Transcript Lipid Pathways to Atherosclerosis - MGH

Genomic Analysis of Stress and
Inflammation
Massachusetts General Hospital
Departments of Medicine and
Genetics
Harvard Medical School
Boston University
PGA components and projects
Projects
 Genetic dissection of
signal transduction
 Host-pathogen
interactions:
Pseudomonas and CF
 Definition of Protein
networks
 Macrophage activation
by metabolic and
pathogen stresses
Component Centers





Microarray and sequencing
Brian Seed, PhD
Education and Training
Fred Ausubel, PhD
Proteomics
Jack Szostak, PhD
Human Tissue and Animal
Models
Mason Freeman, MD
Bioinformatics
Temple Smith, PhD
George Church, PhD
Components : Microarrays

Microarray generation
– Human and mouse cDNA arrays
– Bacterial arrays
– Specialty arrays (e.g., inflammatory gene
subsets)

Sequencing and cDNA library generation
– Array verification and generation
– Identification of genes isolated by RNA
display
– cDNA library production
Education and Training

Genomics training course - hands-on lab
experience
 Web based genomics training
 Visiting scientist project advice
 Seminar series in genomics
 Bioinformatics training, undergrad work
study
Proteomics

Development of RNA display technology
for identification of protein-protein
interactions
– PDZ domains
– Kinase substrate identification
Human Tissue / Animal Models

Acquisition and process of human tissues for gene
expression profiling and immunohistochemistry
– Atherosclerotic lesions (carotid, coronary, periph.)
– Hearts (idiopathic cardiomyopathy, CAD)
– Lungs (cystic fibrosis, emphysematous)
 Mouse and Cell Models
– CD14 and toll receptor null (endotoxin/bacterial
signaling)
– CD36 null (lipid uptake)
– SR-A null (lipid uptake and bacterial interactions)
– Cystic Fibrosis (with Gerry Pier and Fred Ausubel)
Animal Models

Conditional KO mice
– Based on homologous recombination in bacteria
– Conditional alleles generated by site-specific
recombinase action
– Flexible, medium throughput technology

In vivo imaging of transgenic reporter mice
– Fluorescence imaging of cells in living animals
– Ear, dorsal skin chamber and cranial windows
– Provides information about the activation of genes
in the organismic environment
Generation of Mutant Collections

Pseudomonas Transposon Insertion Collection
– High-quality non-redundant collection of multiple
insertions in all non-essential genes

Somatic cell mutant cell lines
– Reporter cell lines that facilitate the rapid
identification of mammalian somatic cell mutations
affecting signal transduction
– Mutant progeny derived from those lines
Bioinformatics

Database design and management
 Software development (data entry and
tracking software)
 Web access of data for internal and external
users
 Data analysis software
 Bioinformatics education
Genetic dissection of signal transduction

To create reporter cell lines that allow identification of
genes promoting activation of stress and inflammation
pathways.

To develop and exploit automated sib selection strategies
to identify new molecules that activate stress and
inflammation pathways.

To use microarray analysis to understand the genotypes of
mutant cell lines bearing lesions in stress and
inflammation signal transduction pathways.
.
Stages in the enrichment of an NF-kB inducer
.
.
.
In Vivo Imaging

Direct visualization of reporter gene output in
transgenic animals
 Can be performed using the same reporters used in
high-throughput discovery screens

Allows responding cell populations to be
identified in vivo
 Responding cells can be culled and phenotyped
.
.
Host/Pathogen interactions
The mucoid derivatives of PA14 from Aim 2 will be used to infect CF
mice. The P. aeruginosa microarray (Aim 1) and a murine microarray
consisting of the currently available UniGene clusters will be used to
determine gene expression profiles for the pathogen and the host,
respectively, before and during the infection process.
 Mucoid derivatives of P. aeruginosa PA14 mutants attenuated for
pathogenicity in model non-vertebrate hosts will be introduced into
CF mice to identify virulence-related factors required for infection of
the CF lung.
 P. aeruginosa PA14 mutants identified in Aim 4 that display defects in
CF lung pathogenesis will be used to infect CF mice and expression of
both P. aeruginosa and mouse genes will be analyzed using the P.
aeruginosa and mouse DNA microarrays.

Multi-Host Pathogenesis System
humans
.
112399-4
mice
P. aeruginosa
Strain PA14
Bacterial Pathogen
plants
nematodes
insects
Pseudomonas aeruginosa Pathogenicity-Related
GenesIdentified by Genetic Screening
051300-1
Host
Assay
# Screened
# Genes
Lettuce
Leaf Infection
C. elegans
Killing
Wax Moth
Killing
All Screens
2500
5500
5000
8000
9
13
11
32
Number of genes identified, which when mutated result
in decreased virulence on the hosts indicated on the left.
9
10
not tested
>18
5
13
5
22
9
11
11
30
8
10
not tested

P. aeruginosa Kills C. elegans
and Colonizes the C. elegans Intestine
P. aeruginosa
% nematodes killed
100
P. aeruginosa
E. coli
80
60
.
40
20
0
0
20
40
60
Hours of Feeding on
P. aeruginosa
80
E. coli
P. aeruginosa Pathogenicity-Related Genes
Identified by Screening in Model Hosts

32 Pathogenicity genes identified out of 8,000 random
transposon insertions screened.

Need to screen 30,000 random insertions to reach
saturation.

Obtain full set of virulence-related genes by screening in
several model invertebrate hosts.

Because screening for mutant phenotypes is rate limiting,
construct non-redundant insertion library (4800
nonessential genes) to screen the rest of the genome.
Advantages of a Non-Redundant Library

Simplifies screening in multiple hosts.

