Hypoxia-induced mitogenic factor (FIZZ1/RELMα) induces endothelial cell apoptosis and subsequent interleukin-4-dependent pulmonary hypertension

Abstract

Pulmonary hypertension (PH) is characterized by elevated pulmonary artery pressure that leads to progressive right heart failure and ultimately death. Injury to endothelium and consequent wound repair cascades have been suggested to trigger pulmonary vascular remodeling, such as that observed during PH. The relationship between injury to endothelium and disease pathogenesis in this disorder remains poorly understood. We and others have shown that, in mice, hypoxia-induced mitogenic factor (HIMF, also known as FIZZ1 or RELMα) plays a critical role in the pathogenesis of lung inflammation and the development of PH. In this study, we dissected the mechanism by which HIMF and its human homolog resistin (hRETN) induce pulmonary endothelial cell (EC) apoptosis and subsequent lung inflammation-mediated PH, which exhibits many of the hallmarks of the human disease. Systemic administration of HIMF caused increases in EC apoptosis and interleukin (IL)-4-dependent vascular inflammatory marker expression in mouse lung during the early inflammation phase. In vitro, HIMF, hRETN, and IL-4 activated pulmonary microvascular ECs (PMVECs) by increasing angiopoietin-2 expression and induced PMVEC apoptosis. In addition, the conditioned medium from hRETN-treated ECs had elevated levels of endothelin-1 and caused significant increases in pulmonary vascular smooth muscle cell proliferation. Last, HIMF treatment caused development of PH that was characterized by pulmonary vascular remodeling and right heart failure in wild-type mice but not in IL-4 knockout mice. These data suggest that HIMF contributes to activation of vascular inflammation at least in part by inducing EC apoptosis in the lung. These events lead to subsequent PH.

pulmonary hypertension (PH) is characterized by elevated pulmonary artery pressure that leads to progressive right-sided heart failure. It is associated with significant morbidity and mortality. Despite major advances in diagnosis and treatment of this disease over the last several decades, the underlying mechanisms of PH remain enigmatic. In humans, severe pulmonary arterial hypertension (PAH, idiopathic PH) is characterized by plexiform lesions that contain phenotypically altered pulmonary smooth muscle and endothelial cells (ECs) (56). Evidence suggests that early EC apoptosis may trigger both degenerative and reactive proliferative events that result in the pulmonary vascular pathology of PH (25). Failure to stop the initial repair response could result in pathophysiological vascular remodeling, such as that which occurs during PH (25). Strong circumstantial clinical evidence supports an immune pathogenesis of PH, but whether the immunological features are a cause or consequence of PH is unknown. It has been shown that vascular endothelial growth factor (VEGF) receptor (VEGFR) blockade with Sugen 5416, a potent inhibitor of VEGFR tyrosine kinase, combined with chronic hypoxia, results in severe angioproliferative PH in rats (52). The mechanism of this response is not clear, but evidence suggests that EC apoptosis increases in response to the loss of survival signaling at the initiation, creating conditions that favor emergence of apoptosis-resistant cells with increased growth potential.

A study by Daley and colleagues (8) demonstrated pulmonary vascular muscularization in mice in response to Aspergillus or ovalbumin (OVA) antigen immunization, suggesting that lung vascular remodeling is regulated by the immune system, specifically a T-helper (Th) type 2 immune response. Our group has shown that hypoxia-induced mitogenic factor (HIMF, also known as FIZZ1 or RELMα), a member of the resistin family of proteins, is dramatically and transiently upregulated in a hypoxia-induced PH model. HIMF has proinflammatory and chemokine actions (53, 60) and is known to be persistently upregulated in the lungs in animal models of allergic inflammation (23), bleomycin-induced (35) or herpes virus-induced (39) pulmonary fibrosis, and schistosomiasis-induced PH (19). Genetic transfer of HIMF in rat lung induces the vascular remodeling and hemodynamic changes of PH, and inhibition of this pathway prevents development of hypoxia-induced PH (2).

HIMF (FIZZ1) is well known as a marker for alternatively activated (M2) macrophages (40). A recent study showed that this protein is produced by M2 macrophages that were recruited to the lung at an early phase of hypoxia, and that inhibition of this phenotype ameliorated hypoxia-induced PH (57). These data suggest that M2 macrophage-derived HIMF (FIZZ1) protein plays an important role in PH development. Because its expression is driven by Th2 cytokines interleukin (IL)-4 and IL-13 and it has a regulatory effect on Th2 immune response, HIMF is considered a novel effector molecule in type 2 inflammation (41). Additionally, we have shown that resistin-like molecule β (RELMβ), a human correlate of HIMF, is upregulated in lungs of patients with scleroderma-associated PH, strongly suggesting an etiologic role of resistin family proteins in PH (5). On the other hand, human resistin (hRETN) is expressed by myeloid cells, and its expression pattern shows a greater similarity to that of murine HIMF (RELMα) than to that of murine resistin. Thus, the putative functions of murine HIMF, specifically its effect on vascular cells, are thought to be shared with resistin in humans (41).

We have published evidence that a single systemic injection of recombinant (r) HIMF protein causes lung inflammation at 7 days that is dependent on the VEGF and Th2 cytokine IL-4 pathways (60, 62). During this early inflammation phase, HIMF significantly upregulated VEGF and downregulated VEGFR2 expression in the lung (60). We also have shown that HIMF induces growth factors and chemokines [such as VEGF, stromal cell-derived factor-1 (SDF-1/CXCL12), and monocyte chemotactic protein 1] in the lung resident cells that contribute to vascular inflammation (60, 62). HIMF also caused significant macrophage recruitment to the lung. This recruitment was completely abolished in IL-4 knockout (KO) mice, suggesting that HIMF's inflammatory response is dependent on IL-4 and that it amplifies Th2 inflammation in the lung (62).

In the present study, we dissected the mechanism by which HIMF and its human homolog hRETN cause inflammation in the pulmonary vasculature. We hypothesized that HIMF can cause pulmonary vascular remodeling and subsequent PH in response to initial pulmonary EC apoptosis and vascular inflammation in a mechanism dependent on the Th2 cytokine IL-4. We investigated 1) whether HIMF promotes EC apoptosis during early lung inflammation and what inflammatory markers are involved; 2) whether HIMF and hRETN promote apoptosis in cultured pulmonary microvascular ECs (PMVECs) and which signaling pathway is involved; and 3) whether HIMF-induced pulmonary vascular remodeling and PH development is IL-4 dependent.

