NK Cells Engineered to Express a Bispecific CD123/NKG2D … · 2019-11-28 · Antigen Receptor...
1
Contact James Trager, PhD [email protected] www.nkartatx.com Background Natural killer (NK) cells have an emerging role in cellular immunotherapy for patients with acute myeloid leukemia (AML). NK cells are highly active and suitable for allogeneic use as they are not HLA-restricted and do not cause GVHD. We sought to demonstrate that engineering of NK cells can improve their ability to recognize and kill AML cells and control tumor burden. Natural killer group 2D (NKG2D) is a transmembrane protein belonging to the CD94/NKG2 family of C-type lectin-like receptors. Ligands of NKG2D (NKG2D- L) are highly expressed in AML tissues but weakly on healthy tissues, making NKG2D an attractive target for NK cell engineering. The interleukin-3 receptor alpha chain (CD123) has also been identified as a potential immunotherapeutic target for high risk AML patients due to its overexpression in AML blasts and leukemia stem cells compared with normal hematopoietic stem cells. We therefore developed a CD123-NKG2D tandem CAR construct that contains human CD123-specific single-chain variable fragment (scFv) and NKG2D. We co-expressed this chimera in NK cells with membrane-bound IL-15 to support prolonged cell survival and proliferation. Results Retroviral transduction of NK cells with tandem CAR (n=6) resulted in 68.4% and 37.8% median CD123 and NKG2D CAR positivity respectively when measured 14 days post-transduction (Figure 4). In a 4 hour cytotoxicity assay, the CD123-NKG2D tandem CAR transduced NK cells were no more effective than NKG2D-CAR alone in killing the CD123 low AML cell line HL60, but showed better cytotoxicity than Mock (GFP)- transduced NK cells, indicating that the tandem CAR can trigger robust cytotoxicity through NKG2D (figure 5). CD123-NKG2D tandem CAR NK cells were effective at killing the CD123 + THP-1 AML cell line (figure 6a), exhibiting significantly greater cytotoxicity than Mock (GFP)- transduced NK cells, or NK cells expressing either CD123-CAR (Figure 6a); in a kinetically resolved cytotoxicity assay, the tandem CAR was more effective than the NKG2D CAR against THP-1 (figure 6b). In addition, the cytotoxicity of CD123- NKG2D tandem CAR transduced NK cells was efficiently blocked by either CD123 peptides (Figure 7) or anti-NKG2D antibodies (Figure 5b), further demonstrating the ability of the tandem receptor to be triggered by both component binding moieties. Studies to test the in vivo anti- tumor efficacy of CD123-NKG2D tandem CAR transduced NK cells in an NSG mouse model of THP-1 are under way. References 1. Chang et al. Cancer Res; 73(6) March 15, 2013. 2. Hilpert et al, Journal of immunology, 2012, 189: 1360–1371. 3. Gill et al, Blood 2014 123:2343-2354. 4. Fernandex-Messina et al, 2012 Front. Immunol., 25 September ;3:299. 5. Carlsten and Childs, Front Immunol. 2015 Jun 10;6:266. Conclusion In summary, our data demonstrate the capacity of CD123-NKG2D tandem CAR NK cells to kill both CD123 + and NKG2D-L + targets and support the clinical use of tandem CAR NK cells for patients with AML. NK Cells Engineered to Express a Bispecific CD123/NKG2D Chimeric Antigen Receptor (CAR) and IL-15 As Off-the-Shelf Therapy for Acute Myeloid Leukemia Chao Guo, PhD 1,2 , Luxuan Buren, PhD 1,2 , Alexander Aronov 1 , James Trager, PhD 1 * 1 Nkarta Inc, South San Francisco CA 94080, USA 2 Contributed equally *Corresponding author CD123/NKG2D Bispecifics: mIL-15 CD123 scFv CD3z CD3z NKG2D mIL-15 CD3z CD123 scFv