Virus Research 181 (2014) 27–34
Contents lists available at ScienceDirect
Virus Research journal homepage: www.elsevier.com/locate/virusres
Characterization of the chemokine response of RAW264.7 cells to infection by murine norovirus Emily Waugh a , Augustine Chen b , Margaret A. Baird a,1 , Chris M. Brown b , Vernon K. Ward a,∗ a b
Department of Microbiology and Immunology, School of Medical Sciences, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand Department of Biochemistry, School of Medical Sciences, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand
a r t i c l e
i n f o
Article history: Received 12 September 2013 Received in revised form 13 December 2013 Accepted 16 December 2013 Available online 24 December 2013 Keywords: Norovirus Calicivirus Chemokines Immunity
a b s t r a c t Noroviruses are an emerging threat to public health, causing large health and economic costs, including at least 200,000 deaths annually. The inability to replicate in cell culture or small animal models has limited the understanding of the interaction between human noroviruses and their hosts. However, an alternative strategy to gain insights into norovirus pathogenesis is to study murine norovirus (MNV-1) that replicates in cultured macrophages. While the innate immune response is central to the resolution of norovirus disease, the adaptive immune response is required for viral clearance. The specific responses of macrophages and dendritic cells to infection drive the adaptive immune response, with chemokines playing an important role. In this study, we have conducted microarray analysis of RAW264.7 macrophages infected with MNV-1 and examined the changes in chemokine transcriptional expression during infection. While the majority of chemokines showed no change, there was specific up-regulation in chemokines reflective of a bias toward a Th1 response, specifically CCL2, CCL3, CCL4, CCL5, CXCL2, CXCL10 and CXCL11. These changes in gene expression were reflected in protein levels as determined by ELISA assay. This virus-induced chemokine response will affect the resolution of infection and may limit the humoral response to norovirus infection. © 2014 Published by Elsevier B.V.
1. Introduction Noroviruses are non-enveloped, positive-sense, single-stranded RNA viruses belonging to the Caliciviridae family. In general, calicivirus genomes comprise three orfs, with orf1 encoding the non-structural polyprotein while orf2 and orf3 encode for the structural capsid proteins (Sosnovtsev et al., 2006), while an additional fourth open reading frame has been discovered in murine norovirus (McFadden et al., 2011). Human norovirus (HuNoV) is the most common cause of non-bacterial gastroenteritis worldwide (Perry et al., 2009) and in developing countries, norovirus infection can be attributed to the death of 200,000 children five years old or younger (Patel et al., 2009). Symptoms of a norovirus infection commonly include vomiting and diarrhea as well as abdominal cramps and
Abbreviations: HuNoV, human norovirus; MNV, murine norovirus; qPCR, quantitative real time PCR; h.p.i., hours post-infection; GO, gene ontology; TLR, Toll-like receptor; GEO, gene expression omnibus; NCBI, National Center for Biotechnology Information; RMA, robust multi-array average. ∗ Corresponding author. Tel.: +64 3 479 9028; fax: +64 3 479 8540. E-mail address: [email protected] (V.K. Ward). 1 Present address: Department of Pathology, University of Otago, Dunedin 9054, New Zealand. 0168-1702/$ – see front matter © 2014 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.virusres.2013.12.025
fever (Estes et al., 2006) and due to the self-limiting nature of the virus, symptoms typically resolve within 24–48 h in immunocompetent hosts (Mumphrey et al., 2007). Despite the large burden of morbidity and mortality caused by norovirus infections, there is currently no vaccine or anti-viral treatment available. This is in part because the study of HuNoV has been limited by the lack of a small animal model or cell culture system that supports HuNoV replication (Duizer et al., 2004). In 2003, a norovirus was identified that infected mice lacking signal transducer and activator of transcription 1 (Karst et al., 2003). This murine norovirus (MNV-1) has a tropism for macrophages and dendritic cells (Wobus et al., 2004), thus it can be effectively propagated in cell culture and is now routinely used as a model system to study HuNoV infection (Belliot et al., 2008; Wobus et al., 2006). The innate immune response to MNV-1 has been well investigated, as it is the predominant response to a norovirus infection (Changotra et al., 2009; Karst et al., 2003). The innate response to MNV-1 is initiated through recognition of the viral double-stranded RNA by retinoic acid-inducible gene 1 (RIG-I)-like receptors, such as melanoma differentiation-associated protein 5 (MDA5) (McCartney et al., 2008). This induction initiates several signaling pathways within the cell that leads to the production of type 1 interferons (IFN-␣ and IFN-) as well as other pro-inflammatory
28
E. Waugh et al. / Virus Research 181 (2014) 27–34
cytokines (Platanias, 2005). This process causes the activation of an antiviral state and control of the virus infection (Loo et al., 2008). In addition, studies using both HuNoV and MNV-1 have demonstrated that the acquired immune response is vital for clearance of a norovirus infection (Chachu et al., 2008) and prolonged HuNoV infection is observed in immunocompromised individuals (Wingfield et al., 2010). Both the humoral and cell-mediated acquired immune response play a role in eliminating norovirus infections (LoBue et al., 2010). The production of antibody and the generation of memory B cells are important in protecting the host, with poor antibody responses allowing reinfection of the host (Bok et al., 2011). Cytotoxic T cells induced and activated by cytokines, kill cells infected with norovirus. Studies have shown that the cytokine profile of norovirus-infected cells includes the production of large amounts of gamma interferon (IFN-␥) and interleukin 2 (LoBue et al., 2010; Souza et al., 2007), which are characteristic of a Th1 response (Lindesmith et al., 2005). Chemokines are an integral and critical step in the immune response to virus infections (Sillanpaa et al., 2008; Zeremski et al., 2007). Few investigations have been carried out on the chemokine response to MNV-1. Chemokines are a family of small (∼8–12 kDa) secreted molecules that exert their effects by binding to seven-transmembrane G-protein coupled receptors to establish a chemokine gradient, allowing directional migration of cells expressing the corresponding chemokine receptor (Deshmane et al., 2009; Zhao et al., 2009). Members of the chemokine family mostly fall into two broad groups; the CC chemokines and the CXC chemokines. Other groups include the C and CX3C chemokines. In addition, chemokines can be divided into constitutive and inflammatory chemokines (Lateef et al., 2009; Sallusto and Baggiolini, 2008). Members of the CC class ( chemokines) recruit the majority of leukocyte cell types (Cole et al., 1998), while the CXC chemokines (␣ chemokines) primarily recruit neutrophils and T lymphocytes (Cole et al., 1998; Mortier et al., 2008). Viruses are known to interact with and manipulate the host chemokine system (Lateef et al., 2010; Luttichau, 2010; Sillanpaa et al., 2008; Webb and Alcami, 2005). However, the ability of MNV-1 to interact with the chemokine network and the role of chemokines important in both the innate and the acquired immune response to MNV-1 has not been studied. This study characterizes the host chemokine response to MNV-1 infection of RAW 264.7 macrophages.
