Enhances Rhinovirus-Induced RANTES Secretion by ... - CiteSeerX

Interferon- Enhances Rhinovirus-Induced RANTES Secretion by Airway Epithelial Cells Shinichi Konno, Kristine A. Grindle, Wai-Ming Lee, Mary K. Schroth, Anne G. Mosser, Rebecca A. Brockman-Schneider, William W. Busse, and James E. Gern Departments of Medicine, Pediatrics, and Molecular Virology, University of Wisconsin-Madison, Madison, Wisconsin

Respiratory viruses, including rhinoviruses, infect respiratory epithelium and induce a variety of cytokines and chemokines that can initiate an inflammatory response. Cytokines, such as interferon (IFN)- and tumor necrosis factor (TNF)-, could enhance epithelial cell activation by inducing virus receptors. To test this hypothesis, effects of IFN- or TNF- on expression of intercellular adhesion molecule (ICAM)-1, rhinovirus binding, and virus-induced chemokine secretion on A549 and human bronchial epithelial cells (HBEC) were determined. The results varied with the type of cell. IFN- was a stronger inducer of ICAM-1 and viral binding on HBEC, whereas TNF- had greater effects on A549 cells. In addition, IFN-, but not TNF-, synergistically enhanced regulated on activation, normal T cells expressed and secreted (RANTES) mRNA expression and protein secretion induced by RV16 or RV49. To determine whether IFN- could enhance RANTES secretion independent of effects on ICAM-1 and RV binding, HBEC were transfected with RV16 RNA in the presence or absence of IFN-. RV16 RNA alone stimulated RANTES secretion, and this effect was enhanced by IFN-. These results demonstrate that IFN- can enhance rhinovirus-induced RANTES secretion by increasing viral binding, and through a second receptor-independent pathway. These findings suggest that IFN-, by upregulating RANTES secretion, could be an important regulator of the initial immune response to rhinovirus infections.

Airway epithelial cells are the principal host cells for respiratory virus infection and play an important role in initiating the immune response to virus. Virus-infected epithelial cells secrete cytokines (e.g., GM-CSF and interleukin [IL]-11) and chemokines (e.g., regulated on activation, normal T cells expressed and secreted [RANTES] and IL-8) that can then recruit and activate inflammatory cells (1–4). Chemokine secretion by airway epithelial cells may be a key event both in innate and acquired immune responses to viral infections. For example, RANTES is a potent chemoattractant for monocytes, NK cells, CD4/CD45RO T cells, eosinophils, and basophils(5–7), whereas IL-8 is one of the principal signals to recruit neutrophils into the airway (8). Through these effects, chemokine secretion has also been linked to the pathogenesis of virus-induced airway inflammation and asthma (9–11).

(Received in original form November 13, 2000 and in revised form January 25, 2002) Address correspondence to: James E. Gern, M.D., K4/918 University of Wisconsin Hospital, Madison, WI 53792-9988. E-mail [email protected] Abbreviations: bronchial epithelial cell growth medium, BEGM; enzymelinked immunosorbent assay, ELISA; fetal calf serum, FCS; intercellular adhesion molecule, ICAM; interferon-, IFN-; human bronchial epithelial cells, HBEC; low-density lipoprotein receptor, LDLR; phosphatebuffered saline, PBS; RNA-dependent protein kinase, PKR; regulated on activation, normal T cells expressed and secreted, RANTES; rhinovirus, RV; tumor necrosis factor-, TNF-. Am. J. Respir. Cell Mol. Biol. Vol. 26, pp. 594–601, 2002 Internet address: www.atsjournals.org

The interaction of respiratory viruses and epithelial cells can be modulated by cytokines that are secreted by mononuclear cells. For example, tumor necrosis factor (TNF)-, which is secreted by macrophages in response to rhinovirus (RV) (12), can induce intercellular adhesion molecule (ICAM)-1 expression, and perhaps through this action can decrease the infectious inocula needed to establish a productive RV infection in vitro (1). Furthermore, interferon (IFN)-, which is secreted by lymphocytes in response to RV, is also a potent inducer of ICAM-1 on a variety of cell types (13, 14). Therefore, cytokine induction of ICAM-1 may have dual effects during RV infections: increased ICAM-1 expression could enhance recruitment of inflammatory cells into the airway, and thereby help to initiate antiviral immune responses in the airway. However, because ICAM-1 is the receptor used by major group RV(15, 16), virus binding to host cells could also be increased, thus potentiating damage to airway epithelial cells. Although the effects of IFN- and TNF- on epithelial cells are complex, their effects on ICAM-1, and independent effects on chemokine production, suggest that these cytokines might be important regulators of RV-induced epithelial cell activation. To evaluate this hypothesis, the effects of IFN- and TNF- on epithelial cell ICAM-1, RV binding, and RV-induced epithelial cell secretion of the chemokines RANTES and IL-8 were determined with A549 cells, a type 2 alveolar epithelial cell line, and with nontransformed cultures of human bronchial epithelial cells (HBEC).

Materials and Methods Cell Culture HBEC for primary culture were isolated from surgical specimens of normal human bronchi or trachea (lung transplant donors) as previously described (2). All primary cell cultures were grown in serum-free, hormonally-supplemented medium (Bronchial Epithelial Cell Growth Medium [BEGM]; Clonetics, San Diego, CA), and were passaged up to four times. Cell viability, as assessed by exclusion of trypan blue dye, was consistently greater than 90%. A549 cells obtained from American Type Culture Collection (CCL 185; Rockville, MD) were grown in Kaign’s nutrient mixture F-12 medium (Irvine Scientific, Santa Ana, CA) supplemented with 10% fetal calf serum (FCS; Hyclone, Logan, UT), 2 mM L-glutamine, 100 U/ml penicillin, 100 g/ml streptomicin, and 20 g/ml gentamycin at 37C and 5% CO2. The Ohio strain HeLa cells were maintained in shake cultures in medium B (17). Recombinant human IFN- and TNF- were purchased from Promega, Inc. (Madison, WI).

