Cancer Cell, Volume 26

Supplemental Information

Postsurgical Adjuvant Tumor Therapy

by Combining Anti-Angiopoietin-2 and Metronomic

Chemotherapy Limits Metastatic Growth

Kshitij Srivastava, Junhao Hu, Claudia Korn, Soniya Savant, Martin Teichert, Stephanie S. Kapel, Manfred Jugold, Eva Besemfelder, Markus Thomas, Manolis Pasparakis, and Hellmut G. Augustin

Pre-surgery

0

5

10

15M

icro

-met

s co

unts

IgG

Ctr

Ang

2 A

b

***IgG Ctr Ang2 Ab

A

2 mm

B

Dor

sal

Vent

ral

Post-surgery

Figure S1, related to Figure 1 (A) Pre and post-surgery bioluminescence images of mice bearing 4T1-luciferase tagged orthotopic mammary tumors. Post-surgery bioluminescence imaging confirmed the absence of detectable metastatic progression prior to primary tumor resection and absence of residual primary tumor after mastectomy. Mice meeting both criteria were randomized into treatment groups (B) Representative low magnification images of HE stained lung sections from IgG and anti-Ang2 antibody treated mice bearing lung metastases originating from orthotopic 4T1 breast cancer. Black arrows mark macro-metastasis and arrowheads mark micro-metastatic lesions (magnification shown in the insets). (C) Quantitation of frequency of micro-metastatic lesions in lungs (n=5; values are mean±SD; ***p≤0.001; the experiment was reproduced three times with similar results. The figure shows representa-tive images from one of the experiment.

C

1

Pre-su

rgery

Post-s

urgery

0

500

1000

1500

2000

Ser

um V

EG

F co

ncen

tratio

n (p

g/m

l) ***

Figure S2, related to Figure 3. Effect of primary tumor resection on levels of circulating VEGF in Lewis lung carcinoma tumor bear-ing mice. Serum concentrations of VEGF were analyzed before and 24 hr after resection of the primary tumor (n=10; values are mean±SD, ***p≤0.001).

2

Paclita

xel

Gemcit

abine

0.0

0.5

1.0

1.5

2.0 **

Ly6

C-h

i cel

ls

(% o

f tot

al c

ells

in lu

ng m

ets)

VEGF Ab Ang2 AbIgG Ctr

100 µm

HIF

1α

100 µm

H&

E

C

D VEGF Ab Ang2 AbIgG Ctr

Figure S3, related to Figure 4. (A-B) Comparative analysis of effect of metronomic Gemcitabine and metronomic Paclitaxel on the recruitment of myeloid derived CD11bGr1 and Cd11bGr1Ly6Chi cells (n=5, values are mean±SD, **p≤0.01). The experiment was performed twice with similar results. The data from one representative experiment are shown in the figure. (C-D) Representative immunohistochemical images of HE and HIF1-alpha stained lung sections. Samples from different treatment groups (IgG, anti-Ang2 antibody and anti-VEGF antibody) were stained to visualize hypoxic areas in lung metastases originating from Lewis lung carcinoma (scale bar: 100µm).

Paclita

xel

Gemcit

abine

**

CD

11b+

Gr1

+ ce

lls

(% o

f tot

al c

ells

in lu

ng m

ets)

0

10

20

30

40

A B

3

IgG

Ang2 A

b0.0

0.5

1.0

1.5

2.0

Tie2

+ CC

R2- fr

actio

nof

mac

roph

ages

(%)

95.8%

0 200 400 600 800 1KFSC-A

0

200

400

600

800

1K

SSC-

A

94.6%

0 200 400 600 800 1KSSC-W

0

200

400

600

800

1K

SSC-

H

3.05%

0 200 400 600 800 1KFSC-A

100

101

102

103

104

CD45

2.51%

100

101

102

103

104

F4/80

100

101

102

103

104

CD11

b

3.87% 1.37%

2.05%92.7%

100

101

102

103

104

Tie2

100

101

102

103

104

CCR2

50.1%

0 200 400 600 800 1KFSC-A

0

200

400

600

800

1K

SSC-

A

43.1% 0.36%

0.152%56.4%

100

101

102

103

104

Tie2

100

101

102

103

104

CCR2

11.7%

100

101

102

103

104

F4/80

100

101

102

103

104

CD11

b

83.5%

0 200 400 600 800 1KSSC-W

0

200

400

600

800

1K

SSC-

H

8.79%

0 200 400 600 800 1KFSC-A

100

101

102

103

104

CD45

IgG

Ang2 Ab

A B

C

0

100

200

300

400

Ser

um C

CL2

(pg/

ml)

