About the use of Protein Models

1
About the use of Protein Models

• Jan 28 2003
• University of Basel

2
Classification of protein models

• Two criteria are important
• Model correctness is dictated by the quality of
  the sequence alignment. If the alignment
  contains errors, then the model will have major
  errors.
• Model accuracy is dictated by the deviation of
  the modelling templates used relative to the
  experimental control structure.

3
Sequence alignment error
Dihydrolipoamide Dehydrogenase
4
Comparison of experimental structures as a
function of sequence identity
Rmsd of common core
Sequence identity
Chothia C, Lesk AM (1986) The relation between
the divergence of sequence and structure in
proteins. EMBO J. 5, 823-826.
5
Model quality will determine its possible
applications

• Incorrect models can only be of use if the area
  of interest is right. This can happen in cases
  where large proteins have both well defined
  regions and areas with little known structural
  similarities.
• Correct, but low accuracy models (lt 70 sequence
  identity) cannot generally be used in detailed
  interaction studies (drug design), but have many
  applications

6
Interpreting the impact of mutations on protein
function. The link to disease - CD40 Ligand

• X-linked hyper-IgM syndrome (XHM) is an
  immunodeficiency caused by mutations in the gene
  encoding the CD40 ligand (CD40L). The disease is
  characterized by severely reduced serum levels of
  IgG, IgA, and IgE, with normal to elevated IgM.
  The disease is caused by mutations of the CD40
  ligand (CD40L) gene. CD40L is a type II
  transmembrane protein that belongs to the tumor
  necrosis factor (TNF) superfamily, and is mainly
  expressed by activated CD4 T cells. Interaction
  between CD40L and its counter-receptor CD40
  (expressed by B lymphocytes) is a key signal in
  memory B cell generation and germinal center
  formation. Defective expression of CD40L leads to
  failure to mount secondary antibody responses to
  T-dependent antigens, accounting for the
  increased susceptibility to bacterial infections
  observed in XHM patients.

7
Case study of the Trp 140 to Gly mutation

• CD40L G140 is present and expressed but non
  functional.
• W140 points into the core of the protein.
• The subunit conformation is thus not correct and
  the resulting CD40L form is not functional.

8
Case of the W140 to R mutation
9
Prioritisation of residues to mutate to determine
protein function

• The discovery of gene function in the
  post-genomic era will require a sustained
  experimental effort, which includes the creation
  of molecular mutants. The prioritisation of
  residues to mutate will be greatly optimised by
  considering the 3-D structure of the target
  protein. In many cases, one is then able to
  predict the nature of the change. These
  predictions can then in be interpreted in the
  light of the model.

10
The Fas - Fas Ligand interaction

• Fas ligand (FasL, also called CD95 ligand) is a
  40 kDa type II membrane protein belonging to the
  tumor necrosis factor (TNF) family of proteins.
  This family consists of trimeric ligands that
  induce defined cellular responses upon binding to
  their respective receptors. Receptors of the TNF
  receptor family are type I membrane proteins.
  They are characterized by the presence of
  cystein-rich motives conferring an elongated
  structure to their extracellular domains.
• FasL is one of the major effector of CD8
  cytotoxic T lymphocytes and natural killer
  cells. It is also involved in the establishment
  of peripheral tolerance, in the
  activation-induced cell death of lymphocytes and
  in the delimitation of immunoprivileged regions
  such as the eye and testis.

11
Ligand receptor interaction I

• P206 and Y218 of FasL are predicted to be very
  close to the Fas receptor.
• Several artificial mutants of both residues
  result in 100 to 500 fold less active FasL
  despite normal expression and glycosylation.
• One mutant (Y218D) was surprisingly active and in
  sharp contrast with Y218R which is 10000 fold
  less active than wt

12
Ligand receptor interaction II

• The environment of Y218 is very polar as well as
  basic (R86, R89, K78 and K217).
• The additional mutant Y218F was created to
  confirm this (loss of hydroxyl group on Y218).
• Y218F presented only intermediate activity,
  compatible with loosing the hydroxyl group.

13
Providing hints for protein function C. elegans
insulin-like proteins

• Insulin and related peptides are key hormones for
  the regulation of growth and metabolism.
  Originally discovered in mammals, insulin-related
  peptides have been identified in chordates,
  mollusks and insects. Then, an insulin
  receptor-like gene, daf-2, was reported in the
  nematode Caenorhabditis elegans. The authors
  showed that, as in mammals, this gene is involved
  in the regulation of metabolism. Interestingly,
  they also showed that this gene affects the
  worms longevity. Thus, their finding not only
  demonstrate that the genetic circuitry that
  regulates glucose metabolism was already present
  in the last common ancestor of mammals and
  nematodes, more than 600 million years ago, but
  also suggests a possible link between aging and
  glucose metabolism. Further understanding of
  insulin-like signaling pathway in C. elegans
  requires the identification of the ligand(s) of
  this receptor. Hence we analyzed protein
  sequences issued from C. elegans genome project
  to search for insulin-related peptides.

14
Profile searches 10 new seqeunces
15
Comparison between M04D8.3 (Q21506) and hIGF-1
16
Towards in silico drug design

• Model or x-ray structures of protein targets
• Finding potential binding sites
• Defining the physicochemical and structural
  features of such binding sites
• Fitting known small molecules into identified
  binding sites in silico screening
• Designing new compounds

17
GPCRs

• G protein-coupled receptors (GPCRs) mediate our
  sense of vision, taste, smell and pain. They are
  also involved in cell communication and
  recognition processes. They are a major class of
  drug targets.
• GPCRs can now be modelled with medium accuracy,
  and the resulting models can be used to assist
  drug discovery projects.

18
Validation of modelling and docking methods using
bovine Rhodopsin

• A) Comparison of predicted (green) and x-ray
  structure (blue) of bovine Rhodopsin (RMS of TM
  regions of 3.1 Å)
• B) Comparison of the docked and x-ray structure
  of cis-retinal.

19
DNA gyrase

• DNA gyrase is involved in the vital process of
  DNA replication, transcription and recombination.
  It is a procaryotic Topoisomerase II with no
  direct mammalian counterpart.
• DNA gyrase is a well established target for
  antibiotics (quinolones, coumarins,
  cyclothialidines).
• Resistances to these compounds are a serious
  issue. Thus new classes of compounds are needed.

20
The Target

• The ATP-binding B subunit of the DNA gyrase was
  used as a target for a rational approach.
• A combination of these methods was used
• in silico screening to reduced the compound set
  to screen,
• A biased HTS of DNA gyrase for further reduction,
• Validation of hits based on biophysical
  properties,
• A 3D guided optimisation process.

21
Results

• The in silico screening allowed the reduction of
  the initial data set containing 350 000 compounds
  to 3000 molecules.
• Testing these 3000 selected compounds in the DNA
  gyrase assay provided 150 hits clustered in 14
  classes. Seven classes could be validated as
  true, novel DNA gyrase inhibitors that act by
  binding to the ATP binding site located on
  subunit B
• The 3D guided optimization provided highly potent
  DNA gyrase inhibitors, e.g., the
  3,4-disubstituted indazole being a 10 times more
  potent DNA gyrase inhibitor than Novobiocin.