Fig 1: Proposed debranching model.(A) Surface depiction of UCH37 showing the canonical S1 (cS1) ubiquitin (Ub)-binding site and the new K48-specific binding sites identified in this study. (B) Proposed mechanism for chain debranching using the K48-specific binding sites. The K48-linked portion of a branched chain engages the K48-specific Ub-binding sites in two different orientations: one with the proximal Ub (Ubprox) docked at the a5–6 motif and the other with the distal Ub (Ubdist) bound to that site. As the docking models show (Figure 4H), the K48 isopeptide bond is less obstructed and closer to the catalytic C88 residue when the K48 Ubdist moiety is bound to a5–6 and Ubprox is bound near the L181 region. In this orientation, the other Ubdist at the branchpoint is positioned near the frontside of the enzyme. With K48 isopeptide bond cleavage occurring on the backside, the a5–6 motif would serve as the noncanonical S1 (ncS1) site and the L181 region would be the ncS1´ site.
Fig 2: Backside mutant impairs K48 chain binding.(A) Deuterium uptake plots showing how the F117A mutation affects hydrogen-deuterium exchange in a peptide corresponding to residues 115–125 located in the a5–6 motif of UCH37. For comparison, the uptake plots corresponding to a region located outside the a5–6 motif (residues 200–218) are also shown. The data on the top represents the rate of exchange with UCH37C88A•RPN13DEUBAD and the data on the bottom corresponds to UCH37C88A/F117A•RPN13DEUBAD. At least two replicates of each experiment were performed. (B)1H13C-methyl TROSY NMR spectra of the Ile region of ILV-labeled K48 di-ubiquitin (Ub) proximal subunit (Ubprox) free in solution (gray) and bound to UCH37C88A•RPN13DEUBAD (the C88A complex; green). Ratio of UCH37 to K48 di-Ub is 1:1.5. Concentrations used: 45 µM UCH37C88A•RPN13DEUBAD and 30 µM K48 di-Ub. (C)1H13C-methyl TROSY NMR spectra of the Ile region of ILV-labeled K48 di-Ub distal subunit (Ubdist) free in solution (gray) and bound to the C88A complex (orange). Ratio of UCH37 to K48 di-Ub is 1:1.5. Concentrations used: 45 µM UCH37C88A •RPN13DEUBAD and 30 µM K48 di-Ub. (D) 1H13C-methyl TROSY NMR spectra of the Ile region of K48-linked Ubprox free in solution (gray) and bound to UCH37C88A/F117A•RPN13DEUBAD (the F117A complex; green). Ratio of UCH37 to K48 di-Ub is 1:1.5. Concentrations used: 45 µM UCH37C88A/F117A•RPN13DEUBAD and 30 µM K48 di-Ub. (E)1H13C-methyl TROSY NMR spectra of the Ile region of K48-linked Ubdist free in solution (gray) and bound to the F117A complex (orange). Ratio of UCH37 to K48 di-Ub is 1:1.5. Concentrations used: 45 µM UCH37C88A/F117A•RPN13DEUBAD and 30 µM K48 di-Ub. (F)1H13C-methyl TROSY NMR spectra of the Ile region of mono-Ub bound to the C88A complex (orange) and the F117A complex (blue). Ratio of UCH37 to mono-Ub is 1:1.5. Concentrations used: 45 µM UCH37 •RPN13DEUBAD complex and 30 µM mono-Ub. Figure 5—source data 1.Full NMR spectra of mono-Ub and K48 di-Ub in presence and absence of UCH37.
