Establishment of a Cell-Based Assay Specific for a Growth Factor that Shares a Common Receptor Chain and Overlapping

Biological Activities with Other Cytokines and Growth Factors

Michael G Tovey,INSERM Director of Research, Laboratory of Biotechnology &

Applied Pharmacology,ENS CACHAN

tovey@vjf.cnrs.fr

Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF)

GM-CSF is a hematopoietic growth factor that plays a central role in in the generation of neutrophils, macrophages, and DCsGM-CSF acts together with IL-3 and IL-5 to regulate the survival proliferation, differentiation, and functional activation of hematopoietic cellsGM-CSF also regulates its own activity by via the induction of CIS, a SOCS family member SH2-domain protein that inhibits Jak2/STAT5 phosphorylation & signaling.CIS and SOCS3 also regulate EPO induced Jak2/STAT5 signaling

Quantification of GM-CSF Activity

Current methods for quantifying human GM-CSF activity, are bioassays based on the ability of GM-CSF to support the proliferation of cell lines such as TF-1, or UT-7, that require GM-CSF for their growthDue to overlapping biological activities, IL-3 or EPO, can also support the proliferation of these cells and act synergistically together with GM-CSF. M-CSF & IL-1 can also enhance GM-CSF dependent cell proliferation TGF-β and IFNα, or IFN β, can antagonize this activityIL-3 & IL-5 share a common receptor βc chain with GM-CSFThus, proliferation based assays for GM-CSF are subject to non specific interference.

Development of a Cell-based Assay Specific for GM-CSF: Strategy

Use a cell line that possesses functional GM-CSF receptors, but does not require GM-CSF or other related growth factors for proliferationU937 cells possess functional GM-CSF receptors but not express functional EPO receptorsU937 cells do not require GM-CSF, EPO or other related growth factors for proliferationTransfect U937 cells with a GM-CSF responsive reporter-gene construct

CM-CSF-SpecificRegulatory Element

LuciferaseSV40

Min. prom.

Reporter-Gene Assay: Construction

SV40Poly A

Intron Human β-globine

However,

GM-CSF, IL-3, & IL-5 share a common receptor βc chain

GM-CSF Receptor

GM-CSF binds to a heterodimeric receptor comprised of a GM-CSF specific α subunit and a common receptor βc chain that is shared with IL-3 & IL-5GM-CSF receptor does not possess intrinsic tyrosine kinase activity; associates with Jak2 required for βc trans-phosphorylation, initiation of signaling, & biological activityReceptors for GM-CSF, IL-3, & IL-5 are expressed at very low levels (100-1,000 receptors/cell)GM-CSF, IL-3, & IL-5 each bind with low affinity to their specific Rα chain (Kd = 0.2 – 100 nM)In the presence of the βc receptor chain each cytokine binds with high affinity (Kd = 100 pM) resulting in dimerization of both sub-units and receptor activation

Development of a Cell-based Assay Specific for GM-CSF: Strategy

Overexpress the GMRα, GM-CSF ligand-specific binding sub-unit, of the human GM-CSF receptorHypothesis, GMRα receptor chain will compete with IL-3 or IL-5 specific binding sub-units for the pool of the βc signaling receptor sub-unit common to the GM-CSF, IL-3, and IL-5 heterodimeric receptors.

GM-CSF Receptor Structure and Signal Transduction

STAT5 STAT5STAT5

P

STAT5P

IL3Rα

IL3

IL5Rα

IL5

GMRα

IL5Rα

IL5

IL3Rα

IL3

TATA

GM-CSF Gene Reporter Assay (FireFly Luciferase)

SV40Poly A

Intron

+1

Stat5 Stat5 GM-CSF

(T/A)TTCnnnGAA(T/A) FireFly Luciferase

STAT5 Consensus Sequence: (T/A)TTCnnnGAA(T/A)

GATTTCTAGGAATTCAAATCGGATCTAGATTTCTAGGAATTCAAATCGGATCTAGATTTCTAGGAATTCAAATCGGATCTAGATTTCTAGGAATTCAAATCG

STAT5 con. Dx4

GATTTCTAGGAATTCttctcagaaGAATTCttctgagaaGAATTCttctcagaaGAATTCttctgagaaGAATTCttctcagaaGGAATTCAAATCG

TTCCCCGAAATGATGAGCTAGGAGCCTGATTTCCCCGAAATGATGAGCTAGGAGCCTGATTTCCCCGAAATGATGCTAGGAGCCTGATTTCCCCGAAATGATctGtTAG

GAGGCTCTGATTTCCGGGAAACTGATTCCCGGAATACGTTTTCCGGGAATACGTTTCCGGGAAACGTATTCCGGGAAAACTGATTCCGGGAAATGATCTGTTAG

STAT5 con. Cx6

STAT5 con. Ax4

STAT5 con. Bx6

0

2

4

6

8

10

12Fo

ld in

du

ctio

n lu

cife

rase

Ctl

500ng/ml rhGM-CSF

STAT 5 Reporter Gene Constructs – Firefly luciferase A, B, C or DTranscient transfection in U937 cells

