1. Dynamics of the Transition between Open and Closed Conformations in a Calmodulin C-Terminal Domain Mutant 
Johan Evenäs, Anders Malmendal, Mikael Akke 
Structure 
Volume 9, Issue 3, Pages (March 2001)
DOI: /S (01)

2. Figure 1 Off-Resonance Rotating-Frame 15N Relaxation Data
Representative data are shown for R86 (filled circles), S101 (filled triangles), and D131 (open squares). (a) R1ρ decay curves obtained at a nominal tilt angle of θ = 54°. The lines show the single-exponential fits of the equation I(t) = I(0)exp(−R1ρt) to the experimental data for each residue. The bars represent the estimated uncertainties (one standard deviation) in the measured intensities. The tilt angles and optimized values of R1ρ are 57.5°, 8.0 ± 0.2 s−1 (R86); 57.5°, 15.8 ± 0.5 s−1 (S101); and 52.1°, 10.4 ± 0.2 s−1 (D131). The rotating-frame 15N relaxation rate constants (R1ρ) are shown as a function of (b) tilt angle, θ, and (c) the effective field squared, ωe2 [c.f. equations (1) and (2)]. R1 is included at θ = 0°, and R2 (inverted symbols) and R02 at θ = 90° [43]. The solid lines represent the nonlinear fits, as described in detail in the text [c.f. equations (9)–(12)]. Error bars represent one standard deviation. The dashed lines show the curves expected in absence of conformational exchange, Rex = 0 [c.f. equation (1)]. The optimized values of τex and φ are 20.3 ± 2.1 μs, (269 ± 27) × 103 s−1 (R86); 16.9 ± 1.7 μs, (980 ± 104) × 103 s−1 (S101); and 23.3 ± 1.5 μs, (484 ± 36) × 103 s−1 (D131)
Structure 2001 9, DOI: ( /S (01) )

3. Figure 2 Exchange Correlation Times: τex
Three different groups identified using cluster analysis are represented by colored symbols: red triangles, τex = 18.5 ± 2.6 μs; green circles, τex = 25.2 ± 2.7 μs; and blue diamonds, τex = 41.2 ± 10.3 μs (see the text for details).
(a) τex values shown as a function of amino acid sequence. Error bars represent one standard deviation. Locations of the helices are indicated at the top.
(b) The distributions of τex values derived from the sums of residue-specific normal distributions given by the optimized values and standard deviations. The total sum is represented by the black line
Structure 2001 9, DOI: ( /S (01) )

4. Figure 3 Populations and Chemical Shift Differences
Comparison of the φ values of (Ca2+)2-E140Q and the Ca2+-induced 15N chemical shift changes in wt-Tr2C at 301 K. The data are color coded (red triangles, green circles, and blue diamonds) according to clusters, based on the values of τex, see Figures (2)–(4) and the text for details.
(a) |φ1/2/(γNB0)| plotted versus |Δδ(wt)|. Black line: linear regression using all data, yielding po = 0.88 ± 0.03 rc = Red line: linear regression using the red group only, yielding po = 0.50 ± 0.17 rc = Open red triangles indicate residues that have not been included in the linear regression of the red group (see the text for details).
(b) The 15N chemical shift differences |Δδ(wt)| (open circles) and |Δδ(E140Q)| (colored symbols) are shown as a function of the amino acid sequence. |Δδ(E140Q)| was calculated from the φ values, assuming an open population of po = 0.50
Structure 2001 9, DOI: ( /S (01) )

5. Figure 4
Location in the Structures of Residues Exhibiting Different Exchange Correlation Times
The residues are color coded according to cluster (see the text for details): red, τex = 18.5 ± 2.6 μs; green, τex = 25.2 ± 2.7 μs; blue, τex = 41.2 ± 10.3 μs; and gray, τex not determined. The backbone Cα trace is shown for all residues. The side chain heavy atoms are shown only for those residues for which reliable estimates of the dynamic parameters could be obtained (i.e., red, green, and blue groups). The closed (a) and open (b) states of (Ca2+)2-E140Q are represented by the C-terminal domain (residues M76–K148) of apo CaM (PDB entry: 1CFC [29]) and (Ca2+)2-bound wt-Tr2C (PDB entry: 1CMG [31]), respectively. The figure was generated using MOLMOL [84] and POVRAY. For clarity, the calcium ions have been omitted
Structure 2001 9, DOI: ( /S (01) )

6. Figure 5
Pulse Sequence for the Off-Resonance Rotating-Frame 15N Relaxation Experiment
Narrow and wide bars indicate 90° and 180° pulses, respectively. Gradient pulses are indicated by hatched bars, except for the gradient pulses used for coherence selection, which are stippled. At point a, a water-selective 1H 90° pulse is applied as a rectangular pulse (ω1 ∼125 Hz). At point b, a 5 ms hyperbolic secant-shaped amplitude-modulated adiabatic pulse is applied off-resonance along the x axis to rotate the z magnetization into alignment with the effective field. During the relaxation delay T, a 15N off-resonance spin-lock field is applied by continuous-wave irradiation. An even number of relaxation cycles are employed with a 1H 180° pulse applied in the middle of each cycle as an ∼540 μs cosine-modulated rectangular pulse [85] with excitation maxima positioned 2 kHz from the carrier (on water). Following the relaxation delay, a second reversed adiabatic pulse is applied to bring the magnetization back to the z axis. To ensure the same amount of heating of the sample in all experiments, a compensating spin-lock field of length T′ = Tmax − T is applied far off-resonance (300 kHz) in the beginning of each recycle delay. In all experiments, Tmax was set to 500 ms and the recycle delay (D) to 2 s. The delays τa, τb, and τc were set to 2.25, 2.75, and 0.5 ms, respectively. The delay τ′b is given by τb + 2pw, where pw is the length of the 1H 90° pulse. The experiment employs gradient and phase cycling schemes as described by Farrow et al. [78]. Gradient levels used were g1 = 1 ms, 5 Gcm−1; g2 = 0.5 ms, 4 Gcm−1; g3 = 1 ms, 10 Gcm−1; g4 = 0.5 ms, 8 Gcm−1; g5 = 1.25 ms, 30 Gcm−1; and g6 = ms, 29.5 Gcm−1. The phase cycling used was φ1 = x, −x; φ2 = y; φ3 = 2(x), 2(y), 2(−x), 2(−y); φ4 = x; and receiver was x, −x, −x, x. For each increment of t1, two free induction decays (FIDs) were collected with the phase of φ4 and the amplitude of g6 inverted for the second FID. For each increment of t1, the phases of φ4 and the receiver were incremented by 180°
Structure 2001 9, DOI: ( /S (01) )