Systems Biology and reconstruction of Cellular Networks is a major challenge to improve the understanding of the different clinical presentation of acute Cellular Rejection. As a collaborative project between our lab and the Columbia Genome Center, directed by Dr. Andrea Califano who is also Mentor of Anshu Sinha, PhD student, they are actively involved in the development of this project. We are currently exploring the CARGO study database to reconstruct the network from peripheral blood mononuclear cells during rejection and quiescence. Ongoing projects are focused on the reconstruction of the CD4 T-cell network of patients undergoing heart transplantation and identification of individuals modules of the PBMC subnetworks.

Reconstructing the CREB-Network from 285 leukocyte array in CARGO yields insight into transcription factor function during quiescence and rejection (abstract as presented at AST, July 2006)

Background To better understand the mechanisms of rejection, highthroughput genomics and computational biology combine to powerful tools to identify specific regulatory key genes and modules. These are potential targets for designing immunomodulatory interventions. Method: Using ARACNe (Algorithm for the Reconstruction of Accurate Cellular Networks), we screened the 285 leukocyte microarrays utilized in the Cardiac Allograft Rejection Gene expression Observational (CARGO)-study to reconstruct the peripheral PBMC-network of heart Tx recipients.
Results Out of the 7300 genes represented on the microarray, we retained those fulfilling a 70% completeness criterion. Missing values were imputed by using K-nearest (10) neighbors and the resulting matrix (4688 genes and 285 arrays) was analyzed using ARACNe. We identified several hubs implicated in the regulation of the immuneresponse including cAMP-responsive-element-binding protein (CREB), one of the best characterized transcription factors. To identify the CREB subnetwork, we selected from the complete network only those genes that had significant pair-wise mutual information with the available probe sets for CREB (P < 1e-7). Twenty-nine genes were directly connected to CREB. Eleven of the 29 (38%, Fisher s extact test 0.0045) predicted first neighbors were previously reported as CREB targets by Chromatin-Immuno-Precipitation (CISH, TOP1, MARK3, SON, FADD, TSG101, ZNF146 ,HSPC111, MDS025, GABPB2), either known to play a role on immuneregulation like CISH and FADD or with unknown function.

Conclusion Computational biological approaches allows the identification of molecular modules from complex cellular networks in transplantation medicine. These may serve as targets for immunomodulatory interventions.

 Putative regulation of CREB during rejection and quiescence.