Hhip haploinsufficiency sensitizes mice to age-related emphysema

Significance

Genome-wide association studies (GWAS) have been very successful in discovering genetic loci associated with complex traits. However, only few studies applied murine models to investigate how GWAS genes contribute to human lung diseases. Motivated by GWAS linking Hedgehog interacting protein (HHIP) to emphysema and impairments in lung function, this study demonstrated that Hhip^+/− mice developed spontaneous emphysema and lung function impairment over time. Moreover, emphysema, associated with increased oxidative stress in Hhip^+/− lungs, was prevented by treating the mice with the antioxidant, N-acetyl cysteine (NAC). This post-GWAS functional study connects aging-related diseases, molecular mechanisms, and potential therapy in a genetic haploinsufficient murine model, which may lead to improvements in understanding pathophysiologic concepts of alveolar loss related to aging.

Abstract

Genetic variants in Hedgehog interacting protein (HHIP) have consistently been associated with the susceptibility to develop chronic obstructive pulmonary disease and pulmonary function levels, including the forced expiratory volume in 1 s (FEV₁), in general population samples by genome-wide association studies. However, in vivo evidence connecting Hhip to age-related FEV₁ decline and emphysema development is lacking. Herein, using Hhip heterozygous mice (Hhip^+/−), we observed increased lung compliance and spontaneous emphysema in Hhip^+/− mice starting at 10 mo of age. This increase was preceded by increases in oxidative stress levels in the lungs of Hhip^+/− vs. Hhip^+/+ mice. To our knowledge, these results provide the first line of evidence that HHIP is involved in maintaining normal lung function and alveolar structures. Interestingly, antioxidant N-acetyl cysteine treatment in mice starting at age of 5 mo improved lung function and prevented emphysema development in Hhip^+/− mice, suggesting that N-acetyl cysteine treatment limits the progression of age-related emphysema in Hhip^+/− mice. Therefore, reduced lung function and age-related spontaneous emphysema development in Hhip^+/− mice may be caused by increased oxidative stress levels in murine lungs as a result of haploinsufficiency of Hhip.

Aging (senescence) increases vulnerability to age-associated diseases, such as chronic obstructive pulmonary disease (COPD), the third leading cause of death in the United States (1). COPD is also strongly influenced by cigarette smoke (CS) and genetic predisposition (2, 3).

Genome-wide association studies (GWAS) have successfully identified susceptibility loci for many complex diseases in the last decade (4, 5). However, understanding the biology underlying these GWAS associations remains a major challenge (6, 7). Mechanistic studies on novel susceptibility genes in complex traits, including COPD and pulmonary function levels, may provide insights into disease pathogenesis and thus highlight potential targets for therapy. Through GWAS, the chromosome 4q31 Hedgehog interacting protein (HHIP) locus has been consistently identified in COPD (8–10) and in spirometric measures, including the forced expiratory volume in 1 s (FEV₁) and ratio of FEV₁ to forced vital capacity in general population samples (11–14). Hhip represses hedgehog signaling by competitive binding with three Hedgehog (HH) ligands: Sonic Hedgehog, Indian Hedgehog, and Desert Hedgehog (15). Deletion of the Hhip gene in mice is neonatally lethal because of defective branching in the lung (14, 15). The reduced expression of HHIP in COPD lungs, and reduced enhancer activity associated with COPD risk alleles at the HHIP locus (16), suggested a protective role of HHIP in COPD pathogenesis.

Cellular redox (reduction-oxidation) homeostasis is necessary to protect cells from internal and external environmental stresses, thus maintaining normal cellular functions. During aging, reactive oxygen species (ROS) production and oxidant burden increase in lungs, contributing to parenchymal lung destruction and emphysema (17, 18). Pulmonary emphysema may result from accelerated and premature aging of the lungs because of cellular senescence (19), consistent with the observation that most COPD patients develop disease at advanced ages. Deficiency of cellular guardian genes, such as SIRT1 (sirtuin 1) (20) and TLR4 (Toll receptor 4) (21), leads to spontaneous pulmonary emphysema in murine models because of increased oxidative stress and accelerated aging.

We have previously shown that Hhip heterozygous knockout mice (Hhip^+/−) developed more severe emphysema than wild-type littermates when exposed to CS (22), suggesting that HHIP protects lung cells from environmental stressors, such as CS exposure. To extend our findings and understand the roles of HHIP in regulating lung function in nonsmokers, Hhip^+/− mice were evaluated in an aging model. Lung quasi-static compliance, pressure–volume (PV) flow loops, and lung morphology were assessed over time, and molecular mechanisms by which Hhip regulates lung redox homeostasis and determines susceptibility to age-related emphysema were investigated.

Results

Hhip Haploinsufficiency Leads to Spontaneous Age-Dependent Airspace Enlargement.

Given that the HHIP locus is consistently associated with pulmonary function levels in the general population, we assessed lung morphology and mechanics in Hhip^+/− and Hhip^+/+ mice at different ages without smoke exposure. Hhip^+/− mice have normal lung development (14) and normal distal airspace size at up to 8 mo of age. However, starting at 10 mo of age, Hhip^+/− mice demonstrated significant age-dependent increases in airspace size in contrast to Hhip^+/+ mice (Fig. 1) (unpaired t test, P < 0.01), suggesting increased susceptibility of Hhip^+/− mice to age-related airspace enlargement. There were also strong interactions between genotype and age of mice, suggesting genotype- and age-dependent airspace enlargement in murine lungs (Fig. 1B) (two-way ANOVA, P < 0.01).

Fig. 1.

Age-dependent airspace enlargement in Hhip^+/− mice. (A) Histology of lung sections from Hhip^+/− vs. wild-type littermate control mice (Hhip^+/+) at different ages as demonstrated by H&E staining. (Scale bars, 60 μm.) (B) Mean chord length (MCL) measurements on airspace size in murine lung sections from Hhip^+/+ (+/+) and Hhip^+/− (+/−) mice. Means ± SEM shown in B are from 5 to 14 mice per group. *P < 0.05 and **P < 0.01 by unpaired t test. There is a significant interaction between genotype and ages of mice by two-way ANOVA (P < 0.01).

Hhip Haploinsufficiency Accelerated Age-Dependent Lung Compliance Increases.

Consistent with airspace enlargement, Hhip^+/− mice showed significantly increased lung compliance and decreased total lung elastance compared with age-matched Hhip^+/+ mice at 10 and 12 mo of age (Fig. 2 A and B) (unpaired t test, P < 0.05). Furthermore, Hhip^+/− mice showed left shifts in the PV loops starting at 10 mo of age, suggesting that Hhip haploinsufficiency led to increased quasi-static lung compliance (Fig. 2C) (unpaired t test at each pressure point, P < 0.05). These age-related impairments in respiratory mechanics were associated with progressively reduced expression of Hhip in Hhip^+/+ murine lungs and even greater reductions in Hhip expression in Hhip^+/− lungs during aging (Fig. S1) (23), supporting protective roles of Hhip during aging.

Fig. 2.

