Index:
After most of the spin systems have been assigned, it is time to start to do some NOE assignments. This is accomplished using the Cross Peak Assignment Tool. This is probably the tool of the Pronto program you will use most of all. This tool has three purposes:
1. Assign a cross peak. If you enter a cross peak from an n-dimensional spectrum, the tool can search along each of the n axes and suggest possible atoms for assignment. You can then list the cross peaks involved in the assignment of these atoms and transfer all the chemical shift values to the contouring setup window. You will use this method for assigning your NOE cross peaks one by one.
2. Locate a cross peak between two or more atoms. This can be used if you're interested in determining if a NOE actually exists between two or more atoms, e.g. if you have seen from the calculated structure that two protons should be close in space, you can then use this facility to locate a possible NOE cross peak in the spectrum.
3. Determining distances between atoms. If a structure has been determined (either by X-ray crystallography or by preliminary NMR data), you can have the Pronto program calculate distances between pairs of atoms. If the atom specification is not unique (HB#, etc.), an interval will be returned. It is also possible to calculate the interval from a set of substructures.
In this manual, only point 1 will be described in detail. Point 3 will be mentioned during the assignment of the NOE peaks.
After having assigned the intra residue NOE peaks, it is time to analyze the inter residue cross peaks. The peaks already assigned can easily be filtered out from the cross peak list using the setup facility in the cross peak window. You can then start from the top of the list and assign the peaks one by one. If you determine that a peak is an artefact from the cross peak search, remove it from the cross peak list.
A cross peak to be assigned is shown in Figure 1. The cross peak assigner is opened from the main menu, the spectrum in which this peak is located is inserted as the main spectrum, and possibly one or more 2D spectra (e.g. COSY or TOCSY) are inserted as well.
After the cross peak to be assigned has been inserted into the cross peak assignment window, the window will look like Figure 2. Pressing the Search button will start a search along the two axes for atoms matching the two chemical shift values at 8.453 and 3.624 ppm. If you want to look for atoms, it is important that none of the atoms in the two browsers at the bottom have been selected. Selecting an atom in one of the lists works like a hold button on a slot machine, thus a search for cross peaks between two atoms is done by selecting the two atoms in the browsers below and releasing the cross peak in the cross peak browser at the top before pressing Search.
If a 3D structure is available for the protein being analyzed, this could be used for measuring distances between the atom candidates suggested by the search. Select the Structure Cluster to be used in the window to the right of the Current Structure File: text, select Dist On on the two axes, and select an atom in each of the two browsers. The distance will be shown to the right of the Current Distance: text. If one or more of the atoms include a wildcard, or the Structure Cluster contains more than one substructure, a maximum and a minimum distance will be displayed.
The result of a search is shown in Figure 3. The first atom in each of the two lists has been selected. When an atom is selected in a browser, a list of cross peaks in which this atom is involved is shown in a browser under the atom browser.
In this example, a 3D structure of the protein analyzed is available, and the distance between the two atoms selected is calculated: 3.01 Å. This distance agrees well with the apparent intensity of the cross peak, making the assignment: HN in Ala-35 and HA in Val-32 very likely.
Of course using a structure file for the assignment is not always possible (or desirable), so you'll have to check the alignment of the NOE cross peak with other cross peaks referring to the atoms. To make a plot as shown in Figure 4, do as follows. Select the relevant cross peaks from the browsers, and press Set, and select Cross Peak Assign in the contouring setup window. Pressing Make Contouring displays a diagram as shown in Figure 4. The cross peak under the cross hair is the one being assigned. The peak in the window below it is an HN-HA peak from Ala-35, and the cross peak to the right of the cross peak being assigned is an HN-HA peak from Val-32. It can be seen that the three NOESY peaks match up quite well. If you are satisfied with this assignment, simply press the Assign peak in the cross peak assignment window.
The result of the search shown in Figure 3 shows three possible assignments for the cross peak for the second atom. In Figure 5, one of the other atoms has been selected, and an HN-HA cross peak for Val-79 has been chosen. As the two atoms have a distance of 10.57 Å, this assignment is not very likely, unless the coordinate file does not fully reflect the structure of the protein in this part of the molecule.
The result of Add'ing this cross peak instead of the cross peak from Val-32 to the contouring setup window is shown in Figure 6. The initial cross peak is now shown in the window to the upper right. Note, that the center of the cross peak in the window to the upper left is located above the cross hair, giving us reason to discard this assignment.
Assignment of 3D cross peaks is done in a similar way as assigning 2D cross peaks, except that each peak is assigned to three atoms.
