Engineering of protein structures
❑ Overview of mutations
❑ Databases of mutations
❑ Missense mutations
❑ Prediction of mutational effects
❑ Rational design of proteins
Outline
2Engineering of protein structures
Overview of mutations
3Overview of mutations
❑ Mutations in DNA or mRNA may occur
▪ Errors in DNA replication during cell division
▪ Exposure to mutagens (physical or chemical agents)
▪ Viral infections
▪ By scientists’ intervention
❑ Mutations can be harmful or not
Overview of mutations
4Overview of mutations
❑ Location in the DNA
▪ Non-coding region –> affect gene expression (transcriptional
regulation, mRNA stability, translation rates, location, etc.)
▪ Coding region (exons) –> may affect protein sequence
Overview of mutations
5Overview of mutations
❑ Types
▪ Point mutations – a single nucleotide is changed in DNA (or RNA)
▪ Substitutions
▪ Single nucleotide polymorphism (SNP – pronounced “snip”)
▪ Genetic variation; occurs in > 1 % of population
▪ About 10,000,000 in the human genome
▪ Insertions or deletions
▪ Codons have triple nature (3 nucleotides → 1 amino acid)
▪ Potential for frameshift (change in the grouping of
codons, resulting in a different translation)
▪ Can be very deleterious
▪ Other types (duplications, translocations, inversions, etc.)
Point mutations at protein level
6Overview of mutations
❑ Types of point mutations
▪ Silent (synonymous SNP) – no effect on protein sequence
▪ Missense (non-synonymous SNP) – substitution of amino acid
▪ Nonsense – introduction of a stop codon -> protein truncation
Databases of mutations
7Databases of mutations
❑ Human Genome Variation Society
▪ http://www.hgvs.org
▪ Lists all the available databases of human mutations by types
❑ Central mutation databases (>20)
▪ Substitutions in all genes
▪ Variability in protein sequences
▪ Data mainly from literature
❑ Locus-specific databases (about 700)
▪ Substitutions in specific genes
▪ Typically manually annotated
Central mutation databases
8Databases of mutations
❑ Database of Single Nucleotide Polymorphisms - dbSNP
▪ https://www.ncbi.nlm.nih.gov/snp/
▪ Repository for both SNP and short deletion and insertion
▪ For human genome
Central mutation databases
9Databases of mutations
❑ Online Mendelian Inheritance in Man – OMIM
▪ http://omim.org/
▪ Comprehensive database of human genes and genetic phenotypes
Central mutation databases
10Databases of mutations
❑ Human Gene Mutation Database - HGMD
▪ http://www.hgmd.cf.ac.uk/ac/index.php
▪ Comprehensive collection of mutations in nuclear genes
that underlie or are associated with human inherited disease
Central mutation databases
11Databases of mutations
❑ UniProtKB/Swiss-Prot
▪ http://www.uniprot.org/UniProtKB/
▪ High-quality manually annotated protein entries with partial lists of
known sequence variants
Locus-specific databases
12Databases of mutations
❑ For information on gene-specific databases
Missense mutations
13Missense mutations
What are they?...
How can they affect
proteins?
