Figure 3: Preparation of segmentally labelled lys11-linked SUMO-2 dimers for NMR.

(a) To increase efficiency of formation of SUMO-2 dimer, two truncated SUMO-2 constructs, termed SUMO-2 ΔN11 and SUMO-2 ΔGG, were designed. SUMO-2 ΔN11 lacks Lys11, and thus cannot be modified by another SUMO-2 molecule. SUMO-2 ΔGG has the C-terminal diglycine motif deleted and cannot be conjugated to target proteins. Together, these two constructs can form SUMO-2 dimers, but not longer chains. (b) Chromatogram showing purification of SUMO-2 dimer by gel filtration chromatography on a Superdex 75 column. (c) Coomassie blue-stained SDS–PAGE gel showing fractions from gel filtration chromatography (left) and purified SUMO-2 dimers for NMR studies (right). Lane 1: SUMO-2 dimer comprises ¹³C,¹⁵N-labelled proximal SUMO-2 ΔGG and unlabelled SUMO-2 ΔN11; lane 2: SUMO-2 dimer containing ¹³C,¹⁵N-labelled distal SUMO-2 ΔN11 and unlabelled SUMO-2 ΔGG. (d) Overlay of ¹H–¹⁵N HSQC NMR spectra for the two SUMO-2 domains (distal SUMO-2 ΔN11 in cyan and proximal SUMO-2 ΔGG in green) in the K11-linked segmentally labelled SUMO-2 dimer. All peaks corresponding to the core residues 17–87 possess nearly identical chemical shifts in either SUMO-2 domain. Peaks that show significant difference correspond to residues either only present in one subunit owing to detection at the termini or very close to the isopeptide bond. These data suggest that the core SUMO domains are identical in structure and behave independently from each other.