POSTER # 6673

B7-H3-targeted 4-1BB activation potentiates CD8 T cell-dependent antitumor immunity without systemic toxicity
Gihoon You1, Yangsoon Lee2, Yeon-Woo Kang1, Han Wook Park1, Kyeongsu Park2, Jaeho Jung2, Seung-Woo Lee1,3
1Division of Integrative Biosciences and Biotechnology, Pohang University of Science and Technology (POSTECH), Pohang, Republic of Korea; 2ABL Bio, Inc., Seongnam, Republic of Korea; 3Department of Life Sciences, POSTECH, Pohang, Republic of Korea

Abstract Absence of irAEs following B7-H3×4-1BB bsAb Treatment Changes in CD8 TILs following B7-H3×4-1BB bsAb Treatment B7-H3×h4-1BB bsAb in Human 4-1BB System
A B
ns
A MCF-7 HCC1954 B
4-1BB is a costimulatory receptor on activated T and NK cells, and the stimulation of 4-1BB Treatments 70 10 **** **** 35 A B **
(B7-H3 high) (B7-H3 high)

Total BM cells (x106)

80 30
by the natural ligand or agonistic mAb enhances cellular proliferation and effector ns *

Cell number (x103/mg)

ns
Treatments ISOTYPE

in femur and tibia

60 8

% CD45+ cells
*** ** *** * 2×10 6 3,000
Day 30 B5×1D8 Urelumab Urelumab
functions. Immunotherapy targeting 4-1BB has been tested for cancer patients; however,

% CD45+ cells
0 7 14 21 28 60
50 6 ns Day 20 analog analog

IFN-γ (pg/ml)
1.5×10 6
dose-limiting toxicities of 4-1BB agonists restrict further clinical development. B7-H3 40 4 ns
*** ** 25 0 7 10 14
40 B5×1A10 2,000 1A10
(CD276) is overexpressed on the cell surface of multiple cancers and tumor-associated ISOTYPE Analysis **** ****
20
*
10
*
1×10 6 B5×1A10
30 2 MC38hB7-H3 TIL Analysis 20 ****
endothelial cells, yet barely on healthy adult tissues. To restrict 4-1BB stimulation activity 1D8 ns *** ns 1,000

Relative Light Unit (RLU)

B5×1D8 20 0 15 inoculation 5×10 5
in tumors, we have developed an FcγR-binding null bispecific IgG1 antibody consisting of NK CD4 T CD8 T CD19+ B 0 0
CD8 T non-Treg Treg NK CD8 T non-Treg Treg NK
B7-H3-targeting IgG (B5) and two anti-4-1BB single-chain variable fragments. The B7- C ns D ns
0
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 -7 -6 -5 -4 -3 -2 -1 0 1 2 3
0
-5 -4 -3 -2 -1 0 1 2
1,600 ns 150 300 100

Cell number (x103/mg)

H3×4-1BB bispecific antibody (B5×1D8) shows a potent in vitro T cell costimulatory activity ns
**** **** CD8 TILs [Ab], nM, log

Serum AST (unit/L)

**** ****
ns
C D MDA-MB-231 PANC-1 KATO Ⅲ

Serum ALT (unit/L)

*
in the presence of B7-H3 on the tumor cells. B5×1D8 rapidly accumulates in B7-H3-
1,400 ns 80 ISOTYPE B5×1D8 C

Weight (mg)
100 *** **** 200 (B7-H3 moderate) (B7-H3 moderate) (B7-H3 low)

TNF-α (pg/mg of lysate)

60 3.55 75.9 100 **** 30 ** 40 **

IFN-γ (pg/mg of lysate)

1,200
overexpressing tumors compared to 4-1BB agonistic mAb, 1D8. B5×1D8 elicits a 4-1BB- 2×10 6 120

% Survival HCC1954
TNF-α+ IFN-γ+ (%)
80
dependent anti-tumor response in three different B7-H3-overexpressing murine tumor 1,000
100
40 30

CD8 T cells
50 ns 1.5×10 6 100
20
models. More importantly, in contrast to 1D8, B5×1D8 does not induce any observable 800 **** **** 20 60
20 1×10 6 80
immune-related adverse events (irAEs). Treatment of B5×1D8 increases the density, 600 0 0 0 40
10
NK CD4 T CD8 T 10
cytokine production, and proliferation of CD8 T cells in the tumor. Characterization of TILs 20 5×10 5 60

TNF-α
indicates that B5×1D8 increases the number of PD-1+TIM-3+ “terminal effector” CD8 T cells E 15
ns
100 ns 20 ns 8 F 25 ns 30.5 15.0
0 0 0 0 40

WBC (x 10 cells/L)
**** **** **** ** ns *** ***
for eliminating tumor cells. Furthermore, a combination of B5×1D8 and immune ** **
95 15 6 20 IFN-γ -7 -6 -5 -4 -3 -2 -1 0 1 2 3 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 -5 -4 -3 -2 -1 0 1 2

