The JI PBL Intereron Source
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rollo, E. E.
Right arrow Articles by Shaw, R. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rollo, E. E.
Right arrow Articles by Shaw, R. D.
Right arrowPubmed/NCBI databases
Medline Plus Health Information
*Rotavirus Infections
The Journal of Immunology, 1999, 163: 4442-4452.
Copyright © 1999 by The American Association of Immunologists

The Epithelial Cell Response to Rotavirus Infection1

Ellen E. Rollo*, K. Prasanna Kumar{dagger}, Nancy C. Reich{dagger}, Jean Cohen{ddagger}, Juana Angel2, Harry B. Greenberg§, Riten Sheth*, Joseph Anderson*, Brian Oh*, Scott J. Hempson*, Erich R. Mackow* and Robert D. Shaw3,*

* Department of Medicine, Northport Veterans Affairs Medical Center, Northport, NY 11768; {dagger} Department of Pathology, State University of New York, Stony Brook, NY 11794; {ddagger} Laboratoire de Virologie et d’Immunologie Moleculaire, Institut National de la Recherche Agronomique, C. R. J. Domaine de Vilvert, Jouy-en-Josas, France; and § Departments of Medicine and Microbiology, Palo Alto Veterans Affairs Medical Center and Stanford University, Stanford, CA 94305


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Rotavirus is the most important worldwide cause of severe gastroenteritis in infants and young children. Intestinal epithelial cells are the principal targets of rotavirus infection, but the response of enterocytes to rotavirus infection is largely unknown. We determined that rotavirus infection of HT-29 intestinal epithelial cells results in prompt activation of NF-{kappa}B (<2 h), STAT1, and ISG F3 (3 h). Genetically inactivated rotavirus and virus-like particles assembled from baculovirus-expressed viral proteins also activated NF-{kappa}B. Rotavirus infection of HT-29 cells induced mRNA for several C-C and C-X-C chemokines as well as IFNs and GM-CSF. Mice infected with simian rotavirus or murine rotavirus responded similarly with the enhanced expression of a profile of C-C and C-X-C chemokines. The rotavirus-stimulated increase in chemokine mRNA was undiminished in mice lacking mast cells or lymphocytes. Rotavirus induced chemokines only in mice <15 days of age despite documented infection in older mice. Macrophage inflammatory protein-1ß and IFN-stimulated protein 10 mRNA responses occurred, but were reduced in p50-/- mice. Macrophage inflammatory protein-1ß expression during rotavirus infection localized to the intestinal epithelial cell in murine intestine. These results show that the intestinal epithelial cell is an active component of the host response to rotavirus infection.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Rotavirus is the leading cause of severe gastroenteritis in infants and young children around the world. The virus infects only the mature villus epithelial cells at the apices of small intestinal villi. Rotavirus replication in many species proceeds without rapidly killing the enterocyte or denuding villi (1), but diarrhea may be severe and may cause fatal dehydration. The intestinal epithelial cell is well positioned to regulate the initial host response to infection, potentially to enhance immune function or contribute to disease by altered absorption, secretion, or barrier function.

The epithelium of the small intestine is topographically organized into crypts and villi (1). The latter project up to 1 mm into the intestinal lumen. Proliferation and upward migration of undifferentiated crypt cells are associated with a gradual assumption of the terminally differentiated enterocyte phenotype. A junctional complex that separates apical from basolateral domains interconnects these polarized cells. Rotaviruses infect and replicate within the mature epithelial cells at the apex of the villus, and viral progeny exit the cell from the lumenal membrane by a process of transcytosis (2). The basal epithelial membrane is firmly affixed to a basement membrane, beneath which is the lamina propria that is populated by the inflammatory mononuclear infiltrate that responds to rotavirus infection.

Intestinal epithelial cells are implicated in the initiation of the host response to infection. Bacterial enteric pathogens elicit a variety of inflammatory cytokines from intestinal epithelial cells that are vectorially secreted from the basolateral surface (3). These include IL-8, monocyte chemotactic protein-1 (MCP-1),4 GM-CSF, and TNF-{alpha}. The functional consequences in the intestine of increased local production of these and other cytokines include chemoattraction and activation of immune cells, although the entire range of activities is incompletely understood. Rotavirus infection causes recruitment and activation of CD4+ and CD8+ T cells (4, 5, 6, 7) and a vigorous mucosal IgA response (8, 9). These effectors of acquired immune responses work in concert with other poorly defined innate host defenses to limit natural disease and prevent reinfection. The extent to which innate or acquired immune responses to rotavirus infection are initially regulated by epithelial cytokines is unknown.

Epithelial dysfunction during rotavirus infection has also been suggested as a cause of diarrhea. Altered regulation of the expression of digestive enzymes (10), enhanced chloride secretion in response to a viral enterotoxin (11), and postischemic damage (12) have been proposed. Recently, epithelial dysfunction caused by cytokines in inflammatory diarrheal diseases has been postulated. TNF-{alpha} and IFN-{gamma} are cytotoxic to intestinal HT-29 cells (13), IFN-{gamma} alters the permeability of tight junctions in T84 cells (14), and T cell products alter ion secretion in murine intestine (15). Receptors on enterocytes for several cytokines have been identified (16). Therefore, the early cellular response of infected epithelium may be important in rotavirus pathophysiology as well as in host immunity.

Chemokines constitute a superfamily of 8- to 10-kDa inducible secreted proinflammatory cytokines that serve as chemoattractants and cell activators in immune and inflammatory responses. IL-8 is a chemokine that has recently been implicated in pathogenesis and immunity in several diseases (17, 18, 19, 20). We previously established the existence of a primary IL-8 response to rotavirus infection by intestinal epithelial cells (21). Increased IL-8 mRNA and IL-8 secretion are primary responses to rotavirus infection and not the secondary result of increased IL-1, TNF-{alpha}, or IFN-{alpha}. The IL-8 gene promoter region includes binding sites for transcription factors NF-{kappa}B, (NF)-IL-6, AP-1, AP-3, and others (22, 23). However, no information is available on cellular transcriptional regulation during rotavirus infection.

In the present study we evaluated the effects of rotavirus infection on NF-{kappa}B and IFN transcriptional regulatory elements. NF-{kappa}B was chosen because of the important role this transcription factor plays in the production of IL-8 in the intestine (24), while IFN was chosen because previous studies identified an IFN response to rotavirus infection (25). Selected intestinal cytokine and chemokine responses to rotavirus were defined, and essential differences were noted from response patterns associated with bacterial pathogens. A well-characterized murine model of rotavirus infection provided an opportunity to demonstrate significant correlation of the in vitro and in vivo cellular responses to rotavirus infection. We found that rotavirus-induced intestinal cytokine responses in mice, as in cells, were at least partially regulated by NF-{kappa}B transcriptional activation. We also demonstrated that murine intestinal chemokine responses are age restricted and correlate temporally with age-restricted diarrhea in mice. These data indicate a rapid and complex epithelial cellular response to rotavirus infection that may be critical to the development of the host response.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Animals

BALB/c pregnant dams and prp-RAG breeding pairs were obtained from Taconic Farms (Germantown, NY) and housed in microisolator cages containing sterile bedding, food, and water. WB/ReJ W/+, and C57BL/6J Wv/+, and NF-{kappa}B1 p50-/- breeding pairs (26) were obtained from The Jackson Laboratory (Bar Harbor, ME). KitW/KitW-v offspring were selected by color and compared with heterozygous littermates and congenic controls. Breeding mice were seronegative for rotavirus. W/Wv offspring are mast cell deficient due to mutations involving the tyrosine kinase receptor c-Kit (27, 28, 29).

Mice were infected by gavage with 50–100 µl of virus solution (~1–2 x 107 focus-forming units (ffu)). Sucklings were isolated from the dam for 1 h before and after inoculation. Sham inoculations were performed with medium 199 without additives. Diarrhea observations were made using modifications of the previously published protocol (30). A four-point qualitative scale was applied as follows: 1, normal brown formed stool; 2, yellow pasty; 3, yellow mixed liquid and solid; and 4, entirely liquid. Numbers were assigned by a blinded rater.