Multiple insertion alleles of certain genes will be useful
for confirming the phenotypes of insertion mutations.

> 80% savings in time for each screen in a model host.
The library can be expanded until it is saturated.


Obtain information about genes that are NOT required
for pathogenesis as well as genes that are.

The PCR products used to construct the library can be
used to synthesize a micro-array of (non-essential) ORFs
for PA14 expression studies.
new sequences
Raw Data Archive
- sequences
- trace files
trash
sequences
present?
processed
new sequences
BLAST/FASTA
PA14 genome
IST sequences
y
present?
n
Processing: - transposon clipping
- N-stripping
archive raw data
y
n
y
contaminant?
n
novel PA14 specific
PAO1
PA14
-specific Sequence
y
present?
n
y
patent
sequences
public
sequences
present?
n
mutant priority definition
1 annotated ORF
2 not annot. ORF, putative ORF
3 annotated ORF, put. Promoter
4 not annot. ORF, not put. ORF, put. Promoter
5 not annot. ORF, not put. ORF, not put. Promoter
Building the non-redundant PA14 mutant library
PA14 genome
IST sequences.
PA14 non-genome
project sequences
select seq. for inclusion
in NR1PA14
PA14 redundant
intermediate library
tested in
any host?
y
n
Sibling
present in
NR1PA14?
n
y
library of NR1PA14
redundancies
combine these criteria for mutant priority definition:
•annotated ORF
•putative ORF
•put. Promoter for annotated ORF
•not annot. ORF, put. Promoter
•not put. ORF, not put. Promoter
•homology to pathogenicity factors
•homology to regulatory proteins
•homology to house keeping genes
•assignment to a pathway
containing pathogenicity factors
•coordinate of the mutation along
the 5’-3’ axis compared to other
mutants targeted in the same gene
•PA14-specific?
PA14 non-redundant
library (NR1PA14)
Definition of protein networks by RNA display

To create a cellular protein-RNA fusion library from
pooled mRNA from normal human and mouse
tissues
 To use isolated domains from proteins transducing
stress and inflammation pathway signals to identify
interaction partners of those proteins
 To automate the detection of interactions between
signal transduction proteins and their target proteins.
 To make slick slides for presentations
RNA-Protein Fusions
mRNA Display
P
P
P
P
Decoding Protein-Protein Interactions
Challenge:to identify all binding
partners of a given “bait” protein
Solution: pass a library of cellular
mRNA fusions over the protein &
identify which fusions bind to the
bait protein
Deconvoluting mRNA-Protein Fusion
Targets with Microarrays
Cellular Library Features

No cloning
 Randomly primed cDNA
 Direct assembly of library in vitro
 Libraries are large enough to contain all
possible start and end points for every
protein fragment
Cellular Libraries: Selections
Cellular protein
domain library
Target
(bait)
Select:
1-4 rounds
PCR
Identification of Kinase Substrates
Library of fusions prepared from cellular RNA
Phosphorylate fusions with kinase in vitro
Immunoprecipitate phosphorylated fusions
with anti-phosphotyrosine Ab
PCR RNA from phosphorylated fusions
Other Selections in Progress
- PDZ domains binding known targets
- coiled-coil partners
- calmodulin binding proteins
Macrophage Activation
To perform comparative gene expression studies assessing the impact
of key proteins in inflammatory and stress response pathways, using
macrophages taken from wild type and knock-out mice.
 To explore concordances between murine and human macrophage
expression, and to establish baseline profiles revealing the
consequences of various sample collection practices.
 To analyze gene expression in aortas taken from normal and apo E
null mice and from coronary arteries of mice following allogeneic
heart transplantation
 To conduct parallel investigations on the gene expression profiles of
human carotid endarterectomy, coronary endarterectomy, and heart
transplant specimens, and to establish, if possible, the characteristic
gene clustering features of these conditions.

PS
Fig. 2 LPS signaling pathways
signa transduction pathway
Fig. 4
Differentiation of Embryonic Stem Cells into Macrophages
Undifferentiated, Pluripotent ES Cells
ES cells are cultured on embryonic fibroblast feeder
layers in media containing LIF-D to inhibit differentiation.
ES cell colony
Embryonic fibroblast layer
Differentiation into Embryoid Bodies
Embryoid Body
ES cells are cultured in LIF-D free medium for 1-3 h
and then trypsinized to form a single cell suspension.
ES cells are cultured in semi-solidmet hyl cellulose
media containing cyt okines for 10 days t o promote
the format ion of embryoid bodies rich in myeloid cells.
Harvesting of Embryoid Bodies
and isolation of macrophages
Embryoid bodies are gently washed from the methyl
cellulose media and cultured in pet ri dishes in media
ES differentiated macrophage containing myeloid growth factors. Macrophages produced
from the embryoid body are isolat ed by their ability to
adhere to bact eriologic plastic.
7
w ild ty pe
CD14 null
6
TNF (ng/ml)
5
4
3
2
1
0
0.01
0.1
1
2.5
5
Bioparti cl es /m acrophage
10
+ OxLDL
B.
125
I-OxLDL degradation
A.
+ LDL
800
600
400
200
0
wt
cl 3
cl 4
cl 5 P388D1
PPARg null
RNA display
purif ied
PSD-95
PSD-95 overlay
PSD-95
binding
82
kDa
ABC1
48
NR
2
wt- B
AB
trun C1
-AB
C1
Ra
b6
Ra
b11
Rab6
total
protein