Redesigning type II’ beta-turn in GFP to type I’: implications for folding kinetics and stability

1. Redesigning type II’ beta-turn in GFP to type I’: implications for folding kinetics and stability Bharat Madan, Govindan Raghunathan, and Sun-Gu Lee* Department of Chemical and Biochemical Engineering, Pusan National University, Busan 609-735, Republic of Korea TEL: +82-51-583-8343, FAX: +82-51-512-8563 Abstract Effect on refolding kinetics and thermodynamic stability of GFP Both type I’ and type II’ beta-turns have the same sense of the β-turn twist that is compatible with the beta-sheet twist. They occur predominantly in two residue beta-hairpins, but the occurrence of type I’ β-turns is two times higher than type II’ β-turns. This suggests that type I’ beta-turns are evolutionarily more conserved than type II’ beta-turns, and type I’ turn sequence and structure can be more favorable for protein folding than type II’ beta-turns. However, this possibility has not been tested seriously. Here, we redesigned the native type II’ beta-turn in GFP to type I’ beta-turn, and investigated its effect on protein folding and stability. The type I’ beta-turns were designed based on the statistical analysis of residues in natural type I’ beta-turns. The substitution of the native ‘GD’ sequence of i+1 and i+2 residues with type I’ preferred ‘(N/D)G’ sequence motif increased the folding rate by 50% and slightly improved the thermodynamic stability. Despite the enhancement of in vitro refolding kinetics and stability of the redesigned mutants, they showed poor soluble expression level compared to wild type. To overcome this problem, i and i+3 residues of the designed type I’ β-turn were further engineered. The mutation of Thr to Lys at i+3 could restore the in vivo soluble expression of the type I’ mutant. This study indicates that type II’ beta turns in natural beta-hairpins can be further optimized by converting the sequence to type I’. • The variants with the designed type I’ β-turn sequences were found to have higher • folding rates than the control with the native type II’ β-turn. • The thermodynamic stability of the variants were slightly higher than GFPcon. Figure 5. Refolding kinetics plots and equilibrium denaturation plots for GFPcon and its variants. Table 1. Refolding kinetics and thermodynamic stability data for GFPcon and its designed mutants Figure 1. Overview of the study. Design Strategy Effect on in vivo proteinexpression and whole cell fluorescence Figure 2. Strategy used in designing the beta-turn mutants. Results Figure 6. Expression profile and relative whole cell fluorescence (WCF) of GFPcon and its variants. . (A) Soluble fractions and (B) Insoluble fractions of the expressed 27kDa proteins shown by an arrow. Lanes 1,2,3,4 and 5 in (A) and (B) indicate fractions for GFPcon, TD1, TD2, TD3 and TD4, respectively. (C) The WCF was measured by normalizing fluorescence by cell OD at 600nm. The whole cell fluorescence was calculated relative to GFPcon by setting its whole cell fluorescence to 100. • The in vivo protein expression was severely affected in case of TD1, TD2 and TD4. • The whole cell fluorescence was found to be reduced by 50% in variants TD1, TD2 and TD4. • In case of TD3 the in vivo soluble expression was similar to GFPcon and the whole cell • fluorescence was about 85% of GFPcon. Figure 3. A. The target type II’ β-turn EGDT (in yellow) in GFP and its interactions with neighboring residues, Lys113 and Arg122 (in red). B. Multiple Sequence Alignment (MSA) of residues at positions i+1 and i+2 of β-turn EGDT. • The turn type conversion from type II’ to I’ resulted in an increased folding kinetics of GFP. • The thermodynamic stability of the protein was not affected, rather the designed mutants • had slightly higher stability than the control. • The current findings clearly show that protein properties such as folding kinetics can be • improved further, even though the sequence appears to be evolutionary conserved. • This study emphasizes that for the design of structural motifs like beta-turns, the • conservation should be considered at the level of different folds across protein families, • rather than the homologous proteins. • The results highlight the use of evolutionary information in protein design. Discussion • Of the 19 homologues, 11 sequences have the conserved sequence motif • ‘GD’ at i+1 and i+2 positions, which are known to be preferred by type II’ β- • turns. • On the other hand, when the homologous proteins were analyzed at the • structural level, 16 turns were identified to be type II’ beta-turns. Effect on protein structure and specific fluorescence activity MadanB, Seo SY, Lee SG. Structural and sequence features of two residue turns in beta-hairpins. Proteins 2014.( doi/10.1002/prot.24526) Pedelacq JD, Cabantous S, Tran T, Terwilliger TC, Waldo GS. Engineering and characterization of a superfolder green fluorescent protein. Nat Biotech 2006;24:79-88. Simpson ER, Meldrum JK, Searle MS. Engineering diverse changes in beta-turn propensities in the N-terminal beta-hairpin of ubiquitin reveals significant effects on stability and kinetics but a robust folding transition state. Biochemistry 2006;45:4220-4230. References Figure 4. Excitation - emission profiles and specific fluorescence activities of variants and control (GFPcon). (A) Excitation spectrum, (B) Emission spectrum and (C) specific fluorescence values in arbitrary units (a.u.) relative to GFPcon, to which 100 was assigned • The excitation-emission profile of the variants were similar to the control. • The specific fluorescence activity was found to be similar for all proteins. ak1denotes the rate for the initial or fast phase of refolding progress curve. bk2denotes the rate for the slow phase of the refolding kinetics. c Free energy of denaturation, calculated using the relation ΔG = ΔG(H2O) – m[urea]. dm denotes measure of dependence of ΔG as a function of urea concentration. # Error range: standard deviation of two independent experiments