Gaucher Disease: Task 08 - Sequence-based mutation analysis
|Reference||Sequence Position||Codon change||Codon Number||Amino Acid change||One letter code|
|rs368786234||656||AGC ⇒ AGA||77||Ser ⇒ Arg||S77R|
|rs374003673||847||AAT ⇒ AGT||141||Asn ⇒ Ser||N141S|
|CM880035||-||CGG ⇒ CAG||159||Arg ⇒ Gln||R159Q|
|rs374591570||1062||CTC ⇒ TTC||213||Leu ⇒ Phe||L213F|
|CM992894||-||GGA ⇒ GAA||241||Gly ⇒ Glu||G241E|
|rs371083513||1470||GTA ⇒ ATA||349||Val ⇒ Ile||V349I|
|CM960697||-||ACG ⇒ ATG||408||Thr ⇒ Met||T408M|
|CM880036||-||AAC ⇒ AGC||409||Asn ⇒ Ser||N409S|
|CM870010||-||CTG ⇒ CCG||483||Leu ⇒ Pro||L483P|
|CM057072||-||AAC ⇒ AGC||501||Asn ⇒ Ser||N501S|
In our analysis we looked closer to the amino acid properties and their changing characteristics by mutation. We analyzed the structural difference between wild type (WT) and mutation. We also considered their secondary structure and distinguished between helix (H), sheet (E) and loop (C). We also took two different substitution matrices into account, BLOSUM62 and PAM250. Point Accepted Mutation matrix has only positive integer values as scores and is not symmetric. The score reflects the probability of an amino acid to mutate into another. In contrast the BLOcks SUbstitution Matrix has also negative integers and is symmetric. A positive score indicates that a substitution occurs more than random. While a score of 0 shows that the substitution occurs randomly, a negative one points to a mutation less frequent than a random mutation. In case one of our selected mutations has the worst possible substitution score for this amino acid we highlighted the score red in <xr id="ana"/>. To consider also evolutionary information we created different PSSM matrices. These position specific scoring matrices are based on alignments. Just as BLOSUM, the PSSM has positive and negative integer values as scores. A positive value shows that the substitution occurs more often than expected. Critical functional residues, like active site residues, have high positive scores. One PSSM was created with a Psi-BLAST search. The other one is based on an alignment consisting of all mammalian homologous sequences.
|Changes of Physiochemical Properties||Structural Properties||Conservation||Effect|
|Mutation||From||To||Pymol Visualization||Secondary Structure||BLOSUM62 score||PAM250 score||PSSM score||PSSM WT frequency||PSSM mutatant frequency||MSA WT frequency||MSA mutant frequency|
|S77R||polar, neutral charge, sulfur-containing||polar, positive charge, basic||E||-1||6||1||11%||9%||64%||2%||slightly negative|
|N141S||polar, neutral charge, acidic||polar, neutral charge, sulfur-containing||H||1||5||0||10%||7%||55%||3%||neutral|
|R159Q||polar, positive charge, basic||polar, neutral charge, acidic||E||1||5||-4||83%||0%||86%||0%||negative|
|L213F||nonpolar, neutral charge, aliphatic, hydrophobic||nonpolar, neutral charge, aromatic, hydrophobic||E||0||13||3||22%||13%||100%||0%||slightly negative|
|G241E||nonpolar, neutral charge, aliphatic||polar, negative charge, acidic||C||-2||9||-1||10%||3%||83%||0%||slightly negative|
|V349I||nonpolar, neutral charge, aliphatic, hydrophobic||nonpolar, neutral charge, aliphatic, hydrophobic||E||3||4||0||14%||5%||97%||3%||neutral|
|T408M||polar, neutral charge, hydroxyl-containing||nonpolar, neutral charge, sulfur-containing||H||-1||5||-1||4%||2%||82%||0%||neutral|
|N409S||polar, neutral charge, acidic||polar, neutral charge, sulfur-containing||H||1||5||1||10%||9%||76%||2%||slightly negative|
|L483P||nonpolar, neutral charge, aliphatic, hydrophobic||nonpolar, neutral charge, cyclic||E||-3||5||-3||29%||1%||100%||0%||high negative|
|N501S||polar, neutral charge, acidic||polar, neutral charge, sulfur-containing||E||1||5||-2||87%||3%||86%||1%||negative|
Based on the analysis summed up in <xr id="ana"/> we interpreted our mutations:
S77R : The biggest change happens in the secondary structure. While serine has a short and neutral side chain, arginine shows a much longer positive side chain, that probably causes a clash with the flexible loops of the environment. Additional to the change in its polarity the residue switches from sulfur containing to basic. This could destabilize its secondary structure, as the parallel located sheet may be not fixed anymore to the sheet of the residue. The PSSM show only no high frequency for the WT as well as the mutant. With scores of -1 and 6, the substitution matrices identifies the point mutation as expectable. We think the only effect comes from the structural change and has a slightly negative effect.
N141S : The mutation causes no change in its charge and polarity. The affected residue is located in a helix on the protein surface, which let us assume that no effect may occur (neutral). The substitution matrices as well as the PSSM score confirm us to this opinion, as the scores (1[BLOSUM], 5[PAM] and 0[PSSM]) indicate the mutation as nearly random. Also the PSSM frequency tells us that the mutation is rare (7% and 3%) and the WT not very distinct (10% and 55%).
