Heme Oxygenase-1 Modulates Human Respiratory Syncytial Virus Replication and Lung Pathogenesis during Infection

The human respiratory syncytial virus (hRSV) is the leading cause of lower respiratory tract illness (LRTI) in infants and children worldwide. hRSV produces reinfections throughout life, generating frequent milder respiratory infections in adults as well as severe LRTIs in pediatric, elderly, and immunocompromised patients (1). LRTI can manifest as bronchiolitis or pneumonia, with the risk of death due to respiratory failure (2). Epidemiological studies suggest that hRSV contributes to nearly 33.8 million new episodes of LRTI in children <5 y of age each year, with 3.4 million annual hospital admissions worldwide (3). Furthermore, hRSV infection is linked to neurologic symptoms in patients, as well as learning impairment in animal models (4, 5)

Currently, there are still no licensed vaccines nor specific antiviral drugs for the prophylaxis of hRSV disease in children and other susceptible populations (6). The only approved therapeutic approaches are palivizumab (7), a humanized mAb that protects against hRSV infection in high-risk infants, and ribavirin, an antiviral nucleoside analogue that is rarely used due to toxicity concerns and questionable benefits (7). For these reasons, intense research has focused on finding novel vaccines or antiviral agents to prevent hRSV infection.

During hRSV disease, epithelial cells in the distal airway respond to viral infection by secreting proinflammatory cytokines and chemokines that promote an exacerbated recruitment of infiltrating cells, mainly neutrophils, leading to inflammation and tissue damage (8). Therefore, the design of new strategies to prevent hRSV diseases must consider parameters such as the inhibition of viral replication and reduction of lung inflammation.

Heme oxygenase-1 (HO-1) is a metabolic enzyme that catalyzes the degradation of heme into carbon monoxide, biliverdin, and free iron (9). This enzyme has anti-inflammatory and antioxidant properties, which modulate host innate and adaptive immune responses (10). The immunomodulatory capacity of HO-1 has been demonstrated in several models, such as the LPS-induced acute lung inflammation in which HO-1 activation decreased the migration of polymorphonuclear leukocytes to the lung, reducing oxidative tissue damage (11). Furthermore, the pharmacological induction of HO-1 inhibits dendritic cell (DC) activation and immunogenicity (12), suppressing cytokine secretion and the capacity to prime T cells (13). Also, recent studies have shown that HO-1 can display important antiviral properties. Specifically, upregulation of HO-1 was shown to diminish infection by several viruses, including Ebola, influenza, enterovirus, hepatitis C virus (HCV), hepatitis B virus, and HIV. During infection by these viruses, HO-1 induction protects infected tissues, such as the liver and lungs, from virus-induced oxidative injury (10). Although the mechanisms of action for HO-1 during viral infection have not been elucidated, available data suggest a direct effect on virus proteins or on the activation of cellular processes that interfere with virus replication, such as the type I IFN response (14).

The HO-1 expression can be highly induced by analogs of heme, such as hemin and other metalloporphyrins. The best-characterized inductor of HO-1 is cobalt protoporphyrin (CoPP), which promotes the upregulation of HO-1 gene expression (15). Thus, the primary mechanism involved in the upregulation of the HO-1 enzyme seems to be the enhancement of gene transcription (15, 16). Allosteric HO-1 inhibitors can also induce upregulation of HO-1 expression, as occurs with tin protoporphyrin IX dichloride (SnPP) (17). However, despite inducing HO-1 expression, SnPP irreversibly inhibits the activity of this enzyme (18).

Taking into consideration the multifunctional properties of HO-1, in this study we examined the effects of HO-1 induction on the pathogenesis caused by hRSV in vitro and a mouse infection model. Specifically, we assessed the modulatory effects of HO-1 on airway epithelial cells infected with hRSV. Additionally, we evaluated whether either CoPP or transgenic induction of HO-1 displayed antiviral and anti-inflammatory effects in the airways of hRSV-infected mice.

C57BL/6J and BALB/cJ wild-type mice were obtained from The Jackson Laboratory for the generation of pIi-TTA-TetO-HO-1 transgenic mice. PIi-TTA mice were a kind gift from Christophe Benoist (19). TetO-HO-1 mice, located upstream of the cDNA sequence and located downstream of the cDNA, was cloned at the Not-I/Xho-I sites into the pBluKSM-tet-O-CMV vector containing the followed by the human β-globin intron and the bovine growth hormone polyA.

TetO-HO-1 transgenic mice were generated by pronuclear microinjection of CBA/C57BL6 eggs with a DNA fragment containing a Tet-responsive element downstream a minimal CMV promoter, the human β-globin intron, the human HO-1 cDNA, and the bovine growth hormone polyA. All mice were maintained at the pathogen-free facility of the Pontificia Universidad católica de Chile and manipulated according to guidelines approved by the institution’s Bioethical Committee. hRSV serogroup A2, strain 13018–8, is a clinical isolate obtained from the Instituto de Salud Pública de Chile. In most experiments, a mock control was included consisting of supernatants collected from uninfected human laryngeal epidermoid carcinoma number 2 (HEp-2) cells kept in culture for the same period of time as infected cells.

HEp-2 cells (CCL-23; American Type Culture Collection) were used to propagate hRSV serogroup A2, strain 13018–8 (clinical isolate obtained from the Instituto de Salud Pública de Chile), as previously described (4). Briefly, HEp-2 cell monolayers were grown in T75 flasks with DMEM (Life Technologies, Invitrogen, Carlsbad, CA) supplemented with 10% FBS. Flasks containing 5 ml of culture medium were inoculated with 2 × 10⁵ PFU of hRSV and incubated at 37°C. After viral adsorption (3 h), supernatants were replaced with fresh medium (DMEM 1% FBS) and incubated for 48 h or until the visible cytopathic effect was observed. Cells were harvested, the flask content was pooled, then spun twice at 300 × g for 10 min to remove cell debris. In parallel, supernatants of noninfected HEp-2 monolayers were collected as previously described, and used as noninfectious control (mock). Viral titer of supernatants was determined by immunohistochemistry. Ultraviolet-inactivated hRSV (ultraviolet-hRSV) was obtained exposing ice-packed virus preparation vials for 45–60 min at 302 nm using a 15 W lamp trans-illuminator.

