abstract target a target b · six different immunization approaches were attempted (4 mice per...

LakePharma’s Hybridoma-based Antibody Discovery to

Support CAR T Development

Yayue (Ashlee) Zheng1, Miao Tan1, Matthew Jang1, Margaret L. Wong1, Abdul Khan2 , David P. Andrew2 , Ivo C. Lorenz2, and Brian A. Zabel1*

1LakePharma, 201 Industrial Rd, San Carlos, CA 94070, USA, 2Tri-Institutional Therapeutics Discovery Institute, Belfer Research Building, New York, NY 10021, USA *Corresponding author E-mail address: [email protected]

ABSTRACT

Chimeric antigen receptor T cell (CAR T cell) therapy is effective in treating blood cancers and holds

promise for treating solid tumors. The FDA has recently approved two types of CD19-targeting CAR T

cells, tisagenlecleucel (Kymriah™, Novartis) in leukemia (August 2017) and lymphoma (May 2018) and

axicabtagene ciloleucel (Yescarta™, Kite Pharma) in lymphoma (October 2017). Nearly 1,000 CAR T

clinical trials are ongoing worldwide (Fig. 1). For the next generation of CARs, novel tumor antigen-

binding mAbs are paramount. Here we describe the discovery of diverse, potent, fully human mAbs for

two CAR T tumor targets.

Figure 1. CAR T clinical trials. Geographic distribution of CAR T clinical trials worldwide (A) and specificallywithin the United States (B), as of October 2019, based on data retrieved from clinicaltrials.gov.

A B

IgG mAb Discovery

scFv

Figure 2. CAR Design and Tumor Cell Killing. (A) Tumor antigen-specific IgG monoclonal antibodies are identifiedby hybridoma-based antibody discovery. Immunizing transgenic mice that express fully human VH and VL

domains reduces potential efficacy-limiting CAR antigenicity. (B) The VH and VL sequences are reformatted as asingle chain fragment variable (scFv) domain, which comprises the CAR ectodomain. A transmembrane domainand intracellular signaling domain (e.g. costimulatory and CD3ζ sequences) complete the CAR. (C) Patient-derivedT cells are typically transduced with a virus encoding the CAR construct, and these CAR T cells are expanded invitro and infused into the patient where they specifically target and kill tumor cells.

L L

HH

LH

A B C

RESULTS

Figure 5. Tumor cell binding potency. FACS potency (EC50) analysis of purified mAbs againsttumor lines that express endogenous target. mAbs with a range of cell binding potencies (0.4 -40.1 nM) were discovered, many exceeding the potency of the comparator antibody. Doseresponse curves are shown for mAb A7 and a (+) control mAb against 2 tumor lines (left), andEC50 values are shown for 10 lead candidates (right).

-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -P 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl

U 9 3 7 P o s c tr l

U 9 3 7 2 n d A b o n ly

S e t2 g L

S e t2 g L N e g c tr l

S e t2 g L P o s c tr l

S e t2 g L 2 n d A b o n ly

-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -B -2 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -3 -P 1 4 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

1 3 6 2 9 -2 -N 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

1 0 0 0 0 0

2 0 0 0 0 0

3 0 0 0 0 0

1 3 6 2 9 -2 -F 1 8 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

8 0 0 0 0 0

1 3 6 2 9 -1 -P 2 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




mAb A7 tumor line 1

mAb A7 tumor line 2

Pos ctrl tumor line 1

Pos ctrl tumor line 2

Neg ctrl stains

-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -P 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -B -2 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -3 -P 1 4 -A

L o g (u g /m l)

MF

IU 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

1 3 6 2 9 -2 -N 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

1 0 0 0 0 0

2 0 0 0 0 0

3 0 0 0 0 0

1 3 6 2 9 -2 -F 1 8 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

8 0 0 0 0 0

1 3 6 2 9 -1 -P 2 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -P 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -B -2 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -3 -P 1 4 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

1 3 6 2 9 -2 -N 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

1 0 0 0 0 0

2 0 0 0 0 0

3 0 0 0 0 0

1 3 6 2 9 -2 -F 1 8 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

8 0 0 0 0 0

1 3 6 2 9 -1 -P 2 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -P 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -B -2 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -3 -P 1 4 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

1 3 6 2 9 -2 -N 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

1 0 0 0 0 0

2 0 0 0 0 0

3 0 0 0 0 0

1 3 6 2 9 -2 -F 1 8 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

8 0 0 0 0 0

1 3 6 2 9 -1 -P 2 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -P 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -B -2 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -3 -P 1 4 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