Lung mechanics measurements in Hhip^+/+ and Hhip^+/− mice at 10 and 12 mo (M) of age. (A) Lung compliance measurements. (B) Tissue elastance measurements. (C) Lung volume–pressure curve measurements. *P < 0.05 and **P < 0.01 by unpaired t test at each pressure point to compare mean volume in Hhip^+/+ and Hhip^+/− mice. Hhip^+/− mice at 10 and 12 mo of age had left shifts in their pressure–volume curves compared with Hhip^+/+ mice at the same age, indicating that the Hhip^+/− mice have higher quasi-static lung compliance. Means ± SEM are shown for each group (n = 5–7 mice per group).

Fig. S1.

Reduced expression of Hhip in murine lungs during aging. (A) Expression levels of Hhip in lungs from C57BL/6J and DBA/2J mice. Data were obtained from the microarray online database (https://www.ncbi.nlm.nih.gov/geo/) under accession no. GSE6591. (B) Expression of Hhip in lungs measured by RT-PCR from Hhip^+/+ and Hhip^+/− mice at different ages. *P < 0.05 and **P < 0.01 by unpaired t test. M, months.

Increased Lymphoid Aggregates in the Lungs of Hhip^+/− Mice During Aging.

Aging lungs may exhibit signs of chronic inflammation (24), and Hhip^+/− mice demonstrated increased numbers of lymphoid aggregates in lungs after chronic CS exposure (22); we therefore assessed lymphoid aggregates in murine lungs at different ages. The number of peri-bronchial lymphoid aggregates was significantly increased in Hhip^+/− mice (Fig. 3 A and B) (unpaired t test, P < 0.05) at 10 and 12 mo of age compared with age-matched Hhip^+/+ mice. Consistently, the expression of CXCR3 (C-X-C chemokine receptor type 3) and CXCR5 were also significantly increased in Hhip^+/− lungs at 10 mo of age as assessed by RT-PCR (Fig. 3 C and D) (unpaired t test, P < 0.05). CXCR5 is crucial for B-cell migration, whereas CXCR3 induces the migration of activated T cells into the lungs (25, 26). Interestingly, CXCL10 (C-X-C motif chemokine 10), a ligand for CXCR3, and MCP1 (monocyte chemotactic protein 1) also showed increased levels in Hhip^+/− compared with Hhip^+/+ lungs at 8 mo of age (Fig. 3 E and F) (unpaired t test, P < 0.05). These results indicate increased recruitment of B and T lymphocytes into the lungs in Hhip^+/− mice at 8–10 mo of age. However, total cell counts in bronchoalveolar lavage did not show significant difference between Hhip^+/− and Hhip^+/+ mice at 10 mo of age (Fig. 3G).

Fig. 3.

Increased lymphoid aggregates in the lungs of Hhip^+/− mice during aging. (A) Quantification of lymphoid aggregates per airway (mean internal diameter of 200–1,000 μm) in Hhip^+/+ mice (+/+) and Hhip^+/− mice (+/−) at different ages. Data are means ± SEM from five to nine mice per group. (B) H&E staining in murine lungs. (Scale bars, 60 μm.) (C and D) Gene expression of CXCR3 and CXCR5 in lungs from Hhip^+/+ (+/+) and Hhip^+/− (+/−) mice. Mean ± SEM are shown. (E and F) Measurements on CXCL10 and MCP1 levels in lungs from Hhip^+/+ (+/+) and Hhip^+/− (+/−) mice at 8 mo (M) of age. Data are means ± SEM from 11 to 15 mice per group. (G) Total cell counts in bronchoalveolar lavage from mice at 10 mo of age. *P < 0.05; **P < 0.01 by unpaired t test. N.S., nonsignificant.

Increased Levels of Matrix Metallopeptidases in the Lungs of Hhip^+/− Mice.

Matrix metallopeptidases (MMPs) promote emphysema development in both human COPD patients and murine emphysema models (27); thus, we compared levels of MMPs in lungs from Hhip^+/+ and Hhip^+/− mice. MMP-9 protein levels were significantly increased in Hhip^+/− lungs at 12 mo of age (Fig. S2A) (unpaired t test, P < 0.05). MMP-12 levels were significantly affected by Hhip genotype with a trend toward increased levels in Hhip^+/− lungs (Fig. S2B) (two-way ANOVA, P < 0.05). Lung levels of the collagenase, MMP-8, also showed interaction between genotype and age time points (Fig. S2C) (two-way ANOVA, P < 0.05). However, no differences in levels of MMP-2 were detected in Hhip^+/− and Hhip^+/+ mice at either 8 or 12 mo of age (Fig. S2D).

Fig. S2.

(A–D) Increased levels of matrix metallopeptidases in the lungs of Hhip^+/− mice. Quantification of the MMPs (MMP-9, -2, -8, and -12) in Hhip^+/+ mice (+/+) and Hhip^+/− mice (+/−). MMP-9 was measured by ELISA and MMP-2, -8, and -12 were measured by Luminex-based assay. Means ± SEM were shown from three to six mice per group. *P < 0.05, unpaired t test. MMP-12 showed genotypic difference by one-way ANOVA analysis, P < 0.05.

Increased Cell Death and Cell Senescence in Hhip^+/− Lungs.

As alveolar septal cell apoptosis contributes to emphysema development (28), we also measured cell death in lungs. Significantly increased cell death as assessed by TUNEL staining was detected in lungs from Hhip^+/− mice compared with Hhip^+/+ mice at 8 mo of age (Fig. S3 A and B) (unpaired t test, P < 0.05). However, no difference in cell proliferation was detected in Hhip^+/− vs. Hhip^+/+ lungs as assessed by Ki67 staining (Fig. S3 C and D). Furthermore, we assessed cellular senescence markers in murine lungs. Increased p53 and p21 levels were detected in Hhip^+/− lungs compared with Hhip^+/+ lungs at 10 mo of age (Fig. S4) (unpaired t test, P < 0.05).

Fig. S3.

Increased cell death in the lungs of Hhip^+/− mice. Quantification (A) and representative images (B) of TUNEL staining (green signals) representing dead cells in Hhip^+/+ (+/+) and Hhip^+/− (+/−) mice. Data are means ± SEM from 14–15 mice per group. *P < 0.05, unpaired t test. Quantification (C) and representative images (D) of the Ki67 staining (green signal) in Hhip^+/+ (+/+) and Hhip^+/− (+/−) mice (n = 3 mice per group). Nuclei were stained with DAPI. (Scale bars, 60 μm.)

Fig. S4.

Increased senescence in lungs from Hhip^+/− mice. Measurements on protein levels of p53 and p21 in lungs from Hhip^+/+ (+/+) and Hhip^+/− (+/−) mice at 2 and 10 mo of age by western blot (A) followed by quantification based on band density (B). *P < 0.05 and **P < 0.01 by unpaired t test. GAPDH was used as the loading control.

Increased Oxidative Burden in Hhip^+/− Lungs.

To identify pathways driving spontaneous emphysema in Hhip^+/− mice, we assessed genes differentially expressed in 8-mo-old (preceding phenotypic changes) Hhip^+/− and Hhip^+/+ murine lungs (22). All 48 up-regulated genes in Hhip^+/− vs. Hhip^+/+ mice were significantly enriched in biological oxidation pathways by gene set enrichment analysis (false-discovery rate is 1.20E-04) (Table S1). A subset of selected genes validated by quantitative RT-PCR included Mt3 (metallothionein 3) and Adh7 (alcohol dehydrogenase 7) (Fig. S5 A–D). Increased expression of genes enriched in xenobiotic pathways suggested a pro-oxidant state in Hhip^+/− murine lungs at 8 mo of age.

