Structure-based mutation analysis

TODO

re-check all pictures
re-reference all pictures
add all references/quotes


DISCUSSION!

General

According to the UniProt entry about HFE_HUMAN are three 3D-structures of the HFE_HUMAN available, which are listed below. We have chosen the '1A6Z' because it has the best resolution, a very good R-Value (it measures the quality of the model obtained from the crystallographic data), a pH near the physiological optimum and is as good as complete. '1DE4' has a slightly better R-Value and pH, but this PDB also includes the transferrin receptor, which we do not need and do not want in our structure. Also the missing residues of chain A are the same as in the structure '1A6Z' which are only the first three positions. '1C42' is only a hypothetical model, so we exclude it from further research.

Figure 1: stereochemistrical properties of 1a6z

All stereochemistrical properties of the structure are shown in Figure 1.<ref>Lebrón JA, Bennett MJ, Vaughn DE, Chirino AJ, Snow PM, Mintier GA, Feder JN, Bjorkman PJ.: Crystal structure of the hemochromatosis protein HFE and characterization of its interaction with transferrin receptor.</ref>.

PDB	Method	Resolution (Å)	Chain	R-Value	R-Free	pH	Temperature	Completeness	Missing residues (Chain:pos)
1A6Z	X-ray	2.60	A/C	0.233	0.277	6.5	-150.15°C (123 k)	98.0	A:1-3	C:1-3	-	-	-	-
1C42	model	-	A	-	-	-	-	-	-	-	-	-	-	-
1DE4	X-ray	2.80	A/D/G	0.231	0.265	8.0	-173.15°C (100 k)	94.3	A:1-3	C:121,757-760	D:1-3	F:121,757-760	G:1-3	I:121,757-760

We used the following 7 mutations, because we were not able to map 3 of them to the 1A6ZA chain mostly because of position errors. We are using the same names as given in Task 6. Mutation 8 had to be split into two parts because of possible mutation into to different amino acids.

Mutation
Mutation 2 [M35T]
Mutation 3 [S65C]
Mutation 4 [I105T]
Mutation 5 [Q127H]
Mutation 6 [A176V]
Mutation 7 [T217I]
Mutation 8a [C282Y]
Mutation 8b [C282S]

Mapping

Because we have no annotation about the active site so we just visualized the mutations at the '1A6Z' structure in Figure 2. Secondly we are using the same mutations as in Task 6 and have therefore the same problemes with their visualization (only 7 of 10 visualizeable).

All mutations are scattered accross the protein with no affinity to some special region or secondary structure element. Mutations shown in red are near glycosylation positions and mutations in yellow are near disulfidbonds.

Figure 2: 1A6ZA with 7 of 10 visualized mutations taken from task 6

Mutation 2 [M35T]		Mutation 3 [S65C]		Mutation 4 [I105T]		Mutation 5 [Q127H]
Mutation	Superposition	Mutation	Superposition	Mutation	Superposition	Mutation	Superposition
PyMOL	PyMOL	PyMOL	PyMOL	PyMOL	PyMOL	PyMOL	PyMOL
SCWRL	SCWRL	SCWRL	SCWRL	SCWRL	SCWRL	SCWRL	SCWRL
Mutation 6 [A176V]		Mutation 7 [T217I]		Mutation 8a [C282Y]		Mutation 8b [C282S]
Mutation	Superposition	Mutation	Superposition	Mutation	Superposition	Mutation	Superposition
PyMOL	PyMOL	PyMOL	PyMOL	PyMOL	PyMOL	PyMOL	PyMOL
SCWRL	SCWRL	SCWRL	SCWRL	SCWRL	SCWRL	SCWRL	SCWRL

The structure shown in green is the reference structure, the white one is the mutation and the yellow the overlap between the structure of the mutation and reference (please have first a glance at the PyMOL pictures to see where the mutation acutally is because all SCWRL mutation structures are shown completly in white). It is clearly visible that the sidechains/rotamers used by SCWRL are often very different to these introduced by PyMOL. PyMOL does only that the sidechains of the mutated amino acid but SCWRL even recalculates sidechains of non mutated amino acids because of non allowed clashes between them. SCWRL introduces mostly small directional changes to prohibit these clashes. We believe the rotamers used by SCRWL are the more correct ones, because of the fact that SCRWL does also check for possible clashes and that PyMOL is more a visulization tool.