Human alveolar type II–like pulmonary epithelial cells (A549 cells) (kindly provided by Dr. P. Piedra, Baylor College of Medicine) were maintained in DMEM medium containing 10% (v/v) FBS, 100 IU/ml penicillin, and 100 μg/ml streptomycin. A549 cells were treated with vehicle (NaOH diluted in media), or 50 μM CoPP (HO-1 inducer) (Frontier Scientific), or 50 μM SnPP (HO-1 inhibitor) (Frontier Scientific), and incubated at 37°C in 5% CO₂ (19). After 2 h, supernatants were removed, and cells were inoculated with infectious hRSV at a multiplicity of infection (MOI) equal to 1 PFU per cell and incubated for 2 h in OptiMem I Reduced Serum Medium (preinfection). As a control, cells were inoculated either with the same media (untreated [UT]) or with mock (supernatant of uninfected HEp-2 or with ultraviolet-inactivated hRSV). Each treatment was also performed during the hRSV infection for 24 h (postinfection). In both cases, after 24 h replication was determined by measuring the nucleoprotein (N) RNA, and copies were quantified by real-time quantitative PCR (qPCR). In addition, supernatants from infected cells for each treatment were collected at 48 h postinfection to determine viral titer by immuno-plaque assay.

The viral titer in supernatents of bronchoalveolar lavage fluid (BALF) was determined by immunocytochemistry. Briefly, infectious supernatants were serially diluted (10-fold dilutions), added to 96-well plates with HEp-2 monolayers (80% confluence), and incubated for 48 h at 37°C. Later, cells were fixed with 2% paraformaldehyde-PBS and permeabilized 20 min with 0.2% saponin-PBS. Intracellular staining was performed with an anti-N-hRSV (clone 1E9/D1) Ab for 1 h (dilution 1:500, 0.2% saponin-PBS). Cells were washed twice and incubated with anti-mouse IgG-HRP (dilution 1:200; Invitrogen, Molecular Probes) for 45 min. After washing the complex twice, the substrate TRUE-BLUE peroxidase (KPL) was added to cells and incubated for 10 min at room temperature.

Briefly, 100 μl of virus-containing supernatants from hRSV-infected A549 cells at an MOI of 1, and treated with CoPP, SnPP, or vehicle (6–100 μM) were used to coat ELISA plates for 2 h at 37°C. Then, plates were blocked for 2 h with PBS containing 10% of FBS. After washing three times with PBS-tween 20 0.01%, goat anti-hRSV-HRP conjugated (Abcam) diluted in 1:500 in PBS FBS 10% was added to the wells for 2 h. Finally, after washing five times with PBS-tween 20 0.01%, bound Ab was detected by addition of tetramethylbenzidine peroxidase substrate (BD), stopped with 1 M H₂SO₄, and analyzed at 450 nm by an ELISA Plate Reader.

HEp-2 cells were detached using versene solution (Sigma-Aldrich), washed, exposed to 50 μM of CoPP, SnPP, or vehicle, and chilled on ice for 30 min. Cells were washed and exposed to hRSV (MOI = 1) concomitantly with CoPP, SnPP, or vehicle for 1 h at 4°C. For flow cytometry analyses, cells were washed, fixed (4% paraformaldehyde), and anti–hRSV F-protein FITC–conjugated Ab (1:200, ab20391; Abcam) was added and incubated for 1 h at 4°C. For Western blotting analyses, hRSV- exposed cells were washed, resuspended in RIPA buffer, and incubated at 95°C for 10 min. SDS-PAGE, on 10% Bis/Tris gels and in MES buffer (Invitrogen), was performed and the indicated protein was transferred to nitrocellulose membranes. The membranes were then exposed to anti-N protein Ab, anti HO-1 (ab13248; Abcam), or anti–β actin (6221102; BioLegend), followed by HRP-labeled secondary Ab (goat anti-mouse; Invitrogen, Molecular Probes). Chemiluminescence (Amersham ECL Prime Western blotting Detection Reagent; GE Healthcare, Little Chalfont, U.K.) was detected using a thermal imaging system.

Bone marrow–differentiated DCs from C57BL/6 mice were prepared as previously described (20, 21). Analyses for expression of surface markers by flow cytometry revealed a typical phenotypic profile for immature DCs (>75% of CD11c⁺ cells). On day 5 of culture, DCs were inoculated with hRSV or ultraviolet-hRSV for 2 h at an MOI equal to 1 PFU per cell. As controls, DCs were left UT or inoculated with similar volumes of supernatants from uninfected HEp-2 cultures (mock). Then 48 h postinfection, the viability of DCs was determined by trypan blue exclusion. In contrast, immature DCs (1 × 10⁶ cells) were pulsed for 2 h with 50 μM CoPP, SnPP, or vehicle control. Cells were then washed twice and cultured 4 h before hRSV inoculation. Then, HO-1 mRNA and protein levels were analyzed by qPCR and flow cytometry or immunofluorescence, respectively.

