SHREE ET AL, SUPPLEMENTAL MATERIALS SUPPLEMENTAL FIGURE AND TABLE LEGENDS Supplemental Figure 1. Derivation and characterization of TS1-TGL and TS2-TGL PyMT cell lines and development of an orthotopic implantation model. (A) Workflow for tumor cell line derivation and orthotopic implantation. (B) Flow cytometry analysis of the established TS1 line for EpCAM and endogenous GFP expression prior to and following labeling with the TGL-imaging vector (Ponomarev et al, 2004) and sorting. (C) Representative brightfield image of the TS1-TGL cell line in culture. (D) Hematoxylin and eosin staining of representative orthotopic and spontaneous PyMT tumors, showing the morphological similarities. (E) Ex vivo bioluminescence imaging of lung and liver from a mouse implanted orthotopically with the imaging vector-labeled TS1-TGL cell line, indicating the presence of metastases to those organs. Supplemental Figure 2. Analysis of cathepsin protein, expression and activity levels in mammary tumors and BMDMs following Taxol treatment. (A) Orthotopically implanted tumors were harvested from mice 7 days after vehicle or Taxol treatment and whole tumor lysates were assayed for cathepsin proteins. Quantitation of bands for the active forms of cathepsins S, X/Z, and H are shown in the upper graphs, and none of these cathepsins showed a 1
significant difference in the lysates from Taxol- versus vehicle-treated tumors. Quantitation of the loading control, actin, on each of the three blots used for the relative quantitations of all cathepsins in Figure 1B and Figure S2 are shown in the lower graphs. Quantitation of all western blots was done using area under the curve (A.U.C.) analysis using exposures within the dynamic range of detection. n=6-7 mice per group. (B) Cathepsin mrna levels also increase after Taxol treatment. Quantitative RT- PCR analysis for all cathepsin genes detectable in whole tumor lysates, relative to an endogenous control gene, Ubiquitin C. n=7-8 mice per group. * p<0.05, ** p<0.01 by the Mann-Whitney test. (C) Cathepsin activity levels in macrophages do not change following Taxol treatment. Macrophage (mφ) cell lysates (left) or conditioned media (cm) (right) were labeled with the biotinylated cathepsin ABP, DCG04 (Greenbaum et al. 2000), showing there is no change in the levels of active cathepsins following treatment with Taxol versus vehicle. Supplemental Figure 3. Macrophage quantitation in PyMT tumors with the cell type specific marker Iba1. The macrophage-specific antibody, Iba1, which has generally been used to stain brain resident macrophages known as microglia, but has also been shown to label other tissue macrophage populations (Kohler 2007; Pyonteck et al. 2011), was used to label macrophages in PyMT tumors. The reason for using this rabbit antibody was to perform co-immunostaining with the pan-leukocyte marker 2
CD45, which is raised in rat, without concern for potential cross-reactivity in the subsequent detection with secondary antibodies, which could happen with other macrophage-specific antibodies such as CD68 or F4/80 that are also raised in rat. (A) Representative image of co-localization of two different macrophage markers: F4/80 and Iba1 in mammary tumors. (B) Quantitation via image analysis of F4/80 and Iba1 co-localization (n=11 tumors) showing that Iba1 labels the vast majority of F4/80 + cells, and is thus an appropriate macrophage marker for image analysis in PyMT tumors. Red line indicates the mean co-localization: 92.6%. (C) Macrophages accumulate in tumors after Taxol treatment. Representative composite images of tumors stained for the macrophage marker, Iba1, 7 days after vehicle or Taxol treatment. Higher magnification images from these composites are shown in Figure 2A. (D) Quantitation of intratumoral Iba1 + cells in tumors of vehicle- and Taxol-treated mice via image analysis shows there is no difference between the two groups at this timepoint. Individual data points represent tumors from each treatment group (n=13 vehicle; 12 Taxol). ns= not significant. Supplemental Figure 4. Co-staining of cathepsin B and S in macrophages in patient samples. Representative images of tumor samples post-treatment from Patients 6 and 8 stained with a CD68 antibody (blue), and antibodies against either cathepsin B or 3
