Modeling of Protein-Small Molecule Complexes

Part I: Choosing a protein-hetero compound complex

I chose the hetero compound Retinol C₂₀H₃₀O, which has the hetero-atom code RTL. This compound is also known as Vitamin-A and has been found in 119 different proteins whose structures have been determined as listed on the Protein Data Bank website.

Figure 1

The protein I chose was Bovine Holo-Plasma Retinol-Binding Protein, which has the PDB code 1KT6. I picked this compound by searching through the HIC-Up website and then I went to the PDB site to pick out one of the proteins that it was contained in. I downloaded the protein as a pdb file from the RCSB Protein Data Bank and opened it using DS Viewer Pro 6.0. This is the image I produced by making the protein backbone a ribbon and the retinol ligand in the ball and stick form.

Part II: Extracting the Hetero compound

Next I extracted the compound from the protein. I did this by copying the retinol in the same conformation as it was in the protein and pasting into a new window. In this new window I added hydrogens the see if the hybridization was correct or not by comparing the molecule to the molecule I obtained by searching both the Chemexper and the Chemfinder chemical databases. I fixed the sp2 carbons by designating double C-C bonds.

Figure 2

I then imported the image into Chem3D. Here I was able to analyze the steric energy that the compound experienced in the conformation that the protein had forced it into. I then performed and energy minimization which would give the coordinates of the molecules if they were in the free gas state.

Figure 3

I then performed another steric energy summary and then compared the two values:

Table 1

Immeditely I noticed the difference in the stretch term. This means that in the molecule the compound’s bonds were distorted compared to how they would ideally be in the gas phase. In fact 160.8 kcal/mol of the 188.8 kcal/mol for this molecule in the protein is from this stretch factor. The bend energy is the second biggest change in the system. So the main differences in the confirmation of the molecule that the intermolecular forces with the protein induced were in the bond length between the atoms of retinol and the bond angles between the atoms.

Part III: Superimposing the extracted and the energy-minimized hetero compounds

Using the Chem3D program I superimposed the structure in the conformation it was in inside of the protein with the energy minimized structure. By then selecting identical atoms on each molecule I overlayed the two structures. This was the resulting image:

Figure 4

or with outlined atoms:

Figure 5

As can be seen there does not appear to be much difference between the two molecules. Only by looking at the oxygen on the end and the methyl substituents can you even see any difference.

By looking at the compound from another view, in this case the plain of the ring, you can again only see a slight difference in the methyl groups:

Figure 6

Part IV: Bibliographic Information

When I searched for the hetero compound in the Protein Data Bank I obtained 119 different proteins in 22 different classification categories. Here are just a few of them:

Table 2

PDB ID	Authors	First Page	PubMed ID	Journal Name	Citation Title	Volume No	Publication Year
1B0O	S.Y.Wu, M.D.Perez, P.Puyol, L.Sawyer	170	9867826	J Biol Chem	beta-lactoglobulin binds palmitate within its central cavity.	274	1999
1B4M	J.Lu, C.L.Lin, C.Tang, J.W.Ponder, J.L.Kao, D.P.Cistola, E.Li	1179	10047490	J Mol Biol	The structure and dynamics of rat apo-cellular retinol-binding protein II in solution: comparison with the X-ray structure.	286	1999
1BM7	S.A.Peterson, T.Klabunde, H.A.Lashuel, H.Purkey, J.C.Sacchettini, J.W.Kelly	12956	9789022	Proc Natl Acad Sci U S A	Inhibiting transthyretin conformational changes that lead to amyloid fibril formation.	95	1998
1BMZ	S.A.Peterson, T.Klabunde, H.A.Lashuel, H.Purkey, J.C.Sacchettini, J.W.Kelly	12956	9789022	Proc.Nat.Acad. Sci.USA	Inhibiting Transthyretin Conformational Changes that Lead to Amyloid Fibril Formation	95	1998
1BSO	B.Y.Qin, L.K.Creamer, E.N.Baker, G.B.Jameson	272	9827560	FEBS Lett	12-Bromododecanoic acid binds inside the calyx of bovine beta- lactoglobulin.	438	1998
1BSY	B.Y.Qin, M.C.Bewley, L.K.Creamer, H.M.Baker, E.N.Baker, G.B.Jameson	14014	9760236	Biochemistry	Structural basis of the Tanford transition of bovine beta-lactoglobulin.	37	1998

Here are the abstracts from just two of these 22 articles:

$Text Box: Structural Basis of the Tanford Transition of Bovine â-Lactoglobulin†,‡ Bin Y. Qin,§ Maria C. Bewley,|,^ Lawrence K. Creamer,# Heather M. Baker,|,4 Edward N. Baker,*,|,4 and Geoffrey B. Jameson*,§ Centre for Structural Biology, Institutes of Fundamental Sciences and Molecular Biosciences, Massey UniVersity, Palmerston North, New Zealand, and New Zealand Dairy Research Institute, Palmerston North, New Zealand ReceiVed May 4, 1998; ReVised Manuscript ReceiVed July 22, 1998 ABSTRACT: The structures of the trigonal crystal form of bovine â-lactoglobulin variant A at pH 6.2, 7.1, and 8.2 have been determined by X-ray diffraction methods at a resolution of 2.56, 2.24, and 2.49 Å, respectively. The corresponding values for R (Rfree) are 0.192 (0.240), 0.234 (0.279), and 0.232 (0.277). The C and N termini as well as two disulfide bonds are clearly defined in these models. The glutamate side chain of residue 89 is buried at pH 6.2 and becomes exposed at pH 7.1 and 8.2. This conformational change, involving the loop 85-90, provides a structural basis for a variety of pH-dependent chemical, physical, and spectroscopic phenomena, collectively known as the Tanford transition.$

Part V: Protein-Ligand Interactions

Finally I looked into how retinol interacted with the protein. First I went to the PDBsum site and found my protein. I obtained the wiring diagram for the primary and secondary structure of the protein which is the amino acid sequence, the hydrogen bonds, and any disulfide bridges. The red dots represent the amino acids that interact with the retinol:

Figure 7

I next obtained the ligand plot for the retinol, this shows all the nine amino acids that interact with it. Using DS Viewer I selected these same amino acids so their structure could be seen next to the ligand:

Figure 9

Figure 8

In the overall protein this is how the ligand’s environment is orientated:

Figure 10

The retinol is nested right in the middle of a cylinder of beta-pleated sheets. It is surrouned by a histadine, two methionines, three leucines, a tyrosine, a phenylanaline, and a glutamine. Tyrosine, phenylalanine, and histadine all have rings. Leucine, methionine, and phenylalanine are all non-polar, but phenylalanine is aromatic. Tyrosine is also uncharged but contains a phenolic group and is aromatic. Histidine is a polar side chain and is basic, but at basic end of the physiological pH range it is neutral. This protein structure was determined at a pH of about 9 so this could explain why even though it should be charged it is near a neutral hydrocarbon part of retinol. Glutamine is uncharged but contains an amide group. It is this residue that is shown in the ligand diagram to have a hydrogen bond with the retinol between retinol’s oxygen and glutamine’s nitrogen. The other groups, except Histidine, are all non-polar and would give favorable reactions with the non-polar hydrocarbon chain of retinol.

Finally this is the protein chain backbone:

Figure 11