Poster #A18

Tumor Immunology and Immunotherapy

Imprime PGG, a soluble yeast b-glucan PAMP, synergizes with anti-PD-1 antibody to enhance CD8 T cell anti-tumor immunity

Immune checkpoint inhibitor (CPI) therapy has provided compelling, durableanti-tumor T cell immunity for some cancer patients, particularly those with pre-existing anti-tumor T cell responses. Unfortunately, the majority of patients showlittle to no pre-existing anti-cancer T cell response and do not benefit from singleagent CPI therapy. For these patients, novel therapeutic approaches thatinitiate or revitalize the anti-cancer T cell responses may be particularlypromising as combination agents with CPIs as these would provide theactivated T cell population upon which CPIs would act.

Proper activation of the innate immune system is essential to generaterobust immune responses. Central to this process are non-self danger signalsknown as Pathogen-Associated Molecular Patterns (PAMPs) that bind to andsignal via conserved pattern recognition receptors (PRRs) expressed by innateimmune effector cells (macrophages, monocytes, dendritic cells, neutrophils).Imprime PGG is a soluble yeast b-1,3/1,6 glucan PAMP currently in multipleclinical trials in combination with pembrolizumab (anti-PD-1). Unlike otherPAMPs that must be locally administered to avoid systemic cytokine storms,Imprime has been administered safely by intravenous infusion to >400 humansubjects (>325 cancer patients). Mechanistically, here we show that Imprimeactivates innate immune cells by binding the PRR dectin-1, eliciting dendriticcell and monocyte expression of co-stimulatory ligands, mobilization of myeloidcells, as well as production of a variety of chemokines and cytokines, includingtype I interferon. To examine if Imprime’s innate modulating effects alter CD8 Tcell priming, we used the ovalbumin (OVA)—specific OT-I CD8 T cell system.Following OT-I adoptive transfer into congenic C57BL/6 recipients, mice wereimmunized with OVA peptide and Imprime. Co-administration of Imprimeincreased OT-I expansion compared to peptide alone and induced theirdifferentiation into polyfunctional Tbet+ effector cells. Based on Imprime’s abilityto link innate and adaptive immune responses, we asked if Imprime wouldenhance the anti-tumor efficacy of anti-PD-1 antibody therapy. C57BL/6 micewere injected s.c. with the murine adenocarcinoma cell line MC38. Once tumorsreached ~50mm3, mice were randomized to treatment groups: vehicle, Imprime(1.2mg twice weekly), anti-PD-1 antibody (RMP1-14 clone, 100µg twiceweekly), or Imprime + anti-PD-1 antibody. While anti-PD-1 treatment alonesignificantly reduced tumor growth compared to vehicle or Imprime groups,combination treatment provided superior tumor control with a statistically higherfrequency of mice completely clearing tumor in the combination groupcompared to the anti-PD-1 group (4/57 vs 14/60, respectively). To assess theimmunobiology elicited by treatment, tumors were harvested after two weeks oftherapy, and tumor-infiltrating leukocytes were stimulated with PMA/ionomycinor tumor-specific peptides. Relative to other treatment groups, CD8 TIL fromcombination-treated mice had superior frequencies of cells capable ofproducing IFN-g, TNF-a, or IL-2. Furthermore, mice from these MC38 studiesthat had previously cleared their tumors were protected against tumor re-challenge in the opposite flank, demonstrating that these mice had durable andprotective adaptive immunity. Together, our data demonstrate that Imprime PGGis a unique, clinically relevant PAMP that can be administered systemically toactivate key innate immune cell populations and to drive T cell based anti-cancer responses. Consequently, Imprime PGG synergizes with anti-PD-1antibody treatment to provide anti-tumor CD8 T cells with superior effectorfunctions. We are currently working to extend these mechanistic findings in ourongoing phase II clinical trials.

Background

Spleen 24hrs

Neutrophils

Ly6Chi monocytes cDC1 (XCR1+)

cDC2 (CD172a+)

Vehicle

Imprime

Imprime

b cDC1 (XCR1+)Migratory

ResidentNeutrophils

Ly6Chi monocytes

Imprime

LCs

skin-draining LNs 24hrscDC2 (CD172a+)

CD11bnegCCR7+ CD11b+CCR7+

CD11b+CCR7neg

Vehicle

Imprime

c

Figure 1. Imprime binds to myeloid lineage cells

Figure 1. Imprime binds to myeloid lineage cells. Naïve C57BL/6 mice were injected with 1.2mg Imprime i.v. and various myeloid populations were examine 24hrs later for Imprime binding. (A) Identification of DC subsets. Ly6Chi monocytes and neutrophils were first gated out and then the bulk DC population was then identified as CD11c+MHC-II+Lineageneg. DC subsets were then further subset based on expression of CD172a (Sirp-a), XCR1, CD24, and CCR7. (B) Imprime binding profiles 24hrs post injection in the spleen (B) and skin-draining LNs (C). Plots are representative of 3-4 experiments with at least n = 3 per experiment.