Table S1.

Differentially expressed genes related to the biological oxidation pathway revealed in microarray analysis in Hhip^+/− vs. Hhip^+/+ mice at 8 mo of age

Gene symbol	Definition
Adh7	Alcohol dehydrogenase 7 (class IV)
Mt3	Metallothionein 3
Fmo3	F containing monooxygenase 3
Cyp3a13	Cytochrome P450, family 3, subfamily a, polypeptide 13
Cyp2a5	Cytochrome P450, family 2, subfamily a, polypeptide 5
Ptgis	Prostaglandin I2 (prostacyclin) synthase
Prdx6	Peroxiredoxin 6
Xdh	Xanthine dehydrogenase
Fmo1	Flavin containing monooxygenase 1
Cyp2b10	Cytochrome P450, family 2, subfamily b, polypeptide 10
Cbs	Cystathionine beta-synthase
Cyp4b1	Cytochrome P450, family 4, subfamily b, polypeptide 1

Fig. S5.

Increased oxidative burden in murine lungs from Hhip^+/− mice. Expression levels of genes associated with xenobiotic pathway: Adh7 (A), Mt3 (B), Hs3st1 (C), and Tef (D) measured by RT-PCR in lungs from 8-mo-old mice. Means ± SEM shown from six mice per group. (E) GSH and GSSG ratio (GSH/GSSG) in murine lungs from 9-mo-old mice. **P < 0.01 and *P < 0.05 by unpaired t test. (F) Total antioxidant capacity measurements in murine lungs. Means ± SEM shown from n = 3–6 mice per group. A significant interaction between genotype and age (two-way ANOVA, P < 0.01) and a significant age-dependent reduction of antioxidant capacity in Hhip^+/− (one-way ANOVA, P < 0.01) were detected.

To further confirm that Hhip^+/− mice have increased lung oxidative stress levels, we measured reduced glutathione (GSH) and oxidized GSH (GSSG) in lung lysates from Hhip^+/+ and Hhip^+/− mice. A markedly lower GSH/GSSG ratio was found in lung lysates from 9-mo-old Hhip^+/− mice compared with Hhip^+/+ mice (Fig. S5E) (unpaired t test, P < 0.05). Consistently, the antioxidant capacity in lungs showed a strong age-related reduction in Hhip^+/− mice (Fig. S5F) (one-way ANOVA, P < 0.01).

Hhip, Expressed in Alveolar Type II Cells, Regulates Cellular Oxidant Stress.

Immunofluorescence staining of lung sections revealed that Hhip colocalizes with SPC (surfactant protein C), a marker of alveolar type II (AT II) cells (Fig. 4A), with a reduced level of Hhip in Hhip^+/− mice (Fig. 4B). However, no differences in the numbers of AT II cells (quantified by SPC staining) were detected between Hhip^+/− and Hhip^+/+ mice at 2 mo of age (Fig. 4C).

Fig. 4.

Hhip regulates oxidative stress in murine alveolar type II (AT II) epithelial cells. (A) Immunofluorescence staining of Hhip in murine lungs from a wild-type C57BL/6 mouse. White arrowheads indicated colocalization of Hhip and SPC in AT II cells. (Scale bars, 170 μm.) (B) Immunoblotting of Hhip in AT II isolated from Hhip^+/+ (+/+) and Hhip^+/− (+/−) mice. (C) Quantification of SPC⁺ cells in lungs from age-matched Hhip^+/− and Hhip^+/+ mice at 2 mo of age (five mice per group). (D) Intracellular ROS levels in AT II cells under H₂O₂ treatment. **P < 0.01, unpaired t test. (E) Cell viability measurements in AT II cells treated with H₂O₂ (0, 50, and 100 μM) for 18 h. Means ± SEM shown are from six replicate wells for each condition. *P < 0.05 and **P < 0.01 by unpaired t test.

We then assessed oxidative stress responses in AT II cells. Increased intracellular ROS levels were detected in AT II cells isolated from 9-mo-old Hhip^+/− mice compared with cells from age-matched Hhip^+/+ mice (Fig. 4D) (unpaired t test, P < 0.01). Consistently, increased cell death rates were detected in AT II cells from Hhip^+/− mice treated with H₂O₂ (Fig. 4E) (unpaired t test, P < 0.05).

HHIP Interacts with GSTP1 and Enhances Glutathione-Conjugating Activity.

HHIP prevents the HH ligands from activating transcription factor Gli1 (29). However, AT II cells having minimal Gli1 expression are less likely to be responsive to HH ligands. To identify the potential molecular mechanisms by which HHIP contributes to redox homeostasis in AT II cells, we applied two screening approaches: (i) we searched for cellular proteins that interact with HHIP using affinity purification followed by MS; and (ii) we used a PCR array-based assay to measure expression changes of genes related to oxidative stress pathways in AT II cells derived from Hhip^+/− mice and age-matched Hhip^+/+ mice.

First, in HEK 293 cells transfected with a C-terminal Flag/HA-tagged HHIP construct, HHIP-interacting proteins were visualized by silver staining after immunoprecipitation (IP) with anti-HA antibody (Fig. 5A). Through MS, 239 proteins were identified as cellular interacting proteins of HHIP after background subtraction of proteins identified from vector-transfected cells. Among novel HHIP interacting partners, we detected the presence of GST π 1 (GSTP1), a GST family member that catalyzes the conjugation of GSH with electrophilic compounds to fulfill its detoxification function. First, we confirmed the interaction between HHIP and GSTP1 by IP in HEK 293 cells transfected with Flag/HA-tagged HHIP. Compared with full-length HHIP, HHIP^1–193 that contains a frizzled domain on the N terminus of HHIP maintains its interaction with GSTP1 but HHIP^194–592 did not (Fig. 5B). Furthermore, overexpression of the full-length HHIP (but not the GSTP1 binding-deficient HHIP^194–592 mutant) improved cell viability (Fig. S6A), increased GSTP1 activity (Fig. S6B), and reduced intracellular ROS accumulation in Beas-2B cells treated with H₂O₂ (Fig. 5C), suggesting that the interaction between GSTP1 and HHIP promotes GSTP1 activity and thereby reduces lung oxidative stress levels. Second, we assessed where HHIP interacts with GSTP1 inside cells. Intracellular HHIP was detectable in the mitochondria of AT II cells (Fig. S7A), whereas GSTP1 interacts with HHIP mainly in the mitochondrial fraction of Beas-2B cells, which was enhanced after H₂O₂ treatment (50 μM for 1 h) (Fig. 5D). To determine whether HHIP modulates GSTP1 function, we measured GST enzymatic activity in AT II cells derived from Hhip^+/− and Hhip^+/+ mice at 9 mo of age. We detected significantly lower GST activity in AT II cells derived from Hhip^+/− vs. Hhip^+/+ mice (Fig. 5E), suggesting that HHIP promotes the activity of GST in murine AT II cells. We also detected increased mitochondrial-derived ROS, represented by increased mitoSOX staining in AT II cells from Hhip^+/− compared with Hhip^+/+ mice at 9 mo of age (Fig. S7B), suggesting a functional impacts of Hhip on mitochondria by limiting mitochondrial-derived ROS accumulation in AT II cells.

