Extended Data Fig. 1: E3^UFM1 selectively modifies and then binds 60S ribosomes.

a, Workflow for UFM1 miniTurbo proximity profiling. b, Covalent modification of uL24 UFM1 by mT-UFM1 depends on expression of UBA5 and is enhanced by disruption of UFSP2. Immunoblot analysis of mT-UFM1 knock-in cell lines in the indicated genetic backgrounds used in the proximity labeling experiment in (a) and Fig. 1b. Note presence of non-specific band just above uL24-UFM1 band visible in UBA5^KO. c, mT-UFM1 is conjugated to uL24 on ribosomes (control for experiment in a and Fig. 1b). Lysates of U2OS mT-UFM1 knock-in cells treated with or without 200 nM anisomycin for 20 min to induce ribosome collisions were analyzed before (input) or after pelleting (ribosome pellet) through a sucrose cushion. d, Proximity labeling with mT-UFM1 shows conjugation-dependent biotinylation of proteins. Time course of UFC1 and DDRGK1 biotinylation in U2OS mT-UFM1 knock-in cells in wildtype or UBA5^KO background (as indicated) showing the conjugation-dependent specificity of biotin labeling. Cells were incubated with biotin for the indicated times prior to lysis, followed by streptavidin pulldowns (for biotinylated proteins), and elution from streptavidin beads by boiling in Laemmli buffer for immunoblot analysis. Based on the continued high selectivity for UFC1 biotinylation over the time course, mass spectrometry analysis was performed with a 4 h incubation with biotin (see Methods). e, Representative elutions from pull-downs as in Fig. 1b, c staining nitrocellulose with total protein stain (LI-COR Revert) or immunoblotted for uL24 to show the capture of ribosomes and enrichment of UFMylated uL24 (~80 % UFMylated). Transiently expressed eL36-SBP used to isolate ribosomes results in characteristic ribosome band patterning seen in SBP-UFM1 pulldowns, but lack bands (black arrowheads) discernable in the SBP-UFM1 pulldown that likely correspond to UFL1 and DDRGK1 (by molecular weight). Untagged UFM1 is used as a negative control. f, Sucrose density sedimentation profile for experiment in (g). g, UFM1 modifies exclusively 60S in vivo. Lysates from wildtype K562 cells were fractionated on sucrose density gradients and analyzed by immunoblot with the indicated antibodies. Sedimentation behavior of UFMylated uL24 parallels that of the obligatory 60S markers NEMF and eIF6. h, Quantification of indicated bands for fractions in (g) showing correlations between UFMylated uL24 and NEMF (upper) and eIF6 (lower). i, Validation of cell lines (lanes 7–9) and UFM1 and UFSP2 distribution in fractions (lanes 1–6) used for the sucrose density sedimentation in Fig. 1d. Clonal K562 cell knockouts of UFSP2 and UFM1 show no detectable expression of UFSP2 and UFM1, respectively. Cell lysates were separated via sequential detergent fractionation into, cytosolic (“C”), and membrane (“M”) fractions and analyzed by immunoblot with indicated antibodies. Non-fractionated whole cell lysate, “WC”. This fractionation distinguishes the cytosolic UFC1-UFM1 adduct (an isopeptide linked conjugate) from the co-migrating uL24-UFM1 conjugate as reported previously^2,6,12. j, Additional fractionation controls as in (i) for samples used in Fig. 1d showing partitioning of ER membrane and cytosolic markers. Membrane fractions are highly enriched for membrane markers (DDRGK1 and SEC61β) and lack cytosolic contaminants (e.g., GAPDH). k, Ribosome collisions increase E3^UFM1–60S association. K562 cells were treated with or without 200 nM anisomycin (ANS) for 1 h to induce ribosome collisions. Lysates were sedimented through 1 M sucrose to isolate ribosomes and analyzed by immunoblotting with the indicated antibodies. l, Quantification of mono-UFMylated uL24, UFL1, DDRGK1, and CDK5RAP3 from biological triplicates in experiment as in (k). Data show mean ± SD for n = 3 biological replicates. m, 60S ribosomes are the preferred target of UFMylation in vitro. Sucrose density sedimentation analysis of in vitro UFMylation reaction containing a 1:2 60S:80S molar ratio showing selectivity for 60S ribosome modification. n, 80S ribosomes are poor substrates of UFMylation in vitro. Sucrose density sedimentation as in (m) with the same concentration of 80S ribosomes as substrate showing strongly reduced UFMylation and E3^UFM1 binding. Source data is available in Supplementary Fig. 5 (for b-e and g-n), Supplementary Table 3 (for h), and Supplementary Table 4 (for l). Data in b-e, g, k, m, and n were replicated at least twice with similar results; for detailed descriptions see “Statistics and reproducibility” section of the Methods. The mobility of molecular weight markers (in kDa) is indicated on the left hand side of the blots in panels b-e, g, i-k, m, n.