Missense mutations
14Missense mutations
❑ Mutations affecting structure
▪ Stability & folding
▪ Aggregation
❑ Mutations affecting function
▪ Binding & catalysis
▪ Transport processes
▪ Protein dynamics
▪ Protein localization
Mutations affecting structure
15Missense mutations - structure
❑ Major pathogenic consequences of missense mutation
▪ Compromised folding – the protein has modified folds or
presents more unfolded states
▪ Decreased stability – the lifetime of the protein is decreased
▪ Increased aggregation
Mutations affecting structure
16Missense mutations - structure
❑ Molecular basis of mutations affecting folding & stability
▪ Introduced clashes – common for small to large mutations in
buried residues
▪ Loss of interactions – most pronounced effects related to H-bonds,
salt bridges and aromatic interactions
Mutations affecting structure
17Missense mutations - structure
❑ Molecular basis of mutations affecting folding & stability
▪ Altered conformation of protein backbone – mutations concerning
residues with specific backbone angles (especially glycine and proline)
▪ Changes in charge/hydrophobicity
▪ Introducing hydrophilic/charged residue into the protein core
▪ Introducing hydrophobic residue onto the protein surface
NOTE:
• Glycine – the most
flexible amino acid
• Proline – the most rigid
Mutations affecting structure
18Missense mutations - structure
❑ Mutations can reduce solubility or increase aggregation
▪ Alterations on the surface residues may affects the solubility
(ex: reduction of charge)
▪ Hydrophobic mutations can increase protein aggregation
▪ Aggregating proteins usually have high level of β-structures
❑ Aggregation modulated by short specific sequences
▪ Aggregation-prone regions (APRs) are sequences of 5-15
hydrophobic residues
▪ They tend to stack and form amyloid fibrils (cross-β spines)
▪ Some mutations can increase the propensity to form such
amyloid structures
Mutations affecting function
19Missense mutations - function
❑ Effect on binding and catalysis
▪ Binding sites are tuned to bind specific molecules and
stabilize transition states
▪ Mutations can disrupt or improve the binding and catalysis
❑ Example – drug-resistance of HIV-1 protease mutants
▪ Loss of interactions with inhibitors
Ile50Val
37x higher Ki
Ile84Val
3x higher Ki
Means: isoleucine in position 84
was mutated to valine
Mutations affecting function
20Missense mutations - function
❑ Effect on ligand transport
▪ Pathways are adjusted to permit transport of specific molecules
▪ Mutations can speed-up or disrupt the transport, or allow the
transport of different molecules
Leu177Trp => tunnel
becomes almost closed
release of products 500x slower
Mutations affecting function
21Missense mutations - function
❑ Effect on protein dynamics
▪ Dynamics enables proteins to adapt to their binding partners and
interchanging between conformations
▪ Mutations can:
▪ Make regions more rigid (targeting hinge or very mobile regions,
ex.: loops ) -> reduced adaptability
▪ Increase flexibility of rigid regions (targeting residues with many
contacts in mobile elements) -> increased adaptability
▪ These change may affect activity, specificity or even recognition
Mutations affecting function
22Missense mutations - function
❑ Effect on protein localization
▪ After translation, the protein must be translocated to the appropriate
cellular compartment
▪ Translocation can be regulated by short sequences (Signal Peptides)
on the N-terminus, by Translocation Complexes, Chaperones, etc.
▪ Mutations can disrupt or alter the signal, or complex formation ->
protein fails to be transported to the correct subcellular location
▪ Missing protein -> inactive reaction pathways or unregulated
signaling cascades
▪ Mislocalized protein -> active in the wrong cellular
compartment, causing harmful effects
Prediction of mutational effects
23Prediction of mutational effects
❑ Identification of mutable residues
❑ Prediction of the effects on structure
❑ Prediction of pathogenicity
Identification of mutable residues
24Prediction of mutational effects - mutable residues
What are
these?
Identification of mutable residues
25Prediction of mutational effects - mutable residues