Fig. 5.

HHIP interacts with GSTP1 and enhances glutathione-conjugating activity. (A) Affinity purification of cellular protein complexes associated with Flag/HA-tagged HHIP (indicated by the black arrow) in HEK 293 cells as indicated by the red arrows. (B) IP of HHIP by anti-HA antibody in HEK 293 cells transfected with HHIP full length, HHIP^1–193, or HHIP^194–592 deletion mutant followed by immunoblotting (IB) with indicated antibodies. IgG-L, IgG light chain. (C) Intracellular ROS measurements in Beas-2B cells transfected with empty vector, full-length HHIP, or HHIP^194–592 deletion mutant after H₂O₂ treatment (200 μM for 12 h). (D) IP of GSTP1 in mitochondrial (Mit) and cytosol (Cyt) fractions by anti-GSTP1 or isotype IgG (control) antibody in Beas-2B cells treated with H₂O₂ (50 μM or 100 μM) for 1 h. Relative amount of HHIP to GSTP1 in the Mit-IP portions was shown in the bottom table. (E) GST activity measurements in AT II cells from 9-mo-old Hhip^+/− and Hhip^+/+ mice. **P < 0.01, unpaired t test.

Fig. S6.

Cell viability and GST activity measurements in Beas-2B cells. Cell viability (A) and GST activity (B) were measured in Beas-2B cells transfected with full-length HHIP (HHIP FL) or GSTP1 binding deficient mutant of HHIP (HHIP^194-592). Cells were treated with 200 μM H₂O₂ for 12 h.

Fig. S7.

HHIP is localized in mitochondria and limits mitochondria-derived ROS accumulation in murine AT II. (A) Detection of HHIP in cytosol (Cyt) and mitochondrial (Mit) fraction of AT II cells isolated from a 3-mo-old wild-type mouse by western blotting. GAPDH and electron transport chain complex IV (COX IV) were used as cytoplasmic marker and mitochondrial marker, respectively. (B) Flow cytometry analysis on mitoSOX staining in mitochondria isolated from AT II cells from Hhip^+/+ mice (+/+) and Hhip^+/− mice (+/−) at 9 mo of age.

Expression Profiling of Genes in the Oxidative Stress Pathway in AT II Cells.

To further identify additional mechanisms by which HHIP regulates redox homeostasis in AT II cells, we screened 84 known genes related to the oxidative stress pathway by RT-PCR (Mouse Oxidative Stress RT² Profiler PCR Array, Qiagen) in AT II cells derived from Hhip^+/− and Hhip^+/+ mice at 7 mo of age. Of 84 genes, 24 were differentially expressed in AT II cells from Hhip^+/− compared with cells from Hhip^+/+ mice (fold change > 1.5) (Fig. S8 A and B). Three of these 24 genes, uncoupling protein 2 and 3 (UCP2 and UCP3) and neutrophil cytosolic factor 1 (NCF1) were significantly increased in AT II cells from Hhip^+/− mice (Fig. S8C), suggesting oxidative status is increased in Hhip^+/− AT II cells. Among these 24 genes, expression of four antioxidant genes were reduced in AT II cells from Hhip^+/− mice (Table S2): dual oxidase 1 (Duox1), glutathione peroxidase 6 (Gpx6), prostaglandin-endoperoxide synthase 1 (Ptgs1), and recombination activating gene 2 (Rag2), suggesting that HHIP promotes other antioxidant functions beyond its capability to bind to and increase the activity of GSTP1.

Fig. S8.

Expression changes of genes related to the oxidative stress pathway in AT II isolated from Hhip^+/− and Hhip^+/+ mice at 7 mo of age. (A) The clustergram of the expression of 84 total genes in the oxidative stress pathway (Qiagen) in AT II cells (Hhip^+/− vs. Hhip^+/+). Red dots indicate genes with increased expression and green dots indicate genes with decreased expression in Hhip^+/− AT II cells. n = 3 mice per genotype. (B) Relative expression of 24 differentially expressed genes (fold change > 1.5) in AT II cells comparing Hhip^+/− vs. Hhip^+/+. Means ± SEM are from three mice per genotype. (C) Representative genes (NCF1, UCP2, and UCP3) are differentially expressed in AT II cells from Hhip^+/− (+/−) and Hhip^+/+ (+/+) mice. *P < 0.05, unpaired t test.

Table S2.

Functional category of differentially expressed genes (Hhip^+/− vs. Hhip^+/+) revealed in PCR-array in AT II cells from mice at 8 mo of age in biological oxidation pathway

Oxidation Pathway	Subpathway	Hhip+/− vs. Hhip+/+
		Up (14)	Down (10)
ROS metabolism (61)	Oxidative stress responsive genes (42)	6	3
	Other ROS metabolism genes (4)	0	2
	Superoxide dismutases (3)	1	0
	Other superoxide metabolism genes (12)	5	3
Oxygen transporters (9)	Oxygen transporters (9)	2	0
Antioxidants (36)	Glutathione peroxidases (GPx) (9)	0	1
	Peroxiredoxins (TPx) (7)	2	0
	Other peroxidases (12)	1	3
	Other antioxidants (8)	1	0

Numbers in bold indicate genes involved in the pathway.

Antioxidant Treatment in Hhip^+/− Mice.

We next asked whether increased oxidant levels induced age-related lung destruction and increased lung compliance in Hhip^+/− mice. We added NAC (N-acetyl-cysteine) (30–33), a thiol antioxidant that previously was shown to improve age-related emphysema in murine models (21), to the drinking water of Hhip^+/− and Hhip^+/+ mice at 5 mo of age. When mice were harvested at 10 mo of age, airspace sizes were similar in Hhip^+/− and Hhip^+/+ mice treated with NAC, in marked contrast to mice without NAC treatment (Fig. 6 A and B). NAC treatment also significantly reduced lung compliance in Hhip^+/− mice compared with age-matched untreated Hhip^+/− mice (Fig. 6C) (two-way ANOVA, P < 0.05). These functional and morphological improvements in Hhip^+/− murine lungs were associated with significant reductions in the number of lymphoid aggregates (Fig. 6D), as well as the activity of collagenase (Fig. S9A) and elastase (Fig. S9B) in Hhip^+/− mice treated with NAC compared with untreated Hhip^+/− mice. As expected, NAC treatment reduced oxidative stress levels in Hhip^+/− mice as assessed by reductions in oxidation of total protein in lung lysates from Hhip^+/− mice (Fig. 6 E and F). These findings supported the notion that the increased oxidant burden in Hhip^+/− murine lungs may drive the development of emphysema and impair pulmonary function during aging through promoting inflammation and MMPs activity, which was inhibited by antioxidant NAC treatment (Fig. S10).

Fig. 6.