Virus Suspensions RV16, RV49, and type III parainfluenza virus (PIV III) were grown in the Ohio strain of HeLa cells. Culture medium with HeLa cells was inoculated with virus and grown in rolling bottles at 33 C.

Konno, Grindle, Lee, et al.: IFN- Enhances Rhinovirus-Induced RANTES Secretion

Growth was allowed to continue until cytopathic effects were well advanced. The resulting fluids were centrifuged (2,000 rpm, 30 min, 20C) to remove cellular debris. The virus suspensions were then purified by centrifugation over a sucrose gradient (17). Aliquots of virus with infectivity of 107–1010 tissue culture infectious dose50 (TCID50)/ml, as measured by inoculating serial dilutions of virus suspension in HeLa (RV16) or WI-38 (RV49) cells, were stored at 70C until needed. Radiolabeled RV16 and RV49 were prepared by infecting HeLa cells (4.0 107 cell/ml) with 10–15 plaque-forming units (PFU) /cell in phosphate-buffered saline (PBS). One hour later, infected cells were diluted 10- to 15-fold in growth medium lacking all amino acids except glutamine and supplemented with 25 mM HEPES and 5 g/ml actinomycin D, and incubated at room temperature. Four hours after inoculation, 35S-methionine (1 mCi/1.8 108 cells) was added and the cells were incubated for an additional 3 h. The cells were then pelleted and resuspended in growth medium. The cells were lysed in Nonidet P-40 (0.5%, Sigma-Aldrich, St. Louis, MO) at room temperature. Cell debris was removed by centrifugation (12 min, 15,000 rpm) in a Sorvall SS34 rotor (Kendro Laboratory Products, Newtown, CT), and the supernatant was treated with RNase A and EDTA at room temperature, layered over a series of 10, 20 and 30% sucrose layers, and then centrifuged (15C, 90 min, 45,000 rpm, 55T rotor, Beckman Coulter, Fullerton, CA). After decanting, the pellet was resuspended, aliquoted and stored at 80C.

Generation of Infectious RV16 RNA To produce a large quantity of infectious viral RNA, a full-length RV16 cDNA plasmid, pR16.11, was used as template for in vitro transcription. The pR16.11 was constructed by cloning the RV16 duplex cDNA fragments used for sequencing (18) into a plasmid vector. To generate infectious RV16 RNA, the plasmid was linearized by incubation (37C, 3 h) of 10 g of pR16.11 with 20 U of SacI (New England Biolabs, Beverly, MA) in a total volume of 40 l. RV16 RNA was synthesized by incubating (37 C, 1 h) 8 l of plasmid digest with 8 l 5 buffer (Life Technologies, Rockville, MD), 32 U RNase inhibitor (Promega), 0.56 mM rNTPs (Promega), and 50 U T7 RNA polymerase in a total volume of 40 l. The transcripts were extracted from the mixture using a phenol/chloroform reagent (RNA Stat-60; Leedo Medical Laboratories, Houston, TX) and precipitated twice in 2 vol 100% ethanol and 1/10 vol 3 M sodium acetate (pH 5.2). The RV16 RNA transcripts were quantitated by comparison to dilution of an RNA (Life Technologies) standard on a 1% agarose gel stained in ethidium bromide.

Inoculation of HBEC with RV HBEC were grown in 24-well plates with BEGM to near confluence. Twenty-four hours before viral inoculation, the medium was changed to BEGM without hydrocortisone, epidermal growth factor, epinephrine, or bovine pituitary extract, to remove factors that could potentially affect cytokine production. The monolayers were then inoculated with either RV16, RV49, or PIV III, with or without indicated reagents, and incubated for 24 h (mRNA) or 48 h (protein) at 37C and 5% CO2. The culture supernatants and the infected cells were then collected for enzyme-linked immunosorbent assay (ELISA) and RT-PCR, respectively.

Transfection of Epithelial Cells with RV16 RNA For transfections, A549 cells were plated at 1–2 105 cells/well in six-well tissue culture plates and allowed to grow to 80% confluence ( 48 h). The cells were then incubated (37 C, 5 h) with 8 l LipofectAMINE reagent (Life Technologies) and either RV16 RNA (0.005–0.5 g), or RNA isolated from the yeast Blastomycetes (Bruce Klein, University of Wisconsin-Madison). Five hours following the start of transfection, the medium containing the trans-

595

fection complexes was removed and replaced with 2 ml fresh medium as described above. Twenty-four hours later, the cells were harvested for RNA and/or luciferase activity, and supernatants were collected and stored at 80C pending analysis for RANTES protein. Transfection efficiencies were monitored by cotransfecting with a plasmid (pGL2) containing the luciferase gene (Promega), and then analyzing luciferase activity in cell lysates with luminometer (Analytical Luminescence, Ann Arbor, MI). Preliminary experiments showed no significant changes in luciferase expression when pGL2 was cotransfected with transfection reagent containing 0–2 g/ml RV RNA.