MATERIALS AND METHODS

Animals and in vivo HIMF instillation.

Male wild-type (WT) and IL-4 KO mice (all on a C57BL/6 background, 8–12 wk old) were purchased from The Jackson Laboratory (Bar Harbor, ME). Housing and procedures involving experimental animals were approved by the Animal Care and Use Committee of the Johns Hopkins University. rHIMF protein was produced in T-REx 293 cells as previously described (53). We injected rHIMF (200 ng/animal in 100 μl saline) intravenously as reported previously (60, 62). FLAG protein (200 ng/animal in 100 μl saline; Sigma-Aldrich, St. Louis, MO) was used as a negative control (rHIMF is FLAG-tagged at its COOH-terminus). Endotoxin level of rHIMF protein was <0.1 ng/μg. Lung tissue was collected at 7 days (early inflammation phase) and 30 days (PH development phase) after the injection.

Histology and laboratory testing.

Once appropriate hemodynamic measurements were made, mice were killed by exsanguination, and the heart and lungs were removed en bloc. For histology, the lung was inflated under constant pressure with 1% low-melt agarose in phosphate-buffered saline (PBS) and placed on ice. The inflated lung was then placed in 4% paraformaldehyde and subsequently processed for histology as described previously (60–62). The heart was then bisected into the right ventricle (RV) and left ventricle (LV) plus septum (S). Each portion of the heart was weighed, and the RV and LV + S ratio was determined (2). The lung and right ventricles were stored at −80°C for use in immunoblot analysis or in RNAlater (Qiagen, Hilden, Germany) for quantitative RT-PCR (QRT-PCR).

Pulmonary vascular remodeling analysis.

Pulmonary vascular remodeling of the mice was assessed as we have previously published (2). For initial analysis, sections were stained with hematoxylin and eosin. Additional lung sections were dual labeled with antibodies to von Willebrand factor (vWF; Dako, Glostrup, Denmark) and smooth muscle actin (SMA; Dako) to stain endothelium and vascular smooth muscle, respectively; next, they were counterstained with hematoxylin as we have described (2). To assess remodeling of the lung arteries and arterioles, an investigator blinded to treatment group examined 100 arteries at random per lung section under ×40 objectives using an Olympus-BHS microscope (Olympus, Tokyo, Japan) attached to a QImaging Retiga 4000RV digital camera (QImaging, British Columbia, Canada). Small arteries with an internal diameter ≤100 μm were then classified as nonmuscular, partially muscular, or fully muscular, according to α-SMA staining. Average data reflect results from five lungs per group. Negative control sections for the immunohistochemical experiments received identical treatments but were not exposed to the primary antibody; they showed no specific staining.

In vivo hemodynamics.

In vivo right ventricular function was assessed by a pressure-volume catheter as described previously (2, 3, 51). Briefly, mice were anesthetized with 1–2% isoflurane, 750–100 mg/kg ip urethane, 5–10 mg/kg ip etomidate, and 1–2 mg/kg ip morphine. Next, they were subjected to tracheostomy and ventilated with 6–7 μl/g tidal volume at 130 breaths/min. An incision was made between the seventh and eighth ribs to expose the right ventricular apex, through which we inserted a 1.4-Fr pressure-volume catheter (SPR 839; Millar Instruments, Houston, TX) as described previously (2, 51).

QRT-PCR analysis.

Total RNA was isolated from lung tissue of mice with the Trizol reagent method (Invitrogen, Grand Island, NY). The quality of total RNA samples was assessed with an Agilent 2100 Bioanalyzer (Agilent Technologies). Reverse transcription was performed by processing total RNA with the Applied Biosystems High-Capacity cDNA Archive kit first-strand synthesis system for RT-PCR according to the manufacturer's protocol. QRT-PCR was carried out with the TaqMan assay system from Applied Biosystems. All PCR amplifications were carried out in duplicate on an ABI Prism 7300 Sequence Detection System by a fluorogenic 5′-nuclease assay (TaqMan probes). Probes and primers were designed and synthesized by Applied Biosystems: mu_gapdh (Mm99999915_g1); mu_actb (Mm00607939_s1); mu_pgk1 (Mm00435617_m1); mu_Ccl5 (Mm01302428_m1); mu_Angpt2 (Mm00545822_m1); mu_Pdgfa (Mm01205760_m1); mu_Mmp2 (Mm00439498_m1); mu_Vcam1 (Mm01320970_m1): mu_Timp2 (Mm00441825_m1); and mu_Bmp4 (Mm01332882_m1). Relative gene expression was calculated by using the 2^−ΔΔC_t method, in which C_t indicates cycle threshold, the fractional cycle number at which the fluorescent signal becomes detectable (36). We calculated the ΔC_t value of each sample using three endogenous control genes (gapdh, actb, and pgk1). Average fold change = 2^{−averageΔΔC_t} for gene expression in treated relative to control samples.

Isolation of PMVECs.

Mouse PMVECs were isolated as described previously (62). Mouse lung parenchyma was removed from a peripheral region devoid of large airways, rinsed with DMEM, minced, and digested in 1 ml of collagenase (1 mg/ml; Sigma-Aldrich) at 37°C for 20 min with occasional agitation. The cellular digest was filtered through sterile mesh and centrifuged (400 g for 7 min). The cell pellet was resuspended in 1 ml of growth medium and plated on gelatin-coated T-25 flasks. After 5–7 days, areas of cells exhibiting cobblestone morphology were selected, treated with trypsin, and replated on gelatin-coated dishes. Next, cells were labeled with fluorescent-labeled, acetylated, low-density lipoprotein (AcLDL; Alexa Fluor 488 AcLDL; Invitrogen), and AcLDL-positive cells were sorted and collected by flow cytometry (FACSAria; BD Biosciences, San Jose, CA). Cells were grown in a humidified atmosphere of 5% CO₂ at 37°C in DMEM enriched with 20% heat-inactivated fetal bovine serum (FBS), 15 μg/ml endothelial cell growth supplement (Millipore, Billerica, MA), 100 U/ml penicillin-100 μg/ml streptomycin, 0.25 μg/ml amphotericin B, and 0.1 mM MEM with nonessential amino acids. PMVECs were used in experiments at passages 3–6 when they were nearly (∼85%) confluent. The rHIMF that was used in this experiment was prepared as described above, and rmIL-4 was purchased from Peprotech (Rocky Hill, NJ).

