Modeling host responses in ferrets during A/California/07/2009 influenza infection

Infection and monitoring of Ferrets

Animal experiments with virus A/California/07/2009 or with A/Brisbane/59/2007 were performed in the AAALAC-accredited ABSL-3 facility, University of Pittsburgh or in the ABSL-2 enhanced facility Animal Resource Centre, University Health Network, Toronto, respectively. All procedures were conducted in accordance with the NRC Guide for the Care and Use of Laboratory Animals, the Animal Welfare Act, and the CDC-NIH Biosafety in Microbiological and Biomedical laboratories and approved by the Institutional Animal Care and Use Committee. All ferrets were proven seronegative by HI assay against A/California/04/2009, A/Brisbane/59/2007 and B/Florida/04/2006. Castrated, descented male fitch ferrets (Mustela putorius furo), 5 – 9 months of age and weighing between 800 to 1500 grams (Triple F Farms Inc., Sayre, PA, USA) were pair-housed in individually HEPA-Filtered biocontainment caging (Allentown Inc., Allentown, NJ, USA). Ferrets were given food and water ad libitum. Animals were infected and monitored as previously described (Zitzow et al., 2002), except using 5% isofluorane anesthesia. Briefly, 29 male fitch ferrets were anesthetized with isofluorane and infected intranasally with 1×10⁶ 50% egg infectious doses (EID₅₀) (~1×10^5.75 TCID₅₀/ml) of A/California/07/2009 in a volume of one milliliter. From these, four animals were sacrificed on days 1, 3, 5, 7, and 14 days post infection (PI). Nasal washes, lung tissue and blood samples were collected from each ferret in order to determine viral load, gross and histopathological findings, innate and adaptive host immune responses, and molecular changes associated with disease course. The remaining nine animals were allowed to go to endpoint. A/California/07/2009 is the World Health Organization recommended isolate for generation of 2009-H1N1 vaccines (http://www.who.int/entity/csr/resources/publications/swineflu/H1N1Vaccinevirusrecommendation26May2009.pdf) and was thus the virus used for ferret studies. An additional 4 ferrets were used as Day 0 uninfected controls. Temperatures were measured through use of an implantable temperature transponder (BMDS, Sayre, PA) and were recorded daily at approximately the same time each day. Pre-infection values were averaged to obtain a baseline temperature for each ferret. The change in temperature was calculated daily for each animal. Weight loss and clinical signs of sneezing and nasal discharge, other respiratory distress, and level of activity were assessed daily. A scoring system was used to assess activity level where 0=alert and playful; 1=alert but playful only when stimulated; 2=alert but not playful when stimulated; 3=neither alert nor playful when stimulated. Based on the daily scores for each animal in a group, a relative inactivity index was calculated (Zitzow et al., 2002).

An additional 15 ferrets were infected with 1×10⁶ EID₅₀ of A/Brisbane/59/2007 and monitored for clinical signs. On days 1, 2, 5, 7 and 14 post infection, 3 ferrets were euthanized and organs collected for pathological examination, virus titration and microarray analysis. A group of 3 uninfected controls was also included.

Histopathology

Lung tissue excised from the middle lobes was formalin-fixed and paraffin embedded. Tissue slides were stained with hematoxilyn/eosin for histopathologic assessment. Immunostain was performed by using a polyclonal goat anti-influenza A 1:500 (#PAB7123P, Maine Biotechnology Services, Portland, ME, USA) followed by incubation with a donkey anti-goat b/c antibody 1:200 (#805-7602, Rockland immunochemical, Gilberstville, PA, USA).