Whole Cell ELISA Assay for ICAM-1 Expression Epithelial cell ICAM-1 expression was measured with a semiquantitative whole-cell ELISA, as previously described (2). Briefly, HBEC monolayers were cultured (48 h, 96-well plates) with or without IFN- (0.1, 1, or 10 U/ml) or TNF- (10 or 100 U/ml), washed, and then incubated (2 h) in blocking solution with anti– ICAM-1 monoclonal Ab (RR1; a gift from Robert Rothlein, Boehringer Ingelheim, Ridgefield, CT) or isotype control antibody (mouse IgG1; Sigma Chemical Co.). After washing, the ELISA was developed using horseradish peroxidase–conjugated sheep anti-mouse IgG1 Ab (Sigma Chemical Co.) and 3,3 ,5,5tetramethylbenzidine. The data are reported as mean values absorbance (@ 450 nm) from three wells. The coefficient of variation was generally less than 10%.

RV Binding Assay Monolayers ( 80% confluent) of A549 cells or HBEC in 12-well plates were incubated with either IFN- (10 U/ml), TNF- (100 U/ml), or medium alone for 24 h. Preliminary experiments had established that these concentrations of cytokines produced maximum ICAM-1 expression without causing cytopathic effects. After incubation, cells were washed once with PBS, overlayed with 0.5 ml PBS containing 0.1% human serum albumin, and then anti–ICAM-1 mAb (final concentration 20 g/ml), anti-low density lipoprotein receptor (LDLR) mAb (final concentration 100 g/ml) or mouse isotype control (IgG2b; PharMingen, San Diego, CA) were added to the wells. The plates were rotated at room temperature for 1 h, and either 35S-methionine-labeled RV16 (1,000 particles/cell) or 35S-methionine-labeled RV49 (100 particles/cell) was added to the wells and incubated at room temperature for a further 2 h. Lower concentrations of RV49 were used because of the lower titers of this virus preparation. The supernatant was sampled for scintillation counting, and unbound virus was removed. The cells were washed twice with 1 ml/well of PBS lacking divalent cations and the cells removed by trypsinization. The trypsinized cells were mixed well and sampled for scintillation counting. Scintillation samples were counted in Bio-Safe scintillation fluid (Research Products International Corp, Mount Prospect, IL).

RV16 Replication Assay When monolayers of HBEC in 24-well plates with BEGM were near confluence, medium (BEGM without hydrocortisone, epidermal growth factor, epinephrine, and bovine pituitary extract) was applied along with either IFN- (500 U/ml), IFN- (10 U/ml), or TNF- (100 U/ml). After incubation (24 h, 37 C, 5% CO2), the cells were inoculated with RV16 (10 7 TCID50/ml), and incubated at room temperature for 90 min on rotating plate. The cells were washed to remove unbound virus, and incubated (35 C, 5% CO2) in fresh medium and cytokine, and this was verified by performing viral titers 2 h after inoculation and after washing. Twentyfour hours after inoculation, the cell suspensions were subjected to three freeze/thaw cycles to release intracellular virus to mea-

596

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 26 2002

sure viral yield. The suspensions were collected, centrifuged to remove cellular debris, and the supernates were collected and stored at 70C pending quantitative viral cultures, as previously described (12). All samples were performed in duplicate, and viral titers were expressed as log (TCID 50 units/ml).

Chemokine ELISA RANTES and IL-8 protein in culture supernatants were measured using sandwich ELISA (2) with the following materials and reagents: capture antibodies, murine anti-human RANTES monoclonal Ab (MAB278; R&D Systems, Minneapolis, MN) or murine anti-human IL-8 monoclonal Ab (MAB208; R&D Systems); detection antibodies, biotinylated goat anti-human RANTES polyclonal Ab (BAF278; R&D Systems) or biotinylated goat anti-human IL-8 polyclonal Ab (BAF205; R&D Systems); recombinant human RANTES (Biosource, Camarillo, CA), recombinant human IL-8 (208-IL; R&D Systems), horseradish-streptavidin peroxidase conjugate (Zymed Laboratories, South San Francisco, CA), and 3,3,5,5-tetramethylbenzidine (Kirkegaard and Perry, Gaithersburg, MD). The samples were assayed in duplicate, and the results expressed in pg/ml. The sensitivity of both the RANTES and IL-8 assays was 3–6 pg/ml and the coefficient of variation was generally less than 3%.

Detection of RANTES mRNA with Semiquantitative RT-PCR RANTES mRNA in bronchial epithelial cells was analyzed with semiquantitative RT-PCR as previously described (2). The number of cycles (33 cycles for RANTES, 24 cycles for G3PDH) of PCR was chosen to maintain a linear relationship between mRNA and the PCR products so that quantitative comparisons of PCR products could be obtained. A reference standard curve consisting of serial 4-fold dilutions of RANTES or G3PDH cDNA template was included in each PCR run. Each PCR run also included negative control sample containing all PCR reagents but no cDNA. G3PDH cDNA was amplified to provide a control for the amount of total RNA in each sample. Southern blotting for RANTES and G3PDH was performed, and the blot image was scanned (Scanjet II P; Hewlett Packard, Greeley, CO) to convert it into a digital image, and mRNA was quantitated by comparing the standard curves of the blot densities of the hybridized PCR products from serial diluted cDNA templates with the aid of computer software (Sigmagel; Jandel Scientific, San Rafael, CA).

Data Analysis The effects of cytokines on epithelial receptor expression or RV binding was evaluated by ANOVA, and multiple pairwise comparisons were performed with the Bonferroni t test. Treatment-

related differences in RANTES protein were evaluated with ANOVA, and Fisher test was used for pairwise comparisons between variables. RANTES mRNA expression data were not normally distributed and were analyzed by ANOVA on ranks. Values for RANTES secretion after transfection in Table 1 were logtransformed to approximate a normal distribution, and were analyzed by three-way ANOVA (factors evaluated were cytokine pretreatment, experiment no., and the dose of poly IC). All calculations were performed with the aid of either Stat View II (Abacus Concepts, Inc., Berkeley, CA) or SigmaStat (SPSS Inc., Chicago, IL) computer software. A P value less than 0.05 was considered significant.