from cells at 9 or 18 h.p.i. using the Qiagen RNeasy mini kit as per manufacturers instructions. Transcriptomic analysis was conducted in triplicate using a GeneChip® Mouse Genome 430 2.0 Array (Affymetrix) at the Otago Genomics Facility. The complete data set obtained from the microarray analysis was submitted to the GEO database (NCBI), accession number GSE50093. 2.3. Expression analysis by microarray Raw microarray intensities from CEL files were normalized using RMA and quantile normalization using the ExpressionFileCreator in Genepattern (Reich et al., 2006) and preprocessed using PreprocessDataset and comparison was done using ComparativeMarkerSelection. Ratios of expression at 9 or 12 h.p.i. were generated. Ratios were not generated for genes with marginally detectable expression of 50 in the mock-infected control. Differentially expressed gene sets were analyzed using DAVID Bioinformatics Resources 6.7 (National Institute of Allergy and Infectious Diseases, NIH) (Huang da et al., 2009a, 2009b). Affymetrix I.D.s of genes altered 2-fold or more were used for functional annotation clustering. This was performed to identify groups enriched in genes associated with similar biological functions. Mouse Genome Informatics (Eppig et al., 2012) was used to ascertain the identity and function of the genes. Raw microarray data (Bok et al., 2009) was accessed through the GEO database at the National Center for Biotechnology Information (accession number GSE12518). This data was analyzed in the same way as the CEL files generated in this study. 2.4. Virus purification
2. Materials and methods
RAW264.7 cells were seeded at 2.8 × 107 in 175 cm2 flasks. The following day, confluent monolayers were infected with MNV-1 at an m.o.i. of 1.0, as previously described with the exception that after absorption, the inoculum was removed and the cells were washed with fresh medium before adding 60 ml of medium and incubating for 28 h. The infected cells were frozen at −20 ◦ C then thawed at 37 ◦ C before clarification of the inoculum by a low speed centrifugation at 1000 × g for 20 min. This unpurified inoculum was purified through a 30% (w/v) sucrose cushion at 112,700 × g for 4 h at room temperature. The pellet was re-suspended in 1 ml Dulbecco’s phosphate buffered saline overnight at 4 ◦ C, then filter sterilized through a 0.45 m filter.
2.1. Cells and virus
2.5. Detection of IFN-ˇ
RAW264.7 cells (obtained from ATCC) were maintained in a 75 cm2 flask at 5 × 106 cells in Dulbecco’s modified Eagle’s medium supplemented with 5% fetal calf serum, 100 U/ml penicillin and 0.1 mg/ml streptomycin. During infection cells were incubated in the same medium without antibiotics. A working stock of MNV-1 was prepared using a previously described reverse genetics method (Ward et al., 2007). Viral titers were determined by crystal violet plaque assay in RAW264.7 cells (Gonzalez-Hernandez et al., 2012).
Levels of IFN- in virus inoculum were determined by ELISA using a mouse IFN- ELISA kit (PBL Interferon Source, Piscataway, NJ, USA). Unpurified MNV-1 and purified MNV-1 inocula dilutions were assayed in triplicate.
2.2. Microarray A 6-well plate was seeded with RAW264.7 cells at 8 × 105 cells per well. The next day confluent monolayers were infected with unpurified MNV-1 at an m.o.i. of 1.0. Medium was removed from wells, replaced with virus inoculum and incubated at 37 ◦ C + 5% CO2 for 1 h, with rocking. Following incubation, 2 ml of fresh medium was added to each well and plates further incubated at 37 ◦ C + 5% CO2 . Mock-infected cells were treated the same way with culture medium added instead of virus inoculum. RNA was extracted
2.6. Quantitative real-time PCR Microarray results were validated by qPCR for selected genes that showed altered expression. Six-well plates were seeded with RAW264.7 cells at 8 × 105 cells per well. The next day 80% confluent monolayers were infected with purified MNV-1 at an m.o.i. of 1.0, or mock-infected. Total RNA was isolated from cells at 12, 15 and 18 h.p.i. using TRIzol reagent (Invitrogen, Auckland, NZ) and RNA purification was performed using the PureLink RNA mini kit (Invitrogen). Isolated RNA was subjected to DNase treatment and cDNA was generated from DNase-treated total RNA via random hexamer reverse transcription using the High Fidelity Transcriptor cDNA synthesis kit (Roche, Auckland, NZ). Gene-specific primers for qPCR were selected using the Universal Probe Library from
E. Waugh et al. / Virus Research 181 (2014) 27–34
29
Table 1 Primer Sequences. Chemokine
Forward primer sequence (5 –3 )
Reverse primer sequence (5 –3 )
CCL2 CCL3 CCL4 CCL5 CCL9 CXCL2 CXCL10 CXCL11 CXCL16
AGGTGTCCCAAAGAAGCTGTA TGCCCTTGCTGTTCTTCTCT ATGAGACCAGCAGTCTTTGCTCCA TCGTGCCCACGTCAAGGAGTATTT TACCAGGCCGGGCATCATCTTTAT ACATCCCACCCACACAGTGAAAGA GCTGCCGTCATTTTCTGC AGATCCAAGCAAGCTCGCCTCATA TGAAAGCATCTTGGAGCCAGAGGA
ATGTCTGGACCCATTCTTCT GTGGAATCTTCCGGCTGTAG GCTGCTCAGTTCAACTCCAAGTCA TCTTCTCTGGGTTGGCACACACTT TGGCAGTTCACACCCTTCTCTTCA TCCTTCCATGAAAGCCATCCGACT TCTCACTGGCCCGTCATC ATGTTCCAAGACAGCAGAGGGTCA TGGTTCCCGAAGAATGACTCGGTT
Roche (Table 1). Primer design requirements included a DNA melting temperature between 58 and 60 ◦ C, a guanine–cytosine content of 40–60% and the primers were across intron spanning regions to eliminate genomic DNA contamination. Where possible, primers validated in published articles were used (Cantini et al., 2012; Griener et al., 2011; Hochrainer et al., 2013; Lu et al., 2013; Ohoka et al., 2011). A serial dilution of cDNA was used for each primer set to determine the effective range of template concentration that gave a consistent and optimal efficiency. qPCR was performed on a LightCycler® 480 platform using 2× SYBR green master mix (Roche). Briefly, 3 l of cDNA was added to 0.5 M each of the corresponding forward and reverse primer and 5 l of 2× SYBR green master mix. Cycling conditions were 95 ◦ C for 5 min followed by 50 cycles at 95 ◦ C, 5 s; 58 ◦ C, 5 s; 72 ◦ C, 8 s; then 65 ◦ C for 5 min. Duplicate samples were set up for each sample and all samples were normalized to -actin in the same sample to correct for efficiency of reverse transcription. The levels of genes of interest were calculated using the efficiency method for relative quantification. The efficiency method employs relative standards prepared using sample cDNA and thus normalizes the amplification efficiency of target and reference genes as well as run-to-run differences. Specificity of the qPCR was confirmed by the melting curve of amplified products. In addition, each qPCR experiment included a standard curve of each primer set to determine primer efficiency. This efficiency value (between 1.8 and 2.0) was used in the normalization and fold change calculation, using Roche LightCycler480 software (version 1.5.0.39).
