GITRL-Fc can significantly reduce tumor growth by stimulating innate and adaptive immunity Hyun-Bae Jie, Minu K. Srivastava, Erin Mayes, Rui Yun, Fumiko Axelrod, Jorge Monteon, Austin Gurney, and Angie In-Kyung Park

OncoMed Pharmaceuticals, Inc., Redwood City, CA

ABSTRACT

MATERIALS AND METHODS

RESULTS AND CONCLUSIONS

SUMMARY

Although the interaction between GITR (Glucocorticoid-Induced Tumor Necrosis Factor

Receptor) and GITRL (its ligand) is important for the development of immune responses, the

cellular mechanisms underlying anti-tumor immunity including both innate and adaptive

immunity is not fully understood. We generated a novel single-gene linkerless GITRL trimer

fused to an immunoglobulin Fc domain (GITRL-Fc), and investigated its effect on controlling

tumor growth and immune responses in preclinical tumor models. Treatment with GITRL-Fc

significantly reduced tumor growth in several preclinical tumor models by inducing Th1 biased

anti-tumor immunity and reducing Treg-mediated immune suppression. Immune cell depletion

studies showed that anti-tumor immunity induced by GITRL-Fc depended on CD8+ T cells as

well as NK cells. Furthermore, the combination of GITRL-Fc with PD-1 blockade significantly

reduced the tumor growth in the Renca murine kidney adenocarcinoma tumor model. Taken

together, these results suggest that GITRL-Fc can improve cancer treatment as a single agent

or in combination therapy by enhancing innate and adaptive cellular immunity.

Cell

membrane

Native form of GITRL

trimer on cell surface

“stalk”

Structure and stimulatory function of GITRL-Fc engineered using a novel single-gene

linkerless GITRL trimer

Single-gene GITRL trimer significantly reduce tumor growth

by inhibiting Treg suppressive function and inducing Th1 response

GITRL-Fc mediates robust anti-tumor activity through T and NK cells

0.00

1000.00

2000.00

3000.00

4000.00

5000.00

6000.00

7000.00

8000.00

0.00 0.01 0.10 1.00 10.00 100.00

GIT

R S

ign

alin

g (

NF

-kB

Lu

cif

era

se

)

Agonist GITR mAb

Control mAb

Cells alone

ELISPOT data measuring antigen-specific interferon gamma production

Control Fc GITRL-Fc

T-cell mediated cytotoxicity

0 5 10 15 20 25 30 35 40 450

100

200

300

400

500

600

Days

Tu

mo

r V

olu

me,

mm

3

0 5 10 15 20 25 30 35 40 450

100

200

300

400

500

600

Days

Tu

mo

r V

olu

me,

mm

3

0 10 20 300

100

200

300

400

500

600

Days

Tu

mo

r V

olu

me,

mm

3

0 5 10 15 20 25 30 35 40 450

100

200

300

400

500

600

Days

Tu

mo

r V

olu

me,

mm

3

Anti-PD1 GITRL-Fc + anti-PD1

GITRL-Fc protein. DNA constructs were generated that encoded murine GITRL extracellular domain

concatamerized to form a single-gene trimer and fused to the N- or C-terminus of a murine IgG Fc domain as

indicated in the diagrams.

Luciferase Reporter assay: The ability of GITRL activation of NFkB signaling was determined by in vitro

luciferase reporter assays. HEK-293 cells were stably transfected with an expression vector encoding a full-

length human or mouse GITR as well as plasmids encoding a NFkB-dependent luciferase reporter construct.

GITR and NFkB luciferase reporter expressing HEK-293 cells were plated to 96 well plates and incubated

overnight. Serially diluted recombinant fusion proteins or antibodies were added to the appropriate wells and

incubated overnight. Luciferase levels were measured 18 hours later using a Steady Glo luciferase assay kit

(Promega).