❑ The effect of mutations on the protein can be predicted
directly from the role of the modified residue
❑ Mutation of evolutionary conserved residues
▪ Residues important for protein function or stability tend to be
highly conserved over evolution
▪ Mutation of highly conserved residues -> often lead to
destabilization or loss of function
▪ Mutation of highly variable residues -> often neutral
26
❑ Mutations affecting stability & folding
▪ Mutation of residues with many contacts or with favorable
interaction energy -> often destabilizing or compromise folding
▪ Mutation of residues in protein core -> often destabilizing
▪ Small residue to large -> steric clashes
▪ Large to small -> loss of contacts (creation of a void)
▪ Polar to non-polar -> loss of H-bond
▪ Neutral to charged -> introduction of isolated charge
▪ Mutation of residues on protein surface (often neutral)
▪ Polar to hydrophobic -> desolvation penalty (destabilizing)
▪ Mutation involving proline or glycine -> altered conformation
Identification of mutable residues
Prediction of mutational effects - mutable residues
27
❑ Mutations affecting function
▪ Mutation of residues in binding or active sites -> modify binding
or catalysis
▪ Mutation of residues in transport pathways -> modify transport
▪ Mutation of hinge or mobile residues, residues on loops with many
contacts -> modify flexibility
▪ Mutation of residues directing protein localization -> mislocalization
of proteins
Identification of mutable residues
Prediction of mutational effects - mutable residues
28
❑ Tools for annotating (identifying) the role of residues
▪ Individual tools for specific analysis
▪ Evolutionary conservation – ex:. ConSurf, …
▪ Residue contacts – ex: Contact Map Web Viewer, …
▪ Residue interactions – ex: Protein Interaction Calculator, …
▪ Accessible surface area – ex: AsaView, Naccess, …
▪ Binding sites – ex: CASTp, metaPocket 2.0, meta-PPISP, …
▪ Transport pathways – ex: CAVER 3.0, POREWALKER, …
▪ Protein dynamics – ex: NMA, molecular dynamics, …
▪ Protein localization – ex: SignalP, TargetP, Phobius, TMHMM, …
Identification of mutable residues
Prediction of mutational effects - mutable residues
29
❑ HotSpot Wizard – meta-server combining several tools
▪ http://loschmidt.chemi.muni.cz/hotspotwizard/
▪ Homology modelling, MSA, conservation, correlation, pockets and
tunnels detection, docking, stability prediction, design of smart library
Identification of mutable residues
Prediction of mutational effects - mutable residues
Functional hot-spots
Stability hot-spots (evolution)
P
P
P
P
P
P
P
P
P
P
P
P
P P P
W
W
- - - -
G
G
G
G
G
G
G
G
G
G
GP
P
P
P
P
P
P
P
P
P
R
R
R
R
R
R
R
R
R
R
R
M
M
M
M
M
M
M
M
M
M
M
I
I
I
I
I
I
I
I
I
I
I
T
T
T
T
T
T
T
T
T
T
Y
Y
Y
Y
Y
Y
Y
Y
Y
A
A
A
A
A
A
A
A
A
A
A
S
S
S
S
S
S
S
S
S
S
W
W
W
W
W
W
W
W
W
W
W
W
C
C
C
C
C
C
C
C
H
H
H
H
H
H
H
H
S
S
S
S
S
S
S
S
S
S
S
L
L
L
L
L
L
L
L
L
L
L
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
L
L
L
L
L
L
L
L
L
L
LI
I
L
L
Y
Y
Y R
Correlated hot-spots
P
P
P
P
P
P
P
P
P
P
P
P
P P P
W
W
- - - -
G
G
G
G
G
G
G
G
G
G
GP
P
P
P
P
P
P
P
P
P
R
R
R
R
R
R
R
R
R
R
R
M
M
M
M
M
M
M
M
M
M
M
I
I
I
I
I
I
I
I
I
I
I
T
T
T
T
T
T
T
T
T
T
Y
Y
Y
Y
Y
Y
Y
Y
Y
A
A
A
A
A
A
A
A
A
A
A
S
S
S
S
S
S
S
S
S
S
W
W
W
W
W
W
W
W
W
W
W
W
C
C
C
C
C
C
C
C
H
H
H
H
H
H
H
H
S
S
S
S
S
S
S
S
S
S
S
L
L
L
L
L
L
L
L
L
L
L
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
L
L
L
L
L
L
L
L
L
L
LI
I
L
L
Y
Y
W
W
W
R
R
R
K
K
K
K
K
W
W
W
D
D
D
E
E
E
E
E
V
V
V
Stability hot-spots (flexibility)
Identification of mutable residues
Prediction of mutational effects - mutable residues 30
Identification of mutable residues
Identification of mutable residues
Prediction of mutational effects - mutable residues 32
33
❑ Prediction of mutant structures – general workflow
▪ Mutated residue and its surroundings represented by rotamers
from rotamer library (conformations derived form X-ray structures)
▪ The best set of rotamers selected by Monte Carlo approach
▪ Optionally – energy minimization, backbone flexibility
▪ Comparing structures of mutant and native protein -> assessment
of the mutational effect (G = GMut - GNative)
❑ Available tools
▪ Geometric: PyMOL; WhatIF
▪ Energy-based: FOLDX, Rosetta-ddG
▪ Homology: Swiss Model, MODELLER, etc.
