A genetic model for in vivo proximity labelling of the mammalian secretome

Abstract

Organ functions are highly specialized and interdependent. Secreted factors regulate organ development and mediate homeostasis through serum trafficking and inter-organ communication. Enzyme-catalysed proximity labelling enables the identification of proteins within a specific cellular compartment. Here, we report a BirA*G3 mouse strain that enables CRE-dependent promiscuous biotinylation of proteins trafficking through the endoplasmic reticulum. When broadly activated throughout the mouse, widespread labelling of proteins was observed within the secretory pathway. Streptavidin affinity purification and peptide mapping by quantitative mass spectrometry (MS) proteomics revealed organ-specific secretory profiles and serum trafficking. As expected, secretory proteomes were highly enriched for signal peptide-containing proteins, highlighting both conventional and non-conventional secretory processes, and ectodomain shedding. Lower-abundance proteins with hormone-like properties were recovered and validated using orthogonal approaches. Hepatocyte-specific activation of BirA*G3 highlighted liver-specific biotinylated secretome profiles. The BirA*G3 mouse model demonstrates enhanced labelling efficiency and tissue specificity over viral transduction approaches and will facilitate a deeper understanding of secretory protein interplay in development, and in healthy and diseased adult states.

Molecular methods to study protein trafficking between organs

“…Upon introduction of a Cre recombinase, loxP‐flanked GFP is excised, which in turn allows expression of BirA*G3. To validate the system, BirA*G3 mice were crossed with the whole‐body Sox2‐Cre driver [73]. Mice fed 2000 PPM biotin for 7 days or subcutaneously injected with 100 μl of 180 mM biotin for 1 hour presented unique biotinylating patterns, via streptavidin western blotting, in a number of organs.…”

“…However, analysis of Sox2‐BirA*G3 versus control mice fed with biotin did not reveal gross morphological differences in organs or markers of cellular ER stress, unfolded protein response, or cell death. Moreover, comprehensive RNAseq analysis of kidney, brain, and liver of these mice only revealed four differentially expressed genes between Sox2‐BirA*G3 and control mice [73]. In addition, whole‐body or fat body biotinylation in Drosophila did not significantly affect lifespan and climbing ability [1].…”

“…Variable mass isobaric tags, such as tandem mass tags (TMT), concurrently label proteins from multiple individual samples. The samples are mixed and simultaneously injected, and enriched proteins are quantified based on the mass tag associated with the tryptic peptide hit [73]. When studying secreted proteomes, homogeneous populations of cells can be subjected to perturbations and protein secretion followed [74,75].…”

“…Proteins released via the non-conventional protein secretion pathway, such as transmembrane proteins that are proteolytically cleaved through shedding at their ectodomain, were also detected. Utilizing BirA*G3 mice, proteins in the low ng/ml range can be detected in the serum with high spatial and temporal accuracy [73]. It is important to note the increase in spatial specificity and secreted proteome coverage when using these characterized BirA*G3 transgenic mice compared to viral approaches described above [118,119].…”

“…Another limitation is access to the organ or specific cell types. For example, in the brain, for the highest efficiency of labeling, the virus should be stereotactically injected to cross the blood-brain barrier [73,122,123] (Table 3). Another group crossed a Rosa26 TurboID mouse (cytoplasmically-localized) with Adlh1l1-Cre ERT and Camk2-Cre ERT to study the cell-type specific cytoplasmic proteomes of glia and neurons.…”