Determination of viral load

MDCK cells plated in 6-well tissue culture plates were inoculated with 0.1 ml of virus serially diluted in Dubecco’s modified Eagle’s medium (DMEM). Virus was adsorbed to cells for 1 h, with shaking every 15 min. Wells were overlaid with 1.6% w/v Bacto agar (DIFCO, BD Diagnostic Systems, Palo Alto, CA, USA) mixed 1:1 with L-15 media (Cambrex, East Rutherford, NJ, USA) containing antibiotics and 0.6μg/ml TPCK-trypsin (Sigma, St. Louis, MO, USA). Plates were inverted and incubated for 4 days. Cells were fixed for 10 minutes using 70% v/v Ethanol and then overlaid with 1% w/v crystal violet. Cells were then washed with deionized water to visualize plaques. Plaques were counted and compared to uninfected cells. Additionally, viral load was assessed by HA activity in tissues which were homogenized 10% w/v in DMEM and serially diluted (1/2 log) from 10-0.5 to 10-6.5 in quadruplicate on MDCK cells. After 2 hours incubation at 37°C, 5% CO2 samples were removed and replaced with DMEM containing 1% BSA, 1ug/ml TPCK-Trypsin and antibiotics and incubated. On day 6 post infection, fifty microliters of supernatant were tested for presence of virus by HA activity against 0.5% turkey erythrocytes. The viral loads were determined as the reciprocal of the dilution resulting in 50% HA positivity. Titers are given as TCID₅₀ per gram of tissue or TCID₅₀/ml for nasal washes.

Hemagglutinin Inhibition (HI) assay

RDE-treated sera were serially diluted in v-bottom 96-well microtiter plates followed by the addition of 4 Hemagglutination units (HAU) of influenza virus. Following an incubation of approximately 15 minutes, 0.5% turkey erythrocytes (TRBC) were added and mixed by agitation. The TRBC were allowed to settle for 30 minutes at room temperature and HI titers were determined by the reciprocal value of the last dilution of sera which inhibited hemagglutination of TRBC by the virus. A negative titer was calculated as 10.

RNA Extraction and Microarray Analysis

Performed as previously published (Cameron et al., 2008) except for the gene classification methods. Briefly, lung sections (~1 g) were obtained during necropsy and immediately put in RNAlater (Ambion, Austin, TX, USA) then homogenized in TriPure reagent, and total RNA was isolated according to the manufacturer’s recommended protocol (Roche, Indianapolis, IN, USA). Total RNA was amplified using MessageAMP kits (Ambion). cRNA (10 g) was labeled and hybridized to Affymetrix GeneChip Canine Genome 2.0 Array (Affymetrix, Santa Clara, CA, USA). Probe-level analysis was performed using the Probe Logarithmic Intensity Error Estimate (PLIER) method in Expression Console Version 1.1 (Affymetrix). The raw intensity values for each target on the Affymetrix arrays were preprocessed with variance stabilization, log₂ transformation, and normalization to the control group (Day 0 samples) excluding elements with median signal intensities of <6 log₂ units across samples. Time series analysis was performed using EDGE version 1.1.291 (Storey et al., 2005) with Q ≤ 0.005. DAVID bioinformatic tool (Huang da, Sherman, and Lempicki, 2009) was used to perform functional classification in differentially expressed genes, using a biological significance filter of ≥1.5-fold, additional custom enrichment of functional categories (ISGs, cytokine/chemokines and leukocyte markers) was performed. Gene hierarchical clustering and heat map generation was done using the Multi Experiment Viewer 4.4 software (Saeed et al., 2006). Ingenuity Pathway Analysis 7.6 (IPA) (Ingenuity Systems, Redwood City, CA, USA) was used to select, annotate, and visualize genes by function and pathway. Datasets are publicly available at the NCBI’s Gene Expression Omnibus. (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=PENDING).

