Predicting 3D Protein Structure using Homology Modeling

1
Predicting 3D Protein Structure using Homology
Modeling

• By Affan Kayser Diadji Wague Hua Yang
  Jordan Liz Tanjina Nadia Adam Nop Kasarah
  Allen W. Light A.
• City College Bioinformatics workshop

2
Introduction

• Proteins play an important roles in
  various diseases. Through analyzing the
  properties and characteristics of proteins, we
  try to find their involvements in diseases.
  Structural analysis of the proteins pave our ways
  to find solutions to certain diseases. In order
  to analyze proteins which structures havent been
  determine yet, we use homology modeling to model
  proteins by using appropriate templates. MOE will
  help us do the homology modeling and be able to
  reveal the significances of the proteins in
  diseases from their structural analysis.

3
What is Homology Modeling?

• Homology modeling is based on the assumption
  that, if two proteins have similar sequences then
  they will have similar structures. If the
  sequence identity is greater than 90 then the
  predicted structure is practically the same as
  the template structure. If the sequence identity
  is lower than 25 then the predicted structure is
  not reliable. If the sequence identity is in the
  range 50-80 then the homology modeling can
  generate reliable and informative structure.
• There are many software for homology modeling,
  the general procedure is
• Search for the template based on sequence
  similarity using Blast (an online software
  provided by the NCBI database)
• Select the candidate template to make sure the
  E-value of the sequence alignment is less than
  10-4 and sequence identity is greater than 50.
• Input the target sequence and the template
  sequence and structure into homology modeling
  software to do modeling, for example SWISS MODEL
  (http//swissmodel.expasy.org/SWISS-MODEL.html)
  ,or MOE.
• Check the accuracy of the generated structure
  model.

4
Sequence Alignment
5
MOE
www.chemcomp.com

• We are using MOE (molecular
  operating environment,) to do the
  modeling and this software will do all the steps
  for you when you input the sequences to be
  modeled.
• We all found our protein sequences from the NCBI
  website and save in FASTA format.
• We copied these sequences to MOE and searched for
  appropriate template.
• After sequence alignment, we start to model the
  target protein structure based on the template
  structure.
• Verify the local geometry of the generated model
  using Ramachandran Plot.

6
About TAL1

• Tumor specific alteration of TAL1 occur in
  patients with T-Cell Acute Lymphoblastic Leukemia
  (T-ALL)
• This alteration is caused by chromosomal
  rearrangement
• Alteration of TAL1 may lead to leukemogenesis
• Related to TAL2 and LYL1. Thus, has an etiologic
  role in leukemia

7
Continue.

• Contain a) amino acid residues 1-331
• b) truncated species ( residues 176-331)
• Both have 1. basic helix-loop-helix motif 2.
  protein dimerization 3. DNA-binding domain that
  were found in transcription factors

As products
subgroup of bHLH protein which is a potential
mediator of T-cell leukemogenesis
8
PDB ID 1MDY As template

• TAL1 polypeptides do not have intrinsic DNA
  binding activity
• cant self-associate to form bHLH homodimers
• May be able to interact with other bHLH protein
  in a stable manner
• Two heterodimers(TAL1-E12 , TAL1-E47) that
  recognize E-box motif
• structure model by MOE

Loop
Beta
Helix
E-box motif a element found in various
eukaryotic transcription enhancers
9
Ramachandran Plot
It is a way to visualize dihedral angles f
against ? of amino acid residues in protein
structure
10
JunB

• Psoriasis is an inflammatory disease of the skin
  and joints
• JunB is an element of the AP-1 transcription
  factor and is known to coordinate cell
  differentiation, stress responses, cytokine
  expression in various organs and proliferation.
  But when a person develops psoriasis, the
  expression levels of JunB lessen.

11
The structure of. CAG33122 protein using MOE,
PDB 1UAX
12
This is the HIF1 protein structure done by MOE
homology modeling. The PBD code 1P97
13
Protein analysis of HIF1

• The predicted structure on the previous page
  displays the HIF1 protein, which is the main
  regulator of hypoxia-induced gene expression as
  well as the DNA binding protein. It is a
  heterodimeric transcription factor composed of
  HIF1A.
• The HIF1 protein expressed by the structure
  concludes that it plays a big role in hypoxia.
  However, even though many proteins are known,
  this protein helps identify what hypoxia is
  really about since it does associate with the
  HIF1A gene.

