Transcript Glucocorticoids and feed forward regulation: a new paradigm for

Glucocorticoid receptor crosstalk with NF-kB in
airway cells – analyzing the cistromes
BIOS6660 Genomic Data Analysis with R and
Bioconductor
Anthony Gerber MD, Ph.D.
October 20, 1015
The Nuclear Receptor family
• Transcription factors
• Many are ligand activated
• Only class of transcription factors that can be
targeted by small molecules in the clinic
• Clinical targets include estrogen, androgen,
mineralocorticoid, Vitamins D, glucocorticoid
and thyroid receptors, RXR, PPAR
• Major interest in developing selective
ligands/modulators to enable improved
therapeutic windows
Glucocorticoids in the clinic: a large footprint
 10-20 million annual prescriptions
for oral glucocorticoids in USA
 > 50 million prescriptions for
localized delivery (inhaled, topical, eye
drops)
Major targets are diverse immunemediated diseases
•Rheumatoid arthritis
•Inflammatory bowel disease
•COPD
•Asthma
•Other lung diseases
o Hypersensitivity
pneumonitis
o BOOP
o NSIP
o vasculitis
•Organ transplants
•RDS of prematurity
“Off target” effects
CNS: Anxiety, insomnia
Ocular: Glaucoma
Cardiovascular: Hypertension
Endocrine: Diabetes, obesity
Skin: Fragility
Muscle: Atrophy
Bone: Osteoporosis
Balancing disease symptoms with glucocorticoid side effects in the clinic:
A 50 year old female with severe, persistent asthma
No oral glucocorticoid use
• lung function < 50% of normal
• short of breath after walking 2 blocks
• unable to go up a flight of stairs
• frequent coughing episodes
• 1-2 ER visits per quarter
• 2-3 hospitalizations per year
Taking oral glucocorticoids
• lung function ~80% of normal
• no shortness of breath after 10 blocks
• able to go up 2 flights of stairs
• no hospitalizations or ED visits
• 20 pound weight gain
• lower extremity edema
• irritability
• high blood sugars
• increased risk of osteoporosis
There is a major unmet need for improved glucocorticoid-based therapies
Background: Glucocorticoids bind to the glucocorticoid
receptor (GR), causing it to regulate gene expression
Glucocorticoids
GR is a basic model of metazoan
transcriptional regulation
-> Recent example: DNA implicated as
regulating GR activity through
allosteric mechanisms (Hudson et al, Nat
Struct Mo Bio, 2013, Meijsing et al, Science 2009)
Therapeutic effects of GR
activation are also intensely
studied
-> >10000 Pubmed citations for
“asthma and glucocorticoid”
Image from http://brainimmune.com/the-glucocorticoid-receptor/
“Transprepression” typically
implicated in mediating
therapeutic effects
Structural considerations
Steve Bilodeau et al. Genes Dev. 2006;20:2871-2886
How do glucocorticoids work?
Pro-inflammatory
6
TNF
Dex
Dex+TNF
4
2
Change in mRNA level (log2)
0
TNFα
Dex
-2
-4
‒
‒
+
‒
‒
+
+
+
TNFAIP3
-6
HBEGF
8
7
6
5
4
3
2
1
0
β-actin
Anti-inflammatory
Glucocorticoids “spare” the
expression of negative
feedback targets of TNF (a
major inflammatory signal)
How do glucocorticoids actually work?
TNFAIP3
5’
Intron 2
(821 bP)
(+5,670 — +6,491)
1
2
Relative luciferase activity
TNFAIP3 reporter
Hela cell GR/NF-kB ChIP-seq
Rao et al, Genome Biology, 2011
2500
pTNFAIP3I2
3’
1
NFᴋB-BS1(CTTGGAAAGTCCAGG)
2
NFᴋB-BS2(CTGGGGAATTCCAGA)
GR-BS(CCAGAACAAAAAGTACAAT)
pIL8
250
2000
200
1500
150
1000
100
500
50
0
0
siCtrl
5
B
siTNFAIP3
4
siCtrl
3
TNFα
Dex
2
1
—
—
+
—
siTNFAIP3
+
+
—
—
+
—
+
+
70 kDa
TNFAIP3
42 kDa
Β-actin
0
-1
Dex+TNFα
Dex
-2
TNFα
Change in IL1a mRNA level (log2)
How do glucocorticoids actually, actually work?
TNFAIP3 contributes
to glucocorticoidmediated cytokine
repression in airway
epithelial cells
How do glucocorticoids work and what prevents
them from working in asthma?