Energy comparison

SCWRL

SCWRL predicts protein side-chain confirmations given a fixed backbone. We are using SCWRL version 4 released in 2009.

Usage:

• use only chain A of backbone pdb: 1A6ZA.pdb
• extract amino acid sequence and change it to lowercase: aa.txt
• introduce each mutation into on seperated aa_x.txt file as capital
  • cmd: scwrl -i 1A6ZA.pdb -s aa_x.txt -o ./mutant_pdbs/1A6ZA_mutant_x.pdb > 1A6ZA_mutant_x.txt

Results:

Mutation	Position	Energy	Energy normalized
Reference	--	247.944	1
Mutation 2 [M35T]	35	252.324	1,017665279
Mutation 3 [S65C]	65	246.695	0,994962572
Mutation 4 [I105T]	105	250.833	1,011651825
Mutation 5 [Q127H]	127	252.368	1,017842739
Mutation 6 [A176V]	176	280.381	1,130823896
Mutation 7 [T217I]	217	260.189	1,049386152
Mutation 8a [C282Y]	282	389.539	1,571076533
Mutation 8b [C282S]	282	255.859	1,031922531

• The energy is normalized by the wild-type structure. A value larger than 1 means that the energy is increased compared to the wild-type. A value smaller 1 shows a decreased energy.

Only mutation 3 shows an decreased energy level which means that this mutation is able to occur more often because it is favoured. All mutations expect 8a are placed around an energy level of 1 what means that mutations should occur at the same amout in population as the reference. Therefore it is very astounding that mutation 8a has the highest increased energy level altough it is the mutation which causes most of all hemochromatosis cases.

Minimise

Minimise is able to minimise the energy of an model.

Usage:

• remove all hydrogen and water atoms from the pdb files with repairPDB
  • cmd: repairPDB 1A6ZA_mutant_x.pdb -nosol > ./repair_pdb/1A6ZA_mutant_x_clean.pdb
• minimise the energy of the models:
  • cmd: minimise 1A6ZA_mutant_x_clean.pdb ./minimised_pdb/1A6ZA_mutant_x_clean_minimised.pdb > 1A6ZA_mutant_x_clean_minimised.txt

Results:

Mutation	Position	Energy	Energy normalized
Reference	--	-3724.153777	1
Mutation 2 [M35T]	35	-5020.465319	1,348082174
Mutation 3 [S65C]	65	-5040.815685	1,353546601
Mutation 4 [I105T]	105	-5028.869826	1,35033893
Mutation 5 [Q127H]	127	-5031.137220	1,350947765
Mutation 6 [A176V]	176	-4957.946411	1,331294761
Mutation 7 [T217I]	217	-5037.718631	1,352714988
Mutation 8a [C282Y]	282	-2596.778899	0,697280256
Mutation 8b [C282S]	282	-5017.057355	1,347167076

• The energy is normalized by the wild-type structure. A value larger than 1 means that the energy is increased compared to the wild-type. A value smaller 1 shows a decreased energy.

The result seems to almost a correct one. According to the energy levels all mutations except 8a occur less than the reference structure because they have an significant increased energy level. But it is not very clear, why mutation 8a shows a radical decreased energy level, which implies that this mutation will occur much more often in the population as the reference structure. This can not be correct and thus should be counted as an error by minimise.

FoldX

FoldX scores the importance of amino acid interactions according to the overall stability of the protein and calculates the energy.

Usage:

• create a runfile tutorial and adjust all default parameters to known (if possible): runfile.txt
• create a listfile of all pdb files that should be included in energy calculation: listfile.txt
• run foldx with runlist
  • cmd: Foldx -runfile runfile.txt > output.txt

Results:

Mutation	Position	Energy	Energy normalized
Reference	--	169.51	1
Mutation 2 [M35T]	35	208.08	1,227538198
Mutation 3 [S65C]	65	206.66	1,219161111
Mutation 4 [I105T]	105	210.39	1,241165713
Mutation 5 [Q127H]	127	205.04	1,209604153
Mutation 6 [A176V]	176	214.31	1,264291192
Mutation 7 [T217I]	217	208.15	1,227951153
Mutation 8a [C282Y]	282	242.23	1,429001239
Mutation 8b [C282S]	282	215.50	1,271311427

• The energy is normalized by the wild-type structure. A value larger then 1 means that the energy is increased compared to the wild-type. A value smaller 1 shows a decreased energy.