In Vivo Studies: The murine colon carcinoma (CT26-WT, ATCC CRL-2638), the murine renal cortical

adenocarcinoma (Renca, ATCC CRL-2947), and the murine skin melanoma (B16F10, ATCC CRL-6475) were

obtained from American Type Culture Collection. Single cell suspensions of CT26 or Renca tumor cells were

injected subcutaneously into the flanks of 7-8 week old Balb/c or C57BL/6J mice. One or two weeks following

tumor inoculation, mice with palpable tumors were injected i.p. individually with GITRL-Fc, anti-GITR (DTA-1),

anti-PD1, or combination according to the schedule described in the results and conclusions. Isotype antibodies

were used as a control. Anti-CD4 (clone: GK1.5), anti-CD8 (clone: YTS169.4), anti-GM1 and anti-NK1.1

antibodies were used to deplete CD4+, CD8+ T cells and NK cells, respectively. Tumor volumes were monitored

by measuring two bisecting diameters of each tumor with electronic calipers. Tumor volumes were calculated

using the formula: V=0.5ab2, with a as the larger diameter and b as the smaller diameter.

ELISPOT: Splenocytes were cultured in the presence and absence of tumor specific CD8 T-cell peptide (1μg/ml

final concentration of AH-1 peptide for CT26 model) in T cell media for 48 hrs followed by the ELISPOT assay as

described in manufacturer’s instructions.

T cell and NK cell Cytotoxicity Assay: Ten million splenocytes were cultured with tumor specific CD8 T-cell

peptide AH1 for 7 days, washed and counted. These effector T cells were co-cultured with calcein AM labeled

CT26 tumor cell targets (E:T of 50:1) for four hours in triplicate wells in a 96-well plate and the supernatants

were collected and measured for the release of calcein from the tumor cells. For NK cell cytotoxicity assay, YAC-

1 cells labeled with calcein AM were used as NK cell target (E:T of 25:1). Specific lysis is calculated using the

formula: Specific lysis (%)=(experiment release-target spontaneous)/(maximum release-target

spontaneous)X100.

Treg Suppression Assay CD4+CD25+ regulatory T-cells (from each treatment group) were isolated in 2 steps

using a Treg isolation kit (Miltenyi). T-cells were isolated from spleens of naïve mice by negative selection and

labeled with VTD dye. T-cells were then stimulated with anti-CD3 and anti-CD28 antibody and incubated with

Tregs for three days. Cells were then stained with anti-CD4 or anti-CD8 and the dilution of VTD was used to

calculate T-cell proliferation.

GITRL-Fc is active and more potent than agonist GITR mAb.

GITRL-Fc induces a potent anti-tumor immunity by promoting Th1 type immune response

and inhibiting Treg suppressive function.

GITRL-Fc significantly reduces tumor growth in a T cell- or NK cell-dependent manner.

GITRL-Fc enables complete eradication of some tumors as a single agent.

Combination of GITRL-Fc with anti-PD1 blockade significantly reduces tumor growth.

Treg mediated suppression

d0 500,000 Renca on Right Flank

d7 d10 d12 d16 d20

GITRL

GITRL-Fc

In vivo tumor growth

AH-1

IFN-g ELISPOT

0

100

200

300

400

500

600

700

800

Control DTA-1 336B3

Sp

ot

Co

un

ts

IFN-g Spots

-AH1 +AH1

p=0.005

p=0.006

GITRL-Fc

GITRL-Fc

-400 -200 0 200 400 600 800

IFN-g

IL-10 GITRL-Fc: 8.3

DTA-1: 3.1

Control: 5.3

IFN-g/IL-10 ratio

400 200 0 200 400 600 800 (ELISPOT #, + AH-1 peptide)

GITRL-Fc and PD-1 blockade significantly reduces tumor growth

d0 30K CT26 on Right Flank

d7 d12 d15

GITRL (336B3):0 .25mg/mouse 2QW, three doses

a-CD4, a-CD8 and a-Ganglioside Ab: 0.5 mg/mouse, 2QW four doses

0

1000

2000

3000

4000

12 16 19

Tu

mo

r vo

lum

e

(mm

3)

