/Products/RNAi (RNA Interference)/shRNA/Supporting Data
Products > RNAi (RNA Interference) > shERWOOD-UltramiR shRNA > Supporting Data

shERWOOD-UltramiR shRNA

Superior Knockdown and Specificity with shERWOOD-UltramiR shRNA

shERWOOD-UltramiR shRNA designs are stringently selected by the shERWOOD algorithm and expressed from an optimized microRNA scaffold for increased small RNA processing.  These designs outperform early generation shRNA libraries providing more efficient knockdown when compared at the individual gene level or in pooled shRNA screens.   

The data shown below is adapted from the Hannon Lab publication describing the development and validation of the shERWOOD algorithm and the shERWOOD-UltramiR shRNA library. See Knott et al. 2014. A computational algorithm to predict shRNA potency (Molecular Cell. December 2014).

Consistent knockdown efficiency relative to early generation shRNA designs

The combination of the shERWOOD shRNA designs and UltramiR scaffold consistently produces potent shRNA even when expressed from a single integration in the genome. Knockdown  efficiencies of shERWOOD-UltramiR hairpins were benchmarked against existing TRC and GIPZ shRNAs targeting three different genes. shERWOOD-UltramiR designs produced very potent and consistent knockdown at single copy relative to  available TRC and GIPZ hairpins targeting the same genes (Knott et al., 2014). 

The consistent performance seen with shERWOOD-UltramiR shRNA provides the clearest results for single gene interrogation and is essential for optimal deconvolution in pooled shRNA screening and greater confidence in results. 

Figure 1. Individual shRNA from the shERWOOD-UltramiR, Hannon-Elledge V.3 (GIPZ) and TRC targeting the mouse genes Mgp, Slpi and Serpine2 were compared based on knockdown efficiency by measuring knockdown at the mRNA level.  Dotted line represents 70% knockdown.  Mouse 4T1 cells were transduced at single copy and knockdown was tested following selection. shRNA from the TRC and Hannon-Elledge V.3 (GIPZ) were expressed from the pLKO.1 and GIPZ lentiviral vectors (respectively) and the shERWOOD-UltramiR shRNA are expressed from the LMN retroviral vector.  (Data adapted from Knott et al. 2014) 

Improved specificity versus classic stem loop shRNA

On target specificity of the shERWOOD-UltramiR shRNA shown in Figure 1 was compared to that of TRC shRNA that showed potent single copy knockdown.  RNA-seq analysis was performed on cell lines expressing shRNA targeting Slpi and Mgp. The graphic below shows a heat map of the number of genes with differential expression (fold change > 2 and FDR <0.05) from each of the pairwise comparisons.  shERWOOD-UltramiR shRNA showed relatively few differences (less than 25 genes) while TRC designs show approximately 250 genes altered in cells expressing shRNA targeting Slpi, and over 500 in the line expressing the Mgp shRNA.  The two TRC shRNA selected for the comparison were the only two targeting those genes which provided significant knockdown (see  shRNA in Figure 1: TRC-Mgp-1 and TRC-Slpi-1 versus all four shERWOOD-UltramiR shRNA for each gene.  No shRNA targeting Serpine2 were compared due to the lack of a TRC shRNA producing significant knockdown for that gene.)

Figure 2. Heat map showing the number of differentially expressed genes (> 2-fold change and FDR <0.05) identified  through pairwise comparisons of the cell lines corresponding to (A) Mgp and (B) Slpi knockdown by the shERWOOD-UltramiR shRNA and the only TRC shRNAs showing significant knockdown for the two genes (TRC-Mgp-1 and TRC-Slpi-1). (Adapted from Knott et al. 2014)

This data is consistent with other publications showing classic stem loop shRNA can cause significant off-target effects and toxicity. Several reports (Beer et al 2010, Castanatto et al 2007, Pan et al 2011, Baek et al 2014, Knott et al 2014) have shown that off-target effects can be ameliorated by expressing the same targeting sequence in a primary microRNA scaffold (shRNA-miR).
Figure 3.  Percentage of shRNA targeting essential genes that depleted in each of the TRC, GIPZ, shERWOOD or shERWOOD-UltramiR shRNA screens.">
Figure 3.  Percentage of shRNA targeting essential genes that depleted in each of the TRC, GIPZ, shERWOOD or shERWOOD-UltramiR shRNA screens.

More potent shRNA per gene enables superior hit stratification

To benchmark the shERWOOD algorithm design against the early generation TRC and Hannon Elledge (GIPZ) shRNA designs, a large scale screen was performed using each of these designs to target 2200 genes that were likely to impact growth and survival based on gene ontology. Inclusion as a hit required that at least 2 shRNA for that gene were depleted. The box plot shows the average percentage of shRNA per gene that were scored as hits. The data shows that the shERWOOD 1U shRNA designs produce a higher percentage of potent shRNA per hit compared to early generation shRNA designs (Knott et al 2014). This  makes for more confidence in screen hits and ultimately fewer false positives and negatives from shRNA screens.