March 30, 2015

CHALLENGES AND VARIABILITY IN

DYSTROPHIN DETECTION:

IMMUNOFLUORESCENCE,

ANTIBODY STAINING

AND IMAGING METHODOLOGIES

CHANTAL BEEKMAN

AFRODITE LOURBAKOS

2

o Existence and variability of trace dystrophin & revertants

o Dystrophin intensity by a reproducible automated image analysis method

o Limitations: variability, linearity, control, biopsy quality

o Clinical study perspective on biopsy dystrophin analysis, next steps

Agenda

3

Counting positive fibers that look like revertants is complicated

Multiple types of revertant fibers present in a biopsy

Counting fibers that look like revertant fibers is difficult to interpret

Ab15277 (C-terminal) MANDYS106 (exon 43) DMD

Exon 45-50

detection of revertant fibers

Revertant fibers: bright muscle fibers with high levels of dystrophin

Revertants without

exon 43

Revertants with

exon 43

4

Variability in trace dystrophin and revertant fibers in DMD

• Trace dystrophin observed in all patients

(> 500 biopsies; 16 deletion mutations)

Essential

• representative images capturing many fibers

• measure dystrophin intensity in the entire membrane

DMD del 49-50

DMD del 49-50

Dystrophin traces: low levels of expression in many fibers throughout a biopsy

• Patches of low dystrophin-positive areas at the sarcolemma (Nicholson et al., 1990, 1993)

5

The low trace signal detected in DMD is dystrophin specific

a genetic control analysis

Beekman et al 2014

MANDYS106 (exon 43) Isotype control

DMD del 45

(epitope present)

DMD del 43

(epitope absent)

6

Dystrophin measurement Staining & imaging

0%

20%

40%

60%

80%

100%

0 500 1000 1500

fiber dystrophin intensity (a.u.)

Automated image analysis

Definiens customised software

% c

um

mu

lati

ve f

ibers

revertants

trace

Dystrophin % m

uscle

fib

ers

0%

5%

10%

15%

20%

25%

0 1000 2000

dystrophin mean

of population

Revertants

Trace dystrophin

Method captures the variation in dystrophin intensity between fibers in DMD

Beekman et al 2014 (PLoS One)

• confocal microscopy (25x magnification)

• identification of individual fibers

minimum 90% fibers

generally 250 - 600 fibers / section

• measure per fiber dystrophin intensity

entire membrane for every fiber

operator independent, objective

high reproducibility

Spectrin

7

Reproducibility in dystrophin measurement in different DMD biopsies

• 4 DMD biopsy samples

• 2-3 experiments (over a 1.5 year period) providing inter-assay precision

• 2-4 sections analysed providing intra-assay precision

• 8 operators randomly involved in staining, imaging and image processing

Beekman et al 2014

Sample

Dystrophin membrane

intensity MANDYS106 (a.u)

Ranking

Precision

Exp 1 Exp 2 Exp 3

Intra assay;

CV%

sections

Inter assay;

CV%

experiments

DMD 3 255 241 208 4 3-10% 10%

DMD 1 286 298 228 3 2-8% 14%

DMD 4 306 427 354 2 2-5% 17%

DMD 5 406 - 397 1 4-6% 2%

8

• Negative control differentiates from DMD and is required

• Reproducible differences over a wide range of dystrophin intensity

• Linearity between intensity signal and dystrophin concentration cannot be established

Immunofluorescence can detect differences but is not quantitative

• Different control muscles and donors vary by 30%

• Relative comparison of dystrophin differences between biopsies is informative

CTRL 2

DMD 1

BMD 4

DMD 2

BMD 1

BMD 3

BMD 2

48-50

Ex.19:c.23

80+3A>C

45-50

45-47

03-07

45-47

9

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 200 400 600 800

0207

0507

002001

002002

DMD 5

DMD 4

DMD 3

DMD 1

mean membrane dystrophin intensity (a.u.) c

um

ula

tive

% f

ibers

Assay detects differences accurately in DMD

Dystrophin varies between DMD patients

thus important to compare with pre-treatment biopsy per patient

Beekman et al 2014

DMD 4 del 45

DMD 3 del 50 DMD 1 del 48-50

DMD 5 del 45

10

Image Method comparison: Biological Outcome Measures 2013

Arechavala-Gomeza et al.

10 circles across membranes; measure min & max; divide by spectrin

Taylor et al.

mean membrane intensity per image; threshold approach based on spectrin;

divide by spectrin

Beekman et al.

mean membrane intensity per fiber, not divided by spectrin

Arechavala-Gomeza Taylor

1

2 3

4 5

6 7

8

9 10

11

12

21

13 14 15

16

17

18 19

20

22

23 24

25 26

27

28 29

30

31 32 33 34 35 36

37

38

39 40

41

42

43

44

45

46

47 48

49 50

51

52

Beekman

Anthony et al. 2014 (Neurology)