1 3 6 2 9 -2 -N 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

1 0 0 0 0 0

2 0 0 0 0 0

3 0 0 0 0 0

1 3 6 2 9 -2 -F 1 8 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

8 0 0 0 0 0

1 3 6 2 9 -1 -P 2 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -P 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -4 -B -2 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

2 0 0 0 0 0

1 3 6 2 9 -3 -P 1 4 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

1 3 6 2 9 -2 -N 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

1 0 0 0 0 0

2 0 0 0 0 0

3 0 0 0 0 0

1 3 6 2 9 -2 -F 1 8 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L



S e t2 g L 2 n d A b o n ly-3 -2 -1 0 1 2

0

2 0 0 0 0 0

4 0 0 0 0 0

6 0 0 0 0 0

8 0 0 0 0 0

1 3 6 2 9 -1 -P 2 3 -A

L o g (u g /m l)

MF

I

U 937

U 9 3 7 N e g c trl


U 9 3 7 2 n d A b o n ly

S e t2 g L




mAb EC50 (nM) mAb EC50 (nM)

A1 2.5 A6 1.1

A2 40.1 A7 6.3

A3 3.2 A8 10.1

A4 0.4 A9 0.9

A5 11.7 A10 0.6

(+) ctrl 3.3

0100000200000300000400000500000600000

M1 M2 M3 M4 Control M

0100000200000300000400000500000600000


Immunization Cohort 1: Target A

protein ECD; HT-HOCKTM

immunization; 4 mice (human

antibody-producing transgenic);

plasma titer checks by ELISA.

0100000200000300000400000500000600000


0100000200000300000400000500000600000


D17 D31

D38D17

ELIS

A R

LU

ELIS

A R

LUEL

ISA

RLU

ELIS

A R

LU

Plasma dilution

Plasma dilution Plasma dilution

Plasma dilution

Figure 3. In-life plasma titer checks over time identify mice with maximum titer for antibodyrecovery. Six different immunization approaches were attempted (4 mice per cohort) usingprotein, DNA, and cellular immunogens. Mice immunized with Target A protein ECD (Cohort 1)or Target A protein ECD limited to the membrane-proximal domain (Cohort 4) using a HT-HOCK™

protocol (High Titer Hock injections) showed the best plasma titers. Plasma titers improved fromD17 to D31/D38 in all mice in cohorts 1 (M1-4) and 4 (M5-8).

Immunization Cohort 4: Target A

protein ECD limited to just the

membrane-proximal domain; HT-

HOCKTM immunization; 4 mice

(human antibody-producing

transgenic); plasma titer checks by

ELISA.

Five cell lines used in 3° screen:Target A (ECD) TransfectantTarget A (ECD with mem prox domain only) TransfectantTumor line (endogenous Target A+)Tumor line (Target A-negative)Parental

A10A7 A2 A4 (+) ctrl

Figure 4. FACS tertiary screen confirms domain-specific Target A cell binding. 96 hits wereidentified in the primary ELISA screen. 33 hits confirmed in the 3° cell binding screen by FACS. 4representative mAbs are shown. A7 and A10 originated from Cohort 1; A2 and A4 originatedfrom Cohort 4. Antibodies arising from Cohort 4 stained cell lines expressing the membraneproximal domain only, as expected (blue staining).

mAb staining

No

rmal

ized

to

mo

de

Figure 10. VH and VL SequenceDiversity among Target B Binders.Clustal alignment and dendrogramstructures show relatedness of VH

and VL sequences. mAb siblingsbased on VH homology are indicatedby blue shading; light chain siblingsindicated by red shading.

VH Dendrogram

No. amino acid changes

B1

B2

B8

B19

B22

B20

B24

B21

B23

B18

B25

B26

B7

B9

B14

B16

B17

B15

B13

B3

B4

B5

B6

B10

B11

B12

VL Dendrogram


B23

B19

B16

B7

B14

B24

B12

B12a

B15

B22

B17

B18

B2

B13

B25

B8

B20

B21

B26

B6

B10

B11

B3

B4

B9

B5

Immunization with

Target B protein ECD;

RIMMS7TM

immunization; 4 mice

(human antibody-

producing transgenic);

plasma titer checks by

ELISA.

D10 D17

ELIS

A R

LU

ELIS

A R

LU

Plasma dilution

Plasma dilution

Figure 7. In-life plasma titer checks over time identify mice with maximum titer forantibody recovery. Mice were immunized using a RIMMS7™ protocol (Rapid ImmunizationMultiple Sites 7 (7 sites)). Plasma titers improved from D10 to D17 in all mice.