FSC

SSC

CD11b

Ly6C

MHC II

Line

age

CD11c

MH

C II

CD172a

XCR

1

CCR7

CD

24

CD24

CC

R7

CD11b

CC

R7

Boolean gating:NOT monocytes and

neutrophilsidentification of

total DCs

cDC1

resident cDC1

dermal (CD103+)migratory

cDC1

cDC2

Langerhans cells

total cDC2

CD11b+

resident

CD11b+

migratory

CD11bneg

migratory

Ly6G

Ly6C

Ly6Chi mono

neutrophils

CD11b

Dum

p

aA

Spleen 24hrs

Neutrophils

Ly6Chi monocytes cDC1 (XCR1+)

cDC2 (CD172a+)

Vehicle

Imprime

Imprime

b cDC1 (XCR1+)Migratory

ResidentNeutrophils

Ly6Chi monocytes

Imprime

LCs

skin-draining LNs 24hrscDC2 (CD172a+)

CD11bnegCCR7+ CD11b+CCR7+

CD11b+CCR7neg

Vehicle

Imprime

c

Figure 1. Imprime binds to myeloid lineage cells

Figure 1. Imprime binds to myeloid lineage cells. Naïve C57BL/6 mice were injected with 1.2mg Imprime i.v. and various myeloid populations were examine 24hrs later for Imprime binding. (A) Identification of DC subsets. Ly6Chi monocytes and neutrophils were first gated out and then the bulk DC population was then identified as CD11c+MHC-II+Lineageneg. DC subsets were then further subset based on expression of CD172a (Sirp-a), XCR1, CD24, and CCR7. (B) Imprime binding profiles 24hrs post injection in the spleen (B) and skin-draining LNs (C). Plots are representative of 3-4 experiments with at least n = 3 per experiment.

FSC

SSC

CD11b

Ly6C

MHC II

Line

age

CD11c

MH

C II

CD172a

XCR

1

CCR7

CD

24

CD24

CC

R7

CD11b

CC

R7

Boolean gating:NOT monocytes and

neutrophilsidentification of

total DCs

cDC1

resident cDC1

dermal (CD103+)migratory

cDC1

cDC2

Langerhans cells

total cDC2

CD11b+

resident

CD11b+

migratory

CD11bneg

migratory

Ly6G

Ly6C

Ly6Chi mono

neutrophils

CD11b

Dum

p

a

B

cDC1 r

eside

nt

cDC1 m

igrato

ry

cDC2 C

D11b- m

igrato

ry

cDC2 C

D11b+ m

igrato

ry

cDC2 C

D11b+ re

siden

tLC

s0.0

0.5

1.0

1.5

2.0

2.5

Fold

cha

nge

in M

FI

over

veh

icle

CD86

* *

***

*****

MFI

fold

cha

nge

over

veh

icle

cDC1 r

eside

nt

cDC1 m

igrato

ry

cDC2 C

D11b- m

igrato

ry

cDC2 C

D11b+ m

igrato

ry

cDC2 C

D11b+ re

siden

tLC

s0.0

0.5

1.0

1.5

2.0CD40

***

**

Imprime matures DCsC

cDC1 r

eside

nt

cDC1 m

igrato

ry

cDC2 C

D11b- m

igrato

ry

cDC2 C

D11b+ m

igrato

ry

cDC2 C

D11b+ re

siden

tLC

s0

1

2

3PD-L1

***

**

***

*****

CD86 CD40 PD-L1

D

Ly6C

hi mon

o

Neutro

phils

Ly6C

hi mon

o

Neutro

phils

Ly6C

hi mon

o

Neutro

phils

0

5

10

15

20

1

Fold

cha

nge

over

veh

icle Spleen

LNBlood

** ***

***

***

* ***

Ly6C

hi mon

o

Neutro

phils

0

5

10

15

20

1

Fold

cha

nge

over

veh

icle

Dectin-1-/-WT

***

***

Imprime mobilizes and matures inflammatory Ly6Chi monocytes

Figure 1. Imprime binds to and activates myeloid lineage cells. Naïve C57BL/6 mice were injected with 1.2mg Imprime i.v. and various myeloid populations in skin-draining LNs (sdLNs) were examine 24hrs later for Imprime binding. (A) Identification of DC subsets in skin-draining LNs. Ly6Chi monocytes and neutrophils were first gated out and then the bulk DC population was then identified as CD11c+MHC-II+Lineageneg. DC subsets were then further subset based on expression of CD172a (Sirp-a), XCR1, CD24, and CCR7. (B) Imprime binding profiles 24hrs post injection in the sdLNs and (C) expression levels of CD86, CD40, and PD-L1. (D) Imprime causes Dectin-1-dependent mobilization of Ly6Chi monocytes in sdLNs (E) and upregulation of activation and co-stimulatory molecules.