Results Effects of IFN- versus TNF- on ICAM-1 Expression To compare the effects of IFN- and TNF- on ICAM-1 expression by HBEC or A549 cells, cell monolayers were incubated with either IFN- (0.1–10 U/ml), TNF- (10– 1,000 U/ml), or medium alone for 48 h, and ICAM-1 expression was measured by whole cell ELISA. On HBEC, ICAM-1 expression was induced by IFN- in a dose-dependent manner, whereas TNF- caused relatively small, but statistically significant increases in ICAM-1 expression (Figure 1A). In contrast to experiments conducted with HBEC, TNF- was a more potent than IFN- in stimulating ICAM-1 expression on A549 cells (Figure 1B). Preliminary experiments had shown that higher doses of either cytokine produced cytopathic effects on HBEC. Effects of IFN- on RV Binding to A549 Cells To confirm that IFN-–induced increases in ICAM-1 would enhance RV binding, the effects of IFN- on binding of major and minor group RV to airway epithelial cells were measured. Either HBEC or A549 cells were incubated either with IFN- (10 U/ml), TNF- (100 U/ml), or medium alone, and then 35S-methionine-RV for 2 h with or without anti-ICAM monoclonal Ab, anti-LDLR monoclonal Ab (C7) or isotype control Ab. Relatively small amounts of the major-group virus RV16 bound to untreated epithelial cells, and blocking antibody had no detectable effect on viral binding (Figures 2A and 2B). IFN- enhanced binding of RV16 to either HBEC (Figure 2A) or A549 cells (Figure 2B), and this binding was blocked by mAb specific for ICAM-1 (Figures 2A and 2B). In contrast to effects of IFN-, TNF- greatly enhanced RV16 binding to A549, but

Figure 1. Effects of IFN- versus TNF- on epithelial cell ICAM-1 expression. Confluent monolayers of (A) HBEC or (B) A549 cells were incubated (48 h) with either IFN- (0.1–100 U/ml) or TNF- (10–1,000 U/ml), and ICAM-1 expression was determined by whole cell ELISA. Results are expressed as mean SE; (A) HBEC, n 3– 5, (B) A549 cells, n 4. *P 0.05.


597

Figure 2. Effects of IFN- versus TNF- on RV binding to airway epithelial cells. Confluent monolayers of either HBEC (A, C) or A549 cells (B) were preincubated (24 h) with either IFN- (10 U/ml), TNF- (100 U/ml), or medium alone. The cells were then incubated with either a specific antibody (ICAM-1 or LDLR; solid bars) to block viral binding, or an isotype control Ab (IgG2b; open bars), washed, and binding of 35S-methionine-labeled RV16 (A, B) or RV49 (C) was determined. Percent viral binding is expressed as mean SE (n 3). *P 0.05, **P 0.01, and †P 0.001 versus samples incubated in the presence of blocking antibody (ICAM-1 or LDLR).

had a relatively small effect on binding to HBEC (Figures 2A and 2B). There was a distinct pattern of binding of the minor group virus RV49 to HBEC (Figure 2C). Binding of RV49 to HBEC cells was relatively high in cells incubated in medium alone, and was significantly blocked by C7 mAb (Figure 2C). Specific RV49 binding to HBEC in the presence of either IFN- or TNF- remained high, but was unchanged compared with cells incubated in medium alone. Effects of Cytokines on RV16 Replication To determine whether cytokines could affect RV16 replication, replication of RV16 in HBEC or A549 cells was determined in the presence or absence of IFN- (10 U/ml), TNF- (100 U/ml), or IFN- (500 U/ml). Cytokine pretreatment had no significant effect on the amount of virus associated with HBEC (3.34 0.54 TCID50/ml) or A549 cells (2.13 0.89 TCID50/ml) after inoculation with RV16 followed by washing (mean SD, P NS, ANOVA). In contrast, evaluation of quantitative viral cultures 24 h after inoculation demonstrated that IFN- significantly suppressed the viral yield in HBEC (Figure 3A), and similar trends were noted in A549 cells (Figure 3B). Neither IFN- nor TNF- significantly affected the viral yield in either type of epithelial cell. Effects of Cytokines on RV16-Induced RANTES Secretion The effects of TNF- (100 U/ml) and IFN- (10 U/ml) on RANTES secretion were determined in HBEC or A549 cells in the presence or absence of RV16 (107 TCID50/ml). IFN- alone was a relatively weak inducer of RANTES in either cell type, but enhanced RV16-induced RANTES secretion (Figures 4A and 4B).

TNF- alone had different effects on the two different cell types: by itself, it had little effect on HBEC RANTES secretion (Figure 4A), but was a potent inducer of RANTES in A549 cells (Figure 4B). In contrast to IFN-, however, TNF- did not enhance RV16-induced RANTES secretion in either cell type. These experiments were broadened to evaluate potential synergy between IFN- and RV49, a minor-group rhinovirus that binds to the LDLR instead of ICAM-1, in stimulating secretion of RANTES and IL-8. HBEC were incubated (48 h) with RV16 (107 TCID50/ml) or RV49 (106 TCID50/ml) in the presence or absence of IFN- (10 U/ml). Inoculation with either RV16 or RV49 alone (48 h) induced low levels of RANTES secretion from HBEC, and this was significantly enhanced by coincubation with IFN- (Figure 5A). A different pattern of IL-8 secretion was observed, however, in that unstimulated cells secreted relatively high amounts of IL-8, and this tended to be enhanced by RV16 or RV49. In contrast to effects on RANTES, however, IFN- did not enhance virus-induced IL-8 (Figure 5B). The synergistic effects between RV16 and IFN- on RANTES production were also demonstrated at the level of RANTES mRNA expression. Inoculation with RV16 or RV49 (24 h) induced low level expression of RANTES mRNA after 24 h (Figure 5C). Coincubation with IFN- enhanced RV16-induced expression of RANTES, and similar trends were observed in experiments conducted with RV49 (Figures 5C and 5D). To test whether IFN- could boost RANTES production stimulated by viruses other than RV, effects of PIV III on RANTES secretion were determined in the presence or absence of either IFN- (10 U/ml) or TNF- (100