Cells, virus production, and purification

Rhesus rotavirus (RRV; a type 3 simian rotavirus strain) was propagated in MA104 cells as previously described (21), and purified by sucrose gradient centrifugation. Virus bands were identified by hemagglutination, dialyzed against Tris-saline-2 mM CaCl2, and reconcentrated by ultracentrifugation. Aliquots were used to prepare psoralen-inactivated virus (PI-RRV) by incubation with 4'-aminomethyl-4,5',8-trimethyl psoralen (40 µg/ml) on ice under A365 UV light for 50 min (31). PI-RRV was characterized by HA titer (1/6400 to 1/12,800) and failure to infect MA104 monolayers as measured by immunohistochemical infectious focus assay 24–96 h postinfection (p.i.). The wild-type murine rotavirus EC (type G3/P16) was prepared from intestinal homogenates as previously described (32). This strain is not tissue culture adapted, but is passaged in mice to maintain a high degree of virulence. Infectious viral titer was determined by oral inoculation of mice with serial 10-fold dilutions. The human intestinal epithelial cell line HT-29 (American Type Culture Collection, Manassas, VA) was grown in DMEM/Ham’s F-12 supplemented with 10% FBS, 100 µg/ml penicillin, and 100 µg/ml streptomycin at 37°C in an atmosphere of 5% CO2. All tissue culture media were free of detectable endotoxin (<0.10 ng/ml) as determined by assay with Limulus amebocyte lysate (ICN, Costa Mesa, CA).

HT-29 cell monolayers were grown to confluence, washed twice with medium without serum, and then infected with rotavirus at various doses, expressed as multiplicity of infectious virus particles per cell (moi).

Virus-like particles (VLPs)

Rotavirus VLPs were made by baculovirus expression of RRV proteins as previously reported (33). VLPs containing rotavirus structural proteins VP2, VP6, and VP7 (2/6/7) or 2/4/6/7 were purified on CsCl gradients, desalted in G25 columns, and assayed for HA activity. Comparable amounts of VLPs were used in these assays. Only 2/4/6/7 VLPs hemagglutinated and HA activity were used to standardize the amount of RRV and VLPs applied to cells. RRV and 2/4/6/7 VLPs were trypsin activated (0.01%) for 15 min at 37 C to cleave VP4. Trypsin cleavage is required to render RRV infectious and to permit VLPs to penetrate cell membranes.

Monolayers were washed with DMEM, and then VLPs or virus were adsorbed for 1 h at 37°C. Comparable HA titers of RRV and 2/4/6/7 VLPs (50 HA units) were applied to duplicate wells of HT-29 cells in six-well dishes. Corresponding microgram amounts of 2/6/7 VLPs or media (mock adsorbed) were applied to duplicate wells. VLPs or viruses were removed, monolayers were washed three times and incubated for 24 h at 37°C. Supernatants were assayed for the presence of secreted IL-8 using a Quantikine ELISA assay (R&D Systems, Minneapolis, MN).

VLPs or virus were assayed for ability to activate NF-{kappa}B. A plasmid containing five NF-{kappa}B binding sites upstream of a luciferase reporter gene was transfected with FuGene6 for 12 h into 60% confluent HEK293 cells in six-well dishes (1 µg/well). Twenty-four hours posttransfection, cells were washed, and virus or VLPs were trypsin activated and adsorbed to monolayers as described above. 2/4/6/7 VLPs (4 µg), HA titer-matched RRV (trypsin-activated), or 2/6/7 VLPs (4 µg) were adsorbed to duplicate wells of previously transfected cells. Following a 90-min adsorption, cells were washed and incubated at 37°C for 5 h. Monolayers were washed three times with PBS and lysed in 0.2 ml of luciferase assay buffer (Promega, Madison, WI). Aliquots of lysate were assayed for luciferase activity in a luminometer, and relative light units were standardized to the protein content of the lysates with a protein assay (Bradford reagent, Bio-Rad, Hercules, CA).

Intestinal fluid collection

The contents of small intestines were collected as previously described (8). Briefly, after removal, the outside was rinsed with cold HBSS. Cold buffer (5 ml) was rinsed through the inside of each intestine and mixed 1/1 with protease inhibitor solution (50 mM EDTA and 0.1 mg/ml soybean trypsin inhibitor, pH 7.4). The tube was vortex-mixed for 1 min and centrifuged (850 x g, 4°C) for 10 min. PMSF (100 µl; 100 mM) was added. The tube was centrifuged (1300 x g, 4°C) for 25 min. One hundred microliters of PMSF, 0.01% sodium azide, and 300 µl of FBS were added to supernatant in a clean tube. Aliquots were frozen at -80°C until testing.

Detection of rotavirus intestinal IgA Abs

IgA-specific Abs to rotavirus were measured in intestinal contents using an ELISA modified from a published method (8). A guinea pig anti-RRV serum was used as capture Ab on Immulon II microtiter plates (Dynatech, Chantilly, VA). Two-fold dilutions of intestinal fluid samples in PBS/1% FBS were added and incubated for 2 h. Bound Abs were detected using biotinylated isotype-specific goat anti-mouse Ig and avidin-HRP. Soluble TMB (KPL, Gaithersburg, MD) was used as a substrate. The reaction was stopped after 10 min with 1 N H2SO4, and A450 was measured. Nonspecific binding was evaluated in wells without RRV, and values were subtracted from RRV-containing wells. The titer was the greatest dilution that exceeded mean background by 2 SD.

EMSA

Cells and virus infection. HT29 cells were grown to 80–90% confluence in 100-mm plates. For virus infection, cells were washed in serum-free medium and overlaid with 2 ml of serum-free medium containing live or inactivated virus at an moi of 8–15. After the specified time of infection, the cells were washed with PBS and lysed.

Preparation of nuclear extracts for EMSA. Nuclear extracts were prepared as previously described (34). To identify the IFN-{alpha}/ß-induced ISGF3, extracts were incubated with 32P end-labeled double-stranded oligonucleotide corresponding to the IFN-stimulated response element (ISRE) site from the ISG15 gene, 5'-GGGAAAGGGAAACCGAAACTGAA. The DNA binding reactions contained 10 µg of nuclear protein in a 15-µl reaction mixture (with 2 µg of poly(dI-dC), 0.5 µg of nonspecific plasmid DNA, and 0.5 µg of herring sperm DNA) (35). The complexes were resolved in a 4.5% polyacrylamide gel with 0.25 x TBE at 4°C. The activation of NF-{kappa}B was similarly assayed using a double-stranded oligonucleotide probe with the sequence 5'-TCAACAGAGGGGACTTTCCGAGAGG (36). STAT1 activation was assayed with a IFN-{gamma}-activated site (GAS) sequence from the insulin response factor-1 site, 5'-GCCTGATTTCCCCGAAATGACGG double-stranded oligonucleotide.

IL-8 promoter/chloramphenicol acetyltransferase (CAT) reporter

The IL-8 promoter region from -450 to 100 bases was inserted into the PstI site in pCAT-Basic vector (Promega) at the -50 site from the CAT gene. The hygromycin resistance gene was inserted into the BamHI site of the vector. This IL-8/pCAT construct (10 µg) was transfected into ECV-304 human endothelial cells with Lipofectamine (Life Technologies, Grand Island, NY). Cells were grown in the presence of 150 µg/ml hygromycin starting 48 h posttransformation. Induction of CAT was measured by ELISA (Boehringer Mannheim, Indianapolis, IN). RRV infected this cell line by immunohistochemical assay (data not shown). Confluent monolayers in a six-well dish were exposed to RRV and PI-RRV (psoralen inactivated) for 24 h. The monolayer was washed and lysed, and protein was measured. One hundred and fifty microliters of total cellular protein was used for the CAT ELISA. Noninfected cells were used to determine the constitutive level of CAT expression, and TNF-{alpha} was used as a positive control for IL-8 induction. Results were correlated to IL-8 levels in the supernatant that were measured by a commercial IL-8 ELISA assay (R&D Systems, Minneapolis, MN) according to the manufacturer’s directions and as previously described (21).