qRT-PCR

Quantitative real-time PCR (qRT-PCR) was performed in triplicate using an ABI-Prism 7900HT (Applied Biosystems, Foster City, CA, USA). PCR reactions were performed in a 10μl reaction volume using SYBR green master mix (Applied Biosystems), 40 amplification cycles and the annealing temperature was 60 degrees. Primer sequences were as follows: for CCL2, 5′-GCTCCCTATTCACTTGCTGTTTC-3′ (forward) and 5′-GATTCGATAGCCCTCCAGCTT-3′ (reverse); for CXCL10, 5′-CTTTGAACCAAAGTGCTGTTCTTATC-3′ (forward) and 5′-AGCGTGTAGTTCTAGAGAGAGGTACTC-3′ (reverse); for IL-1β, 5′-GGACTGCAAATTCCAGGACATAA-3′ and 5′-TTGGTTCACACTAGTTCCGTTGA-3′ (reverse); for TNFα, 5′-CCAGATGGCCTCCAACTAATCA-3′ (forward) and 5′-GGCTTGTCACTTGGAGTTCGA-3′ (reverse). Expression levels were normalized using the house-keeping gene β-actin, results were expressed as arbitrary units. The statistical analysis was performed using the SPSS 16.0 software (SPSS Inc, Chicago, IL, USA) and the Student’s t test was used to assess the statistical differences in the expression levels respect to the control groups.

Pathological features of the infection with A/California/07/2009 in ferrets

Following challenge with California/07, ferrets exhibited classical clinical signs of influenza infection including elevated temperature, weight loss, nasal discharge, sneeze (table 1 and supplementary Fig. 1) and crusty nose (22% of ferrets on day 6 PI). Animals lost weight throughout the first week post-challenge and slowly recovered by day 14 PI. 3 ferrets on day 6 PI and 2 ferrets on day 7 PI, lost more than 20% of their initial weight and they were euthanized. The resulting mortality rate was 27%. Viral kinetics demonstrated early replication in the upper tract at day 1 PI, which decreased over the following four days (Fig. 1A). Viral load was low on day 1 in the trachea and progressively increased until day 5 PI (Fig. 1B). Viral titers in lung tissue showed great variability, it was initially detected on day 1 and peaked on day 5 PI (Fig. 1C). No virus was detected in any airway tissues on day 7 or later (Fig. 1).

Hemagglutining inhibition assay (HAI) in serum samples revealed the early presence of hemagglutinin inhibiting A/California/07 (HAI) antibodies as soon as day 7 PI (Fig. 5C). This is mostly an IgM response with IgG following at later times (Data not shown).

The lungs of California/07 infected ferrets had severe and rapid histopathological changes (Figs. 1D and 2A-L). Early in infection, atelectasis was widespread with intra-alveolar exudate varied from mild to severe (Fig. 2A-C). Bronchial epithelium showed moderate intracellular edema with occasional supra-epithelial exudate containing acute inflammatory cells (Fig. 2E-F). At day 1 post infection, marginating leukocytes invaded the vascular walls and the intense inflammatory infiltration over the ensuing week led to vascular adventitial destruction (Fig. 2K). Early bronchial epithelial necrosis was observed focally on day 1, was widespread by days 2 and 3, peaked between days 3 and 5 and was relatively quiescent by day 7 PI (Fig. 2D-G and I-K). Bronchial submucosal glands also had acute inflammation that was maximal by day 3, but persisted for 1-2 weeks (Data not shown). Two weeks PI, chronic inflammation was mostly restricted to submucosal regions (Data not shown). Intra-alveolar exudate and hemorrhage was pervasive by days 3-5 and were replaced by cellular consolidation with hyperplasia of type-II pneumocytes between days 5-7. At one-week PI, central lung regions were more consolidated while peripheral alveoli contained exudate and inflammatory cells. Gram stain of the tissue did not reveal any organisms in the lung at any time after infection.

