Scorpion toxin II isolated from the scorpion Androctonus australis hector and insect toxin alpha isolated from the scorpion Leiurus Quinquestriatus Hebraeus share many similarities in both sequence and structure. The insect toxin is a 65 amino acid protein and the scorpion toxin II contains 64 residues. The two sequences are easily alligned. In the alignment below 27 of amino acids are different (out of the 65 total) including the missing N-terminal methionine (in the case of scorpion toxin II).

	

 MET1=VAL2=ARG3=ASP4=ALA5=TYR6=ILE7=ALA8=LYS9=ASN10=TYR11=ASN12=CYS13= 	
     VAL1=LYS2=ASP3=GLY4=TYR5=ILE6=VAL7=ASP8=ASP9 =VAL10=ASN11=CYS12=

     VAL14=TYR15=GLU16=CYS17=PHE18=ARG19=ASP20=ALA21=TYR22=CYS23=ASN24=
     THR13=TYR14=PHE15=CYS16=GLY17=ARG18=ASN19=ALA20=TYR21=CYS22=ASN23=

     GLU25=LEU26=CYS27=THR28=LYS29=ASN30=GLY31=ALA32=SER33=SER34=GLY35=
     GLU24=GLU25=CYS26=THR27=LYS28=LEU29=LYS30=GLY31=GLU32=SER33=GLY34=

     TYR36=CYS37=GLN38=TRP39=ALA40=GLY41=LYS42=TYR43=GLY44=ASN45=ALA46=
     TYR35=CYS36=GLN37=TRP38=ALA39=SER40=PRO41=TYR42=GLY43=ASN44=ALA45=

     CYS47=TRP48=CYS49=TYR50=ALA51=LEU52=PRO53=ASP54=ASN55=VAL56=PRO57=
     CYS46=TYR47=CYS48=TYR49=LYS50=LEU51=PRO52=ASP53=HIS54=VAL55=ARG56=

     ILE58=ARG59=VAL60=PRO61=GLY62=LYS63=CYS64=ARG65
     THR57=LYS58=GLY59=PRO60=GLY61=ARG62=CYS63=HIS64

For this exercise, you are going to model the scorpion toxin II protein based on the structure of the insect toxin alpha using the structure data in the Brookhaven Protein Data Bank file pdb1lqh.ent. Subsequently, you will then compare the modeled structure to the known X-ray structure (PDB file pdb1ptx.ent). The approach we will use DOES NOT WORK VERY WELL. Several better approaches have been published. Some of the issues are as follows:

1) It is better to minimize in steps. For example, first allowing only the changed residues to move, followed by the sidechains (i.e. fixing or constraining the backbone positions) and, only after the sidechains have been optimized, minimizing the entire structure.

2) It seems to be better to include hydrogens, even if the x-ray structure was not of very high resolution.

3) Molecular dynamics can sample more conformational space than minimization and will often lead to a lower energy structure.

4) It makes more sense in a comparison like the one below to minimize both the model and the comparison structure with the same force field, under the same conditions.

WHAT MIGHT I EXPECT TO GET FROM HOMOLOGY MODELING?

See a comparison of experimental structures here

Also note the comparison of 3 structures of a mutant of the protein Barstar from the 3 molecules in the crystal unit cell on this slide.

In other homology modeling tests in Dr. Wampler's laboratory using simple minimization approaches, the results are typically between 1 and 2 Angstroms RMSD for the all-heavy atom comparison. In these cases, a stepwise minimization approach (as described in item 1 above) was used.

Section 1: Preparations

1) Copy the insect toxin alpha structure file pdb1lqh.ent from the class directory (/11/users/BCMB8200/lab3) to a working directory (say lab3).

2) Copy the modified structure of the scorpion toxin II (file 1ptx.pdb) to your working directory.

3) Copy the lab report template (file lab3-tmp.html) to your working directory and change its name as appropriate.

4) Use either Web Majic or the jot editor to begin editing your lab-report file changing it to include your name, etc.

Section 2: Mutating the Insect Toxin Alpha structure

NOTE: KEEP GOOD NOTES! You need to record any variations to the procedure given below. For example, if you make a change out of order or incorrectly and then fix it, note these things in your report.

1) Load the structure, pdb1lqh.ent into Insight II.

2) In order to monitor the progress of your exercise, you may want to:

• color by atom type (Molecule Menu, Color, Color Method = By Atom)
• display only the back bone atoms (Molecule Menu, Display, Only, atom set=backbone, execute)
• Label by residue number and type (Molecule Menu, Label, Monomer, label property=Type_and_number)

3) Delete the methionine, residue #1

• Biopolymer menus
• Residue menu
• Delete
• enter "LQH:1" in the text box, then click "Execute"

NOTE: Residue selections from the command line my include comma's or dashes. E.G. LQH:1-10 selects the first 10 residue of the molecule LQH. LQH:1,3,5,7 selects residues 1, 3, 5 and 7.

4) Replace each residue in turn so that the sequence of the 1lqh file becomes that of the 1ptx protein (see list below):

• Biopolymer menus
• Residue menu
• Replace
• Enter the residue specification in the Residue to Replace box (e.g. LQH:3)
• Select the replacement residue name (e.g. LYS)
• Repeat this process for each of the 26 residues that must be changed

	List of replacements:
	
	ARG3->LYS	VAL14->THR	GLY31->LYS	ALA51->LYS	LYS63->ARG
	ALA5->GLY	GLU16->PHE	ALA32->GLY	ASN55->HIS	ARG65->HIS
	ALA8->VAL	PHE18->GLY	SER33->GLU	PRO57->ARG
	LYS9->ASP	ASP20->ASN	GLY41->SER	ILE58->THR
	ASN10->ASP	LEU26->GLU	LYS42->PRO	ARG59->LYS
	TYR11->VAL	ASN30->LEU	TRP48->TYR	VAL60->GLY
		

NOTE 1: You may monitor your results at any time by using the List function from the Molecule menu.