Figure 3. Effects of cytokines on RV16 replication. Confluent HBEC monolayers of (A) HBEC (n 5) or (B) A549 cells (n 3) were preincubated with either IFN-, IFN-, or TNF- for 24 h, and were then inoculated with RV16 (1 107 TCID50/ml, 90 min), and extracellular virus was washed away. Viral titers of cell lysates were determined 24 h after inoculation. *P 0.05 versus cells preincubated in medium alone.

598


Figure 4. IFN- synergistically enhances RV16-induced RANTES secretion. Confluent monolayers of (A) HBEC or (B) A549 cells were incubated (48 h, 37C) with RV16 (1 107 TCID50/ml) or medium alone (open bars) in the presence of TNF- (100 U/ml; striped bars) or IFN- (10 U/ml; solid bars). Supernatant fluids were analyzed for RANTES secretion. Results are expressed as mean SE, n 11 for HBEC and n 4 for A549 cells. *P 0.05 compared with cells incubated in medium alone, †P 0.05 compared with cells incubated with IFN- alone.

U/ml). Although PIV III (105 or 106 TCID50/ml) alone was a potent stimulus of RANTES secretion, neither IFN- nor TNF- significantly affected PIV-induced RANTES production by HBEC (Figure 6). Effects of IFN- on RANTES Secretion by A549 Cells Transfected with RV16 RNA To determine whether IFN- effects on RV-induced RANTES production were independent of changes in the number of surface receptors, infectious RV16 RNA was transfected into A549 cells in the presence or absence of IFN-. Transfected RV RNA (0.5 g/well) alone stimulated RANTES secretion in A549 cells in a time-dependent manner, but relatively little RANTES was secreted by cells that were transfected with yeast RNA (Figure 7A). Incubation of transfected cells with IFN- (10 U/ml) caused a small additional increase in RANTES secretion (Figure 7A). To determine whether effects of IFN- depends on the quantity of intracellular RV RNA, A549 cells were transfected with 0.005–0.5 g/well of RV16 RNA in the presence or absence of IFN- (10 U/ml). IFN- enhanced RANTES secretion after transfection of small amounts (0.005–0.05 g/well) of RV16 RNA, but had less effect when cells were transfected with 0.5 g/well of RV16 RNA (Figure 7B). Transfection of A549 cells with RV16 RNA caused new synthesis of infectious RV16 particles (100.5–102 TCID50/ml). To exclude the potentially confounding effects of newlysynthesized viral particles binding to surface ICAM-1, these experiments were repeated in the presence or absence of monoclonal Ab (RR1) to block RV attachment to ICAM-1. Although the anti–ICAM-1 Ab tended to reduce the amount of secreted RANTES, IFN- (10 U/ml) and RV16 RNA (0.005 g/well) acted synergistically to induce RANTES secretion even in the presence of anti–ICAM-1 mAb (Figure 7C). Synergistic Effects of Double-Stranded RNA (dsRNA) and IFN- on RANTES Secretion Double-stranded RNA is formed during RV replication (19), and can activate epithelial cells and fibroblasts to secrete cytokines such as IFN- and IL-6 (20). To investigate potential mechanisms for the upregulation of RANTES secretion by RV and IFN-, HBEC were either transfected with synthetic dsRNA (poly IC, 1–100 ng/ml) or synthetic single-stranded RNA (ssRNA, poly U, 1–100 ng/ml), or

were mock transfected. The samples were next incubated (24 h, 37C) with IFN- (10 U/ml), TNF- (100 U/ml), or medium alone, and RANTES secretion was determined. Transfection of poly IC alone induced RANTES secretion in a dose-dependent fashion, whereas IFN- or TNF- alone induced low levels of RANTES (Table 1). The combination of dsRNA transfection and IFN- incubation synergistically

Figure 5. IFN- enhances RANTES production in response to either a major (RV16)- or a minor (RV49)-group RV. Confluent monolayers of HBEC were incubated (48 h, 37C) with either RV16 (1 107 TCID50/ml), RV49 (1 106 TCID50/ml), or medium alone, in the presence (solid bars) or absence (open bars) of IFN- (10 U/ml). Supernatant fluids were analyzed for RANTES (A, n 9) and IL-8 (B, n 11) 48 h after inoculation. (C) Cell pellets of samples treated as described above were analyzed by semiquantitative RT-PCR and Southern blotting for RANTES and GAPDH mRNA. RANTES mRNA was undetectable in cells incubated in medium alone (lane 1), but was induced by either RV16 (lane 2), RV49 (lane 3), or IFN- (lane 4). Furthermore, the combination of IFN- along with either RV16 (lane 5) or RV49 (lane 6) induced the highest levels of RANTES mRNA. (D) The RANTES/G3PDH mRNA ratios from five separate experiments are expressed as mean SE, *P 0.05 compared with cells incubated with RV16 alone.


