A LAG-3/PD-L1 bispecific antibody inhibits tumour growth in two syngeneic colon carcinoma models

Matthew Kraman, Katarzyna Kmiecik, Carlo Zimarino, Mustapha Faroudi, Katy Everett, Mateusz Wydro, Jacqueline Doody

F-star, Cambridge, United Kingdom

POSTER

128

CONCLUSIONS

The anti-mouse LAG-3/PD-L1 mAb2 bispecific antibody:

Maintains biophysical characteristics of IgG

Co-engages both antigens at nanomolar affinities

Potently activates T cells in vitro

Inhibits tumour growth more efficiently than combination of two antibodies

ABSTRACTA bispecific antibody against murine LAG-3 and PD-L1 was generated which binds bothantigens simultaneously with nanomolar affinities. The anti-LAG-3/PD-L1 bispecificantibody (mAb²) inhibits LAG-3 binding to MHC II and PD-L1 binding to PD-1 and CD80,thereby enhancing T cell activation in an in vitro assay. This potency translates into in vivoefficacy, where the anti-LAG-3/PD-L1 bispecific antibody decreased tumour burden in anMC38 colon carcinoma tumour model. At the end of the study tumour-free animals weremore numerous in the LAG-3/PD-L1 bispecific group than in the group treated with acombination of anti-LAG-3 and PD-L1 mAb. Efficacy was also seen in a CT26 murine coloncancer model, where the anti-LAG-3/PD-L1 bispecific antibody also demonstrated morepotent anti-tumour activity compared to the mAb combinations. Thus, the preclinicaldata supports developing an anti-human LAG-3/PD-L1 bispecific for the treatment ofcancer patients.

BACKGROUNDF-star’s Modular Antibody Technology™ platform introduces a novel antigen binding site into the constant(Fc) region of an antibody to create a so-called Fcab™ (an Fc-domain with antigen binding activity).

Fcab mAb²

LAG-3 FCAB BINDING SITES LAG-3 FCAB BINDING SITES

PD-L1 FAB BINDING SITESThe resulting Fcab is then used as a building block for otherdrug formats. In particular, an Fcab can be easily combinedwith the variable region (Fab) of an existing antibody togenerate a full-length bispecific antibody or mAb2™.

This simple “plug-and-play” capability of F-star’s highlyefficient discovery engine creates virtually limitlessbispecific product opportunities.

SITC 2016, November 9-13, 2016

mAb2 inhibits tumour growth in syngeneic colon carcinoma models

LAG-3/PD-L1 mAb² suppresses tumour growth in the MC38 syngeneic tumour model.

CohortNumber of tumour free animals at end of study

LAG-3/PD-L1 mAb² 6/8

LAG-3 mAb + PD-L1 mAb 0/7

LAG-3/mock mAb² 0/7

PD-L1 mAb 1/7

LAG-3 mAb 0/6

IgG Control 0/9

0 7 8 11 14

Figure 13. CT26 tumour cells were injected subcutaneously in Balb/c mice until a palpable tumour formed. Three doses of antibody mixes were administered at day 8, day 11, and day 14. In all cohorts a total of 20mg/kg (10mg/kg antibody A + 10mg/kg antibody B or IgG control) total antibody was administered.

LAG-3/PD-L1 mAb2 is superior to a combination of monoclonal antibodies in suppressing tumour growth in the CT26 syngeneic tumour model.

1500

1000

500

2000

LAG-3/mock mAb²

LAG-3/PD-L1 mAb²LAG-3 mAb + PD-L1 mAb

PD-L1 mAbLAG-3 mAbIgG Control

Tum

ou

r V

olu

me

(m

m3)

CT26 Colon Carcinoma Growth Curves

3

2

1

0

We

igh

t (g

)

Figure 12. MC38 tumour cells were injected subcutaneously into C57/Bl6 mice and grown until a palpable tumour formed. Three doses were administered at day 8, day 11, and day 14. In all cohorts a total of 20mg/kg (10mg/kg antibody A + 10mg/kg antibody B or IgG control) of total antibody was administered.

1500

1000

500

20Time (days)

LAG-3/mock mAb²

LAG-3/PD-L1 mAb²

LAG-3 mAb + PD-L1 mAb

PD-L1 mAb

LAG-3 mAb

IgG Control

MC38 Colon Carcinoma Growth Curves

*

3

2

1

0

We

igh

t (g

)

MC38 End Tumour Weights

0 7 8 11 14 17

Time (days)

Tum

ou

r V

olu

me

(m

m3)

Figure 11. DO11.10 cells were incubated with theindicated mAb or mAb2 while LK35.2 cells wereincubated with OVA peptide and then combined.Cells were incubated at 37oC for 24 hours prior tomeasuring IL-2 levels in the supernatant by ELISA.

