Switchable CAR-T Cells Outperformed Traditional Antibody-

Redirected Therapeutics Targeting Breast Cancers

Supporting Data

Yu J. Cao1*, Xuechun Wang1, Zhidong Wang1, Lijun Zhao1, Shuhong Li1, Zhuxia
Zhang1, Xiaoyi Wei1, Hwayoung Yun2, Sei-hyun Choi3, Zhong Liu4, Lili Zhao5,
Stephanie A. Kazane6
1 State Key Laboratory of Chemical Oncogenomics, Key Laboratory of Chemical Genomics,
Peking University Shenzhen Graduate School, Shenzhen, Guangdong, 518055, China

2 College of Pharmacy, Pusan National University, Busan 46241, Republic of Korea

3 Department of Chemistry, The Scripps Research Institute, La Jolla, CA, 92037, USA

4 Shandong New Time Pharmaceutical Co., Ltd, No.1 North Outer Ring Road, Feixian County,
Shandong, 273400, China

5 State Engineering Laboratory of High Expression of Mammalian Cells, No.1 North Outer Ring
Road, Feixian County, Shandong, 273400, China

6 California Institute for Biomedical Research, 11119 N Torrey Pines Rd, La Jolla, CA, 92037,
USA

* To whom correspondence should be addressed: joshuacao@pku.edu.cn

S1
 Materials and Methods

Cell lines and cell culture. The breast cancer cells SKBR3, HCC1954, MDA MB453,
MDA MB361, BT-20, MDA MB231, MDA MB468 were obtained from American Type
Culture Collection. The cell line MDA MB435 and its Her2-transfected cell line, MDA
MB435/Her2 were supplied from Dr. Brunhide H. Felding (The Scripps Research
Institute, La Jolla, CA). All cell lines were maintained in DMEM medium supplemented
with 10% heat-inactivated fetal bovine serum, plus 1% antibiotics and 2mM L-glutamine.

Synthesis of FITC-linker
5-(3-(1-((1R,8S,9s)-bicyclo[6.1.0]non-4-yn-9-yl)-3-oxo-2,7,10,13-tetraoxa-4-
azapentadecan-15-yl)thio-ureido)-2-(6-hydroxy-3-oxo-3H-xanthen-9-yl)benzoic acid
(FITC linker, 16)
S O
H
O O
HN N O N O
O 1) H H
O O H
H H2N NH2 , DIPEA, DMF
N 3
O O CO2H
O 2) FITC, DIPEA, DMF
H

15 16
O O OH

To a solution of 1,11-diamino-3,6,9-trioxaundecane (515 mg, 2.68 mmol) and N,N-

diisopropyloethylamine (DIPEA, 0.39 mL, 2.23 mmol) in N,N-dimethylformamide (DMF, 5
mL) was added (1R,8S,9s)-Bicyclo[6.1.0]non-4-yn-9-ylmethyl N-succinimidyl carbonate
(15, 130 mg, 0.446 mmol) in DMF (3 mL) dropwise over 13 min. After 30 min, the
reaction mixture was concentrated in vacuo. The residue was diluted with CH2Cl2 (100
mL) and washed with 1N NaOH (10 mL  2) and H2O (10 mL). The organic layer was
dried over anhydrous Na2SO4, filtered, and concentrated in vacuo.
The residue was dissolved in DMF (5 mL) and DIPEA (0.23 mL, 1.32 mmol) was added.
To this solution was added fluorescein isothiocyanate (FITC, 171 mg, 0.39 mmol) in
DMF (4 mL) dropwise over 5 min. The reaction mixture was stirred for 2 h, and purified
by preparative HPLC, lyophilized to afford 16 (215 mg, 0.284 mmol, 64% for 2 steps): 1H
NMR (DMSO-d6, 500 MHz) δ 10.11 (s, br, 1H), 10.03 (s, br, 1H), 8.27 (s, 1H), 8.09 (s, br,
1H), 7.74 (d, J = 9 Hz, 1H), 7.17 (d, J = 9 Hz, 1H), 7.06 (t, br, J = 6Hz, 1H), 6.67 – 6.66
(m, 2H), 6.61 (s, 1H), 6.60 (s, 1H), 6.57 – 6.55 (m, 2H), 4.02 (d, J = Hz, 2H), 3.70 – 3.68
(m, 2H), 3.63 – 3.61 (m, 2H), 3.60 – 3.57 (m, 4H), 3.52 – 3.48 (m, 4H), 3.40 – 3.38 (m,
2H), 3.13 – 3.09 (m, 2H), 2.24 – 2.22 (m, 2H), 2.14 – 2.11 (m, 4H), 1.52 – 1.50 (m, 2H),