The result seems to be a mixture of the result of SCWRL and Minimise. All mutations have an increased energy level, which means they do not occur as often as the reference. But again mutation 8a has the highest energy level, which seems to some strange behaviour associated with this mutation.

Gromacs

• We used the -ignh mode to ignore all hydroxen atom.
• As forcefield, we chosed the AMBER03, CHARMM27 and AMBERGS model.

Energy table for the AMBER03 forcefield.

Mutation	Total Energy Bond	Difference Bond	Total Energy Angle	Difference Angle	Total Energy Potential	Difference Potential
Wild-Type	848,392	1	2707,42	1	-32380,6	1
[M35T]	774,332	0,912705447	2707,35	0,999974145	-32738	1,011037473
[S65C]	695,044	0,819248649	2699,65	0,997130109	-32872,2	1,01518193
[I105T]	0	0	0	0	0	0
[Q127H]	835,05	0,984273779	2782,21	1,027624085	-32213,2	0,994830238
[A176V]	760,573	0,896487709	2734,66	1,010061239	-33030,1	1,020058307
[T217I]	727,061	0,8569871	2712,54	1,001891099	-32609,5	1,007069048
[C282Y]	851,692	1,003889711	2754,62	1,017433571	-31431,1	0,970676887
[C282S]	852,244	1,004540354	2706,07	0,99950137	-32036,4	0,989370178

Energy curve of the AMBER03 forcefield with nstep = 500

Energy curve of the CHARMM27 forcefield with nstep = 500

Energy curve of the AMBERGS forcefield with nstep = 500

time versus nstep plot of the three different forcefields

Wild-Type force field comparisson

Forcefield	Bond	Angle	Potetial
AMBER03	848,392	2707,42	-32380,6
CHARMM27	1064,95	---	-37356,1
AMBERGS	724.545	2785.47	-40390.8

Energy for the Wild-Type

Energy	Average	Err.Est.	RMSD	Tot-Drift (kJ/mol)
Bond	848.392	380	-nan	-2320.49
Angle	2707.42	22	-nan	-96.9545
Potential	-32380.6	1200	-nan	-7696.01

Energy for the Mutation [M35T]

Energy	Average	Err.Est.	RMSD	Tot-Drift (kJ/mol)
Bond	774.332	300	2156.92	-1864.67
Angle	2707.35	16	130.232	-51.7455
Potential	-32738	1100	3905.23	-7119.34

Energy for the Mutation [S65C]

Energy	Average	Err.Est.	RMSD	Tot-Drift (kJ/mol)
Bond	695.044	230	1873.95	-1383.66
Angle	2699.65	9.8	113.651	-42.4234
Potential	-32872.2	890	3424.15	-5775.47

Energy for the Mutation [I105T]
For this mutation, gromacs faild to calculate energies.

Energy for the Mutation [Q127H]

Energy	Average	Err.Est.	RMSD	Tot-Drift (kJ/mol)
Bond	835.05	370	-nan	-2189.82
Angle	2782.21	20	-nan	-94.6097
Potential	-32213.2	1100	-nan	-7254.92

Energy for the Mutation [A176V]

Energy	Average	Err.Est.	RMSD	Tot-Drift (kJ/mol)
Bond	760.573	290	-nan	-1676.2
Angle	2734.66	20	-nan	-125.203
Potential	-33030.1	1000	-nan	-6485.15

Energy for the Mutation [T217I]

Energy	Average	Err.Est.	RMSD	Tot-Drift (kJ/mol)
Bond	727.061	260	1980.17	-1567.84
Angle	2712.54	12	119.206	-50.0057
Potential	-32609.5	980	3633.19	-6435.18

Energy for the Mutation [C282Y]

Energy	Average	Err.Est.	RMSD	Tot-Drift (kJ/mol)
Bond	851.692	370	2500.48	-2249.54
Angle	2754.62	25	158.776	-152.741
Potential	-31431.1	2100	16296.8	-13588.8

Energy for the Mutation [C282S]

Energy	Average	Err.Est.	RMSD	Tot-Drift (kJ/mol)
Bond	852.244	380	2424.32	-2374.34
Angle	2706.07	24	145.766	-110.811
Potential	-32036.4	1200	4277.91	-7896.74

Discussion

TODO: still missing..

References

<references />