Index:
After having inserted spectra into Pronto, contoured regions to get familiar with the intensity of the peaks in the spectra, and having identified cross peaks (either manually or automatically), it is now time to start building spin systems from the data in the cross peak catalog.
A spin system in Pronto consists of a list of atoms. Normally, each spin system in Pronto will contain all the atoms of one amino acid residue, i.e. both the peptide part and the ring part of an aromatic residue will be assigned to one spin system. Each atom has one or more links to one of the chemical shift values of a set of cross peaks. The atom itself is named according to the chemical element, and possibly from the chemical shift range in which it appears (HN, aromatic, etc.). A real atom name is first assigned to the connection between the spin system and the atom, i.e. in the process of assigning atoms it is possible to assign the same atom to for instance an HB in one spin system and an Hg in another spin system.
The whole principle about assigning peaks relies on the fact that two cross peaks referring to the same atom in one dimension should be aligned and have identical appearance. Therefore, a close connection between the spin system buildup tool and the contouring setup window is necessary.
In this chapter examples will be given for how you assign 2D and 3D spin systems from scratch, and how you assign 3D spectra when you already have made the 2D assignments. Emphasis will be put on assigning 2D COSY and TOCSY spectra, and 3D HMQC TOCSY and NOESY spectra of 15N labelled proteins.
The starting point of the spin system assignment is: One cross peak. Normally one would start with a potential HN-HA cross peak, and start to build the rest of the spin system, but it is also possible to do it in other ways, for instance from strong peaks in the methyl group part of the spectrum. If you start from the HN part and build a partial spin system, and start from a methyl group and build a partial spin system and later find out that they in fact belong to the same amino acid residue, then you can merge the two spin systems.
A list of possible HN-HA cross peak candidates that have not been assigned can easily be sorted out by using the setup facility in the cross peak window: You select the spectrum to be analyzed, the region in which you expect HN-HA peaks, and that you only want to look at unassigned peaks. Then you simply can start from one end of the list, and assign until the list is empty. If you need a full list of all the cross peaks, or a list using another selection criterion, simply open another cross peak window.
Let us go through the assignment of a peak in a 2D TOCSY spectrum:
The starting point in our search will be the peak shown in Figure 1. Select the peak as the current peak, and open the Spin System Buildup tool window.
The top of the window contains the name of the spin system we are working with. Press the New SS button to start with a new, empty spin system. The line below contains three fields, and each can contain a reference to a spectrum. The first spectrum is the main search spectrum, i.e. the spectrum for which you want to assign peaks. The two other spectra are used to get additional chemical shift information, but they are first useful if you already have assigned the spin system using one of these spectra.
The rest of the window is divided into four major parts.
A The Spin System.
The top part of the window consists of a browser with the atoms currently assigned to this spin system. The right part of the table contains the chemical shift values of the atom, as they have been defined along the different axes. If you have two atoms, as defined by one cross peak between atom 1 in the fast dimension and atom 2 in the slow dimension, you will by default only see two chemical shift values, one in the cell for atom 1/fast dimension and one in the cell for atom 2/slow dimension. Pressing the Auto Set ppm's button will make the Pronto program look for the two missing chemical shift values using other sources: If a peak on the other side of the diagonal has been assigned in one of the extra spectra selected, the chemical shift values from that will be inserted into the two other cells. The cells will then be marked with either X1 or X2. If a peak on the other side of the diagonal has not been assigned in any of the spectra, the program will duplicate chemical shift values between axes referring to the same nuclei. Duplicated chemical shift values from the other side of the diagonal are marked with SP.
The buttons marked I are used to include a specific chemical shift value for the search in that dimension. Pressing more I's in one column includes additional chemical shift values. The I's are also used to select chemical shift values to be transferred to the contouring setup.
The extent of the search depends upon the number of dimensions of the main spectrum, and on the number of columns that have I's selected. This is summarized below:
2D spectrum, one column: This selects a line search. If the Ha chemical shift value is selected, this could be used to look for peaks to new atoms in this spin system.
2D spectrum, two columns: This is a point search, i.e. only peaks occurring within a small square (its size depends upon the tolerances selected) are found. If the duplicated chemical shift values from a HN-HA peak are selected, this will search for a peak on the other side of the diagonal.
3D spectrum, one column: This locates all the peaks in one particular plane in the 3D spectrum.
3D spectrum, two columns: This initiates a line search. If you select the 15N value and the chemical shift value of a proton attached to that nitrogen, and search in an HMQC-TOCSY spectrum, you can find the whole spin system in one search!