S2
1.29 – 1.23 (m, 1H), 0.87 – 0.83 (m, 2H); 13C NMR (DMSO-d6, 125 MHz) δ 181.40,
169.37, 157.29, 152.78, 129.89, 124.96, 116.50, 116.12, 113.50, 110.63, 103.10,
100.00, 99.84, 70.63, 70.61, 70.53, 70.40, 70.00, 69.30, 62.23, 44.56, 41.31, 40.95,
29.45, 21.71, 20.41, 18.52; HR-MS (ESI+) calcd. for C40H44N3O10S+ [M + H+]+ 758.2742;
found 758.2743.

Preparation of anti-Her2-FITC by site-specific conjugation. The expression of anti-

Her2 Fab with and without UAAs, and the synthesis of bifunctional cross-linkers were
essentially as previously described1, 2. The preparation of site-specific anti-Her2-FITC
were detailedly described below: Anti-Her2 Fab with pAcF was buffer exchanged into
100mM acetate buffer pH4.5 and the concentrations were adjust to 3mg/mL. The oxime
ligation was conducted with 50-fold molar excess of bifunctional linkres with an alkoxy-
amine on one terminus and an azide at the other. The reaction was complete after 16hr,
and monitored by liquid chromatography-mass spectroscopy (LCMS). Excess linker was
removed by size exclusion filtration, and the buffer was exchange to PBS. The azide
linker- Fab conjugates were mixed with FITC-linker at a 1: 50 ratio at 5mg/mL
concentration, and anti-Her2-FITC site-specific switch was obtained through a copper-
free [3+2] Huisgen cycloaddition (“Click” reaction). The unconjugated FITC-linker was
removed by size-exclusion chromatography (Superdex 200), and the final conjugates
were stored in PBS (pH7.4).

Preparation of bispecific antibodies and ADC. Bispecific antibodies were constructed

by two-step process of anti-Her2 Fab and anti-CD3 Fab by using genetically encoded
unnatural amino acids with orthogonal chemical reactivity1. The homogeneous product
has an overall purity >95%. ADC was prepared by the site-specific conjugation of anti-
Her2 IgG to auristatin using pAcPhe incorporation (HA121X) and oxime ligation3. One
molecule of anti-Her2 IgG has two auristatin molecules with >95% purity.

Lentivirus vector generation and T cell transduction. The second generation of

lentivirus vector plasmid, pRRL-SIN-EF1-WPRE and the three additional plasmids
(pMDL, pRev and pVSVG) were kindly provided by Dr. Inder M. Verma (Salk Institute for
Biological Studies, La Jolla, CA). The anti-FITC scFv (FITC-E2), anti-Her2 scFv (4D5)
and anti-Myc scFv (9E10) was respectively linked to the hinge and transmembrane
regions of human CD8 chain and the cytoplasmic regions of the human 4-1BB and

S3
CD3 molecules. Lentivirual vectors were produced after transfection of HEK293FT
human embryonic kidney cells in DMEM medium supplemented with 10% FBS, 6mM
glutamine, 1% penicillin/streptomycin, 0.1mM MEM Non-essential amino acid, 1mM
MEM sodium pyruvate. Transduction of human T cells was conducted as previously
described2, 4. HEK 293FT Cells were seeded at 5  106 per 10cm tissue culture plate 24h
before transfection. Cells were transfected with 7.5g pRRL-SIN-EF1-WPRE, 6g
pMDL, 6g pREV and 2.5g pVSVG plasmids using FugeneHD (Promega). Medium
was changed 6h after transfection and the viral supernatant was harvested at 48 post
transfection. Virus particles were concentrated 10-fold by ultracentrifugation for 2h at
27,000rpm with a Beckman Ti70 rotor (Beckman Coulter). Human T cells were purified
from peripheral blood mononuclear cells (PBMC) using EasySep human T cell
enrichment kit (StemCell Technologies Inc), and further activated for 24h with
CD3/CD28-coated magnetic beads (Life Technologies) before infection. Concentrated
lentivirus were incubated with the activated human T cells in the presence of 5g/mL
protamine sulfate and 50 IU/mL IL-2, and centrifuged at 1000rpm for 1h at 32C. The
surface expression of CAR on transduced T cells was determined by flow cytometry
after staining with APC labelled anti-human IgG antibody and APC labelled anti-mouse
IgG antibody.