11

Image Analysis Method comparison by Prosensa

• Analysis Methods are in agreement on the relative comparison between biopsies

• Negative control (PBS) can be 50% of DMD trace values

• Positive control dystrophin intensity vary for different muscles

• Linearity cannot be established due to lack of dystrophin standards

Relative comparison between biopsies is most informative

CTRL 1 CTRL 2 CTRL 3 BMD A DMD B DMD C BMD D DMD E BMD F PBS

0

50

100

Arechavala-Gomeza et alTaylor et al.Beekman et al (Definiens)

Rela

tive d

ystr

op

hin

in

ten

sit

y (

%)

12

Biopsies and DMD biomarkers in clinical programmes

Biopsies are valuable for assessment of disease progression and drug effect

• Dystrophin protein detection

• Exon skip

• Drug concentration in muscle

• Explore other biomarkers of disease

~500 biopsies from multicenter studies (80 sites) covering 16 different DMD deletions

• analysis of biopsies under Good Clinical Laboratory Practices for reliability and

traceability

Potential for surrogate endpoints, but issues to consider:

• Biopsy sampling: which muscle, biopsy size, disease progression

• Quality of handling, shipping and storage (procedures and training)

13

DMD Healthy control muscle

Connective

tissue

Disease progression and heterogeneous sampling can affect analysis

Connective tissue

Muscle fibers Adipose cells

DMD -Extensive disease

progression:

• Sufficient number of muscle fibers required

• Small biopsies and low muscle content are prone

to experimental artifacts

14

H&E staining

Spectrin staining

good intermediate poor

Freeze artefacts influence staining and ability to compare biopsies

No reliable comparison can be made between biopsies that differ in quality

% of subjects with evaluable biopsies

phase II DMD114117 (13 sites) 88%

phase II DMD114876 (13 sites) 66%

phase III DMD114044 (46 sites) 50%

15

• Dystrophin immunofluorescence analysis using intensity measurements

• dystrophin signal localisation at entire membrane

• variable presence of trace dystrophin and revertant fibers

• requires good quality biopsies

• reproducibly and accurately detects differences between DMD biopsies

- intra-assay precision typically <10%; inter-assay precision <25%

• Advancing immunofluorescence methods for quantification requires

development and purified dystrophin protein control with a signal dilution

relevant to DMD patients

• Further effort and development to correlate dystrophin to clinical outcome in

randomized placebo controlled clinical trials is required

CONCLUSIONS

16

• Biomarin/Prosensa: Stavros Giannakopoulos, Janwillem Testerink, Tjadine Holling, Niki Heuvelmans, Ronald

Jan Baptiste, Dyonne Kreuger, Dirk Duinsbergen, Umut Akinci Corbacioglu, Roelof Kuper, Jessica Sipkens,

Petra de Bruijn, Simone van Wijngaarden, Karin Boekhoorn, Carla Persoon-Deen, Nico Yau, Karolina

Kozaczynska, Anneke Janson, Alison Foster, Claire Wardell, Judith van Deutekom, Afrodite Lourbakos, Sjef de

Kimpe, Giles Campion

• LUMC: Jan Verschuuren for BMD samples, VUMC: H.W. Niessen and Asterand for control samples

• GSK: Steve Hood, Rob Geeske, Katja Remlinger, John Kraus

• BOM consortium: Anthony V, Arechavala-Gomeza LE, Taylor BS, Vulin A, Kaminoh BS, Torelli S, Feng L,

Janghra L, Bonne G, Beuvi, M, Barresi R, Henderson, M, Laval S, Lourbakos A, Campion G, Straub V , Voit T,

Sewry C , Ellis J, Morgan JE , Flanigan KM, Muntoni F.

• Clinical Investigators : Alexandra Araujo , Elena Belousova , Enrico Bertini , Peter Born, Craig Campbell ,

Claude Cances, Brigitte Chabrol, Jong-Hee Chae, Jaume Colomer Oferil, Giacomo Pietro Comi , Jean-Marie

Cuisset , Guy D'Anjou Isabelle Desguerre , Alberto Dubrovsky, Ricardo Erazo Torricelli, Raúl Escobar, David

Feder , Alessandra Ferlini , Roberto Giugliani, Ágnes Herczegfalvi , Yuh-Jyh Jong , Shigemi Kimura , Jan-Bernd

Kirschner, Karin Kleinsteuber , Hirofumi Komaki , Anna Kostera-Pruszczyk , Martin Kudr , Eugenio Maria

Mercuri , Valentina Ricotti, Wolfgang Mueller-Felber, Erik Niks, Katsuhisa Ogata, Ignacio Pascual , Yann Pereon,

Salmo Raskin, Jan Verschuuren, Michela Guglieri, Viera Darin, Magnhild Rasmussen, Laurent Servais,

Umbertina Reed, Ulrike Schara, Kathryn Selby, Claudia Sobreira, Yasuhiro Takeshima, Haluk Topaloglu, Juan

Jesus Vilchez Padilla, Ayse KaradumanGiuseppe Vita, G. Buyse, Kevin Flanigan, K., Petr Vondracek, Gert

Wiegand, Ekkehard Wilichowski, Nicholas Deconinck, Brenda Wong, Francesco Muntoni,Volker Staub, Haluk

Topaloglu, Greg MacDonald, Thomas Voit, Mar Tulinius, Nathalie Goemans

• DMD patients and families

Acknowledgements