Male 6–8 wk old BALB/cJ wild-type mice were pretreated i.p. (7.6 μmol/kg) either with CoPP to induce the HO-1 expression, or with SnPP to inhibit the activity of HO-1 as previously described (22), 24 h before viral challenge. Mice, treated with NaOH (diluted in PBS), were included as the vehicle control. After 24 hr mice were anesthetized with ketamine or xylazine (80 and 8 mg per kg, respectively) and challenged intranasally with either 1 × 10⁶ PFU of hRSV or an equal volume of mock (as non-infectious control). Animal body weight was recorded daily postinfection. At day 4 postinfection, mice were terminally anesthetized by i.p. injection with a mixture of ketamine and xylazine. BALF and lung tissue samples were collected for further analyses.

Female or male 6–8 wk old rtTA-HO-1 mice and littermate mice were treated with 800 μg/ml doxycycline (DOX) and 36 mg/ml sucrose in drinking water, protected from the light, to induce the expression of human HO-1 in MHC class II positive (MHC-II⁺) cells. Then 48 h later, mice were anesthetized with ketamine/xylazine (20 and 1 mg/kg, respectively) and challenged intranasally with either 1 × 10⁶ PFU of hRSV or an equal volume of mock (as non-infectious control). Animal body weight was recorded daily postinfection. At day 4 postinfection, mice were terminally anesthetized by i.p. injection with a mixture of ketamine and xylazine. BALF and lung tissue samples were collected for further analyses.

Mice were bled from the cheek into 100 μl of heparin (125 UI/ml). For DNA purification, we used the DNeasy Blood and Tissue Kit following the manufacturer’s instructions. Consequently, we performed a PCR for each transgene inserted in the transgenic mice: rtTA and tHO-1 using GoTaq G2 Flexi DNA Polymerase (Promega). Then, samples were loaded in an agarose electrophoresis. The mice that only had the tHO-1 gene, but lacked the tTA gene were named littermate (LM), which were used as controls.

Total RNA was isolated from tissues or cell cultures by using the Trizol reagent (Life Technologies, Invitrogen), according to the manufacturer’s instructions. cDNA synthesis from total RNAs was performed using the ImProm-II Reverse Transcription kit (Promega) and random primers. qRT-PCR reactions were carried out using a StepOne plus thermocycler (Applied Biosystems). The abundance of HO-1 and Nrf2 mRNAs was determined by relative expression to the respective housekeeping gene by the 2-ΔΔ threshold cycle (ΔΔCt) method. For N-gene expression, absolute quantification data were expressed as the number of hRSV N-gene copies for each 5 × 10³ copies of β-actin transcript, as previously described (4). The following primers were used: hRSV N Forward 5′-GCTAGTGTGCAAGCAGAAATC-3′ and Reverse 5′-TGGAGAAGTGAGGAAATTGAGTC-3′, mouse HO-1 5′-CCTCTGACGAAGTGACGCC-3′ and Reverse 5′-CAGCCCCACCAAGTTCAAA-3′, human HO-1 Forward 5′-AGGCAGAGGGTGATAGAAGAGG-3′ and Reverse 5′-TGGGAGCGGGTGTTGAGT-3′, mouse Nrf-2 Forward 5′-TTC TTT CAG CAG CAT CCT CTC CAG-3′ and Reverse 5′-ACA GCC TTC AAT AGT CCC GTC CAG-3′, mouse IFN-α Forward 5′-TCC TGA ACC TCT TCA CAT CAA A-3′ and Reverse 5′-ACA GGC TTG CAG GTC ATT GAG-3′, mouse IFN-β Forward 5′-AGC TCC AAG AAA GGA CGA ACA-3′ and Reverse 5′-GCC CTG TAG GTG AGG TTG AT-3′, mouse β-actin Forward 5′-ACCTTCTACAATGAGCTGCG-3′ and Reverse 5′-CTGGATGGCTACGTACATGG-3′.

DCs (1 × 10⁶ cells) were produced as described above. Briefly, DCs were grown over 12 mm microscope cover glasses (Marienfeld-Superior, DE). Then, DCs were inoculated, as mentioned above, with hRSV, ultraviolet-hRSV or mock, and cultured for 48 h at 37°C. Inoculated DCs were prepared for confocal microscopy analysis, as previously described (23). Briefly, DCs were washed and fixed with 2% p-formaldehyde for 15 min at 4°C. Then, cells were permeabilized with 0.05% saponin-PBS for 15 min at 4°C. Next, cover glasses were passed to a cold chamber and DCs were double-stained with 1/200 anti-mouse HO-1 mAb (Abcam), already dissolved in 0.05% saponin-PBS. These preparations were incubated overnight at 4°C in darkness. The next day, cells were washed with PBS and stained with 1/200 goat anti-mouse IgG-Alexa Fluor 488 (Invitrogen) and 1/200 goat-anti rabbit IgG-AF555 (Invitrogen) secondary Abs, and incubated for 3 h at 4°C in darkness. After washing the complex, cells were dried and mounted with DABCO mounting medium for confocal microscopy. Fluorescence measurements were performed on a FluoView FV1000 confocal microscope (Olympus). After image recording (at 40× magnification), each channel (HO-1, nuclei, and transmission) was analyzed separately using Olympus Fluoview version 3.0 software.

