Inflammation and lung injury in an ovine model of fluid resuscitated endotoxemic shock

A description of the model and methodology has been published previously [24]. Briefly, hyperdynamic endotoxemic shock was induced by administration of an escalating dose of lipopolysaccharide (LPS; E coli serotype O55:B5; Sigma-Aldrich, St Louis, USA) up to 4 μg/kg/h over 4 h (total LPS dose 11.25 μg/kg). Adequate endotoxemia was confirmed by the occurrence of systemic hypotension with a mean arterial pressure (MAP) less than 60 mmHg after 3 h of endotoxin infusion. In the final hour of endotoxemia animals received either fluid resuscitation with 40 mLs/kg of 0.9% saline over an hour (FR; n = 8) or commenced protocolised vasopressor support (NFR; n = 8). Noradrenaline was started 60 μg/mL in 5% dextrose (Hospira, Lake Forest, IL, USA) to maintain a MAP between 60 and 65 mmHg. If noradrenaline reached a predetermined 20 μg/min, vasopressin (PPC, Richmond Hill, ON, Canada) was commenced at 0.8 units/h and increased to a maximum of 1.6 units/h if hypotension persisted. The administration protocol was the same for both groups. Post-fluid resuscitation, both groups were monitored for 12 h after the end of endotoxin infusion.

Bronchoalveolar lavage (BAL) samples were collected from the lower lobe of the left lung at baseline, pre-LPS infusion (or sham) (T1), post-endotoxemia/pre-resuscitation (T2) and 0 and 12 h post-saline resuscitation for ELISA analysis and aliquots cytospun onto slides for inflammatory cell counts. Arterial blood samples were collected for flow cytometry at baseline, T2, 0, 2 and 12 h post-saline resuscitation. Full blood counts were performed using the veterinary mode of the AcT diff™ haematology analyser (Beckman Coulter Australia Pty Ltd., NSW, Australia). Blood films were prepared, stained with Quick Dip (POCD Scientific, NSW, Australia) and used for a manual white blood cell differential count. Neutrophil numbers were calculated based upon the total white cell count from the AcT diff™ and the white blood cell differential count. Tissue samples were removed for histopathology, gene and protein expression studies. Pulmonary oedema was determined by measuring the wet-to-dry-weight of post-mortem left and right upper and lower lobe lung tissue. Serum albumin and total protein were assessed using commercially available kits on the COBAS Integra 400 blood chemistry analyser (Roche Diagnostics, Australia) while total protein was performed on BAL samples using the BCA (bicinchoninic acid) protein assay kit (Pierce Technology, USA).

The concentration of IL-6, IL-1β, IL-8, IL-10 and TNF-α in lung tissue homogenate and BAL fluid was quantified by in-house ELISAs using methods published previously [25‐27]. Positive internal controls were used to ensure that inter- and intra- plate variability was < 10%, and confirm the precision and accuracy of all ELISA assays. Hyaluronan was measured in BAL fluid using a standard sandwich ELISA (R&D Systems, Minneapolis, USA).

Total RNA was isolated from lung tissue using the RNeasy Mini Kit (Qiagen, VIC, Australia). Real-time Quantitative PCR was performed using primers for IL-6, IL-1β, IL-8, TNF-α, IL-10, MMP-1, 2, 8, 9, 13 and TIMP-1, 2, (PrimerDesign, Southampton, UK). The stability of six candidate normalization genes (ACTB, GAPDH, HGPRT, PGK1, PPIA and RPLP0) was evaluated using real-time quantitative PCR. The Qbase PLUS program was used to identify the most stably expressed housekeeping genes and all data was subsequently normalized to a geomean of PGK1 and ACTB (Life Technologies, Grand Island, USA).

EDTA whole blood collected at each time-point was stained with a fluorescein isothiocyanate (FITC) conjugated monoclonal antibody against CD11b (Bio-Rad, CA, USA). Red blood cells were then lysed (FACS-Lyse; BD Biosciences, San Jose, California) and samples were analysed using a FACS Canto flow cytometer (BD Biosciences). Flow cytometry data was analyzed using FlowJo (Tree Star Inc., Ashland, Oregon) with granulocytes identified based upon forward and side scatter gating. The granulocyte CD11b mean fluorescent intensity (MFI) of each sample was determined as a ratio of the MFI of the stained sample to the MFI of its unstained control.