Prediction of effects on structure
Prediction of mutational effects - structure
34
❑ PyMOL
▪ https://pymol.org/
▪ Mutagenesis module
▪ User can choose rotamers and visualize potential clashes
▪ Very fast; fixed backbone; no mutational scoring
Prediction of effects on structure
Prediction of mutational effects - structure
36
❑ FOLDX
▪ http://foldxsuite.crg.eu/
▪ Stand alone, with plug-in to Yasara modeling tool
▪ Fast (minutes)
▪ Fixed backbone conformation
▪ Construction of single or multiple mutants
▪ Empirical scoring function for calculation of stability change (ΔΔG)
Prediction of effects on structure
Prediction of mutational effects - structure
37
❑ FOLDX
Prediction of effects on structure
Prediction of mutational effects - structure
Prediction of effects on structure
38
❑ Rosetta-ddG
▪ Under https://www.rosettacommons.org/
▪ Stand alone with bash and python scripts available
▪ Slow (hours-days)
▪ Fixed or flexible backbone conformation
▪ Construction of single or multiple mutants
▪ Empirical force field for calculating structure and stability of
wild-type and mutant
▪ Construction of PDB and prediction of stability change (ΔΔG)
❑ AlphaFold 3, ESM Fold, etc. (ML-based)
▪ Only structural prediction (no stability score)
Prediction of mutational effects - structure
Prediction of pathogenicity
39
❑ Prediction of impact of mutation on protein function
▪ Tools employ machine learning approaches
▪ Trained on functional experimental data
▪ Predictions can be based on sequence only
▪ Qualitative results – i.e. deleterious versus neutral
▪ Primarily intended for pathogenicity prediction (leading to disease)
❑ Available tools
▪ MutPred, SNAP, PhD-SNP, SIFT, MAPP …
▪ PredictSNP – meta server combining a pipeline of many tools
Prediction of mutational effects - pathogenicity
40
❑ PredictSNP:
❑ http://loschmidt.chemi.muni.cz/predictsnp/
❑ Combines many tools for Protein or DNA assessment of SNPs
Prediction of pathogenicity
Prediction of mutational effects - pathogenicity
41
Prediction of pathogenicity
Prediction of mutational effects - pathogenicity
Prediction of pathogenicity
42
❑ There are many more tools out there
Prediction of mutational effects - pathogenicity
43Rational design of proteins
❑ Protein engineering: sometimes we can use mutagenesis to
rationally design proteins according to our needs
❑ Properties that can be modified by mutagenesis
Rational design of proteins
Such as?...
44Rational design of proteins
❑ Protein engineering: sometimes we can use mutagenesis to
rationally design proteins according to our needs
❑ Properties that can be modified by mutagenesis
▪ Stability
▪ Function
▪ Binging site (catalytic activity or substrate specificity)
▪ Macromolecular interface
▪ Molecular tunnels/channels
▪ Solubility
Rational design of proteins
45
❑ Prediction of stability change upon mutation
▪ Structure of mutant protein may not be produced
▪ Tools often employ
▪ Empirical scoring functions
▪ Evolutionary conservation analysis (ex: back-to-consensus)
▪ Machine learning approaches
❑ Available tools
▪ Energy-based: Rosetta-ddG, FOLDX 
▪ Evolution-based: FireProtASR
▪ Hybrid approaches: FireProt, PROSS
Rational design of proteins - stability
Rational design: stability
46
❑ FireProt
▪ https://loschmidt.chemi.muni.cz/fireprotweb
▪ In silico analysis of all possible
mutations
▪ Energy- and evolution-based analyses
▪ Multiple-point mutants for
gene synthesis
Rational design of proteins - stability
Rational design: stability
47
❑ FireProt
Rational design of proteins - stability
Rational design: stability
48
❑ FireProt
Rational design of proteins - stability
Rational design: stability
49
❑ PROSS
▪ https://pross.weizmann.ac.il/step/pross-terms/
▪ Combination of mutations “allowed” by conservation analysis and
Rosetta calculations (energy)
Rational design of proteins - stability
Rational design: stability
50
❑ FireProtASR
▪ https://loschmidt.chemi.muni.cz/fireprotasr
▪ Ancestral sequence reconstruction (ASR)
▪ Automated ancestral inference & phylogenetic tree
▪ Useful to find stable ancestral enzymes
Rational design of proteins - stability