DCs, A549, and HEp-2 cells were inoculated, as mentioned above, and cultured for 24 or 48 h, respectively, in the presence of hRSV, ultraviolet light-inactivated hRSV, and mock. DCs were then stained with anti-CD11c-PE-Cy7 (clone HL3; BD Pharmingen), anti-IA/IE-PerCP-Cy7 (clone M5; BD Pharmingen), and anti-hRSV N-AF647 conjugated (clone 1E9/D1) in 10% of FBS in PBS as a blocking solution. For HO-1 intracellular staining, fixed cells were incubated with anti-mouse HO-1 mAb (Abcam) in permeabilization buffer (1% saponin, 10% FBS in PBS) for 45 min at 4°C. Then, cells were washed and stained with goat anti-mouse IgG-AF488 (Invitrogen). For cell infiltration analysis, lung samples were homogenized and filtered using a 40 μm cell strainer. BALF was centrifuged at 300 × g for 5 min, washed, and stained with anti-CD11b-APC (clone CBRM1.5; BD Pharmingen), anti-CD11c-PE (clone CBRM1.5; BD Pharmingen), anti-IA/IE-APC-Cy7 (clone M5; BD Pharmingen), anti-Ly6C-PercCP 5.5 (clone AL-21; BD Pharmingen), and anti-Ly6G-FITC (clone RB6-8C5; BD Pharmingen). In addition, lung samples were stained for HO-1 mAb, as described above. For the A549 cell line, cells were stained with anti hRSV F-Alexa Fluor 647, conjugated, and washed for further analysis. All samples were acquired on a FACS Canto II flow cytometer (BD Biosciences, San Jose, CA) and analyzed using FlowJo 7.6 software.

To perform histopathology analyses without losing significant tissue architecture, before BALF collection, the major bronchus of the left lung was clamped using 10 cm Kelly hemostatic forceps. After BALF of the right lung, the left lung was fixed with 4% paraformaldehyde, then paraffin embedded using a Leica ASP300S enclosed, automatic tissue processor (Leica Microsystems, Wetzlar, Germany). Then, 4 μm-thick tissue sections were obtained using a Microm HM 325 Rotary Microtome (Thermo Scientific), before being mounted and stained for histopathology analyses using H&E.

Immunoreactive CXCL1/KC were quantified by using a double Ab ELISA kit (DuoSet; R&D Systems, Minneapolis, MN). IL-6, IL-4, IL-10, and IFN-γ were quantified by using Ab ELISA kit (OptEIA; BD Pharmingen), and CCL3/MIP-1α detection was performed following the manufacturer’s protocol (Ready Set Go!; Affymetrix).

Lymph nodes from infected and mock BALB/cJ mice or tTA-HO-1 conditional transgenic mice of each treatment were removed and mechanically homogenized in 1 × PBS. After erythrocyte lysis with ACK buffer (150 mM NH₄CL, 10 mM KHCO₃, 0.15 mM EDTA), cells were resuspended at a final concentration equal to 5 × 10⁶ cells per ml in RPMI 1640 medium, supplemented with 10% FBS, 1 mM nonessential amino acids, 2 mM glutamine, 1 mM pyruvate, 10 μg/ml penicillin G, 100 μg/ml streptomycin, 50 μg/ml gentamicin, and 50 μM 2-ME. Then, single-cell suspensions were left untreated or stimulated with ultraviolet-hRSV or anti-CD3ε/CD28. After 72 h of incubation (37°C, 5% CO₂), culture supernatants were analyzed for IFN-γ production by sandwich ELISA, and the expression of activation marker CD69 was measured on the surface of cells by flow cytometry (FACS Canto II flow cytometer [BD Biosciences, San Jose, CA] and analyzed using FlowJo 7.6 software).

Immature DCs from conditional transgenic mice were incubated for 2 h with 50 μM CoPP or 1.5 μg/ml DOX (for all the incubation period). Then, cells were washed once and incubated with 700 μl of fresh RPMI 1640 medium. Maturation of DCs was induced by LPS treatment (1 μg/ml) for 16 h (Escherichia coli 0111; Invitrogen) and the surface expression of CD40, CD86, and CD80 molecules was measured by flow cytometry. Also, cells were pulsed with 50 μg/ml of OVA protein, non-OVA pulsed cells ± LPS were used as a control for antigenic presentation. After overnight incubation, cells were washed twice, detached, counted, and plated in round-bottom 96-well plates. Then, 50,000 DCs were incubated with 100,000 purified (over 90%; Miltenyi Biotec) OVA-specific TCR OT-II T CD4⁺ cells (relationship 2:1 T cells:DCs). After 48 h of coculture, supernatants were collected for IFN-γ secretion, and T cells were stained for CD69 activation marker.

All statistical analyses were performed using GraphPad Prism Software version 6.1. Statistical significance was assessed using the Student t test and one-way ANOVA test with a posteriori Bonferroni test or Kruskal–Wallis ANOVA test with a posteriori Mann–Whitney U test. Differences were considered significant when p < 0.05.