NKG2D mIL-15 CD123 CAR: NKG2D CAR: Figure 2. Targeting AML with CAR-NK cells and CD123-NKG2D-CAR design Schematic of retroviral vector encoding CD123-NKG2D-CD3ζ and membrane bound IL-15 Figure 4. CD123/NKG2D bispecifics can be expressed at high levels in NK cells. D a y 7 0 2 0 4 0 6 0 8 0 1 0 0 % C D 1 2 3 -F la g E x p re s s io n G F P C D 1 2 3 C A R C D 1 2 3 /N K G 2 D B is p e c ific s N K G 2 D C A R d a y 7 0 2 0 4 0 6 0 8 0 1 0 0 % N K G 2 D E x p re s s io n Summary of %CD123 expression (n=6 donors) Summary of %NKG2D expression (n=6 donors) Figure 6. CD123-NKG2D bispecific CAR confers best durable cytotoxicity of NK cells for at least 2 weeks after NK transduction. Cytotoxicity of NK-CARs and GFP-transduced NK cells against the AML cell line 2 weeks post-transduction. (A) 4 hour cytotoxicity assay. (B) In vitro cytotoxicity assay was monitored for up to 4 days in culture with IncuCyte ® live-cell imaging. E:T ratio is 1:8. 1 :1 1 :2 1 0 2 0 3 0 4 0 5 0 T H P -1 4 h rs p o s t tra n s d u c tio n d 1 4 % C y to to x ity 0 .0 0 .3 0 .5 0 .8 1 .0 1 .3 1 .5 1 .8 2 .0 2 .3 2 .5 2 .8 3 .0 3 .3 3 .5 3 .8 0 2 0 0 4 0 0 6 0 0 8 0 0 E :T = 1 :8 c e ll n u m b e rs G F P C D 1 2 3 -C A R N K G 2 D -C A R C D 1 2 3 -N K G 2 D -C A R T H P -1 o n ly Low residual lysis of THP-1 when CD123-CAR is blocked by CD123 peptides (1ug/ml) G F P C D 1 2 3 -C A R C D 1 2 3 -N K G 2 D -C A R N K G 2 D -C A R 0 2 0 4 0 6 0 8 0 % C y to to x ity G F P C D 1 2 3 -C A R C D 1 2 3 -N K G 2 D -C A R N K G 2 D -C A R + - +- +- +- CD123 peptide Figure 7. Selective cytotoxicity against CD123high THP-1 line demonstrates that bispecific constructs can fire through CD123. A significant difference (P < 0.05) between lysis in the presence and absence of CD123 peptide is indicated by an asterisk. Figure 1. Nkarta Natural Killer((NK) cell platform showing ex vivo natural killer (NK) cell activation and expansion procedures Figure 3. NKG2D-CAR (NKX101) shows dose-dependent efficacy in an AML model d a y 6 d a y 1 2 d a y 1 8 d a y 2 4 d a y 3 0 d a y 3 6 0 1 ´ 1 0 9 2 ´ 1 0 9 3 ´ 1 0 9 4 ´ 1 0 9 5 ´ 1 0 9 T im e (d a y s ) B L I (p h o to n /s e c ) G F P N K X 1 0 1 (5 M ) N K X 1 0 1 (1 0 M ) N K X 1 0 1 (2 0 M ) 1 2 3 4 5 Day 6 Day 12 Day 18 Day 24 Day 30 Day 36 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 GFP NKX101(5M) NKX101(10M) NKX101(20M) NKG2D Ligands Hilpert et al, Journal of immunology, 2012 Gill et al, Blood, 2014 D0 D6 D12 D18 D24 D30 THP-1 (IV) (2 x 10 5 ) Imaging Imaging Imaging Imaging Imaging IL-2 Infuse NKs (IV) D3 (A) Bone marrow aspirate showing AML (B) SFI levels of the various NKG2DL obtained with each single patient of the different leukemia entities. (C) Primary patient AML samples express CD123. (D) CD123-NKG2D-CAR design A. B. C. D. Antitumor activity of NKG2D transduced NK cells in THP-1 AML NSG xenograft model (A) Schematic timeline of the AML NSG xenograft model. (B) Luciferase labeled THP-1 cells (2x10 5 ) were injected intravenously in NSG at Day 0. Control mice (GFP; n=5) received 5x10 6 NK cells. The remaining 15 mice received a single intravenous injection of 5x10 6 , 10x10 6 and 20x10 6 NK cells, transduced with NKG2D-CAR followed by IL-2 injections every other day for 1 week. Photoluminescence signals were measured at weekly intervals by IVIS Spectrum In Vivo Imaging system. A. B. C. (A) CD123 expression and (B) NKG2D expression in expanded NK cells from 6 donors transduced with a retroviral vector containing GFP only or a vector encoding a CAR construct. A. B. B. A. Days: * * EXPAND