CD11c MHC II

CD86 CD40 PD-L1

PBSImprime

ELy6Chi monocytes

LNs 24hrs post Imprime

Peripheral LNs 16hrs post Imprime i.v.

Cd274Cd40Cd80

Clec7a

0

1

2

3

4

5

Rel

ativ

e m

RN

A ex

pres

sion

*

**

*

Ifng

Tnfa Il1a

Il1b Il6

Il23a

(p19

)

Il12b

(p40

)0

4

8

12

16

202040

Rel

ativ

e m

RN

A ex

pres

sion

**

*

*

*

**

Ccl2Ccl3Ccl4Cxcl1Cxcl2

0

5

10

15

20

Rel

ativ

e m

RN

A ex

pres

sion

*

*

*

*

*

Irf5Irf7Irf9

Socs1

Socs3Stat1Stat2Stat3

0

1

2

3

4

5

Rel

ativ

e m

RN

A ex

pres

sion

*

*

*

*

*

**

*

Mx1

Oas1a

Oasl1

Usp18

0

2

3

4

1

Rel

ativ

e m

RN

A ex

pres

sion

**

**

* *

WTDectin-1-/-

Ifnar1-/-

Activation-associated Cytokines Chemokines

Signaling IFN stimulated genesImprime binds to antigen-presenting cells

Imprime induces a Dectin-1-dependent pro-inflammatory profileA

B

PBS

Impri

me0

100

200

300

400

500

pg/1

0mg

tissu

e

TNF-α

***

PBS

Impri

me0

200

400

600

800

1000

pg/1

0mg

tissu

e

IFN-γ

*

PBS

Impri

me0

20

40

60

80

pg/1

0mg

tissu

e

GM-CSF

***

PBS

Impri

me0

500

1000

1500

2000

pg/1

0mg

tissu

e

IL-6

**

PBS

Impri

me0

20

40

60

80

100

pg/1

0mg

tissu

e

IL-1β

**

PBS

Impri

me0

1000

2000

3000

4000

5000

pg/1

0mg

tissu

e

CCL2*

PBS

Impri

me0

100

200

300

400

pg/1

0mg

tissu

e

CCL3***

PBS

Impri

me0

50

100

150

pg/1

0mg

tissu

e

CCL4

*

PBS

Impri

me0

10

20

30

40

pg/1

0mg

tissu

e

CXCL2

*

In vivo cytokine production in LNs: 24hrs post Imprime

Figure 2. Imprime induces Dectin-1-dependent production of pro-inflammatory cytokines and chemokines. (A) Naïve WT or Dectin-1-/- C57BL/6 mice were injected with PBS or 1.2mg Imprime i.v. and 16hrs later skin-draining lymph nodes were harvested and analyzed for transcriptional changes using a custom QuantiGene Plex. Identified genes were categorized into Activation-associated, Cytokines, Chemokines, Signaling and IFN-stimulated genes. IFNaR1-/- mice were used to confirm the dependence of ISG’s on type I IFN. (B) Naïve C57BL/6 mice were injected with PBS or 1.2mg Imprime i.v. Skin-draining LNs were harvested 24hrs later and cell lysates were analyzed for cytokine and chemokine proteins using the Luminex platform.

OVA peptide +/- Imprime (1.2mg)

Exp 9996

Transfer 1x105 naïve OT-I CD8 T cells specific to OVA257-264 peptide

C57BL/6 Day 7 Harvest spleen/LN Assess OT-I response

(peptide and Imprime are NOT conjugated)

WT Dectin-1-/-0

1

2

3

4

5

% O

T-I o

f tot

al C

D8

d5 PBL

PBSImprimeLPS

*** ***

ns

**

Statistics were done using a one-way ANOVA, Kruskal-Wallis nonparametric test (because of combining multiple experiments there were significant differences in stdev using a parametric test).

WT Dectin-1-/-0.0

0.5

1.0

1.5

% O

T-I o

f tot

al C

D8

d7 spleen PBSImprimeLPS

******

ns

**

Statistics were done using a one-way ANOVA, Kruskal-Wallis nonparametric test (because of combining multiple experiments there were significant differences in stdev using a parametric test).