up-regulated 2-fold or more during MNV-1 infection and these genes were further clustered into 187 Gene Ontology annotation clusters using DAVID functional annotation tools (Huang da et al., 2009b). The most noteworthy clusters included toll-like signaling (e.g. toll-like receptor (TLR) signaling P = 9.1 × 10−11 ) programmed cell death (e.g. apoptosis P = 1.3 × 10−5 ), as well as immune molecules (P = 1.5 × 10−4 ) (Fig. 1). The most significant
2.7. Chemokine quantification Confluent monolayers of RAW264.7 cells were mock-infected or infected with purified MNV-1 at an m.o.i. of 1.0. Cell culture supernatants were collected and DuoSet ELISA Development kits (R&D Systems, Minneapolis, MN, USA) were used to determine the amount of CCL5, CXCL2, CXCL10 and CXCL11 produced during MNV-1 infection. OptEIA Mouse ELISA set (BD Biosciences Pharmingen, New Jersey, USA) was used to quantify the amount of CCL2 produced during infection. 3. Results and discussion 3.1. Overall changes in cellular gene expression during MNV-1 infection To measure changes in gene expression in macrophages (RAW264.7) in response to infection by MNV-1, cells were mockinfected or infected with virus and mRNA levels measured by microarray at 9 and 18 h post infection (h.p.i). Duplicate microarrays using two biological replicates on an Affymetrix Mouse 430 2.0 array, which represented over 34,000 genes, were carried out. The results showed that the expression of approximately 33,000 genes (97%) remained relatively constant during MNV-1 infection. This analysis identified 766 genes whose expression was
Fig. 1. Heat map of selected genes up-regulated 3-fold or more during MNV-1 infection. RAW264.7 cells were infected with MNV-1 at m.o.i. of 1. Total cellular RNA was harvested at 9 and 18 h.p.i. and analyzed by Affymetrix Mouse 430 2.0 microarray. Heat map image of fold changes was generated using Rstudio software (Version 0.96.330, ©Rstudio, Inc.). Affymetrix IDs and gene names are indicated. (A) Genes associated with Toll-like receptor signaling. (B) Genes associated with apoptosis. (C). Genes associated with the immune response.
30
E. Waugh et al. / Virus Research 181 (2014) 27–34
Fig. 3. Up-regulation of chemokine mRNA expression in MNV-1 infected RAW264.7 cells. Cells were infected with MNV-1 at an m.o.i. of 1.0, total cellular RNA was isolated at 18 h.p.i. and cDNA generated. qPCR analysis was carried out using SYBR green signal transmission on a Roche LightCycler 480 system. Chemokine levels were normalized to -actin. Data shows the mean fold increases over mock-infected cells of three experiments performed in duplicate for all chemokines except CXCL10, which is one representative experiment.
Fig. 2. Heat map of selected down-regulated genes during MNV-1 infection. RAW264.7 cells were infected with MNV-1 at m.o.i. of 1. Total cellular RNA was harvested at 9 and 18 h.p.i. and analyzed by Affymetrix Mouse 430 2.0 microarray. Heat map image of fold changes was generated using Rstudio software (Version 0.96.330, ©Rstudio, Inc.). Affymetrix IDs and gene names are indicated. (A) Genes associated with cell cycle. (B) Genes associated with programmed cell death. (C) Genes associated with selected biosynthetic pathways. (D) Genes associated with nucleotide metabolism.
