Using computational and experimental methods to understand protein function:

Application to proprotein convertases and other disease-related proteins

Johannes Elferich

June 4th 2015

  1. Using computers to describe proteins
    1. Sequence (Sec61 translocon complex)
    2. Structure (Otub1-monoubiquitination)
  2. How propeptides of proprotein convertases sense pH
    1. Propeptides contain histidines
    2. Histidine protonation destabilizes propeptide structure
    3. Histidine pKa determines pH of activation
  3. Conclusions & Acknowledgments
  • Protein sequences can be represented as strings 
    
>Protein
NLYIQWLKDGGPSSGRPPPS
  • Protein structures can be represented as a list of atoms and their coordinates 
    
ATOM    1  N   MET A   1   27.340  24.430   2.614
ATOM    2  CA  MET A   1   26.266  25.413   2.842
ATOM    3  C   MET A   1   26.913  26.639   3.531
ATOM    4  O   MET A   1   27.886  26.463   4.263
ATOM    5  CB  MET A   1   25.112  24.880   3.649
ATOM    6  CG  MET A   1   25.353  24.860   5.134
ATOM    7  SD  MET A   1   23.930  23.959   5.904
ATOM    8  CE  MET A   1   24.447  23.984   7.620
ATOM    9  N   GLN A   2   26.335  27.770   3.258
ATOM   10  CA  GLN A   2   26.850  29.021   3.898
ATOM   11  C   GLN A   2   26.100  29.253   5.202
ATOM   12  O   GLN A   2   24.865  29.024   5.330
Pairwise comparison
1/7 14% Identical
Multiple sequence alignment
Hidden Markov model
sequence profiles
Park and Rapoport. Annual Review of Biophysics 2012
Insertion of a zinc-finger domain blocks translocation in the presence of zinc
Hypothesis:
Secreted proteins select against small folded domains at their N-termini
Approach:
Use HMM sequence profiles to quantify the number of secreted and non-secreted proteins with small domains at their N-termini
Conti, Elferich et al. Nature Structural and Molecular Biology 2014
Results
Conti, Elferich et al. Nature Structural and Molecular Biology 2014
Type of domains
Disulfides
Conti, Elferich et al. Nature Structural and Molecular Biology 2014
Molecular dynamics
  • Coordinates change according to velocities
  • Velocities change according to forces
  • Forces are calculated according to energy function
Monte-Carlo sampling
  • Structure is modified by a random move
  • New structure is evaluated by energy function
  • If the move leads to less favorable energy it is reversed with a probability based on the energy increase
Otub1 stabilizes p53 by inhibiting the E2-Ligase that polyubiquitinates p53
Monoubiquitination of Otub1 is critical for p53 stabilization
E2-Ligase preferentially binds to monoubiquitinated Otub1
Li, Sun, Elferich et al. Journal of Biological Chemistry 2014
  • E2-Ligase
  • Ubiquitin charged to E2-Ligase
  • Otub1
  • Ubiquitin ligated to Otub1
Li, Sun, Elferich et al. Journal of Biological Chemistry 2014 Juang, Landry et al. Molecular Cell 2012
Lys59
Lys109
  • E2-Ligase
  • Ubiquitin charged to E2-Ligase
  • Otub1
  • Ubiquitin ligated to Otub1
Increased interaction of monoubiquitinated Otub1 with the E2-Ligase cannot be explained by direct interactions within the known E2~Ubq/Otub1 complex.
Li, Sun, Elferich et al. Journal of Biological Chemistry 2014
  • E2-Ligase
  • Ubiquitin
  • Otub1
  • Ubiquitin ligated to Otub1
  • Ubiquitin bound to E2 backside
Mutagenesis of Ser22 blocks binding of monoubiquitinated Otub1 to E2-Ligase
Mutagenesis of Ile44 can rescue binding
Monoubiquitinated Otub1 interacts with E2-Ligase through "backside" interaction of the conjugated Ubiquitin
Li, Sun, Elferich et al. Journal of Biological Chemistry 2014 Sakata, Satoh et al. Structure 2010
  1. Sequences
    • Secreted proteins do not select against small folded domains at their N-terminus
    • Testing of genome-wide hypotheses is possible due to:
      • Abundance of available protein sequences
      • Homology detection and domain annotation using Hidden Markov model sequence profiles
  2. Structures
    • Monoubiquitinated Otub1 binds to E2-Ligase through "backside" interaction
    • Visualization of structures is critical for understanding of protein function and for formulating hypotheses
    • Various algorithms can be used to generate structural ensembles:
      • Molecular Dynamics is biophysically meaningful
      • Monte-Carlo sampling allows for more extensive exploration of conformational space
POMC
Bone morphogenic protein
Cadherin
Matrix metalloprotease
Anthrax toxin
HIV gp160
The Cell. 4th Edition
Name Specificity Cellular localization Tissue localization Involved in disease Knockout phenotype
PC1/3 Dibasic Secretory granules Neurons, endocrine cells Obesity, Diabetes Growth retardation, abnormal hormone levels
PC2 Dibasic Secretory granules Neurons, endocrine cells Obesity, Diabetes Growth retardation, abnormal hormone levels, developmental abnormalities
Furin Dibasic Golgi apparatus, cell surface, extracellular Ubiquitous Cancer, Heart disease, Influenza, HIV, Anthrax Die on embryonic day 11
PC4 Basic motif Granules, cell surface Germ cells Infertility Infertile males
PC5/6 Dibasic Cell surface Development, Neurons Colon tumors Death at birth, bone defects
PACE4 Dibasic Cell surface Development, neurons, glial cells Ostheoarthritis 25% death rate before birth, bone defects
PC7 Dibasic Secretory pathway Ubiquitous Immune defects(?) None
Seidah. Annals of the New York Academy of Science 2011
Mutation of His69 can block furin activation
Homology model of furin propeptide
Histidine protonation states
Feliciangeli et al. Journal of Biological Chemistry 2006
✂
Structures of bacterial subtilisins or animal proprotein convertases in complex with their propeptides. Tube thickness indicates conservation. Occurence of histidine at each position is color-coded in blue.
Hypothesis:
Eukaryotic subtilases that need to sense pH for activation have more histidines in their propeptides than prokaryotic subtilases
Elferich et al. FASEB 2013

