i like this post (click again to cancel)
i dont like this post (click again to cancel) remove favorite mark from this question (click again to restore mark)

Hello everyone, I have a stupid question and often confuse me. We often meet peak missing in 15N H HSQC,but usually how many percent peak enough to solve NMR solution structure? I ask a lot of people and check in literature. Some guy suggested at least 80 except proline, other answer is at least 90%?For my current project, only appear 80% peak except proline. Now I am trying to optimize sample condition to make more peak. But I need to know a percent that I must optimize for my sample peak. Check my biology result I know that most missing peak lie in a very important disorder, which play an important function role. May I have a understand, If we can check missing peak. If miss peak have no important function, we don't care and directly to assign other peak and solve structure. If missing peak have function, it is a totally different condition. For my protein, If I can't make the disorder(misssing peak), I will not get the correct structure. Thanks a lot!

Chengkun Wang

asked Jan 14 '15 at 18:29

chengkun%20wang's gravatar image

chengkun wang

One Answer:
i like this answer (click again to cancel)
i dont like this answer (click again to cancel)

Unfortunately your question cannot be answered by providing a simple number, but here are a few guidelines to consider.

Critical information regarding the size of your protein and it concentration has not been included in your question. The following is applicable to proteins in the range 10000 - 25000 daltons.

In general I would recommend that you spend sufficient time and effort for optimizing the conditions in an attempt to obtain >98% of expected peaks in HSQC before you proceed to the next steps of data acquisition. The reason is that, the 1H-15N HSQC is the most sensitive 2D NMR spectrum among the standard set of spectra that are generally used for obtaining the data required for structure determination. If you are not able to get a good 1H-15N HSQC, then your chances of obtaining adequate quality data for complete structure determination are lower.

If time is a limitation or if even after optimization you are unable to obtain more than 80% of expected peaks in the 1H-15N HSQC then the following may be considered.

Complete structure determination is not necessary for obtaining biologically relevant information. It is possible that ALL the missing peaks belong to residues in disordered N-terminus or C-terminus, and in such a case you will have no difficulty in obtaining a nearly complete structure if you have more than >90% expected peaks, and with some hard work even 80% would be sufficient, provided that the quality of the remaining spectrum is good.

If the missing peaks are not necessarily due to residues in the disordered N-terminus or C-terminus, but are present in external loops, then also you will not have much problem in structure determination. One way to check if disorder is restricted to external loops and unstructured N/C-termini is to add a non-specific protease to your sample and monitor the 1H-15N HSQC spectra as a function of time, initially for half a day at approximately 1 hour intervals, and then for about a week at 1-2 day intervals. If the distribution of chemical shifts of the original peaks remains similar to your initial spectrum and if a new cluster of sharp peaks appear then it indicates the the protease has left the core and removed the non-core segments which were in intermediate exchange in your original spectrum and are now in fast exchange (because of smaller size) -- (you have to make sure that you are using a version of 1H-15N HSQC that does not use pre-saturation ). If this is the result, then I would expect that you will be able to proceed further and obtain a fairly good structure with your original sample. However, these data may not be conclusive, because many stable proteins are also subject to protease cleavage.

It is important to remember that absence of rigid 3D-structure is not a necessity for biological function. Some functionally active proteins DO NOT have a rigid 3D structure and are disordered.

Another useful experiment to decide weather it will be worthwhile to try to proceed without further optimization of experimental conditions is to obtain a 2D homonuclear NOESY - if the NOESY is of good quality then further data required for assignment could be acquired and it would be likely that you will obtain useful structural information regarding the 'rigid regions' . The partial structure obtained in this manner can then be used in conjunction with molecular modeling to deduce additional information regarding the conformational space of the residues that do not give rise to visible peaks in your 1H-15N HSQC spectrum. A single mobile aromatic residue in the active site, under certain conditions, could be responsible for variation of chemical shifts of many other neighboring residues, leading to intermediate exchange and line broadening.

It is possible that motion of a fragment/domain/set of residues is NECESSARY for physiological function in your case. If you label your protein with both 15N and 13C, then you may be able to obtain sequential resonance assignments for all of the visible (80%) peaks. This will give you information regarding the location of the residues responsible for missing peaks in the sequence - this information will be of biological relevance.

However, as I said at the beginning, it is strongly recommended that you first try to optimize the conditions to ensure that the percentage of expected peaks in your 1H-15N HSQC is as high as possible. In addition to the changes in temperature, pH, concentration, buffer, magnetic field strength, addition of inhibitor could prove to be useful, as many inhibitors reduce the conformational flexibility. See answers to related question in this forum. Unless you have access so some form of high throughput screening, it is not unusual to spend about one year to optimize the conditions - although it takes only about a week to acquire the data that will ultimately be used for the structure determination process.

Note of caution: the set of conditions that are best for obtaining the best 1H-15N HSQC spectra are not always the best (though most of the time they are the best) for obtaining the best possible NOESY type spectra that are used for obtaining distance restraints. Conditions that reduce linewidths always increase the intensity of the 1H-15N HSQC spectra, however, reduction of linewidth may either increase or decrease the intensity of cross peaks in NOESY type spectra. 1H-15N HSQC spectra are used for optimization of conditions because they can be acquired much more rapidly than NOESY spectra and also because the number of 'expected peaks' in 1H-15N HSQC spectra is easy to estimate.


answered Feb 14 '15 at 22:01

sekhar%20Talluri's gravatar image

sekhar Talluri

updated Feb 14 '15 at 23:21

Your answer
Please start posting your answer anonymously - your answer will be saved within the current session and published after you log in or create a new account. Please try to give a good answer, for discussions, please use comments and please do remember to vote (login to vote)
toggle preview

powered by CNPROG