Binding of antigen specific switches to target cells. Flow cytometry was performed
to evaluate Her2 or IGF1R binding of different FITC or Myc-based switches on breast
cancer cells (SKBR3, MDA MB453, MDA MB231 and MDA MB468). A fixed
concentration of 20nM switches were incubated with cells at 4C for 2h. After washes,
the cells labelled with FITC conjugates were analyzed directly on BD LSR II flow
cytometer (Becton Dickinson Immunocytometry System), and the cells labelled with Myc
fusions were further incubated with FITC labelled anti-Myc antibody, followed by flow
cytometry analysis.

In vitro cytotoxicity assay. Co-culture containing 1×104 target cells and PBMCs or
1×105 CAR-T cells with different concentrations of ADC, bsAb or antibody switches were
incubated at 37°C for 24h. To test the effects of E: T ratio, 1×104 target cells were
incubated with 100pM antibody switches with different numbers of the corresponding
CAR-T cells (0.01 - 10 ×105) at 37°C for 24h. Cytotoxic activity was determined by
measuring lactate dehydrogenase (LDH) levels in the cultured supernatant using the

S4
Cytotox-96 nonradioactive cytotoxicity assay kit (Promega). Percent cytotoxicity was
calculated by: % cytotoxicity = [(absorbance experimental – absorbance spontaneous
average)/(absorbance maximum killing average – absorbance spontaneous average)] ×
100.

CAR-T cell activation analysis. CAR-T cells were incubated with switch bound target
cells for 24h at 37°C. T cell activation was analyzed by flow cytometry with APC
conjugated anti-human CD69 and PerCP/Cy5.5 conjugated anti-human CD25. In
addition, the levels of cytokines (IL-2, IFN-γ and TNF-α) produced in the culture medium
were measured by human IL-2, IFN-γ and TNF-α test kits (Thermo Fisher Scientific). All
the tests were conducted in triplicate and results are shown as mean ± SD.

In vivo efficacy study. The efficacy studies targeting Her2 expressing breast cancers
were conducted in 8-week old female NSG mice. The tumor models including Her2 3+
(HCC1954), 2+ (MDA MB453) and 1+ (MDA MB435) were investigated for the in vivo
efficacy of switchable and conventional CAR-T cells, bsAbs and ADC. On day 0, 5  106
(HCC1954 or MDA MB453) or 2  106 MDA MB435 cells in 50% Matrigel were s.c.
implanted into the right flank. For sCAR-T cells groups, on day 8, anti-FITC CAR-T cells
(15 106 cells) were injected into the peritoneal cavity. Two days later, mice were
received i.v. injection of anti-Her2-FITC as treatment and anti-Her2 Fab as control at
0.5mg/kg every other day. For conventional CAR-T, on day 10, anti-Her2 CAR-T cells
(15 106 cells) were administrated by once i.v. injection. For bsAbs group, ex vivo-
expanded T cells (15 × 106 cells) were i.p. injected three times for Her2 1+ tumor model
on day 8, 14 and 20. On day 10, mice were received i.v. injection of PBS or bsAbs
(1mg/kg) every other day. For ADC group, once 5mg/kg dose of ADC was i.v. injected
on day 10. Animals were monitored and tumors were measured twice weekly using
caliper for an additonal 40 days.

Bioluminescence imaging. MDA MB435 tumor growth was also monitored by

Bioluminescent imaging (BLI). BLI was done using Xenogen IVIS imaging system and
the photons emitted from Luc-expressing cells within the animal body were quantified
using Living Image software (Xenogen). Briefly, mice bearing MDA MB435 Luc+ tumor
cells were injected intraperitoneally with D-luciferin (150mg/kg, 200L per mouse)
suspended in PBS and imaged under isoflurane anesthesia after 10min. A pseudocolor

S5
image representing light intensity (blue, least intense; red, most intense) was generated
using Living Image.

