It seems an increase in some of the peak amplitudes from spin-labeled protein+ ligand would be greater than the unlabeled protein+ligand. The assumptions are:

• the pulse sequence repetition rate is similar in both data set acquisitions and is less than ~3 x T1 for the resonances involved.
• the T1 of spin-labeled protein+ ligand spins are significantly shorter than the unlabeled protein+ligand.

Any spins in the vicinity of the paramagnetic, electron(s) have an addition T1 relaxation pathway. After ~3 sets of the pulse sequence, the equilibrium Z magnetization would be greater due to the decrease in T1. The increase in T1 should be a function of the average spin to unpaired electron distance. The protein-ligand exchange rate could also have an effect.

When the global molecular motion of the unpaired electron is highly asymmetrical, it’s possible the effect on T1 is a function of the electron to spin average angle.

Dear Paul,

William gave an excellent explanation...nonetheless three questions:

* Are ALL peaks in 2D (presumably 1H-15N HSQC?) affected or only a subset (in line with Williams explanation).

* How do 1D 1H (e.g. watergate) with long interscan delays (>10s) of the two samples compare?

* Have you deterined the 90° pulses for both samples? Absence or presence of salt (coming or not comping with the protein constructs) may be another explanation.

Alfred

Alfred Ross thanks for your reply. Yes, all the 2D peaks are affected.

Dear all,

I agree with the answers before.

You could increase the d1 delay in both cases in order to circumvent the T1 effect.

If you have the time, you could also aquire side-specific T2 or R1roh values using series of 2D or pseudo3D experiments for the MTSL labeled and unlabeled sample respectively. Conparison of therse values might be less error prone than comparison of peak intensities.

For analysis like this the protein part of the Bruker Dynamics Center could be used.

Many Greetings,

Kristof