in the lung. The lung inflammation score (right panel) was calculated as described in Methods. Unpaired Student’s t test; *p < 0.05, **p < 0.01 compared to WT group. c, Representative images from the lungs of HDM-exposed WT or LXR-DKO mice. d, 8- week- old WT mice were separated in three groups: non-sensitized mice (Saline group, N = 11) received daily an intranasal administration of physiological saline and an intraperitoneal administration of vehicle (DMSO in PBS); HDM-sensitized mice received daily an intranasal administration of HDM and an intraperitoneal administration of either vehicle (DMSO in PBS) (HDM group, N = 10) or of the LXR agonist GW3965 (20 mg/kg) (GW3965 + HDM group, N = 11). Administration of the LXR agonist or vehicle was initiated one day before the start of sensitization. Lung resistance (RL) to increasing doses of methacholine was assessed 24 h after the last HDM exposure. Pooled data from two independent experiments. The effect of treatment was analyzed using one-way ANOVA with a post hoc Bonferroni test. Mean ± SEM; ###P < 0.0001 compared to saline group; *p < 0.01 compared to HDM group