NOTE 2: YOU SHOULD SAVE YOUR INTERMEDIATE RESULTS TO A FILE PERIODICALLY.

5) "Cap" the ends of the protein (fix the protonation of the N-terminal and the terminal oxygen of the C-terminal)

• Biopolymer
• Modify
• Hydrogens
• Select capping mode, "Charged"
• Select execute

6) Renumber the sequence

• Biopolymer menus
• Protein menu
• Rename
• In "Old Range" box, enter "LQH:2-65"
• In "New Range Start" box, enter "LQH:1"
• "Cancel"

7) Check your new sequence against the table at the top of this document.

8) Fix any problems in sequence.

9) Fix the hydrogens and protonation for pH 7.0

• Biopolymers menus
• Modify
• Hydrogens
• Set pH
• Enter 7.0
• Execute

10) Save your altered protein to a PDB file.

11) Save a Biosym format file with the proper cvff potentials.

• Select the "Builder" Menus
• Select ForceField menu
• "Select"
• Click the "Clear_Charges" button
• Select cvff.frc if not already the default
• Select "Execute"

• Select the Forcefield Menu again
• Select Potentials
• "Fix" for all three items
• Select "Execute"

  NOTE: the total charge should now be 1.0

• Save this structure in Biosym format (Molecule Menu, Put,...)

Section 3. Initial Comparison.

The comparison structure has been modified from the Brookhaven structure. The Brookhaven file, pdb1ptx.ent, containing the scorpion toxin II protein contains a number of residues that have alternate positions. In each case the lowest occupancy position has been edited out of the file. This structure, unlike the insect toxin alpha structure, does not contain hydrogen atoms. Thus, in order to compare the modeled structure with it, you will have to add hydrogens.

1) First, load the modified scorpion toxin II structure from the file, 1ptx.pdb (located on the class directory).

2) Add hydrogens to the 1PTX structure (Biopolymer menus, Modify, Hydrogens, Set pH, enter "7.0", select PTX, execute)

3) Compare the two structures (Transform, Superimpose, select your model then select PTX, select "End_Definition", then "EXECUTE").

4) Record the initial heavy atoms RMSD in your report.

5) Repeat the comparison for the backbone atoms and list that value in your report as well.

6) Examine and document the comparison of your initial model with the x-ray structure in your report.

Section 4. Refinement by Minimization.

The goal of this section is to refining your initial model using molecule mechanics minimization. Some intial comments and preparations might help you have a successful modeling effort.

A. Start each sequence of work by clearing the display of all structures and loading your saved structure (Biosym archives file) of your initial model. This way you will begin at a well defined starting point each time.

B. When in doubt save a working copy of your structure.

C. Make careful notes so that you can communicated just what you did.

1) Optimize the structure from the Builder menu and save the optimized structure.

NOTES:

1) Use the program defaults except where you are told to change them.

2) You might want to set the iteration count to 100 for the first few optimizations just to make sure everything is working. Examine the .out file to see any problems that might have arisen.

3) Note the change from "Steepest Descents" to "Conjugate Gradient" after about 450 iterations.

4) Note the change from "Conjugate Gradient" to VA09A minimization toward the end (~1000 iterations and energy <245). VA09A stands for the BFGS, quasi Newton-Raphson method WHICH IS MUCH SLOWER.

5) YOU WILL NOT HAVE TIME TO MINIMIZE TO FULL CONVERGENCE (~7000 iterations), but the structure when the energy drops below 245 (~1200 iterations) will change very little in terms of the comparison below.

6) IF YOU WANT TO MINIMIZE MORE FULLY, TRY USING THE DISCOVER MODULE AND RESTRICTING THE METHOD TO CONJUGATE GRADIENT

2) After you have stopped minimizing, document the change in both the ALL-heavy atom and backbone RMSD relative to the experimental structure in the 1ptx.pdb file (Remember to add pH 7.0 hydrogens to the 1ptx file before comparing). Each step in minimization will generate a .cor (coordinates file) and you may load each one and see how the RMSD's change as the optimization progresses.

3) DID OPTIMIZATION IMPROVE THE MODEL BASED ON THE RMSD'S?

4) Document the comparison between this structure and the X-ray structure.

Within your descussion of the comparison, be sure to point out and discuss the following:

• Why some of the outer sidechain groups don't flip over into the positions found in the X-ray structure.
• The difference between the backbone RMSD and that for all of the heavy atoms.
• Be sure you illustrate your discussion with some appropriate images.

NOTE: THIS IS AN APPLES AND ORANGES COMPARISON, since the X-ray structure was optimized under different environmental conditions and with a different force field than you are using.

Section 5. Finishing the lab and turning it in.

1) Examine your lab report using the Netscape browser.

2) Clean up your directory of any unnecessary files, BUT DO NOT REMOVE THE MINIMIZED STRUCTURE OR THE VARIOUS FILES FROM THE MINIMIZATION THAT GO WITH IT. You'll have to pay attention to the automatic file names created by the minimizer. If you have to restart a calculation in a new session, be sure you load the .car file (not the .cor file).

3) Be sure your permissions allow Dr. Wampler access to your lab report and its accompanying files.

4) Send Dr. Wampler an e-mail message indicating the location of your lab report.