REFERENCES

1. Cao, Y., Axup, J. Y., Ma, J. S., Wang, R. E., Choi, S., Tardif, V., Lim, R. K., Pugh, H.
M., Lawson, B. R., Welzel, G., Kazane, S. A., Sun, Y., Tian, F., Srinagesh, S.,
Javahishvili, T., Schultz, P. G., and Kim, C. H. (2015) Multiformat T-cell-engaging
bispecific antibodies targeting human breast cancers, Angew Chem Int Ed Engl 54,
7022-7027.
2. Cao, Y., Rodgers, D. T., Du, J., Ahmad, I., Hampton, E. N., Ma, J. S., Mazagova, M.,
Choi, S. H., Yun, H. Y., Xiao, H., Yang, P., Luo, X., Lim, R. K., Pugh, H. M., Wang, F.,
Kazane, S. A., Wright, T. M., Kim, C. H., Schultz, P. G., and Young, T. S. (2016) Design
of Switchable Chimeric Antigen Receptor T Cells Targeting Breast Cancer, Angew
Chem Int Ed Engl 55, 7520-7524.
3. Axup, J. Y., Bajjuri, K. M., Ritland, M., Hutchins, B. M., Kim, C. H., Kazane, S. A.,
Halder, R., Forsyth, J. S., Santidrian, A. F., Stafin, K., Lu, Y., Tran, H., Seller, A. J.,
Biroc, S. L., Szydlik, A., Pinkstaff, J. K., Tian, F., Sinha, S. C., Felding-Habermann, B.,
Smider, V. V., and Schultz, P. G. (2012) Synthesis of site-specific antibody-drug
conjugates using unnatural amino acids, Proc Natl Acad Sci U S A 109, 16101-16106.
4. Ma, J. S., Kim, J. Y., Kazane, S. A., Choi, S. H., Yun, H. Y., Kim, M. S., Rodgers, D.
T., Pugh, H. M., Singer, O., Sun, S. B., Fonslow, B. R., Kochenderfer, J. N., Wright, T.
M., Schultz, P. G., Young, T. S., Kim, C. H., and Cao, Y. (2016) Versatile strategy for
controlling the specificity and activity of engineered T cells, Proc Natl Acad Sci U S A
113, E450-458.

S6
 Supplementary Figures

anti-Her2 ADC (Trastuzumab-auristatin F)

O O
H H
N N N O O
N N N O NH2
H
O O O O O

Auristatin F-linker
O N
O N
Acetate buffer pH 4.5
O
O

37C
1-4 days

anti-Her2 IgG anti-Her2 IgG-auristatin F conjugates

(A121X)

Figure S1. Site-specific conjugation of alkoxy-amine–derivatized auristatin to

anti-Her2 IgG with pAcF. The noncleavable auristatin derivatized with a terminal
alkoxy-amine is coupled by oxime ligation to antibodies through pAcF residues.

S7
 anti-Her2/anti-CD3 bsAb

anti-Her2 Fab anti-CD3 Fab

(LS202X) O
O (HK138X)
O
H
O O N3 O O O
H 2N O O O N O O NH2
Acetate buffer pH 4.5 H
H

37°C
N3-PEG3-ONH2 16-24 hr
O
H
O O O
N O O O N O O N
O O N3 H
H

PBS pH 7.4
37°C
1-3 days

O
H
O O O
O N O O N
H
N O O H
O O N
N N

anti-Her2/anti-CD3 bsAb

Figure S2. General scheme for the generation of bispecific antibodies: the
antibodies with pAcF were first selectively coupled with bifunctional ethylene
glycol linkers, and subsequently conjugated to each other by “Click” reaction.

S8
 anti-Her2/FITC Switch
O

anti-Her2 Fab
(LS202X)

O O N3
H 2N O O

N3-PEG3-ONH2 Oxime ligation

O O N3
N O O

O O OH

CO2H
“Click” chemistry
H
H H
HN N O N O
O O
H
S O

FITC-linker

HO O O

N N HO2C
O N
N O
3 H H
O N N NH
O
O 3 S

anti-Her2/FITC

Figure S3. General scheme to generate site-specific FITC-antibody conjugates.

Mutant antibodies incorporated with pAcF were conjugated with bifunctional
ethylene glycol linkers, followed with BCN-PEG4-FITC by “Click” reaction.

S9
 anti-Her2/FITC

anti-Her2 Fab MW: 47849

(LS202X)

anti-Her2–azide
MW: 48066

anti-Her2/FITC
MW: 48823

Figure S4. Mass spectrometric analysis of anti-Her2 (LS202X) mutant, anti-

Her2-azide, and the final anti-Her2/FITC product.

S10
 MW kDa
245
180
135
100
75
63
48
35
25
20
17

11

Figure S5. SDS-PAGE analysis of anti-Her2/FITC switch before and after

conjugation under non-reducing conditions.

S11
 SKBR3 MDA MB453 MDA MB231 MDA MB468
(Her2 3+) (Her2 2+) (Her2 1+) (Her2 0)

anti-Her2/FITC
anti-Her2 Fab

Figure S6. In vitro binding activity of anti-Her2/FITC switches. Different Her2

expressing breast cancer cells SKBR3 (Her2 3+), MDA MB453 (Her2 2+), MDA
MB231 (Her2 1+) and MDA MB468 (Her2 0) were incubated with 100nM anti-
Her2/FITC and validated by flow cytometry.