Since GR interactions with DNA define GR
activity
 study GR interactions with DNA in airway
cells
Since GR interactions with inflammatory factors
are important for GC efficacy
 Study DNA-based interactions between GR
and NF-kB
No current data on GR cistrome in airway cells…
ChIP - overview
Cross-link
Chromatin shear and prep
IP
Purify DNA
ChIP- downstream assays
ChIP-Seq example summary data
GR PEAKS
Airway epithelial ChIP-Seq experimental design
Cells:
Beas-2B
Treatments:
dexamethasone
(dex; 100 nM)
tumor necrosis factor-α
(TNF; 20 ng/ml)
Sequencing:
Illumina Hi-Seq;
performed in biological
duplicate
Conditions:
Treatment (1 hr) IP Antibody
control
dex
TNF+dex
control
TNF
TNF+dex
control
dex
TNF
TNF+dex
GR
NFkB-p65
RNAP2
Pattern 1: GR+NFkB co-occupancy & reduced RNAP2
IL8
locus
50
50
75
75
125
125
125
dex
GR IP
TNF+dex
GR IP
TNF
p65 IP
TNF+dex
p65 IP
dex
RNAP2 IP
TNF
RNAP2 IP
TNF+dex
RNAP2 IP
Pattern 1: GR+NFkB co-occupancy & reduced RNAP2
CCL2
locus
45
45
55
55
80
80
80
dex
GR IP
TNF+dex
GR IP
TNF
p65 IP
TNF+dex
p65 IP
dex
RNAP2 IP
TNF
RNAP2 IP
TNF+dex
RNAP2 IP
Pattern 1 summary
• GR binds pro-inflammatory gene in absence of TNF
• GR occupancy is maintained/enhanced in presence
of NFkB
• GR+NFkB co-occupancy reduces RNAP2 recruitment
NET EFFECT = repression of pro-inflammatory
transcription
Pattern 2: NFkB-mediated GR occupancy & reduced RNAP2
ICAM1
locus
20
20
50
50
75
75
75
dex
GR IP
TNF+dex
GR IP
TNF
p65 IP
TNF+dex
p65 IP
dex
RNAP2 IP
TNF
RNAP2 IP
TNF+dex
RNAP2 IP
Pattern 2 summary
• GR binds pro-inflammatory gene only in presence of
TNF
• Role of NFkB in GR recruitment unclear, possibly
indirect
• GR+NFkB co-occupancy reduces RNAP2 recruitment
NET EFFECT = repression of pro-inflammatory
transcription
Pattern 3: GR+NFkB co-occupancy & enhanced RNAP2
SERPINA3
locus
165
165
40
40
10
10
10
dex
GR IP
TNF+dex
GR IP
TNF
p65 IP
TNF+dex
p65 IP
dex
RNAP2 IP
TNF
RNAP2 IP
TNF+dex
RNAP2 IP
Pattern 3: GR+NFkB co-occupancy & maintained RNAP2
TNFAIP
3
locus
60
60
35
35
35
35
35
dex
GR IP
TNF+dex
GR IP
TNF
p65 IP
TNF+dex
p65 IP
dex
RNAP2 IP
TNF
RNAP2 IP
TNF+dex
RNAP2 IP
Pattern 3 Summary
• GR binds anti-inflammatory genes with
dex+TNF
• GR+NFkB co-occupancy does not appear
antagonistic
• GR+NFkB co-occupancy enhances or maintains
RNAP2 recruitment
NET EFFECT = activation (or sparing) of antiinflammatory transcription
What do we want from our ChIP data?
BASIC
1. Identification of peaks for GR and p65 under
each condition and r values
2. Identification of differential GR binding
vehicle vs. dex
3. Identification of differential binding p65
binding vehicle vs. TNF treatment
What do we want from our ChIP data?
BASIC
1. Identification of peaks for GR and p65 under
each condition and r values
2. Identification of differential GR binding
vehicle vs. dex
3. Identification of differential binding p65
binding vehicle vs. TNF treatment
What do we want from our ChIP data?
ADVANCED
1. Differential GR binding (dex vs dex + TNF)
2. Differential p65 binding (TNF vs TNF + dex)
3. RNAP2 patterns
1.
2.
3.
4.
5.
6.
Increased at TSS with dex vs vehicle
Closest GR peak
Increased at TSS with TNF and TNF + dex > TNF
Increased at TSS with TNF and TNF + dex < TNF
Closest p65 peak to TSS for patterns 3 and 4
Closest GR peak to TSS for patterns 3 and 4
What do we want from our ChIP data?
Collaboration Level!
1. Compare GR binding between Gerber lab
data set and paper from Rao et al (Genome
Biology, 2011)
2. Compare p65 binding between Gerber lab
data set and paper from Rao et al
(Identification of differential GR binding
vehicle vs. dex
3. Compare regulatory outcomes – i.e.
correlate with RNAP2 occupancy
QUESTIONS?