HOW TO MAKE YOUR OWN KINEMAGES
The tutorial example described on the following pages lets you practice by constructing a kinemage file to illustrate the structural features of ricin (PDB file 2AAI). That file can be obtained from the Protein Data Bank (http://www.rcsb.org), if you do not have it locally. 2AAI, as well as any of the other files mentioned below, can also be obtained from the Richardson lab Web or FTP site at kinemage.biochem.duke.edu.
This tutorial (file Make_kin.txt) leads you through most of the authoring functions for making your own kinemages, including stylistic suggestions on how to choose views, colors, button layouts, ribbons, etc. in ways that will help communicate your 3-D ideas effectively. Several more specialized functions, such as rotatable bonds, are included in the later parts.
Afterward, you may want to read through the explanations in file Pkin_5_4.txt, and use that as additional reference information on how to do things you want in your kinemage. For color names and suggestions on color usage, see Kin.1 of Demo5_4b.kin; Kin.2 explains the kinemage file format for a simple example. If needed, refer to file KinFmt54.txt for a formal description of the various keywords, fields, and parameters of the kinemage file format. If you want to use a kinemage feature you have seen in a previous example, use a word-processor to look at how it was done in that file and construct yours by analogy.
Background reading (optional):
On kinemages: Richardson & Richardson 1992, Protein Sci. 1: 3
Richardson & Richardson 1994, Trends Biochem. Sci. 19: 135
Richardson & Richardson 1999, Int. Tables Crystallogr., vol. F, 25.2.13
On ricin: Rutenber, Robertus, et al. 1991, Proteins 10: 240, 251, 260
KINEMAGE-CONSTRUCTION TUTORIAL - RICIN:
Ricin is a complex toxin capable of crossing membranes and inactivating ribosomes. It is composed of two unrelated polypeptide chains, with the catalytic site in the alpha/beta A chain, two similar antiparallel-beta domains in the B chain, and four bound carbohydrates. Its crystal structure was determined by Robertus and colleagues at the Univ. of Texas, and is PDB file 2AAI.
To start - a simple Calpha-backbone kinemage
For an initial look at the ricin molecule, we will run a simple default Prekin script on file 2AAI. Either drag-and-drop the 2AAI icon onto Prekin, or launch Prekin and open the file from its menu. The output file can be 2aai.kin. When the first dialog box of choices comes up, accept the default "Backbone browsing script", which will execute a simple script producing Calphas, disulfides, and non-water het groups for all subunits in the file. When Prekin is done, it will launch Mage so you can look at the resulting kinemage.
You should see Calpha backbones for the two ricin chains in different colors, with yellow disulfides and several bound sugars (pink). Move it around by dragging with the mouse. Such a simple, default kinemage shows many of the features of the structure, and is useful for many purposes. Note, however, that the viewpoint is arbitrary and the default colors and the names and arrangements of buttons are not ideal. If you want to show particular details and get your point across to someone unfamiliar with the structure, then there are many ways to make the kinemage more informative and persuasive.
Choosing and saving views
First move around to find a view that spreads out the 3 domains in the plane of the screen, with the A chain (the white one) at the top. See if you can enlarge the zoom factor by one or two arrow-clicks without going off the screen edges. Type "s" on the keyboard to toggle into stereo, to make sure your zoom and orientation allow seeing most of the important parts in stereo (you can check for that, even if you can't see stereo yourself). Try improving the depth-cueing by clicking in the top half of the Zclip scroll-bar until something gets clipped away in the back or front, and then back off a step or two by clicking in the bottom half (a zclip of 130 or so will probably work well in this case). Recenter using 'pickcenter' if needed (remember to turn it off again afterward), and try using the keyboard keys "z" and "x" to move the center backward or forward. Once satisfied, choose 'Keep Current View' under the Edit menu,, and save this as View 1, giving it a viewID such as "overview". Move the image, then choose View1 under the View menu, which should reproduce the view you just saved.
Pickcenter near the middle of the A chain and zoom in somewhat. Turn on the "Find" function under the Tools menu. In the dialog box, turn on centering and ask to search for " 177 ", which is the active-site Glu of ricin; Mage will center on Glu 177 and mark it. Choose a view for the A chain that shows both the central beta sheet and this active-site Glu, and save it as View2, with a viewID of "A chain". Now pickcenter between the two domains of the B chain, zoom in, and save a view that shows the domains fairly equivalently, in a vertical orientation to allow for stereo. Locate the place where the chain moves between the two similar halves of the B chain, and make a note of that residue number (which you will use to put the two domains under the control of separate buttons).
