Whole-cell current evoked by 10 μM NMDA and 10 μM glycine (open bars) at −80 mV before and after exposure to 30 μM DHA (solid bar) in cells co-transfected with N1/K2(R)+N2B/K2(Q)L614A (a), N1/K2(R)L614A+N2B/K2(Q)L614A (b), N1/K2(Q)+N2B/K2(R)L614A (c), or N1/K2(R)L614A+N2B/K2(R)L614A (d). (e) Plot of current evoked immediately after exposure to DHA as a fraction of control current before DHA (mean ± s.e.m.) for the 12 chimaeric constructs and, for comparison, homomeric L614A mutants of GluK2(Q) and (R). Sample size is shown for each bar. Note the logarithmic scale. DHA had minimal effect when L614A and Q to R editing were not present on the same chimaeric subunit; green horizontal bar indicates the 95% confidence interval (C.I.) for N1/K2(Q)L614A +N2B/K2(Q)L614A. Significant potentiation was observed when one of the two subunits included the L614A substitution and Q to R editing (P < 0.0001, one way ANOVA with post hoc Student-Newman-Keuls comparison); 95% C.I. for N1/K2(Q)L614A +N2B/K2(R)L614A (yellow bar) Potentiation was greatest when both chimaeric subunits were edited (R) and mutated (L614A); 95% C.I. for N1/K2(R)L614A+N2B/K2(R)L614A (pink bar).