Human cell culture and EC activation.

Human PMVECs (HMVEC-L; Lonza, Walkersville, MD) were cultured in endothelial cell basal medium-2 (VEM-2; Lonza) supplemented with 5% FBS, human epidermal growth factor, human VEGF, human fibroblast growth factor (with heparin), long R3 insulin-like growth factor-I, hydrocortisone, ascorbic acid, gentamicin, and amphotericin B (Bulletkit CC-3202; Lonza). Human pulmonary artery smooth muscle cells (HPSMCs; Lonza) were cultured in smooth muscle basal medium (SmBM; Lonza) supplemented with 5% FBS, human epidermal growth factor, human fibroblast growth factor-B, insulin, gentamicin, and amphotericin B (Bulletkit CC-4149; Lonza). Only PMVECs and HPSMCs from passages 4–6 were used. For EC conditioned medium preparation, human PMVECs were serum- and growth factor-starved (VEM-2 media with 1% FBS) for 24 h and then treated with rhRETN (100 ng/ml; Peprotech), rhIL-4 (50 ng/ml; Peprotech), or vehicle as described previously (4). Preliminary studies were performed to optimize both incubation time and the recombinant protein concentrations used (data not shown). After 24 h, the medium was collected and used for HPSMC proliferation analysis.

Cytokine analysis.

We measured IL-4 level in the cell-free bronchoalveolar lavage fluid (BALF) from each mouse and endothelin-1 (ET-1) level in the human PMVEC culture medium by using individual ELISA kits for mouse IL-4 (eBioscience, San Diego, CA) and human ET-1 (Enzo Life Sciences, Farmingdale, NY), respectively.

Cell viability and signaling.

Cell viability and cell signaling were analyzed as described previously (60, 63). Briefly, mouse PMVECs (2 × 10⁴) grown in a 96-well plate were serum-starved overnight and then stimulated with vehicle (bovine serum albumin) or rHIMF (1–50 nM) for 24 h. Viable cells were quantified with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) cell proliferation kit (Roche Applied Science, Upper Bavaria, Germany) according to the manufacturer's directions. For cell signaling analysis, cells were grown to ∼80% confluence before being serum-starved and treated with rHIMF (20 nM) or vehicle for 30 or 60 min. The cells were then lysed with lysis buffer (Cell Signaling Technology, Danvers, MA) containing 1 mM phenylmethylsulfonyl fluoride and phosphatase inhibitor cocktail (Sigma-Aldrich). The cell lysates were used for immunoblot analysis as described below.

HPSMC proliferation assay.

HPSMCs (1 × 10⁴) were cultured in 96-well culture plates and serum-starved for 24 h before the proliferation assay. The medium was replaced with SmBM, VEM-2 (1% FBS), or PMVEC-conditioned medium (100 μl/well), and cells were cultured for 24 h. Cell proliferation was determined by the MTT method as described above. Cells exposed to SmBM with 5% FBS were used as positive controls, and those exposed to SmBM without serum or growth factors were considered negative controls. Serum-reduced VEM-2 medium (1% FBS) was used as a negative control for the cells treated with conditioned medium from human PMVECs.

Immunofluorescence and confocal microscopy.

Immunofluorescence staining was carried out as described previously (1). Briefly, the paraffin sections were blocked with appropriate blocking serum (Vector Laboratories, Burlingame, CA) for 1 h at room temperature and then treated with anti-vWF (Dako), anti-SMA (Dako), anti-thrombomodulin (Abcam, Cambridge, UK), or anti-cleaved caspase 3 (Casp3; Cell Signaling Technology) antibodies. Next, the sections were incubated with the appropriate fluorochrome-coupled secondary antibodies (Jackson ImmunoResearch, West Grove, PA). Finally, the sections were washed in PBS, mounted with ProLong Gold antifade reagent with DAPI (Invitrogen), and covered and sealed with a glass cover slip. Negative control sections for the immunohistochemical experiments received identical treatments but were not exposed to the primary antibody; they showed no specific staining. Apoptotic cells in mouse lung tissue were detected by using the FragEL DNA Fragmentation Detection Kit (EMD Millipore). Staining was imaged with a Zeiss 510 Meta confocal microscope (Carl Zeiss Microscopy, Thornwood, NY), and image quantification was carried out with Volocity image software (PerkinElmer, Waltham, MA) at the Johns Hopkins School of Medicine Microscope Facility. For each image, we selected intra-alveolar vessels (≤80 μm; ×1,000 magnification), and the software automatically quantified mean fluorescence intensity of the apoptotic cells and DAPI in the vasculature. Values were normalized by the total cell density. We also quantified mean fluorescence of TUNEL-positive cells (human and mouse PMVEC) on lower-magnification photomicrographs (×10 objective) and normalized by the cell density in each photomicrograph.

Immunocytochemistry.

Human and mouse PMVECs were grown as described above, and immunocytochemistry was performed. Cells were cultured in Lab-Tek Chamber slides (Thermo Scientific, Rochester, NY). To exclude possible apoptosis from serum starvation, we used 1% FBS in the medium. After 6 h of stimulation with various agents, cells were fixed with 4% paraformaldehyde and permeabilized with 0.5% Triton X-100 in PBS. Click-iT TUNEL Alexa Fluor Imaging Assay reagents (Invitrogen) were used to detect apoptotic cells. For angiopoietin 2 (Ang2) expression analysis, cells were fixed and permeabilized with methanol for 5 min and then incubated with the corresponding serum for 1 h. Cells were then exposed to anti-Ang2 (Santa Cruz Biotechnology, Santa Cruz, CA) or anti-vWF (Dako) antibodies at 4°C overnight, extensively washed, and incubated with appropriate fluorochrome-coupled secondary antibodies (Jackson ImmunoResearch) for 45 min. Slides were sealed and analyzed with a Zeiss 510 Meta confocal microscope as described above.