S12
 SKBR3 (Her2 3+) EC50 (nM)
Mean Fluorescence Intensity Control N.D.
500 ADC 0.95±0.11
bsAb 0.79±0.14
400
anti-Her2/FITC 1.16±0.24
300

200

100

0
10 -3 10 -2 10 -1 10 0 10 1 10 2 10 3
Concentration (nM)

MDA MB468 (Her2 0) EC50 (nM)

Mean Fluorescence Intensity

Control N.D.
500 ADC N.D.
bsAb N.D.
400 N.D.
anti-Her2/FITC
300

200

100

0
10 -3 10 -2 10 -1 10 0 10 1 10 2 10 3
Concentration (nM)

Figure S7. Binding of ADC, bsAb and anti-Her2/FITC to SKBR3 (Her2 3+) and
MDA MB468 (Her2 0) cancer cells. Cells were incubated with indicated switches
and detected with Alexa Fluor647 conjugated anti-human kappa chain antibody.
The EC50 value was calculated with Graphpad Prism software.

S13
 SKBR3 (Her2 3+) HCC1954 (Her2 3+) MAD MB435/Her2 (Her2 3+)
100 100 100
Cytotoxicity (%)

Cytotoxicity (%)

Cytotoxicity (%)
75 75 75

50 50 50

25 25 25

0 0 0
0.1 1 10 0.1 1 10 0.1 1 10
E: T ratio E: T ratio E: T ratio

MDA MB453 (Her2 2+) MDA MB361 (Her2 2+)

100 100

Cytotoxicity (%)
Cytotoxicity (%)

75 75

50 50

25 25

0 0
0.1 1 10 0.1 1 10
E: T ratio E: T ratio

BT20 (Her2 1+) MDA MB231 (Her2 1+) MAD MB435 (Her2 1+)
100 100 100
Cytotoxicity (%)

Cytotoxicity (%)
Cytotoxicity (%)

75 75 75

50 50 50

25 25 25

0 0 0
0.1 1 10 0.1 1 10 0.1 1 10
E: T ratio E: T ratio E: T ratio
MDA MB468 (Her2 0)
100
CAR-T
4D5-BBZ
Cytotoxicity (%)

75 100
Control CAR-T
FITC-E2-BBZ
Cytotoxicity (%)

50 75 sCAR-T + aHer2-FITC
FITC-E2-BBZ

25
50
0
0.1 25 1 10
E: T ratio
0
Figure S8.
0.1 In vitro1 comparison
10 of sCAR-T cells with conventional anti-Her2 CAR-
E: T ratio
T cells on different Her2 expressing cancer cells. The cytotoxicity of anti-FITC
CAR-T cells in the presence of 100pM anti-Her2/FITC was assessed at the
indicated E: T ratios. In the presence of switch, anti-FITC CAR-T showed the
comparable activity as anti-Her2 CAR-T against different Her2 expressing cancer
cells.

S14
 MDA MB435 (Her2 1+) IC50 (pM)
100
0h 59.4±2.5
Cytotoxicity (%)

75 24 h 65.2±2.7
48 h 69.4±1.7
50 72 h 122.4±3.8

25

0
10 -2 10 0 10 2 10 4 10 6
Concentration (pM)

Figure S9. Functional stability analysis of the anti-Her2/FITC by cytotoxicity on

MDA MB435 (Her2 1+) breast cancer cells. The switches were incubated in
human plasma at 37°C for up to 72h before analysis.

S15
 HCC1954 (Her2 3+)
Tumor volume (mm3)
1500
1500
Tumor volume (mm3 )

CAR-T
Legend
CAR-T Control
Legend sCAR-T
CAR-T
1000 sCAR-T
Legend
1000

500
500

0
0 0 10 20 30 40 50
0 10 20 30 40 50 Days
Days

Figure S10. In vivo efficacy comparison of CAR-T approaches (switchable and

conventional) in Her2 3+ human breast tumor xenografts. Female NSG mice
were s.c. inoculated with HCC1954 tumors in the right flank. After 8 days, the
mice were treated as indicated: group 1 (anti-Her2 CAR-T), group 2 (anti-Her2
Fab + anti-FITC CAR-T), group 3 (anti-Her2/FITC + anti-FITC CAR-T). Mean
tumor volume was calculated by W L H as measured by calipers.