Improving the colors and button names, and saving your modified kinemage
Turn on "Change Color" (Edit menu), then click on a sugar atom. Pull down the color-choice list and release on white, and Accept. Turn on "Show Object Properties" (Edit menu), and again pick a sugar. The resulting dialog box will show you the program's internal data structure for this item. Edit the group name to "sugars", and click on the "dominant" option below it (which will hide its subgroup and list buttons on the button panel). Accept the result, and see how the buttons have changed. Then edit the group names for the protein, from 2aai 1 and 2 to "Ricin A ch" and "B chain" (remember that you only get about 11 characters for a group button name, and 9 for a vectorlist name). Click "nobutton" for the subgroups. Now you need to change the colors of the two subunits, A to differ from the white sugars and B to contrast better with the yellow SS bonds (turn on the check box for 'changecolor' on the button panel). The "tint" colors work best for Calpha backbones, because they can be distinguished without overwhelming small features you want to emphasize. Make the A chain yellowtint and the B chain greentint (try out some other possibilities, too).
Choose "Save as", under the File menu; you will be given a dialog box to locate and name the saved file, with a default name that ends in ".kip" rather than ".kin". Quit out of Mage.
Making more pieces to add to the kinemage
The carbohydrates in ricin are "N-linked" - bound to Asn sidechain N atoms. In order to find and show those linkages, and produce other useful vectors for your kinemage, run Prekin again to produce 2 sets of output:
a) vectors for the active-site Glu 177 sidechain in the A chain
b) all the Asn sidechains of the B chain
To achieve this, do the following steps. Name the output file something like "2aai.sc.kin". In the initial dialog box of Prekin, choose "New Ranges". In the range dialog specify residue 177 to 177, check the 'sc' (sidechain) and 'at' (balls for non-C atoms) boxes, the "OK accepts and ends ranges" box, and OK. In the following dialogs accept the defaults for no focus and 0.2A balls, and ask for "only first subunit". When Prekin is done, choose "New pass" (File menu); ask for "New Ranges" again, replace the 3 dots by "asn" for "3-letter residue code", check 'sc', accept and end ranges, no focus and this time subunits in the range 2 to 2 (note that you must check the radio button as well as editing the range numbers). When it's done, quit Prekin.
Using Drawline to draw sugar-Asn bonds, and Prune to delete unused Asn
Launch Mage with your older 2aai.kip file, then choose "Append" on the File menu to add in the new 2aai.sc.kin. Change the color of the Glu 177 sidechain vectors to something bright, contrasting, and oxygenish, such as pink or hotpink.
Choose the B chain view and turn off everything but the Asn sidechains and the sugars. Turn on "Draw New" on the Edit menu. For each of the two long carbohydrate chains, find an Asn whose Nd2 atom is close enough to be covalently bonded (around 1.4A) to a sugar atom; use 'drawline' to add a line for the bond, by picking the two atoms. Now select "Draw New Setup" (Edit menu), and in the dialog box set 0.7 in the "shorten lines" box and check "line ends unpickable". Then, for each of the two-sugar chains, find an Asn whose Nd2 is H-bonded to a sugar atom (2.5-3A distance), and click on the two atoms to add a shortened line to represent the Hbond. (Turn off 'drawline' while measuring the distance, but if you draw an unwanted bond remove it with 'eraselast'.)
Select "Prune" on the Edit menu, which turns on 3 new buttons: 'punch' deletes single vectors, or two that join at the point you picked; 'prune' deletes a connected string of P, L, L vectors; and 'undo p' will undo previous punches and prunes in reverse order, up to about 10 steps back. Find an Asn far from any sugars, and try out these functions. Then remove all the unconnected Asn's, using two prunes at the Cgamma branch point for each one. (Don't prune the pink Glu 177.) Turn things back on and save the modified kin.
Using a word-processor to edit the kinemage file
First be sure to look at Kin.2 of Demo5_4b.kin for an introduction to the kinemage file format. Then use a word-processor to rearrange the vectorlists in the file you just saved. First, cut-and-paste Glu 177 (from the subgroup line down thru the balllist and the vectorlist) after the Calphas of the A chain, with {Glu 177} as its subgroup name, followed by the word "dominant". (And delete the line for its group.)
Set up two subgroups under the B chain group, one called {domain 1} and the other {domain 2}. Look up the residue number you identified as being the switch point between domains in the B chain, and divide the B chain Calphas into two vectorlists, with the first one in bluetint and the second in greentint (copy the vectorlist line and paste it in at the junction point). Remember that you need to duplicate the switch-point Calpha, so it can be both the end of domain 1 and also the start of domain 2 (for the d2 Calpha, put a "P " between the pointID and the coordinates, deleting the "L" if there is one). Put the correct vectorlist in each subgroup. Paste in the unpruned Asn sidechains as two vectorlists, each under the correct subgroup (dividing them at the residue number you had determined). Edit the color to "sky" for the domain 1 Asns (to go with the bluetint Calphas) and "sea" for the domain 2 ones, to go with the greentint.