PCR amplification of chemokine mRNA

The small intestines from untreated mice and those that had been infected with RRV for various times were removed, rinsed twice with sterile PBS, pH 7.4, homogenized in 3 ml of Tri-Reagent (Molecular Research Center, Cincinnati, OH). Total cellular RNA was extracted according to the manufacturer’s instructions. Briefly, after homogenization, the samples were allowed to stand for 5 min at room temperature. 0.6 µl of chloroform was added, and samples were allowed to stand for 15 min at room temperature. The mixture was centrifuged at 12,000 x g for 15 min at 4°C. The aqueous phase was removed, 1.5 ml of isopropanol was added, the mixture was allowed to stand at room temperature for 10 min. It was then centrifuged at 12,000 x g for 10 min at 4°C. The pellet was washed with 3 ml of 75% ethanol, dried, and resuspended in 200 µl of DEPC-treated 10 mM Tris-HCl and 1 mM EDTA (pH 7.5). SDS was added to 0.5%, proteinase K was added to 0.5 mg/ml, and the mixture was incubated at 50°C for 1 h, phenol/chloroform extracted, and ethanol precipitated. The pellet was air-dried and resuspended in DEPC-treated water. The RNA was fractionated on a 1.5% agarose formaldehyde gel and stained with 1 mg/ml of ethidium bromide to confirm that the RNA was not degraded.

The human PCR primers were: IP-10 (341-bp product): sense, 5'-GAGAAAGAGATGTCTGAATCC-3'; antisense, 5'-ATAGCACCTCAGTAGAGC-3'); IFN-ß (420-bp product); sense, 5'-ACGACAGCTCTTTCCATG-3'; antisense, 5'-TTGGCCTTCAGGTAATGCAG-3'); RANTES (254-bp product), sense, 5'-TCATCCTCATTGCTACTG-3'; antisense, 5'-CTCCATCCTAGCTCATCTCCAAA-3'); MIP-1{alpha} (526-bp product): sense, 5'-TCTGCATCACTTGCTGCTGACA-3'; antisense, 5'-TGCCTACACAGGCTGATGACA-3'); MIP-1-ß (331-bp product): sense, 5'-TCTATGATGACATCGCCCT-3'; antisense, 5'-GACTTACTTGGTGCTGAGAG-3'); Gro-{gamma} (271-bp product): sense, 5'-AGTGCTTGCAGACACTGCA-3'; antisense, 5'-TTTCAGCTCTGGTAAGGGCA-3'); IL-1{alpha} (408-bp product): sense, 5'-GTAAGCTATGGCCCACTCCAT-3'; antisense, 5'-TGACTTATAAGCACCCATGTC-3'); IL-1ß (408-bp product): sense, 5'-GACCTGGACCTCTGCCCTCTG-3'; antisense, 5'-AGGTATTTTGTCATTACTTTC-3'); GM-CSF (165-bp product): sense, 5'-CTGCTGAGATGAATGAAACAG-3'; antisense, 5'-AGTGTAGTGGCT-3'); TNF-{alpha} (702-bp product): sense, 5'-ATGAGCACTGAAAGCATGATC-3'; antisense, 5'-TCACAGGGCAATGATCCCAAAGTAGACCTGCCC-3'); GAPDH (600-bp product); sense, 5'-CCACCCATGGCAGCAAATTCCATGGCA-3'; antisense, 5'-AGTGGACCTGACCTGCCGTGTAGA-3'); ß-actin (304-bp product): sense, 5'-TCCTGTGGCATCCACGAAACT-3'; antisense, 5'-GAAGCATTTGCGGTGGACGAT; IFN-{alpha} (315-bp product): sense, 5'-CAGGAGGAGTTTGATGGCAACCAG-3'; antisense, 5'-GACAACCTCCCAGGCACAAGGGC-3'); IL-8 (253-bp product): sense, 5'-GGCTCTCTTGGCAGCCTTCCTG-3'; antisense, 5'-CTTCTCCACAACCCTCTGCACCCA-3'); MCP-1 (576-bp product): sense, 5'-CTGAAGCTCGCACTCTCGCCTCC-3'; antisense, 5'-CAAAACATCCCAGGGGTAGAACTG-3'); and lymphotactin (400-bp product): sense, 5'-GCGGGACCTCAGCCATGAGA-3'; antisense, 5'-ATGAGCTGGCTGGCTGGAGA-3'.

The murine primers included: TNF-{alpha} (309-bp product): sense, 5'-GGCAGGTCTACTTTGGAGTCATTGC-3'; antisense, 5'-ACATTCGAGGCTCCAGTGAATTCGG-3'); IFN-{alpha} (308-bp product): sense, 5'-TCTCTCCTGCCTGAAGGAC-3'; antisense, 5'-ACACAGTGATCCTGTGGAA-3'); IFN-ß (390-bp product): sense, 5'-CAGCTCCAAGAAAGGACGAA-3'; antisense, 5'-GTAGCTGTTGTACTTCATGAG-3'); IL-1{alpha} (288-bp product): sense, 5'-CTCTAGAGCACCATGCTACAGAC-3'; antisense, 5'-TGGAATCCAGGGGAAACACTG-3'); IL-1ß (381-bp product): sense, 5'-GCAACTGTTCCTGAACTCA-3'; antisense, 5'-CTCGGAGCCTGTAGTGCAG-3'); GM-CSF (430-bp product): sense, 5'-TTCCTGGGCATTGTGGTCT-3'; antisense, 5'-TGGATTCAGAGCTGGCCTGG-3'); GAPDH (357-bp product): sense, 5'-CGGAGTCAACGGATTTGGTCGTAT-3'; antisense, 5'-AGCCTTCTCCATGGTGGTGAAGAC-3'); RANTES (254-bp product): sense, 5'-TCATCCTCACTGCAGCCGCCC-3'; antisense, 5'-CTCTATCCTAGCTCATCTCCAAA-3'); IP-10 (431-bp product): sense, 5'-CGCACCTCCACATAGCTTACAG-3'; antisense, 5'-CCTATCCTGCCCACGTGTTGAG-3'); MCP-1 (582-bp product): sense, 5'-GGAAAAAT GGATCCACACCTTGC-3'; antisense, 5'-TCTCTTCCTCCACCACCATGCAG-3'); MIP-1{alpha} (561-bp product): sense, 5'-GAAGAGTCCCTCGATGTGGCTA-3'; antisense, 5'-CCCTTTTCTGTTCTGCTGACAAG-3'); MIP-1ß (540-bp product): sense, 5'-CCACAATAGCAGAGAAA CAGCAAT-3'; antisense, 5'-AACCCCGAGCAACACCATGAAG3'); KC (530-bp product): sense, 5'-GACGAGACCAGGAGAAACAGGG-3'; antisense, 5'-AACGGAGAA AGAAGACAGACTGCT-3'); MIP-2 (377-bp product): sense, 5'-CTTCCTCGGGCACTCCAGA-3'; antisense, 5'-GGACAGCAGCCCAGGCTCCT-3'); lymphotactin: sense, 5'-CCCAGCAAGACCTCAGCC-3'; antisense, 5'-GAGGCTGTTACC CAGTCAGGG-3'; and ß-actin (245-bp product): sense, 5'-GTGGGCCGCTCTAGGCACCA-3'; antisense, 5'-CGGTTGGCCTTAGGGTTCAGGGGG-3'. The primers were complementary to mRNA sequences within the chemokine gene of interest.

cDNA was made at each time point in a total reaction volume of 100 µl combining 0.2 mg of oligo(dT) and 8 µg of RNA in DEPC-treated water, boiling for 1 min, chilling on ice, adding 1x PCR buffer (10 mM Tris-HCl and 50 mM KCl, pH 8.3, 20°C), 2.5 mM MgCl2, 1 mM of each deoxynucleotide, 1 U/ml of RNase inhibitor, and 50 U of Moloney murine leukemia virus reverse transcriptase (Epicentre Technologies, Madison, WI). After 1 h of incubation at 37°C, then 10 min at 95°C, 5 µl of each RT reaction was combined with MgCl2 to 1.5 mM, 1 µl of each chemokine-specific sense and antisense primer, and 0.5 U of Tfl DNA polymerase (Epicentre Technologies). The cDNAs were amplified in a Perkin-Elmer/Cetus DNA thermal cycler (Palo Alto, CA). Each PCR product was then fractionated on a 1.7% agarose gel containing 0.5 mg/ml of ethidium bromide, viewed with a transilluminator, and photographed using Polaroid 665 instant film (Polaroid, Cambridge, MA). Mouse ß-actin or GADPH served as the internal reverse transcription-PCR control. The appropriate number of PCR cycles was selected for each sample to ensure that serial dilutions of the internal control provided proportionate changes in densitometry.