NK1.1 + Iso

IgG2a + Iso

0

1000

2000

3000

4000

12 16 19

Tu

mo

r vo

lum

e (

mm

3) NK1.1 + 336B3 (Low)

IgG2a + 336B3 (Low)

p<0.02

GITRL(Low)

GITRL(Low)

0

1000

2000

3000

4000

12 16 19

Tu

mo

r vo

lum

e (

mm

3)

NK1.1 + DTA

IgG2a + DTA

DTA-1

DTA-1

0

1000

2000

3000

4000

12 16 19

Tu

mo

r vo

lum

e (

mm

3)

NK1.1 + 336B3 (High)

IgG2a + 336B3 (High)

GITRL(High)

GITRL(High)

p<0.0005

D0. B16F10 injection

GITRL-Fc (Low= 0.1, high= 0.25 mg per mouse

or DTA-1 Ab 0.25mg per mouse

-2 -1 +1

NK1.1: 0.5 mg per mouse

+16 +12 +7

** *

0

2

4

6

8

10

12

14

16

18

Iso DTA-1 GITRL (High)

Sp

ec

ific

lys

is (

%)

NK1.1

IgG2a

*: p<0.005, **: p<0.0005

The roles of T cells in GITRL-Fc-mediated

anti-tumor immunity

The role of NK cells in GITRL-Fc-mediated

anti-tumor immunity

NK cell cytolytic activity enhanced by

GITRL-Fc and DTA-1 antibody

Control DTA-1 336B3

-10

-5

0

5

10

15

target: CT26WTE:T=25:1

Sp

ecif

ic L

ysis

(%

)

GITRL-Fc

NF-kB luciferase assay

0

500

1000

1500

2000

2500

0 5 10 15 20 25

Me

an

Tu

mo

r V

olu

me

(m

m3

)

Days

a-CD4 + GITRL-Fc a-CD8 + GITRL-Fc

a-CD4 + a-CD8+ GITRL-Fc a-Ganglioside + GITRL-Fc

LFT-2+ GITRL-Fc LFT-2+ IgG2A

*

** *: p < 0.0001

**: p<0.00001

Concentration (μg/ml)

Figure 1. Structure of GITRL-Fc (A) and GITR Signaling

assay measuring NF-kB dependent luciferase activity

(B). Both monovalent and bivalent GITRL-Fc fusion

proteins are more active than agonist GITR mAb DTA-1.

A

B

Figure 2. GITRL trimer significantly reduces tumor growth by inhibiting Treg suppressive function and inducing Th1 response. (A) The effect of DTA-1 and GITRL-Fc

on tumor growth in vivo. Tumor injection and drug treatment was performed as described in the Materials and Methods section. T cell medicated cytotoxicity (B), Treg

suppression assay (C), IFN-g ELISPOT assay with AH-1 peptides (D), IFN-g/IL-10 ratio calculation (E) were performed using splenocytes from the mice treated with

drugs as indicated.

A

Figure 3. GITRL-Fc mediates robust anti-tumor immunity in a T or NK

cell dependent manner. (A) Significant tumor growth was observed in

mice treated with anti-CD8/-CD4 depleting antibodies compared with

anti-CD8 Ab alone in CT26-WT model, (B) Significant tumor growth

was observed in mice treated with NK cell-depleting Ab (NK1.1)

compared with isotype control in B16F10 melanoma model. The drug

dose and injection schedule is indicated in the figure, (C) NK cell

cytolytic activity was measured using splenocytes from mice from the

experiment described in B.

Figure 4. GITRL-Fc displays single

agent activity in promoting rejection of

murine Renca kidney adenocarcinoma

cell line and combination activity with

anti-PD1. The drug dose and schedule

is indicated in the figure.

B C

D E

A B

C

d0 30K CT26.WT on Right Flank

d7 d10 d14 d17

Dose: 0.25mg/mouse 2QW, four doses

Anti-PD1

d23

d19

0

200

400

600

800

1000

1200

1400

1600

6 10 14 17 21 24 28

Me

an

Tu

mo

r V

olu

me

(m

m3

)

Days

Cont DTA-1 336B3

P<0.02