S (red/brown color) to visualize tumor cells. Macrophages showing co-staining with cathepsin B or S are indicated by white arrows. Patient information can be found in Table S1. Images captured using a Zeiss Z1 Axioimager and 40x objective lens. Supplemental Figure 5. Co-culture and conditioned media assays for assessing effects of macrophages on Taxol-induced tumor cell death. (A) Representative flow cytometry plots showing separation of macrophages and tumor cells in co-culture assay (plot on left) and DAPI/ Annexin V staining to analyze tumor cell death (right). (B) Schematic of experimental design for conditioned media experiments. (C) BMDM conditioned media, treated with DMSO or JPM (10µM), was concentrated then labeled with the biotinylated cathepsin ABP, DCG04. There is a clear reduction in active cathepsin species following JPM treatment. (D) Tumor cell death in the BMDM conditioned media experiments was confirmed by western blot analysis of cleaved caspase 3 (CC3) at 24 hours after Taxol treatment. A representative western blot is shown on the left, with quantification of three independent experiments graphed on the right, * p<0.05, ns= not significant. Supplemental Figure 6. Analysis of cathepsin deficient macrophages. (A) Representative flow cytometry plots showing no major differences in differentiation of WT versus cathepsin-deficient BMDMs in response to colony stimulating factor-1 (Csf-1) in the macrophage derivation protocol. 4
(B) Representative images of bead phagocytosis assay of WT versus cathepsindeficient BMDMs, showing no differences in FluoSphere bead uptake in the absence of individual cathepsins or following incubation with the cathepsin inhibitor JPM. FluoSphere beads are labeled in red. All images were taken with an x20 objective lens on a Zeiss Axioimager M1 microscope. Supplemental Figure 7. JPM inhibits cathepsin activity in PyMT tumors. (A) Whole tumor lysates from vehicle or JPM-treated PyMT tumors (representative tumors, n=5 each) were labeled with the biotinylated cathepsin ABP, DCG04. (B) Quantification of active cathepsin bands relative to the loading control GAPDH shows there is a significant decrease in the levels of active cathepsins following JPM treatment administered twice daily at 100 mg/kg/day (p=0.0016). Supplemental Figure 8. Continuous low-dose cyclophosphamide does not affect tumor vascularity at end-stage, but does result in a significant increase in tumor cell apoptosis. (A) Graph of total vessel area, as determined by staining of endothelial cells with an antibody against CD34, relative to the total DAPI+ tumor area. There is no significant difference between tumors from vehicle and CLD Cyclo-treated animals at the end of the 9-14 week regression trial. Microvascular density (B) and vessel length (C) were similarly unaffected by CLD Cyclo treatment. n=10 mice per group for A-C. 5
(D) Graph of vessel functionality, assessed by colocalization of CD34 with FITCconjugated lectin, which was intravenously injected into tumor-bearing mice. Graph shows mean + SEM of individual tumors, n=4 mice per group, with no significant difference between groups. (E) Graph showing percentage of cleaved caspase 3 (CC3)-positive cells. There is a significant increase in apoptosis in tumors following CLD Cyclo treatment compared to vehicle (* p<0.05). (F) Representative images of CC3 staining in combination with either a tumor cell marker (EpCAM) or endothelial cell marker (CD34), showing that most CC3+ cells are in the tumor bulk, not blood