Weight (mg)
[Ab], nM, log [Ab], nM, log

MON (%)
10

GRA (%)
checkpoint blockade (ICB), anti-PD1, synergistically inhibits tumor growth. The human 4-

LYM (%)
15
E F

GzmB+ (%) CD8 T cells

65.7 96.7 20 ns
1BB-targeting bispecific antibody also induces B7-H3-dependent 4-1BB costimulation and
90 10 4 150 * **
50
D

Ki-67+ CD8 T cells

10

Ki-67+ (%) CD8 T

5 40 ISOTYPE B5×1A10
2,500
inhibits tumor growth in human 4-1BB knock-in (KI) mice. In sum, our data suggest that 85 15 6,000

Tumor Volume (mm3)

5 2

Tumor Volume (mm3)

ISOTYPE

(x103/mg)
5 100
30
the B7-H3×4-1BB bispecific antibody represents an alternative form of IgG-based 4-1BB 0 80 0 0 0 10
2,000 B5×1A10
4,000
agonistic mAb for effective and safe cancer immunotherapy against B7-H3 positive 50
20 1,500

cancers as monotherapy and combination therapy with other immunotherapy, like ICB.
G 200 ns 25 ns 120 ns ns 5
10 1,000

CD8α
* ** **** **** **** **** 2,000
20 100

% CD45 cells
500 ****

% CD8 T cells
150 ns 0 0 0

Weight (mg)
ns **** **** 80 Ki-67
15 0 0

+
100 60

Introduction of B7-H3×4-1BB bsAb

0 5 10 15 20 25 0 10 20 30 0 10 20 30
10
40 G 42.2 68.8 100 25

CD8 T cells (x103/mg)

ns ***
50 **
Days after tumor inoculation Days after tumor inoculation

PD-1+ TIM-3+ (%)

5 **** ****
20 80 20

PD-1+ Tim-3+
CD8 T cells
B7-H3 (CD276) 0 0 0 Figure 6. (A) Dose-dependent costimulatory activity of Urelumab analog and B5×1A10 on Jurkat-NFκB-luc2/h4-
60 15
NK CD4 T CD8 T CD11b+ CD62L+ CD62L– 1BB reporter cells. Luminescence was measured 6 hours after stimulation with indicated cancer cells. (B and C)
• Members of B7 family
• TAA-targeting arm CD44– CD44+ 40 10 Dose-dependent costimulatory activity of Urelumab analog, 1A10, and B5×1A10 on PBMCs stimulated with
• D265A/N297A mutant • Function Figure 2. (A) Experimental scheme. C57BL/6 naïve mice (n = 6-7 per group) were treated with indicated 20 5 anti-human CD3 (5 µg/ml) and HCC1954 cells. IFN-γ secretion by ELISA (B) and optical cellular density by cell

Tim-3
for avoiding FcγR interaction - Acts as immune checkpoint antibodies once a week. Systemic alterations in each organ of antibody-treated mice were addressed 7 days counting kit (C) were analyzed 72 hours after stimulation. (D) MC38hB7-H3 tumor-bearing h4-1BB KI mice (n =
to block cross-linking of 4-1BB 0 0
- Involved in tumor metastasis/angiogenesis after the last treatment. (B) Number of bone marrow (BM) cells (left), NK-, CD4 T-, and CD8 T cell frequency 8/group) were treated with 2.25 mg/kg of hIgG1 isotype of 3 mg/kg of B5×1A10. Black arrows (↑) indicate
• Clone: B5 PD-1
- Correlated with bad prognosis, poor clinical (middle), B cell frequency (right) from femur and tibia. (C) Serum AST (left) and ALT (right). (D) Liver weight (left), treatment points. Tumor growth curves of individual mice are shown on the right. ****P < 0.0001, two-way
outcome liver-infiltrated NK-, CD4 T-, and CD8 T cell number (right). (E) Peripheral blood cell population analyzed by the Figure 4. (A) MC38hB7-H3 tumor-bearing C57BL/6 mice (n = 4-7 per group) were intraperitoneally treated with ANOVA with Bonferroni posttests for (D). ns, not significant. Data presented as mean ± SD for (A to C) and
CBC counter. WBC; white blood cells, LYM; lymphocytes, GRA; granulocytes, MON; monocytes. (F) Weight of 10.0 µg of hIgG1 isotype or 13.3 µg of B5×1D8. and tumor tissues were analyzed 4 days after last treatment. (B mean ± SEM for (D).
• Expression inguinal lymph node. (G) Spleen weight (left), NK-, CD4 T-, CD8 T-, and CD11b+ myeloid cell population (middle), and C) Flow cytometric analysis of TIL composition (B, left), cell count per mg of tumor (B, right), and TNF-α and
• 4-1BB agonistic arm - Overexpressed in a wide range of solid and subtypes of CD8 T cell (right). The immune population in BM, liver, and spleen, was analyzed by flow IFN-γ in restimulated CD8 TILs (B). (D) The protein level of TNF-α and IFN-γ in the tumor lysate by ELISA. (E to G)
• Clone: 1D8, for mouse 4-1BB
Clone: 1A10, for human 4-1BB
cancers (Colon, Lung cancer, melanoma, etc.)
- Expressed in tumor-associated endothelial
cytometry. *P < 0.05; **P < 0.01; ***P < 0.001; and ****P < 0.0001, one-way ANOVA with Bonferroni’s multiple
comparison test for (B to G). ns, not significant. Data presented as mean ± SD. All icons were “Created with
Flow cytometric analysis of Ki-67 (E), GzmB (F), and PD-1/Tim-3 (G) expression in CD8 TILs. *P < 0.05; **P < 0.01;
***P < 0.001; and ****P < 0.0001, unpaired Student’s t-test for (B to G). ns, not significant. Data presented as
Conclusion
and stromal cells BioRender.com.” mean ± SD.
• B7-H3×4-1BB bsAb elicits potent T cell
costimulatory activity and antitumor
Functional Characterization of B7-H3×4-1BB bsAb Antitumor Efficacy of B7-H3×4-1BB bsAb Synergistic Effect of B7-H3×4-1BB bsAb with PD-1 Blockade efficacy without apparent systemic irAEs.

