-6 -4 -2 2 4 6
-6
-4
-2
2
4
6
R e p lic a te 1
Re
pli
ca
te2
Myofibroblast to fibroblast transition assay (MFT) Figure 3. MFT assay format (A). Inhibition of ACTA2 mRNA measured by FISH in a 384w plate format (B) (ViewRNA, Affymetrix®). High content imaging with IN Cell 2200 (GE Healthcare). Quantification using in-house algorithm measuring ACTA2 expression normalized by expression of EEF1A1 (housekeeping gene).
Introduction • Systemic sclerosis (SSc) is a rare autoimmune and fibrotic disease
characterized by a severe skin thickening. It affects connective tissues of multiple organs causing structural and functional abnormalities (e.g. lung, heart, kidney). To date, no effective therapy has been identified.
• Myofibroblasts play a significant role in SSc by secreting profibrotic factors, leading to excessive synthesis and deposition of extracellular matrix. They are characterized by overexpression of alpha smooth muscle actin (αSMA).
Aim of the study • Our goal was to identify and validate novel candidate drug targets using RNAi
in phenotypic assays in SSc patient-derived fibroblasts. In collaboration with the Leiden University Medical Center, dermal fibroblasts and keratinocytes were isolated from 4 mm skin biopsies collected from 14 patients with diffuse cutaneous SSc. Patient were recruited according to ACR/EULAR 2013 criteria.
ww
w.c
river
.com
Target Discovery program in Systemic sclerosis Loubna Chadli, Britt Sotthewes, Marc Cheung, Stefan N. Andersen, Guilherme G. Dos Santos, Jamil Aarbiou, Jeroen DeGroot# and David F. Fischer. Charles River Laboratories Nederland B.V. ; Darwinweg 24, 2333 CR Leiden, the Netherlands. #Corresponding author: [email protected].
Cha
rles
Riv
er S
ympo
sium
- A
dvan
cing
Rar
e D
isea
se D
rug
Dis
cove
ry
06/2016
A
B
Figure 1. Reversion of myofibroblast phenotype by adenoviral transduction of shRNAs.
Primary screen: ~21,000 shRNAs
Rescreen: ~ 550 shRNAs
3D contractility & toxicity counter screen
On-target
RNA profiling
3D skin
Expression
Targets
MFT assay development & SSc patient recruitment in collaboration with LUMC
Target Discovery & Validation Approach
• RNA profiling comparing SSc and healthy biopsies allowed the selection of the most relevant SSc donors (data not shown).
• A high-throughput assay measuring the reversion of myofibroblasts phenotype was developed to screen Charles River’s Silence Select library (> 21,000 adenoviral shRNAs).
• Inhibition of mRNA expression of αSMA was quantified in a 384 assay format using fluorescence in situ hybridization (FISH).
• In addition, fibroblast contractility and impedance assays and a 3D skin equivalent model were developed to validate the potential targets identified during the screening phase.
Figure 2. Target discovery approach using human primary SSc dermal skin fibroblasts
ACTA2 shRNA Neg. ctrl
ACTA2 (Cy3)
Nuclei (DAPI) ACTA2 (Cy3)
EEF1A1 (Cy5)
Untransduced
D0 D2
Serum deprivation
Seed 10% FBS
D3
Transduction with adenoviral
shRNAS
D4
Refresh
D7
Fix plate & Staining
High content imaging
Primary screening of SilenceSelect® shRNA library Figure 4. Primary screening performed in one SSc donor, n=2 (~ 21,000 shRNAs, 95 plates). ACTA2 and aveGFP shRNAs used as positive and negative controls. Data represent robust Z score based on sample distribution. Example heat maps show plate distribution (A). Data distribution shown in dot plot (B), box plot (C) and correlation plot (D). Hit calling performed at < -1.5. Duplicate hits at -1.5, single hits at -2.3 and activator shRNAs (>2.0): ~550 shRNAs selected for rescreen. Statistical results shown in F.
