The Ontario Structural Genomics Initiative

1
The Ontario Structural Genomics Initiative
2
REFERENCES

• Nature Structural Biology, 7, 903-909, 2000
• Journal of Molecular Biology, 302, 189-203, 2000
• Nature Structural Biology, SG supplement, Nov
  2000
• Structure, 6, 265-267, 1998
• Nature Structural Biology, 6, 11-12, 1999
• Current Opinion in Biotechnology, 11, 25-30, 2000
• Nature Genetics, 23, 151-157, 1999

3
STRUCTURAL GENOMICS

• The determination of the three-dimensional
  structures of the proteins encoded by the genes
  from an entire genome.
• The complete DNA sequences of many organisms are
  known and there are 100 ongoing genomic
  sequencing projects.
• The natural extension of sequencing projects is
  the determination of the corresponding protein
  structures.
• The goals of current genomics projects are to
  understand the cellular and molecular functions
  of all the gene products. Ultimately to help in
  the design of diagnostics and therapeutics.

4
SEQUENCED GENOMESNCBI Genome Database

• A. aeolicus (1522) M. thermoautotrophicum
  (1855)
• A. fulgidus (2407) M. jannaschii (1715)
• B. subtilis (4100) M. tuberculosis (3918)
• B. burgdorferi (850) M. genitalium (467)
• C. elegans (19 099) M. pneumoniae (677)
• C. trachomatis (1052) P. horikoshii (1979)
• C. pneumoniae (894) R. prowazekii (834)
• E. coli (4289) S. cerevisiae (5885)
• H. influenzae (1709) Synechocystis sp.(3169)
• H. pylori (1566) T. pallidum (1031)
• A. thaliana (15 000 ) H. sapiens (?30 000)
• LEGEND Archaea Bacteria Eucarya

5
THE PROTEOMICS CHALLENGE
What do all those proteins do?
Similar Function
Known Function

• Any Genome

Unknown Function
6
FUNCTIONAL PROTEOMICS
uncovering the function of all genes/proteins

• Genome Wide Analysis
• protein-protein interactions
• protein expression/localization
• biochemical assays
• protein structure

Known Function
Unknown Function
7
BEYOND SEQUENCING PROJECTS
GENOME
DNA Microarray
Genetic Screens
PROTEOME
Protein Ligand Interactions
Protein-Protein Interactions
Protein Structure
8
THE POST-GENOMIC ERA

• Functional proteomics currently exploits
• several complementary technologies
• DNA Microarray Technology
• For genome-wide transcription profiling
• Protein-Ligand Interactions
• To discover small molecule inhibitors of proteins
• To discover function
• Protein-Protein Interactions
• To define the network of regulatory interactions
• To discover function

9
PROTEINS WITH 3D HOMOLOGS
of Proteins
10
MAKING STRUCTURAL GENOMICSA REALITY

• Initially the rate determining step in SG
• was preparing suitable protein samples.
• - Need faster methods in protein production
• - Must overcome bottleneck of growing crystals
• - Initiated program directed solely at this
  issue

11
GOALS OF STRUCTURAL GENOMICS

• to develop improved methods that will result in
• high-throughput biology and protein structure
  determination
• robots, robots, robots
• cloning
• expression
• purification
• crystallization
• to determine new protein folds
• to determine the functions of unknown proteins

12
STRUCTURAL GENOMICS
The early years

• A move away from hypothesis driven researcha
  system where structures are solved first followed
  by asking questions about the protein later.
• A large number of targets are required from which
  high-throughput methods must be implemented for
  such a project to be successful
• Cloning, expression and purification are
  important!!
• What targets?
• What is the priority of targets?

13
STRUCTURAL GENOMICS PROJECTS

• A. Edwards U of T 20 M. thermoautotrophicum
• S.H. Kim Berkeley 12 Methanococcus jannaschii
• S. Yokoyama Tokyo U 10 Thermus thermophilus
• J. Moult CARB 10 Haemophilus influenzae
• D. Eisenberg UCLA 8 Pyrobaculum aerophilum
• A. Sali BNL 3 S. cerevisiae

14
SG CONSORTIUMS

• The NIH/NIGMS have funded 7 SG centers with each
  center obtaining about 4 million US per year in
  funding.
• New York SG Consortium (www.nysgrc.org)
• Midwest Center for SG (UHN/UofT)
• The Berkeley SG Center
• Northeast SG Consortium (UHN/UofT) (www.nesg.org)
• Tuberculosis SG Consortium (www.doe-mbi.ucla.edu/T
  B)
• The Southeast Collaboratory for SG
• The Joint Center for SG (www.jcsg.org)