Improved lung function and reduced airspace in Hhip^+/− mice by antioxidant treatment. H&E staining of lung sections (A) and MCL measurements (B) on murine lung sections from Hhip^+/+ and Hhip^+/− mice at 10 mo of age with or without NAC treatment for 5 mo. (Scale bars, 60 μm.) (C) Lung volume–pressure measurements. (D) Quantifications on lymphoid aggregates per airway in Hhip^+/+ and Hhip^+/− mice at 10 mo of age treated with NAC or H₂O. **P < 0.01, unpaired t test. (E) Protein oxidation measurements by Oxyblot. DNPH, 2,4-dinitrophenylhydrazine. (F) Densitometry quantifications on oxidized protein relative to α-Tubulin. Means ± SEM shown from three mice per group. *P < 0.05, unpaired t test.

Fig. S9.

Measurements of collagenase and elastase activity in lungs from Hhip^+/− (+/−) and Hhip^+/+ (+/+) mice at 10 mo of age with or without NAC treatment. Means ± SEM shown were from five to seven mice per group. **P < 0.01, unpaired t test.

Fig. S10.

A schematic model to summarize effects of Hhip haploinsufficiency on lungs that may contribute to age-related emphysema and lung function impairments, which were prevented by antioxidant NAC treatment.

Discussion

COPD is more prevalent in the elderly (34) and lung function levels normally decline with aging. Understanding the biological mechanisms underlying aging-induced FEV₁ decline and emphysema development is an urgent unmet need. Motivated by human GWAS studies showing that HHIP is associated with both COPD affection status in case-control studies and FEV₁ levels in general population samples (12, 13), we demonstrate that HHIP protects aging-related airspace enlargement and increases in lung compliance in mice. Reduction in HHIP expression in COPD lungs (16), as well as in lungs of aging rodents (23), increases susceptibility to develop emphysema during aging. Furthermore, reduced antioxidant capacity and increased oxidant levels in Hhip^+/− mice preceded emphysema development.

HHIP inhibits the HH pathway, a critical lung developmental pathway, through preventing three ligands from binding to its receptor, protein patched homolog 1, and eventually activating Gli1 (29). However, age-related emphysema in Hhip^+/− mice might depend on a noncanonical HH pathway given that the HH pathway was regarded as an antagonist of aging in Alzheimer’s disease and diabetes (35) because of its critical roles in organ-specific stem cells (36, 37). We would expect a slower aging process because of partial de-repression of the HH pathway in Hhip^+/− mice (22), whereas we observed accelerated aging-associated emphysema and increased cellular senescence in Hhip^+/− mice. By searching for additional interacting proteins of HHIP, we identified GSTP1, a detoxification gene, which protects murine lung epithelial cells from H₂O₂-induced cell death (38). Therefore, the interaction between HHIP and GSTP1 in the cellular mitochondrial fraction suggests that HHIP may reduce oxidative stress by binding to and possibly increasing the activity of GSTP1. Interestingly, overexpressing GSTP1 in Caenorhabditis elegans extended its lifespan (39), suggesting a potential antiaging role of GSTP1, consistent with reduced GSTP1 activity in Hhip^+/− mice.

Based on its interaction with GSTP1, HHIP may have an autonomous effect in AT II cells. However, Pdgfrα⁺ (platelet-derived growth-factor receptor α) mesenchymal cells but not endothelial cells and alveolar type I cells in the alveolar space also express HHIP (Fig. S11). Therefore, increased cell death and senescence detected in lungs from Hhip^+/− mice may result from a combination of autonomous and paracrine effects in multiple cell types. Although more work is needed to characterize autonomous and paracrine effects of HHIP in lungs during aging, it is noteworthy that the increased oxidant burden and increased lymphoid aggregates accompanied by lung function abnormalities in adult Hhip^+/− mice recapitulate those seen in human COPD patients carrying risk alleles in the HHIP GWAS locus (16, 22).

Fig. S11.

Expression of Hhip in different cell types in murine lungs. (A) PECAM1 (platelet endothelial cell adhesion molecule), (B) podoplanin/gp36, and (C) PDGFRα (platelet-derived growth factor receptor α) immunohistochemistry in lung sections with β-gal staining in Hhip^+/− mice. Hhip^+/+ lung sections stained with isotype IgG were used as negative controls for immunohistochemistry and LacZ staining. Hhip^+/− mice were generated with insertion of bacterial lacZ gene to replace the initiation codon and the rest of the downstream sequences in the first exon of the murine Hhip gene, so that expression of lacZ is driven by the endogenous Hhip promoter. X-gal staining was performed to indicate expression of Hhip in murine lungs. Hhip^+/− and Hhip^+/+ mice are at 2 mo of age. Cells with blue color indicate Hhip positive cells. Cells with colocalization signals are marked with arrows, while cells negative for Hhip expression are marked with arrowheads. (Scale bars, 50 μm.)

Senile lungs also exhibit increased inflammation that is likely pathologically important for the development of COPD, including increased levels of CXCR3 and CXCR5 to promote migration of B and T lymphocytes (24). Consistent with increased levels of cytokines, lymphoid aggregates, previously associated with severe emphysema in human COPD lungs (40), were also detected in lungs from Hhip^+/− mice. An increased number of lymphoid aggregates may contribute to the greater airspace enlargement observed in Hhip^+/− mice as (i) lymphoid aggregate number and size correlate with emphysema severity in human COPD patients (41); (ii) activated T lymphocytes (which are present in lymphoid aggregates) not only secrete MMPs (42) that can promote lung destruction, but also induce greater MMP production by macrophages (43); (iii) B cells (also present in lymphoid aggregates) promote CS-induced emphysema in mice (44); and (iv) B-cell products (autoantibodies) have been linked to emphysema in CS-exposed mice and human COPD patients (45, 46). Thus, Hhip may protect mice from spontaneous airspace enlargement in mice by inhibiting adaptive immunity (including lymphoid aggregates formation).

The cause of reduced levels of Hhip in lungs during aging, which may include reduced numbers of AT II cells during aging or reduced number of transcripts of HHIP in each AT II cell, needs further investigation. Nonetheless, a moderate reduction of Hhip in Hhip^+/− mice was sufficient to provoke oxidative stress, lymphoid aggregates, alveolar loss, lung function abnormalities, and cellular senescence related to emphysema at 10 mo of age, suggesting that normal levels of Hhip are critical to maintain lung homeostasis.

However, linking HHIP with aging definitively will require further comprehensive and sophisticated investigations. Therefore, we conclude that HHIP is important for maintaining lung homeostasis, but its plausible role in aging more generally will need additional evidence. Focusing on earlier age time points is instrumental to identify molecular pathways that serve as potential drivers for phenotypic changes in Hhip^+/− lungs. Thus, we only used mice with maximum age of 18 mo in this study, unlike most other traditional aging studies.

Oxidative stress has been a therapeutic target for multiple diseases. Despite the well-accepted contribution of oxidative stress to pulmonary emphysema, the efficacy of antioxidant therapy in pulmonary disease has been controversial (27, 47), possibly because of genetic complexity in COPD patients. Given that Hhip^+/− mice, which mimic human HHIP risk allele carriers with respect to HHIP expression levels, clearly derived more benefit from NAC treatment for pulmonary morphology and lung function improvement than Hhip^+/+ mice, additional studies are needed to determine whether COPD patients carrying HHIP risk alleles benefit more from antioxidant therapies.

