Authors:	Ivanova, Elena
Title:	Dissecting tail-anchored protein localization signals with iPAL (imaging pooled-to-array library) scanning
Online publication date:	30-Nov-2022
Year of first publication:	2022
Language:	english
Abstract:	Most proteins in a eukaryotic cell are synthesized by ribosomes in the cytosol and then need to be selectively targeted to different subcellular compartments to fulfill their functions. The core of protein targeting is the recognition of a targeting signal within the nascent protein by a targeting factor for the destination organelle. It is a complex process that involves multiple overlapping pathways and errors in protein targeting occur even under optimal conditions, leading to protein mislocalization. Mutations in localization signals can also result in protein mislocalization and cause severe genetic disorders. Thus, it is important to understand the nature of localization signals, the mechanisms of protein targeting, and the systems involved in the quality control (QC) of protein targeting. In this study, I describe iPAL scanning (imaging pooled-to-arrayed libraries), a deep mutational scanning approach to dissect protein localization signals. In iPAL, pooled libraries of protein variants with alterations in a potential localization signal are constructed by homologous recombination. Each protein variant is fused to a fluorescent protein (mNeonGreen) to read out its localization with fluorescence microscopy. Using in vivo barcoding and deep sequencing, pooled libraries are converted into verified arrayed libraries, followed by high-throughput fluorescence microscopy to determine the localization of each variant. Importantly, iPAL is not limited by the number of phenotypic bins and thus can be used to dissect complex localization signals. Due to the arrayed library format, all generated protein variants can be directly used for downstream applications, e.g., co-localization experiments or genetic screens. I applied iPAL scanning to investigate localization determinants of tail-anchored (TA) proteins in budding yeast. TA proteins carry a single α-helical transmembrane domain (TMD) at their C-terminus, which enables them to localize to the endoplasmic reticulum (ER), mitochondria, Golgi, nuclear envelope, vacuole and plasma membrane (PM). However, the features of these TMDs crucial for targeting specificity are not well defined. Using iPAL, I generated 1350 unique TMDs with variations in amino acid composition, TMD length and charge of the flanking residues. By analyzing the localization and biophysical properties of TMD variants, I found that TMDs of most ER-resident TA proteins of average hydrophobicity are alone sufficient for correct localization. In contrast, a combination of properties ensures mitochondrial and PM localization. Mitochondrial localization requires low hydrophobicity of the TMD and a high positive charge at the C-terminal extension. Two different types of features lead to PM localization: long TMDs of very high hydrophobicity flanked with positively charged residues at the N-terminus or short cysteine-rich transmembrane modules (CYSTM) of low hydrophobicity. Long length and high hydrophobicity of the TMDs from the first group might be related to the greater thickness of the PM and positively charge flanking residues might be required for the electrostatic interaction with phosphatidylinositides at the PM. It is possible that cysteine residues are important for the oligomerization of CYSTM proteins or for their interaction with TMDs from other PM proteins in the bilayer. However, it remains a possibility that CYSTM proteins are not integrated into the PM but are associated only peripherally. In a second line of research, I combined iPAL with genetic screens to demonstrate the downstream application of iPAL and investigate which TMD features are recognized by protein targeting and QC systems. I performed a screen for protein turnover factors by introducing the knockouts of Get3 and Emc3 targeting factors Tul1, Hrd1, Asi1 and Doa10 E3 ligases and Msp1 and Spf1 protein dislocases into an mNeonGreen-TMD library. I observed that substrates of ER and mitochondrial protein QC, namely of Doa10, Msp1 and Spf1, tend to have TMDs of low hydrophobicity. TMDs of low hydrophobicity are generally not recognized by the SRP (signal recognition particle) and GET (guided entry of TA protein) targeting factors and, thus, might be more frequently subjected to QC. Altogether, my work reveals that different combinations of TMD properties, including hydrophobicity, length and presence of positively charged flanking residues, provide the targeting specificity for ER, mitochondrial and plasma membranes. I anticipate that the iPAL approach will have a wide range of applications and help to understand the interplay between protein sequence and visual phenotypes.
DDC:	500 Naturwissenschaften 500 Natural sciences and mathematics 570 Biowissenschaften 570 Life sciences
Institution:	Johannes Gutenberg-Universität Mainz
Department:	FB 10 Biologie
Place:	Mainz
ROR:	https://ror.org/023b0x485
DOI:	http://doi.org/10.25358/openscience-8039
URN:	urn:nbn:de:hebis:77-openscience-23228f98-cb49-429b-8448-ee4107bad58a5
Version:	Original work
Publication type:	Dissertation
License:	In Copyright
Information on rights of use:	http://rightsstatements.org/vocab/InC/1.0/
Extent:	vi, 127 Seiten; Illustrationen, Diagramme
Appears in collections:	JGU-Publikationen