IL-2

re

leas

e (

pg/

mL)

2,000

1,500

1,000

500

00.01 0.1 1 10 100

IL-2

re

leas

e (

pg/

mL)

600

400

200

00.01 0.1 1 10 100

0 .01 0 .1 1 10 100

0

5 0 0

1 ,0 0 0

1 ,5 0 0

2 ,0 0 0

2 ,5 0 0

3 ,0 0 0

[co n c e n tra tio n ] (n M )

[m I

L-2

] (p

g/m

l)

2,000

1,500

1,000

500

0

IL-2

re

leas

e (

pg/

mL) 2,500

3,000

0.01 0.1 1 10 100

mAb or mAb² concentration (nM)

LAG-3/PD-L1 mAb²

LAG-3/mock mAb² + PD-L1 mAb

LAG-3/mock mAb²

LAG-3 mAb

PD-L1 mAb

LAG-3 mAb + PD-L1 mAb

mAb or mAb² concentration (nM)

mAb2 potently activates T cells in vitro

Figure 10. DO11.10 murine CD4 T cell hybridoma specific for ovalbumin are mixed with LK53.2 B lymphoma cells,leading to the activation and IL-2 release by the DO11.10 cells. Transduced murine LAG-3 in DO11.10 cells binds tonaturally occurring MHC II on the LK35.2 cells, leading to inhibition of IL-2 release. Inclusion of an anti-LAG-3 mAb ormAb2 to the cells releases the LAG-3 inhibition and increases IL-2 secretion. Alternatively, transduced PD-L1 in LK35.2cells binds to naturally expressed PD-1 on DO11.10 cells resulting in inhibition of IL-2 release that can be relieved byaddition of anti-PD-L1 mAb or mAb2. Investigation of dual inhibition can be accomplished by incubating LAG-3expressing DO11.10 cells with PD-L1 expressing LK35.2 cells.

LAG-3/PD-L1 mAb² blocks immune suppression mediated by LAG-3 and PD-L1.

DO11.10 T cell - LAG-3

LK35.2- PD-L1

0 .0 1 1 1 0 0

0

1 0 0 0

2 0 0 0

3 0 0 0

4 0 0 0

m L A G -3 /P D -L 1 m A b2

m L A G -3 /m o c k m A b2

m P D -L 1 P C I

m L A G -3 P C 1

Ig G C o n tro l

FITC

MFI

4,000

3,000

2,000

1,000

00.01 1 100

mAb or mAb² concentration (nM)

LAG-3/PD-L1 mAb²LAG-3/mock mAb²PD-L1 mAbLAG-3 mAbIgG Control

Binding to recombinant LAG-3 Binding to LAG-3 expressed on cells Blocking LAG-3-binding to MHC II

Re

sp. D

iff

(RU

)

0 80 160 240 320 400 480 560 640 720 800

Time (sec)

0

20

40

60

80

100

120

Ka (1Ms) Kd (1/s) KD (M)

LAG-3 2.7E+06 1.8E-03 9.8E-10

Ka (1Ms) Kd (1/s) KD (M)

PD-L1 7.0E+05 6.3E-05 9.0E-110 .0 0 1 0 .0 1 0 .1 1 1 0 1 0 0

0

1 0 0 0

2 0 0 0

3 0 0 0

4 0 0 0

5 0 0 0

B in d in g o f m L A G -3 /P D -L 1 m A b2 to H E K c e lls e x p re s s in g P D -L 1

L o g c o n c [n M ]

FIT

C M

FI

m L A G -3 /P D -L 1 m A b2

m L A G -3 /m o c k m A b2

m P D -L 1 P C I

m L A G -3 P C 1

Ig G C o n tro l

FITC

MFI

4,000

3,000

2,000

1,000

0

5,000

0.001 0.01 0.1 1 10 100

LAG-3/PD-L1 mAb²LAG-3/mock mAb²PD-L1 mAbLAG-3 mAbIgG Control

Figure 5. Flow cytometry analysis to measure the bindingof LAG-3/PD-L1 mAb2 to HEK-293 cells transduced withmouse PD-L1.

