BioProject

Background: Circulating polymers of alpha1-antitrypsin (α1AT) are chemo-attractant for neutrophils and contribute to inflammation in pulmonary, vascular and adipose tissues. Cellular factors affecting the intracellular itinerary of mutant polymerogenic α1AT remain obscure. Methods: Here, we report on an unbiased genome-wide CRISPR/Cas9 screen for regulators of trafficking of the polymerogenic α1AT-H334D variant. Single guide RNAs targeting genes whose inactivation enhanced accumulation of polymeric α1AT were enriched by iterative construction of CRISPR libraries based on genomic DNA from fixed cells selected for high polymer content by fluorescence-activated cell sorting. This approach bypassed the limitation to conventional enrichment schemes imposed by cell fixation. Results: Our screen identified 121 genes involved in polymer retention at false discovery rate < 0.1. From that set of genes, the pathway ‘cargo loading into COPII-coated vesicles’ was overrepresented with 16 significant genes, including two transmembrane cargo receptors, LMAN1 (ERGIG-53) and SURF4. LMAN1 and SURF4-disrupted cells displayed a secretion defect extended beyond α1AT monomers to polymers, whose low-level secretion was especially dependent on SURF4 and correlated with SURF4-α1AT-H334D physical interaction and with enhanced co-localisation of polymeric α1AT-H334D with the endoplasmic reticulum (ER). Conclusions: These findings suggest that ER cargo receptors co-ordinate intracellular progression of α1AT out of the ER and modulate the accumulation of polymeric α1AT not only by controlling the concentration of precursor monomers but also through a previously-unrecognised role in secretion of the polymers themselves. Overall design: CHO-K1 Tet-on_α1AT-H334D_Cas9 cells were initially transduced with a genome-wide CRISPR/Cas9 knockout library (Lib0, also called L0) comprising 125,030 single guide RNAs. α1AT-H334D expression was then induced with doxycycline followed, 24 hrs later by fixation, permeabilisation and staining with the polymer-specific monoclonal antibody 2C1 (Mab2C1). Cells were FACS sorted into 3 bins based on Mab2C1-dependent fluorescence intensity: ‘brightest, B’, ‘medium-bright, M’ and ‘dull, D’. Cell fixation precluded conventional enrichment schemes through successive rounds of phenotypic selection and expansion of the pooled cells. To circumvent this impasse, we implemented an approach based on recovery of sgRNA sequences from phenotypically-selected cell populations. Genomic DNA from the ‘brightest’-sorted cells was extracted and fragments covering integrated sgRNA sequences were PCR-amplified and used to generate a derivative CRISPR library. The derivative library (Lib1, also called L1), enriched in viral particles bearing phenotype-linked sgRNA sequences, was transduced into parental CHO-K1 Tet-on_α1AT-H334D_Cas9 cells followed by further phenotypic selection and generation of a second, enriched derivative library (Lib2, also called L2). Genomic DNA, pooled from sorted cells in the different bins at different stages of the phenotypic enrichment process and from unsorted control cells, was subjected to high-throughput sequencing and MAGeCK bioinformatics analysis (Li et al., 2014) to determine sgRNA sequence enrichment and the corresponding gene ranking list.