Figure 6. Influence of IFN- and TNF- on PIV-induced RANTES secretion in HBEC. Confluent monolayers of HBEC were inoculated with PIV (1 105 or 1 106 TCID50/ml), along with either IFN- (10 U/ml; solid bars), TNF- (100 U/ml; striped bars), or medium alone (open bars). Culture supernates collected 48 h later were analyzed for RANTES protein. Results are expressed as mean SE, n 5. *P 0.05 versus corresponding values obtained with uninfected cells.

enhanced RANTES secretion (P 0.05). TNF- also boosted dsRNA-induced RANTES (P 0.05 versus control), although to a lesser extent than IFN- (P 0.05). Transfection of the ssRNA analog poly U, or mock transfection, elicited little or no RANTES secretion ( 15 pg/ml, data not shown).

Discussion Epithelial cell–derived chemokines such as RANTES contribute to the initiation of the immune response to viral infections, and our studies suggest that in the case of RV, IFN- may be an important regulator of this process. Specifically, our studies demonstrate that IFN- can synergistically enhance RANTES mRNA and protein secretion from RV16 (major serotype RV)- or RV49 (minor serotype RV)-infected HBEC. These studies provide evidence of two distinct mechanisms for the effects of IFN-. First, IFN- can induce cellular receptors, i.e., ICAM-1 and LDLR, used by RV to gain entry into the host epithelial cell, and increasing the amount of receptor-mediated RV binding was associated with increased RANTES secretion. Second, these experiments provide several lines of evidence that IFN- can also enhance RV-induced RANTES secretion independent of effects on viral receptors. Finally, because IFN- promoted virus-induced RANTES, but not IL-8, this effect may be an important mechanism

599

TABLE 1

DsRNA and IFN- synergistically enhance RANTES secretion (pg/ml) Co-stimulus Poly IC (ng/well)

None

TNF- (100 U/ml)

IFN- (10 U/ml)

0 1 10 100

10 140 577 1,920

32* 167* 912* 3,119*

129† 541† 1,946† 5,220†

*P 0.05 versus no co-stimulus. † P 0.05 versus no co-stimulus or samples treated with TNF-.

by which mononuclear cells can modulate epithelial cell responses to viral infection. IFN- could potentiate responses to major-group RV by enhancing expression of ICAM-1. Accordingly, our experiments confirmed previous observations to indicate that IFN- is more potent than TNF- as an inducer of ICAM-1 on nontransformed human airway epithelial cells. The comparative effects of TNF- and IFN- on HBEC ICAM-1 expression mirrored effects of these cytokines on RV-induced RANTES secretion: these data suggest the possibility of a causal relationship whereby IFN- enhances RV-induced RANTES secretion by increasing viral binding to HBEC. In addition to enhancing ICAM-1 expression and RV binding to airway epithelial cells, several lines of evidence indicate that IFN- can promote RV-induced RANTES through a receptor-independent mechanism. First, TNF- did not augment RV-induced RANTES secretion in either HBEC or A549 cells, despite the fact that TNF- induced ICAM-1 expression, particularly on A549 cells. Further evidence of an intracellular effect of IFN- was provided by experiments conducted with the minor group RV49. Binding of RV49 to HBEC was high in cells incubated in medium alone, and there was no further enhancement by either IFN- or TNF-. This finding is consistent with the finding that RV49 also induces greater cytopathology in HBEC compared with RV16 (2). The binding was blocked by C7 mAb, indicating that the virus bound to

Figure 7. Effects of IFN- on RANTES secretion by A549 cells transfected with RV16 RNA. (A) A549 cells were transfected with either RV16 RNA (0.5 g/ well) or RNA isolated from the yeast Blastomycetes (0.5 g/well), washed, and incubated (24–48 h) with fresh medium in the presence (solid bars) or absence (open bars) of IFN- (10 U/ml). RV16 and IFN- had additive effects to stimulate RANTES secretion (n 5). In contrast, yeast RNA had relatively little effect. (B) The effects of IFN- were related to the amount of RV RNA transfected, and were greatest in combination with lower doses of RV RNA (n 4). (C) RV RNA (0.005 g/well) and IFN- had additive effects on RANTES secretion even in the presence of anti–ICAM-1 mAb (RR1 10 g/ml), which was added to prevent the binding of newly synthesized viral particles. Results are expressed as mean SE, *P 0.05 (n 3).