IgG C

tr

Ang2 A

b

***

Grou

pIg

GAn

tiAng

2Fo

ldp-

Sign

ifi-

12

31

23

chan

geva

lue

canc

e

TREM

-165

44.57

822

611.4

8511

404.6

175

1966

.0515

1002

.9995

643.8

9711

.230.0

307

*

CXCL

1022

86.30

537

00.78

898

23.77

414

89.08

4589

1.19713

8.989

56.2

80.0

660

ns

TIMP

-127

433.6

305

5333

2.905

5315

0.501

1722

9.100

525

28.08

241

80.09

555.5

90.0

100

*

CCL2

4037

3.363

550

327.8

425

4093

3.69

2365

9.449

3848

.074

3992

.4015

4.18

0.005

2**

CXCL

154

76.89

1515

521.7

515

6683

.9765

2723

.9035

2193

.1425

3213

.143

3.40

0.054

9ns

CCL1

1710

.018

496.6

93769.4

99457.3

26728.5

225

-28.0

375

2.57

0.115

3ns

IL-1

b36

58.70

738

69.76

5515

87.61

912

94.10

314

31.60

3595

7.791

2.47

0.035

5*

IL-1

769

48.28

124

7.029

527

41.09

1521

87.99

220

14.74

3552

8.897

2.10

0.219

9ns

CCL1

712

090.9

925

4257

.381

7202

.4025

6394

.398

3348

.7555

2123

.4055

1.98

0.105

2ns

CXCL

212

29.96

910

78.70

5589

2.96110

22.9911

71.27

0529

6.887

1.29

0.228

0ns

Il-6

1200

.573

23.04

25798.8

575

281.3

42717.8

02632.7

735

1.24

0.371

5ns

GM-C

SF27

720.3

935

1689

4.797

520

898.5

1624

911.6

1612

206.0

495

1839

5.521

1.18

0.264

5ns

CXCL

913

492.8

125

6828

.827

3415

7.293

1992

0.065

513

986.5

105

1922

1.71

1.03

0.480

0ns

BLC

9100

.224

3675

.001

8174

.123

1206

4.205

3376

.4645

5077

.079

1.02

0.482

8ns

Cal3

5639

1.914

554

919.8

0256

426.0

1955

190.9

915

5563

7.317

5497

9.213

51.01

0.147

0ns

Cal1

5642

8.529

554

779.0

1456

566.5

4855

733.0

4355

798.0

5855

340.2

9251.0

10.3

189

ns

Cal2

5621

1.933

5601

4.161

556

311.3

175

5645

3.274

556

720.3

4256

205.6

731.0

00.0

894

ns

sICAM

-155

969.8

4755

590.1

3155

842.6

385

5604

8.613

556

414.0

805

5599

1.181

0.99

0.056

5ns

CCL3

1941

.0945

858.7

085

4197

.5065

2704

.9725

2037

.5025

2489

.703

0.97

0.470

8ns

C5/C

5a42

901.1

9144

271.5

505

4463

6.647

553

213.3

7444

237.2

8344

674.7

190.9

30.1

556

ns

CXCL

1173

0.64211

7.62689

7.22382

8.41610

03.95

926

6.272

50.8

30.3

678

ns

CCL1

272

44.23

3545

39.87

935

19.49

1559

25.61

7584

51.06

2542

76.58

10.8

20.2

674

ns

TNFa

3108

3.552

522

599.3

629

385.5

0933

777.9

838

208.7

6440

058.4

5750.7

40.0

195

*

IL-4

1343

.5815

69.21

551047

.4925

1452

.544

1444

.9225

524.4

495

0.72

0.275

6ns

IL-1

a28

435.922

166.9

875

8046

.703

2733

3.458

2162

7.957

3274

0.32

0.72

0.161

8ns

IL-2

716

967.3

435

2145

.5905

59.79

78859

.938

6041

.9965

1197

6.633

50.71

0.334

9ns

CCL1

156

67.41

9517

73.2424

24.82

457

72.5849

93.532

36.44

60.7

00.1

930

ns

IL-3

926.4

46435.6

91708.3

691479

.4705

1174

.391

690.2

020.6

20.0

956

ns

IL-1

ra15

863.3

633

889.8

615

9831

.068

4168

4.397

2528

0.547

532

689.9

4850.6

00.0

986

ns

M-CS

F46

52.81

5531

93.90

337

48.07

541

02.86

6520

31.50

614

459.5

10.5

60.2

406

ns

IL-1

014

97.64

9511

6.14191

4.94516

49.92

720

17.09

590

0.941

50.5

50.1

298

ns

IL-2

4162

.7305

2830

.7035

1611

.4705

7132

.1565

4073

.776

5129

.9055

0.53

0.045

3*

IFN-

gam

ma

1719

9.469

7784

.4465

7553

.429

1817

2.583

518

559.1

1934

306.4

6250.4

60.0

535

ns

G-CS

F31

1.19275

.861532

7.774

511

49.66

7552

8.806

538

.9595

0.42

0.185

2ns

IL-1

2 p70

950.4

41278

.262

945.1

295

2294

.222

2492

.109

3060

.204

0.40

0.001

8**

IL-2

358

38.52

3534

41.60

585

7.21770

59.91

8510

354.2

5293

04.41

40.3

80.0

167

*

IL-1

615

144.1

375

1965

6.642

527

63.65

9550

099.4

1839

995.3

685

4690

0.523

50.27

0.002

4**

CCL4

4704

.722

5297

.558

4247

.419

1982

0.674

516

549.2

0222

862.4

9650.2

40.0

006

***

CCL5

1577

2.665

543

91.80

3-1

61.64

5536

291.4

5531

525.7

855

2161

1.755

0.22

0.011

3*

IL-1

387

13.41

635

84.26

0518

17.30

9520

304.7

345

2479

2.158

2677

8.595

50.20

0.001

2**

IL-7

1804

.077

577.9

975

1072

.758

2302

1.117

580

24.37

222

227.7

2850.0

60.0

136

*

CXCL

1234

8.969-1.12

15-175

.4105

1596

.2545

360.8

5872.7

140.0

60.0

428

*

IL-5

-33.8

76-141

.7045

216.0

365

630.1

175

486.8

075

-56.9

945

0.04

0.110

5ns

D

<1000 1000-2000 2000-4000 4000-8000 8000-16000 16000-32000 <32000

Signal intensity

0 50K 100K 150K 200K 250K

FSC-A

0

50K

100K

150K

200K

250K

SSC

-A

88.6

0 50K 100K 150K 200K 250K

SSC-W

0

50K

100K

150K

200K

250K

SSC

-H

96.3

0 50K 100K 150K 200K 250K

FSC-A

0102

103

104

105

CD45

2.72

TREM

1

0 102 103 104 105

CD11b

0102

103

104

105

42.1

E

0 50K 100K 150K 200K 250K

SSC-W

0

50K

100K

150K

200K

250K

SSC

-H

93.6

TREM

1

0 102 103 104 105CD31

0102

103

104

1050.0569 3.35e-3

24.275.80 50K 100K 150K 200K 250K

FSC-A

089.5

50K

100K

150K

200K

250K

SSC

-A

0 50K 100K 150K 200K 250K

FSC-A

0102

103

104

5

CD45

-Pdp

n-Ly

ve-

7.13

10

4