S16
 MDA MB453 (Her2 2+)

Tumor volume (mm3) 500

1500
Tumor volume (mm3 )

CAR-T
Legend
400 CAR-T
Control
Legend sCAR-T
CAR-T sCAR-T
Legend
1000 300
200
500 100
0
0 0 10 20 30 40 50
0 10 20 30 40 50 Days
Days

Figure S11. In vivo efficacy comparison of CAR-T approaches (switchable and

conventional) in Her2 2+ human breast tumor xenografts. Female NSG mice
were s.c. inoculated with MDA MB453 tumors in the right flank. After 8 days, the
mice were treated as indicated: group 1 (anti-Her2 CAR-T), group 2 (anti-Her2
Fab + anti-FITC CAR-T), group 3 (anti-Her2/FITC + anti-FITC CAR-T). Mean
tumor volume was calculated by W L H as measured by calipers.

S17
 Days
50 40 30 20 10 0
MDA MB435 (Her2 1+)
Saline
Saline + T cells

8 ADC
ADC
T cells
(total counts  109)
bsAb
Legend
Luminescence

CAR-T
Legend
6 CAR-T cells
CAR-T
Control sCAR-T
Legend

sCAR-T
Legend T cells
4

0
0 10 20 30 40 50
Days

Figure S12. In vivo comparison of CAR-T approaches (switchable and

conventional), bsAb and ADC in Her2 1+ MDA MB435 breast tumor xenografts.
Tumor burden was monitored by weekly bioluminescence imaging (BLI), and
quantified by the radiance detected in the region of interest (ROI). Results are
derived from six mice per group, and error bars represent SD.

S18
 HCC1954 (Her2 3+)

CAR-T
Tumor volume (mm3)
30

25
1500
Tumor volume (mm3 )

CAR-T
Legend
20 Control
Legend sCAR-T
CAR-T sCAR-T
Legend
1000
15
500 0 10 20 30 40 50
Days

0
0 10 20 30 40 50
Figure S13. Body weight changes of mice during treatment of CAR-T and sCAR-
Days
T against HCC1954 (Her2 3+) xenograft models. No significant changes in body
weight or other treatmentrelated toxicities were observed during the studies.

S19
 MDA MB453 (Her2 2+)

CAR-T
Tumor volume (mm3)
30

25
1500
Tumor volume (mm3 )

CAR-T
Legend
Control
Legend sCAR-T
CAR-T 20
1000 sCAR-T
Legend

15
500 0 10 20 30 40 50
Days
0
0 10 20 5030 40
Days
Figure S14. Body weight changes of mice during treatment of CAR-T and sCAR-
T against MDA MB453 (Her2 2+) xenograft models. No significant changes in
body weight or other treatmentrelated toxicities were observed during the
studies.

S20
 MDA MB435 (Her2 1+)
T cells

Tumor volume (mm3)

30
CAR-T

25

Days
20
50 40 30 20 10 0
15 0

Tumor volume (mm3 )

0 10 20 30 40 50
Days
500
Saline
Saline + T cells

ADC
ADC
1000
bsAb
Legend

sCAR-T
Legend
CAR-T 1500
Control sCAR-T
Legend

sCAR-T
Legend T cells 2000

Figure S15. Body weight changes of mice during treatment of different

immunotherapeutics against MDA MB435 (Her2 1+) xenograft models. No
significant changes in body weight or other treatmentrelated toxicities were
observed during the studies.

S21
 anti-Her2/Myc
Thrombin site

6His 4D5 scFv Linker Myc

anti-IGF1R/Myc
Thrombin site

6His Myc Linker IGF1R Adnectin

anti-Her2/Myc/anti-IGF1R
Thrombin site

6His 4D5 scFv Linker Myc Linker IGF1R Adnectin

Figure S16. Schematic diagram of Myc-based recombinant fusions. The anti-

Her2/Myc construct contained 4D5 scFv, 218 peptide linker and Myc; the anti-
IGF1R/Myc construct consisted of Myc, 218 peptide linker and anti-IGF1R
Adnectin; the bivalent fusion was composed of N-terminal 4D5 scFv and C-
terminal Adnectin, and Myc tag grafted between them separated with 218 peptide
linkers. All the constructs were placed on the C-terminal of histidine affinity tag,
followed by a thrombin cleavage site.

S22
 ani-Her2/Myc

kDa kDa
180 180
135 135
100 100
75 75
65 65
45 45

35 35

25 25

18 18

SDS-PAGE Western Blot

Figure S17. SDS-PAGE and western blot analysis of anti-Her2/Myc before and
after thrombin cleavage under nonreducing conditions. Western blot was probed
anti-Myc mouse monoclonal antibody, followed by chemiluminescence using
HRP labelled goat anti-mouse IgG secondary antibody.