clusters of down-regulated genes (779) included those involved in cell cycle regulation (e.g. cell cycle, P = 2.9 × 10−5 ), programmed cell death (e.g. apoptosis, P = 1.0 × 10−2 ) and nucleotide binding (e.g. ATP binding, P = 5.4 × 10−2 ) (Fig. 2). These changes indicate that MNV-1 may be modulating the cell cycle, cell metabolism and/or regulating apoptosis. A similar microarray performed by Bok et al. (2009) also identified down-regulation in apoptosis in RAW264.7 cell during a MNV-1 infection showed that MNV-1 down-regulates survivin, a key regulator of apoptosis. 3.2. Genes associated with interferon stimulation and the immune response are significantly up-regulated by MNV-1 infection A large number of up-regulated genes clustered into immune response categories including cytokine-cytokine receptor signaling (P = 1.0 × 10−9 ), defense response (P = 4.0 × 10−11 ), inflammatory response (P = 1.9 × 10−4 ) and chemotaxis (P = 6.6 × 10−3 ) (Fig. 1 and Table 2). This result was not unexpected as it is known that norovirus infection induces an innate immune response as well as a Th1-type acquired response (Chachu et al., 2008; Lindesmith et al., 2005; Wobus et al., 2006). In addition, further analysis of the microarray performed by Bok et al. (2009) supported our findings. Of the 50 most up-regulated genes, 11 (22%) were associated with, or stimulated by type I interferon, suggesting that interferon present in the inoculate may be responsible for some of the up-regulation seen in the microarray data. Although IFN-
quantification revealed that the virus inoculum contained at least 1000 pg/ml of IFN-, the microarray performed using unpurified virus yielded similar results to the purified virus microarray data of Bok et al. (Supplementary data). Additionally, in all subsequent qPCR and ELISA analyses the viral inoculum was purified by ultracentrifugation through a 30% sucrose cushion to remove any residual interferon. The removal of immunostimulatory molecules such as interferon was confirmed by IFN ELISA, in which interferon levels were below the detection limit of the assay. 3.3. The host chemokine response is up-regulated during MNV-1 infection Analysis of the microarray data revealed that a significant number of chemokine (P = 1.9 × 10−3 ) and chemokine receptor genes (P = 1.6 × 10−3 ) were up-regulated by MNV-1 infection. Of the 35 chemokine probes on the array, seven (four CC chemokines and three CXC chemokines) showed high levels of expression and were up-regulated 2-fold or greater (Table 2), whilst the majority of chemokines showed no change in expression during MNV-1 infection. 3.4. Specific chemokine mRNA expression is up-regulated during MNV-1 infection To validate the microarray results, CCL2, CCL3, CCL4, CCL5, CXCL2, CXCL10 and CXCL11 were examined for changes in mRNA levels by qPCR. In addition, CCL9 and CXCL16, which showed high signal strengths but no difference between mock and MNV-1 infected samples, were used as controls. Relative cDNA abundance, was measured at 12, 15 and 18 h.p.i., and peaked at 18 h.p.i. (data not shown). The changes in expression of seven of the nine chemokine genes, measured by qPCR, were consistent with changes seen by microarray (Fig. 3). CCL3 and CCL4 showed modest changes of 3- and 5-fold respectively, and CCL5 by 20-fold. Additionally, CCL2 showed a substantial 7-fold increase, despite a constitutively high basal level within macrophages (Maurer and von Stebut, 2004). CXCL10 showed constant up-regulation of at least 36-fold, however due to the very low abundance of CXCL10 in mock-infected cells there was some inconsistency in the fold changes obtained (data not shown). Similarly, levels of CXCL2 and CXCL11 determined by qPCR showed that they were up-regulated up to 6000-fold during MNV-1 infection (data not
E. Waugh et al. / Virus Research 181 (2014) 27–34
31
Table 2 Chemokine genes showing up-regulation after infection of RAW264.7 cells by Murine Norovirus for 18 h. Affymetrix probe ID
Chemokine
Signal strength
Fold change
Mock
MNV
11
12
1.1
C chemokine ligands
1419412 at
XCL1
CC chemokine ligands
1421688 a at 1420380 ata 1419561 at 1421578 at 1418126 at 1420249 s at 1421228 at 1419684 at 1448898 atb 1417789 at 1419282 at 1419413 at 1449277 at 1422029 at 1445238 at 1417925 at 1450488 at 1418777 at 1434962 x at 1450217 at
CCL1 CCL2 CCL3 CCL4 CCL5 CCL6 CCL7 CCL8 CCL9 CCL11 CCL12 CCL17 CCL19 CCL20 CCL21b CCL22 CCL24 CCL25 CCL27 CCL28
11 495 2567 859 59 22 31 10 6373 5 14 24 23 16 11 10 12 45 47 24
12 1733 7357 7700 1230 20 83 10 8031 5 39 22 21 20 12 11 16 36 7 20
1.1 3.5 2.9 9.0 20.8 0.9 2.7 1.0 1.3 1.0 2.8 0.9 0.9 1.3 1.1 1.1 1.3 0.8 0.1 0.8
CXC chemokine ligands
1441855 x at 1449984 at 1438148 at 1419728 at 1418652 at 1418930 at 1419697 at 1448823 at 1448859 at 1418456 at 1456428 at 1449195 s at 1451610 at
CXCL1 CXCL2 CXCL3 CXCL5 CXCL9 CXCL10 CXCL11 CXCL12 CXCL13 CXCL14 CXCL15 CXCL16 CXCL17
48 51 4 4 10 116 37 9 26 11 11 446 29
55 1909 4 4 20 6791 2175 9 24 9 12 276 28
1.1 37.4 1.0 1.0 2.0 58.5 58.8 1.0 0.9 0.8 1.1 0.6 1.0
CX3C chemokine ligands
1415803 at
CX3CL1
11
10
0.9
Chemokine receptors
1419609 1421186 1422957 1424727 1423466 1421920 1449925 1421420 1448710 1422812 1417625 1450019
CCR1 CCR2 CCR3 CCR5 CCR7 CCR9 CCR10 CXCR3 CXCR4 CXCR6 CXCR7 CX3CR1
31 48 8 6 14 7 57 220 41 5 5 151
271 101 10 7 12 7 29 174 39 5 6 25
8.7 2.1 1.3 1.2 0.9 1.0 0.5 0.8 1.0 1.0 1.2 0.2
a b
at at at at at a at at at at at s at at
Bolded text indicates chemokines of interest. Italicized text indicates control chemokines.