Figure S4, related to Figure 5. (A-B) FACS analysis of CCR-negative Tie-2-positive macrophages in the 4T1 model. (A) Repre-sentative FACS plots from flow cytometric analyses of lung metastases of mice treated in the adjuvant settings with IgG or anti-Ang2 antibody after resection of orthotopic mammary primary tumor (4T1) for the identification and quantification of different macrophage subsets recruited to the metastatic niche. (B) Quantitation of CCR2-negative Tie2-positive macrophages as fraction of total macrophages in lung metastases of mice treated in the adjuvant setting with IgG or anti-Ang2 antibody after resection of orthotopic mammary primary tumor (4T1) (n=5 mice). The experiment was performed twice with similar results. The data from one representative experiment are shown in the figure. (C-D) Evaluation of the effect of Ang2 neutralization on local and systemic concentration of CCL2. (C) Effect of treatment in adjuvant setting with IgG or Ang2 Ab on serum protein levels of CCL2 as analysed by ELISA (n=4 mice) (D) Heatmap showing comparative fold change of cytokines in protein lysates from lung metasta-ses following IgG or Ang2 Ab treatment of mice (LLC tumor model) as analysed by Mouse Cytokine Antibody Array. CCL2 has been highlighted in red font. Quantification was based on signal intensity measured on Gray levels using FIJI. (E) FACS analysis of TREM-1 presentation in stromal compartment of lung metastases. (E-top panel) Representative FACS plots from flow cytomet-ric analyses of lung metastases three weeks after resection of sub-cutaneous Lewis lung carcinoma tumor for identification of TREM1 presentation on myeloid cells recruited to metastatic niche. (E-bottom panel) Representative FACS plots from flow cytometric analyses of lung metastases three weeks after resection of sub-cutaneous Lewis lung carcinoma tumor for identifica-tion of TREM1 presentation on endothelial cells in metastatic niche. (F) Evaluation of specificity and efficiency of Mx1Cre mediated CCR2 depletion. (F-Top panel) Representative plot of FACS analysis of CCR2 depletion in myeloid cells from spleen of PolyIC injected Mx1Cre+ or Mx1Cre- mice. CD45+ cells were gated for CD11b, F4/80 and CCR2. (F-middle panel) Representative plot of FACS analysis of CCR2 depletion in myeloid cells from lungs of PolyIC injected Mx1Cre+ or Mx1Cre- mice. CD45+ cells were gated for Cd11b, F4/80 and CCR2. (F-bottom panel) Representative FACS plot analysis of CCR2 depletion in endothelial cells from lungs of Poly IC injected Mx1Cre+ or Mx1Cre- mice. CD45- cells were gated for CD31 and CCR2. Mx1Cre induction by polyIC specifically depleted CCR2 expression in myeloid cells.Values are mean±SD; ***p≤0.001, ns=non-significant.

0 50K 100K 150K 200K 250KFSC-A

0

102

103

104

1051.71

0 50K 100K 150K 200K 250KFSC-A

0

102

103

104

105 14.3

0 50K 100K 150K 200K 250KFSC-A

0

102

103

104

1052.22

0 50K 100K 150K 200K 250KFSC-A

0

102

103

104

105 9.59

0 102 103 104 105

0

102

103

104

105 4.1

0 50K 100K 150K 200K 250K

0

102

103

104

10566.3

0 102 103 104 105

0

102

103

104

105 2.91

0 50K 100K 150K 200K 250KFSC-A

0

102

103

104

10517.8

0 102 103 104 105

0

102

103

104

105 1.37

0 50K 100K 150K 200K 250K

0

102

103

104

105

47.8

0 102 103 104 105

0

102

103

104

105 1.86

0 50K 100K 150K 200K 250K

0

102

103

104

105

7.76

F4/80 FSC-A F4/80 FSC-A

CD

11b

CC

R2

CD

11b

CC

R2

CD

31

CC

R2

CD

31

CC

R2

Ccr2fl/fl::Mx1Cre- Ccr2fl/fl::Mx1Cre+

CD

11b

CC

R2

CD

11b

CC

R2

Lung

mac

roph

age

(gat

ed fo

r CD

45+)

Spl

een

mac

roph

age

(gat

ed fo

r CD

45+)

Lung

EC

(gat

ed fo

r CD

45-)

F

F4/80 FSC-A F4/80

5

p-IκBα (Ser32)

IκB

Actin

Ctr Ang2

Ang2 +

low TNFα

High TNFα

Low TNFα

A

H

unst

imul

ated

high

TN

Fαlo

w T

NFα

Ang

2A

ng2+

low

TN

Fα

50µm

PhalloidinDAPI p65

Nor

mal

ized

Vca

m-1

expr

essi

on

Contro

lIgG

Ang2 A

b0

2

4

6

8 **7

5

3

1

B

Figure S5, related to Figure 6. (A) qRT-PCR analysis of Vcam-1 expression in tissue lysates from lung metastases following IgG or anti-Ang2 antibody treatment. Mice were treated for 3 weeks post-resection of the primary tumor (Lewis lung carcinoma model). Whole lung lysates from mice sacrificed immediately post-resection of the primary tumor were included as control group to evaluate basal levels of expression (n=5 mice). The experiment was reproduced three times with similar results. The data from one representative experiment are shown in the figure. (B) Patient dataset (GSE3494) analyses to evaluate correlation of Ang2 expression with expression of adhesion molecules in context of cancer progression. (A) Correlation analysis between ANG2 and ICAM1 expression profiles in non-metastatic breast cancer samples (Grade 1) or advanced metastatic breast cancer samples (Grade 2&3) using the GSE3494 data set (n=67).(C) Representative Western blot analysis of ICAM-1 after stimulation of HUVEC with rhAng2 in the presence of a cell permeable STAT3 inhibitor peptide or inactive control peptide. Cells were pretreated with respective peptide for 1 hr before stimulation (n=3 replicates). (D) Quantification of ICAM1 protein expression after stimulation of HUVEC with rhAng2 in the presence of a STAT3 inhibitor peptide or inactive control peptide (n=3 replicates). (E) Representative Western blot analysis of ICAM-1 upon stimulation of HUVEC with rhAng2 and/or low concentration of rhTNFα compared to high rhTNFα (classical inducer of NF-kB signaling). (F) Quantification of ICAM1 protein expression after stimula-tion of HUVEC with rhAng2 and/or low concentrations of rhTNFα compared to high rhTNFα (n=3 replicates). (G) Cell lysates obtained from stimulated HUVEC were blotted to detect phospho-IkBα and total IkB-α. Prestarved HUVEC were stimulated by PBS (Ctrl), Ang2 (400ng/ml), low TNFα (1ng/ml) , high TNFα (10ng/ml) or a combination of Ang2 (400ng/ml) and low TNFα (1ng/ml) for 15 min prior to preparation of cell lysates. The experiment was reproduced three times with similar results. Data from one representative experiment are shown in the figure. (H) Representative images of prestarved HUVEC after stimulation with PBS (Ctrl), Ang2 (400ng/ml), low TNFα (1ng/ml) , high TNFα (10ng/ml) or a combination of Ang2 (400ng/ml) and low TNFα (1ng/ml) for 15 min to visualize nuclear translocation of p65 subunit (red). Phalloidin (green) was stained to visualize the cytoskeleton, whereas nuclei (blue) were stained with DAPI. The experiment was reproduce three times with similar results. Data from one representative experiment are shown in the figure. Values are mean±SD, *p≤0.05, **p≤0.01, ***p≤0.001, ns=non-significant.