Immunohistochemistry

Immediately after sacrifice, 24-h rotavirus-infected and sham-infected mouse pup intestines were infused with 10% neutral-buffered formalin, excised, dehydrated in a series of graded ethanol solutions, and embedded in paraffin. Sections were deparaffinized, rehydrated, and incubated with 3% H2O2 for 30 min to eliminate endogenous peroxidase activity. After blocking with 10% normal rabbit serum in PBS, pH 7.4, for 1 h, the sections were incubated for 60 min with anti-MIP-1 polyclonal Ab (R & D Systems). The sections were rinsed in PBS and incubated with biotinylated goat anti-mouse IgG for 30 min and subsequently with the Vectastain ABC reagent (Vectastain ABC kit, Vector Laboratories, Burlingame, CA) and the peroxidase substrate solution diaminobenzidine. For rotavirus detection, rabbit hyperimmune antiserum was used as the primary Ab. The sections were photographed on a Nikon Labophot-2 microscope (Melville, NY) equipped with a MTI 3CCD camera (DAGE-MTI, Michigan City, MI) and captured using National Institutes of Health Image version 1.6 on a Macintosh computer with a Scion CG-7 image capture board.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Cell studies

Activation of transcription factors in response to rotavirus infection. Rotavirus infection generates the production of viral dsRNA, which can activate latent cellular transcription factors. The latent NF-{kappa}B transcription factor is normally resident in the cytoplasm of the cell in association with inhibitory proteins, I{kappa}Bs (recently reviewed in Ref. 37). Phosphorylation of I{kappa}Bs by activated kinases can result in degradation and release of NF-{kappa}B that is now free to translocate to the nucleus. In the nucleus it binds to specific target sites and can activate the transcription of responsive genes. The gene encoding the IL-8 chemokine has NF-{kappa}B target sites within its promoter. To demonstrate the activation of NF-{kappa}B in response to rotaviral infection, an EMSA was used. Nuclear extracts were prepared from the HT29 human colon cell line before and after infection and were used in an DNA binding assay with an NF-{kappa}B target site (Fig. 1GoA). Activation of NF-{kappa}B was observed early in infection and was readily detected 2 h after viral infection.



View larger version (77K):
[in this window]
[in a new window]
  		FIGURE 1. EMSA analysis of transcription factor activation during the course of rotavirus infection. A, Nuclear extracts were prepared from mock-infected HT29 cells or cells infected with RRV for 2 or 4 h and used in a DNA binding reaction with the NF-{kappa}B oligonucleotide. NF-{kappa}B complexes are indicated, and specificity is demonstrated by inclusion of a 100-fold molar excess of unlabeled oligonucleotide (NF-{kappa}B comp) in the DNA binding reaction. B, Nuclear extracts were prepared from mock-infected HT29 cells or cells infected with RRV for 2, 3, or 4 h and used in a DNA binding reaction with the ISRE oligonucleotide. ISGF3 complexes are indicated, and specificity is demonstrated by inclusion of a 100-fold molar excess of unlabeled oligonucleotide (ISRE competitor) in the DNA binding reaction.

 
Cells are also know to respond to viral infection with the activation of the type I IFN (IFN-{alpha}/ß) genes (reviewed in Ref. 38). IFNs are produced and secreted from the infected cell and can function in an autocrine or paracrine fashion. IFNs act by binding to cell surface receptors and stimulating latent Janus tyrosine kinases that, in turn, phosphorylate members of the transcription factor family, the STATs (reviewed in Ref. 39). The response to type I IFN is the activation a cytoplasmic transcription factor termed IFN-stimulated gene factor-3 (ISGF3) and its subsequent translocation to the nucleus. ISGF3 contains STAT1, STAT2, and a non-STAT molecule, p48. Activated ISGF3 binds to a DNA target site within stimulated genes designated ISRE. To determine whether rotavirus infection induced the production and action of IFN-{alpha}/ß, an EMSA analysis was performed with the ISGF3 target site (Fig. 1GoB). Activation of ISGF3 was detected in infected HT29 cells, but with delayed kinetics compared with those of NF-{kappa}B. ISGF3 was not apparent at 2 h p.i. , but appeared by 3 h p.i. . The fact that the kinetics of activation of NF-{kappa}B precede ISGF3 supports our findings indicating that NF-{kappa}B is activated in a primary manner during rotavirus infection. Because NF-{kappa}B is a transcription factor that has been shown to be involved in the transcriptional response of the IFN-ß gene and many other cytokine genes, it may be a causative inducer of type I IFNs (40).

To determine whether rotavirus replication was required for NF-{kappa}B activation, experiments were first performed with PI-RRV. This was prepared from an RRV preparation of 4 x 108 ffu/ml with an HA titer of 1/6400. The inactivated sample contained no detectable infectious virus, but maintained an HA titer of 1/6400. Infections were performed with infectious or inactivated virus standardized by HA titer (Fig. 2Go). Both infectious and inactivated RRV caused nuclear translocation of NF-{kappa}B. However, activation of NF-{kappa}B by infectious RRV was significantly greater (the PI-RRV activation was 18.7% by densitometry of the RRV result; Fig. 2Go). To evaluate the effects of inactivated rotavirus on the IFN signal pathway we tested the appearance of a STAT transcription factor. The response to autocrine IFN not only activates ISGF3 as described above, but also activates a STAT1 dimer that can bind to a distinct target site referred to as GAS (39). We used an EMSA to assess the autocrine response of cells to inactivated rotavirus and found that psoralen-inactivated rotavirus could still elicit the autocrine activity of IFN (Fig. 2GoB).



View larger version (62K):
[in this window]
[in a new window]
  		FIGURE 2. EMSA analysis of transcription factor activation in response to inactivated or live rotavirus infection. A, Nuclear extracts were prepared from uninfected HT29 cells or from cells infected for 6 h with psoralen-inactivated rotavirus (i-RRV) or live rotavirus (RRV) matched by HA titer. Extracts were incubated with the NF-{kappa}B oligonucleotide, and the migration of NF-{kappa}B complexes is indicated. The specificity of the complexes is shown by inclusion of a 100-fold molar excess of unlabeled NF-{kappa}B oligonucleotide (RRV plus competitor). B, Extracts prepared in A were used in a DNA binding reaction with the GAS oligonucleotide. The migrations of STAT1 complexes are indicated, and the specificity is shown by inclusion of a 100-fold molar excess of unlabeled GAS oligonucleotide (RRV plus competitor).

 
Because nonreplicating viral particles activated transcription factors, we sought to determine whether viral capsid structures lacking genomic RNA would be capable of activating NF-{kappa}B. VLPs were assembled by recombinant expression of rotavirus proteins in combinations of either 2 and 6 (2/6), 2, 6, and 7 (2/6/7), or 2, 4, 6, and 7 (2/4/6/7). Because HA is a viral property conferred by the outer capsid protein VP4 and is representative of intact capsids, VLPs and RRV were standardized by HA titer. The HA titer of VLP 2/4/6/7 was 1/12 and was tested with an equivalent RRV sample of the same titer. The amount of recombinant VLPs was small compared with the previous studies of NF-{kappa}B activation by RRV. To assay NF-{kappa}B activity induced by this concentration of viral particles, a luciferase reporter construct with an NF-{kappa}B promoter was transfected into cells before viral particle adsorption. Fig. 3Go shows that 2/4/6/7 VLPs, but not 2/6 or 2/6/7, caused a 6-fold increase in luciferase activity in Vero cells. This indicated that NF-{kappa}B activation occurred following entry of complete viral particles into cells. To confirm the functional consequences to HT-29 cells of NF-{kappa}B activation by VLPs, supernatants from cultures exposed to VLPs were assayed for IL-8 by ELISA (R & D Systems). Fig. 4Go shows that 2/4/6/7, but not 2/6/7, VLPs caused increased IL-8 secretion. However, infectious RRV caused ~10 times greater IL-8 secretion.



