1. CPMG relaxation dispersion NMR measurements under two static magnetic fields (600 and 800 MHz here) are strongly suggested, in order to faithfully estimate the dynamics parameters for a two-site slow (kex < Δω), fast (kex > Δω) or intermediate (kex ≈ Δω) exchange process. Because one set of results at one single static field is not enough to get unique solutions for some dynamics parameters by Levenberg-Marquardt non-linear least squares fitting.
2. Here we use pulse sequences from Kay lab to measure backbone N-H or side-chain CH3 CPMG relaxation dispersion profiles, and use the program CPMGFit from Palmer lab to do the non-linear least squares fitting to extract dynamics parameters. The relation between 1/τcp (1/ms) and νcpmg (Hz) is: 1/τcp (1/ms) = 0.002*νcpmg .
3. The general steps for getting cpmg profile curves between R2,eff ~ νcpmg :
-
Measure a series of 2D spectra with ncyc_cp = 0,2,4,6,8,10,12,14,16,18,20,24,28,34,40 by the “interleave” method; and constant time T = 40ms here,
-
Process all 2D spectra with the same macro, and get the intensities for the peaks you focus on,
-
Using excel, calculate the R2,eff values based on the equation: R2,eff = -25ln(I(n)/I(0)); and νcpmg = ncyc_cp/T = ncyc_cp/0.04 (Hz);
4. Run CPMGFit (Prof. Palmer Lab) to do the non-linear fitting with the proper model equations. For example, for apo-CupA M116,
The input file looks like the following:
title test
fields 1 14.1
function CPMG
R2 7.0 10 3
Rex 1 5 3
Tau 0.1 1.0 3
xmgr
@ XAXIS LABEL “1/tcp (1/ms)”
@ YAXIS LABEL “R2(tcp) (1/s)”
@ XAXIS TICKLABEL FORMAT DECIMAL
@ YAXIS TICKLABEL FORMAT DECIMAL
@ XAXIS TICKLABEL CHAR SIZE 0.8
@ YAXIS TICKLABEL CHAR SIZE 0.8
@ WORLD XMIN 0.0
data
0.1 13.35693384 0.20000 14.1
0.2 11.70070816 0.20000 14.1
0.3 10.52861231 0.20000 14.1
0.4 10.68790081 0.20000 14.1
0.5 10.36373427 0.20000 14.1
0.6 9.833991237 0.20000 14.1
0.7 9.872712395 0.20000 14.1
0.8 9.689739742 0.20000 14.1
0.9 9.366542289 0.20000 14.1
1.0 9.317617646 0.20000 14.1
1.2 9.042098019 0.20000 14.1
1.4 8.791702186 0.20000 14.1
1.7 8.437310403 0.20000 14.1
2.0 8.095035264 0.20000 14.1
Here we use fast exchange limit model, and the fitting equation is:
y=R2+Rex*(1 – 2*Tau*x*tanh(1/(2*Tau*x)))
And type the command like this:
/path/cpmgfit –grid –debug –f input_filename > output_filename
So the output file looks like the following:
# CPMGfit 1.43
#
# title test
# function CPMG
# equation y=R2+Rex*(1 – 2*Tau*x*tanh(1/(2*Tau*x)))
# points 14
# X2 41.9488
# X2(red) 3.8135
#
# Parameter Fitted_Value Fitted_Error Sim_value Sim_error
# R2 8.4806 0.1239 8.4669 0.1278
# Rex 5.2159 0.2441 5.2525 0.2520
# Tau 0.7933 0.0762 0.7973 0.0812
#
# Field Rex Rex_error
# 14.1000 5.2159 0.2520
#
# %tile X2
# 0.0500 4.4406
# 0.1000 5.6343
# 0.1500 6.2493
# 0.2000 7.0048
# 0.2500 7.6012
# 0.3000 8.2517
# 0.3500 8.7359
# 0.4000 9.1695
# 0.4500 9.7304
# 0.5000 10.2241
# 0.5500 10.7436
# 0.6000 11.3981
# 0.6500 11.9411
# 0.7000 12.8253
# 0.7500 13.5048
# 0.8000 14.3279
# 0.8500 15.3564
# 0.9000 16.9108
# 0.9500 18.9777
#
