Novel anti-tumour necrosis factor receptor-1 (TNFR1) domain antibody prevents pulmonary inflammation in experimental acute lung injury

Background Tumour necrosis factor alpha (TNF-α) is a pleiotropic cytokine with both injurious and protective functions, which are thought to diverge at the level of its two cell surface receptors, TNFR1 and TNFR2. In the setting of acute injury, selective inhibition of TNFR1 is predicted to attenuate the cell death and inflammation associated with TNF-α, while sparing or potentiating the protective effects of TNFR2 signalling. We developed a potent and selective antagonist of TNFR1 (GSK1995057) using a novel domain antibody (dAb) therapeutic and assessed its efficacy in vitro, in vivo and in a clinical trial involving healthy human subjects. Methods We investigated the in vitro effects of GSK1995057 on human pulmonary microvascular endothelial cells (HMVEC-L) and then assessed the effects of pretreatment with nebulised GSK1995057 in a non-human primate model of acute lung injury. We then tested translation to humans by investigating the effects of a single nebulised dose of GSK1995057 in healthy humans (n=37) in a randomised controlled clinical trial in which subjects were subsequently exposed to inhaled endotoxin. Results Selective inhibition of TNFR1 signalling potently inhibited cytokine and neutrophil adhesion molecule expression in activated HMVEC-L monolayers in vitro (P<0.01 and P<0.001, respectively), and also significantly attenuated inflammation and signs of lung injury in non-human primates (P<0.01 in all cases). In a randomised, placebo-controlled trial of nebulised GSK1995057 in 37 healthy humans challenged with a low dose of inhaled endotoxin, treatment with GSK1995057 attenuated pulmonary neutrophilia, inflammatory cytokine release (P<0.01 in all cases) and signs of endothelial injury (P<0.05) in bronchoalveolar lavage and serum samples. Conclusion These data support the potential for pulmonary delivery of a selective TNFR1 dAb as a novel therapeutic approach for the prevention of acute respiratory distress syndrome. Trial registration number ClinicalTrials.gov NCT01587807.

cell attachment and achieved a steady state when endothelial cells became confluent (72 hours). Thus, experiments were conducted after the electrical resistance achieved a steady state of 3000 ohms. Resistance data were normalized to the initial TER (ohms) and recorded over 20 hours.
Transendothelial resistance increased over 72 hours, suggestive of monolayer integrity, and this effect was associated with lack of permeability to FITC-labelled albumin. After monolayer integrity was confirmed, Iscove's modified Dulbecco's medium (IMDM: 700 μL) was placed in the lower chamber. Cells were washed twice with PBS before the addition of GSK1995057 or dummy dAb (10nM) or vehicle control in medium for 1 hour prior to stimulation with TNF (10ng/mL)for 4 hours at 37C, 5% CO2. All conditions were performed in triplicate. Following a further 2 washes in PBS (to remove basal chemokine release from stimulated HMVEC-L), the medium in the lower chamber was changed to IMDM containing IL-8 (50 ng/mL; R&D Systems). Peripheral blood neutrophils were isolated from healthy volunteers using dextran sedimentation and discontinuous plasma-Percoll® gradient centrifugation, as previously described (E1).
Neutrophils were suspended at 1x106 cell/mL in serum-free IMDM, and Leukotracker™ (Cambridge Bioscience, UK) added at the manufacturer's recommended dilution. Cells were incubated at 37C, 5% CO2 for 1 hour, washed twice and re-suspended at 1x10 6 cells/mL in serum-free media; 200μL of this solution was added to the upper chamber, and neutrophils left to migrate for 3 hours. After 3 hours, the medium in the upper chamber was discarded and non-migratory cells removed with a cotton swab. The Transwell filter was then placed in a well containing lysis buffer and medium from the lower chamber containing migrated neutrophils. The plate was gently shaken for 5 minutes at room temperature before the fluorescence was read, in triplicate, at 480 nm/520 nm.