ENGINEER THAW & ADMINISTER NK cell γ-retrovirus Tumor targeting – NKG2D, other targets Homing – Tumor microenvironment evasion Persistence – mbIL-15 Stimulatory cell line ISOLATE EXPAND Donor CRYOPRESERVE Patient E:T Figure 5. Gain of function and loss of function studies demonstrate that bispecific constructs can fire through NKG2D (A) Results of 4-hour cytotoxicity assays performed by against the CD123 low HL60 cell line. (B) A significant difference (P < 0.05) between lysis in the presence and absence of NKG2D blockade is indicated by an asterisk. 1 ;1 1 :2 2 0 4 0 6 0 8 0 H L 6 0 4 h rs 1 4 d p o s t tra n s d u c tio n E :T % C y to to x ity G F P C D 1 2 3 -C A R C D 1 2 3 -N K G 2 D -C A R N K G 2 D -C A R Low residual lysis of HL60 when NKG2D is blocked GFP CD123-CAR CD123-NKG2D-CAR NKG2D-CAR 0 20 40 60 80 % Cytotoxity GFP CD123-CAR CD123-NKG2D-CAR NKG2D-CAR + - +- +- +- anti-NKG2D blocking * * * * A. B.
Transcript of NK Cells Engineered to Express a Bispecific CD123/NKG2D … · 2019-11-28 · Antigen Receptor...
Contact
JamesTrager,[email protected]
Background Natural killer (NK) cells have an emerging role in cellular immunotherapy for patients with acute myeloid leukemia (AML). NK cells are highly active and suitable for allogeneic use as they are not HLA-restricted and do not cause GVHD. We sought to demonstrate that engineering of NK cells can improve their ability to recognize and kill AML cells and control tumor burden. Natural killer group 2D (NKG2D) is a transmembrane protein belonging to the CD94/NKG2 family of C-type lectin-like receptors. Ligands of NKG2D (NKG2D-L) are highly expressed in AML tissues but weakly on healthy tissues, making NKG2D an attractive target for NK cell engineering. The interleukin-3 receptor alpha chain (CD123) has also been identified as a potential immunotherapeutic target for high risk AML patients due to its overexpression in AML blasts and leukemia stem cells compared with normal hematopoietic stem cells. We therefore developed a CD123-NKG2D tandem CAR construct that contains human CD123-specific single-chain variable fragment (scFv) and NKG2D. We co-expressed this chimera in NK cells with membrane-bound IL-15 to support prolonged cell survival and proliferation.
Results Retroviral transduction of NK cells with tandem CAR (n=6) resulted in 68.4% and 37.8% median CD123 and NKG2D CAR positivity respectively when measured 14 days post-transduction (Figure 4). In a 4 hour cytotoxicity assay, the CD123-NKG2D tandem CAR transduced NK cells were no more effective than NKG2D-CAR alone in killing the CD123low AML cell line HL60, but showed better cytotoxicity than Mock (GFP)-transduced NK cells, indicating that the tandem CAR can trigger robust cytotoxicity through NKG2D (figure 5). CD123-NKG2D tandem CAR NK cells were effective at killing the CD123+ THP-1 AML cell line (figure 6a), exhibiting significantly greater cytotoxicity than Mock (GFP)-transduced NK cells, or NK cells expressing either CD123-CAR (Figure 6a); in a kinetically resolved cytotoxicity assay, the tandem CAR was more effective than the NKG2D CAR against THP-1 (figure 6b). In addition, the cytotoxicity of CD123-NKG2D tandem CAR transduced NK cells was efficiently blocked by either CD123 peptides (Figure 7) or anti-NKG2D antibodies (Figure 5b), further demonstrating the ability of the tandem receptor to be triggered by both component binding moieties. Studies to test the in vivo anti-tumor efficacy of CD123-NKG2D tandem CAR transduced NK cells in an NSG mouse model of THP-1 are under way.