View larger version (58K):
[in this window]
[in a new window]
  		FIGURE 3. NF-{kappa}B-luciferase activation by nonreplicating VLPs. VLPs produced from recombinant RRV proteins (containing either three or four capsid proteins) were standardized to HA titers of 1/12. Virus and VLPs were applied to 293 cell monolayers that had previously been transfected a NF-{kappa}B-reporter plasmid containing five NF-{kappa}B binding sites upstream of a luciferase gene. Cells were lysed and assayed for luciferase activity at 6 h p.i. . Results were standardized to the protein content of the cell monolayer and are reported as fold increases greater than control values. The results shown represent the average ± SEM of four monolayers (*, p < 0.05 vs control monolayer, by t test).

 


View larger version (50K):
[in this window]
[in a new window]
  		FIGURE 4. IL-8 Production induced by VLPs. VLPs were applied to HT-29 cell monolayers. Twenty-four hours p.i. supernatants were assayed by an enzyme immunoassay for IL-8. The result induced by VLP 2/4/6/7, analyzed by ANOVA and Fisher’s protected least significant difference post-hoc comparison, was significantly different from the medium control value (p < 0.008) and the VLP 2/6/7 value. The VLP 2/6/7 value was not significantly different from the medium control value (p = 0.14). The response of the RRV-infected cells was significantly (~10 times) greater than the VLP 2/4/6/7 response for similar HA-titrated samples (p < 0.0001 for comparison to all other groups).

 
These results indicate that rotavirus infection activates NF-{kappa}B and increases IL-8 secretion. To confirm that transcriptional regulation of IL-8 production by NF-{kappa}B activation occurs during exposure to both infectious and inactivated RRV, a plasmid containing the IL-8 promoter linked to a CAT reporter gene was employed. ECV 304 cells were transfected with the IL-8 promoter-CAT construct 24 h before RRV or PI-RRV adsorption. Increased CAT activity and increased IL-8 concentrations were identified in the culture supernatants following inoculation of cells with either RRV or PI-RRV. Untreated cells produced 1.0 ± 0.09 ng CAT/mg cell protein and 69 pg/ml IL-8. RRV infection resulted in 2.6 ± 0.15 ng CAT and 340 pg/ml IL-8, PI-RRV infection produced 4.05 ± 0.25 ng CAT and 680 pg/ml IL-8, while TNF-{alpha} yielded 5.6 ± 0.61 ng CAT and 890 pg/ml IL-8.

Effects of protein kinase inhibitors. Kinase-dependent signaling pathways regulate cellular control of the transcription of many cytokine genes. Therefore, we evaluated the effects of protein kinase inhibitors on the IL-8 secretion stimulated by rotavirus. Cells were pretreated with 100 µM genistein (a protein kinase inhibitor) or 25 µM H-7 (inhibits kinases protein kinases A, C, and G). Viability during inhibitor exposure was confirmed by trypan blue exclusion after 24 h. Baseline IL-8 secretion from control HT-29 cells produced a supernatant concentration of 81 ± 20 pg/ml. Noninhibited monolayers infected with RRV for 24 h (moi of 10) increased IL-8 to 9233 ± 1628 pg/ml. The IL-8 concentration was less markedly increased by infection in the presence of H-7 (100 µM) at 5273 ± 986 pg/ml (-43% compared with RRV alone) and genistein (100 µM) at 538 ± 219 pg/ml (-94%). Therefore, protein kinases are crucial to rotavirus-induced IL-8 secretion from intestinal epithelial cells.

Cytokine gene expression in response to rotavirus infection of epithelial cells. We previously reported that secretion of IL-8 (a C-X-C chemokine), but not TNF-{alpha}, IL-1{alpha}, IFN-{gamma}, or IL-6, was induced in HT-29 cells following rotavirus infection. Nevertheless, the activation of NF-{kappa}B and ISGF3 suggested that other cytokines may also be induced by rotavirus. Therefore, we analyzed RNA in rotavirus-infected HT-29 cells by RT-PCR for evidence of induction of selected cytokines, including chemokines of the C-X-C and C-C families. We also observed the effect of rotavirus on RNA specific for the inflammatory cytokines TNF-{alpha}, IL-1{alpha}, and IL-1ß, which have been shown to increase in HT-29 cells following exposure to invasive bacterial pathogens. Additionally, IFN-{alpha} and IFN-ß, which are widely produced by mammalian cells in response to viral infection, were evaluated.

Rotavirus infection (moi of 10) induced mRNA specific for several chemokines (Fig. 5Go). Consistent with NF-{kappa}B activation during infection, rotavirus induced cytokines with promoters containing NF-{kappa}B binding sites (40, 41, 42, 43, 44, 45, 46, 47). C-X-C chemokines with and without the ELR motif (IL-8 and GRO{gamma}, similar to murine MIP2) (48, 49, 50) were increased by infection. Rotavirus infection markedly increased the level of mRNA of the C-X-C chemokine IP-10 that is regulated synergistically by NF-{kappa}B and IFN (51). The C-C chemokines RANTES and MCP1 were also increased. In addition, GM-CSF was induced, but IL-1{alpha}/ß and TNF-{alpha} were not increased. Variable increases were noted in specific IFNs, as IFN-{alpha}, but not IFN-ß, increases were detected during infection. The peak stimulation occurred between 3 and 16 h p.i. and persisted beyond 24 h as previously reported for IL-8 (21).



View larger version (126K):
[in this window]
[in a new window]
  		FIGURE 5. Stimulation of cytokine mRNAs in HT-29 cells by RRV. HT-29 cell monolayers were exposed to RRV (moi of 10). RNA for the indicated cytokines was evaluated by RT-PCR before infection (0 h p.i.) and up to 16 h p.i. The internal standard was ß-actin, shown in row 1. All primers were used in preliminary amplification of DNA from human blood leukocytes stimulated with LPS and the phorbol ester PMA and yielded products of the predicted number of base pairs. These results are representative of five separate experiments on individually prepared monolayers.

 
Murine cytokine response during rotavirus infection and diarrhea

The murine model of rotavirus infection and diarrhea is well characterized and has been used for many studies of pathogenesis and immunity (52). Pups are susceptible to RRV infection and diarrhea up to 15 days of age. We sought to determine whether the HT-29 cell cytokine response to rotavirus infection was replicated in a murine model of rotavirus infection and diarrhea.

Chemokine responses to rotavirus infection of BALB/c, pfp-Rag-2, and Kit W/Wv mice. BALB/c pups were infected with RRV, and intestinal chemokine mRNA levels were determined. Significant increases were noted in mRNA specific for several chemokines (Fig. 6GoA). To investigate the possible increases in chemokine mRNA in mast cells or lymphocytes responding to rotavirus, we also assayed chemokine mRNA levels in KitW/KitW-v mice (mast cell deficient) and pfp-Rag-2 mice (T cell deficient). Whole intestinal tissues from mice were examined by semiquantitative PCR to evaluate rotavirus-stimulated chemokine mRNA changes (Fig. 6GoA). GAPDH mRNA PCR controls were used as internal standards for mRNA comparisons. Data were obtained from BALB/c, W/Wv, and control +/+ littermates and from pfp-Rag-2 mice. Responses were similar among all strains tested. Diarrhea and rotavirus-specific intestinal IgA responses of mice deficient in mast cells or lymphocytes also varied little from those of BALB/c controls. Differences were not seen in the diarrhea score after RRV infection, nor did mast cell deficiency decrease virus-specific mucosal IgA (Table IGo). As expected, intestinal IgA responses were absent in Rag-2 mice that lack T and B lymphocytes. Therefore, we concluded that murine chemokine responses to rotavirus infection are not dependent on lymphocytes or mast cells, nor is activation of these cells by epithelial cell-derived cytokines important in rotavirus diarrhea.



View larger version (74K):
[in this window]
[in a new window]
  		FIGURE 6. A, Comparison of intestinal cytokine responses among conventional, mast cell-deficient, and lymphocyte-deficient mice. BALB/C, WBB6F1/J-+/+, WBB6F1/J-W/Wv, and pfp-RAG-2 mice were inoculated with 107 ffu of RRV (or sham inoculated with medium alone) at 10 days of age. At designated times p.i. the small intestines were removed and prepared for RT-PCR. DNA samples were standardized by GAPDH DNA, which was ±10% by densitometry for all samples shown above. Baseline levels of chemokine RNA were established in pups sham-inoculated with medium alone. The response to rotavirus was measured at 6–72 h after virus introduction into the stomach by gavage. Data are shown for 6 h p.i., which was usually the peak of the mRNA levels, but levels greater than those in sham-inoculated mice persisted from 24–72 h. Delayed responses were not noted. These results are representative of three distinct experiments. For each RNA, sample material was pooled from three mice. B, Representative chemokine responses in p50-/- mice. p50-/- and BALB/c mice were inoculated with RRV and evaluated as described in A. Intestines were removed at 6 and 24 h p.i. Data shown are representative of findings in three distinct experiments. Each RNA sample is pooled from three mice.