Immunoblotting.

Protein samples were prepared with extraction buffer as described previously (60, 61). Briefly, mouse lung tissue lysates or cell lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and protein was transferred to nitrocellulose membrane. Membranes were incubated with primary antibodies against mouse vascular cell adhesion molecule-1 (VCAM-1; Santa Cruz Biotechnology), VEGF-A (Santa Cruz Biotechnology), cleaved Casp3 (Cell Signaling Technology), Jun NH₂-terminal kinase (JNK; Cell Signaling Technology), p38 mitogen-activated protein kinase (MAPK; Cell Signaling Technology), or β-actin (Sigma-Aldrich). After membranes were incubated with HRP-conjugated secondary antibody (Bio-Rad Laboratories, Hercules, CA), the signal was visualized by an ECL substrate kit (GE Healthcare, Buckinghamshire, UK). Each band was quantified by Image J software (National Institutes of Health, Bethesda, MD).

Statistical analysis.

All data are presented as means ± SE. Differences between multiple groups were compared by ANOVA followed by Bonferroni's multiple-comparison tests. Two-group analysis was performed by Student's t-test. A value of P < 0.05 was considered statistically significant.

RESULTS

HIMF induces early EC apoptosis in the pulmonary vasculature in vivo.

We have previously shown that HIMF induces pulmonary vascular inflammation in a VEGF (60)- and IL-4 (62)-dependent manner at 7 days after injection. In the present study, we found that HIMF injection caused a significant increase in apoptosis-positive cells in the lung vasculature of both WT and IL-4 KO mice compared with that in vehicle-treated controls; however, the degree of EC apoptosis was significantly less in IL-4 KO mice (Fig. 1, A and B). Similarly, the number of Casp3-positive cells in the intra-alveolar vessels was increased in HIMF-treated WT and IL-4 KO mice compared with that in vehicle-treated controls (Fig. 1C). These data suggest that HIMF treatment promotes early EC apoptosis in the pulmonary vasculature in both WT and IL-4 KO mice but that presence of IL-4 enhances the effect of HIMF.

Fig. 1.

Hypoxia-induced mitogenic factor (HIMF) induces pulmonary endothelial cell (EC) apoptosis in vivo. A: HIMF significantly increases apoptotic (TUNEL-positive) cells in the lung vasculature. Apoptotic cells in the lung vasculature of wild-type (WT) and interleukin (IL)-4 knockout (KO) mice 7 days after treatment with HIMF or vehicle were detected as green fluorescence, and nonapoptotic cells were detected by DAPI (blue) staining. PV, pulmonary vessel. B: quantification of the fluorescence intensity of apoptotic cells. *P < 0.05 vs. vehicle-treated controls of each genotype; #P < 0.05 vs. WT HIMF. Values are means ± SE (n = 6–8 images from 4 animals/group, by ANOVA). C: vascular endothelium and cleaved caspase 3 (Casp3)-positive cells colocalized in the pulmonary vasculature of WT and IL-4 KO mice 7 days after treatment with HIMF or vehicle (arrows). EC [von Willebrand factor (vWF)-positive cells, green] and cleaved Casp3-positive cells (red) are shown. Scale bars: 50 μm. VS, vessel.

HIMF induces EC apoptosis in mouse PMVECs.

We next examined whether HIMF causes EC apoptosis in cultured PMVECs. We have shown previously that, at concentrations <10 nM, HIMF significantly increases EC proliferation (60). However, here we found that, at higher concentrations (50 nM, 24 h), HIMF significantly decreased cell viability (Fig. 2A). HIMF treatment significantly activated Casp3 (Fig. 2B) and increased TUNEL-positive cells (Fig. 2, C and D). Because we have previously shown that HIMF promotes reactive oxygen species production in cultured cells, we tested whether HIMF causes EC apoptosis by activating stress-activated kinases such as JNK and p38 MAPK. These kinases are generally referred to as stress-stimulated MAPKs and are required for the induction of apoptosis by diverse stimuli, including oxidant stress (54). Here, we found that HIMF promoted significant phosphorylation of both JNK and p38 MAPK in mouse PMVECs within 30 min (Fig. 2, E and F). These results suggest that HIMF promotes apoptosis through Casp3 and its downstream stress-activated kinases in mouse PMVECs.

Fig. 2.

HIMF induces pulmonary EC apoptosis in vitro. A: HIMF stimulated cell growth at lower concentrations and cell death at higher concentrations (>10 nM) in mouse pulmonary microvascular ECs (PMVECs) (24 h). *P < 0.05 vs. vehicle-treated controls. Values are means ± SE (n = 5/group, by ANOVA). B: HIMF stimulation (20 nM) significantly increased cleaved caspase 3 activation in mouse PMVECs. C and D: the number of TUNEL-positive cells was significantly greater after HIMF stimulation (20 nM, 6 h) than after vehicle treatment. Quantification of mean fluorescence intensity of TUNEL is shown as a ratio of DAPI staining (D, analysis by Student's t-test, n = 8 images from each group). E and F: HIMF stimulation (20 nM) significantly increased phosphorylation of stress-activated kinases Jun NH₂-terminal kinase (JNK) and p38 mitogen-activated protein kinase (MAPK) in mouse PMVECs. Values are means ± SE (n = 3/group, ANOVA). ***P < 0.001 and ****P < 0.0001 vs. vehicle-treated controls. Values are means ± SE (n = 3/group, by ANOVA). These experiments were repeated three times with similar results.

HIMF-induced pulmonary vascular inflammation is IL-4-dependent.