Upregulation of HO-1 suppresses infection of several viruses, including HIV, HCV, hepatitis B virus, Ebola virus, and influenza virus (16). To evaluate whether HO-1 induction reduces hRSV infection in vitro, we pretreated the human airway epithelial cell line A549 with increasing concentrations of CoPP (an HO-1 inducer), vehicle, or SnPP (an allosteric HO-1 inhibitor) for 2 h (Fig. 1A, 24). Supernatants of cells were then removed and cells were washed with fresh medium and infected with hRSV at an MOI equal to 1 for 2 h. After viral adsorption, cells were washed with medium to remove unbound virus. For postinfection treatment, cells were infected as described above, and CoPP was added to the cells and incubated for 24 or 48 h. Induction of HO-1 by CoPP was assessed by flow cytometry and Western blotting (Fig. 1D, Supplemental Fig. 1), respectively. Supernatants were harvested 48 h postinfection, and infective virus particle production was measured by detection of the hRSV F protein by ELISA. We observed that production of virus particles in supernatants was inversely correlated with increasing CoPP doses postinfection in A549 cells (Fig. 1A, white circles). Conversely, treatment with increasing doses of SnPP was directly associated with hRSV production in the supernatant of the treated A549 cells (Fig. 1A, black squares). Because treatment with 50 μM of CoPP resulted in more than a 60% reduction in hRSV production, as compared with cells treated with vehicle, we selected this concentration for further experiments. Moreover, to address whether HO-1 upregulation affects the generation of the viable progeny of hRSV viral particles, supernatants from infected A549 cells treated with CoPP, vehicle, or SnPP were harvested to quantify virus titers by immune-plaque assays. We found that hRSV-infected A549 cells treated with 50 μM of CoPP postinfection showed reduced virus titers by 3 log units as compared with cells treated with vehicle (Fig. 1B). Furthermore, because N transcription can be considered a measurement of hRSV viral replication within infected cells (25), we evaluated this transcript in cells treated as indicated above. We observed a significant reduction in viral N mRNA amounts in cells treated with CoPP, both added preinfection and postinfection (Fig. 1C). Conversely, treatment with SnPP increased the amounts of hRSV N-transcripts in infected cells. Equivalent to A549, virus particle production and viral replication were similarly influenced by the HO-1 expression in HEp-2 cells (data not shown). In addition, to address the stage of the hRSV replication cycle that was affected by HO-1, we evaluated the effects of HO-1 induction, mediated by CoPP, on hRSV binding and entry. Previous reports have described that the active fusion process of hRSV is inhibited in HEp-2 epithelial cells at 4°C, whereas viral binding still occurs (26, 27). Viral binding was assessed by Western blotting, measuring the hRSV-N protein over total protein collected from these cells (Fig. 1D), as well as by flow cytometry analyses of surface expression of the hRSV F-protein (Fig. 1E, 1F). Interestingly, no significant changes were observed for the expression of hRSV-N (Fig. 1D, middle panel) nor the surface expression of the hRSV-F protein (Fig. 1E, 1F), suggesting that HO-1 induction has no effect on hRSV binding. To evaluate viral entry, cells were infected at 4°C for 1 h, as described above, washed and treated with CoPP, vehicle, or SnPP. Cells were then incubated at 37°C for 5 h. Viral entry was measured by detecting intracellular expression of hRSV-F protein by flow cytometry. When CoPP-treated cells were compared with vehicle-treated or SnPP-treated cells, no significant differences were observed (Fig. 1G, 1H). Virus particle production and viral replication in HEp-2 cells were affected similarly as for A549 cells (data not shown).

Thus, these results suggest that although HO-1 induction negatively modulates hRSV replication in human alveolar epithelial cells, the inhibition of HO-1 activity promotes replication of this virus in these cells. However, HO-1 induction does not affect hRSV binding nor fusion in infected cells.

Because HO-1 can modulate the function of APCs (12, 28) and previous reports have shown that hRSV infects DCs, impairing their capacity to prime T cells (14, 29), we evaluated whether HO-1 was expressed in DCs upon hRSV infection. Bone marrow–derived DCs were exposed to hRSV, ultraviolet-hRSV, or mock. HO-1 expression was determined by qPCR, flow cytometry, and immunofluorescence. As shown in Fig. 2A, a significant increase in HO-1 mRNA expression levels was observed in hRSV-infected DCs, as compared with mock or untreated cells at 48 h postinfection. On the contrary, ultraviolet-hRSV–inoculated DCs did not show an upregulation of HO-1 mRNA levels, suggesting that HO-1 induction depends on hRSV gene transcription. Consistent with this observation, flow cytometry analyses exhibited a significant increase in the expression of HO-1 in hRSV-inoculated DCs (Fig. 2B). Confocal microscopy experiments further confirmed these findings, showing significantly higher HO-1 levels in hRSV-infected DCs, as compared with control cells (Fig. 2C). The HO-1 expression levels seen in hRSV-infected DCs were equivalent to the HO-1 levels induced by CoPP (Fig. 2C), suggesting that hRSV is a potent stimulus for HO-1 expression in DCs. Of note, the subcellular location of HO-1 was mainly situated near the plasma membrane in CoPP-treated DCs, but displayed a homogeneous distribution throughout the cytoplasm and plasma membrane in hRSV-infected DCs, suggesting that the differential distribution of HO-1 could be associated with different functions.

Considering hRSV infection was significantly reduced in vitro as a result of HO-1 induction in airway epithelial cells and that this enzyme can display a potent anti-inflammatory activity (30–32), we evaluated whether the pharmacological induction of HO-1 in vivo can modulate hRSV infection in mice. Thus, BALB/cJ mice were treated either with CoPP or SnPP 24 h before hRSV infection. Then, animals were challenged intranasally with hRSV (1 × 10⁶ PFU) or with equivalent volumes of a mock solution (as a non-infectious control). Changes in body weight were monitored as a parameter of disease progression for 7 d postinfection (4). As shown in Fig. 3A, a noticeable weight loss was observed after hRSV infection in vehicle-treated mice. Remarkably, CoPP-treated mice displayed accelerated kinetics of body weight recovery after hRSV infection, as compared with vehicle-treated mice (Fig. 3A). Consistently, significant differences between CoPP-treated and vehicle-treated mice were observed at days 2, 3, and 4 postinfection with hRSV. In agreement with the body weight recovery data, a significant reduction of hRSV-N RNA levels was observed at day 4 postinfection in the lungs of hRSV-infected mice that were pretreated with CoPP (Fig. 3B). In contrast, mice receiving SnPP as pretreatment showed high viral loads in the lungs (Fig. 3B). Furthermore, viral titers in BALF from mice at 4 d postinfection were analyzed by immuno-plaque assays. CoPP-treated mice showed a 2-log reduction in viral titers, as compared with vehicle-treated mice (Fig. 3C). Further, qPCR analyses confirmed a significant increase in HO-1 mRNA levels in the lungs of CoPP-treated, hRSV-infected, and CoPP-uninfected mice as compared with control mice at day 4 postinfection (Supplemental Fig. 2A). These results suggest that the administration of CoPP was effective at inducing HO-1 expression in the lungs of mice. Furthermore, mRNA expression data were supported by flow cytometry analyses, which showed that CoPP treatment increased the amount of HO-1 protein in epithelial Epcam-positive cells (Supplemental Fig. 2B). Therefore, our data suggest that pharmacological induction of HO-1 expression, before virus challenge, reduces viral replication and virus particle production.