STAT5-luc STAT5-luc

A B C D

STAT5 Reporter Gene – Firefly luciferase (C or D) relative to Renilla luciferase expression

Transcient transfection in K562 and U937 cells

0

200

400

600

800

1000

1200

1400

1600

1800

Rat

io (

Fire

fly

/ R

enill

a lu

cife

rase

)

Ctl

500ng/ml rhGM-CSF

K562 U937STAT5 is expressed

constitutively in K562 cells

STAT5 - luc C D C D

Effect of rhEPO and rhIFNγ on STAT5 – luciferase (C or D) Reporter GeneTranscient transfection in U937 cells

STAT5 C STAT5 D0

2

4

6

8

10

12

Fold

ind

uct

ion

Ctl

500ng/ml rhGM-CSF

500ng/ml rhEPO

10000 IU/ml IFNg

0,0

2,0

4,0

6,0

8,0

10,0

12,0

14,0

16,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Fire

Fly Fo

ld In

duc

on

Clone N°

Analysis of Clones Transfected with the US5‐Luc Construct

Selection of stable clones transfected by STAT5-luc C in U937 cell line

1

3

5

7

9

11

13

0.01 0.1 1 10

Fold

ind

uct

ion

Log rh GM-CSF (ng/ml)

Clone 2

Clone 3

Clone 6

0

5

10

15

20

25

30

35

40

45

50

0.0005 0.005 0.05 0.5 5 50 500 5000

Fire

fly

Luci

fera

se F

old

In

du

ctio

n

ng/ml

rh GM-CSF 4h

rh GM-CSF 18h

rh EPO 4h

rh EPO 18h

Firefly Luciferase induction in U937/STAT5-luc stable cell line

0.E+00

5.E+05

1.E+06

2.E+06

2.E+06

3.E+06

3.E+06

4.E+06

4.E+06

0.0E+00 5.0E+04 1.0E+05 1.5E+05 2.0E+05 2.5E+05 3.0E+05 3.5E+05 4.0E+05

Fire

fly

luci

fera

se R

LU

cell number

Cell number optimisation for the GM-CSF assayStimulation with 200ng/ml rhGM-CSF (Gibco®)

0

2

4

6

8

10

12

14

0.000001 0.00001 0.0001 0.001 0.01 0.1 1 10 100 1000

Fold

In

du

ctio

n

GM-CSF ng/ml

GM-CSF Standard Curves

StandardNIBSC

TATA

GM-CSF Gene Reporter Assay (FireFly / Renilla)

SV40Poly A

Intron

+1

Stat5 Stat5 GM-CSF

(T/A)TTCnnnGAA(T/A) FireFly Luciferase

CMVSV40Poly A

Intron

+1 Renilla LuciferaseIRES

GM-CSF-Rα+1

Double transcient transfection STAT5-luc (C or D) and pCMV-GM-CSF-Rα in U937 cells

Control GM-CSF-Rα Control GM-CSF-Rα

STAT5 DSTAT5 C

0

2

4

6

8

10

12

14

16

18

20Fo

ld in

du

ctio

n

Ctl

500ng/ml rhGM-CSF

0

2

4

6

8

10

12

14

16

18

20

22

Ctl 100 500 1000 100 500 1000 100 500 1000

Fire

fly

Luci

fera

se f

old

ind

uct

ion

STAT5-luc STAT5-luc + GM-CSF-Ralpha

rh GM-CSF Gibco® (ng/ml)

rh IL-3Immunotools® (ng/ml)

rh IL-5Immunotools® (ng/ml)

Firefly Luciferase induction in the U937 cell line transciently tranfected with STAT5-luc and GM-CSF-Rα treated with GM-CSF, IL-3 or IL-5

0

500

1000

1500

2000

2500

3000

3500

4000

4500

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55

RLU

Ren

illa

Clone N°

Analysis of US5-Luc Clones Expressing Renilla Luciferase and the GM-CSFRα Receptor Chain

0

2

4

6

8

10

12

14

16

18

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55

Fire

Fly

Fold

In

du

ctio

n

Clone N°

Analysis of RUS5-Luc clones Expressing RenillaLuciferase and GM-CSFα Receptor Chain

0

5

10

15

20

25

30

0.001 0.01 0.1 1 10 100

Fold

ng/ml

US5 Dose Response Curve (16h Incubation)

GM-CSF

IL-3

IL-5

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0.00001 0.001 0.1 10

Fold

GMCSF ng/ml

US5 GM-CSF (4h incubation)

0

2

4

6

8

10

12

0.00001 0.001 0.1 10Fo

ldIn

duct

ion

GMCSF ng/ml

US5 GM-CSF (16h incubation)

TATA

GM-CSF Gene Reporter Assay (FireFly / Renilla)