At the end there should be a "New group" with the 4 bonds you drew between the Asn's and the sugars. Delete the group, subgroup, and list lines, which should leave those bonds suitably at the end of the list of sugar vectors.
Look at the resulting kinemage to see if you did what you intended and whether you like the results. Modify views, if needed. Note anything that needs to be changed, for future editing.
Write text and caption
[If on the PC, do the following in a word-processor, since only the Mac and Unix Mages have an editable text window.]
(In Mage) set "Text Editable" on the Edit menu. Replace the default caption for Kinemage 1 with a short 2 or 3-line description of what is shown and what the colors mean. In the text section, put a title at the top in caps, your name(s) as author(s), then a blank line, "Kin.1- Ricin A and B chain, with Glu 177, SS, and sugars", another blank line, and a paragraph or so about what a reader should look for in the kinemage. (Omit carriage returns except to end paragraphs, to let the text flow properly when windows are resized.) Save the file once more, but don't quit. Examine the kinemage again. If you were the reader, would you learn significantly more from this version than from the original default kinemage? This example was more complex than most, to introduce you to a variety of authoring tools. We hope it convinced you that worrying about how the kinemage looks is worthwhile, and that the process is fun.
KINEMAGE 2 - A RIBBON SCHEMATIC of a B chain BETA TREFOIL structure
Prekin can make either a simple, default ribbon with no secondary-structure information, or else it can do arrows for beta ranges, spiral ribbons for helices, and widened single splines for coil. [To try out the first kind, choose 'Built-in scripts', 'Ribbon', accept defaults, and do just one subunit of ricin.] In order to do the more interesting ribbons for one of the B chain domains, you will need assignments for the beta strands. These are often found in the PDB header, but in this case they were given only for the A chain (look at the file header, to see those assignments). If you want to assign them yourself from the structure, then run the built-in script called 'Main chain and H-bonds' on the second subunit and work from the resulting kinemage. When judging strands from the H-bonding, keep in mind that different definitions are acceptable, but for this purpose it doesn't hurt to be generous. Alternatively, you can use the 12 strands listed here:
139-147, 150-155, 158-163, 172-177, 180-183, 189-194, 201-206, 215-220, 223-227, 230-236, 244-249, 257-262.
Run Prekin with "New Ranges". First do a range with "Ribbon: entire range" and residues 137-262 (to get just domain 2; if you wanted a whole subunit, you'd leave the ranges open). Then check the button for "Ribbon: BETA range" and do 12 ranges, one for each strand (note that for each range you can type a number, tab, number, return). On the last one, use "OK accepts and ends ranges" (but if you forget, then use "OK ends at previous range"). Do no focus, accept the default ribbon controls, and edit the subunit range to 2,2 and also check the "subunits in range above" radio button.
Look at the ribbon kinemage and try to find a view with approximate 3-fold symmetry (if you have trouble, start from View1 and drag toward the upper right, to get 3 big loops around the edge and 3 strands making a triangle around an open center). Save that 3-fold view. This structure is called a beta trefoil (one of Thornton's "superfolds" that occurs in several non-homologous families of proteins). Each third of it has 4 beta strands in 2 hairpins, the first of which form a highly twisted barrel (however, the first 3 strands in sequence go with the last strand). We can use the end of each big loop as the border between thirds (residues 172, 215, and 257 - see if you agree). Save the file, with your view.
In a word-processor, look at the ribbon kinemage, which has vectorlists for the edges of strands, vectorlists with "width4" for the splines in coil, and ribbonlists for the beta strands (and for alpha, if there were any). Search for the first occurrence of " 172 ", which should be a coil vectorlist followed by a beta vectorlist. From where you are onward, do a replace of "sea" with "cyan", to change the strand color. Go back to the top, find " 215 ", and replace "cyan" with "sky" from there down. Then from the top find " 257 " and replace "sky" with "sea" for the last strand. At the top, change '@kinemage 1" to "@kinemage 2". Give the group a name like {ricin B ch} and the subgroup {domain 2}. Just above that is a list of the masters; you can change the order of those lines in order to control the order of master buttons on the screen.
Look at the kinemage, which should now have its 3-fold nature emphasized by color (the 3-fold is only approximate - note the differences caused by the break between N- and C-terminus). Edit the text and caption to say something suitable. If you like, try to find views from the side that show each of the 3 pieces of the trefoil in similar orientations. Save the kinemage again.