15
SG COMPANIES

• Integrative Proteomics Inc. Toronto
• (www.integrativeproteomics.com)
• Structural Genomix Inc. San Diego
• (www.stromix.com)
• Syrrx Inc. La Jolla
• (www.syrrx.com)
• Astex Inc. Cambridge
• (www.astex-technology.com)
• Structure-Function Genomics Piscataway

16
CRYSTALLOGRAPHIC DEVELOPMENTS

• Multiwavelength Anomalous Dispersion
• Synchrotron Radiation
• Cryocrystallography
• CCD Detectors and Image Plates
• Software

17
STRUCTURAL BIOLOGY OVER THE YEARS
Structural biology on a genomic scale
1998
Target
Sample
Structure
TIME
18
Overview of Structural Proteomics

• Genome Analysis and Target Selection
• Cloning, Expression and Purification
• Crystallography NMR
• Structure
• Fold and Functional Analysis

FAST
SLOW
FAST
19
STRUCTURE SHOW AND TELL

• The structure will reveal the fold of the
  protein.
• TIM barrel, Rossmann fold

20
STRUCTURE SHOW AND TELL

• The structure will reveal the active site.
• protease (Ser-His-Asp)

21
STRUCTURE SHOW AND TELL

• The structure may reveal evolutionary links
• between proteins lacking sequence similarity.

22
STRUCTURE SHOW AND TELL

• The structure may reveal the function of the
  protein.

23
TARGET SELECTION

• Groups are focusing on complete organisms
• thermophilic, mesophilic or halophilic
• eukaryotic or prokaryotic
• classes of proteins from different organisms
• There isnt a coordinated international group
  that assigns targets (yet!).
• Some groups may solve the same structures
  (redundant).
• two SG pilot projects solved factor 5A first!!!
• Membrane proteins and proteins whose structures
  are already solved are eliminated.

24





















TARGET SELECTION



700
600
s
s
Transmembrane
e
e
n
n
500
e
e
Known 3D structure
g
g


f
f
400
Number of ORFs
o
o
Genes not selected


r
r
e
e
300
Genes targeted
b
b
m
m
u
u
200
N
N
100
0
lt16
16 - 31
31 - 51
51 - 71
71 - 100
gt100
Protein size (kDa)
25
DRUG DISCOVERYANTIBIOTICS

• Targets in this area of structural genomics are
  bacterial proteins that are essential for growth
  and survival.
• cell wall biosynthesis
• aromatic amino acid biosynthesis
• The development of a broad spectrum antibiotic
  would encompass the structures of a single
  protein from different bacterial organisms.

26
DRUG DISCOVERYHUMAN DISEASE

• Targets in this area of structural genomics are
  G-protein coupled receptors, ion channels and
  kinases etc.
• -GPCRs and ion channels are membrane proteins
• and are more difficult to purify and
  crystallize
• The development of techniques to allow
  over-expression, purification and crystallization
  of these targets is required and in progress.

27
AIMS OF PILOT PROJECT

• determine feasibility of a Structural Genomics
  Project
• develop technologies necessary for large-scale
  initiatives
• develop high-throughput (HTP) cloning
• develop high-throughput expression
• develop high-throughput purification

28
Methanobacterium thermoautotrophicum

• isolated in 1971
• thermophile (optimal growth T is 65C)
• methanogen (grows on methane as a carbon source)
• sequenced (Smith, DL et al., 1997, J. Bact., 179,
  7135)
• 1 751 377 bp and 1855 orfs
• 13 are similar to eucaryal sequences
• proteins in DNA metabolism, transcription and
  translation
• archaeal proteins are smaller and more stable
• than bacterial and eukaryal homologs

29
PROTEIN FUNCTION
Assigned Function Sequence Homology 45
Conserved Function Sequence Homology 28
Unknown Function No Sequence Homology 27
30
CLONING OF MT GENES

• PCR amplification of gene of interest
• purification of PCR product
• ligation into pET15b expression vector
• T7 promoter
• induced with IPTG
• cleavable hexahistidine fusion tag
• transformation into DH5? E. coli cells
• plasmid prep
• transformation into BL21(DE3) E. coli cells
• expression and purification

31
LIMITED PROTEOLYSIS

• single domain proteins and proteins less
• than 40 kDa can be expressed in E. coli
• multi-domain proteins and proteins greater than
• 40 kDa are quite difficult to express in E. coli
• these proteins may be expressed in yeast or
  baculo
• OR
• these proteins must be broken down into domains