B
D
aveGFP neg. control
ACTA2 pos. control
Cut-off -1.0
Cut-off -2.3
Cut-off -1.5
Samples
-6 .0
-4 .0
-2 .0
0 .0
2 .0
4 .0
6 .0
AC
TA
2/
EE
F1A
1
Rob
ust
Z sc
ore
AC
TA2/
EEF1
A1
Rob
ust
Z sc
ore
AC
TA2/
EEF1
A1
C
ρ: 0.6
One plate Stack of 80 plates A
-6 -4 -2
-6
-4
-2
R e p lic a te 1
Re
pli
ca
te2
-2.3 -1.5 -1.0
~ 550 selected hits
E
Screen statistics Z prime Assay
window Spearman rank Kappa value
Average 0.27 5.45 0.60 0.37
Hit performance
% false positive
% false negative
Robust Z score of duplo hits
% inhibition in duplo hits
Duplo hit rate
Average 0.09% 0.39% -2.07 71.58% 2.18%
F
Fibroblasts
Myofibroblast
Adenoviral transduction of
shRNAs
Myofibroblast
Hit: potential target gene
No hit
Migration ECM aSMA
a v e G F P _ v 1 7 AC T A2 _ v 3 S a m p le s
-6 .0
-4 .0
-2 .0
0 .0
2 .0
4 .0
6 .0
AC
TA
2/
EE
F1A
1
Conclusion SSc fibroblasts constitute a relevant model to identify novel targets involved in SSc pathogenesis. In this program, we successfully identified ~ 550 shRNAs to be used in rescreen. 2D and 3D assays are in development to validate the targets.
Acknowledgments We would like to thank Dr. J. De Vries-Bouwstra and Pr. T.W.J. Huizinga for their role in patient recruitement (Department of Rheumatology, Leiden University Medical Center). We also thank our collaborators at Pluriomics B.V. for access and guidance on the xCelligence system (Leiden, The Netherlands).
Validation assays: Gel contractility
Validation assays: Impedance assay
Figure 6. Effect of TGFβ1 on impedance of human healthy fibroblasts. Cell impedance (electrical resistance) measured in 48 well E-plates (A, electrode plates) using xCelligence® ECR system (ACEA Biosciences) (B). Assay in development.
Validation assays: 3D skin equivalent model
Figure 7. 3D full-thickness skin equivalent (SE) model. Fibroblasts seeded in rat tail collagen to reconstitute a dermal equivalent. Keratinocyte seeded on dermal equivalent to induce epidermal reconstruction (A). Pictures of SE model and histology with hematoxylin eosin staining (B). Assay in development.
Untransduced Neg. ctrl Pos. ctrl
SSc donor 01
Healthy donor 01
SSc donor 03
SSc donor 02
Healthy donor 02
0
2 .01 0 -2
4 .01 0 -2
6 .01 0 -2
8 .01 0 -2
1 .01 0 -1
1 .21 0 -1
Are
a(a
.u.)
H e a lth y d o n o rs S S c d o n o rs
3 .0 x
2 .2 x
1 .7 x
0 2 0 4 0 6 0 8 0 1 0 00
2
4
6
8
1 0
1 2
1 4
N o c o a t in g - 8 K
T im e (h o u rs )
Ce
llin
de
x
8 K - 0 T G F B - N o C o a t in g8 K - 2 T G F B - N o C o a t in g8 K - 1 0 T G F B - N o C o a t in g
Serum starvation
TGFβ1 addition
Seeding
x1.84 x1.87
Reconstructed epidermis
Dermal equivalent
A B
2 ng/mL TGFβ1 No TGFβ1
10 ng/mL TGFβ1
Impedance index
A
B
A
B Transwell with fibroblast populated
collagen gel
Fibroblast Epidermal sheet
Fibroblasts in collagen
Keratinocytes
Reconstructed epidermal sheet Histology
ACTA2 pos. ctrl. aveGFP neg. ctrl.
Untransduced
Figure 5. Effect of adenoviral shRNAs on healthy and SSc fibroblasts contractility. Collagen gel contraction measured 4 days after transduction (A). Gel area quantified using Image J software (B). Assay in development.