0100000200000300000400000500000600000700000800000

1:200 1:800 1:3200 1:12800 1:51200 1:1024001:2048000

100000200000300000400000500000600000700000800000

1:200 1:400 1:800 1:1600 1:3200 1:6400 1:12800

M1 M2 M3

M4 Control M1 Control M2

Figure 9. Identification of desired non-internalizing mAbs. Target B+ cell lines wereincubated with discovered mAbs or positive control(10 μg/mL) at 4 °C, washed, and then incubated atinternalization-restrictive (4 °C) or permissive (37°C) temperatures for 1 h. Secondary antibody wasthen added to assess levels of surface staining byFACS. Surface staining with the (+) ctrl mAb, aknown internalizer, was lost following 37 °Cincubation, whereas surface staining withdiscovered mAbs B19 and B22 was retained.

(+) Ctrl, known

internalizer

4°CInternalization-

restrictive

37°C Internalization-

permissive

mAb B19

mAb B22

mAb staining

2nd onlymAb

2nd onlymAb

2nd onlymAb

2nd onlymAb

2nd onlymAb

2nd onlymAb

Figure 6. VH and VL sequence diversity amongTarget A binders. Clustal alignment anddendrogram structures show relatedness of VH

and VL sequences. mAb siblings based on VH

homology are indicated by shading.

VH Dendrogram


A1

A1a

A8

A9

A10

A2

A3

A4

A7

A5

A6

A7

A4

A9

A10

A2

A3

A5

A6

A1

A8

VL Dendrogram


Figure 8. ELISA primary screen identified 384 hitswith RLU >225,000. Ten (10) 384-well microplateswere screened by ELISA against Target B protein.104 hits confirmed in the secondary screen tobind Target B-positive cells by FACS.

0

100000

200000

300000

400000

500000

600000

700000

800000

ELIS

A R

LU

Well IDBACKGROUNDCARs are synthetic chimeric proteins introduced into patient T cells that commandeer antigenic

specificity and redirect T cells to recognize tumor antigens and kill tumor cells. CARs typically comprise

a single chain variable fragment (scFv) ectodomain originally derived from a tumor antigen-binding

mAb; a transmembrane domain; a CD3ζ signaling domain, and usually one or two costimulatory

domain(s) (Fig. 2). For a given CAR, it is difficult to predict a priori the optimal scFv tumor cell binding

potency, affinity or target epitope specificity within a tumor antigen that will drive maximal CAR T

efficacy. It is known that a wide range of scFv affinities (KD 1 nM – 1 μM) can support effective anti-

tumor CAR T-dependent killing. The position of the scFv target epitope within the tumor antigen can

impact killing activity as well (e.g. a membrane-proximal epitope was superior to a membrane-distal

epitope for a CD22-targeted CAR T). It may be advantageous to select scFvs from mAbs that do not

trigger receptor internalization. It is also known that scFvs based on fully human variable domain

sequences are less antigenic than mouse-derived antibodies and thus less susceptible to efficacy-

killing anti-CAR responses in vivo. To maximize success and mitigate risk in CAR T development, a

diverse selection of fully human target-binding mAbs is desirable.

TARGET A TARGET B

REFERENCES1Watanabe et. al “Expanding the Therapeutic Window for CAR T Cell

Therapy in Solid Tumors: The Knowns and Unknowns of CAR T Cell

Biology”, Front Immunol Oct. 2018.

Dotti et. al “Design and Development of Therapies using Chimeric Antigen

Receptor-Expressing T cells”, Immunol Rev Jan. 2014.

ClinicalTrials.gov

CONCLUSIONS• For Target A, we discovered many unique and diverse domain-specific

leads. These IgGs are currently being reformatted as ectodomain scFvs for CAR development programs.

• For Target B, we discovered many unique diverse non-internalizing leads. These IgGs are currently being reformatted as ectodomain scFvs for CAR development programs.

• AAV, lentivirus engineering and production

• CAR T and stable cell line generation

Multiple discovery platforms available to support CAR T development

LakePharma Provides Integrated Solutions for Biologics Development• Given the large number of worldwide CAR T cell clinical trials

(906), many of which are in states within or near LakePharmasites (CA, TX, MA) (Fig. 1), LakePharma is well-positioned toprovide mAb discovery and CAR T cell developmentresources to help create new and effective cancertreatments.

abstract target a target b · six different immunization approaches were attempted (4 mice per...

Documents