A CD25 CD44 CD11c B IFN-γ A MC38hB7-H3 B C MC38hB7-H3 • B7-H3×4-1BB bsAb increases

15,000 1,000 3,000 250 ISOTYPE
2,000 120 2,000
A ISOTYPE or B5×1D8
B C proliferation and cytokine production of
1D8

Tumor Volume (mm3)

Tumor Volume (mm3)

ISOTYPE CD8 T cells within TME.

****
******** **** 2,000 120
**** ****
800 **** 2,500 ********
200 B5 100 (13/14) (13/14) ISOTYPE + ISO
1D8 WT
IFN-γ (ng/ml)

****

% Tumor Growth
10,000 **** **** ****
B5×1D8 1,500 1,500 Day 100 B5×1D8 + ISO
% Survival

80 B5×1D8 1,500
600 2,000 150 ISOTYPE 0 14 17 19 20 23

% Survival
****
ns 80 ISOTYPE + αPD-1 • 4-1BB stimulation on CD8 T cells induce
MFI

**** **** ****

**** ******** **** **** 1,000 60 1D8 1,000 1D8 ****
400 **** 1,500 100 ns KO 1,000 B5×1D8 + αPD-1
PD-1+ Tim-3+ terminally differentiated
******** ****
5,000 B5×1D8 60
****
*
40 B5 **** ****
500 500 MC38hB7-H3 ISO or αPD-1 ****
population.
* * ***
200 1,000 50 **** B5×1D8 40
500 ****
20 inoculation ****
0 0 500 0 **** ns
**** * 20 **
0 0 **** ns
0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 0 0
[Ab], pg/ml, log [Ab], pg/ml, log
0 5 10 15 20 25 30 35 0 20 40 60 80 100 0 5 10 15 20 25 30 35
0 5 10 15 20 25 30 35
0
0 20 40 60 80
• B7-H3×4-1BB bsAb shows synergistic
Days after tumor injection Days after tumor injection Days after tumor injection effect with PD-1 blockade.
C CD25 CD44 D S T Days after tumor injection Days after tumor injection
Normal cells (no B7-H3) B7-H3-expressing tumor cells
1,440 **
5,000 10,000
6 ** ns D Rechallenge E B16-F10 hB7-H3 F CT26 hB7-H3
• B7-H3×h4-1BB bsAb can activate human
Tumor/Spleen ratio

**** MC38 1D8 MC38hB7-H3

D
ns
4,000 ####
8,000 **** MC38hB7-H3 120 2,000 1,500 PBMC and induce antitumor activity in
Tumor Volume (mm3)

Tumor Volume (mm3)

$$$$ #### $$$$

4 MC38 ISOTYPE ISOTYPE ISOTYPE + ISO B5×1D8 + ISO ISOTYPE + αPD-1 B5×1D8 + αPD-1
3,000 6,000 ####
**** 100 B5×1D8-treated
1,500 1D8 1D8 6,000 h4-1BB KI mice.