600


the LDLR, as is expected with a minor group RV. Our HBEC monolayers were grown in serum-free medium in vitro, and studies in hepatocytes suggest that growing cells in the absence of serum induces the expression of LDLR (21). Studies using bronchoscopy and bronchial lavage suggest that HBEC are also exposed to low concentrations of serum proteins in vivo, because bronchial secretions normally contain very little serum proteins such as IgG or albumin. The fact that IFN- augments RV49-induced RANTES secretion without affecting viral binding further supports the concept that IFN- can augment RV-induced RANTES production independent of effects on receptor expression. Finally, additional evidence for receptor-independent augmentation of RV-induced RANTES was provided by experiments in which RV RNA was transfected into cells without viral binding to cell surface receptors. For example, IFN- boosted RANTES secretion even when infectious RV RNA was transfected into epithelial cells, thus bypassing engagement of cellular receptors. This effect was independent of the binding of newly-synthesized viral particles to receptors on the cell surface, as demonstrated by performing the same experiment in the presence of blocking anti–ICAM-1 mAb. IFN- enhanced RANTES secretion after transfection of low doses of RV RNA, but had less effect when the cells were transfected with large doses (0.5 g) of RV RNA. One possible explanation for this effect is that RV RNA-induced RANTES production may depend on activation of double-stranded RNA-dependent protein kinase (PKR) (22) as it has recently been demonstrated that RV can activate PKR in B cells (23). The dose response curve for PKR activation by dsRNA is bell-shaped (24), and this pattern is similar to the dose–response curve of RV RNAinduced RANTES secretion. Our experiments support the concept that generation of dsRNA during RANTES replication may be an important factor in RV-induced RANTES production, in that transfection of poly IC was a potent stimulus for RANTES. Moreover, IFN- enhanced dsRNA-induced RANTES secretion, suggesting a possible intracellular mechanism for this effect. This finding suggests that IFN- and RV, or RV dsRNA, may activate signal transduction pathways that are complementary in their ability to activate RANTES gene transcription. In support of this hypothesis, RV infection causes nuclear factor- B (NF- B) activation (3), and the RANTES gene promoter contains four potential binding sites for NF- B (25, 26), as well as several IFN-–activated sites (TTNCNNNAA) (25, 27). Complementary effects of IFN- could include either induction of additional transcription factors (e.g., STAT-1) that enhance RANTES promoter activity, or stabilization of RANTES mRNA stability (28). We are now conducting additional experiments to test these hypotheses. The effect of IFN- on virus-induced RANTES generation appears to vary with the type of virus. It has previously been reported that IFN- enhances RSV-induced RANTES (29), but we did not find IFN- to enhance PIVinduced RANTES secretion. This finding, plus the fact that PIV alone was a better inducer of RANTES compared with RV, suggests that these different single-stranded RNA vi-

ruses activate distinct intracellular signaling pathways. Moreover, the lack of synergism with IFN- suggests that mechanisms used by PIV and IFN- to enhance RANTES production may overlap. In conclusion, these in vitro experiments demonstrate that IFN- is a potent positive regulator of RANTES secretion by RV-infected airway epithelial cells. This finding suggests that the secretion of IFN- by T cells and NK cells within airway tissues during the early stages of RV infection could regulate interactions between viral particles and host epithelial cells, and subsequent secretion of RANTES. These findings suggest that IFN- may have a dual role in the immune response to infections with RV and perhaps other respiratory viruses. First, studies in animals (30) and in humans (31) indicate that strong IFN- responses are associated with an effective antiviral response. On the other hand, because RANTES is a potent chemoattractant for memory T cells, monocytes, eosinophils, and basophils, augmentation of RANTES secretion in vivo could enhance airway inflammation and promote airway obstruction, and this may be of particular significance in patients with asthma. Further understanding of this pathway could lead to new insights into the pathogenesis of RV-induced inflammation during common colds and virus-induced exacerbations of asthma. Acknowledgments: The authors thank Claire R Dick for excellent technical assistance, and Dr. Robert Love for providing the bronchial tissue used to prepare the epithelial cell cultures. This work was supported by grants from the National Institutes of Health (AI-40685 and AI-34891).

References 1. Subauste, M. C., D. B. Jacoby, S. M. Richards, and D. Proud. 1995. Infection of a human respiratory epithelial cell line with rhinovirus: induction of cytokine release and modulation of susceptibility to infection by cytokine exposure. J. Clin. Invest. 96:549–557. 2. Schroth, M. K., E. Grimm, P. Frindt, D. M. Galagan, S. Konno, R. Love, and J. E. Gern. 1999. Rhinovirus replication causes RANTES production in primary bronchial epithelial cells. Am. J. Respir. Cell Mol. Biol. 20: 1220–1228. 3. Einarsson, O., G. P. Geba, Z. Zhu, M. Landry, and J. A. Elias. 1996. Interleukin-11: stimulation in vivo and in vitro by respiratory viruses and induction of airways hyperresponsiveness. J. Clin. Invest. 97:915–924. 4. Zhu, Z., W. Tang, Y. Wu, O. Einarsson, M. L. Landry, J. M. Gwaltney, Jr., and J. A. Elias. 1996. Rhinovirus stimulation of interleukin-6 in vivo and in vitro: evidence for nuclear factor kappa B-dependent transcriptional activation. J. Clin. Invest. 97:421–430. 5. Rot, A., M. Krieger, T. Brunner, S. C. Bischoff, T. J. Schall, and C. A. Dahinden. 1992. RANTES and macrophage inflammatory protein 1 alpha induce the migration and activation of normal human eosinophil granulocytes. J. Exp. Med. 176:1489–1495. 6. Schall, T. J., K. B. Bacon, K. J. Toy, and P. S. Goedegebuure. 1990. Selective attraction of monocytes and T lymphocytes of the memory phenotype by cytokine RANTES. Nature 347:669–671. 7. Zhang, L., A. E. Redington, and S. T. Holgate. 1994. RANTES: A novel mediator of allergic inflammation? Clin. Exp. Allergy 24:899–904. 8. Baggiolini, M., B. Dewald, and M. Moser. 1994. Interleukin-8 and related chemotactic cytokines. Adv. Immunol. 55:97–179. 9. Teran, L. M., S. L. Johnston, J. M. Schroder, M. K. Church, and S. T. Holgate. 1997. Role of nasal interleukin-8 in neutrophil recruitment and activation in children with virus-induced asthma. Am. J. Respir. Crit. Care Med. 155:1362–1366. 10. Teran, L. M., M. C. Seminario, J. K. Shute, A. Papi, S. J. Compton, J. L. Low, G. J. Gleich, and S. L. Johnston. 1999. RANTES, macrophage-inhibitory protein 1, and the eosinophil product major basic protein are released into upper respiratory secretions during virus-induced asthma exacerbations in children. J. Infect. Dis. 179:677–681. 11. Grünberg, K., M. C. Timmers, H. H. Smits, E. P. A. De Klerk, E. C. Dick, W. J. M. Spaan, P. S. Hiemstra, and P. J. Sterk. 1997. Effect of experimental rhinovirus 16 colds on airway hyperresponsiveness to histamine and interleukin-8 in nasal lavage in asthmatic subjects in vivo. Clin. Exp. Allergy 27:36–45. 12. Gern, J. E., E. C. Dick, W. M. Lee, S. Murray, K. Meyer, Z. T. Handzel, and W. W. Busse. 1996. Rhinovirus enters but does not replicate inside monocytes and airway macrophages. J. Immunol. 156:621–627.