G2+G3

4 5 6 7 8

4

6

8

10

Correlation coefficient (Spearman’s r)=0.2246p value (two-tailed)=0.0461

G1

4 5 6 7 8

4

6

8

10

Correlation coefficient (Spearman’s r)= -0.1236p value (two-tailed)=0.3192

Ang2Peptide

+- +- - C

tr

STA

T3-i

+

ICAM-1

Actin

D

E

0.00.51.01.52.0

Ang2Peptide

+- +- -

Ctr

STA

T3-i

+

**

ns

ICA

M-1

leve

l

ICAM-1

Actin

TNFα

Ang2

- -- + +- -

10 n

g

1 ng

1 ng

F

G

0.00.51.01.52.0

Ang2 - + +- -TNFα - -

10 n

g

1 ng

1 ng

ICA

M-1

leve

l *****

*** ns

***

Relative ANG2 expression

Rel

ativ

e IC

AM

-1ex

pres

sion

Relative ANG2 expression

Rel

ativ

e IC

AM

-1ex

pres

sion

Correlation analysis between ICAM-1 and ANG2 in breastcancer samples using the GSE3494 data set (n=67)

C

6

Movie S1, related to Figure 6. Movie footage of macrophages flowing over a HUVEC coated µ-Slide in a parallel flow chamber. The HUVEC coated chambers were treated with IgG for 12 hr and then subjected to continuous flow of macrophages at a rate of 30 µl/min over a period of 1 min.

Movie S2, related to Figure 6, Movie footage of macrophages flowing over a HUVEC coated µ-Slide in a parallel flow chamber. The HUVEC coated chambers were treated with Ang2 Ab for 12 hr and then subjected to continuous flow of macrophages at a rate of 30 µl/min over a period of 1 min.

Movie S3, related to Figure 6. Movie footage of microspheres flowing over a HUVEC coated µ-Slide in a parallel flow chamber. The HUVEC coated chambers were treated with IgG for 12 hr and then subjected to continuous flow of microspheres conjugated to either IgG antibodies at a rate of 30 µl/min over a period of 1 min.

Movie S4, related to Figure 6. Movie footage of microspheres flowing over a HUVEC coated µ-Slide in a parallel flow chamber. The HUVEC coated chambers were treated with Ang2 Ab for 12 hr and then subjected to continuous flow of microspheres conjugated to ICAM-1 antibodies at a rate of 30 µl/min over a period of 1 min.

7

58.4%

0 50K 100K 150K 200K 250K

FSC-A

0

50K

100K

150K

200K

250K

SSC

-A

88.5%

0 50K 100K 150K 200K 250K

SSC-W

0

50K

100K

150K

200K

250K

SSC

-H

11.9%

0 102

103

104

105

Gr1

010

2

103

104

105

CD

11b 89.0%

7.97%

0 102

103

104

105

Ly6G

010

2

103

104

105

Ly6C

74.9%

0 50K 100K 150K 200K 250K

FSC-A

0

50K

100K

150K

200K

250K

SSC

-A

92.7%

0 50K 100K 150K 200K 250K

SSC-W

0

50K

100K

150K

200K

250K

SSC

-H

4.85%

0 102

103

104

105

Gr1

010

2

103

104

105

CD

11b 95.9%

0.210%

0 102

103

104

105

Ly6G

010

2

103

104

105

Ly6C

Ang

2 A

b +

VE

GF

Ab

Ang

2 A

b +

VEG

F A

b +

Met

(GEM

)

Figure S6, related to Figure 7. Effect of different adjuvant therapeutic regimens on mobilization of CD133+VEFR2+Ter119- myeloid cells. Mice were treated for three weeks after mastectomy for removal of the primary tumor (4T1 orthotopic breast cancer). The treatment groups were as follows: IgG = control IgG group; MTD (PTX) = maximum tolerable dose Paclitaxel chemotherapy; Ang2 Ab = anti-Ang2 antibody treatment; LDMC (PTX) = Low dose metronomic chemotherapy; LDMC (PTX) + Ang2 Ab = combination of low-dose metronomic chemotherapy and anti-Ang2 antibody treatment (n=4 mice). The experiment was performed twice with similar results. Data from one representative experiment are shown in the figure. (B) Effect of antiangiogenic therapies or their combination with metronomic Gemcitabine on recruitment of immature myeloid cells to metastatic niche. Representative FACS plots from flow cytometric analyses of lung metastases of mice treated in the adjuvant setting with a combination of anti-VEGF antibody and anti-Ang2 antibody or anti-VEGF antibody, anti-Ang2 antibody and metronomic Gemcitabine for quantitation of recruited bone marrow derived cells (CD11b-positive Gr1-positive Ly6Chi). The mice were treated for three weeks after resection of syngenic Lewis lung carcinoma primary tumors (n=5 mice). The experiment was performed twice with similar results. Data from one representative experiment are shown in the figure. (C) Effect of Ang2 Ab treatment on the local concentrations of Bv8 (ELISA of lung metastatic foci) in postsurgically treated tumor bearing mice (n=5 mice). (D) Effect of Ang2 neutralization on Bv8 mediated angiogenesis and branch formation. Tube formation assays were performed using HCMEC (human cardiac microvascular endothelial cells) and Bv8 (500ng/ml), a major proan-giogenic cytokine from immature myeloid cells, was used as stimulus. Three increasing concentrations of IgG or Ang2 antibody were used to evaluate the effect on the average number of branches and the cumulative branch length after 6 h time point (n=5); values are mean±SD; *p≤0.05, ***p≤0.001, ns=non-significant). The experiment was reproduced three times with similar results. Data from one representative experiment are shown in the figure.