32
PROTEINS DESTINED FOR NMR
Protein lt20 kDa
N15 Label
NMR
Protein-Protein Interactions
Aggregated, Unfolded Folded
Co-Expression
Structure
33
COMPARISON OF N15 NMR SPECTRA
Excellent
Poor
34
IDENTIFICATION OF A FOLDED DOMAIN
Before
After Proteolysis
35
PROTEINS FOR CRYSTALLOGRAPHY
36
STRUCTURE DETERMINATION STEPS

• Clone Gene
• Purify Protein
• Crystallize Protein
• Collect X-Ray Diffraction Data
• Identify Selenium Sites
• Calculate Phases using MAD
• Calculate Electron Density Map
• Build Model of Protein in Electron Density
• Refine and Rebuild Protein Model

37
PROTEIN CRYSTALLIZATION

• A crystal is an ordered three-dimensional array
  of molecules in the same orientation held
  together by non-covalent interactions.
• Crystals are grown by slow-controlled
  precipitation from crystallization conditions
  that do not denature the protein.
• These conditions can contain precipitants such as
  salts (NaCl, AmSO4), organic solvents
• (EtOH, MPD) or polymers (PEG), buffers,
• additives and ions.

38
PROTEIN CRYSTALLIZATION contd

• Each protein has its own empirically determined
  crystallization condition.
• pH
• ionic strength
• protein concentration
• temperature
• ions
• precipitant
• We cannot sample complete crystallization
  matrices.
• We start off with approximately 200 different
  crystallization solutions and hope for the best.

39
PROTEIN CRYSTALLIZATION contd
CRYSTAL TRIALS Crystallization solutions
used to screen for protein
crystallization conditions
1
2
3
40
PROTEIN CRYSTAL
100 microns
41
X-Ray DIFFRACTION
42
X-RAY DIFFRACTION IMAGE
43
PROGRESS TOWARDS HTP CLONING

• Initial Rate
• 24 clones per person per week
• Current rate
• 96 clones per person per week

44
PROGRESS TOWARDS HTP PROTEIN EXPRESSION

• Established conditions to maximize number of
  soluble clones
• bacterial strain
• induction conditions
• magic plasmid

45
PROGRESS TOWARDS HTP PURIFICATION

• Initial Rate
• 1 protein/person/week
• Current Rate
• 8 proteins/person/week
• Target Rate
• 16 proteins/person/week

46
ACHIEVEMENTS

• We have optimized HTP cloning.
• We have optimized HTP expression
• and purification.
• We are in the process of automating cloning
  and purification.

47
SUMMARY OF MT PROTEINS
Cloned
Expressed
Soluble
Purified
Microcrystals/Promising HSQC
Well diffracting crystals/excellent HSQC
48
KNOWN FUNCTION BUT UNKNOWN STRUCTURE
49
UNKNOWN BUT STRUCTURE SUGGESTS FUNCTION
MTH152 FMN-binding protein Ni2 binding
MTH150 Nicotinamide mononucleotide adenylyltransfe
rase
MTH1615 Nucleic acid binding
MTH538 Phosphorylation -independent 2-component
signaling protein
50
STILL UNKNOWN
MTH1184
MTH1175
51
CONCLUSIONS FROM FEASIBILITY STUDY
Crystallization is now rate limiting NMR can
play a significant role Solubility presents a
major hurdle Small, single domain proteins
behave better Low hanging fruit 20 of
proteome Must develop HTP methods for
recalcitrant proteins
52
STRATEGIES FOR TACKLING RECALCITRANT PROTEINS
1. Focus on domains 2. Empirical
bioinformatics 3. Identification of binding
partners (proteins and ligands)
53
TAKE HOME LESSON

• think about biology on a genomic scale

54

• PROTEINS Structure, Function and Genetics has
  inaugurated a new short format of Structure
  Notes designed to provide brief accounts of
  structures that contain too little new
  information to warrant a full length article
• what can you expect from robots!!!
• - Bill L Duax

55
THE TEAM
A. Edwards / C. Arrowsmith Steven Beasley Asaph
Engel Brian Li Anthony Semesi Emil Pai
Vivian Saridakis Ning Wu Aiping Dong
Akil Dharamsi
Adelinda Yee
Dinesh Christendat
56
1999 OCI SUMMER STUDENTS
Stephanie Fung
Joanne Loo
Hedyah Javidni
Gundula Min-Oo
Ashleigh Tuite
Fred Hsu
57
2000 OCI SUMMER STUDENTS

• Ashleigh Tuite
• Fred Cheung
• Laura Faye
• Toni Davidson

58
COLLABORATORS

• Lawrence McIntosh (UBC)
• Cameron Mackereth
• Mike Kennedy (PNNL)
• John Cort
• Mark Gerstein (Yale)
• Yuval Kluger
• Kalle Gehring (McGill)
• G. Kozlov