PF00082: Subtilase family

  • Dataset of 6533 sequences with start and end position of the protease domain
Protease domain (PF00082) Propeptide

Calculated values:

  • [His]Pro: Histidine content in propeptide
  • [His]Mat: Histidine content in protease domain
  • Δ[His]=[His]Pro-[His]Mat: Enrichment of histidines in propeptide
Elferich et al. FASEB 2013
Δ[His]=[His]Pro-[His]Mat: Enrichment of histidines in propeptide
Positive Δ[His] Negative Δ[His]
Elferich et al. FASEB 2013
Difference of Δ[AA] between prokaryotes and eukaryotes for all residues
Elferich et al. FASEB 2013
Secondary structure content of various subtilase propeptides is similar at neutral pH
Far-UV circular dichroism spectra of the propeptide of subtilisin, aqualysin, PC1/3 and furin
Only histidine containing propeptides show structure loss at acidic pH
Ellipticity at 222nm as a function of pH
Dillon, Williamson, Elferich et al. Journal of Molecular Biology 2012
Activity of propeptide:protease complexes after incubation at various pH values
Dillon, Williamson, Elferich et al. Journal of Molecular Biology 2012
Dillon, Williamson, Elferich et al. Journal of Molecular Biology 2012
Hypothesis:
Furin and PC1/3 activate at different pH values due to differences in histidine pKa values within their propeptides
Paroutis et. al. Physiology 2004
m/z profile of peptide YHF
m/z profile after exchange at pH 9.0
Rate constant of deuterium uptake
3.03.54.04.55.05.56.06.57.07.58.08.59.0pH0.0000.0050.0100.0150.020
Furin
Histidine pKa kmax
H52 6.07±0.02 0.0124±0.0001
H66 5.98±0.03 0.0069±0.0001
H69 6.04±0.05 0.0063±0.0002
H80 6.02±0.03 0.0094±0.0002
H84 6.04±0.07 0.0073±0.0002
PC1/3
Histidine pKa kmax
H67 6.31±0.03 0.0068±0.0001
H72 5.61±0.06 0.0034±0.0001
H75 5.97±0.03 0.0057±0.0001
H85 5.85±0.04 0.0052±0.0001
Elferich et al. under review 2015
NMR structure of PC1/3 propeptide
Predicted pKa values for all 20 structures
Elferich et al. under review 2015
  • Propeptides encode histidines
  • Upon reaching the correct pH for activation the conserved histidine is protonated
    • Furin: H69 (pH 6)
    • PC1/3: H72 (pH 5.5)
  • This triggers a conformational change, exposing the cleavage loop
  • Cleavage of the loop in cis or in trans results in activation
Shinde Lab
  • Ujwal Shinde
  • Danielle Williamson
  • Stephanie Dillon
Committee
  • Larry David
  • Caroline Enns
  • David Farrens
  • Eric Gouaux
  • Mike Harms
Collaborators
  • Mushui Dai
  • Yuhuang Li
  • Bill Skach
  • Brian Conti
Collaborators
  • Buddy Ullman
  • Peter Rotwein
  • Scott Landfear
  • Monika Davare
  • Gary Thomas
  • Jimmy Dikeakos
Biochemistry Department
  • Amber Jones Brunette
  • Christopher Schafer
  • Jon Fay
  • Jessica Martin
  • Gregory Martin
  • Jean Summerton
  • Tony Caps
  • Susan Kozak
  • Valerie Scott
Funding
  • American Heart Association
  • Tartar Fund
  • Vertex Pharmaceuticals
Family
  • Henrike Elferich
  • Christa Elferich
  • Werner Elferich
  • Nick Javidi-Sharifi
  • Mirjam Javidi-Sharifi
  • Bijan Javidi-Sharifi
  • Nathalie Javidi-Sharifi