14
1AQN- Structure modeled by MOE
15
Subtilisin Mutant

• Acts as an Intramolecular Chaperone
• Subtilisin-like serine Proteinase
• It facilitates folding by acting as a template
  for its protease domain, which does not form A
  part of it.
• There are two Subtilisin

16
Subtilisin Mutant cont.

• Altered and wild type Subtilisin
• They have identical amino-acid sequences
• Differs in thermostability and substrate
  specificities.
• Through a mutated intramoleculer chaperone
• An identical polypeptide can fold into an altered
  conformation
• It also maintains memory of the folding process.

17
LDLR Structure

• This is the back bone to the structure of the LDL
  receptor protein.

• This structure was created by Moe. Its the
  template for the LDLR protein.

18
Space Filling Model

• Space filling molecular models show the relative
  atomic sizes of the atoms of the molecule.
• This model shows the carbons (gray), hydrogen's
  (light gray), oxygen (blue), carbon monoxide's
  (red), and sulfur (yellow) of the protein.

19
LDL Receptor

• Knowing the structure of the LDL receptor
  scientists can better understand the effects it
  has on Familial Hypercholesterolemia patients.
• For example, scientist can find why the LDL
  receptor fails to activate and bind to the
  surface and can learn how to provide LDL
  receptors to FH patients.

20
What Is HER4?

• HER4 , also known as ErbB4, is able to lower the
  expression level of breast and prostate tumors.
• ErbB4 is a transmembrane receptor, which is a
  protein that covers the plasma membrane of the
  cell, that controls cell reproduction and
  differentiation.
• It is apart of the HER family.
• Has to be enzymatically separated which may
  change the function of its Intercellular
  Domain(ICD) as stated in several reports

21
ABG35747 HER4 MOE Structure
22
Conclusion

• The predicted structures of our target proteins
  help us analyze the significances of the proteins
  and its involvements in diseases. Thus, we have
  better understandings of the diseases.

23
Further Research
We will continue to create models of
unstructured proteins using modeling softwares
other then MOE to see if they yield similar
results.
24
References

• http//www.epitomics.com/products/1200-1.htm
• lhttp//www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db
  proteinval66393602
• http//www.ncbi.nlm.nih.gov/BLAST/Blast.cgiR
• http//breast-cancer-research.com/content/8/2/R19
• http//www.findarticles.com/p/articles/mi_m0DFY/is
  _03_2001/ai_70738971
• MOE- Molecular Operating Environment
• http//www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db
  proteinval31077211
• http//www.wrongdiagnosis.com/medical/hif1_protein
  .htm
• http//www.genecards.org/cgi-bin/carddisp.pl?gene
  HIF1A
• Bacterial lipopolysaccharide induces HIF-1
  activation in human monocytes via p44/42 MAPK and
  NF-kappaB.Frede S, Stockmann C, Freitag P,
  Fandrey J.
• Wegele H, Muller L, Buchner J. (2004). Hsp70 and
  Hsp90 - a relay team for protein folding. Rev
  Physiol Biochem Pharmacol 1511-44
• uhttp//www.ncbi.nlm.nih.gov/BLAST/Blast.cgi
• uhttp//www.ebi.ac.uk/cgi-bin/clustalw/result?tool
  clustalwjobidclustalw-20060809-16204016treendi
  sphidetreetypenewsortbyseqnocoloryes
• uhttp//www.ncbi.nlm.nih.gov/gquery/gquery.fcgi
• uhttp//ad2004.com/Biblecodes/Hebrewmatrix/leukemi
  a.html
• http//www.humpath.com/IMG/jpg/psoriasis_nail_20xp
  asd-2.jpg
• http//www.nature.com/nature/journal/v437/n7057/fu
  ll/nature03963.html
• http//www.nycornell.org/dermatology/research/img/
  psoriasis.jpg
• http//www.chemcomp.com/software-pro.htm

25
Acknowledgements
Special Thanks To

• Dr. Sat Bhattacharya
• Dr. Yuying Q. Gosser
• Joe Wu
• Harlem Children Society
• Bioinformatics workshop at City College of New
  York