To assess whether the pharmacological induction of HO-1 modulates hRSV-mediated airway inflammation, histopathological analyses were performed in mice for each treatment at day 4 postinfection. Consistent with body weight loss and viral titration data, CoPP-treated mice displayed reduced inflammation after hRSV infection, as compared to vehicle-treated control animals (Fig. 4A, middle panel). Furthermore, CoPP-treated mice exhibited reduced inflammatory infiltration in both bronchoalveolar airspaces and the lung interstitium, suggesting that HO-1 induction protects mice against bronchopneumonia and interstitial pneumonia. Conversely, SnPP treatment caused a slight increase in lung inflammation (Fig. 4A, bottom panel). BALF collected at day 4 postinfection and evaluated by flow cytometry for leukocyte infiltration were consistent with previous data (21), showing that hRSV challenge caused significant infiltration of inflammatory cells into the airways of vehicle-treated mice (Fig. 4B, 4C). Further, we analyzed the frequency of total leukocytes (CD45⁺) and neutrophils (MHC-II negative [MHC-II⁻] CD11c⁻ CD11b⁺ Ly-6G⁺/Ly-6C⁺) in the BALF of hRSV-infected mice (Fig. 4B, 4C). A significant decrease in the total cell counts of total leukocytes and neutrophils in BALF was observed for CoPP-pretreated mice upon hRSV challenge, at day 4 postinfection (Fig. 4B, 4C). Contrarily, SnPP-treated mice showed a higher amount of neutrophils in BALF at day 4 post hRSV infection (Fig. 4C). Additionally, hRSV infection increased the production of several cytokines and chemokines in BALF, which are associated with lung inflammation (23). Thus, to evaluate whether the observed anti-inflammatory effects of HO-1 in the course of hRSV infection were due to the modulation of cytokines or chemokines in BALF, protein levels of IL-6, IL-4, IFN-γ, IL-10, CCL3/MIP-1α, and CXCL1/KC (an IL-8 homolog) were measured by ELISA. Overall, the protein concentrations of all hRSV-inducible cytokines (IL-6, IL-4, and IFN-γ; Fig. 5A–C) and chemokines (CCL3/MIP-1α and CXCL1/KC; Fig. 5D, 5E) were lower in mice that were treated with CoPP, as compared with vehicle-treated mice. Conversely, CoPP-treated mice displayed increased secretion of the immunomodulatory cytokine IL-10, even in mock controls (Fig. 5F), as previously described (33). Interestingly, SnPP-treated mice showed increased levels of CCL3/MIP-1α (Fig. 5D), suggesting that the increase in neutrophil recruitment can be mediated by a higher concentration of this chemokine.

Taken together, these data suggest that the pharmacological induction of HO-1 expression before virus challenge prevents the development of pulmonary inflammation by modulating the airway immunological milieu during hRSV infection, as evidenced by a reduction in lung inflammation and neutrophil infiltration in BALF.

Previous reports indicate that HO-1 can regulate early innate immunity by modulating type I IFN production (14), as evidenced by the inhibition of HCV replication after activation of this enzyme (34). Therefore, we evaluated whether CoPP-mediated HO-1 induction could activate an antiviral type I IFN response in vivo during hRSV infection. Expression of IFN-α/β was measured in lungs from both hRSV-infected and control mice. As shown in Fig. 6, CoPP treatment enhanced mRNA levels of IFN-α (Fig. 6A) and IFN-β (Fig. 6B) in the lungs of hRSV-infected mice at day 4 postinfection. In addition, CoPP treatment also increased IFN-α/β mRNA expression in the lungs of mock (uninfected) control animals. Interestingly, SnPP treatment failed to modulate this type I IFN response. These data suggest that CoPP by itself induces an antiviral state in airway cells and that the upregulation of IFN-α/β requires HO-1 activity.

In addition to the important immunomodulatory properties of HO-1 on the innate immune response, this enzyme also exerts immunomodulatory effects on T cell–mediated adaptive response, by impairing T cell activation, proliferation, and their effector functions (16, 35). To examine whether HO-1 induction affects T cell responses, single-cell suspensions were obtained from mediastinal lymph nodes of each experimental group and were stimulated for 72 h with ultraviolet-hRSV or with anti-CD3ε/CD28 (1 μg/ml), or left untreated. T cell activation in single-cell suspensions was determined by CD69 expression in CD4⁺ and CD8⁺ T cells as a parameter for early activation (36). A significant increase in CD69 expression was observed in both CD4⁺ (Fig. 7A) and CD8⁺ (Fig. 7B) T cells stimulated with a polyclonal stimulus (anti-CD3ε/CD28) for all treatments. However, CoPP treatment led to a reduction of CD69 expression on anti-CD3ε/CD28-stimulated cells (Fig. 7A, 7B). No significant changes were observed in ultraviolet-hRSV stimulation, as evidenced by a non-significant increase in CD69 expression, which had been previously described by our group (36). In agreement with these data, a significant amount of IFN-γ was secreted in response to anti-CD3ε/CD28 stimulation with a mild decrease for the CoPP treatment (Fig. 7C).