shown), however the actual fold change in expression of CXCL2 and CXCL11 could not be accurately determined due to the low abundance of these chemokines in mock-infected macrophages that were at the detection limits of the assay. Low levels of CXCL2 in unstimulated RAW264.7 cells has been described previously (Kim and Zhang, 2003). The qPCR data indicated that the results from the microarray analysis accurately reflected the overall changes in chemokine gene expression that occurred during MNV-1 infection of RAW264.7 cells. 3.5. Chemokine protein concentration in medium from MNV-1 infected cells is up-regulated Six of the nine chemokines of interest were assayed using commercially available kits. The concentrations of the chemokine proteins present in the MNV-1 infection supernatant were measured at multiple time points throughout infection. Cells began
showing apparent cytopathic effects after 21 h of infection at the m.o.i. used in these studies. The ELISA data established that CCL2, CCL9 and CXCL2 in particular were secreted at high levels from MNV-1-infected RAW264.7 cells, with concentrations of 7019 pg/ml, 22,285 pg/ml and 8698 pg/ml, respectively being detected at 21 or 24 h.p.i. CCL2 and CXCL2 showed modest upregulation of 4- and 4.5-fold, respectively. CXCL10 and CCL5 secretion increased considerably in response to MNV-1 infection with a 38- and 8-fold increase at 18 and 21 h.p.i. respectively. Concentrations of 665 pg/ml and 290 pg/ml were detected for CXCL10 and CCL5 (Fig. 4). CXCL11 was below the detection limit (
Contents lists available at ScienceDirect
Virus Research journal homepage: www.elsevier.com/locate/virusres
Characterization of the chemokine response of RAW264.7 cells to infection by murine norovirus Emily Waugh a , Augustine Chen b , Margaret A. Baird a,1 , Chris M. Brown b , Vernon K. Ward a,∗ a b
Department of Microbiology and Immunology, School of Medical Sciences, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand Department of Biochemistry, School of Medical Sciences, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand
a r t i c l e
i n f o
Article history: Received 12 September 2013 Received in revised form 13 December 2013 Accepted 16 December 2013 Available online 24 December 2013 Keywords: Norovirus Calicivirus Chemokines Immunity
a b s t r a c t Noroviruses are an emerging threat to public health, causing large health and economic costs, including at least 200,000 deaths annually. The inability to replicate in cell culture or small animal models has limited the understanding of the interaction between human noroviruses and their hosts. However, an alternative strategy to gain insights into norovirus pathogenesis is to study murine norovirus (MNV-1) that replicates in cultured macrophages. While the innate immune response is central to the resolution of norovirus disease, the adaptive immune response is required for viral clearance. The specific responses of macrophages and dendritic cells to infection drive the adaptive immune response, with chemokines playing an important role. In this study, we have conducted microarray analysis of RAW264.7 macrophages infected with MNV-1 and examined the changes in chemokine transcriptional expression during infection. While the majority of chemokines showed no change, there was specific up-regulation in chemokines reflective of a bias toward a Th1 response, specifically CCL2, CCL3, CCL4, CCL5, CXCL2, CXCL10 and CXCL11. These changes in gene expression were reflected in protein levels as determined by ELISA assay. This virus-induced chemokine response will affect the resolution of infection and may limit the humoral response to norovirus infection. © 2014 Published by Elsevier B.V.
1. Introduction Noroviruses are non-enveloped, positive-sense, single-stranded RNA viruses belonging to the Caliciviridae family. In general, calicivirus genomes comprise three orfs, with orf1 encoding the non-structural polyprotein while orf2 and orf3 encode for the structural capsid proteins (Sosnovtsev et al., 2006), while an additional fourth open reading frame has been discovered in murine norovirus (McFadden et al., 2011). Human norovirus (HuNoV) is the most common cause of non-bacterial gastroenteritis worldwide (Perry et al., 2009) and in developing countries, norovirus infection can be attributed to the death of 200,000 children five years old or younger (Patel et al., 2009). Symptoms of a norovirus infection commonly include vomiting and diarrhea as well as abdominal cramps and
Abbreviations: HuNoV, human norovirus; MNV, murine norovirus; qPCR, quantitative real time PCR; h.p.i., hours post-infection; GO, gene ontology; TLR, Toll-like receptor; GEO, gene expression omnibus; NCBI, National Center for Biotechnology Information; RMA, robust multi-array average. ∗ Corresponding author. Tel.: +64 3 479 9028; fax: +64 3 479 8540. E-mail address: [email protected] (V.K. Ward). 1 Present address: Department of Pathology, University of Otago, Dunedin 9054, New Zealand. 0168-1702/$ – see front matter © 2014 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.virusres.2013.12.025
fever (Estes et al., 2006) and due to the self-limiting nature of the virus, symptoms typically resolve within 24–48 h in immunocompetent hosts (Mumphrey et al., 2007). Despite the large burden of morbidity and mortality caused by norovirus infections, there is currently no vaccine or anti-viral treatment available. This is in part because the study of HuNoV has been limited by the lack of a small animal model or cell culture system that supports HuNoV replication (Duizer et al., 2004). In 2003, a norovirus was identified that infected mice lacking signal transducer and activator of transcription 1 (Karst et al., 2003). This murine norovirus (MNV-1) has a tropism for macrophages and dendritic cells (Wobus et al., 2004), thus it can be effectively propagated in cell culture and is now routinely used as a model system to study HuNoV infection (Belliot et al., 2008; Wobus et al., 2006). The innate immune response to MNV-1 has been well investigated, as it is the predominant response to a norovirus infection (Changotra et al., 2009; Karst et al., 2003). The innate response to MNV-1 is initiated through recognition of the viral double-stranded RNA by retinoic acid-inducible gene 1 (RIG-I)-like receptors, such as melanoma differentiation-associated protein 5 (MDA5) (McCartney et al., 2008). This induction initiates several signaling pathways within the cell that leads to the production of type 1 interferons (IFN-␣ and IFN-) as well as other pro-inflammatory
28
E. Waugh et al. / Virus Research 181 (2014) 27–34
cytokines (Platanias, 2005). This process causes the activation of an antiviral state and control of the virus infection (Loo et al., 2008). In addition, studies using both HuNoV and MNV-1 have demonstrated that the acquired immune response is vital for clearance of a norovirus infection (Chachu et al., 2008) and prolonged HuNoV infection is observed in immunocompromised individuals (Wingfield et al., 2010). Both the humoral and cell-mediated acquired immune response play a role in eliminating norovirus infections (LoBue et al., 2010). The production of antibody and the generation of memory B cells are important in protecting the host, with poor antibody responses allowing reinfection of the host (Bok et al., 2011). Cytotoxic T cells induced and activated by cytokines, kill cells infected with norovirus. Studies have shown that the cytokine profile of norovirus-infected cells includes the production of large amounts of gamma interferon (IFN-␥) and interleukin 2 (LoBue et al., 2010; Souza et al., 2007), which are characteristic of a Th1 response (Lindesmith et al., 2005). Chemokines are an integral and critical step in the immune response to virus infections (Sillanpaa et al., 2008; Zeremski et al., 2007). Few investigations have been carried out on the chemokine response to MNV-1. Chemokines are a family of small (∼8–12 kDa) secreted molecules that exert their effects by binding to seven-transmembrane G-protein coupled receptors to establish a chemokine gradient, allowing directional migration of cells expressing the corresponding chemokine receptor (Deshmane et al., 2009; Zhao et al., 2009). Members of the chemokine family mostly fall into two broad groups; the CC chemokines and the CXC chemokines. Other groups include the C and CX3C chemokines. In addition, chemokines can be divided into constitutive and inflammatory chemokines (Lateef et al., 2009; Sallusto and Baggiolini, 2008). Members of the CC class ( chemokines) recruit the majority of leukocyte cell types (Cole et al., 1998), while the CXC chemokines (␣ chemokines) primarily recruit neutrophils and T lymphocytes (Cole et al., 1998; Mortier et al., 2008). Viruses are known to interact with and manipulate the host chemokine system (Lateef et al., 2010; Luttichau, 2010; Sillanpaa et al., 2008; Webb and Alcami, 2005). However, the ability of MNV-1 to interact with the chemokine network and the role of chemokines important in both the innate and the acquired immune response to MNV-1 has not been studied. This study characterizes the host chemokine response to MNV-1 infection of RAW 264.7 macrophages.