Tumor Volume (mm )

####
This figure was “Created with BioRender.com.”
MFI

(0/10) (2/10) (1/10) (7/10)

3
**
% Survival

80 B5×1D8 1,000 B5×1D8

2,000 4,000 ns ns B5 2 ns
60 1,000 4,000
ns ns
1,000 2,000 * ns 500
40 1D8-treated

Acknowledgement
0 500 ns
0 0 ****
20 **** ns 2,000
B5×1D8
8
8

B5 5
D
1D

Naïve 0 0
PE

8
PE

8
8
8

****
×1

0
1D
1D
1D
1D

0
TY
TY

0 5 10 15 20 0 10 20 30 40
This research was supported by the BK21Plus Program funded by the Ministry of
×
×

0 50 100 150 200

B5
B5

0
O
O

IS
IS

Days after 1st inoculation Days after tumor injection Days after tumor injection
Figure 1. (A and B) Dose-dependent costimulatory activity of human IgG1 isotype, 1D8, B5, and B5×1D8 on
0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80
Education, Korea (10Z20130012243), and by the grants from ABL Bio, Inc., Korea,
Days after tumor injection
CD8 T cells stimulated with anti-CD3ε (1 µg/ml) and irradiated MC38hB7-H3 cells. Flow cytometric analysis of Figure 3. (A to C) MC38hB7-H3 tumor-bearing C57BL/6 or 4-1BB KO mice (n = 7-14 per group) were treated with and Dong-A ST Co., Ltd., Korea.
surface expressions on CD8 T cells (A) and IFN-γ secretion by ELISA (B) 72 hours after stimulation. (C) Flow indicated antibodies on day 7 and 10 after tumor injection and analyzed for tumor growth (A), survival (B) for
cytometric analysis of surface expressions on CD8 T cells stimulated with anti-CD3ε (1 µg/ml) and irradiated C57BL/6 mice, and tumor growth for C57BL/6 or 4-1BB KO mice (C). Numbers in survival curves indicate tumor-
MC38 or MC38hB7-H3 cells with indicated antibodies (1 µg/ml) 72 hours after stimulation. (D) Representative ex free mice/total mice at the end of the experiment. (D) Long-term survivors (n = 6-8 per group) from 4-1BB Figure 5. (A) Experimental scheme of combination therapy of B7-H3×4-1BB bsAb and anti PD-1 in MC38hB7-H3

Disclosures
vivo fluorescence images of spleen (S) and tumor (T) (left), and tumor-to-spleen ratio (right) from MC38hB7-H3 agonist treatments (A) were rechallenged with MC38 and analyzed for survival. (E) B16-F10hB7-H3 tumor-bearing tumor-bearing C57BL/6 mice (n = 10/group). 37.5 µg of hIgG1 ISOTYPE or 50.0 µg or bsAb were administered
tumor-bearing mice 24 hours after intravenous injection of 37.5 µg of 680XL-labeled mAb (1D8 and B5) or 50.0 C57BL/6 mice (n = 10 per group) were treated with indicated antibodies on day 6, 9, 12, and 15 after tumor intraperitoneally with 200 µg of rat IgG2a isotype (ISO) or anti-PD-1 (αPD-1) from day 14 after tumor injection
µg of 680XL-labeled B5×1D8 (n = 3/group). */$/#P < 0.05; **/$$/##P < 0.01; ***/$$$/###P < 0.001; and ****/$$$$/####P injection and analyzed for tumor growth. (F) CT26hB7-H3 tumor-bearing BALB/c mice (n = 11 per group) were (when the tumor reached an average volume of 100-200 mm3). (B to D) Tumor growth (basal tumor volume at
< 0.0001, two-way ANOVA with Bonferroni posttests compared with hIgG1 isotype group (A and B); two-way treated with indicated antibodies on day 7 and 10 after tumor injection and analyzed for tumor growth. 10.0 day 14) curves (B), survival curves (C), and tumor growth curves for individual mice (D). Numbers in in each Yangsoon Lee, Kyeongsu Park, and Jaeho Jung are employees of ABL Bio, Inc.,
ANOVA with Bonferroni posttests (C); and one-way ANOVA with Bonferroni’s multiple comparison test (D). ns, µg for mAb and 13.3 µg for bsAb were intraperitoneally administered in all experiments. *P < 0.05; **P < 0.01; plots in (D) indicate tumor-free/total mice ratios. *P < 0.05; **P < 0.01; ***P < 0.001; and ****P < 0.0001, two-way which develops and supports the supply of recombinant antibodies presented in
not significant. For (D), * compares two cell lines, $ (for MC38) and # (for MC38hB7-H3) compare each treatment ***P < 0.001; and ****P < 0.0001, two-way ANOVA with Bonferroni posttests for (A, E, and F); and Log-rank ANOVA with Bonferroni posttests for (B); and Log-rank (Mantel-Cox) test for (C). ns, not significant. Data
in one cell line. Data presented as mean ± SD. (Mantel-Cox) test for (B and D). ns, not significant. Data presented as mean ± SEM. presented as mean ± SEM.
this manuscript. The other authors declare no potential conflicts of interest.