Portions of left and right lower and upper lobe lung tissue were fixed in 10% buffered formalin for at least 24 h, processed and embedded in paraffin. All sections (5 μm) were stained with haematoxylin and eosin and examined by light microscopy by two independent trained investigators blinded to the slide identity. Pulmonary oedema, vascular and alveolar features, and bronchiole pathology was graded: 0 (normal), 1 (mild), 2 (moderate) and 3 (severe). Features examined included intravascular obstruction, inflammatory cell infiltration, pulmonary congestion, thickening of the alveolar septa, presence of amorphous material and detachment of the bronchiole lining as detailed in Table 1. The upper lobe did not differ significantly from the lower lobe. Lung sections for immunohistochemistry were dewaxed, rehydrated in a series of ethanol and water, then stained with macrophages antibody [MAC 387] (Genetex, CA, USA), MMP-2, MMP-9 (Abcam, Cambridge, UK), TIMP-1 and TIMP-2 (Abbiotec, CA, USA). Primary antibodies were applied and incubated overnight at 4 °C. The sections were then washed and the Vector universal ABC secondary antibody kit (Vector Laboratories, CA, USA) was used to link the primary antibody to the chromogen for 1 h. A positive reaction was detected with 3,3′-diaminobenzidine tetrahydrochloride (Sigma-Aldrich, St Louis, USA), which produces a brown colour at the site of the reaction. Sections were counterstained with haematoxylin, dehydrated in ethanol and xylene and mounted in DePex (ThermoFisher Scientific, Scoresby, Australia). Negative controls were processed using the same methodology, but without the primary antibody. A rabbit IgG polyclonal isotype control (Abcam, Cambridge, UK) was used to ensure staining was not caused by non-specific interactions of immunoglobulin molecules with the sample.

Images for immunohistochemistry were photographed at a magnification of 250x and visualized using the AxioVision 4.7 Image Analysis system (Carl Zeiss, Germany). Semi quantitative analysis of macrophage number was performed on 20 fields of photographed images.

Lung wet-to-dry weight ratio was used as a measure of increased capillary permeability and extravascular lung fluid. There was no significant difference between the left and right lobe of lung tissue. FR resulted in a significant increase in lung fluid compared to NFR (Fig. 1a, p = 0.005). Serum levels of albumin and total protein (Fig. 1b and c) decreased throughout the experiment with significantly lower levels in the FR group compared to NFR from T0 onwards (p < 0.05 and p < 0.01 respectively). BAL total protein levels were unchanged.

Linear regression analysis demonstrated that levels of IL-6 BAL increased significantly from baseline to 12 h in the FR group (Fig. 2a, p = 0.002). Levels of IL-6 in the tissue homogenate (Fig. 2b) were significantly elevated after 12 h compared to NFR (p < 0.001). IL-8 levels in BAL fluid of the FR group rose significantly from baseline to 12 h (Fig. 2c, p = 0.007) while there was a trend for increased expression in the tissue homogenate (Fig. 2d, p = 0.06). IL-1β levels in BAL (Fig. 2e) and tissue homogenate (Fig. 2f) were significantly increased in FR compared to NFR (p < 0.01 and p < 0.05 respectively). From baseline to 12 h glycocalyx GAG hyaluronan levels were also significantly elevated in BAL (Fig. 2g) with NFR (p = 0.04) and FR (p < 0.001). Although there was no significant difference at 12 h there was a trend for increased levels in the FR group (p = 0.09). There was no significant difference in IL-10 or TNF-α levels in either tissue homogenate or BAL.

Gene expression levels of pro- and anti-inflammatory cytokines and extracellular matrix proteins were measured in post-mortem lung tissue. Quantitative real-time PCR showed significant increases in IL-6 (Fig. 3a, p = 0.03) and IL-8 (Fig. 3b, p = 0.01) with fluid resuscitation. There was no change in the levels of IL-1β, TNF-α, MMP-1, MMP-2, MMP-9, TIMP-1 or TIMP-2. Gene expression levels of IL-10, MMP-8 and MMP-13 were below the limits of detection.

The number of circulating neutrophils in whole blood, as calculated from blood smear analysis and total white cell count, decreased significantly in both groups between baseline and T0 before recovering post resuscitation (Fig. 4a, p < 0.05) with the NFR group recovering more quickly. Flow cytometry data revealed the levels of the neutrophil activation marker CD11b increased significantly in both groups pre-resuscitation (Fig. 4b, p < 0.001). In the post-resuscitation period CD11b levels fell significantly with NFR (p = 0.007) but not in the FR group. Neutrophil counts in BAL cytospins, as a percentage of total cells, were increased at 12 h with NFR (Fig. 4c, p = 0.02). There were no significant differences in BAL macrophages, lymphocytes or eosinophils between the two groups. Additionally, immunohistochemical staining was used to stain lung sections with a macrophages antibody (Fig. 4d). There was no significant difference between NFR (12.4 ± 0.78%) and FR (12.1 ± 0.97%).

The left lobe from NFR (1.1 ± 0.4, Fig. 5a, c) and FR (1.2 ± 0.3, Fig. 5e, g) groups had little inflammatory infiltrate or oedema. Right lobe tissue in the NFR (1.7 ± 0.4, Fig. 5b, d) and FR (1.7 ± 0.4, Fig. 5f, h) groups had some additional damage to bronchioles with detachment of the cell lining but this was minimal and did not affect overall scoring. Immunohistochemical staining in the lung also revealed there was no change in secretion of MMP-2, MMP-9 or their inhibitors TIMP-1 and TIMP-2 with saline resuscitation.