Day 5 blood Day 7 spleen

Figure 3. Imprime enhances expansion and effectordifferentiation of antigen-specific CD8 T cells. (A)Vaccination model. Naïve WT or Dectin-1-/- mice received OT-ICD8 T cells. The following day they were injected with 100µgOVA257-264 peptide i.v. and either PBS, Imprime, or LPS. Themagnitude of the OT-I response was assessed on day 5 in theblood (B) and day 7 in the spleen (C). (D) OT-I CD44 and Tbetexpression on day 7 in the spleen. (E) On day 7, splenocyteswere stimulated with OVA peptide for 5hrs with brefeldin A. Cellswere then stained for intracellular IFN-g, TNF-a, and IL-2.

A

B C

E

PBS

Impri

meLP

S0

20

40

60

80

100

% o

f OT-

I

SPDP

none

TP

*** **

***

*****

***ns

PBS

Impri

meLP

S0

20

40

60

80

100

% o

f OT-

I

SPDP

none

TP

*** **

***

*****

***ns

Cytokine production:combinations of IFN-g, TNF-a, IL-2

none1 cytokine2 cytokines3 cytokines

% o

f OT-

I

D

Tbet

CD

44

+PBS +Imprime

HostOT-I

Vehicl

e

Impr

ime

PD1

Impr

ime+

PD10

10

20

30

40

50

% m

ice

with

no

palp

able

tum

or

4/57

14/60

1/57 0/59

**

Figure 4. Imprime synergizes with anti-PD-1 antibody to reduce MC38 tumor growth. C57BL/6 mice were injected with MC38 s.c. When tumors were ~50mm3, mice were treated with PBS (Vehicle), Imprime, anti-PD-1 (clone RMP1-14), or anti-PD-1 + Imprime. (A) Tumor growth kinetics. Mean tumor volume is shown without error bars. (B) Tumor volumes on d21 post tumor inoculation. Each symbol represents a single mouse and data is shown as mean +/- SEM. (C) Percentage of MC38 tumor-bearing mice that cleared tumor by d21 post inoculation. Fractions show the number of mice without palpable tumors out of the total number of mice in each treatment group pooled from 3 experiments. (D) Comparison of granzyme B expression by CD8 TIL after 2 weeks of treatment. (E) TIL were isolated from MC38 tumors after 14 days of treatment and stimulated with PMA/ionomycin. Tumor size versus % CD8 IFN-g, TNF-a, or IL-2 production was plotted for each individual mouse. Linear correlations (solid lines) and Pearson correlations were calculated for each treatment group.

Imprime+PD1

Imprime

PD1

Vehicle

DAPICD8GrzB

Vehicle

ImprimePD1

Combo

0

20

40

60

% o

f CD

3

CD8+GrzB+

**

Fig. 6d

Imprime+PD1

Imprime

PD1

Vehicle

DAPICD8GrzB

Vehicle

ImprimePD1

Combo

0

20

40

60

% o

f CD

3

CD8+GrzB+

**

Fig. 6dVehicle Imprime

PD-1 Imprime + PD-16 9 12 15 18 21

0100200300400500600700

Day post tumor

Mea

n Tu

mor

Vol

ume

(mm

3 )

VehicleImprimePD1Imprime+PD1

day post tumortreatment initiation Veh

icle

Impr

ime

PD1

Impr

ime+

PD10

200

400

600

800

1000

1200

Tum

or V

olum

e (m

m3 )

*

A B

C

D

0 500 1000 15000

20

40

60

80

100

tumor mm3

% IF

N-γ

+ of

CD

8 TI

L(P

MA

/iono

myc

in)

VehicleImprimePD-1Combo

0 500 1000 15000

10

20

30

40

50

tumor mm3

% IL

-2+

of C

D8

TIL

(PM

A/io

nom

ycin

)

0 500 1000 15000

20

40

60

80

100

tumor mm3

% T

NF-α

+ of

CD

8 TI

L(P

MA

/iono

myc

in)

IFN-g TNF-a IL-2E

A general structure of yeast-derived Imprime

PGG

Conclusions• Imprime binds to a wide variety of professional

antigen-presenting cells (APCs).• Imprime matures APCs via Dectin-1, initiating

production of localized pro-inflammatory cytokines • Imprime enhances CD8 T cell expansion and their

capacity to produce effector cytokines.• Imprime synergizes with anti-PD-1 antibody

therapy in the murine MC38 tumor model. Combination therapy results in increased clearance of established tumors, increased granzyme B+ CD8 TIL, and improved TNF-a/IL-2 production by CD8 TIL.

• These data provide rationale for Imprime and checkpoint inhibitor therapy which is currently being tested in clinical trials.

Ross B. Fulton, Steven M. Leonardo, Kathryn A. Fraser, Takashi O. Kangas, Adria B. Jonas, Anissa S.H. Chan, Nandita Bose, Keith B. Gorden, Jeremy R. Graff, and Mark Uhlik. Biothera Pharmaceuticals, Inc., Eagan MN, 55121 rfulton@biothera.com