Rational design: stability
51
❑ FireProtASR
▪ https://loschmidt.chemi.muni.cz/fireprotasr
▪ Ancestral sequence reconstruction (ASR)
▪ Automated ancestral inference & phylogenetic tree
▪ Useful to find stable ancestral enzymes
Rational design of proteins - stability
Rational design: stability
❑ RosettaDesign
▪ http://rosettadesign.med.unc.edu/
▪ Monte Carlo sampling (random search) to predict minimum-energy
structure of mutants
▪ Predicts free energy changes upon mutations (G)
▪ Helps design mutations to optimize the binding site and increase
interactions with a ligand/substrate
52Rational design of proteins - function
Rational design: function
❑ PocketOptimizer
▪ https://github.com/Hoecker-Lab/pocketoptimizer/
▪ Aimed at maximizing the affinity of a binding site towards a ligand
▪ Modular pipeline with different tools
▪ Flexibility, docking, mutagenesis, energy calculation
▪ Predicts global minimum-energy designs
53
Rational design: function
Rational design of proteins - function
❑ FuncLib
▪ https://funclib.weizmann.ac.il
▪ To redesign and/or optimize binding site
▪ Utilizes evolution (conservation) and Rosetta calculations (energy)
to introduce multiple-point mutations to modify the properties of
the binding site
▪ Can be used to improve the binding affinity towards a ligand
▪ Outputs up to 50 multiple-point mutants for protein synthesis
54
Rational design: function
Rational design of proteins - function
❑ FuncLib
55
Rational design: function
Rational design of proteins - function
❑ AffiLib
▪ https://affilib.weizmann.ac.il
▪ To optimize protein-protein interface
▪ Utilizes evolution (conservation) and Rosetta (energy) to introduce
mutations and optimize macromolecular interface
▪ Suggests mutations on the interface residues to improve the
binding affinity
▪ Outputs up to 50 multiple-point mutants for protein synthesis
56Rational design of proteins - function
Rational design: function
57
❑ Mutation Cutoff Scanning Matrix (mCSM-PPI2)
▪ http://biosig.unimelb.edu.au/mcsm_ppi2/
▪ To optimize protein-protein interface
▪ Based on machine learning, evolutionary data and energy (FoldX)
▪ Provides mutational G
▪ Modes of calculations
▪ Single mutation – single point mutations on interface
▪ Mutation list – single mutations accordingly to a user
▪ Alanine scanning (all interface residues are mutated to alanine)
▪ Systematic – position saturation (all interface residues are
mutated to all other 19 amino acids)
Rational design of proteins - function
Rational design: function
❑ Aggrescan3D; SoluProt (see lecture 7 - Analysis of protein structures)
❑ SolubiS
▪ https://solubis.switchlab.org/
▪ To identify stabilizing mutations that reduce the aggregation tendency
of a protein
▪ 1) Identifies exposed APRs
▪ 2) Introduces “gatekeeper” residues (P, R, K, D and E) into APRs
▪ 3) Assesses the stability changes of mutations (ΔΔG)
58Rational design of proteins - solubility
Rational design: solubility
References I
❑ Ng, P. C. & Henikoff, S. (2006) Predicting the effects of amino acid
substitutions on protein function. Annual Review of Genomics and
Human Genetics 7: 61-80.
❑ Thusberg, J. & Vihinen, M. (2009) Pathogenic or not? And if so, then
how? Studying the effects of missense mutations using bioinformatics
methods. Human Mutation 30: 703-714.
❑ Potapov, V. et al. (2009) Assessing computational methods for
predicting protein stability upon mutation: good on average but not in
the details. Protein Engineering, Design & Selection 22: 553-560.
References 59
References II
❑ Khan, S. & Vihinen, M. (2010) Performance of protein stability
predictors. Human Mutation 31: 675-684.
❑ Bendl, J. et al. (2016) PredictSNP2: A Unified Platform for Accurately
Evaluating SNP Effects by Exploiting the Different Characteristics of
Variants in Distinct Genomic Regions. PLOS Computational Biology 12:
e1004962.
❑ Musil, M. et al. (2019) Computational Design of Stable and Soluble
Biocatalysts. ACS Catalysis 9: 1033−1054.
❑ Planas-Iglesias, J. et al. (2021) Computational design of enzymes for
biotechnological applications. Biotechnology Advances 47:107696
References 60