Endothelial ligand cell surface expression
HMVEC-L were seeded into 6-well plates (1.2 x 10 5 cells/well) and grown to confluence over 3 days (37°C, 5% CO2). The effect of TNFR1 antagonism on adhesion molecule expression was assessed following 1 hour preincubation with GSK1995057 or dummy dAb (10nM) or vehicle control followed by TNF (10ng/mL) stimulation for 4 hours. Cells were washed three times with Hanks Balanced Salt Solution, and harvested using non-enzymatic cell dissociation buffer (Sigma-Aldrich). The cell suspension was placed on ice and centrifuged at 400 g (5 minutes, 4C). The supernatant was removed and the cell pellet re-suspended in blocking buffer (5% fetal bovine serum, 0.5% bovine serum albumin in PBS) for 10 minutes at 4C. Antibodies to E-selectin (clone HCD62E, Biolegend), VCAM-1 (Clone STA, Biolegend) and ICAM-1 (Clone HA58, Biolegend), along with appropriate isotype and unstained controls were added and the cell incubated in the dark (30 minutes, 4C). Cells were washed twice and re-suspended in stain buffer (PBS +1% FBS + 0.09% sodium azide) before being transferred to FACS tubes. Samples were analyzed on a FACSCanto™ II flow cytometer (BD Biosciences).

Study Design & Procedures
Adult male Cynomolgus monkeys weighing between 3.6-5.7kg, were chosen for this study because it is a species that has shown pharmacologic responses to GSK1995057, and is commonly used for efficacy studies in the LPS-challenge model. Determination of group size was based on feasibility but with a target of including at least n=5 animals per treatment group based on an assessment of the variability of planned BAL endpoints from historical data in this model. Animals were assigned to treatment groups by the study director based on baseline response to LPS in Session 1 to ensure a balance of LPS responsiveness across treatment groups following the design schematic below: Sedation was induced with intramuscular ketamine (5-10 mg/kg) to allow placement of a venous catheter, and induction of anesthesia with a bolus of intravenous propofol (3mg/kg). Anesthesia was maintained with a continuous intravenous infusion of propofol (0.1-0.2 mg/kg/min). Treatments were administered 1 hour before LPS challenge. The doses of vehicle (Group 2) and GSK1995057 (Groups 3-5) were delivered by aerosol inhalation using a Pari LC jet nebulizer (Pari, USA) at a dose volume of 0.5 mL/kg. Each treatment was delivered at a calculated dose time determined for each animal based on body weight. Therefore, based on mean body weights at the time of dosing, the total delivered doses (i.e. nominal doses) in this study were 1.2mg of fluticasone propionate in Group 1, and 0.043mg, 0.45mg, and 4.7mg of GSK1995057 in Groups 3, 4, and 5, respectively. Dose times ranged between 9 and 14.25 minutes and an additional 3.5 mL of the dosing formulation was added to the nebulizer chamber for each dose to compensate for dead space and possible foaming during the nebulisation.

Bronchoalveolar lavage
A paediatric fiberoptic bronchoscope was wedged into a segmental bronchus, and 3 aliquots of sterile 0.9% saline (20 mL) instilled. Bronchoalveolar lavage return was measured, and samples placed on ice until transferred to the laboratory for processing. BALF was centrifuged and the supernatant removed and stored at -80C for subsequent analysis. The cell pellet was re-suspended in phosphate buffered saline (2 mL). Cytospins were prepared and stained with Wright-Giemsa, before a minimum of 200 nucleated cells were counted to determine the differential cell counts.

Subjects
Healthy subjects were recruited by advertising. Screening consisted of a history and physical examination, blood investigations, ECG, and spirometry (the full inclusion and exclusion criteria and study schedule are outlined in Supplementary Table E1 and Supplemental data E1).