KINEMAGE 3 - Active site, with ATOM BALLS, Hbonds, and a ROTATABLE SIDECHAIN
The Prekin 'Focus' option, and other active-site details
Run Prekin on file 2AAI again, with output file 2aaiA.kin. Choose "built-in scripts", then the "caSS" script, and only first subunit. When it's done, choose "New Pass" and "Focus only". In the next dialog box, choose to do the focus on a residue by number. In the "Focus Point Values" dialog, specify residue 177; for radii try 8A for sidechains, 12A for main chain, and 0 for everything else; ignore the special logic controls. Do only first subunit. When Prekin is done, choose "New pass", and ask for ranges. Set the residue range as 208 to 208, replace the 3 dots with "glu", and check the 'ro' box (rotation or mutation); this tells Prekin to set up rotatable bonds for a sidechain. [If you are using an earlier Prekin, such as 2.5, then you can get a rotatable sidechain by checking none of the boxes and giving the 3-letter code.] OK and end; no focus; first subunit. When Prekin asks for a file that contains an example of that amino acid, open PREKINAA.LIB. When it finishes, launch the kinemage.
Pruning sidechains and main chain, and drawing Hbonds
Choose and keep a view that gives a good closeup of the active site cleft. Prune away all the sidechains except for the Glu 177 and 208, the Gln, the two Tyr, the Arg, and the Trp. Turn off the Calphas. In "Draw New Setup" set 'shorten lines' to 0.7 and check "line ends unpickable". Turn on "Draw New", and draw in the sidechain Hbonds: there should be 3, to main chain COs from the Arg, Trp, and one Tyr. Now, prune away all main chain except for one peptide's worth (from Calpha to Calpha) at each of those 3 Hbonds: use 'prune' to take out a residue at a time until you are close, then use 'punch' to trim down to the Calpha (this allows a smooth joining, later, between main chain and Calpha backbone). Turn the Calphas back on. Choose "Change Color" and pick an atom in the Glu 177 sidechain - but this time check the "point" color box before setting the color to, say, hotpink. Do this for each atom in the sidechain, to change its color without creating a new list.
Rotating the sidechain of Glu 208
Glu 177 was known to be the catalytic residue for ricin's nuclease activity on ribosomal RNA, and mutating it to Asp lowers that activity. However, much to everyone's consternation, the "control" mutation to Ala had nearly normal activity. Use the sliders at far right to rotate the sidechain dihedral angles of Glu 208 (there will still be a "ghost" sidechain at its original position), and see if you can get its carboxyl group close to the Glu 177 carboxyl position; could this happen in the Glu177Asp mutant? in the Glu177Ala mutant? An x-ray structure of the E177A mutant showed that this hypothesis is correct. Save the modified kinemage.
Editing the backbone, and adding balls
Edit the kinemage you just saved, changing its number to 3. Delete subgroup line before the Calphas, and delete the group, subgroup, and vectorlist lines between the Calphas and main chain ('mc'). For each of the 3 mainchain peptides, cut-and-paste it between the correct Calphas above. Do the sidechains as a dominant subgroup (note that the end of each vector in Glu 177 has a color name). At the end of the file, delete the "new group" group line, rename its subgroup as {sc Hbonds} and add the parameter "dominant" to it, color its vectorlist purple, and move that whole subgroup up above the rotation group.
Between the sidechain subgroup and its vectorlist, add the following two lines (note that the word "balllist" really does have 3 l's):
@balllist {sc O} color= red radius= .2
@balllist {sc N} color= sky radius= .2
Then COPY and paste all the sidechain oxygen atoms (from Tyr, Gln and Glu) under the first balllist and all the sidechain nitrogen atoms (from Gln, Arg, and Trp) under the second one. This will produce a ball&stick representation for those parts. It is important to put the balls BEFORE their vectors, so that Mage can shorten those vectors appropriately to make the balls look like balls. [Note that Prekin can also generate non-C balllists directly, as you did in Kin.1.] Look at the resulting Kin. 3, and edit text and caption for it. Make a note of anything that needs to be fixed, and fix it. [Note that the N and O of Gln 173 were actually assigned backwards, given the Hbond to Glu 208; if you want, reverse their atom names and ball colors.]
Assembling the final kinemage file
Paste the texts for Kins 2 and 3 into the text section of your original kinemage, with Table-of-Contents entries. Paste first Kin.2 and then Kin. 3 onto the end of your previous file (for each, start with "@kinemage n" and go to the end).
You have now made a quite complicated kinemage file, which illustrates a number of interesting things about ricin. Congratulations!