Our results herein, coupled with our recent finding that Hhip protected mice from CS-induced emphysema (22), support the hypothesis that Hhip is essential for maintaining lung homeostasis during the aging process in response to environmental insults by protecting cells from oxidative stress-induced injury. Our studies expand knowledge about the molecular mechanisms by which HHIP contributes to age-related decline in lung function in humans. These new insights into aging-related emphysema may eventually facilitate the development and testing of more targeted therapies to reduce morbidity in aging subjects carrying risk variants at the HHIP locus by limiting decline in FEV₁ during aging.

Materials and Methods

Full experimental procedures and any associated references are available in SI Materials and Methods.

Animals.

Hhip^+/− mice in C57BL/6J background were described previously (22). This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals (48) of the National Institutes of Health.

Airspace Size Analysis.

For alveolar morphometric analysis, at least 15 images per mouse lung were randomly taken for analysis using methods described previously (22). Images were processed using Scion imaging software and analyzed for mean alveolar chord length (49).

Statistical Analysis Methods.

We used two-way ANOVA analyses followed by Student two-sample unpaired t tests. See details in SI Materials and Methods.

SI Materials and Methods

Animals.

Hhip^+/− mice in C57BL/6J background were described previously (22), whereas the exon 1 of Hhip was replaced by a lacZ reporter gene. All mice were housed in the animal facility of Harvard Medical School with a 12-h-light/12-h-dark cycle. Hhip^+/+ and Hhip^+/− mice were harvested at different ages for lung mechanics and morphometric analysis as described below. This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals (48) of the National Institutes of Health.

Airspace Size Analysis.

For alveolar morphometric analysis, mouse lungs were fixed with formaldehyde for 48 h after inflation with PBS with a constant pressure of 25 cm H₂O. The mean alveolar chord length was measured after slides were stained with Gill’s solution. At least 15 images per mouse lung were randomly taken for analysis using methods described previously (22). Images were processed using Macro imaging software and analyzed for mean alveolar chord length.

Antioxidant Treatment in Mice.

Female Hhip^+/− and littermate mice (∼20-wk-old) were supplemented with NAC (1 g/kg body weight; Sigma) in drinking water for 5 mo. In the control group, mice were supplied with regular drinking water for the same duration. At the end of the treatment, mice were killed by CO₂ narcosis and cervical dislocation before removing lungs for other studies. Mice were assessed for lung mechanics and morphometric analysis.

Cell Lines and Plasmid Construction.

Human Beas-2B bronchial epithelial cells (#CRL-9609), HEK293 cells (#CRL-1573), and mouse MLE12 cells (#CRL-2110) were purchased from ATCC and cultured in complete DMEM supplemented with 10% (vol/vol) FBS, penicillin (50 units/mL), streptomycin (50 μg/mL), and gentamicin (10 µg/mL).

HHIP and HHIP deletion mutants (HHIP^1–193 and HHIP^194–592) were cloned into pC-FLAG-HA CMV vector with BamHI and EcoRV restriction sites to generate HA-FLAG-HHIP constructs, and then confirmed by sequencing.

Measurement of Lung Mechanics.

To measure respiratory mechanics, mice were anesthetized with a mixture of 200 mg/kg body weight ketamine, 10 mg/kg xylazine, and 3 mg/kg acepromazine. A tracheostomy was performed, and an 18-gauge cannula was inserted and secured in the trachea using sutures. Mice were then connected via the cannula to a digitally controlled mechanical ventilator (Flexi Vent device; Scireq). Ventilator settings were f = 150/min, FiO₂ = 0.21, tidal volume =10 mL/kg body weight, and positive end-expiratory pressure (PEEP) = 3 cm H₂O. Murine lungs were inflated to total lung capacity 25 cm H₂O three times before volume history. Tissue resistance (G) and tissue elastance (H) at PEEP of 3 cm H₂O were then measured, followed by stepwise quasi-static compliance (Cst) and volume–pressure curve measurements.

Oxidative Stress Measurement in Murine Lungs (GSH/GSSG).

Total glutathione content in lung protein lysate (GSH/GSSG) was measured using the OxiSelect Total Glutathione Assay Kit (cat# STA-312) (Cell Biolabs). For measuring GSH, 10 μL of lung lysate sample was precipitated with 5% (wt/vol) metaphosphoric acid and levels of GSH were determined by Ellman’s reagent (5,5′-dithiobis-2-nitrobenzoic acid or DTNB), which reacts with GSH to form a spectrophotometrically detectable product at 412-nm wavelength. For measuring GSSG, 20 μL of lung protein lysate sample was pretreated with 2 μL of 1-methyl-2-vinylpyridinium trifluoromethanesulfonate1 (M2VP), a thiol-scavenging reagent, to rapidly scavenge GSH. Subsequently, oxidized GSH was determined using Ellman’s reagent as described above. The total GSH and GSSG content was determined by comparing with a glutathione standard curve. The original amount of GSH was determined by subtracting GSSG from total GSH. Raw readings were normalized to the protein amount in each sample.

Lung Antioxidant Capacity Measurement.

To measure total antioxidant capacity in murine lungs, we used OxiSelect Total Antioxidant Capacity Assay kit (#STA-360, Cell Biolabs) according to the manufacturer’s instructions. This assay is based on the reduction of copper (II) to copper (I) by antioxidants. Under reduction, the copper (I) ion reacts with a coupling chromogenic reagent that produces a color with a maximum absorbance at 490 nm. The absorbance value of each sample was converted to antioxidant capacity by comparing with the standard curve with known concentration of uric acid standards.

AT II Cell Isolation from Murine Lungs.

Murine lung AT II cells were isolated and used for experiments within 5 d of culture without passage. In brief, mice were anesthetized, and the lungs were perfused with 20 mL of Ca²⁺/Mg²⁺-free PBS. An intravenous catheter (20-G) was inserted into the trachea. Dispase solution (2 mL; Gibco, Invitrogen) was infused through the tracheal catheter and 1 mL of 1% low-melting agarose (Sigma-Aldrich) was slowly infused through the catheter. Immediately, the lungs were covered with ice for 2 min. The lungs were removed and transferred to 2 mL of dispase and incubated for an additional 45 min at room temperature. The lungs were then transferred to 7 mL of DMEM with 0.01% DNase I (Sigma-Aldrich) in a 60-mm culture plate. The digested lung tissue was teased and filtered through 70-μm cell strainer and the pellet was collected after centrifugation (130 × g for 8 min). The pellet was resuspended in DMEM supplemented with 25 mM Hepes 8 buffer, pH 7.4, 10% FBS, and 1% penicillin and streptomycin. The cell suspension was incubated in plates precoated with anti-CD16/32 (BD Biosciences) and anti-CD45 (BD Biosciences) for 1 h at 37 °C. Unbound cells were collected and inoculated into noncoated culture plates for 4 h at 37 °C. The cell suspension was harvested and centrifuged (130 × g for 8 min). The pellet was resuspended in DMEM containing 10% FBS and antibiotics and inoculated to fibronectin-coated plates (BD Biosciences). AT II cells were subjected to experiments after confirmation of the purity >95% by immunofluorescence staining for SPC.

Detection of β-Gal Activity: X-Gal Staining.