Study Design
The study was a randomized, placebo-controlled clinical trial. The participants, site investigators and the staff performing the laboratory assays were blinded to the treatment allocation. The site clinical trials pharmacist and central GSK study team were unblinded to treatment allocation. Eligible subjects received a unique subject number.
Subjects were assigned to receive a total nebulised dose of GSK1995057 (26 mg) or identical placebo (1:1) in accordance with the randomization schedule generated by Clinical Pharmacology and Biometrics at GSK. When an eligible subject was recruited, the clinical trials pharmacist allocated the subject to the designated treatment group, maintaining blinding. Details of the study medication are given in Supplementary table 2. The nebulizer solutions had an identical appearance. Study drug was administered using a Pari eFlow nebuliser. One hour after the study medicine was taken, subjects inhaled LPS (total dose 50 μg) using a large-volume reservoir delivery system as described (E2). BAL was performed 6 hours after LPS inhalation (7 hours after dosing GSK1995057). Blood was collected as defined in the schedule of events in the online supplement (Supplementary Table 1). LPS inhalation (Escherichia coli serotype O26:B6; Sigma Chemicals, Poole, Dorset, UK) was dissolved in endotoxin free sterile 0.9% saline and inhaled via an automatic inhalation-synchronized dosimeter nebuliser (Spira, Hameenlinna, Finland). This delivers particles of a mass median aerodynamic diameter (MMAD) of 10 μm as described previously (E3). The dosimeter produces a calibrated aerosol of 8 μL at each slow inhalation starting from functional residual capacity (FRC) to total lung capacity (TLC). Each subject performed five successive inhalations of the LPS solution (1.25 mg/mL) through the mouthpiece with a nose clip in place. The total dose of inhaled LPS was 50 μg.
The BAL procedures, as well as the processing of BALF and blood, were performed as described below. Subjects stayed overnight in a clinical research unit following dosing with LPS.

Bronchoalveolar lavage
Bronchoalveolar lavage was performed 6 hours after LPS inhalation according to standard guidelines (E4). Three successive 60 mL aliquots of 0.9% saline were instilled into a sub-segment of the right middle lobe and each aspirated immediately with low suction. Bronchoalveolar lavage fluid (BALF) return was measured and immediately placed on ice until transferred to the laboratory for processing. BALF was centrifuged at 900g for 5 minutes at 4°C.
The supernatant was removed and stored at -80C for subsequent analysis. The cell pellet was re-suspended in phosphate buffered saline (PBS: 10 mL) supplemented with 1% bovine serum albumin (BSA) (Sigma Chemicals).
Total cell count was determined using a hemocytometer. Cell viability was measured by the ability of live cells to exclude Trypan blue. Cytospins were prepared and stained with Speedy-Diff (Clin-Tech Ltd, UK). A single cytologist who was blinded to treatment allocation performed the differential cell counts. Epithelial lining fluid (ELF) concentrations of GSK1995057 were derived by normalising to concentrations of ELF urea (E5).

Plasma samples
Blood was collected into K3 EDTA tubes, mixed gently, and placed immediately on ice until processed. Blood was centrifuged at 1500g for 10 minutes at 4°C. The plasma was removed and stored at -80C for subsequent analysis.

Serum Samples
Blood was collected into serum separator tubes, mixed gently and maintained at ambient temperature for 30 minutes to 1 hour to allow for complete clotting before centrifugation at 1500g for 15 minutes at 4°C. The serum was removed and stored at -80C for subsequent analysis.
This GSK clinical trial (GSK study ID: 116236) was prospectively registered on ClinicalTrials.gov (NCT01587807) and the clinical trial protocol and results summaries are available at the GSK clinical trial registry website (https://www.gsk-clinicalstudyregister.com).              Free and total sTNFR1 were measured to monitor the interaction of GSK1995057 with sTNFR1. The assay for free sTNFR1 detects sTNFR1 not bound to GSK1995057, whereas the assay for total sTNFR1 detects both free sTNFR1 and sTNFR1 bound to GSK1995057. No error bar shown in panel (a) for GSK1995057 group as all values measured were below the lower limit of quantification of the assay (61.73 pg/mL).