To confirm the HO-1 protective role against hRSV infection through CoPP administration, conditional transgenic mice overexpressing the HMOX-1 gene in MHC-II⁺ cells (tTA-tHO-1) were instilled intranasally either with hRSV (1 × 10⁶ PFU) or mock. Then 48 h before infection, 800 μg/ml DOX and 36 mg/ml sucrose were added to drinking water to induce HO-1 expression in these transgenic mice. To address whether the treatment with DOX induces the expression of exogenous human HO-1 in MHC-II⁺ cells in the conditional transgenic mice tTA-tHO-1 (Supplemental Fig. 3A), qPCR analyses of the human HO-1 gene were performed from lung tissue obtained from mice of each experimental group (Supplemental Fig. 3B). The detection of human HO-1 in sorted pulmonary MHC-II⁺ cells further validated this model (Supplemental Fig. 3C). Disease parameters were evaluated as described above, in this transgenic mouse model expressing human HO-1. Consistent with the effects observed in CoPP treatment, the selective expression of human HO-1 induced by DOX resulted in an accelerated kinetic of body weight recovery at days 3 and 4 postinfection, as compared with littermate controls and untreated transgenic mice (Fig. 8A). No significant unspecific effects were observed for DOX treatment in uninfected mice (Fig. 8A). These results are consistent with the observations described above, as at day 4 postinfection. We found a decreased neutrophil infiltration in the airways of tTA-HO-1 mice treated with DOX, as compared with littermates treated with DOX and infected with hRSV (Fig. 8C). Furthermore, DOX-induced transgenic mice also showed reduced viral loads in the lungs, as compared with infected littermates at day 4 postinfection with hRSV (Fig. 8B). Finally, lung histological analyses were performed to evaluate whether the transgenic expression of HO-1 could prevent lung inflammation. As shown in Fig. 8D, transgenic mice treated with DOX displayed significantly less inflammation than hRSV-infected littermate controls and untreated transgenic mice (Fig. 8D). Therefore, these data suggest that exogenous expression of human HO-1 in MHC-II⁺ cells decreases neutrophil infiltration and lung inflammation in the conditional tTA-HO-1 transgenic mice.

Next, we assessed whether exogenous expression of HO-1 in MHC-II⁺ cells in conditional transgenic mice could affect the function of APCs and their ability to process and present Ags to T cells. First, we evaluated the maturing capacity of DCs from tTA-HO-1. tTA-HO-1 DCs were treated with vehicle, CoPP, or DOX for 2 h. Cells were washed, and LPS (1 μg/ml) was added to the medium to induce maturation (22). After 16 h, DC maturation was measured by surface expression of CD80, CD40, and CD86 molecules by flow cytometry. Overall, all assessed surface markers were upregulated in LPS-pulsed DCs, suggesting that HO-1 had no effect in the maturation process of DCs in conditional transgenic mice after HO-1 induction by DOX (Supplemental Fig. 4A). To evaluate whether the ability of DCs to process and present Ags to T cells could be altered by the transgenic expression of HO-1, OVA-pulsed mature DCs from tTA-HO-1 mice were cocultured with OT-II CD4⁺ T cells, which recognize the OVA-derived peptide OVA323–339/I-Ab complex as a cognate ligand. Thus, LPS-DCs treated with CoPP showed an impairment in T cell activation after 48 h of coculture, suggesting that CoPP pretreated DCs were unable to prime OT-II T cells (Supplemental Fig. 4B). On the contrary, mature DCs from the conditional transgenic mice that were treated with DOX were capable of processing and presenting OVA-derived peptides, resulting in OT-II T cell priming. However, the activation of OT-II T cells in the coculture with DCs from conditional transgenic mice treated with DOX was lower compared with cocultures between vehicle-treated DCs and OT-II cells (Supplemental Fig. 4B), suggesting that the expression of HO-1 in DCs from conditional transgenic mice slightly affects the T cell priming. Furthermore, to assess whether expression of HO-1 in MHC-II⁺ cells affects T cells priming in vivo during hRSV infection, littermate, untreated or DOX-treated tTA-tHO-1 mice were infected with hRSV and 4 d postinfection, single-cell suspensions were obtained from mediastinal lymph nodes and stimulated for 72 h with ultraviolet-hRSV or with anti-CD3ε/CD28 (1 μg/ml) or left untreated. To evaluate T cell activation in the cells suspensions, CD69 expression was measured in CD4⁺ T cells. No significant increases in CD4 or CD69 expression were observed for cell suspensions from anti-CD3ε/CD28-stimulated DOX-induced transgenic mice as compared with unstimulated DOX-induced transgenic mice, suggesting that T cell activation is affected by HO-1 overexpression in MHC-II⁺ cells (Fig. 9A). In agreement with these data, no significant IFN-γ secretion could be measured in supernatants of cell suspensions derived from DOX-induced transgenic mice (Fig. 9B), suggesting that T cell function can also be modulated by HO-1 overexpression in MHC-II⁺ cells.

HO-1 is a stress-inducible enzyme that catalyzes heme degradation into carbon monoxide, biliverdin, and free iron (37). HO-1 induction has been involved in several pathophysiological conditions, and HO-1 is generally regarded as a cytoprotective enzyme (38–40). Recently, there has been a growing interest about the regulatory role that HO-1 may play in immunity and tolerance (41, 42). Several studies support the beneficial effects that HO-1 and its products can have during inflammatory processes (42–44). However, little is known about the contribution of HO-1 expression during viral infections in vivo. Previous evidence described that gene overexpression of HO-1 can attenuate disease severity of influenza infection in mice (45). Moreover, induction of HO-1 inhibited influenza replication by activating the type I IFN system with the subsequent induction of IFN-stimulated genes (46). Further, HO-1–deficient (ho-1^−/−) mice displayed decreased survival rates after influenza infection, as compared with wild-type mice (47).