Beginning at day 1 post infection, immunocytochemistry showed influenza viral antigens predominantly along the apical surfaces of bronchi and bronchial submucosal glandular epithelium (Fig. 2J-L). Immunocytochemical staining is pronounced at day 1, peaked at day 3 and rapidly declined by days 5 and 7 (data not shown). Viral infection was associated with epithelial necrosis and sloughing into the airways. This process was sufficiently rapid and severe that many airways were obstructed by necrotic plugs debris (Fig. 2L). Immunocytochemical staining of gut, lymphoid tissue, pancreas, and brain demonstrated no evidence of viral antigens or inflammatory lesions (data not shown).

Pathological features of the infection with A/Brisbane/59/2007

Ferrets infected with Brisbane/59 displayed mild clinical symptoms such as 5% maximum weight loss, nasal discharge, sneeze and protracted fever, and without lethality (table 1 and Supplementary Fig. 1). The virus was detected in the nasal wash of infected ferrets and the viral load peaked on day 3 post infection. Hemagglutinin inhibiting A/Brisbane/57 antibodies were detected in serum samples on day 7 PI, and higher titers on day 14 PI were found (Fig. 5C). Histological assessment in the nasal turbinates displayed inflammatory infiltrate on days 2 and 5 post infection, with sloughing in some areas (data not shown). In the lung tissue of Brisbane/59 infected ferrets, the viral load was found to be below detection levels (table 1, Fig. 1A and Fig. 1C) and no histological abnormalities were found (Fig. 1E).

Gene expression profiles in lung tissue of infected animals

To determine the temporal relationship between host immune responses and the progression and resolution of 2009-H1N1 clinical disease, we conducted a longitudinal study of expression patterns of host immune genes in the lungs of infected animals using high density microarrays (Figs. 3A and 4A-B). On day three post infection, there was a significant increase in interferon stimulated genes (ISG) (Fig. 3A), a functional group that included numerous genes for antiviral proteins (e.g. OAS1&2) and transcriptional regulators (e.g. STAT1) (Malmgaard, 2004). In contrast to the robust expression pattern observed for 2009-H1N1, lungs from animals infected with seasonal H1N1 Brisbane/59 expressed a restricted pattern of ISGs. On day 5, only 12 ISGs were upregulated (mean expression levels were at least 1.5-fold respect to the control group), while lungs from ferrets infected with California/07 displayed 26 upregulated ISGs. Concomitant with early expression of ISGs, California/07 induced the expression of multiple chemokines including CCL2, CCL8, CCL13, CCL19, CXCL7 and CXCL10, and also inflammatory cytokines such as TNFα and IL-1β (Figs. 4A and 5A). These mediators participate in the recruitment of multiple leukocyte subsets including NK cells, monocyte/macrophages, T cells and granulocytes (Charo and Ransohoff, 2006). We observed that their expression patterns coincided with the infiltration of inflammatory mononuclear cells and polymorphonuclear leukocytes in the lungs during infection with California/07 (Fig. 2A-L). Lungs from Brisbane/59-infected animals showed involvement of a limited number of cytokine/chemokines (Fig. 4A) and lower expression levels compared with California/07 (Fig. 5A). The modest lung response during the infection with Brisbane/59, in terms of ISGs, cytokines and chemokines, is likely due to a lack of local host immune activation as the lungs had no detectable viral load (table 1). In the lungs of California/07-infected ferrets, a striking decrease in the expression of ISGs and the chemokines CCL2, CCL8, and CXCL10 takes place on day 7 (Figs. 3A, 4A and 5A). Clinically, on day seven, animals reached the nadir of weight loss and began a slow, but progressive increase, in weight. In contrast to the reduction in ISG and chemokine expression on day seven, expression of genes associated with the effector phase of adaptive immunity (e.g. Ig lambda, CD3, Granzyme A, Granzyme B, and Granzyme H) are increased in 2009-H1N1 infected lungs (Fig. 4B). Similarly, increased expression of granzyme A, B, and H are also observed in peripheral blood on Day 7 (Fig. 5B). The shift from innate to adaptive immune responses is further supported by the appearance of hemagglutinin inhibiting serum antibodies on day 7 (Fig. 5C).