References 1. Chang et al. Cancer Res; 73(6) March 15, 2013. 2. Hilpert et al, Journal of immunology, 2012, 189: 1360–1371. 3. Gill et al, Blood 2014 123:2343-2354. 4. Fernandex-Messina et al, 2012 Front. Immunol., 25 September ;3:299. 5. Carlsten and Childs, Front Immunol. 2015 Jun 10;6:266.
Conclusion In summary, our data demonstrate the capacity of CD123-NKG2D tandem CAR NK cells to kill both CD123+ and NKG2D-L+ targets and support the clinical use of tandem CAR NK cells for patients with AML.
NKCellsEngineeredtoExpressaBispecificCD123/NKG2DChimericAntigenReceptor(CAR)andIL-15AsOff-the-ShelfTherapyforAcuteMyeloidLeukemiaChao Guo, PhD 1,2, Luxuan Buren, PhD1,2, Alexander Aronov1, James Trager, PhD1* 1Nkarta Inc, South San Francisco CA 94080, USA 2Contributed equally *Corresponding author
CD123/NKG2DBispecifics:
mIL-15CD123scFv CD3z
CD3zNKG2D mIL-15
CD3zCD123scFv NKG2D mIL-15
CD123CAR:
NKG2DCAR:
Figure 2. Targeting AML with CAR-NK cells and CD123-NKG2D-CAR design
Schematic of retroviral vector encoding CD123-NKG2D-CD3ζ and membrane bound IL-15
Figure 4. CD123/NKG2D bispecifics can be expressed at high levels in NK cells.
Da
y 7
0
2 0
4 0
6 0
8 0
1 0 0
%C
D1
23
-F
la
g
E
xp
re
ss
io
n
G F P
C D 1 2 3 C A R
C D 1 2 3 / N K G 2 D B i s p e c i f i c s
N K G 2 D C A R
da
y7
0
2 0
4 0
6 0
8 0
1 0 0
%N
KG
2D
E
xp
re
ss
io
n
Summaryof%CD123expression(n=6donors) Summaryof%NKG2Dexpression(n=6donors)
Figure 6. CD123-NKG2D bispecific CAR confers best durable cytotoxicity of NK cells for at least 2 weeks after NK transduction. Cytotoxicity of NK-CARs and GFP-transduced NK cells against the AML cell line 2 weeks post-transduction. (A) 4 hour cytotoxicity assay. (B) In vitro cytotoxicity assay was monitored for up to 4 days in culture with IncuCyte® live-cell imaging. E:T ratio is 1:8.
1: 1
1: 2
1 0
2 0
3 0
4 0
5 0
T H P - 1 4 h r s p o s t t r a n s d u c t i o n d 1 4
%
Cy
to
to
xi
ty
0. 0
0. 3
0. 5
0. 8
1. 0
1. 3
1. 5
1. 8
2. 0
2. 3
2. 5
2. 8
3. 0
3. 3
3. 5
3. 8
0
2 0 0
4 0 0
6 0 0
8 0 0
E : T = 1 : 8
ce
ll
n
um
be
rs
G F P
C D 1 2 3 - C A R
N K G 2 D - C A R
C D 1 2 3 - N K G 2 D - C A R
T H P - 1 o n l y
LowresiduallysisofTHP-1whenCD123-CARisblockedbyCD123peptides(1ug/ml)
GF
P
CD
12
3- C
AR
CD
12
3- N
KG
2D
- CA
R
NK
G2
D- C
AR
0
2 0
4 0
6 0
8 0
%
Cy
to
to
xi
ty
G F P
C D 1 2 3 - C A R
C D 1 2 3 - N K G 2 D - C A R
N K G 2 D - C A R
+-+-+-+-CD123peptide
Figure 7. Selective cytotoxicity against CD123high THP-1 line demonstrates that bispecific constructs can fire through CD123. A significant difference (P < 0.05) between lysis in the presence and absence of CD123 peptide is indicated by an asterisk.