Assigning cross peaks in a 3D 15N-HMQC-NOESY experiment is more simple than assigning peaks in a homonuclear 1H experiment, as two of the axes are correlated: Cross peaks assigned to an HN atom must also be assigned to the 15N atom at which this HN proton is attached. This simplifies the assignment of two of the three axes. However, the Pronto program does not know about the correlation of the two axes, so it is up to you to select the proper atom pair.
Figure 7 shows an example of a 3D 15N-HMQC-NOESY cross peak to be assigned. Open the cross peak assignment window, select the 3D HMQC-NOESY as the primary spectrum, a 3D HMQC-TOCSY and a 2D COSY as secondary spectra, and choose this cross peak from the cross peak list at the upper left of this window.
Figure 8 shows the initial setup of the cross peak assignment window in this case. Note that it is necessary to change the tolerance value in the 15N dimension. In this case 0.5 ppm seems to be a good choice. Before any search has been done, the three atom browsers will contain all atoms defined so far, i.e. the program does not pick out the 15N atoms for the browser used in the 15N dimension. Press the Search button to start the search for atoms in the three dimensions.
The result of the search is displayed in Figure 9. The two proton axes have been selected for the search in the PDB structure file, and the first atom of each of the three browsers have been selected. The search has found several candidates in the 15N dimension (the third dimension), however only one of the match the atoms found in the HN dimension (the first dimension). Two of the atoms have thus been assigned uniquely, unless of course errors were made in the spin system assignment.
The difficult part is now to assign the proton in the second dimension. If you determine that a likely candidate is a proton in the same spin system as the one that contains the HN-15N pair, use the spin system buildup window to assign the cross peak to the spin system instead. Start by checking the first candidate in the second atom browser. You can see from the distance between the protons determined by a lookup in the structure file that this might be a good candidate. First Set the cross peak information in the contouring setup window for the cross peak to be assigned. Add the cross peak information for a cross peak in the first cross peak browser having the first chemical shift value identical to the first chemical shift value of the cross peak to be assigned. This will also set the chemical shift value in the 15N dimension, eliminating the need to pick cross peaks in the third browser. Select a cross peak in the second browser, where the second chemical shift value is identical with the second chemical shift value of the cross peak to be assigned.
The result of plotting these cross peaks together is shown in Figure 10. The cross peak at the center of the upper, left window is the cross peak to be assigned (under the center of the cross hair). The cross peak in the window below, at the same HN and 15N values, is the HA cross peak from spin system 35. The cross peak in the upper, right window is the HB proton of spin system 32. Both of the cross peaks seem to match quite well.
In Figure 11, another atom has been chosen from the second atom browser. Add a cross peak from the browser below it to the contouring setup window.
Figure 12 shows the contour diagram displayed when two candidates for assignment in the second dimension have been added to the contouring setup. Note, as we did not Remove the previous cross peak displayed, both candidates will be displayed. The cross peak to be assigned is located at the center of the cross hair. The cross peak at the upper, right window is the same as tested in Figure 10. The cross peak at the upper left window has been assigned to the second atom to be tested. Due to the lack of colors in the printing of this manual, it is difficult to see the exact center of the peak displayed, as the display contains contours both from a 15N-HMQC-NOESY and a 15N-HMQC-TOCSY 3D spectrum, and the two spectra have not been completely aligned. The cross peak in the upper left window is present in both the NOESY and the TOCSY spectrum, however the peak in the NOESY spectrum is located at a little lower field than the peak to be assigned, thus excluding this as a candidate for the assignment. Also as seen in Error! Bookmark not defined., the distance between the protons is 13.4 Å, making this assignment unlikely.
When you have gone through all the possible atoms and their corresponding cross peaks, select the right three atoms from the browsers and assign the cross peak by pressing the Assign button. If three good candidates cannot be found, it is still possible to partially assign the cross peak to the atoms that match, and later make the correct assignment. This could typically be the case if you have not made a complete assignment of all the spin systems. The check facility will give a list of the cross peaks that have only been assigned partially.
Assignment of HN-HN 15N-HMQC-NOESY peaks is especially easy. If the HN proton of spin system i is close to the HN proton of spin system j, then the HN proton of spin system j must also be close to the HN proton of spin system i. This will give a characteristic four cross peak pattern when the right candidates are contoured. If the wrong pair of protons is selected, only three cross peaks will show up. An example of this is shown in Figure 13 (right), and Figure 14 (wrong).
Note the missing HN-HN cross peak in the lower, right window of Figure 14.