IgG

Ang2A

b0

100

200

300

400ns

Bv8

conc

entr

atio

n(p

g/m

l lun

g m

ets

lysa

tes)

IgG

Ang2 A

b

MTD (PTX)

LDMC (P

TX)

LDMC (P

TX)+ Ang

2 Ab

0

1000

2000

3000C

D13

3+V

EG

FR2+

Ter1

19-

mey

loid

cel

ls p

er 1

x107

** *

*

NS

A

B

C D

Aver

age

num

ber o

f br

anch

es p

er m

m2

0

5

10

15

20

IgGanti-Ang2

*

***

***

5 µg 10 µg 30 µg

8

Supplemental Experimental Procedures

Antibody reagents: The murine-chimeric antibody anti-Ang2 was generated as a derivate

of previously described human anti-Ang2 LC06 antibody (anti-murine/human Ang-2)

(Thomas et al., 2013). Murine-chimeric anti-VEGF antibody was generated based on the

human anti-VEGF B20-4.1 (anti-murine/human VEGF-A) (Liang et al., 2006). Both murine-

chimeric antibodies were generated by molecular fusion of human variable domains to

constant antibody domains of murine IgG2a.

Antibody genes were ordered as gene syntheses and cloned via unique restriction sites

using standard cloning procedures into separate expression vectors enabling secretory

expression in HEK cells growing in suspension. HEK293-F cells (Invitrogen) were

transfected according to the cell supplier’s instructions using Maxiprep (Qiagen)

preparations of the antibody vectors, Opti-MEM I medium (Invitrogen) and 293fectin

(Invitrogen) and the recombinant antibodies were then produced in serum-free FreeStyle

293 expression medium (Invitrogen) during 6 days.

For transient expression of murine-chimeric anti-Ang2 and anti-VEGF, transfected HEK cells

were cultured in stirred 10 l fermenters. Antibody containing cell culture supernatants were

harvested by centrifugation and sterile filtrated. Proteins were purified from supernatants

referring to standard protocols - Protein A, ion exchange and size exclusion chromatography

were applied. Purified antibodies were concentrated and diafiltrated by membrane-based

tangential flow filtration. The protein concentration was determined by measuring the OD at

280 nm, using a molar extinction coefficient calculated according to Pace et al (1995). Each

antibody was analytically characterized by SDS-PAGE, size-exclusion chromatography and

mass spectrometry. In addition, Surface Plasmon Resonance analysis was used to

demonstrate that target specificity and high affinity of anti-Ang2 and anti-VEGF parent

antibodies (Thomas et al., 2013; Liang et al., 2006) was maintained in the murine-chimeric

antibodies. Antibodies were stored in 20 mM histidine, 140 mM NaCl (pH 6.0) at −80 °C at a

concentration of 3 mg/ml. Endotoxin levels of both stock solutions was below 0.2 EU/ml.

Treatment regimens: Mice were treated intraperitoneally (i.p.) with the Ang2 Ab or control

IgG or VEGF Ab (10mg/kg) twice a week. For combination therapy, toxicity evaluation and

survival studies in 4T1 orthotopic breast cancer model, animals received Abraxane™

9

(Paclitaxel) for low-dose metronomic chemotherapy (6mg/kg, i.p, qd) or for maximum

tolerable dose (MTD) (30mg/kg, i.p., qdx5, 1 cycle) as previously described (Ng et al.,

2006).   Mice within Lewis lung carcinoma model received either Abraxane™ at low-dose

metronomic chemotherapy (6mg/kg, i.p, qd) (Ng et al, 2006) or Gemcitabine (Sigma) at

metronomic dosing (1mg/kg, i.p. qd) or at MTD (150mg/kg, i.p, twice a week) (Tran Cao et

al., 2010). CCL2 neutralizing antibodies (R&D) or control IgG were administered at dosage

of 2mg/kg twice a week for accessing the role of CCL2 in adjuvant settings as described

before (Zhu et al., 2011). All treatments were initiated after primary tumor resection and

randomization into treatment groups. CCR2 flox/flox::Mx1 Cre positive or negative mice

(Willenborg et al., 2012) were treated with three injections of Poly IC (Invivogen, 250µg, i.p.)

for Cre induction after primary tumor resection at day 0, 2 and 4 as previously described

(Pajerowski et al., 2010).

Additional animal experiments

Matrigel plug assay: 1) To analyze the effect of Ang2 neutralization on recruitment of

CCR2+ macropahges: LLC (1x106) cells were implanted in C57BL6 mice and the primary

tumor was removed after 2 weeks. Post-resection mice were treated with either IgG or Ang2

Ab. Serum was collected from mice 10 days post-resection of the primary tumor. Matrigel

(BD Bioscience, 354230) plugs containing 25% serum from either IgG or Ang2Ab treated

mice were implanted subcutaneously in the flanks of WT SCID mice. Another group to

evaluate the effect of addition of recombinant CCL2 to serum of Ang2 Ab treated mice was

also included in the Matrigel plug implanted animals. 2) To analyze the effect of Ang2

neutralization on Bv8 mediated angiogenesis. Matrigel was mixed at 4 °C with mouse

bFGF (R&D, 3139-FB-025/CF) dissolved in NSS at a final concentration equal to 1.0 µg/ml

or Bv8 (Peprotech, 100-46) dissolved in NSS at final concentration equal to 500ng/ml and

injected subcutaneously (0.5 ml/mouse) into the flank of 6–8 week-old SCID mice (Charles

River). Matrigel with NSS alone was used as negative control. Mice implanted with Bv8

containing Matrigel plus were treated with either IgG or Ang2 Ab. The plugs were collected

for further processing after 10 days of implantation.

