$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
There is an important need for tools that can isolate biological samples from specific subcellular compartments to better investigate how subcellular organization and trafficking factor into cellular processes. mRNA translation is a key example of such a process: growing evidence suggests that translation is highly compartmentalized and, in particular, that many mRNAs are localized to the organelles where their encoded proteins will reside1,2. However, less is known about the subcellular localization of translational machinery. Ribosomes begin their existence in the nucleolus and nucleus, where assembly from ribosomal proteins (RPs) and ribosomal RNAs (rRNAs) begins. They are then exported to the cytoplasm for the final stages of assembly after which they begin translating mRNAs3,4. Canonical sites of translation include the cytoplasm and endoplasmic reticulum; the latter is where integral membrane proteins and proteins destined for membrane-bound organelles or extracellular secretion are synthesized5. More recently identified sites of localized translation include the mitochondria, peroxisome, and neuronal processes6,7,8,9,10,11,12. Interestingly, evidence is also emerging that translational machinery features such as ribosome-associated proteins, post-translational modifications, rRNA modifications, and non-canonical stoichiometry of core ribosomal proteins are involved in regulatory mechanisms that govern translation of specific mRNAs and parameters such as translation fidelity and rate13,14,15,16,17,18,19,20. For example, hundreds of proteins have been identified that interact with the ribosome and could potentially lead to differences in subcellular ribosome composition13. However, techniques for isolating subcellular ribosome populations have been limited to date; thus, much remains unknown about how the composition of translational machinery contributes to regulating localized mRNA translation. The advent of versatile methods for isolating subcellular ribosome populations, particularly those compatible with mammalian model systems, will allow for exploration of these translation machinery features at the subcellular level.
The earliest methods for isolating subcellular fractions containing ribosomes were based on principles of sequential lysis and centrifugation21,22. Typically, these methods are limited to certain organelles and often require extensive processing times. A more recent method is based on the biotin ligase BirA, which specifically biotinylates an AviTag peptide substrate10,23,24. The BirA can be genetically localized to the organelle of interest in cells where ribosomes have been tagged with AviTag. However, because BirA is active at physiologic concentrations of biotin, it is necessary to culture cells under conditions of biotin depletion in order to avoid constitutive and non-specific biotinylation of AviTag-bearing ribosomes10,24. This limits the use of BirA in cell types that are sensitive to biotin depletion25. Finally, promiscuous biotinylation methods have been developed that use engineered enzymes such as BioID, TurboID, and APEX26,27,28. Upon induction, these enzymes produce a reactive biotin intermediate that covalently modifies any lysine residues within a radius of the enzyme. Promiscuous biotinylation methods are highly useful for mapping subcellular proteomes, but less informative for distinguishing proteins in the general vicinity of localized ribosomes as opposed to proteins that are in closer proximity with them.
Here, we demonstrate our recently developed method for labeling and isolating subcellular ribosome populations, called AviTag-specific Location-restricted Illumination-enhanced Biotinylation (ALIBi). This method leverages an engineered split BirA fused to optogenetic enhanced Magnet (eMag) switches (Figure 1A) that heterodimerize under blue light illumination29; and a peptide tag consisting of FLAG epitope, tobacco etch virus (TEV) protease cleavage site, and AviTag (FTA) fused to the Rpl31 ribosomal protein (Figure 1B). The split enzyme can be genetically targeted to selected organelles, is inactive at physiologic concentrations of biotin, and can be reconstituted within minutes upon blue light illumination and the addition of supplemental biotin (Figure 1C-E). Subsequent cell lysis, affinity purification, and non-denaturing enzymatic elution via TEV protease cleavage yields ribosomes and co-purifying molecules such as mRNAs and ribosome-associated proteins, which are suitable for downstream analytic methods such as RNA sequencing and mass spectrometry proteomics (Figure 1A). Therefore, this method reveals not only ribosome composition but also the mRNAs to which ribosomes are bound. In a recent article describing the seminal application of ALIBi to characterize subcellular ribosome populations, we successfully targeted the split enzyme to the cytoplasm, nucleolus, nucleus, endoplasmic reticulum, plasma membrane, and mitochondria (Table 1)30. The resulting comparison of translational machinery between these subcellular compartments led to novel and unexpected findings regarding translation at the endoplasmic reticulum and mitochondria, as well as potential leads for future investigation of ribosome biogenesis factors in the nucleus and nucleolus.
This protocol contains a detailed demonstration of transfection and activation of the ALIBi system, cell lysis, affinity purification, and sample elution. We also discuss how to verify successful affinity purification via western blot and make recommendations for performing downstream RNA sequencing or mass spectrometry (MS) proteomic analysis (see Supplemental File 1). This protocol is written for experiments using mouse embryonic stem cells (mESCs) expressing Rpl31 fused to FTA (L31-FTA) homozygously from the endogenous Rpl31 locus; this cell line is available from the authors upon request. The strategy for generating this cell line is described in Supplemental File 1 and in Supplemental Figure S1 and Supplemental Figure S2. Plasmids encoding the split enzymes targeted to several organelles are available on Addgene (catalog numbers #235608 to 235643). The design principles for targeting the split enzymes to these organelles is described in Supplemental File 1. The protocol assumes that each sample originates from mESCs grown in a 10 cm plate. We have found that the 10 cm scale yields sufficient material for mass spectrometry proteomics, while the 6-well scale is sufficient for western blots or RNA sequencing.