 

View this table:
[in this window]
[in a new window]
  		Table I. Diarrhea score and intestinal IgA titer for BALB/c, WBB6F1/J-+/+, WBB6F1/J-W/WV, and RAG-2 mice following inoculation with rhesus rotavirus

 
Overall, the murine chemokine responses to rotavirus corresponded well to the responses noted in HT-29 cells. A few exceptions were noted. For instance, GM-CSF was induced in HT-29 cells, but not in mice. IL-1{alpha} was induced in W/Wv and pfp-Rag-2 mice, but not in BALB/c mice. Increased IFN-{alpha} occurred in cells, but IFN-ß was seen in mice. Notably, mRNA for the inflammatory cytokine TNF-{alpha} was not increased by rotavirus in cells or mice. While epithelial cells produce chemokines in response to rotavirus, contributions from macrophages, dendritic cells, or other stromal cells cannot be excluded.

Chemokine responses to rotavirus infection of p50-/- mice. The importance of NF-{kappa}B in modulation of the murine chemokine response to RRV infection is shown by comparison of representative chemokine mRNA in p50-/- mice and normal BALB/c control mice (Fig. 6GoB). p50-/- mice showed a marked diminution in the induction of mRNA specific for the C-C chemokine MCP-1 and the C-X-C chemokine IP-10 in comparison to BALB/c controls. It is interesting to note, however, that an increase in chemokine mRNA was still observed in p50 mutants. p50-/- mice may not be entirely deficient in NF-{kappa}B activity, as other NF-{kappa}B and Rel proteins may provide some redundancy. Combined mutants of Rel proteins p50 and p65 are severely abnormal and die prematurely, precluding studies of the response to rotavirus in these mice (53). In addition, other transcriptional regulatory elements such as the IFNs may be capable of inducing lesser chemokine responses in the absence of NF-{kappa}B activity. The p50-/- mice infected with RRV had diarrhea scores similar to those of infected C57 control mice (Table IGo). Further studies will be needed to determine whether chemokines (NF-{kappa}B regulated or others) may have a role in viral diarrhea.

The cellular localization in the mouse intestine of a representative NF-{kappa}B-regulated cytokine MIP-1 (47) was investigated by immunohistochemistry in tissue obtained from murine EC rotavirus-infected mice. Staining of MIP-1 was localized to the intestinal epithelial cells in infected intestine (Fig. 7Go).



View larger version (68K):
[in this window]
[in a new window]
  		FIGURE 7. MIP-1ß is expressed in murine intestinal epithelium during rotavirus infection. Sections obtained from intestinal preparations of rotavirus-infected murine intestine and additional sham-infected control were stained for the presence of MIP-1ß as described in Materials and Methods. The rotavirus strain used to infect these mice is the murine EC strain. This strain was used because it causes widespread epithelial infection in mice (see Fig. 9Go). These photographs are representative results (reproduced in three other mice) that demonstrate the location of MIP-1ß in epithelial cells.

 
Murine chemokine responses to rotavirus is age restricted. BALB/c pups were inoculated with medium or RRV (107 ffu) at 9, 11, 13, 15, 18, and 30 days of age and evaluated for the chemokine mRNA responses. Cytokine mRNA coding TNF-{alpha}, IL-1{alpha}, and IFN-ß was detectable in the small intestines of mice of 9–18 days of age, but did not vary in amount during rotavirus infection. However, mRNA for five of six chemokines shown in Fig. 8Go was present in small amounts or not at all in sham-inoculated pups and was increased by rotavirus infection. The chemokine increases tapered at 13 days of age and did not occur on day 15, the age at which mice no longer develop diarrhea with rotavirus infection.



View larger version (64K):
[in this window]
[in a new window]
  		FIGURE 8. Cytokine mRNA in RRV-infected murine intestine as a function of age. BALB/c mice of the indicated ages (9–30 days) were inoculated with RRV as described in Fig. 6GoA. Intestines were removed at 6 h p.i. Data shown are representative of findings in three distinct experiments. Each RNA sample is pooled from three mice.

 
We considered that reduced infection in older mice might be a possible explanation for age-related declines in murine responses to rotavirus. Previous studies showed that RRV infected fewer enterocytes than the homologous murine wild-type strains (54), However, neither strain caused diarrhea in mice older than 15 days of age. We reasoned that infection with a murine virus, unlike infection with the simian virus RRV, would cause readily detectable virus Ag by immunohistochemistry in large numbers of enterocytes and permit us to test the hypothesis that chemokine induction did not occur in response to rotavirus in adult mice despite abundant infection of enterocytes. We undertook experiments with the homologous murine rotavirus EC to determine whether increased chemokine mRNA levels occurred in adult mice infected with rotavirus. We confirmed that EC infection of pups (<15 days of age) showed chemokine responses similar in specificity and duration to those induced by RRV (data not shown). We then infected 5 wk adult BALB/c mice with EC rotavirus and assessed infection of intestinal epithelial cells in pups and adults by rotavirus immunohistology. Infected epithelial cells were readily detected at villus tips at all doses in both pups and adults, which showed similar numbers of infected epithelial cells (Fig. 9Go, A–D). Small intestines were removed at 0–72 h p.i. and were evaluated by RT-PCR for chemokine mRNA levels. EC-infected adult mice demonstrated no increases compared with sham-inoculated controls of chemokine mRNAs specific for MIP-1ß or IP-10 (Fig. 10Go). Increases in chemokine mRNA were evident, as before, in similarly infected 9-day-old pups. These data show that despite significant infection with homologous host murine rotavirus, adult mice do not up-regulate chemokine mRNA levels in response to infection.



View larger version (147K):
[in this window]
[in a new window]
  		FIGURE 9. Immunohistochemical assay of murine rotavirus EC (wild type) in the intestines of BALB/c pups (A and C) and adults (B and D). Inoculation in A and B was with DD50 x 105, while the dose in C and D was DD50 x 103. Rotavirus Ag is readily observed in epithelial cells at villus tips in similar frequency in pups and adults 24 h p.i. DD50, virus dose causing diarrhea in 50% of mice.

 


View larger version (65K):
[in this window]
[in a new window]
  		FIGURE 10. Representative chemokine responses from young and mature RRV-infected mice. BALB/c mice of either 9 days or 5 wk of age were infected with murine rotavirus strain EC at a dose of DD50 x 105. Intestines were removed and prepared for RT-PCR at 6 h p.i. Densitometry results were as follows: IP-10: sham pups, 1001; infected pups, 4366; sham adults, 0; infected adults, 0; MIP1-ß: sham pups, 257; infected pups,1635; sham adults, 0; infected adults, 0. Results at 24 h p.i. (data not shown) were similar. These results are representative of three distinct experiments.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Intestinal epithelial cells secrete cytokines with diverse cell chemotactic and activating properties in response to infection by bacteria (55, 56), protozoa (17, 57), or parasites (58). Chemokines are well known for the attraction and activation of polymorphonuclear and mononuclear inflammatory cells and lymphocytes. In addition, the epithelial cytokine response to infection may include growth factors to ameliorate tissue damage. For instance, GM-CSF has mitogenic properties, and other cytokines, including IL-8, have been shown to promote cell proliferation and epithelial repair. Some cytokines, such as IL-1, IL-10, IFNs, and others, have previously been shown to regulate intestinal transport and permeability (14, 59, 60). Receptors for several cytokines, including chemokines, have now been shown on intestinal epithelial cells, revealing the potential for autocrine signaling among infected epithelial cells (16, 61). It is apparent that epithelial cytokines constitute an important element of the host response to viral infection of the intestinal epithelium.