R159Q : The substitution changes the residue from basic and positive charged into a acidic residue without charge. The mutant has a much longer side chain which extends deep into the protein. This structural characteristics as well as the great pH change let us assume that a clash or effect on the structure around this amino acid cannot be avoided. The scores, especially the evolutionary based position specific score, affirms us in our assumption. In the end, the frequency of the WT (83% and 86%) as well as the absence of the mutant in the MSAs, leave us no doubt of the mutation severeness.
L213F : The amino acid is located in a sheet and turns from aliphatic to aromatic. Although, this is a great structural change, there seem to be no clashes or other influences on the neighborhood. Both substitution matrices indicate this to occur on random. Even though, the PSSM shows an mutation appearance more than random and the mutation can be seen in 13% of the alignment sequences, the MSA of homologous sequences consists only of leucine at this position. However, we think that the mutation may be neutral, but uncommon between the homologous sequences. As we are not quite sure, we defined it as possible damaging.
G241E : There is great change in the physicochemical properties caused by the mutation from glycine to glutamine acid, especially for the charge and pH. The residue appears in a loop on the protein surface. Its side chain does not extend into the protein. That is why the property change will not have an effect on the protein structure. The substitution scores deviate from each other. While BLOSUM shows an occurrence less than random, PAM indicates this mutation to happen more frequent. Also the different MSAs show frequencies that make the data interpretation difficult. Due to the combination of the structural and the property characteristics, we would assume the mutation as neutral. But because of the remaining data, we are not sure and declare it as slightly damaging.
V349I : The mutation from valine to isoleucine makes no difference in the properties. Also, both branch chain amino acids differ only slightly in their structure. The PAM declares the substitution as rare, as the score of 4 is the worst for valine, but occurs for several amino acids. Based on evolutionary information the PSSM score defines the mutation as random. This is not surprising concerning their similarity. In the MSAs both amino acids appear at position 349, but valine more often. Based on this observation, we are convinced that this is a neutral mutation between similar amino acids, which occurs once in a while.
T408M : The mutant and the WT differ in their polarity. While threonine is hydroxyl containing, methionine has a sulfur atom in its structure. The side chain causes no clashes with other residues. This substitution happens little less than random. Both amino acids are rare at position 408 in the PSSM alignment. Considering only homologous sequences the WT occurs way more often (82%). The mutant is not observable. We interpret this data as a neutral mutation.
N409S : The mutation causes no difference in charge and polarity. It occurs in a helix and lies close to a sheet, which lets us assume that they stabilize each other. However, the mutation may not destabilize the structure very much. The substitution happens nearly random. In the PSSM MSA, the mutant appears nearly as often as the WT (difference 1%). But the frequencies in the homologous alignment differ much more (difference 73%). We think that this mutation has a minimal effect on the protein that can cause a slight form of the disease, but definitely not deadly.
L483P : There seem to be no great changes in the pysicochemical properties except the cyclic characteristic of proline. The residue is located at the end of a sheet and may be not important for the structure stabilization, but it is obvious that the ring clashes with following residues of the adjoining loop which can be severe for the structure. Especially the BLOSUM and PSSM scores (both -3) show an rare occurrence of the mutation. The frequency balance between mutation and wild type is very dissimilar (29%>>1%). By looking at the evolutionary information of the homologs, we can see that the WT is present in all sequences. This can give as a hint to the severeness of the mutation. The reason for never showing up as well as a rare appearance, can be that the mutation causes not only a disease but death. All information about this mutation let us identify a high negative effect.
N501S : The pysicochemical properties have been described before for mutation N409. The amino acid is located at the beginning of a beta sheet. The side chain does not clash with residues in its neighborhood. The substitution matrix scores show that the mutation appears more than randomly. But the PSSM score shows a lower probability for the mutation. Also the alignments show that the WT occurs much more often (87% and 86%) than the mutant. This frequency tells us that the mutation happens seldom. We suppose this as an observation of selection pressure. So, we assume that the mutation is disease causing.
Comparison of different approaches
First of all we interpreted our collected data from analysis of <xr id="ana"/>. After that we run several predictors of mutation effects. All these results are summarized in <xr id="app"/>. Then we reassessed our analysis by considering the prediction results. In the end we validated our new interpretation (consensus in <xr id="app"/>) against the databases dbSNP and HGMD. In two cases, both databases had contrary mutation information. While HGMD identifies them as disease causing, dbSNP classifies the two mutations as non disease causing. That is why we marked them as possibly damaging. For three mutations (N141S, R159Q, N501S) all predictions, our interpretations, and the validation totally agree with each other. So, we can say for sure that the mutation of asparagine to serine on position 141 has no effect. However, the mutation on position 159 from arginine to glutamine is defintely disease causing as well as the SNP at position 501 from asparagine to serine. In the end, we classified the mutation N409S as disease causing, because the score of SIFT is marginal to disease causing, which would answer our assumption of a slight disease.
|Summary of different prediction approaches|
|Mutation||Analysis of <xr id="ana"/>||SIFT||Polyphen2 (HumDiv)||MutationTaster||SNAP||Consensus||Validation|
|L213F||pdc||0||0.79||100%||0||dc||HGMD[dc] vs dbSNP[ndc]|
|G241E||pdc||0.01||0.89||100%||0||dc||HGMD[dc] vs dbSNP[ndc]|