Tail vein metastasis model: Tail vein metastasis experiments were performed in order to

study how Ang-2 affects expression of endothelial specific adhesion molecules, which may

10

enhance homing of metastases enhancing macrophages. LLC cells (1x106) were injected i.v.

and anti-Ang-2 antibody or IgG was administered i.p 1 hr after tumor cell injection. Lungs

were collected after 24 hr for cryo-preservation until further processing.

Frequency of metastatic lesion quantification: Number of metastatic lesions in lung and

bones were quantified by bioluminescent imaging and histological analysis. Frequency of

lymph node metastases were quantified on basis of bioluminescent imaging and confirmed

by necropsy.

Ex vivo analytical assays

Histological analyses and immunohistochemical procedures: For bone histology,

femurs were harvested, fixed overnight in 2% paraformaldehyde at 4°C, and decalcified in

0.5 M EDTA (pH 7.4) for 14 days at 4°C prior to paraffin embedding and sectioning at 5 µm.

Sections were stained with hematoxylin and eosin (H&E). PFA-fixed (4%) ovaries and Zinc

fixed Lungs were paraffin-embedded, sectioned (7µm) and stained with H&E. Sections of

Zinc fixed lung and Matrigel plugs were immunostained with antibodies against CD31

(1:100, BD553370/557355), SMA (1:200, Sigma C6198), F4/80 (1:100, Invitrogen

MF48004), CCR2 (1:150, ThermoScientific PA1-27409), F4/80(1:100, AbD Serotec

MAK0497R), HIF1alpha (1:100, Abcam ab8366). The primary antibodies were detected by

fluorescent-conjugated secondary antibodies (1:500, Invitrogen A11006 or Invitrogen

A21085), biotinylated secondary antibody (1:200, VectorLab BA 4001) or HRP-conjugated

secondary antibody (1:250, DAKO K403). Biotinylated and HRP secondary antibodies were

detected using DAB substrate kits (Liquid DAB+ Substrate (Dako K3468) or DAB peroxidase

Substrate Kit (VectorLab,SK 4100). Imaging was done on an Olympus IX81 microscope.

Image analysis

Vasculature analyses in metastases: Images of stained samples against CD31, SMA and

DAPI were acquired with the Zeiss Cell Observer. Analysis of CD31-positive areas within

each DAPI-enriched region of interest was performed using the Fiji image processing

software package. Vessel size distribution was determined by filtering identified CD31+

vessels according to area in 4 categories (I: 0-100 µm2; II: 100-400 µm2; III: 400-1600 µm2;

IV: >1600 µm2). The resulting number was set to proportion to the total vessel number.

SMA+ pericyte coverage was determined by analyzing the colocalization of CD31+ and

11

SMA+ areas in double stained samples. By the process of Gaussian Blurring a sufficient

overlap was created and compared to the number of total CD31+ vessels.

Mean metastatic lung fraction: H&E stained images of lung sections were acquired with

Zeiss Cell Observer. Five FOV per section and 5 sections per sample were analyzed.

Metastatic area fraction was analyzed using Trainable Weka Segmentation plugin of Fiji

image processing software package and expressed as percentage of lung cellularity area.

Matrigel plugs: Images of stained samples against F4/80, CCR2 and DAPI were acquired

with Zeiss Cell Observer. Analysis of F4/80 and CCR2 dual positive cells within DAPI

enriched region was performed using the Fiji image processing software package. Images of

double stained samples against CD31 and using DAPI were acquired with the Zeiss Cell

Observer. Analysis of CD31-positive areas within each DAPI-enriched region of interest was

performed using the Fiji image processing package.

Flow cytometry: For FACS analysis, lung metastases were harvested, briefly washed with

cold PBS and subsequently minced on ice. The tissue was digested with 3mg/ml

Collagenase A (Sigma) in DMEM at 37° for 30 min. A single cell suspension was prepared

by passing the cells through a syringe with an 18 G cannula and filtering them through a

70µm cell strainer. For macrophage identification, cells were stained with the following

antibodies: rat anti-mouse CD45-FITC (1:400, BD Pharmingen 30-F11), rat anti-mouse

CD11b-PE-Cy7 (M1/70) (1:000, eBioscience 25-0112-82), Tie2-PE (TEK4), (1:200,

eBioscience 12-5987-81); CCR2-APC (1:50, R&D systems FAB5538-A); F4/80-Pacific blue

(1:100, AbD Serotec MCA497PBT) and TREM-1-PE (1:100, R&D systems FAB1187P).

Similarly for endothelial cell identification, cells were stained with following antibodies rat

anti-mouse CD45-FITC (1:400, BD Pharmingen 30-F11); Lyve-1-FITC (1:250, eBioscience

53-0443); TREM-1-PE (1:100, R&D systems FAB1187P); CD31-APC (1:100, BD

Pharmingen 551262) and hamster anti-mouse Podoplanin-Alexa488 (1:100, e-bioscience

53-5381-82) Similarly for myeloid subset identification, cells were stained with following

antibodies rat anti-mouse CD11b-PE (1:000, eBioscience 12-0112); Gr1-FITC (1:100, BD

Pharmingen 553127); Ly6G (1A8)-Pacific Blue (1:100, Biolegend 127612), Ly6C (HK1.4)-

APC (1:25, Biolegend 128016).

12

For VEGFR2+CD133+Ter119- myeloid cells evaluation blood was collected in heparinized

tubes. Cell suspensions were evaluated after red cell lysis and labeling by flow cytometry

according to the protocol described earlier (Goon et al., 2006). At least 105 events per

sample were analyzed. BMRC were defined as CD45-(1:400, BD Pharmingen, 561086)

/Ter119-(1:200, BD Pharmingen, 553673) /VEGFR2+ (1:200, BD Pharmingen, 561252)

/CD133+(1:25, ebioscience, 11-1331). FACS analysis was performed with a FACS Canto™

(Becton Dickinson) and data were processed using FlowJoTM (TreeStar, OR).