S23
 anti-IGF1R/Myc

kDa
kDa
150
100
116
70
66
50
45
35 35

25 25
20
18
14 15

SDS-PAGE Western Blot

Figure S18. SDS-PAGE and western blot analysis of anti-IGF1R/Myc before and
after thrombin cleavage under nonreducing conditions. Western blot was probed
anti-Myc mouse monoclonal antibody, followed by chemiluminescence using
HRP labelled goat anti-mouse IgG secondary antibody.

S24
 ani-Her2/Myc/anti-IGF1R

kDa kDa

150
116 100
70
66
50
45
35 35
25 25
20
18
14 15

SDS-PAGE Western Blot

Figure S19. SDS-PAGE and western blot analysis of anti-Her2/Myc/anti-IGF1R
before and after thrombin cleavage under nonreducing conditions. Western blot
was probed anti-Myc mouse monoclonal antibody, followed by
chemiluminescence using HRP labelled goat anti-mouse IgG secondary
antibody.

S25
 SKBR3 MDA MB453 MDA MB231 MDA MB468
Her2 3+ Her2 2+ Her2 1+ Her2 0

Her2

PE

IGF1R

APC

Figure S20. FACS analysis of Her2 and IGF1R expression on different breast
cancer cells. Her2 and IGF1R expression was determined with PE-conjugated
anti-Her2 antibody and APC conjugated anti-IGF1R antibody.

S26
 SKBR3 MDA MB453 MDA MB231 MDA MB468
(Her2 3+) (Her2 2+) (Her2 1+) (Her2 0+)
Normalized
Control
anti-Her2/Myc
anti-IGF1R/Myc
anti-Her2/Myc/anti-IGF1R

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

FITC

Figure S21. Flow cytometry analysis of Myc-based recombinant fusions binding

to different Her2/IGF1R expressing breast cancer cells. Cells were consecutively
labeled with different switch antibodies (25 nm) and secondary FITC conjugated
anti-Myc antibody (R&D).

S27
 sCAR-T-Her2 +
Control sCAR-T-Her2 sCAR-T-IGF1R sCAR-T-IGF1R Bi-sCAR-T

SKBR3
Her2 3+

MDA MB453
Her2 2+

MDA MB231

CD25
Her2 1+

MDA MB468
Her2 0

CD69

Figure S22. sCAR-T activation by various Myc-based recombinant fusion

switches against different Her2/IGF1R expressing cancer cells. Myc-based
sCAR-T cells were cocultured with SKBR3, MDA MB453, MDA MB231 or MDA
MB468 cells at E: T = 10: 1 with 100 pM of the corresponding switches for 24h. T
cell activation was evaluated by flow cytometry with staining for CD69 and CD25
redirected sCAR-T cells.

S28
 SKBR3 (Her2 3+) MDA MB453 (Her2 2+)
*** 200 ****
400 ns ****
*** ****
** 150 ****
IFN- (pg/mL)

IFN- (pg/mL)
300

100
200

100 50

0 0

tin

tin
tin

tin
tin

tin
e

yc

yc
l

l
ic

ic
M

M
ec

ec
ec

ec
ec

ec
h

h
5/

5/
Ve

Ve
dn

dn
dn

dn
dn

dn
4D

4D
/A

/A
/A

/A
/A

/A
yc

yc
yc

yc
yc

yc
M

M
M

M
M

M
5/

5/
+

+
4D

4D
yc

yc
M

M
5/

5/
4D

4D
MDA MB231 (Her2 1+) MDA MB468 (Her2 0)
125 *** 125 ns
*** ns
100 *** 100
*** ns
IFN- (pg/mL)

IFN- (pg/mL)

ns
75 75

50 50

25 25

0 0
tin

tin
tin

tin
tin

tin
e

yc

yc
l

l
ic

ic
M

M
ec

ec
ec

ec
ec

ec
h

h
5/

5/
Ve

Ve
dn

dn
dn

dn
dn

dn
4D

4D
/A

/A
/A

/A
/A

/A
yc

yc
yc

yc
yc

yc
M

M
M

M
M

M
5/

5/
+

+
4D

4D
yc

yc
M

M
5/

5/
4D

4D

Figure S23. sCAR-T activation by various Myc-based recombinant fusion

switches against different Her2/IGF1R expressing cancer cells. Myc-based
sCAR-T cells were cocultured with SKBR3, MDA MB453, MDA MB231 or MDA
MB468 cells at E: T = 10: 1 with 100 pM of the corresponding switches for 24h.
IFN-γ levels from the incubation medium were measured by ELISA kit. Error bars
represent standard deviation of duplicate samples.