We and others have shown that treating mice with HIMF increases pulmonary vascular inflammation by upregulating lung VEGF-A (60) and VCAM-1 (55). VEGF-A has been shown to be chemotactic for monocytes and macrophages, and VCAM-1 mediates the adhesion of inflammatory cells to the vascular endothelium (49). We also have shown that HIMF treatment increases IL-4 expression, suggesting that HIMF amplifies Th2 response in the lung (62). In this context, we further examined whether HIMF-induced pulmonary vascular inflammation is dependent on Th2 cytokine IL-4. HIMF-induced VEGF-A and VCAM-1 expression was completely abolished in lungs from IL-4 KO mice compared with that in WT groups (Fig. 3, A and B). Based on our microarray analysis, we also analyzed PH-related gene markers of vascular inflammation and found that HIMF-induced expression levels of Ccl5 (RANTES), Ang2, platelet-derived growth factor-α (PDGFA), matrix metalloproteinase-2 (MMP2), VCAM-1, and tissue inhibitor of metalloproteinase 2 (TIMP2) were significantly suppressed in IL-4 KO mice compared with those in WT mice (Fig. 3C). These data suggest that HIMF requires endogenous IL-4 to cause vascular inflammation in the lung.

Fig. 3.

HIMF-induced early pulmonary vascular inflammation is IL-4 dependent. A: HIMF treatment significantly elevated protein expression of vascular cell adhesion molecule-1 (VCAM-1) and vascular endothelial growth factor (VEGF)-A in the lungs of WT mice but not of IL-4 KO mice. B: quantification of VEGF-A and VCAM-1 is shown as a ratio of each protein to β-actin. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 vs. WT control; ##P < 0.01 and ####P < 0.0001 vs. WT HIMF (n = 4 animals/group). C: quantitative RT-PCR showed that, 7 days after HIMF injection, the expression of vascular inflammatory marker genes was significantly increased in lung tissue of WT mice but not in lung tissue of IL-4 KO mice. Values are means ± SE (n = 3/group).

Both HIMF and IL-4 activate PMVECs.

Next, we examined whether HIMF directly induces vascular inflammatory molecules in cultured mouse PMVECs. Of the HIMF-induced inflammatory genes listed in Fig. 3C, we chose to analyze VCAM-1 and Ang2 because they are known to be expressed by ECs. It is also well known that Ang2, in the presence of other stimuli such as VEGF, induces inflammation and vasculogenesis (16), and we have shown that HIMF enhances the VEGF signaling pathway (60). In the present study, both HIMF and rmIL-4 increased VCAM-1 protein expression (Fig. 4A) and Ang2 gene and protein expression in mouse PMVECs (Fig. 4, B and C), suggesting that both HIMF and IL-4 activate PMVECs.

Fig. 4.

HIMF and IL-4 activate VCAM-1 and angiopoietin 2 (Ang2) expression in mouse PMVECs. A–C: primary mouse PMVECs were treated with or without HIMF (20 nM) or rmIL-4 (50 ng/ml) for 24 h. A: VCAM-1 protein expression was analyzed in cell lysates (left). The ratio of VCAM-1 to β-actin is shown (right). ***P < 0.001 vs. vehicle control; n = 3/group. B: cells were stained with fluorescent antibodies to Ang2 and vWF. DAPI was used to stain the nuclei. Images are shown at ×630 magnification. C: Ang2 gene expression was analyzed by quantitative RT-PCR. *P < 0.05 and **P < 0.01 vs. control (n = 3/group). These experiments were repeated at least three times with similar results.

Human RETN and IL-4 induce apoptosis and activate Ang2 expression in human PMVECs.

To validate the results from mouse cells in human cells, we examined whether the effect of hRETN on human PMVECs was similar to that of mHIMF on mouse PMVECs. Both hRETN and rhIL-4 treatment significantly increased the number of PMVECs that underwent apoptosis (Fig. 5, A and B) and the level of Ang2 expression (Fig. 5, C and D). Unlike in mouse PMVECs, Ang2 and vWF colocalized in human PMVECs, a finding that supports previous observations (16).

Fig. 5.

Human resistin (RETN) and IL-4 induce EC apoptosis and activation, and the conditioned media from activated ECs enhances human pulmonary artery smooth muscle cell (HPSMC) proliferation. A: both RETN (200 ng/ml, 6 h) and IL-4 (50 ng/ml, 6 h) caused TUNEL-positive reactions in human PMVECs (×630 magnification). B: quantification of mean fluorescence intensity of RETN- and IL-4-induced TUNEL-positive cells is shown as a ratio of DAPI staining [n = 8 low-magnification (×100) images from each group]. C: colocalization of RETN- or IL-4-induced Ang2 and vWF in PMVECs. Primary human PMVECs were treated with or without rhRETN (100 ng/ml) or rmIL-4 (50 ng/ml) for 24 h. Images are shown at ×630 magnification. D: quantification of mean fluorescence intensity of RETN- and IL-4-induced Ang2 expression is shown as a ratio of DAPI staining [n = 8 low-magnification (×200) images from each group]. E: PMVECs were treated with 100 ng/ml rhRETN or 50 ng/ml rhIL-4 for 24 h, and the conditioned medium was added to serum- and growth factor-starved HPSMCs for 24 h. Serum-free smooth muscle basal medium (SmBM) was used as a negative control. rhRETN (100 ng/ml) or rhIL-4 (50 ng/ml) was added directly to HPSMCs for comparison. F: PMVECs were treated with 100 ng/ml rhRETN for 24 h, and ET-1 level in the conditioned medium was measured by ELISA. Values are means ± SE (n = 4 separate culture cells/group, Student's t-test). Results are shown as means ± SE. ****P < 0.0001 compared with serum-free SmBM medium; ####P < 0.0001 compared with EC medium control (VEM-2 media with 1% FBS); n = 6/group. These experiments were repeated at least three times with similar results.

Media from human RETN- and IL-4-activated PMVECs enhances HPSMC proliferation.

It has been suggested that pulmonary ECs are involved in pulmonary vascular remodeling not only through their ability to control vascular tone but also through the production and release of smooth muscle cell (SMC) growth factors (13). Pulmonary ECs that undergo apoptosis are known to produce factors that enhance SMC growth and inhibit SMC apoptosis (48). In this context, we treated HPSMCs with conditioned media from PMVECs treated with or without rhRETN (100 ng/ml) or rhIL-4 (50 ng/ml) and then analyzed the degree of SMC proliferation. Conditioned medium from PMVECs that had been treated with rhRETN significantly enhanced HPSMC proliferation, but exogenous administration of the same concentration of rhRETN did not affect HPSMC proliferation. However, both conditioned medium from rhIL-4-stimulated PMVECs and directly applied rhIL-4 promoted significant increases in HPSMC proliferation (Fig. 5E). Next, we tried to determine the factors that are produced by PMVECs in response to hRETN. Because hRETN is thought to induce ET-1 in the cardiovascular system (58) and ET-1 is a highly potent vasoactive and SMC mitogenic peptide present at high concentrations in patients with PH (50), we analyzed ET-1 level in the PMVECs with or without hRETN stimulation. hRETN (100 ng/ml, 24 h) significantly increased ET-1 production in PMVECs (19.57 ± 9.11 pg/ml increase in response to hRETN; Fig. 5F).