from cells at 9 or 18 h.p.i. using the Qiagen RNeasy mini kit as per manufacturers instructions. Transcriptomic analysis was conducted in triplicate using a GeneChip® Mouse Genome 430 2.0 Array (Affymetrix) at the Otago Genomics Facility. The complete data set obtained from the microarray analysis was submitted to the GEO database (NCBI), accession number GSE50093. 2.3. Expression analysis by microarray Raw microarray intensities from CEL files were normalized using RMA and quantile normalization using the ExpressionFileCreator in Genepattern (Reich et al., 2006) and preprocessed using PreprocessDataset and comparison was done using ComparativeMarkerSelection. Ratios of expression at 9 or 12 h.p.i. were generated. Ratios were not generated for genes with marginally detectable expression of 50 in the mock-infected control. Differentially expressed gene sets were analyzed using DAVID Bioinformatics Resources 6.7 (National Institute of Allergy and Infectious Diseases, NIH) (Huang da et al., 2009a, 2009b). Affymetrix I.D.s of genes altered 2-fold or more were used for functional annotation clustering. This was performed to identify groups enriched in genes associated with similar biological functions. Mouse Genome Informatics (Eppig et al., 2012) was used to ascertain the identity and function of the genes. Raw microarray data (Bok et al., 2009) was accessed through the GEO database at the National Center for Biotechnology Information (accession number GSE12518). This data was analyzed in the same way as the CEL files generated in this study. 2.4. Virus purification
2. Materials and methods
RAW264.7 cells were seeded at 2.8 × 107 in 175 cm2 flasks. The following day, confluent monolayers were infected with MNV-1 at an m.o.i. of 1.0, as previously described with the exception that after absorption, the inoculum was removed and the cells were washed with fresh medium before adding 60 ml of medium and incubating for 28 h. The infected cells were frozen at −20 ◦ C then thawed at 37 ◦ C before clarification of the inoculum by a low speed centrifugation at 1000 × g for 20 min. This unpurified inoculum was purified through a 30% (w/v) sucrose cushion at 112,700 × g for 4 h at room temperature. The pellet was re-suspended in 1 ml Dulbecco’s phosphate buffered saline overnight at 4 ◦ C, then filter sterilized through a 0.45 m filter.
2.1. Cells and virus
2.5. Detection of IFN-ˇ
RAW264.7 cells (obtained from ATCC) were maintained in a 75 cm2 flask at 5 × 106 cells in Dulbecco’s modified Eagle’s medium supplemented with 5% fetal calf serum, 100 U/ml penicillin and 0.1 mg/ml streptomycin. During infection cells were incubated in the same medium without antibiotics. A working stock of MNV-1 was prepared using a previously described reverse genetics method (Ward et al., 2007). Viral titers were determined by crystal violet plaque assay in RAW264.7 cells (Gonzalez-Hernandez et al., 2012).
Levels of IFN- in virus inoculum were determined by ELISA using a mouse IFN- ELISA kit (PBL Interferon Source, Piscataway, NJ, USA). Unpurified MNV-1 and purified MNV-1 inocula dilutions were assayed in triplicate.
2.2. Microarray A 6-well plate was seeded with RAW264.7 cells at 8 × 105 cells per well. The next day confluent monolayers were infected with unpurified MNV-1 at an m.o.i. of 1.0. Medium was removed from wells, replaced with virus inoculum and incubated at 37 ◦ C + 5% CO2 for 1 h, with rocking. Following incubation, 2 ml of fresh medium was added to each well and plates further incubated at 37 ◦ C + 5% CO2 . Mock-infected cells were treated the same way with culture medium added instead of virus inoculum. RNA was extracted
2.6. Quantitative real-time PCR Microarray results were validated by qPCR for selected genes that showed altered expression. Six-well plates were seeded with RAW264.7 cells at 8 × 105 cells per well. The next day 80% confluent monolayers were infected with purified MNV-1 at an m.o.i. of 1.0, or mock-infected. Total RNA was isolated from cells at 12, 15 and 18 h.p.i. using TRIzol reagent (Invitrogen, Auckland, NZ) and RNA purification was performed using the PureLink RNA mini kit (Invitrogen). Isolated RNA was subjected to DNase treatment and cDNA was generated from DNase-treated total RNA via random hexamer reverse transcription using the High Fidelity Transcriptor cDNA synthesis kit (Roche, Auckland, NZ). Gene-specific primers for qPCR were selected using the Universal Probe Library from
E. Waugh et al. / Virus Research 181 (2014) 27–34
29
Table 1 Primer Sequences. Chemokine
Forward primer sequence (5 –3 )
Reverse primer sequence (5 –3 )
CCL2 CCL3 CCL4 CCL5 CCL9 CXCL2 CXCL10 CXCL11 CXCL16
AGGTGTCCCAAAGAAGCTGTA TGCCCTTGCTGTTCTTCTCT ATGAGACCAGCAGTCTTTGCTCCA TCGTGCCCACGTCAAGGAGTATTT TACCAGGCCGGGCATCATCTTTAT ACATCCCACCCACACAGTGAAAGA GCTGCCGTCATTTTCTGC AGATCCAAGCAAGCTCGCCTCATA TGAAAGCATCTTGGAGCCAGAGGA
ATGTCTGGACCCATTCTTCT GTGGAATCTTCCGGCTGTAG GCTGCTCAGTTCAACTCCAAGTCA TCTTCTCTGGGTTGGCACACACTT TGGCAGTTCACACCCTTCTCTTCA TCCTTCCATGAAAGCCATCCGACT TCTCACTGGCCCGTCATC ATGTTCCAAGACAGCAGAGGGTCA TGGTTCCCGAAGAATGACTCGGTT
Roche (Table 1). Primer design requirements included a DNA melting temperature between 58 and 60 ◦ C, a guanine–cytosine content of 40–60% and the primers were across intron spanning regions to eliminate genomic DNA contamination. Where possible, primers validated in published articles were used (Cantini et al., 2012; Griener et al., 2011; Hochrainer et al., 2013; Lu et al., 2013; Ohoka et al., 2011). A serial dilution of cDNA was used for each primer set to determine the effective range of template concentration that gave a consistent and optimal efficiency. qPCR was performed on a LightCycler® 480 platform using 2× SYBR green master mix (Roche). Briefly, 3 l of cDNA was added to 0.5 M each of the corresponding forward and reverse primer and 5 l of 2× SYBR green master mix. Cycling conditions were 95 ◦ C for 5 min followed by 50 cycles at 95 ◦ C, 5 s; 58 ◦ C, 5 s; 72 ◦ C, 8 s; then 65 ◦ C for 5 min. Duplicate samples were set up for each sample and all samples were normalized to -actin in the same sample to correct for efficiency of reverse transcription. The levels of genes of interest were calculated using the efficiency method for relative quantification. The efficiency method employs relative standards prepared using sample cDNA and thus normalizes the amplification efficiency of target and reference genes as well as run-to-run differences. Specificity of the qPCR was confirmed by the melting curve of amplified products. In addition, each qPCR experiment included a standard curve of each primer set to determine primer efficiency. This efficiency value (between 1.8 and 2.0) was used in the normalization and fold change calculation, using Roche LightCycler480 software (version 1.5.0.39).