Rotavirus causes acute gastroenteritis while infecting exclusively epithelial cells at the apex of mucosal villi. The histology of this self-limited disease in mice and humans shows relatively mild changes of enterocyte vacuolation, villous blunting, and predominantly mononuclear cell infiltrates of the lamina propria (52). It has recently become clear that enterocytes neither cease protein synthesis nor rapidly lyse following infection (2, 10, 21, 62). It is possible that rotavirus activation of NF-{kappa}B even prolongs cell viability by suppressing the apoptosis normally exhibited by terminally differentiated villus tip enterocytes (63, 64). Prolonged viability of rotavirus-infected enterocytes provides an opportunity for these cells to secrete cytokines important in immune regulation or pathogenesis.

We previously discovered that rotavirus increased IL-8 mRNA levels and caused increased IL-8 secretion from HT-29 cells (21). This finding was recently confirmed by others and expanded to include two other chemokines, RANTES and GRO-{alpha} (62). We now show that the rotavirus-increased IL-8 production is dependent on protein kinase activity and regulated transcriptionally by NF-{kappa}B. We also obtained evidence that rotavirus activates IFN transcriptional elements. Activation of these, and perhaps other, transcriptional regulators leads to increases in mRNA of a wide range of cytokines, especially chemokines. These findings establish intestinal epithelial cells as a primary source of chemokines in the early host response to rotavirus infection.

We showed in several different experiments that viral replication is not required to initiate a cellular response to infection. Genetically inactivated RRV was capable of activating NF-{kappa}B and STAT1 (Fig. 2Go). VLPs containing four viral proteins VP 2/6/4/7 activated NF-{kappa}B and caused IL-8 secretion (Figs. 3Go and 4Go). However, infectious RRV more effectively stimulated NF-{kappa}B activation and IL-8 secretion. These findings suggest that rotavirus may activate NF-{kappa}B by diverse mechanisms.

It was recently reported that simian SA11 rotavirus inactivated by the psoralen-UV method did not appear to induce IL-8 secretion from HT-29 cells, nor did VLPs assembled from baculovirus-expressed SA11 VP 2, 6, 4, and 7 (62). It was thus concluded that viral replication is required for chemokine stimulation. It remains a possibility that the use of different viral strains (RRV vs SA-11) may explain the differences between these data. However, several caveats should be noted. Particle integrity of SA-11 was not documented by HA after psoralen-UV inactivation, and only a single relatively small estimated dose of inactivated virus was used (moi of 1), a limiting condition given the dose dependence of the IL-8 response. Furthermore, VLPs in the prior study were not activated with trypsin, which is required to permit penetration of cell membranes following cell attachment (65). Therefore, this finding confirms a prior report that membrane attachment in the absence of penetration is insufficient to induce chemokine production (21). Our present results indicate that viral particles capable of membrane penetration activate NF-{kappa}B and cause IL-8 secretion in the absence of viral replication or the introduction of viral RNA into cells.

Replicating virus generates abundant dsRNA, a potent inducer of protein kinase R. Protein kinase R has been reported to directly phosphorylate I{kappa}B, which results in activation of NF-{kappa}B (66, 67). As noted above, we demonstrated that rotavirus infection activates IFN transcriptional elements, which are also linked to protein kinase R. Following IFN-{alpha}/ß stimulation, two specific transcription factors are activated, ISGF3, which binds the ISRE, and STAT1 which binds GAS (38). Activation of both of these factors can be identified following rotavirus infection (Fig. 1Go for ISGF3; Fig. 2Go for STAT1). It is interesting to note that the kinetics of the appearance of these factors lagged behind the appearance of nuclear NF-{kappa}B (<2 h p.i. ), suggesting the possibility that IFN signaling is either an independent or secondary phenomenon. Cooperation between these two transcriptional regulatory systems is well documented, especially during viral infections (68, 69, 70, 71).

Another possible sustained stimulus of NF-{kappa}B during rotavirus infection may emanate from internal cell stress, which has been reported to occur during distention of the endoplasmic reticulum in which rotavirus maturation occurs (72, 73). This may be considerable during viral replication due to the accumulation and assembly of newly synthesized viral proteins. The appearance of activated transcription factors in vitro suggests, but does not prove, the function of these factors in vivo.

Infection of conventional 10-day-old BALB/c mice with RRV resulted in increased mRNA for C-C and C-X-C chemokines and IFN-ß in intestinal tissue 6 h p.i. . The pattern of chemokine responses was analogous to that in HT-29 cells. Notably, TNF-{alpha} induction was not increased by rotavirus infection either in vitro or in vivo. There were some differences in the chemokine responses between the human cells and murine intestine, but overall these findings are strikingly similar. The congruence of these two models favors the relevance of epithelial chemokines in response to intestinal infection. However, cytokines other than those listed were not measured, and many differences not here noted may exist.

Increases in IFN-ß and IL-1{alpha} mRNA were detected after rotavirus infection in W/Wv and pfp-RAG-2 mice, but not in HT-29 cells, Perhaps cells absent from cultures may participate in the response to infection in murine intestine. For instance, activated mast cells contribute to diarrhea caused by Clostridium difficile toxin (74), which also stimulates chemokine secretion from intestinal epithelial cells (75). Lymphocytes, which chemokines attract and activate, may also contribute to diarrheal disease (15). However, mast cell- or T cell-deficient mice had diarrheal responses to rotavirus infection similar to those of BALB/c mice (Table IGo), and patterns of chemokine mRNAs during infection were similar in the three mouse strains (Fig. 6GoA). Therefore, we concluded that epithelial cytokine activation of neither mast cells nor lymphocytes is important in rotavirus diarrhea. However, we cannot exclude the possibility that production of cytokines from macrophages, dendritic cells, or other stromal cells occurs during rotavirus infection.

The importance of NF-{kappa}B as a regulator of chemokine mRNA in vivo during rotavirus infection was demonstrated in the p50-/- mouse model. NF-{kappa}B consists of heterodimers constructed from a group of NF-{kappa}B/Rel proteins including NF-{kappa}B1 (p50 and the precursor p105), NF-{kappa}B2 (p52 and the precursor p100), RelA, RelB, and c-Rel. These subunits are associated with I{kappa}B inhibitors in the cytoplasm and permit nuclear translocation of NF-{kappa}B after I{kappa}B phosphorylation. The importance of NF-{kappa}B during rotavirus infection in vivo is apparent due to the markedly diminished MCP-1 and IP-10 mRNA increases that rotavirus induced in the p50-/- mouse (Fig. 6GoB). It is interesting to note that a small increase in chemokine mRNA is seen in p50-/- mutants (Fig. 6GoB), possibly due to the activation of other Rel proteins or other transcriptional regulatory elements such as the IFNs. It is possible that redundant signaling mediates a chemokine response to promote diarrhea. Conclusions concerning the effects on epithelial transport and barrier functions of the chemokine response to viral pathogens should not be drawn from this mutant alone. Combined mutants of Rel proteins p50 and p65 are severely abnormal and die prematurely, precluding studies of the response to rotavirus in these mice (53). Other alternatives will be needed to definitively address these issues.

Epithelial cells are a likely source of the rapid increase in chemokine mRNA that follows rotavirus infection in vivo because of the similarity of the chemokine responses of epithelial cell lines and murine intestine and the independence of the chemokine response from modulation by T cells or mast cells. We confirmed by immunohistochemistry that a representative chemokine, MIP-1ß, was produced in murine epithelial cells following rotavirus infection (Fig. 7Go). This confirms that the RNA transcripts that we monitored are actually translated into proteins in intestinal epithelial cells.

We demonstrated that several chemokines were induced by rotavirus infection as well as other cytokines, such as IFN and GM-CSF. The pattern is distinct from invasive bacterial diarrheal pathogens, as we did not identify stimulation of TNF-{alpha} or IL-1{alpha}. Initially, investigators viewed chemokines narrowly as chemoattractants for polymorphonuclear leukocytes, which are not prominent in the host response to rotavirus infection, although some influx of polymorphonuclear leukocytes probably occurs in humans, lambs, and birds if not in mice (76, 77, 78, 79). It is now recognized that chemokines are elements of a complex cell-signaling network that regulates kinesis, activation, and proliferation of a large number of cells, especially those with important roles in Ag-specific immunity, such as B lymphocytes and dendritic cells (80, 81, 82). Other cytokines, noted above, also regulate intestinal transport and permeability. It is likely that many other cytokines not measured in this work are also altered during rotavirus infection. Therefore, studies are needed to evaluate the potential roles of cytokines in the pathogenesis of rotavirus disease, the regulation of the host immune response, and perhaps the maintenance of epithelial integrity during viral intestinal infection.