HIMF causes IL-4-dependent PH development.

Recent studies have suggested that pulmonary vascular inflammation and augmented Th2 immune response contribute to the pathogenesis of severe PH (47), and we have shown that HIMF can induce EC apoptosis and PH-related vascular inflammatory genes in the lung (Figs. 1, 2, and 3). In this context, we further evaluated whether one injection of HIMF can cause PH development in mice as a consequence of lung vascular inflammation. In addition, based on our findings that PH-related vascular inflammatory markers were completely suppressed in the lungs of HIMF-treated IL-4 KO mice (Fig. 3C), we investigated whether PH development is affected by early IL-4-mediated lung inflammation. At 30 days after HIMF injection (PH development phase), WT mice exhibited significantly greater right heart hypertrophy, higher right heart pressure, and lower cardiac output than did vehicle-injected WT mice (Fig. 6, A–C). These hemodynamic changes were absent in IL-4 KO mice, indicating that HIMF causes development of PH by an IL-4-dependent mechanism.

Fig. 6.

HIMF-induced pulmonary hypertension (PH) development is significantly suppressed in IL-4 KO mice. WT and IL-4 KO mice were administered one iv injection of HIMF or vehicle (control). Hemodynamics and the right ventricular ratio were analyzed 30 days postinjection. A: right ventricular weight/(left ventricular + septal weight) [RV/(LV + S)]; n = 4–7/group. B: right ventricular end systolic pressure (RVSP). C: cardiac output (CO). Data are shown as means ± SE; n = 3–5/group. *P < 0.05 and ***P < 0.001 vs. WT control. ##P < 0.01 vs. WT HIMF.

HIMF-induced pulmonary vascular remodeling is significantly suppressed in IL-4 KO mice.

HIMF treatment caused pulmonary vascular media hypertrophy in medium vessels (100–150 μm) and increased perivascular accumulation of inflammatory cells in WT mice. However, these phenomena were completely suppressed in IL-4 KO mice (Fig. 7A). In addition, HIMF treatment caused a significant decrease in the number of nonmuscularized vessels and increases in the number of partially muscularized and fully muscularized vessels in WT mice. However, this HIMF-induced pulmonary vascular remodeling was significantly diminished in IL-4 KO mice (Fig. 7B). We also analyzed IL-4 levels in the BALF from WT mice 30 days after stimulation with or without HIMF. We confirmed that HIMF treatment caused a significant and persistent increase in IL-4 in the lung BALF during the PH development phase (Fig. 7C). Immunohistochemical analysis revealed that HIMF increased SMA- and thrombomodulin-positive cells and cellular activities (shown as DAPI-positive cells) in WT mice. However, these changes were strongly suppressed in IL-4 KO lung vasculature (Fig. 7D). These results suggest that HIMF-induced pulmonary vascular remodeling is, at least in part, mediated by an IL-4-dependent mechanism.

Fig. 7.

HIMF-induced pulmonary vascular remodeling is significantly suppressed in IL-4 KO mice. A: hematoxylin- and eosin-stained lung tissue from WT and IL-4 KO mice 30 days after iv injection with HIMF or vehicle. Scale bars: 100 μm. B: bar graph showing the percentage of small pulmonary arteries in WT and IL-4 KO mice that are nonmuscular (NM), partially muscular (PM), or fully muscular (FM). **P < 0.01 and ****P < 0.0001 vs. WT control; ##P < 0.01 and ####P < 0.0001 vs. WT HIMF. More than 500 vessels were counted in each group (3–5 animals/group). C: ELISA was used to measure IL-4 level in bronchoalveolar lavage fluid (BALF) from WT mice 30 days after treatment with or without HIMF. Values are means ± SE (n = 6/group, Student's t-test). D: systemic injection of WT mice with HIMF increased smooth muscle actin (SMA)-positive cells (identifying vascular media) and thrombomodulin-positive cells (identifying vascular intima). These increases were strongly suppressed in IL-4 KO mice. Scale bars: 50 μm (×1,000 magnification).

DISCUSSION

The present study showed that a single systemic injection of rHIMF can cause EC apoptosis, lung vascular inflammation, and subsequent PH development. HIMF promoted apoptosis of ECs and the induction of vascular inflammatory molecules such as Ang2 and VCAM-1, which are known to recruit inflammatory cells to the vascular beds. These molecules were significantly increased in the lungs of HIMF-treated WT mice, suggesting that induction of EC apoptosis and activation by HIMF, at least in part, triggers an inflammatory response in the lung that requires IL-4. Release of endogenous IL-4 from the immune cells is then critical to causing early pulmonary vascular inflammation and subsequent PH (Fig. 8).

Fig. 8.

Proposed pathway by which HIMF mediates EC apoptosis, pulmonary vascular inflammation, and PH development. Schematic diagram showing our conceptualization of the pathway by which HIMF activates pulmonary EC apoptosis (IL-4 independent) and subsequent vascular inflammation that leads to PH development (IL-4 dependent). Augmented HIMF expression in the pulmonary vasculature causes EC apoptosis via the caspase 3-stress-activated kinase pathway and IL-4-dependent vascular inflammation in the early phase (day 7). These changes lead to subsequent PH (day 30). Human RETN, a human homolog of HIMF, also causes pulmonary EC apoptosis and promotes EC-mediated alteration of pulmonary vascular smooth muscle cells to a proliferative phenotype. Results from previous studies are shown in the dash-lined box (60, 62), and results from the current study are shown in solid-lined boxes.