up-regulated 2-fold or more during MNV-1 infection and these genes were further clustered into 187 Gene Ontology annotation clusters using DAVID functional annotation tools (Huang da et al., 2009b). The most noteworthy clusters included toll-like signaling (e.g. toll-like receptor (TLR) signaling P = 9.1 × 10−11 ) programmed cell death (e.g. apoptosis P = 1.3 × 10−5 ), as well as immune molecules (P = 1.5 × 10−4 ) (Fig. 1). The most significant
2.7. Chemokine quantification Confluent monolayers of RAW264.7 cells were mock-infected or infected with purified MNV-1 at an m.o.i. of 1.0. Cell culture supernatants were collected and DuoSet ELISA Development kits (R&D Systems, Minneapolis, MN, USA) were used to determine the amount of CCL5, CXCL2, CXCL10 and CXCL11 produced during MNV-1 infection. OptEIA Mouse ELISA set (BD Biosciences Pharmingen, New Jersey, USA) was used to quantify the amount of CCL2 produced during infection. 3. Results and discussion 3.1. Overall changes in cellular gene expression during MNV-1 infection To measure changes in gene expression in macrophages (RAW264.7) in response to infection by MNV-1, cells were mockinfected or infected with virus and mRNA levels measured by microarray at 9 and 18 h post infection (h.p.i). Duplicate microarrays using two biological replicates on an Affymetrix Mouse 430 2.0 array, which represented over 34,000 genes, were carried out. The results showed that the expression of approximately 33,000 genes (97%) remained relatively constant during MNV-1 infection. This analysis identified 766 genes whose expression was
Fig. 1. Heat map of selected genes up-regulated 3-fold or more during MNV-1 infection. RAW264.7 cells were infected with MNV-1 at m.o.i. of 1. Total cellular RNA was harvested at 9 and 18 h.p.i. and analyzed by Affymetrix Mouse 430 2.0 microarray. Heat map image of fold changes was generated using Rstudio software (Version 0.96.330, ©Rstudio, Inc.). Affymetrix IDs and gene names are indicated. (A) Genes associated with Toll-like receptor signaling. (B) Genes associated with apoptosis. (C). Genes associated with the immune response.
30
E. Waugh et al. / Virus Research 181 (2014) 27–34
Fig. 3. Up-regulation of chemokine mRNA expression in MNV-1 infected RAW264.7 cells. Cells were infected with MNV-1 at an m.o.i. of 1.0, total cellular RNA was isolated at 18 h.p.i. and cDNA generated. qPCR analysis was carried out using SYBR green signal transmission on a Roche LightCycler 480 system. Chemokine levels were normalized to -actin. Data shows the mean fold increases over mock-infected cells of three experiments performed in duplicate for all chemokines except CXCL10, which is one representative experiment.
Fig. 2. Heat map of selected down-regulated genes during MNV-1 infection. RAW264.7 cells were infected with MNV-1 at m.o.i. of 1. Total cellular RNA was harvested at 9 and 18 h.p.i. and analyzed by Affymetrix Mouse 430 2.0 microarray. Heat map image of fold changes was generated using Rstudio software (Version 0.96.330, ©Rstudio, Inc.). Affymetrix IDs and gene names are indicated. (A) Genes associated with cell cycle. (B) Genes associated with programmed cell death. (C) Genes associated with selected biosynthetic pathways. (D) Genes associated with nucleotide metabolism.
