The shRNAmir libraries containing plasmid DNA were arrayed in 96-well plates such that each well contained one unique and identifiable shRNAmir. The library matrix was introduced into RE-luc2P-HEK293 GS-1101 price cells using a high-throughput transfection method: 100–200 ng shRNA plasmid DNA was incubated at RT for 20 min in 20 μl serum-free MEM containing 600 nl TransIT-Express reagent (MirusBio, Pittsburgh, PA) and transfected into 2×104 HEK293 cells in 100 μl DMEM/10% FBS. Approximately 30 h after transfection, culture media was replaced with DMEM/10% FBS containing 1 μg ml-1 puromycin. After 72 h of selection, during which >80% of the mock-transfected cells died, the selection media was removed, cells
were washed with PBS, and then re-suspended in 200 μl serum-free DMEM containing 1 μg ml-1 trypsin. The cell suspension (50 μl) was aliquoted to four white, clear bottom replica plates containing 50 μl DMEM/20% FBS. Cells were incubated 24h at 37°C prior to bacterial infection. For a more precise estimation of multiplicity of infection (MOI), one of the replica plates was used to calculate the number of host cells with the Cell Titer-Glo assay (Promega, Fitchburg, WI). A standard curve that correlates the ALUs to cell number (5000, 10000, 15000, 20000, 25000, and 30000 cells per well) was determined for every batch of substrate.
Two of the three remaining replica plates were infected with Y. enterocolitica WA at MOI 5 by addition of bacteria in 5 μl DMEM/10% FBS, followed by centrifugation selleck chemicals llc at 200 g for 5 min at RT. The remaining replica plate was used as a reference control (MOI 0). After 1h at 37°C, 20 μl DMEM/10% FBS containing 800 μg ml-1 of the bacteriostatic antibiotic chloramphenicol was added to each well in the plates to limit further Y. enterocolitica growth and to avoid activation of apoptotic pathways. Applying Cell Titer-Glo (Promega), we determined that the HEK293 cells infected with Y. enterocolitica at MOI 5 exhibited maximal inhibition of NF-κB-driven gene expression in response to TNF-α stimulation with no or minimal cellular toxicity. At 5 h post-infection, 25 μl DMEM/10% FBS containing
50 nM TNF-α was added to all culture plates. The screen was run once in duplicate plates. At 20h post-infection, the Cell Titer-Glo assay was used to normalize NF-κB-driven luciferase activity however to the cell titer. Arbitrary luciferase units (ALUs) were measured using the Synergy2 Multi-Mode Microplate Reader (BioTec, Winooski, VT). The relative percentage of NF-κB inhibition (R%I) by Yersinia infection was determined using the formula, R%I = [1-(ALU:MOI 5/ALU:MOI 0)]×100, where ALU:MOI 5 corresponds to the luciferase activity in bacteria-infected cells relative to ALU:MOI 0, the luciferase activity in no infection control. Hit selection criteria and validation assays Genes with at least two shRNAmir constructs that resulted in ≥40% (≥ 2 SD) decrease in R%I of NF-κB reporter gene activity were chosen for further validation.