Functional screening and molecular characterisation of small proteins

Abstract

Proteins are one of the main drivers of cellular function. Most known proteins constitute various hundred amino acids, but growing evidence suggests that additionally thousands of short open reading frames (sORFs) are translated, posing the exciting possibility that many small unknown proteins are yet to be discovered. In Paper I, we designed a knockout screening pipeline to survey 11,776 curated human sORFs for a role in cancer cell line growth. We identified 17 candidates and selected the top six hits for further study. We showed that two of these candidates alleviated their knock-out phenotype when supplied in trans as GFP-fusion proteins. However, knock-out induced transcriptome alterations as assessed via RNA-seq could only partially be rescued and illustrated the difficulties in separating RNA and protein function. We described distinct interaction partners and subcellular localisations for several of our hits and finally verified translation of a pseudogene originating sORF. Overall, this study provides insights into novel putative microproteins.

Microprotein characterisation can be aggravated by the fact that many biochemical tools such as tags are tailored to canonical proteins. In Paper II, we established a minimal singleresidue terminal label (STELLA) strategy. Terminal non-canonical amino acid (ncAA)- tagging was previously not possible, due to inefficient amber suppression. Here, we overcame this issue by generating precursors that are subsequently cleaved, allowing for N- or C-terminal protein labelling with just a single ncAA, which can then be conjugated to fluorescent dyes for in-gel or (live) cell imaging applications. We successfully demonstrated this approach on a range of small proteins with different subcellular localisations. Overall, this paper provides a labelling alternative to larger conventional tags.

Finally, various known small proteins still possess unknown mechanisms of action. For example, it was unclear how viral fusion-associated small transmembrane (FAST) proteins induce cell-cell fusion given the small size of their ectodomains. A previous study showed that the p14 FAST family member exploits actin networks to drive syncytia formation, but it was unknown whether this extends to other FAST proteins. In Paper III, we showed through Co-IP, mutagenic and pharmacological assays, that aquareovirus p22, a FAST protein evolutionary divergent from p14, also exploits N-WASP mediated branched actin networks to induce fusion, albeit by utilising different molecular pathways. A p14/p22 chimera was still functional, even when N-WASP was replaced with the parallel actin nucleator formin. This suggested that the two proteins are modular, and that mechanical pressure coupled to a fusogenic ectodomain is sufficient for syncytia formation. Overall, this study illustrates a more general fusion mechanism for these minimal fusogens.

Taken together, this thesis investigated different aspects of small protein biology, developed microprotein-suitable tools, and contributed to our growing understanding of small protein-mediated function.