Cellular and biochemical studies

Cell culture: LLC (Lewis lung carcinoma; obtained from ATCC) and 4T1-luc (mammary

cancer cell line; kindly provided by Dr. Gary Sahagian, Tufts University, Boston, USA) were

cultured at 37°C and 5% CO2 in MEM (Invitrogen) and RMPI 1640 (Gibco). The medium

was supplemented with 10% FCS, 1% streptomycin and penicillin, 1% glutamine, 1%

sodium pyruvate, and 1% nonessential amino acid. HUVE (human umbilical vein

endothelial) cells and HCME (human cardiac microvascular endothelial) cells were cultured

at 37°C, 100% humidity and 5% CO2 using Endopan3 medium (Pan Biotech GmbH) and

Endothelial cell growth medium MV (Promocell) respectively, supplemented with 1%

streptomycin and penicillin, 5% glutamine, 10% FCS, and 50 Ag/mL gentamycin. U937 cells

were cultured at 37°C and 5% CO2 in RPMI 1640 (Gibco) containing 10% FCS, 1%

glutamine, and 1% streptomycin and penicillin.

Flow chamber experiments: First, microsphere were prepared by covalently conjugating

yellow-green and PE carboxylated fluorescent microspheres (2 µm, Polysciences, Inc.,

Warrington-PA) to protein G (Sigma P4689) using a carbodiimide-coupling kit (Polysciences,

Inc). PE-conjugated microspheres were incubated with nonimmune mouse IgG (mIgG,

Southern Biotech, Birmingham, AL, catalog no. 0102-14), while yellow-green microspheres

were bound to anti-ICAM 1 mouse monoclonal antibody (MEM-111 Abcam, ab2213,

Cambridge, MA) for 2 hr at room temperature on a rotary shaker. Subsequently, the

microspheres were washed with PBS and BSA (0.1%). Prior to use, the microspheres were

washed and sonicated. Second, the flow chamber assay was performed using a µ-Slide 1

0.2 Luer Flow Kit (Ibidi GmbH) and a REGLO compact cassette pump (ISMATEC, IDEX

Health & Science GmbH) with a flow rate of 30µl/min. The µ-Slides were coated with

13

HUVEC and were preincubated overnight with Anti-Ang-2 Ab or IgG. The video capture

system consisted of an Olympus IX81 inverted epifluorescence microscope (Olympus) with

a MCV52 Full-HD camera system (Martin Microscope Company). The capture resolution

and rate was 1080p at 29.97 fps. The captured video was stabilized using iMovie 11v.9.04

(Apple Inc.), exported as uncompressed image sequence (29.97fps) and subsequently

processed using FIJI (Fiji is just image J, released under the General Public License). The

video footage (1 min per flow chamber) was quantified using the MTrack2 plug-in. Tracked

objects consisted either of rolling U937 cells on a HUVEC matrix or a mix of IgG-conjugated

R-PE microspheres and anti-ICAM-1-conjugated yellow-green microspheres.

Endothelial cell stimulation experiments: HUVEC (1.5x105) were seeded in 6-well plates

(Becton Dickinson 35-3046) followed by overnight starvation in endothelial basal media

supplemented with 0.5% FCS (starvation medium). Cells were stimulated with either rhAng2

(400ng/ml, R&D 623AN-CF) or PBS in starvation medium for 24 h. RNA was isolated, cDNA

synthesized and CCL2 expression was evaluated by TaqMan qRT-PCR .

Stimulations with rhAng2 and/or TNFα (low 1ng/ml or high 10 ng/ml, R&D 210-TA-CF) were

carried out for 15 min after overnight starvation. Cells were lysed using RIPA buffer, protein

concentrations were determined by BCA Protein Assay Kit (Pierce) followed by SDS-PAGE

and Western Blotting. The primary antibodies used to probe blots were rabbit anti-human

ICAM-1 (4915S, Cell Signaling), rabbit mAb anti-Phospho-IκBα (Ser32) (2859, Cell

Signaling), rabbit mAb anti-IκBα (4812, Cell Signaling) and rabbit anti-beta Actin (sc-16,

Santa Cruz Biotechnology). The membranes were incubated with horseradish peroxidase-

conjugated (HRP) secondary antibody (DAKO) and detected using enhanced

chemiluminescence substrate (ECL, Pierce). Blots were quantified using Fiji software.

For p65 nuclear translocation studies stimulated cells were fixed with 4% PFA,

permeabilized with Triton X-100 and stained with rabbit anti-human p65 (8242, Cell

Signaling), which was detected by Alexa 546 conjugated goat anti rabbit secondary antibody

(Life technologies, A11071). Phalloidin was stained with Alexa Fluor® 488 Phalloidin

(A12379, Invitrogen) probe to visualize the cytoskeleton while nuclei were stained using

DAPI. Images were takes using LSM 710 confocal microscope.

14

Stimulations with rhAng2 were carried out for 5 and 10 min after overnight starvation for

evaluation of STAT3 phosphorylation or nuclear translocation of p-STAT3. For studying

direct effect of Ang2 stimulation on phosphorylation of STAT-3, cells were lysed using RIPA

buffer, protein concentrations were determined by BCA Protein Assay Kit (Pierce) followed

by SDS-PAGE, Western Blotting and quantification using GE Amersham WB system. The

primary antibodies used to probe blots were mouse anti-human STAT3 (9139, Cell

Signaling) and rabbit anti-human p-STAT3 (Tyr 705) (9145, Cell Signaling). These were

detected by Amersham wb goat anti mouse cy3 29038275 and wb goat anti rabbit Cy5

29038278.

For p-STAT3 nuclear translocation studies stimulated cells were fixed with 4% PFA,

permeabilized with Triton X-100 and stained with rabbit anti-human p-STAT3 (Tyr 705)

(9145, Cell Signaling), which was detected by Alexa 546 conjugated goat anti rabbit

secondary antibody (Life technologies, A11071). Phallaoidin was stained with Alexa Fluor®

488 Phalloidin (A12379, Invitrogen) probe to visualize the cytoskeleton while nuclei were

stained using DAPI. Images were takes using Zeiss Cell observer microscope.

Stat3 inhibition experiment: HUVEC (1.5x105) were seeded in 6-well plates and starved

overnight in endothelial basal media supplemented with 0.5% FCS. Cells were incubated for

1 hr with 200 µM of either Stat3 inhibitor (573095, Millipore) or control peptide (573105,

Millipore) prior to rhAng2 (400ng/ml) stimulation for 1 h. Cell lysates were subjected to SDS-

PAGE followed by Western Blotting. Antibodies used were rabbit anti-human ICAM-1 and

rabbit anti-beta actin as described before.