In this study, we evaluated whether the induction (or inhibition) of HO-1 could affect the development of the disease caused by hRSV in mice. Our findings indicate that preadministration of CoPP can promote a significantly faster body weight recovery after hRSV infection, as compared with control mice. Similar results were observed in transgenic tTA-HO-1 mice that overexpress human HO-1 in MHC-II⁺ cells. However, inhibition of HO-1 activity mediated by SnPP preadministration did not increase body weight loss as a parameter of disease, suggesting that HO-1 enzymatic activity is not necessarily the unique element involved in the protective effect observed in the body weight recovery induced by CoPP. In addition, we found that CoPP preadministration reduced viral mRNA and the titers of infective viral particles in the lungs of hRSV-infected mice, an outcome that coincided with an increase in HO-1 expression in lungs of mice at day 4 postinfection (Supplemental Fig. 2). Interestingly, we observed that SnPP pretreatment caused a slight increase in viral loads in the lungs during the same period of time, however, viral progeny (PFU) was not raised. These data suggest that the HO-1 enzymatic activity or its products are not necessarily involved in the modulation of viral replication and clearance. The increased expression of HO-1, observed in Epcam-positive cells from the lungs of CoPP-pretreated mice, is consistent with the notion that a reduction in viral loads could be related to the promotion of an antiviral response, mediated by HO-1 in airway epithelial cells. This concept is supported by the observation that although CoPP reduced viral loads and virus particle production in hRSV-infected A549 cells, SnPP-mediated inhibition of HO-1 activity increased both parameters. However, the initial steps of the life cycle of the virus, such as hRSV binding and entry, were not affected by HO-1 induction. Furthermore, we observed that CoPP treatment induced an upregulation of IFN-α/β in the lungs of hRSV-infected mice, indicating that HO-1 plays an important role during the development of the antiviral type I IFN response in the airways. All these data suggest that HO-1 induction could be promoting an antiviral state that prevents or limits viral replication and propagation of new virus particles.

In contrast, during acute lung infections, the host inflammatory response requires tight regulation (48) to promote pathogen clearance without evoking an exaggerated inflammatory response that could damage the infected airways. Our results indicate that HO-1 upregulation results in the suppression of lung inflammation, associated with a decrease in the inflammatory cell infiltration and an inhibition of proinflammatory cytokine or chemokine secretion during hRSV infection. Several studies have reported that hRSV infection actively induces activation of NF-κB in the lung tissue, leading to the secretion of proinflammatory mediators (49, 50). Because HO-1 has been found to inhibit NF-κB activity (51), we hypothesized that HO-1 induction decreases lung inflammation through the inhibition of NF-κB, which results in lower levels of proinflammatory cytokines and chemokines. Furthermore, HO-1 induction increases IL-10 secretion in BALF, supporting an anti-inflammatory role for HO-1 during hRSV infection.

Consistent with the observations described in this study for hRSV, decreased neutrophil infiltration and lung injury were described during influenza virus H1N1 infection as a result of adenoviral-mediated HO-1 gene transfer (45). These findings support the notion that HO-1 activity may be important for modulating lung inflammation during viral pulmonary pathologies (47, 52). This idea was further supported by the observation that mice deficient for Nrf2 (nrf2 ^−/−), a transcription factor controlling HO-1 gene expression, displayed a significantly reduced hRSV clearance, a higher bronchopulmonary inflammation, and a reduced body weight gain as compared with control mice (53).

The results observed for pharmacological induction of HO-1 mediated by CoPP are similar. The conditional expression of HO-1 in MHC-II⁺ subsets in vivo promoted a reduction in lung disease in hRSV-infected tTA-HO-1 transgenic mice. These data suggest that the HO-1 enzyme is responsible for the protective effect of CoPP treatment. Importantly, although the MHC-II molecule is constitutively expressed by APCs, its expression is not limited to immune cells. In fact, intestinal and pulmonary epithelial cells can also express this molecule (54). Then, the conditional transgenic model used in this work suggests that the upregulation of HO-1 in epithelial cells and immune cells expressing MHC-II can contribute to suppressing the inflammation triggered by the hRSV infection.

It is important to mention that HO-1 was highly induced in DCs exposed to hRSV and it has been described that infection of DCs by hRSV impairs the ability of prime T cells (55), suggesting that HO-1 induction in response to hRSV could contribute to this process. Furthermore, HO-1 upregulation mediated by CoPP treatment or the exogenous expression of human HO-1 in MHC-II⁺ cells in the conditional transgenic model slightly affects T cell activation and function, indicating that Ag processing and presentation by DCs was impaired by HO-1, in agreement with previous data supporting a role of HO-1 during adaptive immunity (35).

Treatment based in HO-1 upregulation is currently used as a therapeutic approach. The systemic hemin therapy, an HO-1 inducer, has been approved by the Food and Drug Administration to treat acute intermittent porphyria (56). Hemin also has been used to treat thalassemia intermedia, myelodysplastic syndrome, and liver allograft failure in erythropoietic protoporphyria (56). However, this strategy has not been explored for the treatment of infectious diseases.

In summary, the current study demonstrates that hRSV infection can be modulated by the expression of HO-1 both in vitro and in vivo. Thus, HO-1 activity may play a critical role during hRSV infection. HO-1 induction could protect the host from the pulmonary pathology developed upon hRSV infection, by reducing viral replication and lung inflammation, thus favoring disease resolution. Therefore, our results shed light on the potential role of the therapeutic induction of HO-1 in this viral pneumonia and suggest new avenues for the immunomodulatory treatment of hRSV-infected patients.

We thank Dr. Pedro Piedra (Baylor College of Medicine), Dr. George Kollias, and Dr. Christophe Benoist for kindly providing the A549 cells, pIi-TTA-TetO-HO-1, and pIi-TTA transgenic mice, respectively. We also thank María José Altamirano for breeding the mouse colonies that were used in this work.

The authors have no financial conflicts of interest.