clusters of down-regulated genes (779) included those involved in cell cycle regulation (e.g. cell cycle, P = 2.9 × 10−5 ), programmed cell death (e.g. apoptosis, P = 1.0 × 10−2 ) and nucleotide binding (e.g. ATP binding, P = 5.4 × 10−2 ) (Fig. 2). These changes indicate that MNV-1 may be modulating the cell cycle, cell metabolism and/or regulating apoptosis. A similar microarray performed by Bok et al. (2009) also identified down-regulation in apoptosis in RAW264.7 cell during a MNV-1 infection showed that MNV-1 down-regulates survivin, a key regulator of apoptosis. 3.2. Genes associated with interferon stimulation and the immune response are significantly up-regulated by MNV-1 infection A large number of up-regulated genes clustered into immune response categories including cytokine-cytokine receptor signaling (P = 1.0 × 10−9 ), defense response (P = 4.0 × 10−11 ), inflammatory response (P = 1.9 × 10−4 ) and chemotaxis (P = 6.6 × 10−3 ) (Fig. 1 and Table 2). This result was not unexpected as it is known that norovirus infection induces an innate immune response as well as a Th1-type acquired response (Chachu et al., 2008; Lindesmith et al., 2005; Wobus et al., 2006). In addition, further analysis of the microarray performed by Bok et al. (2009) supported our findings. Of the 50 most up-regulated genes, 11 (22%) were associated with, or stimulated by type I interferon, suggesting that interferon present in the inoculate may be responsible for some of the up-regulation seen in the microarray data. Although IFN-
quantification revealed that the virus inoculum contained at least 1000 pg/ml of IFN-, the microarray performed using unpurified virus yielded similar results to the purified virus microarray data of Bok et al. (Supplementary data). Additionally, in all subsequent qPCR and ELISA analyses the viral inoculum was purified by ultracentrifugation through a 30% sucrose cushion to remove any residual interferon. The removal of immunostimulatory molecules such as interferon was confirmed by IFN ELISA, in which interferon levels were below the detection limit of the assay. 3.3. The host chemokine response is up-regulated during MNV-1 infection Analysis of the microarray data revealed that a significant number of chemokine (P = 1.9 × 10−3 ) and chemokine receptor genes (P = 1.6 × 10−3 ) were up-regulated by MNV-1 infection. Of the 35 chemokine probes on the array, seven (four CC chemokines and three CXC chemokines) showed high levels of expression and were up-regulated 2-fold or greater (Table 2), whilst the majority of chemokines showed no change in expression during MNV-1 infection. 3.4. Specific chemokine mRNA expression is up-regulated during MNV-1 infection To validate the microarray results, CCL2, CCL3, CCL4, CCL5, CXCL2, CXCL10 and CXCL11 were examined for changes in mRNA levels by qPCR. In addition, CCL9 and CXCL16, which showed high signal strengths but no difference between mock and MNV-1 infected samples, were used as controls. Relative cDNA abundance, was measured at 12, 15 and 18 h.p.i., and peaked at 18 h.p.i. (data not shown). The changes in expression of seven of the nine chemokine genes, measured by qPCR, were consistent with changes seen by microarray (Fig. 3). CCL3 and CCL4 showed modest changes of 3- and 5-fold respectively, and CCL5 by 20-fold. Additionally, CCL2 showed a substantial 7-fold increase, despite a constitutively high basal level within macrophages (Maurer and von Stebut, 2004). CXCL10 showed constant up-regulation of at least 36-fold, however due to the very low abundance of CXCL10 in mock-infected cells there was some inconsistency in the fold changes obtained (data not shown). Similarly, levels of CXCL2 and CXCL11 determined by qPCR showed that they were up-regulated up to 6000-fold during MNV-1 infection (data not
E. Waugh et al. / Virus Research 181 (2014) 27–34
31
Table 2 Chemokine genes showing up-regulation after infection of RAW264.7 cells by Murine Norovirus for 18 h. Affymetrix probe ID
Chemokine
Signal strength
Fold change
Mock
MNV
11
12
1.1
C chemokine ligands
1419412 at
XCL1
CC chemokine ligands
1421688 a at 1420380 ata 1419561 at 1421578 at 1418126 at 1420249 s at 1421228 at 1419684 at 1448898 atb 1417789 at 1419282 at 1419413 at 1449277 at 1422029 at 1445238 at 1417925 at 1450488 at 1418777 at 1434962 x at 1450217 at
CCL1 CCL2 CCL3 CCL4 CCL5 CCL6 CCL7 CCL8 CCL9 CCL11 CCL12 CCL17 CCL19 CCL20 CCL21b CCL22 CCL24 CCL25 CCL27 CCL28
11 495 2567 859 59 22 31 10 6373 5 14 24 23 16 11 10 12 45 47 24
12 1733 7357 7700 1230 20 83 10 8031 5 39 22 21 20 12 11 16 36 7 20
1.1 3.5 2.9 9.0 20.8 0.9 2.7 1.0 1.3 1.0 2.8 0.9 0.9 1.3 1.1 1.1 1.3 0.8 0.1 0.8
CXC chemokine ligands
1441855 x at 1449984 at 1438148 at 1419728 at 1418652 at 1418930 at 1419697 at 1448823 at 1448859 at 1418456 at 1456428 at 1449195 s at 1451610 at
CXCL1 CXCL2 CXCL3 CXCL5 CXCL9 CXCL10 CXCL11 CXCL12 CXCL13 CXCL14 CXCL15 CXCL16 CXCL17
48 51 4 4 10 116 37 9 26 11 11 446 29
55 1909 4 4 20 6791 2175 9 24 9 12 276 28
1.1 37.4 1.0 1.0 2.0 58.5 58.8 1.0 0.9 0.8 1.1 0.6 1.0
CX3C chemokine ligands
1415803 at
CX3CL1
11
10
0.9
Chemokine receptors
1419609 1421186 1422957 1424727 1423466 1421920 1449925 1421420 1448710 1422812 1417625 1450019
CCR1 CCR2 CCR3 CCR5 CCR7 CCR9 CCR10 CXCR3 CXCR4 CXCR6 CXCR7 CX3CR1
31 48 8 6 14 7 57 220 41 5 5 151
271 101 10 7 12 7 29 174 39 5 6 25
8.7 2.1 1.3 1.2 0.9 1.0 0.5 0.8 1.0 1.0 1.2 0.2
a b
at at at at at a at at at at at s at at
Bolded text indicates chemokines of interest. Italicized text indicates control chemokines.
shown), however the actual fold change in expression of CXCL2 and CXCL11 could not be accurately determined due to the low abundance of these chemokines in mock-infected macrophages that were at the detection limits of the assay. Low levels of CXCL2 in unstimulated RAW264.7 cells has been described previously (Kim and Zhang, 2003). The qPCR data indicated that the results from the microarray analysis accurately reflected the overall changes in chemokine gene expression that occurred during MNV-1 infection of RAW264.7 cells. 3.5. Chemokine protein concentration in medium from MNV-1 infected cells is up-regulated Six of the nine chemokines of interest were assayed using commercially available kits. The concentrations of the chemokine proteins present in the MNV-1 infection supernatant were measured at multiple time points throughout infection. Cells began
showing apparent cytopathic effects after 21 h of infection at the m.o.i. used in these studies. The ELISA data established that CCL2, CCL9 and CXCL2 in particular were secreted at high levels from MNV-1-infected RAW264.7 cells, with concentrations of 7019 pg/ml, 22,285 pg/ml and 8698 pg/ml, respectively being detected at 21 or 24 h.p.i. CCL2 and CXCL2 showed modest upregulation of 4- and 4.5-fold, respectively. CXCL10 and CCL5 secretion increased considerably in response to MNV-1 infection with a 38- and 8-fold increase at 18 and 21 h.p.i. respectively. Concentrations of 665 pg/ml and 290 pg/ml were detected for CXCL10 and CCL5 (Fig. 4). CXCL11 was below the detection limit (
