Mpounded chemical shift differences between the apo-Wt and the apo-mutants (Fig. 2A) were computed as previously described [27]. In brief, the compounded chemical shift difference between the apo-Wt and the apo-mutants was calculated as the magnitude 
of vector A in Figure 2A. Similarly, the compounded chemical shift differencecAMP Binding MeasurementsThe dissociation constant (KD) for cAMP from Wt and de305 were measured through the saturation transfer difference (STD) CAL 120 manufacturer amplification factor (STDaf) [24,40]. All STD measurements were carried out with a solution of 25 mM of Wt or 15 mM de305 in 20 mM phosphate buffer, pH 7.6, 50 mM NaCl, 99.9 D2O andAuto-Inhibitory Hinge HelixTable 1. Total Variance Breakdown in the SVD Analysis of Deletion Mutants and L273W.Mutant EPAC149?12 (de312)Principal Components (PCs) PC1 PCPercentage of Total Variance 52.4 44.8 (97.2 )* 59.8 36.3 (96.1 )* 57.3 35.8 (93.1 )* 72.3 23.4 (95.7 )EPAC149?10 (de310)PC1 PCEPAC149?05 (de305)PC1 PCEPAC149?18 (L273W)PC1 PC*The purchase Ergocalciferol percentages reported in parentheses are the cumulative contribution of PC1 and PC2 for each SVD analysis involving a mutant. doi:10.1371/journal.pone.0048707.tat 25uC. The 1D-STD spectra were acquired at total cAMP concentrations of 25, 50, 75, 100, 150, 200 and 300 mM [24]. Separate reference 1D (STR) experiments were also acquired. The STD amplification factor (STDaf) was calculated as the product of the STD/STR ratio (measured for the well resolved cAMP ribose H1′ at 6.2 ppm) and of the ratio of the total cAMP and protein concentrations. The STDaf values were then normalized relative the STDaf plateau value reached at high cAMP concentrations ([cAMP]Tot. 150 mM). The normalized STDaf values were then analyzed with the binding isotherm equation: Normalized STDaf = 1?(1/ (1+ ([cAMP]/KD))), where [cAMP] 23977191 is the concentration of free cAMP [24,40].Results and Discussion CHESPA analysis of de305, de310 and deTo investigate the effects of the C-terminal deletion mutations, we purified and assigned de305, de310 and de312 in the apo states and compared them to the Wt(apo) and cAMP-bound states (Fig. 2A). We first analyzed the de312 truncation mutant (i.e. EPAC1149?12), which leaves the hinge region (residues 296?10) to a large extent intact but removes the C-terminal tail of the Wt construct, EPAC1149?18. The residue profile of the compounded chemical shift differences between Wt(apo) and de312(apo) (Figure 3A, red bars) exhibits local maxima in the regions most affected by cAMP-binding (Fig. 3A, grey regions) [9,21]. In addition, the [15N-1H]-HSQC spectral comparison of the de312(apo) mutant relative to the Wt(apo) and cAMP-bound states for well dispersed and isolated peaks (Fig. 2B) reveals a slight but consistent shift for de312 towards the active state. However, in order to systematically assess at residue resolution the effect of the de312 mutation on the apo/inactive vs. apo/active auto-inhibitory equilibrium, we took advantage of the recently developed chemical shift projection analysis (CHESPA) (Fig. 2A; Fig. 3B, 3C, red bars). While the compounded chemical shifts quantify only the size of the perturbation, the fractional activation X obtained from the projection analysis (Fig. 3B) together with the cosine H values (Fig. 3C) reflect both the direction and extent of the mutational perturbation toward the apo/active state. The fractional shifts obtained though the projection analysis reflect four main effects: (a) nearest neighbour effects exp.Mpounded chemical shift differences between the apo-Wt and the apo-mutants (Fig. 2A) were computed as previously described [27]. In brief, the compounded chemical shift difference between the apo-Wt and the apo-mutants was calculated as the magnitude of vector A in Figure 2A. Similarly, the compounded chemical shift differencecAMP Binding MeasurementsThe dissociation constant (KD) for cAMP from Wt and de305 were measured through the saturation transfer difference (STD) amplification factor (STDaf) [24,40]. All STD measurements were carried out with a solution of 25 mM of Wt or 15 mM de305 in 20 mM phosphate buffer, pH 7.6, 50 mM NaCl, 99.9 D2O andAuto-Inhibitory Hinge HelixTable 1. Total Variance Breakdown in the SVD Analysis of Deletion Mutants and L273W.Mutant EPAC149?12 (de312)Principal Components (PCs) PC1 PCPercentage of Total Variance 52.4 44.8 
(97.2 )* 59.8 36.3 (96.1 )* 57.3 35.8 (93.1 )* 72.3 23.4 (95.7 )EPAC149?10 (de310)PC1 PCEPAC149?05 (de305)PC1 PCEPAC149?18 (L273W)PC1 PC*The percentages reported in parentheses are the cumulative contribution of PC1 and PC2 for each SVD analysis involving a mutant. doi:10.1371/journal.pone.0048707.tat 25uC. The 1D-STD spectra were acquired at total cAMP concentrations of 25, 50, 75, 100, 150, 200 and 300 mM [24]. Separate reference 1D (STR) experiments were also acquired. The STD amplification factor (STDaf) was calculated as the product of the STD/STR ratio (measured for the well resolved cAMP ribose H1′ at 6.2 ppm) and of the ratio of the total cAMP and protein concentrations. The STDaf values were then normalized relative the STDaf plateau value reached at high cAMP concentrations ([cAMP]Tot. 150 mM). The normalized STDaf values were then analyzed with the binding isotherm equation: Normalized STDaf = 1?(1/ (1+ ([cAMP]/KD))), where [cAMP] 23977191 is the concentration of free cAMP [24,40].Results and Discussion CHESPA analysis of de305, de310 and deTo investigate the effects of the C-terminal deletion mutations, we purified and assigned de305, de310 and de312 in the apo states and compared them to the Wt(apo) and cAMP-bound states (Fig. 2A). We first analyzed the de312 truncation mutant (i.e. EPAC1149?12), which leaves the hinge region (residues 296?10) to a large extent intact but removes the C-terminal tail of the Wt construct, EPAC1149?18. The residue profile of the compounded chemical shift differences between Wt(apo) and de312(apo) (Figure 3A, red bars) exhibits local maxima in the regions most affected by cAMP-binding (Fig. 3A, grey regions) [9,21]. In addition, the [15N-1H]-HSQC spectral comparison of the de312(apo) mutant relative to the Wt(apo) and cAMP-bound states for well dispersed and isolated peaks (Fig. 2B) reveals a slight but consistent shift for de312 towards the active state. However, in order to systematically assess at residue resolution the effect of the de312 mutation on the apo/inactive vs. apo/active auto-inhibitory equilibrium, we took advantage of the recently developed chemical shift projection analysis (CHESPA) (Fig. 2A; Fig. 3B, 3C, red bars). While the compounded chemical shifts quantify only the size of the perturbation, the fractional activation X obtained from the projection analysis (Fig. 3B) together with the cosine H values (Fig. 3C) reflect both the direction and extent of the mutational perturbation toward the apo/active state. The fractional shifts obtained though the projection analysis reflect four main effects: (a) nearest neighbour effects exp.