13. Look, D. C., S. R. Rapp, B. T. Keller, and M. J. Holtzman. 1992. Selective induction of intercellular adhesion molecule-1 by interferon-gamma in human airway epithelial cells. Am. J. Physiol. 263:L79–L87. 14. Rothlein, R., M. Czajkowski, M. M. O’Neill, S. D. Marlin, E. Mainolfi, and V. J. Merluzzi. 1988. Induction of intercellular adhesion molecule 1 on primary and continuous cell lines by pro-inflammatory cytokines. J. Immunol. 141:1665–1669. 15. Greve, J. M., G. Davis, A. M. Meyer, C. P. Forte, S. C. Yost, C. W. Marlor, M. E. Kamarck, and A. Mcclelland. 1989. The major human rhinovirus receptor is ICAM-1. Cell 56:839–847. 16. Staunton, D. E., V. J. Merluzzi, R. Rothlein, R. Barton, S. C. Marlin, and T. A. Springer. 1989. A cell adhesion molecule, ICAM-1, is the major surface receptor for rhinoviruses. Cell 56:849–853. 17. Lee, W. M., S. S. Monroe, and R. R. Rueckert. 1993. Role of maturation cleavage in infectivity of picornaviruses: activation of an infectosome. J. Virol. 67:2110–2122. 18. Lee, W. M., W. S. Wang, and R. R. Rueckert. 1995. Complete sequence of the RNA genome of human rhinovirus 16, a clinically useful common cold virus belonging to the ICAM-1 receptor group. Virus Genes 9:177–181. 19. Rueckert, R. R. 1996. Picornaviridae and their replication. In B. N. Fields, D. M. Knipe, P. M. Howley, R. M. Chanock, J. L. Melnick, T. P. Monath, B. Roizman, and S. E. Straus, editors. Fields Virology, 3rd ed. LippencottRaven Publishers, Philadelphia. 609–654. 20. Williams, B. R. G. 1999. PKR: a sentinel kinase for cellular stress. Oncogene 18:6112–6120. 21. Grove, R. I., C. Mazzucco, N. Allegretto, P. A. Kiener, G. Spitalny, S. F. Radka, M. Shoyab, M. Antonaccio, and G. A. Warr. 1991. Macrophagederived factors increase low density lipoprotein uptake and receptor number in cultured human liver cells. J. Lipid Res. 32:1889–1897. 22. Clemens, M. J. 1997. PKR–a protein kinase regulated by double-stranded RNA. Int. J. Biochem. Cell Biol. 29:945–949. (Review)

601

23. Rager, K. J., J. O. Langland, B. L. Jacobs, D. Proud, D. G. Marsh, and F. Imani. 1998. Activation of antiviral protein kinase leads to immunoglobulin E class switching in human B cells. J. Virol. 72:1171–1176. 24. Manche, L., S. R. Green, C. Schmedt, and M. B. Mathews. 1992. Interactions between double-stranded RNA regulators and the protein kinase DAI. Mol. Cell. Biol. 12:5238–5248. 25. Nelson, P. J., H. T. Kim, W. C. Manning, T. J. Goralski, and A. M. Krensky. 1993. Genomic organization and transcriptional regulation of the RANTES chemokine gene. J. Immunol. 151:2601–2612. 26. Moriuchi, H., M. Moriuchi, and A. S. Fauci. 1997. Nuclear factor-kappa B potently up-regulates the promoter activity of RANTES, a chemokine that blocks HIV infection. J. Immunol. 158:3483–3491. 27. Seidel, H. M., L. H. Milocco, P. Lamb, J. E. Darnell, Jr., R. B. Stein, and J. Rosen. 1995. Spacing of palindromic half sites as a determinant of selective STAT (signal transducers and activators of transcription) DNA binding and transcriptional activity. Proc. Natl. Acad. Sci. USA 92:3041–3045. 28. Koga, T., E. Sardina, R. M. Tidwell, M. Pelletier, D. C. Look, and M. J. Holtzman. 1999. Virus-inducible expression of a host chemokine gene relies on replication-linked mRNA stabilization. Proc. Natl. Acad. Sci. USA 96:5680–5685. 29. Olszewska-Pazdrak, B., A. Casola, T. Saito, R. Alam, S. E. Crowe, F. Mei, P. L. Ogra, and R. P. Garofalo. 1998. Cell-specific expression of RANTES, MCP-1, and MIP-1alpha by lower airway epithelial cells and eosinophils infected with respiratory syncytial virus. J. Virol. 72:4756–4764. 30. Openshaw, P. J. M. 1995. Immunity and immunopathology to respiratory syncytial virus: the mouse model. Am. J. Respir. Crit. Care Med. 152(Suppl.): S59–S62. 31. Parry, D. E., W. W. Busse, K. A. Sukow, C. R. Dick, C. A. Swenson, and J. E. Gern. 2000. Rhinovirus-induced peripheral blood mononuclear cell responses and outcome of experimental infection in allergic subjects. J. Allergy Clin. Immunol. 105:692–698.