Figure 1. Nkarta Natural Killer((NK) cell platform showing ex vivo natural killer (NK) cell activation and expansion procedures
Figure 3. NKG2D-CAR (NKX101) shows dose-dependent efficacy in an AML model
da
y6
da
y1
2
da
y1
8
da
y2
4
da
y3
0
da
y3
6
0
1 ´ 1 09
2 ´ 1 09
3 ´ 1 09
4 ´ 1 09
5 ´ 1 09
T i m e ( d a y s )
BL
I
(p
ho
to
n/
se
c)
G F P
N K X 1 0 1 ( 5 M )
N K X 1 0 1 ( 1 0 M )
N K X 1 0 1 ( 2 0 M )
12345
Day6
Day12
Day18
Day24
Day30
Day36
12345 12345 12345
GFP NKX101(5M) NKX101(10M) NKX101(20M)
NKG2DLigands
Hilpertetal,Journalofimmunology,2012 Gilletal,Blood,2014
D0 D6 D12 D18 D24 D30
THP-1(IV)(2x105)
Imaging
Imaging Imaging Imaging ImagingIL-2
InfuseNKs(IV)
D3
(A) Bone marrow aspirate showing AML (B) SFI levels of the various NKG2DL obtained with each single patient of the different leukemia entities. (C) Primary patient AML samples express CD123. (D) CD123-NKG2D-CAR design
A. B.C.
D.
Antitumor activity of NKG2D transduced NK cells in THP-1 AML NSG xenograft model (A) Schematic timeline of the AML NSG xenograft model. (B) Luciferase labeled THP-1 cells (2x105) were injected intravenously in NSG at Day 0. Control mice (GFP; n=5) received 5x106 NK cells. The remaining 15 mice received a single intravenous injection of 5x106 , 10x106 and 20x106 NK cells, transduced with NKG2D-CAR followed by IL-2 injections every other day for 1 week. Photoluminescence signals were measured at weekly intervals by IVIS Spectrum In Vivo Imaging system.
A.
B. C.
(A) CD123 expression and (B) NKG2D expression in expanded NK cells from 6 donors transduced with a retroviral vector containing GFP only or a vector encoding a CAR construct.
A. B.
B.A.
Days:
* *
EXPAND ENGINEER THAW&ADMINISTERNKcell
γ-retrovirusTumortargeting–NKG2D,othertargetsHoming–TumormicroenvironmentevasionPersistence–mbIL-15
Stimulatorycellline
ISOLATE EXPAND
Donor
CRYOPRESERVE
Patient
E:T
Figure 5. Gain of function and loss of function studies demonstrate that bispecific constructs can fire through NKG2D (A) Results of 4-hour cytotoxicity assays performed by against the CD123low HL60 cell line. (B) A significant difference (P < 0.05) between lysis in the presence and absence of NKG2D blockade is indicated by an asterisk.
1; 1
1: 2
2 0
4 0
6 0
8 0
H L 6 0 4 h r s 1 4 d p o s t t r a n s d u c t i o n
E : T
%
Cy
to
to
xi
ty
G F P
C D 1 2 3 - C A R
C D 1 2 3 - N K G 2 D - C A R
N K G 2 D - C A R
LowresiduallysisofHL60whenNKG2Disblocked
GFP
CD123-C
AR
CD123-N
KG2D-C
AR
NKG2D-C
AR0
20
40
60
80
% C
ytot
oxity
GFPCD123-CARCD123-NKG2D-CARNKG2D-CAR
+-+-+-+-anti-NKG2Dblocking
*
*
**
A.
B.