We have shown that HIMF induces apoptosis and activates Casp3 in vivo and in vitro. In cell culture studies, we confirmed that PMVECs do not release lactate dehydrogenase (a hallmark of necrosis), suggesting that this is not necrotic cell death (data not shown). Notably, HIMF activated kinases JNK and p38 MAPK, which induce apoptosis under stress-related conditions. It also has been shown that long-term overexpression of resistin in the rat heart induces myocardial dysfunction and remodeling associated with a complex phenotype of oxidative stress, inflammation, and cardiomyocyte apoptosis (7). That report supports our current findings that resistin causes apoptosis in certain cell types.

Our finding that HIMF induces significant EC apoptosis in both WT and IL-4 KO mice is not unexpected because IL-4 is expressed mostly in immune cells, not in lung resident cells such as ECs. Although it is well known that IL-4 is produced by antigen-specific Th2 lymphocytes, it can also be produced by cells of the innate immune system such as mast cells, basophils, eosinophils, macrophages, and innate lymphoid-like cells (or nuocytes). Early production of Th2 cytokines IL-4 and IL-13 by innate cells is thought to be important for the initiation phase of Th2-dominant inflammation (9). However, ECs do not express or produce IL-4; therefore, endogenous IL-4 is not required for EC apoptosis. Gene expression data from our laboratory also have confirmed that PMVECs do not express IL-4 gene at baseline or after HIMF stimulation (data not shown). ECs do express type I and type II IL-4 receptors, though, and our in vitro study showed that exogenous IL-4 could induce apoptosis in these cells. Moreover, in vivo, HIMF injection caused significantly less EC apoptosis in IL-4 KO mice than in WT mice. These data suggest that IL-4 produced by immune cells in response to HIMF amplifies EC apoptosis in the lung.

We have shown previously that HIMF injection causes a significant increase in IL-4 expression and amplifies Th2 response in the lung (62). In the present study, we also confirmed that IL-4 is augmented in BALF of HIMF-treated WT mice during the PH development phase (day 30; Fig. 7C). This finding suggests that HIMF-induced amplification of Th2 inflammation is persistent. Thus, it is possible that induction of IL-4 by HIMF stimulation also promotes persistent EC activation and vascular inflammation in the lung.

We also confirmed that HIMF promotes EC apoptosis, Th2-mediated vascular inflammation, and PH development without costimulation by chronic hypoxia. Similar to VEGFR blockade-initiated PH in animal models, VEGFR2 expression was downregulated in the lungs of HIMF-treated animals during the early inflammation phase (60).

In our previous study, systemic injection of rHIMF caused IL-4-dependent early lung inflammation in part by recruiting macrophages to the lung (62). Here, we focused on dissecting the mechanisms of vascular inflammation that are involved in human and animal models of PH. HIMF induced PH-related vascular inflammatory marker genes Ccl5 (RANTES) (12, 15, 34), Ang2 (30), PDGFA (5, 24, 26), MMP2 (28, 33), VCAM-1 (34, 46), and TIMP2 (46) in WT mice, but not in IL-4 KO mice (Fig. 3C). The chemokine gene Ccl5 was particularly elevated in WT mice (10.7 ± 1.30-fold; Fig. 3C). CCL5/RANTES is a key member of the IL-8 superfamily of cytokines and an important chemoattractant for monocytes/macrophages and lymphocytes that constitute the main cell population within the perivascular infiltrate of PH (43). In addition, it has been reported that intranasal administration of rHIMF/FIZZ1 protein in naive WT mice leads to airway eosinophilia, peribronchial fibrosis, and airway inflammation (11). These data suggest that HIMF has a potent immunomodulatory effect in the lung.

Our results confirmed that HIMF promotes EC apoptosis, Th2-mediated vascular inflammation, and PH development without costimulation by chronic hypoxia. It has been shown that EC apoptosis induced by VEGFR blockade with SU-5416, combined with chronic hypoxia, produces severe angio-obilterative PH in adult rats (52). That landmark study led to the concept that EC apoptosis may trigger both degenerative and reactive proliferative events that ultimately cause increased pulmonary vascular resistance and the typical vascular pathology of PH. More specifically, the results imply that EC apoptosis increases in response to the loss of survival signaling at initiation, creating conditions that favor the emergence of apoptosis-resistant cells with increased growth potential. The data suggest an underlying mechanism whereby VEGFR blockade by SU-5416 causes initial apoptosis of lung ECs as the “first hit,” and, in the “second hit,” chronic hypoxia exposure generates severe PH in rats (52). Daley and colleagues (8) demonstrated that allergic inflammation after OVA immunization causes muscularization of pulmonary arteries, but not development of PH. HIMF/RELMα was upregulated in the lungs of animals exposed to either OVA immunization or chronic hypoxia, suggesting that it is a common mediator in both PH models (3, 8). This commonality may be one of the reasons that HIMF treatment alone is sufficient to cause development of PH. The angiopoietin-Tie ligand receptor system has a key role in regulating vascular integrity and quiescence (16). Constitutive Ang1 expression and low-level Tie2 phosphorylation in the adult vasculature suggest that Ang1-Tie2 signaling functions as the default pathway to control vascular quiescence (10). However, Ang2 is an antagonist for Ang1-mediated Tie2 functions; Ang2 mRNA is almost absent in the quiescent resting vasculature and dramatically upregulated in response to stimuli during vasculogenesis and inflammation (16). Ang2 also is known to facilitate vasculogenesis when it functions in concert with VEGF (20). We have shown previously that HIMF induces vascular sprouting in a VEGF-dependent manner (60). Here, we showed that HIMF and hRETN significantly increase Ang2 expression in PMVECs. Thus, our data suggest that VEGF and Ang2 are important mediators of HIMF-induced vasculogenesis. Moreover, because Ang2 expression has been implicated in human and experimental PH and heart failure (14, 29, 30, 38), it is considered to be a promising biomarker of disease severity and treatment response in patients with PH.