3D spectrum, three columns: This is a point search. This will look for peaks inside the small boxes with side lengths set by the tolerances.
B The Search Parameters.
When searching peak lists for candidates for references to the atoms selected, tolerances for the search are necessary. Two tolerances are used for each axis, one involving the main spectrum (currently labelled: Dev: (ppm)), and one involving the extra spectra (labelled: Xsp Dev:(ppm)). Normally a larger value is specified for the tolerances for the extra spectra. Note that the default values supplied by the program are too low when searching in 15N or 13C spectra. It is also possible to limit the search to a specific ppm range.
Pressing the Start search initiates a new search.
C The List of New Atoms.
If the result of the search includes references to atoms not already a part of the spin system, they will appear in this list. The chemical shift values in this list can also be selected, and transferred to the contouring setup window.
D The List of Cross Peaks Found.
The cross peaks found by the search are listed at the bottom of the window. To the right, the atoms connected by the cross peaks are listed by a reference either to the A-list or to the C-list. You can also manually add the current peak to this list by pressing the Add CP button: If nothing happens when you press this button, it is either because no spin system has been selected (press New SS to start a new spin system), or because the peak to be added is not found in the spectrum selected as the main spectrum at the top of the window.
The items found by the search can be removed, either by removing cross peaks or by removing new atoms: If a new atom is removed, all cross peaks referring to that atom will be deleted from the bottom list, and if all cross peaks referring to a new atom are deleted, that new atom will also disappear.
When the Update SS button is pressed, all atoms from the new atom list will be added to the current spin system, i.e. all entries from the C list will be added to the A list. If you added a wrong atom by mistake, you will have to open the spin system zoom window and manually remove the atom from that.
To build a spin system using the cross peak in Figure 1 as a seed, do it as follows:
Open the Spin System Buildup tool, start on a new spin system by pressing New SS, insert the spectrum (in this case a TOCSY spectrum). You may add other spectra, but this is not necessary at this stage. Select the cross peak to be used as seed, either from the cross peak window or from the contouring window. Add the peak to the list at the bottom of the window, by pressing the Add CP button. Unless the peak added is a diagonal peak, two new atoms will appear in the new atom list.
The actions above will lead to a window as shown in Figure 2.
Pressing the Update SS button transfers the two new atoms to the spin system. Press the Auto Set ppm's button to duplicate the ppm values to the other side of the diagonal. You can now assign names to the atoms in the spin system, if you believe that the initial peak is between an HN and an Ha. Select the first peak. Hold down the Change At index button until a menu with atom names appear. This can take a few seconds the first time this menu is popped up. Select the proper atom name from the list. Currently, a naming strategy similar to the one used in X-PLOR is used, except that it is possible to distinguish two atoms if they give rise to two different NMR signals, i.e. if you have two signals from beta-protons in an amino acid residue, but you haven't assigned them stereospecifically, both are named HB#, but as they give rise to two NMR signals, one of the atoms can be called HB#a, the other HB#b.
To start a search for additional peaks in the spin system, select for instance the ppm value of the alpha-proton in the fast dimension. Pressing Start search will now make a search in the cross peak list for cross peaks having a chemical shift value in the fast dimension of 4.694 ppm.
The result of the search is shown in Figure 4: Two cross peaks are found, and one new atom. Because a tolerance is needed during the search in the cross peak catalog, it is always necessary to check carefully all the cross peaks found as candidates to this spin system. Therefore, select the new atoms for plotting one by one, by pressing the I button to the left of the ppm value in the new atom's browser, and include some or all the atoms currently in the spin system. If there are many new atoms, take only a few at a time, and delete the ones you immediately can see does not fit the spin system. If the spin system itself contains many atoms, select only a few of them, at least those used for the search operation. In the example shown in Figure 4, all atoms could be selected, as it involves only two chemical shift values in the fast dimension and three values in the slow dimension. Highlight the I's as shown on the figure, hold down the Set button in the contouring setup window, and select Spin System BuildUp. The selected chemical shift centers will then be transferred to the contouring setup window, using the widths specified in the Set Width fields in the contouring setup window.
It is possible to change the assignment of each of the axes of the new peaks located. Click on the atom assignment to change, and you can assign this axis to:
Pressing Make Contouring generates a plot as shown in Figure 5. Now it is important to understand the three classes of markers available in the spectrum contouring setup window. The markers belonging to the three classes can be distinguished by their size and their color.
0 Class zero contains all the peaks not belonging to any of the other two classes.
1 Class one peaks are those peaks originally in the spin system, i.e. those peaks belonging to the atoms listed at the top of the spin system buildup window, or all the peaks of the spin system if you Set or Add a whole spin system to the contouring setup window.