SV40Poly A

Intron

+1

Stat5 Stat5 GM-CSF

(T/A)TTCnnnGAA(T/A) FireFly Luciferase

CMVSV40Poly A

Intron

+1 Renilla LuciferaseIRES

GM-CSF-Rα+1

Transient Cotransfection of U937 Cells

FireFly Activity

Renilla Activity

0

10

20

30

40

50

60

Control GM-CSF 100ng/ml

Fold

Indu

ctio

n

STAT5Luc+RGM-CSF/hRL

STAT5-Luc+EGFP

0

10

20

30

40

50

60

Control GM-CSF 100ng/ml

Fold

Indu

ctio

n

STAT5Luc+RGM-CSF/hRL

STAT5-Luc+EGFP

Normalized Reporter Gene Assay for GM-CSF

Unexpectedly, constitutive expression of Renilla luciferase is influenced (2x) by STAT5 activation in response to GM-CSF treatmentConsequence effective dynamic range of assay reducedMechanism unclear, no Stat5 recognition sequences in promoter construct Solution change promoter use SV40 constitutive promoter & Herpes simplex TK promoterUse humanized Gaussia luciferase gene instead of Renilla luciferase

GM-CSF Induced FL Expression Normalized Relative to Gaussia Expression (4 H Induction)

0

5

10

15

20

25

30

0.000001 0.00001 0.0001 0.001 0.01 0.1 1 10 100

Nor

mal

ized

Fol

d In

duct

ion

GM-CSF ng/ml

US5 17.5.21 SV

US5 17.5.8 TK

GM-CSF Induced FL Expression Normalized Relative to Gaussia Expression (16 H Induction)

0

5

10

15

20

25

30

35

40

45

50

0.000001 0.00001 0.0001 0.001 0.01 0.1 1 10 100

Nor

mal

ized

Fol

d In

duct

ion

GM-CSF ng/ml

US5 17.5.21 SV

US5 17.5.8 TK

Relationship Between Drug Induced FL Expression and Cell Number

0

50000

100000

150000

200000

250000

300000

0.00001 0.001 0.1 10

RLU

(FL)

Drug (ng/ml)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.00001 0.001 0.1 10

RLU

(FL/

RL)

Drug (ng/ml)

80.000 cells

40.000

20.000

10.000

Response of Reporter Cells to Serum Matrix Effects

0

5

10

15

20

25

30

35

40

45

0.0001 0.01 1 100

Fold Indu

ction

Drug (ng/ml)

RA Sera 10%

RA sera 20%

Normal Sera 10%

Control

0

10

20

30

40

50

60

70

80

0.0001 0.01 1 100

Fold

Indu

ctio

n

Drug (ng/ml)

RA Sera 10%

RA sera 20%

Normal Sera 10%

Control

Stability Studies: Sensitivity

Time (Hrs) Passage # EC50 (pg/ml) LLOQ (pg/ml)4 10 40 104 20 80 204 40 20 518 10 400 20018 20 400 20018 40 200 100

Stability Studies: Proliferation

Passage # Doubling Time (Hrs)

Max Cell Density (x106/ml)

4 27 1.0

10 25 0.95

20 24 1.2

40 24 1.1

Conclusions - I

A cell-based assay specific for a growth factor can be established by the use of a cell-line that does not require the growth factor or other related grow factors for proliferationThe cell does, however, contain a functional receptor/signal transduction system for the growth factor of interestThe assay can be rendered specific by over-expression of a growth factor specific binding receptor sub-unit

Conclusions - II

When a growth factor signals through multiple pathways (MAPK, NFkB, STAT1-5 etc) it may be more useful to reconstitute a complete receptor-signaling transduction system in a cell that does not respond to the cytokine of interest.

0

2

4

6

8

10

12

14

16Fo

ld I

nd

uct

ion

Stat4‐kozak ‐ + ‐ ‐ + ‐ +

CX2RB1 ‐ ‐ + ‐ ‐ + +

CX2RB2 ‐ ‐ ‐ + + + +

STAT4‐Luc + + + + + + +

Reconstitution of a Functional Cytokine Signaling Pathway in Human U937 cells

Stat4‐kozak ‐ + ‐ ‐ + ‐ +

CX2RB1 ‐ ‐ + ‐ ‐ + +

CX2RB2 ‐ ‐ ‐ + + + +

STAT4‐luc + + + + + + +

Reconstitution of a Functional Cytokine Signaling Pathway in Human HEK cells

0

2

4

6

8

10

12

14

16Fo

ld I

nd

uct

ion

Conclusions - III

Reconstitution of a cellular receptor-signal transduction system in a cell that does not respond to the cytokine or growth factor of interest is a powerful tool for the establishment of Specific cell-based assays for the quantification of the activity and neutralizing antibody response to therapeutic proteins

www.coralgablessymposia.org

Gables Symposium 2012 provides a unique forum for thought leaders to address the principal concerns regarding the immunogenicity of biopharmaceuticals; in their development, regulation, and clinical use.

Scientific Organizing Committee

Dr. Michael Tovey, ChairDr. Shalini Gupta, AmgenDr. Susan Kirshner, FDADr. Daniel Kramer, Merck SeronoDr. Robin Thorpe, NIBSC

www.coralgablessymposia.org