Fig 3: Hydrogen-deuterium exchange mass spectrometry (HDX-MS) uncovers a cryptic K48 chain-specific binding site.(A) Differential deuterium uptake plot comparing mono-ubiquitin (Ub) bound UCH37C88A•RPN13DEUBAD to free UCH37C88A •RPN13DEUBAD. (B) Differential deuterium uptake plot comparing UCH37•RPN13DEUBAD covalently linked to Ub propargylamine at the active site Cys (Ub~UCH37•RPN13DEUBAD) to free UCH37C88A •RPN13DEUBAD. (C) Differential deuterium uptake plot comparing K48 di-Ub-bound UCH37C88A•RPN13DEUBAD to mono-Ub-bound UCH37C88A•RPN13DEUBAD. (D) Differential deuterium uptake plot comparing K48 di-Ub-bound Ub ~UCH37•RPN13DEUBAD to free Ub ~UCH37•RPN13DEUBAD. (E–G) Structure of UCH37 (PDB ID: 4UEL) showing regions with statistically significant differences in exchange upon noncovalent binding to mono-Ub (E), covalent attachment of Ub to the active site (F), and noncovalent binding to K48 di-Ub. (G) Data correspond to 2 hr of deuterium labeling. Other highlighted regions include the catalytic Cys (C88), the crossover loop (CL), and the canonical S1 site (pink). (H) Structure of UCH37 (PDB ID: 4UEL) showing statistically significant differences in exchange upon noncovalent binding of K48 di-Ub to Ub ~UCH37•RPN13DEUBAD. Data correspond to 2 hr of deuterium labeling. (I–K) Differential deuterium uptake plot comparing the effects of mono-Ub (I), K48 di-Ub (J), and K6/K48 tri-Ub (K) binding to Ub ~UCH37•RPN13DEUBAD on the exchange of residues in RPN13DEUBAD. (L–M) Heat map showing regions of RPN13DEUBAD with statistically significant differences in deuterium exchange upon noncovalent binding to K48 di-Ub (L), and noncovalent binding to K6/K48 tri-Ub (M). Data correspond to 2 hr of deuterium labeling. Figure 2—source data 1.HDX-MS analysis of UCH37 and RPN13DEUBAD.
Fig 4: Chemical crosslinking confirms the presence of a K48 chain-specific binding site on the backside of UCH37.(A–E) A photolabile diazirine-based crosslinker was appended to individual ubiquitin (Ub) subunits of K48 chains. Crosslinked peptides were mapped onto the structure of UCH37 (PDB: 4UEL) according to their relative abundance based on the area under the curve of the extracted ion chromatogram. (A) Map of the crosslinking data for K48 di-Ub binding to UCH37C88A•RPN13DEUBAD with the proximal Ub subunit labeled with the diazirine colored by normalized relative abundance of crosslinked peptide. (B) Map of the crosslinking data for K48 di-Ub binding to UCH37C88A•RPN13DEUBAD with the distal Ub subunit labeled with the diazirine colored by normalized relative abundance of crosslinked peptide. (C) Map of the crosslinking data for K6/K48 tri-Ub binding to UCH37C88A•RPN13DEUBAD with the proximal Ub subunit labeled with the diazirine colored by normalized relative abundance of crosslinked peptide. (D) Map of the crosslinking data for K6/K48 tri-Ub binding to UCH37C88A•RPN13DEUBAD with the K48-linked distal Ub subunit labeled with the diazirine colored by normalized relative abundance of crosslinked peptide. (E) Map of the crosslinking data for K6/K48 tri-Ub binding to UCH37C88A•RPN13DEUBAD with the K6-linked distal Ub subunit labeled with the diazirine colored by normalized relative abundance of crosslinked peptide.
Fig 5: Docking models of the K48 di-ubiquitin (Ub):UCH37•RPN13DEUBAD complex.(A–B) HADDOCK docking models show two poses corresponding to the interaction between K48 di-Ub and UCH37 along with their fit to experimental small-angle X-ray scattering data of the K48 di-Ub:UCH37C88A•RPN13DEUBAD complex. The goodness of fit to the experimental intensity is represented by ?2 values. In the first pose (A), the proximal Ub (Ubprox; green) interacts with a5–6 of UCH37. In the second pose (B), the distal Ub (Ubdist; orange) interacts with a5–6 of UCH37. (C) Residues highlighting the interaction between the aromatic-rich region of UCH37 a5–6 and the I44 patch of Ubprox in pose 1 (A). (D) Residues highlighting the interaction between the aromatic-rich region of UCH37 a5–6 and the I44 patch of Ubdist in pose 2 (B). (E) Polar contacts between Ubprox and UCH37 a5–6 in pose 1 (A). (F) Contacts between Ubdist and residues of UCH37 located outside the a5–6 motif in pose 1 (A). (G) Contacts between Ubprox and residues of UCH37 located outside the a5–6 motif in pose 2 (B). (H) The relative location of active site and the scissile, K48 isopeptide bond in molecular docking poses 1 and 2. In pose 1, residues of a6 and the loop leading into the catalytic Cys (C88) form a barrier for the isopeptide bond. In pose 2, only Q82 of UCH37 blocks the K48 isopeptide bond.
Supplier Page from DNASU for UCHL5 (Homo sapiens) in pVP16 (His and MBP-tagged bacterial expression vector)