   Footnotes
 
1 This work was supported by Department of Veterans Affairs medical research funds. Back

2 Current address: Pontificia Universidad Javeriana, Instituto de Genetica Humana, Cra. 4 #40-62, Santafe de Bogota, Colombia. Back

3 Address correspondence and reprint requests to Dr. Robert D. Shaw, Research Service, Building 62 (151), Northport Veterans Affairs Medical Center, Northport, NY 11768. E-mail address: Back

4 Abbreviations used in this paper: MCP-1, monocyte chemotactic protein-1; RRV, rhesus rotavirus; moi, multiplicity of infection; VLP, virus-like particle; HA, hemagglutinin; GAS, IFN-{gamma}-activated site; CAT, chloramphenicol acetyltransferase; DEPC, diethylpyrocarbonate; ISRE, IFN-stimulated response element; ISGF3, IFN-stimulated gene factor-3; ffu, focus-forming units; PI-RRV, psoralen-UV-inactivated rotavirus; p.i., postinfection; IP-10, IFN-stimulated protein-10. Back

Received for publication January 11, 1999. Accepted for publication July 23, 1999.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Keljo, D. J., R. H. Squires. 1998. Anatomy and anomalies of the small and large intestine. ed. In Gastrointestinal and Liver Disease Vol. 2:1419. W. B. Saunders Co., Philadelphia.
  2. Jourdan, N., M. Maurice, D. Delautier, A. M. Quero, A. L. Servin, G. Trugnan. 1997. Rotavirus is released from the apical surface of cultured human intestinal cells through nonconventional vesicular transport that bypasses the Golgi apparatus. J. Virol. 71:8268.[Abstract]
  3. McCormick, B. A., S. P. Colgan, C. Delp-Archer, S. I. Miller, J. L. Madara. 1993. Salmonella typhimurium attachment to human intestinal epithelial monolayers: transcellular signalling to subepithelial neutrophils. J. Cell Biol. 123:895.[Abstract/Free Full Text]
  4. Offit, P. A., K. I. Dudzik. 1989. Rotavirus-specific cytotoxic T lymphocytes appear at the intestinal mucosal surface after rotavirus infection. J. Virol. 63:3507.[Abstract/Free Full Text]
  5. Rott, L. S., J. R. Rose, D. Bass, M. B. Williams, H. B. Greenberg, E. C. Butcher. 1997. Expression of mucosal homing receptor {alpha}4ß7 by circulating CD4+ cells with memory for intestinal rotavirus. J. Clin. Invest. 100:1204.[Medline]
  6. McNeal, M. M., M. N. Rae, R. L. Ward. 1997. Evidence that resolution of rotavirus infection in mice is due to both CD4 and CD8 cell-dependent activities. J. Virol. 71:8735.[Abstract]
  7. Offit, P. A., E. J. Hoffenberg, E. S. Pia, P. A. Panackal, N. L. Hill. 1992. Rotavirus-specific helper T cell responses in newborns, infants, children, and adults. J. Infect. Dis. 165:1107.[Medline]
  8. Merchant, A. A., W. S. Groene, E. H. Cheng, R. D. Shaw. 1991. Murine intestinal antibody response to heterologous rotavirus infection. J. Clin. Microbiol. 29:1693.[Abstract/Free Full Text]
  9. Coffin, S. E., M. Klinek, P. A. Offit. 1995. Induction of virus-specific antibody production by lamina propria lymphocytes following intramuscular inoculation with rotavirus. J. Infect. Dis. 172:874.[Medline]
  10. Jourdan, N., J. P. Brunet, C. Sapin, A. Blais, J. Cotte-Laffitte, F. Forestier, A. M. Quero, G. Trugnan, A. L. Servin. 1998. Rotavirus infection reduces sucrase-isomaltase expression in human intestinal epithelial cells by perturbing protein targeting and organization of microvillar cytoskeleton. J. Virol. 72:7228.[Abstract/Free Full Text]
  11. Ball, J. M., P. Tian, C. Q. Zeng, A. P. Morris, M. K. Estes. 1996. Age-dependent diarrhea induced by a rotaviral nonstructural glycoprotein. Science 272:101.[Abstract]
  12. Osborne, M. P., S. J. Haddon, K. J. Worton, A. J. Spencer, W. G. Starkey, D. Thornber, J. Stephen. 1991. Rotavirus-induced changes in the microcirculation of intestinal villi of neonatal mice in relation to the induction and persistence of diarrhea. J. Pediatr. Gastroenterol. Nutr. 12:111.[Medline]
  13. Deem, R. L., F. Shanahan, S. R. Targan. 1991. Triggered human mucosal T cells release tumour necrosis factor-alpha and interferon-{gamma} which kill human colonic epithelial cells. Clin. Exp. Immunol. 83:79.[Medline]
  14. Madara, J. L., J. Stafford. 1989. Interferon-{gamma} directly affects barrier function of cultured intestinal epithelial monolayers. J. Clin. Invest. 83:724.
  15. McKay, D. M., M. A. Benjamin, J. Lu. 1998. CD4+ T cells mediate superantigen-induced abnormalities in murine jejunal ion transport. Am. J. Physiol. 275:G29.[Abstract/Free Full Text]
  16. Panja, A., S. Goldberg, L. Eckmann, P. Krishen, L. Mayer. 1998. The regulation and functional consequence of proinflammatory cytokine binding on human intestinal epithelial cells. J. Immunol. 161:3675.[Abstract/Free Full Text]
  17. Rasmussen, S. J., L. Eckmann, A. J. Quayle, L. Shen, Y. X. Zhang, D. J. Anderson, J. Fierer, R. S. Stephens, M. F. Kagnoff. 1997. Secretion of proinflammatory cytokines by epithelial cells in response to Chlamydia infection suggests a central role for epithelial cells in chlamydial pathogenesis. J. Clin. Invest. 99:77.[Medline]
  18. Bonecchi, R., G. Bianchi, P. P. Bordignon, D. D’Ambrosio, R. Lang, A. Borsatti, S. Sozzani, P. Allavena, P. A. Gray, A. Mantovani, et al 1998. Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (Th1s) and Th2s. J. Exp. Med. 187:129.[Abstract/Free Full Text]
  19. Steiner, T. S., A. A. Lima, J. P. Nataro, R. L. Guerrant. 1998. Enteroaggregative Escherichia coli produce intestinal inflammation and growth impairment and cause interleukin-8 release from intestinal epithelial cells. J. Infect. Dis. 177:88.[Medline]
  20. Z’Graggen, K., A. Walz, L. Mazzucchelli, R. M. Strieter, C. Mueller. 1997. The C-X-C chemokine ENA-78 is preferentially expressed in intestinal epithelium in inflammatory bowel disease. Gastroenterology 113:808.[Medline]
  21. Sheth, R., J. Anderson, T. Sato, B. Oh, S. J. Hempson, E. Rollo, E. R. Mackow, R. D. Shaw. 1996. Rotavirus stimulates IL-8 secretion from cultured epithelial cells. Virology 221:251.[Medline]
  22. Mukaida, N., Y. Mahe, K. Matsushima. 1990. Cooperative interaction of nuclear factor-{kappa}B- and cis-regulatory enhancer binding protein-like factor binding elements in activating the interleukin-8 gene by pro-inflammatory cytokines. J. Biol. Chem. 265:21128.[Abstract/Free Full Text]
  23. Mastronarde, J. G., B. He, M. M. Monick, N. Mukaida, K. Matsushima, G. W. Hunninghake. 1996. Induction of interleukin (IL)-8 gene expression by respiratory syncytial virus involves activation of nuclear factor (NF)-{kappa} B and NF-IL-6. J. Infect. Dis. 174:262.[Medline]
  24. Jobin, C., L. Holt, C. A. Bradham, K. Streetz, D. A. Brenner, R. B. Sartor. 1999. TNF receptor-associated factor-2 is involved in both IL-1ß and TNF-{alpha} signaling cascades leading to NF-{kappa}B activation and IL-8 expression in human intestinal epithelial cells. J. Immunol. 162:4447.[Abstract/Free Full Text]
  25. McKimm, B. J., I.