Hhip^+/− mice were generated with insertion of bacterial lacZ gene to replace the initiation codon and the rest of the downstream sequences in the first exon of the murine Hhip gene, so that expression of lacZ is driven by the endogenous Hhip promoter (23). X-gal staining was performed to indicate expression of Hhip in murine lungs. Fresh mouse lung from Hhip^+/− mice (2-mo-old) were perfused with PBS to remove blood through the right ventricle of the heart and then fixed with fixation buffer (0.25% glutaraldehyde with 2 mM MgCl₂, pH 7.4) for 15 min at room temperature. After inflation with PBS two to three times, lungs were inflated with X-gal solution [5 mM K-Ferricyanide, 5 mM K-Ferrocyanide, 2 mM MgCl₂, X-gal (0.1%) in PBS] at room temperature for 2–8 h followed by fixation overnight in fresh 4% paraformaldehyde before sectioning. Age-matched Hhip^+/+ mice were used as controls.

Cell Viability Analysis.

AlamarBlue (Thermo Fisher Scientific) was used to assess cell viability. AT II cells or Beas-2B cells were seeded into a 96-well plate at a density of 5,000 cells per well in triplicate or quadruplicate. Next morning, H₂O₂ at different concentrations was added into wells. After H₂O₂ treatment, cell viability was determined by AlamarBlue assays in a fluorescence plate reader (Biotek Instruments). Percentages of live cells upon H₂O₂ treatment were calculated based on fluorescence signals from each well normalized to cells without H₂O₂ treatment that was defined as 100%.

Intracellular ROS Measurement.

Beas-2B or mouse AT II cells at 60% confluency were treated with H₂O₂ for 16 h. After treatment, 100 μL PBS with DCFHDA (Cayman Chemical; 1:2,000 dilution, final concentration 20 μM) and 10% (vol/vol) AlamarBlue (Thermo Fisher Scientific) were added to cells in each well at 37 °C for 120 min. ROS levels were measured at Ex/Em: 485 nm/530 nm in a fluorescence plate reader (Biotek, FLX800), and cell viability was measured at Ex: 560 nm/Em: 590 nm. Intracellular ROS levels were normalized to relative number of viable cells as measured by AlamarBlue assay.

Affinity-Purification Followed by MS.

HEK293 cells were transfected with a C-terminal Flag/HA-tagged human HHIP construct. Forty-eight hours posttransfection, total cell lysate in Nonidet P-40 cell lysis buffer (1% Nonidet P-40, 150 mM NaCl, 20 mM Tris⋅HCl, 1 mM EDTA, 1 mM EGTA, and proteinase inhibitor mixture) was subjected to immunoprecipitation with anti-HA agarose beads (Sigma) at 4 °C for 4 h, antibody-bound HHIP protein complexes were eluted with 100 mM Glycine (pH 2.5), and 30% of the eluate was resolved by SDS/PAGE followed by silver staining (Silver Quest, Life Technologies) to visualize HHIP-interacting proteins. Seventy-percent of the eluates were then precipitated with 20% trichloroacetic acid and sent for MS analysis (The Taplin Biological Mass Spectrometry Facility, Harvard Medical School, Boston, MA).

Immunoprecipitation Assay.

HEK 293T cells (3 × 10⁶ cells per 10-cm plate) were transfected with 7 μg of HA/FLAG-tagged human HHIP constructs or empty vector. After 48 h, cells were lysed with immunoprecipitation buffer [50 mM Tris–HCl, 300 mM NaCl, 1% Triton-X-100, 5 mM EDTA, 50 mM NaF, 1 mM Na₃VO₄, and Protease Inhibitor Mixture (Roche Applied Science)]. HA-FLAG-HHIP proteins were immunoprecipitated by using anti-HA agarose gel (Thermo Fisher Scientific), followed by immunoblotting with various antibodies. GSTP1 protein was immunoprecipitated by using rabbit anti-GSTP1 antibody (#311, MBL international) and protein A Dynabeads (Thermo Fisher Scientific).

Mitochondrial Isolation.

Two days after transfection with Flag-tagged HHIP, Beas-2B cells were collected by centrifugation at 370 × g for 10 min and then resuspended in NKM buffer (pH 7.4, 1 mM Tris HCl, 0.13 M NaCl, 5 mM KCl, 7.5 mM MgCl₂). Cells were then pelleted again and supernatant was decanted and resuspended in homogenization buffer (10 mM Tris⋅HCl, 10 mM KCl, 0.15 mM MgCl_2, 1 mM PMSF, 1 mM DTT). Subsequently, cells were homogenized 100 times using a tight pestle. Cell suspensions were then placed in sucrose solution (2M) followed by centrifugation at 1,200 × g for 5 min to spin down the unbroken cells, nuclei, and large debris. Supernatants with mitochondria were centrifuged at 7,000 × g for 19 min to spin down mitochondria, then washed by mitochondria suspension buffer (10 mM Tris HCl, pH 6.7, 0.15 mM MgCl₂, 0.25 mM sucrose, 1 mM PMSF, and 1 mM DTT) and centrifuged at 9,500 × g for 5 min to repellet the mitochondria for Western blotting or IP. Supernatant was diluted in Nonidet P-40 cell lysis buffer (1:3 ratio) and was used for cytosolic-IP; mitochondrial pellets were used for mitochondrial IP in Nonidet P-40 cell lysis buffer.

Immunological Methods.

Primary antibodies used in this study include anti-p53 (#9282, Cell Signaling Technology), anti-SP-C(sc-7706), anti-p21 (sc-397), and anti-Ki67 (sc-15402) antibodies (all from Santa Cruz Biotechnology), anti-COX IV (ab16056) (Abcam), anti-HHIP (#WH0064399M1, Sigma-Aldrich), β-actin (#MAB1501, EMD Millipore), GSTP1 (#ab47709, Abcam) and FLAG (#F1804, Sigma-Aldrich). Secondary antibodies were horseradish peroxidase-linked anti-mouse or anti-rabbit IgG (GE Healthcare). Signals were detected with enhanced chemi-luminescence kit (Perkin-Elmer Life Sciences) followed by a Bio-Rad Imaging autodeveloping system. Band densities were quantified by ImageJ (NIH) software.

Murine lung AT II cells were quantified by staining paraffin-embedded lung sections with SPC (Santa Cruz, #sc-7706) antibody. Sections from five to six mice per group were stained with anti-SPC antibody and DAPI. Numbers of SPC and DAPI double-positive cells were quantified using ImageJ software.

For triple immunofluorescence staining in lung cryosections in wild-type mice, slides were fixed in 4% (wt/vol) paraformaldehyde for 8 min. After washing with PBS twice for 5 min, slides were incubated with 0.05% Triton-X 100 for 6 min. Slides were further blocked with 10% (vol/vol) normal donkey serum for 1 h and then incubated overnight at 4 °C with mouse acetylated-tubulin antibody (Sigma; diluted 1:100), rabbit HHIP antibody (Novus Biologicals; diluted 1:50), and goat anti-SPC (Santa Cruz Biotechnology; diluted 1:50) antibody followed by incubation with second antibodies (Life Technologies), including Alexa 647-conjugated donkey anti-goat IgG (diluted 1:500), Alexa 488-conjugated donkey anti-mouse IgG (diluted 1:500), and Alexa 546-conjugated donkey anti-rabbit IgG (diluted 1:500). After washing and counterstaining with DAPI, images of the stained lung sections were analyzed using a confocal microscope (Leica Microsystems).