Our data also show that HIMF-activated primary pulmonary ECs release vWF, a marker for EC activation, as well as Ang2. Both are biomarkers for human PAH (30, 37). In humans, these molecules are exocytosed from the EC-specific secretory organelle, the Weibel-Palade body (21). Our data are consistent with observations that both molecules are released from the same area (Fig. 5C). However, our results using mouse PMVECs show that Ang2 and vWF are produced in the perinuclear region rather than in the EC membrane (Fig. 4B). These data support those of Hol et al. (22), who reported that Weibel-Palade bodies may not to be involved in regulated secretion of chemokines in the mouse; instead, chemokines may be released from cytoplasmic granules. EC exocytosis components are known to mobilize stem and progenitor cells for the regeneration of an injured organ (31). Indeed, we have shown previously that HIMF causes strong proliferative activity in human mesenchymal stem cells (27), suggesting that it has separate and distinct effects on precursor cells and lung resident endothelium in response to injury.

hRETN has been shown to activate ECs in cardiovascular diseases (17, 58); however, this is the first report to show its effect on apoptosis, Ang2 induction, and EC-mediated SMC proliferation. Our results show that both hRETN, one of the human homologs of HIMF, and IL-4 induce pulmonary EC apoptosis. hRETN is produced mainly by myeloid cells, particularly macrophages (45), and its expression pattern shows a greater similarity to that of murine HIMF (RELMα) than to that of murine RETN. The typical serum level of hRETN is 5–20 ng/ml under normal conditions (42, 58), and it increases under inflammatory conditions such as atherosclerosis, diabetes, and cancer (17). By comparison, circulating levels of RELMβ, the other human homolog of HIMF, range from 0.087 to 0.167 ng/ml under both normal and pathological conditions (42). IL-4 also is known to induce EC (systemic origin) apoptosis through the Casp3-dependent pathway during the pathophysiological progression of atherosclerosis (32). Apoptotic cells are detected in neointima of atherosclerotic vessels, and ECs are thought to undergo apoptosis in response to the proinflammatory cytokines produced by activated macrophages and T lymphocytes as a consequence of the ongoing local immune and inflammatory response characteristic of atherosclerosis (18). It is reasonable to speculate that a similar scenario applies to the pulmonary vasculature under pathological conditions. In response to injury, the interaction between ECs and SMCs contributes to new vessel formation and vascular remodeling as a part of the wound repair process. When PMVECs were treated with SU-5416 to induce apoptosis, the conditioned medium from apoptosed ECs was characterized by elevated TGF-β and VEGF, and it increased SMC proliferation compared with that produced by conditioned medium from nonapoptosed ECs (48).

In the present study, we showed that ECs stimulated with hRETN produced significantly more ET-1 than did untreated ECs (Fig. 5E). ET-1 was reported to have a mitogenic effect on pulmonary vascular SMCs (59), and a recent study showed that ET-1 causes pulmonary vascular SMCs to release inflammatory cytokines that contribute to pulmonary vascular remodeling (64). These data suggest that increased ET-1 release from apoptotic ECs in response to hRETN can transform pulmonary vascular SMCs to a proliferative and proinflammatory phenotype. Thus, an hRETN-mediated EC apoptosis-SMC phenotype alteration loop could result in the progression of PH.

One limitation of our mouse PH model is that we did not observe occlusive lesions in the remodeled small vessels, even though we saw dramatic thickening of the media and perivascular cell infiltrate (Fig. 7D). We also performed the in vivo mouse study using a combination of HIMF injection and chronic hypoxia to establish PH; however, the pulmonary pathology and right heart hypertrophy neither altered nor worsened (data not shown). In a previous study, we found that systemically administered exogenous HIMF colocalized with highly proliferative cells in the pulmonary vasculature during the early inflammation phase (day 7 after injection) and amplified IL-4 expression (62). However, in this study, we did not detect HIMF expression in the lung during the PH development phase (data not shown).

Although here we found that HIMF caused EC apoptosis, we have reported previously that HIMF promotes proliferation of specific cell types by mediating the cell survival proteins Akt and ERK (27). Indeed, we have shown that low concentrations of HIMF (<10 nM) have proliferative effects on ECs (60), suggesting that HIMF has dual and separate actions depending on its concentration. Thus, HIMF may be critical to EC fate and subsequent immune response.

Although our previous study in mouse PMVECs showed that mHIMF significantly increases production of VEGF, a potent HPSMC growth factor (60), rhRETN did not induce VEGF expression in human PMVECs or lung epithelial cells (data not shown). Based on our current studies, we suggest that mHIMF and hRETN have both similar and distinct effects on PMVECs and SMCs. Hence, it is important to dissect these differences of RELM proteins in rodents and humans.

In conclusion, we have shown that systemic injection of rHIMF protein promotes EC apoptosis and production of factors that stimulate immune cell activation in the pulmonary endothelium. These actions lead to vascular inflammation and development of PH via an IL-4-dependent mechanism. HIMF/FIZZ1/RELMα is likely a shared mediator of Th2 inflammation and chronic hypoxia (3, 8). Therefore, dissecting the pathogenic role of the HIMF-mediated immune response during PH development might contribute to the identification of novel diagnostic and therapeutic targets for pulmonary arterial remodeling and right heart dysfunction. More detailed, cell-specific analysis of the mechanisms by which HIMF/hRETN mediates Th2 inflammation is warranted and is ongoing in our group.

GRANTS

This work was supported by National Institutes of Health (NIH) Specialized Centers of Clinically Oriented Research (SCCOR) Grant P50-HL-084946 (to R.A.J.), NIH Centers for Advanced Diagnostics and Experimental Therapeutics in Lung Diseases (CADET I) Grant P50-HL-107182 (to R.A.J.), and PHA/ATS/Pfizer Research Fellowship in Pulmonary Arterial Hypertension (to K.Y.-K.).

DISCLOSURES

No conflicts of interest, financial or otherwise are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: K.Y.-K. and R.A.J. conception and design of research; K.Y.-K., E.T., A.Z., N.C.W., L.W.M., and C.C. performed experiments; K.Y.-K., E.T., A.Z., and A.E.B. analyzed data; K.Y.-K. and E.T. interpreted results of experiments; K.Y.-K. prepared figures; K.Y.-K. drafted manuscript; K.Y.-K., E.T., and R.A.J. edited and revised manuscript; K.Y.-K., E.T., A.Z., N.C.W., L.W.M., A.E.B., C.C., and R.A.J. approved final version of manuscript.