Incidence of primary graft dysfunction after lung transplantation is altered by timing of allograft implantation

The importance of circadian factors in managing patients is poorly understood. We present two retrospective cohort studies showing that lungs reperfused between 4 and 8 AM have a higher incidence (OR 1.12; 95% CI 1.03 to 1.21; p=0.01) of primary graft dysfunction (PGD) in the first 72 hours after transplantation. Cooling of the donor lung, occurring during organ preservation, shifts the donor circadian clock causing desynchrony with the recipient. The clock protein REV-ERBα directly regulates PGD biomarkers explaining this circadian regulation while also allowing them to be manipulated with synthetic REV-ERB ligands.


Suppl.
ex-vivo lung perfusion. All inclusion and exclusion criteria were assigned before analysis began. Lung allocation score (LAS) is assigned to patients in the Toronto transplant program by a multidisciplinary team assessment. A score of 1 is assigned where the recipient's condition is stable or deteriorating slowly, a score of 2 is assigned where the recipient is deteriorating rapidly, a score of 3 is assigned when the patient has been admitted to intensive care. PGD score was used from the same database, if this was unclear or had not been assigned then the patient's record was examined directly from the electronic record. Organ reperfusion time was defined as the earliest time-point after organ retrieval that the donor lung(s) were reperfused with recipient blood, this was usually recorded on the anesthetic chart. Patients were split into two groups according to organ reperfusion time defined by the pilots study performed at a Transplant center in the U.K. Primary outcome was defined as the difference in incidence for primary graft dysfunction (grades 2&3) recorded at four time-points after transplantation for patients whose organs were reperfused during the high risk window (0400-0759) compared to organs reperfused between 0800-0359. Secondary analysis repeated the analysis for primary graft dysfunction grade 3 alone, and also for a sub-cohort of double lung transplant recipients that were not relatively contra-indicated due to BMI or age. The latter analysis was performed to control for the effects of operation length, age and obesity on circadian rhythms.
Ethical Approval: Ethical approvals for both human and animal studies were obtained. and Donor Age. LOESS regression was implemented using R to smooth the PGD incidence data taking into account the individual patient reperfusion times (suppl. fig. 1b). The span parameter for the LOESS was chosen using the elbow method on the residual sum of squares.
Generation of monocyte derived macrophages for microarray analysis: 50 mls of blood was taken from six human volunteers. These volunteers did not have any active health complaints, had not taken any medication for 6 weeks prior to sampling and were not involved in any other clinical studies. The blood was diluted 1:2 with RPMI 1640 (Gibco) and layered over ficoll-paque. This solution was then spun for 30 minutes at 4⁰C after which the interphase layer was taken off. This was then centrifuged for 5minutes at 300g and the pellet was re-suspended in RPMI 1640 (supplemented with penicillin/ streptomycin and 10%FBS) at 5x10 6 cells/ml. All the cells were then plated out into 24 well plates and left to adhere for 1 hour at 37⁰C, following which each well was washed three times and the plates were incubated overnight. Media was changed in all the wells the following morning and also every alternate day until the cells had become monocyte derived macrophages after 10 days. The cells were then stimulated with LPS (10μg/ml) in the presence or absence of GSK 4112 (10μM). After 6 hours the cells were lysed using a RNAeasy kit (Qiagen) and analysed on a microarray platform using the affymetrix gene chip U133 plus 2.0 array and normalized using the RMA algorithm. Significant gene changes were defined by an average fold change of ±2 over all the six patients. Gene changes were separated into repressed genes (n=120) and induced genes (n=22). String analysis, DAVID and Ingenuity analysis was used on the dataset to infer biologically relevant pathways.
Peritoneal exudate Cells: 10mls of phosphate buffered saline was injected into the peritoneal cavity and recovered. The fluid was spun down at 500xg for 8 minutes, after which cells were resuspended in RMP 1640 (10% FBS, penicillin and streptomycin) at a seeding density 1x10 6 cells/ml. 0.5mls was added to a 24 well plate, after which the cells were left to adhere and washed three times the following morning after which the cells were stimulated with LPS for six hours and then lysed for RNA extraction.

Gene Array:
The RNA was amplified, reverse transcribed and then hybridised onto an Affymetrix U133 plus 2.0 chip. Data was normalised using the RMA algorithm 5 and log (2-based) transformation was performed. Paired t-test with limma correction was used to identify those genes significantly regulated by GSK 4112. DAVID (v. 6.7) 6 7 was used to analyse the function of the significantly regulated genes. Only genes that were differentially expression >2-fold were included. Results have been uploaded to array express with the id E-MTAB-6478

Animals
All animals were maintained in 12h:12h light:dark (LD) with food and water supplied ad libitum.
Male C57BL/6J mice were purchased from Charles River (UK). mPER2::luc transgenic mice and global Reverbα KO mice were previously described.

Chemicals
Lipopolysaccharide expressed by the E.Coli strain 026:B6 was obtained from Sigma-Aldrich Corp (Missouri USA). The REVERB agonists, GSK 4112, 2667 and 2945 have been previously described 8 9 and were a kind gift from GSK.