For immunohistochemistry, goat polyclonal PECAM1 (platelet endothelial cell adhesion molecule) Ab (1:50; #sc-1506, Santa Cruz), rabbit polyclonal PDGFRα Ab (1:250; #sc-338, Santa Cruz), rabbit polyclonal αSMA Ab (1:1,000; #5694, Abcam), and hamster monoclonal podoplanin Ab (1:10; #11936, Abcam) were used. Biotinylated secondary antibodies were used at 1:200 then incubated with horseradish peroxidase streptavidin complex (#SA-5004, Vector laboratories). DAB substrate (#SK-4015, Vector laboratories) was incubated for 2–10 min for signal development. Control sections were incubated with the same isotype-specific IgG antibodies at the same concentration for each of the above primary antibodies, followed by an identical staining procedure.

GST Activity Assay.

Total GST enzymatic activity was evaluated in lungs and primary AT II cells from Hhip^+/+ and Hhip^+/− mice, using a GST activity kit (Abcam) that uses monochlorobimane (MCB), a dye that reacts with glutathione. MCB changes from nonfluorescent to fluorescent (Ex/Em = 380/461 nm) when the free form of MCB reacts with glutathione. Briefly, 48 h after seeding, cells were collected for protein lysate. The reaction was initiated by incubation reaction mixture (100 µL) that contained 20 µL of protein lysates and 2 mM reduced GSH in MCB (final concentration 1 mM) at room temperature. The fluorescence signals (Ex/Em = 380/461 nm) were recorded every 5 min for 1 h. The increase rate in fluorescence intensity is directly proportional to the GST enzymatic activity in each sample. Each assay was performed in duplicate, and enzyme activity was recorded as nanomole per minute per milligram of protein after normalization to total protein amount.

PCR-Array in Oxidative Stress Pathway.

Expression levels of genes related to the oxidative stress pathway were assessed in the murine alveolar type II cells by using RT² Profiler PCR Array: Mouse Oxidative Stress (Qiagen). Briefly, murine AT II cells were cultured for 3–4 d and total RNA was isolated for cDNA synthesis using RT² First Strand Kit (Qiagen). RT-PCR was performed in Quantstudio (Life Technology). Data were analyzed with RT² Profiler PCR Array Data Analysis software 3.5 (pcrdataanalysis.sabiosciences.com/pcr/arrayanalysis.php) with four house-keeping genes as reference genes. Fold change of gene-expression changes comparing Hhip^+/− vs. Hhip^+/+ mice bigger than 1.5 was defined as significantly differentially expressed genes.

Detection of Differential Gene Expression by Real-Time RT-PCR.

Gene expression was measured by real-time PCR with gene-specific primers using TaqMan probes. Relative gene expression was calculated based on the standard 2^−ΔΔCT method, using GAPDH and control cells as reference. Gene-expression comparison was carried out using unpaired t test analysis.

Protein Oxidation Detection.

Detection of protein oxidation in murine lungs was performed with OxyBlot Protein Oxidation Detection Kit (EMD Millipore) based on the manufacturer’s instructions. Briefly, 20-μL murine lung protein lysates were incubated with 10 μL of DNPH at room temperature in 6% (wt/vol) SDS denaturing solution for 15 min. After neutralization, protein samples were loaded to a poly-acrylamide gel and further transferred to PVDF membrane. DNP-labeled proteins were detected using anti-DNP antibody in Western blotting.

Quantification of Lymphoid Aggregates in Murine Lungs.

The number of lymphoid aggregates was determined by counting the total number of lymphoid aggregates around airways with 200- to 1,000-μm diameter in each murine lung section that was stained with H&E. The number of lymphoid aggregates per airway was compared in lung sections from Hhip^+/+ and Hhip^+/− mice at ages of 2, 8, and 12 mo by unpaired t test analysis.

Measurement of MMP-9 by ELISA.

MMP-9 levels in murine lungs were measured using mouse MMP-9 Quantikine ELISA Kit (R&D) based on the manufacturer’s protocol. Briefly, after 2 h of incubation of lung protein lysates with assay buffer in adhesive strips at room temperature, each well was washed and incubated with mouse MMP-9 conjugates for 2 h followed by signal development with 100 μL of substrate solution. Optical density was measured in a microplate reader at wavelength of 450 nm. MMP-9 levels in samples were calculated based on amount of MMP-9 standards and then normalized to total lung protein amount.

Measurements on MMPs, Cytokines, Cell Death, and Cell Proliferation in Murine Lungs.

Multiplexed immunoassays with fluorescent microspheres were used to measure cytokines and MMPs using the Milliplex-MAP Mouse Cytokine/Chemokine magnetic kit (#MCYTOMAG-70k, EMD Millipore) and Milliplex-MAP Mouse MMP magnetic kit (#MMMP3MAG-79k, EMD Millipore), respectively, according to the the manufacturer’s instructions by Luminex multiplexing instruments (EMD Millipore).

The level of apoptosis was assessed in paraffin-embedded sections by fluorescent TUNEL assay kit according to the manufacturer’s instructions (DeadEnd Fluorometric TUNEL system, Promega).

The level of proliferation was assessed in paraffin-embedded sections by anti-Ki67 antibody and DAPI in lungs from 2- and 10-mo-old mice. Ki67⁺ and DAPI⁺ cells were quantified using ImageJ software. Proliferating cells were indicated by number of Ki67 and DAPI double-positive cells.

Collagenase Activity Assay.

The enzymatic assay of collagenase (specifically EC 3.4.24.3) was assessed using FALGPA [N-(3-[2-furyl]acryloyl)-Leu-Gly-Pro-Ala; Sigma] as the substrate by measuring the spectrophotometric determined at a wavelength of 345 released by the product FAL [N-(3[2-furyl]acryloyl)-Leu]. In the measurement, 78.3 mg of lung protein lysate was mixed with 290 mL FALGPA solution at a final concentration of tricine (48.3 mM), calcium chloride (9.67 mM), sodium chloride (387 mM), and FALGPA (0.967 mM). The absorbance readings at a wavelength of A345 can be stabilized after 5 min of incubation. The activity of collagenase was determined by the standard curve achieved by gradient concentrated collagenase enzyme.

Elastase Activity Assay.

The enzymatic activity of elastase was assessed using EnzChek Elastase assay kit (Life Technologies) based on the protocol provided by the manufacturer. A murine protein sample of 40 mg was used in the measurement.

Statistical Methods.

We tested data for normality before performing unpaired t-tests. Any data that were not normally distributed will be analyzed using a nonparametric test. However, most in vitro data have limited sample size and showed normal distribution. In all age-related phenotypic examinations in Hhip^+/− and Hhip^+/+ mice (MCL, lung mechanics, lymphoid aggregates, intracellular ROS in AT II cells treated with H₂O₂, expression of Hhip and MMPs measurements), we always started our analysis by using two-way ANOVA if data are normality distributed to determine whether there are significant genotype-by-age interaction effects. If such interaction effects were detected, we then performed unpaired t tests to compare the impact of genotype status at a given age. We also performed one-way ANOVAs to compare age effects within each genotype.