Skip to main content
. 2015 Apr 1;290(22):13654–13666. doi: 10.1074/jbc.M115.651208

TABLE 1.

Effect of C. thermocellum cohesin mutations on type I cohesin-dockerin binding affinity

Mutations in both the hydrophobic and hydrophilic interface patches (Figs. 2 and 3) can significantly reduce binding. Predicted ΔΔG values were calculated based on the Protein Data Bank structure 1OHZ (12) using the Rosetta interface protocol (33). Predicted hot spots (mutations to alanine with ΔΔG >1.0 kcal/mol) within the interface patches were experimentally validated by iELISA. Fig. 4 summarizes the predicted effect on binding of mutations as calculated by additional computational protocols.

Mutation Calculated ΔΔG Rosetta 2.3a Experimental ΔΔG Experimental IC50b
REU kcal/mol nm
WT 0 1.1b

Hydrophobic patch
    L83A 2.2 3.0 123c
    L83S 2.4 3.4 257c,d

Polar patch
    N37A 2.2 0 1
    N37D 2.1 2.5 70
    N37L 1.9 >4e >103
    D39A 3.0 >4 >103c
    D39N 0.3 >4 >103f
    E131A 1.3 1.8 20
    N37A/E120A 2.2 1.2 7c
    N37A/D39A 5.2 >4 >103
    N37A/E131A 3.5 1.0 6
    N37A/D39A/E131A 6.4 >4 >103

Additional positions
    V41A 0.0
    V41Y 0.3 0.7 4f
    D70A 0.1
    D70S 0.2 −0.1 1f
    Y74A 1.8 −0.1g
    V81S 0.1 1.9 30f
    A85L 0.5 −0.1 1f
    E86A 1.6
    E86S 1.6 0.4 g
    E120A 0.0
    D39S/Y74A 4.4 1.1g
    D39S/Y74A/E86S 6 2.8g
    D39N/D70S/V81S 0.5 >3 >200f

a In Rosetta energy units (REU); optimized to correlate with kcal/mol.

b IC50 values in this study are normalized relative to the binding of wild-type C. thermocellum cohesin to C. thermocellum dockerin (i.e. 1.1, first row).

c Reported in Ref. 26.

d The standard error of the mean of logarithmic values of IC50 exceeds the half-log value for this measure.

e Very weak binding could not be fitted and was therefore assigned ΔΔG values of >4 kcal/mol corresponding to the largest measured concentration.

f Measured also previously by enzyme-linked interaction assay (29).

g Measured by isothermal titration calorimetry (65).