Preview

PDF, English - main document Download (7MB) | Lizenz: Single-population transcriptomics as a method to identify a network regulating Golgi structure by Singh, Sanjana underlies the terms of Creative Commons Attribution-NonCommercial-NoDerivatives 4.0

Abstract

This study was aimed at identifying regulators of Golgi organization based on their functional interactions with already established Golgi regulators. The groundwork to achieve this was laid by the observation of heterogeneity in Golgi morphology was observed upon siRNA treatments of proteins known to disrupt Golgi structure, whereby a population of cells could avoid the disruption and retain normal Golgi morphology. The aforementioned observation was usually attributed to inefficient siRNA knockdown in these cells. Here, this assumption was challenged, with the finding that the variability in phenotypes was not just due to transfection efficiency. Following this finding, a likely hypothesis was formulated proposing the presence of a compensation mechanism operating in these cells that enabled them to avoid Golgi disruption. Given such a scenario, comparing gene expression profiles of the differing phenotypes would highlight the proteins involved in such a compensatory mechanism, which were likely to be interactors of the protein being depleted by RNAi. Moreover, these would presumably be involved in regulation of Golgi organization at a steady state. This provided an impetus to devise a method for single-phenotype transcriptome analysis.

With this purpose, a pipeline was assembled that allowed for automated recognition of Golgi phenotypes followed by selective marking of phenotypic cells and subsequent transcriptome analysis. This was achieved by combining automated feedback microscopy, selective marking using photo-activation, cell sorting and low-input RNA sequencing. This pipeline was implemented on one candidate protein -USO1. Two phenotypes were observed upon siRNA treatment with USO1: cells displayed either a fragmented or non-fragmented/intact Golgi structure. Cells with these two phenotypes were collected and the gene expressions for these two phenotypes were compared using a differential expression analysis.

The differential expression analysis between the two phenotypes revealed about 700 genes significantly differently expressed between them, although USO1 was knocked-down to a similar extent in both populations. On further examination, a number of signaling proteins were observed to be up-regulated in cells which displayed intact Golgi morphology. These included proteins involved in chemokine mediated signaling, in addition to many kinases and phosphatases. Of particular interest was the upregulation of proteins involved in clathrin-mediated endocytosis in non-fragmented cells, in addition to two direct interactors, SEMA4F and PRKACA also being highly expressed in the same phenotype. This data suggests that there is likely a large signaling network that allows non-fragmented cells to maintain Golgi structure despite USO1 knockdown, and these pathways probably interact with USO1 at a physiological state to regulate Golgi organization. In conclusion, this study provides a base for future characterization of networks regulating Golgi architecture.