S29
 SKBR3 (Her2 3+) MDA MB453 (Her2 2+)
**** 800 ****
1500 **** ****
TNF- (pg/mL)

TNF- (pg/mL)
**** 600 ****
**** ****
1000
400

500 200

0 0

tin
tin

tin

tin
e

yc

tin

tin
e

yc
l

l
ic

ic
M

M
ec
ec

ec

ec
ec

ec
h

h
5/

5/
Ve

Ve
dn
dn

dn

dn
dn

dn
4D

4D
/A
/A

/A

/A
/A

/A
yc
yc

yc

yc
yc

yc
M
M

M
M

M
5/

5/
+

+
4D

4D
yc

yc
M

M
5/

5/
4D

4D
MDA MB231 (Her2 1+) MDA MB468 (Her2 0)
**** ns
300 **** TNF- (pg/mL) 300 ns
TNF- (pg/mL)

****
ns
****
200 200 ns

100 100

0 0

tin
tin

tin
e

yc
tin
tin

tin
e

yc

l
l

ic
ic

ec
ec

ec
M

ec
ec

ec

h
h

5/
5/

Ve

dn
dn

dn
Ve

dn
dn

dn

4D
4D

/A
/A

/A
/A
/A

/A

yc
yc

yc
yc
yc

yc

M
M

M
M
M

5/
5/

+
+

4D
4D

yc
yc

M
M

5/
5/

4D
4D

Figure S24. sCAR-T activation by various Myc-based recombinant fusion

switches against different Her2/IGF1R expressing cancer cells. Myc-based
sCAR-T cells were cocultured with SKBR3, MDA MB453, MDA MB231 or MDA
MB468 cells at E: T = 10: 1 with 100 pM of the corresponding switches for 24h.
TNF-α levels from the incubation medium were measured by ELISA kit. Error
bars represent standard deviation of duplicate samples.

S30
 Supplementary Tables

Table S1. Comparative IC50 values and maximal killing efficacy of Her2 targeted therapeutics including ADC, bsAbs and
sCAR-T against various human breast cancer cell lines
IC50 (pM)/Maximal Killing (%)
Cell line Her2 level
ADC bsAb sCAR-T
SKBR3 3+ 16.11.8/61.563.7 2.00.4/60.45.2 0.90.1/69.93.8
HCC1954 3+ 105.24.6/27.24.2 15.22.3/45.54.7 8.41.1/67.75.2
MDA MB435/Her2 3+ 51.53.7/47.42.8 2.90.7/65.13.7 0.40.1/85.83.3
MDA MB361 2+ 244.913.4/26.63.5 31.33.0/49.63.1 8.61.1/77.03.6
MDA MB435 2+ 12.01.8/50.62.2 3.50.5/60.24.1 1.10.2/75.75.8
BT20 1+ 452.715.9/12.41.8 88.53.1/37.12.3 26.42.9/56.63.8
MDA MB231 1+ N.D. 59.42.0/49.61.8 32.71.2/57.73.9
MDA MB435 1+ N.D. 101.99.4/43.75.2 42.72.5/47.53.5
MDA MB468 0 N.D. N.D. N.D.
Abbreviations: IC50, half-maximal inhibitory concentration; N.D., not determined.
Data shown are a mean of duplicate samples ± SD.

S31
Table S2. Comparative IC50 values and maximal killing efficacy of monospecific and bispecific sCAR-T against various
human breast cancer cell lines
IC50 (pM)/Maximal Killing (%)
Cell line Her2 level
sCAR-T-Her2+
sCAR-T-Her2 sCAR-T-IGF1R Bi-sCAR-T
sCAR-T-IGF1R
SKBR3 3+ 58.55.8/85.07.7 N.D./21.63.6 55.20.1/78.25.8 21.21.9/82.52.9
MDA MB435 2+ 181.47.2/87.35.2 N.D./48.76.3 196.63.2/84.15.3 21.63.3/81.67.1
MDA MB231 1+ 880.26.6/45.27.1 N.D./43.82.8 117.94.2/35.04.9 112.73.8/65.00.9
MDA MB468 0 N.D. N.D. N.D. N.D.
Abbreviations: IC50, half-maximal inhibitory concentration; N.D., not determined.
Data shown are a mean of duplicate samples ± SD.

S32