Figure 4. Fitted sensorgrams for affinity determination.Binding affinity of LAG-3/PD-L1 mAb2 to immobilizedrecombinant murine PD-L1 was measured using SurfacePlasmon Resonance (SPR).

Binding to recombinant PD-L1 Binding to PD-L1 expressed on cells Blocking PD-L1-binding to PD-1

0 80 160 240 320 400 480 560 640 720 800

Time (sec)

020406080

100120140160180

Re

sp. D

iff

(RU

)

PD-L1 mAb

0 .0 1 0 .1 1 1 0 1 0 0 1 0 0 0

0

2 0 0

4 0 0

6 0 0

8 0 0

1 0 0 0

mAb concentration (nM)

1000

800

400GM

FI

10000.01 0.1 1 10 100

600

200

0

mAb or Fcab concentration (nM)mAb or mAb² concentration (nM)

Monomeric expression of LAG-3/PD-L1 mAb2

m L A G - 3 / P D - L 1 m A b2

s i m u l t a n e o u s b i n d i n g

2 0 0 4 0 0 6 0 0 8 0 0

- 5 0

0

5 0

1 0 0

1 5 0

2 0 0

T i m e ( s )

Re

sp

. D

iff.

m P D - L 1 - F c c o a t e d c h ip

I n j e c t i o n o f

m L A G - 3 / P D - L 1 m A b2

I n j e c t i o n o f

m L A G 3 - F c

Re

sp. D

iff

(RU

)

0

50

100

150

200

0 200 400 600 800

Injectionof LAG-3-Fc

Injection of LAG-3/PD-L1 mAb²

Time (sec)

104 6 8 12 14 16 18

8.562

0 2

mA

U

350

250

200

100

150

300

50

0

Time (min)

0 .1 1 1 0 1 0 0 1 0 0 0

0 .0

0 .5

1 .0

1 .5

m L A G 3

m F c

m C D 4

m P D L 1

m O X 4 0

F c a b c o n c e n tra tio n (n M )

GM

FI

10000.1 1 10 100

1.0

0.5

0.0

1.5 LAG-3FcCD4PD-L1OX40

OD

(4

50

nm

)

Fcab concentration (nM)

Specific binding of anti-LAG-3 Fcab to LAG-3 Simultaneous binding to LAG-3 and PD-L1

Figure 6. Flow cytometry analysis to determine the binding ofrecombinant mouse PD-L1 to PD-1 expressing HEK293 cells inthe presence of an IgG1 containing the anti-PD-L1 Fab portionof the bispecific mAb2

Figure 1. Fitted sensorgrams for affinity determination.Binding affinity of LAG-3/PD-L1 mAb2 to immobilizedrecombinant murine LAG-3 was measured using SurfacePlasmon Resonance (SPR).

Figure 2. Flow cytometry analysis to measure the bindingof LAG-3/PD-L1 mAb2 to HEK-293 cells transduced withmouse LAG-3.

Figure 3. Flow cytometry analysis to determine the bindingof recombinant mouse LAG-3 to MHC II expressing A375cells in the presence of anti-LAG-3 Fcab. Anti-LAG-3antibody used does not block MHC II binding to LAG-3.

IgG ControlPD-L1 mAb

Figure 7. Sensorgram of LAG-3/PD-L1 mAb2 binding to twoantigens simultaneously, with PD-L1-Fc immobilised on anSPR chip and dual injection of the mAb2 and LAG-3-Fc.

Figure 8. Specified antigens were coated on an ELISAplate and incubated with the anti-LAG-3 Fcab. Bindingwas detected using anti-CH2-HRP.

Figure 9. Data collected using SEC. No significantaggregation or fragmentation is observed.

0 . 0 1 0 . 1 1 1 0 1 0 0 1 0 0 0

0

5 0

1 0 0

1 5 0

L A G -3 F c a b

m A b o r F c a b c o n c e n t r a t io n (n M )

Bin

din

g (

% o

f u

nc

om

pe

ted

)

L A G -3 m A b

Anti-mouse mAb2 binds both murine LAG-3 and PD-L1 with nanomolar affinities

DO11.10 T cell Activation Assay

Decrease in IL-2 production Increase in IL-2 production

CT26 End Tumour Weights

DO11.10 LAG-3 + LK35.2 cells DO11.10 + LK35.2 PD-L1 cells

DO11.10 LAG-3 + LK35.2 PD-L1 cells 800