2 This class contains the cross peaks found during the search in the spin system buildup window, or peaks added one at a time to the contouring setup window, using the Cross Peak field in the Set or Add popup menus.
Selecting proper colors or sizes to the three classes makes it easy to spot the cross peaks of interest in the contouring diagram. However, as this manual is not printed in color, you can't distinguish the cross peak classes in Figure 5. The chemical shift value in the slow dimension of the cross peak in row 2, column 1 (counting from the upper left), was used for a search in the fast dimension. Two cross peaks were located, one in row 1, column 2, and one in row 3, column 2. The peak at the bottom is the cross peak on the other side of the diagonal, and the peak at the top is a candidate for a new atom in this spin system. Apparently, the peaks match up well. It could also be seen that there is another TOCSY effect, directly from the HN, to the candidate. Also note that the cross peak from the HA to the candidate is quite intense, and the chemical shift value of the candidate is in the methyl area of the spectrum, giving us reason to assume (at least as a working hypothesis) that this spin system is an alanine.
Pressing Update SS inserts the new atom in the spin system. It could then be selected at the top and named: HB#. Press Auto Set ppm's and Set All I fill the table with all the chemical shift values. Use them for a search for the remaining cross peaks, as shown in Figure 6. As we have selected chemical shift values from both columns and we are searching a 2D spectrum, we are only performing point searches, thus the search should not result in new atoms. The three cross peaks found can simply be added to the spin system, all the I fields can be enabled, the chemical shift values can be transferred to the contouring setup window, and a full plot of the spin system can be made.
The full plot of the spin system built is shown in Figure 7.
The buildup of spin systems starting from one 3D cross peak works in principle in the same way as in the 2D case. The main difference is that since two of the chemical shift values are fixed, the searches performed will generally have a much higher hit rate.
An example of building a spin system starting from a possible HN-HA-N cross peak from a 15N-HMQC-TOCSY spectrum will be described below.
The starting point for the search is shown in Figure 8. The x axis is the dimension of the spectrum containing signals from protons connected to a nitrogen atom, the y axis contains signals from all protons, and the z axis is the 15N axis.
The setup of the spin system buildup tool is shown in Figure 9. Note, that it is necessary to change the defaults for the chemical shift tolerances for the 15N dimension, and also (due to the low resolution in 3D spectra) for the two proton dimensions. The spectrum used in this example has not been calibrated in the 15N dimension, this makes it necessary to increase the chemical shift range as well to include the whole spectral width. In this window, the cross peak to be used as seed for the search has been inserted in the list at the bottom. Because the peak inserted is a non-diagonal 3D peak, three new atom names appear in the new atom's browser. Pressing Update SS moves the atoms to the top browser, and you can insert atom names, if you wish.
Press Auto Set ppm's and select the chemical shift value of the HN atom in the first dimension and the chemical shift value of the N in the third dimension and press the Start search will result in 6 new peaks for 5 new atoms being found.
Select the chemical shift values of all the new atoms as well as the values from the list above, and transfer the values to the contouring setup window by selecting the Spin System Buildup from the Set popup menu. The result of contouring these ranges is shown in Figure 11. The cross peak shown in the middle section is the cross peak used as starting point, and the peak at the bottom is the diagonal peak in this plane.
Now it is impossible to determine the connectivity between the peaks just by looking at a TOCSY spectrum. To determine what type of amino acid this spin system could be, add a 2D COSY spectrum to the contouring setup window and plot it out. This we will not show in this manual, as it is impossible to distinguish the different contours without colors. But from the contour diagrams it would be clear that the peak at 4.775 ppm is an artifact in the water band, the small peak at 1.905 does not belong to this spin system, and that there is a connectivity from the Ha to the cross peak at 2.027 ppm and from 2.027 to two intense peaks at 0.994 and 0.907 ppm. This makes this spin system a candidate for a Valine.
If the new atoms at 1.905 and 4.775 ppm are deleted, Update SS is pressed, and the proper atom names are inserted, the spin system will look like Figure 12.
Selecting the chemical shift values shown in Figure 12, and transferring them to the contouring setup will result in a contouring window as shown in Figure 13.
Sometimes, pressing the Set button for the Spin System BuildUp tool in the contouring setup window would result in a lot of regions being contoured. This is due to the fact that all combinations of axes are considered. To eliminate this, use the Spin System BuildUp CP entry instead. This transfers only the cross peaks from the buildup tool.