Sprouting assay: Spheroids obtained by hanging drop method from cultured HCMEC were

seeded in collagen matrix in 24 well plates. The spheroids were stimulated with either

VEGF+bFGF (25ng/ml, R&D) (positive control), recombinant Bv8 (500ng/ml) or PBS

(negative control). The Bv8 treated spheroids were also treated with either IgG or Ang2 Ab

(1ug/ml). The cytokines were replenished after every 24 hr and the assay was stopped after

3 days. Images were takes using Olympus IX-81 microscope and sprout length was

analyzed using Cell F® software. 10 spheroids were analyzed for each experimental

condition.

15

Tube formation assay: Single cell suspension containing 25,000 HCMEC was added to each

well of collagen matrix coated 96 well plates. The seeded cells were stimulated with either

VEGF+bFGF (R&D) (positive control), recombinant Bv8 (500ng/ml) or PBS (negative

control). The Bv8 treated cells were also treated with either IgG or Ang2 Ab (5,10,20ug/ml).

Images were takes using Zeiss Cell observer microscope and number of branches were

analyzed using Fiji. 10 wells were analyzed per condition.

Western blot of murine tissue: For Western immunoblotting, lung metastases were

dissected, chopped and homogenized in RIPA buffer using a manual tissue homogenizer.

Protein concentrations were determined and protein lysates were separated by SDS-PAGE

followed by Western Blotting. Mouse ICAM-1 was detected using goat anti-mouse ICAM-1

(AF796, R&D) and actin was used as loading control. Blots were quantified using Fiji

software and signals were normalized to respective loading controls.

qRT-PCR: Total RNA from lung or HUVEC lysates was extracted using the RNeasy mini kit

(Qiagen) and subjected to a reverse transcriptase reaction using QuantiTect Rev.

Transcription Kit (Qiagen). cDNA was used in three replicates for qRT-PCR using pre-

designed TaqMan gene expression assay primers (Applied Biosystems) (mouse ICAM1:

Mm00516023_m1; mouse VCAM1: Mm01320970_m1; mouse ß-actin: Mm00607939_S1;

human CCL2: Hs00234140_m1; human ß2M: Hs00984230_m1). Expression data were

acquired and analyzed using an Applied Biosystems StepOne Plus Real time PCR system.

ELISA & cytokine array: Circulating levels of serum mCCL2, mGCSF and mVEGF or CCL2

and Bv8 in protein lysates from metastatic lesions were measured using commercial ELISA

kits (mCCL2,mVEGF and mGCSF from R&D and Bv8 from Cusabio®) as described by the

manufacturer.

Mouse cytokine array was purchased from R&D and level of cytokines was assessed as per

directions of the manufacturer. Briefly, the array membranes were blocked with blocking

buffer at room temperature for 1 h, and media and the detection antibody cocktail were

added and incubated at 4 °C overnight. After washing thrice with 2 ml of wash buffer at room

temperature, streptavidin- HRP was added to each membrane and incubated at room

temperature for 30 min. After washing, the cytokines were detected by a chemiluminescence

reaction. Spots were quantified by using FIJI® after background subtraction. Normalized

16

data were analyzed for treatment specific differences as processed and results per set were

used to determine the mean differences in cytokine abundance and significance was tested

using Students t-test.

References: Goon, P. K., Lip, G. Y., Boos, C. J., Stonelake, P. S., and Blann, A. D. (2006). Circulating endothelial cells, endothelial progenitor cells, and endothelial microparticles in cancer. Neoplasia 8, 79-88. Liang, W. C., Wu, X., Peale, F. V., Lee, C. V., Meng, Y. G., Gutierrez, J., Fu, L., Malik, A. K., Gerber, H. P., Ferrara, N., et al. (2006) Cross-species vascular endothelial growth factor (VEGF)-blocking antibodies completely inhibit the growth of human tumor xenografts and measure the contribution of stromal VEGF. J Biol Chem 281, 951–61 Ng, S. S., Sparreboom, A., Shaked, Y., Lee, C., Man, S., Desai, N., Soon-Shiong, P., Figg, W. D., and Kerbel, R. S. (2006). Influence of formulation vehicle on metronomic taxane chemotherapy: albumin-bound versus cremophor EL-based paclitaxel. Clin Cancer Res 12, 4331-4338. Pace, C. N., Vajdos, F., Fee, L., Grimsley, G., Gray, T (1995) How to measure and predict the molar absorption coefficient of a protein. Protein Sci 4, 2411–2423. Pajerowski, A. G., Shapiro, M. J., Gwin, K., Sundsbak, R., Nelson-Holte, M., Medina, K., and Shapiro, V. S. (2010). Adult hematopoietic stem cells require NKAP for maintenance and survival. Blood 116, 2684-2693. Thomas, M., Kienast, Y., Scheuer, W., Bahner, M., Kaluza, K., Gassner, C., Herting, F., Brinkmann, U., Seeber, S., Kablie, A., et al. (2013) A novel angiopoietin-2 selective fully human antibody with potent anti-tumoral and anti-angiogenic efficacy and superior side effect profile compared to Pan-Angiopoietin-1/-2 inhibitors. PLoS ONE 8:e54923. Tran Cao, H. S., Bouvet, M., Kaushal, S., Keleman, A., Romney, E., Kim, G., Fruehauf, J., Imagawa, D. K., Hoffman, R. M., and Katz, M. H. (2010). Metronomic gemcitabine in combination with sunitinib inhibits multisite metastasis and increases survival in an orthotopic model of pancreatic cancer. Mol Cancer Ther. 9, 2068-2078. Willenborg, S., Lucas, T., van Loo, G., Knipper, J. A., Krieg, T., Haase, I., Brachvogel, B., Hammerschmidt, M., Nagy, A., Ferrara, N., et al. (2012). CCR2 recruits an inflammatory macrophage subpopulation critical for angiogenesis in tissue repair. Blood 120, 613-625. Zhu, X., Fujita, M., Snyder, L. A., and Okada, H. (2011). Systemic delivery of neutralizing antibody targeting CCL2 for glioma therapy. J Neuro-oncol. 104, 83-92.

17