vessels, of CLD Cyclo-treated mice. Arrow points to a rare CC3-positive endothelial cell. Supplemental Figure 9. Macrophages are protective against additional chemotherapies in other tumor cell lines. Percentage of DAPI + (dead) tumor cells in mono- or co-culture of the TS2 (A), Met-1(B), and AT-3 (C) cell lines treated with Etoposide (20 µm) and Doxorubicin (300 nm) with or without JPM-OEt. Co-culture with BMDMs significantly reduced cell death in response to both drugs in all lines tested, with different degrees of cathepsin dependence. For all graphs, data is from 3 independent experiments, each performed in triplicate; * p<0.05, ** p<0.01, ns =not significant. Supplemental Table 1. Information on patient samples used in this study. 6
Relevant information for pre- and post-chemotherapy matched samples from breast cancer patients used for CD68 immunostaining and analysis in Figures 2G and S4. Supplemental Table 2. Statistical results from 5-week combination drug trials. Supplemental Table 3. Statistical results from analysis of lung metastases at the end of 5-week trials. Supplemental Table 4. Statistical results from analysis of long-term survival data in MMTV-PyMT trials. 7
Shree et al, Supplemental Table 1 Supplemental Table 1. Information on patient samples used in this study. Patient No. Tumor grade ER Status PR Status HER2 Status Chemotherapy Clinical Response to Chemotherapy Pathological Response to Chemotherapy Survival Status CD68(+) cells after treatment 1 II/III Negative Negative Negative DC-T Partial Partial NED 2 III/III Negative Negative Negative DC-T Partial Partial NED no change 3 III/III Positive Negative Positive DC-T, Trastuzumab Minimal Minimal NED 4 I/III Positive Positive Negative DC-T Minimal Minimal NED 5 III/III Positive Negative Negative DC-T Partial Partial DOC 6 III/III Negative Negative Positive DC-T Partial Partial DOD 7 III/III Negative Negative Positive EC-T Partial Partial NED Table key: Chemotherapy given: DC= Doxorubicin and Cyclophosphamide; T= Paclitaxel; EC= Epirubicin and Cyclophosphamide NED= no evidence of disease; DOC= died of other cause; DOD= died of disease; = increase
Shree et al, Supplemental Table 2 Supplemental Table 2. Statistical results from 5-week combination drug trials. Comparison p-values Group 1 Group 2 Tumor volume curves (Area-under-curve analysis) Volume of excised tumors (Unpaired t-test) F-test for equal variance 1 Vehicle JPM 0.2309 0.2464 0.1502 Vehicle Taxol 0.0037 0.0883 0.8824 Vehicle Taxol+JPM < 0.0001 0.0002 0.0570 Vehicle Cyclo 0.0438 0.0354 0.9347 Vehicle Cyclo+JPM 0.0438 0.0988 0.4905 Vehicle Taxol+Cyclo < 0.0001 < 0.0001 0.0396 Vehicle Triple < 0.0001 < 0.0001 < 0.0001 Taxol Taxol+JPM 0.0468 0.0441 0.0413 Taxol Taxol+Cyclo 0.0513 0.0188 0.0283 Taxol Triple < 0.0001 < 0.0001 < 0.0001 Taxol+JPM Triple 0.0170 0.0003 < 0.0001 Taxol+Cyclo Triple 0.0234 0.0014 < 0.0001 1 Analysis conducted on data from excised tumors.
Shree et al, Supplemental Table 3 Supplemental Table 3. Statistical results from analysis of lung metastases at the end of 5-week trials. Comparison p-values (Mann-Whitney test) Group 1 Group 2 Total Metastatic Burden Metastasis Incidence Average Metastasis Size Vehicle JPM 0.2890 0.5383 0.4728 Vehicle Taxol 0.0836 0.4887 0.2382 Vehicle Taxol+JPM 0.0124 0.0673 0.0277 Vehicle Taxol+Cyclo 0.0703 0.1653 0.1347 Vehicle Triple 0.0054 0.0111 0.0812
Shree et al, Supplemental Table 4 Supplemental Table 4. Statistical results from analysis of long-term survival data. Comparison p-values (Log-rank test) Group 1 Group 2 Full Survival Early Survival 1 Late-stage Survival 2 Vehicle JPM 0.4208 0.6166 0.1329 Vehicle Taxol 0.0015 <0.0001 0.0024 Vehicle Taxol+JPM 0.0006 0.0437 0.0146 Vehicle Taxol+Cyclo <0.0001 <0.0001 <0.0001 Vehicle Triple <0.0001 0.0005 <0.0001 Taxol Taxol+JPM 0.078 0.3712 0.0146 Taxol Taxol+Cyclo 0.0002 0.0001 0.0069 Taxol Triple <0.0001 0.0077 0.0114 Taxol+JPM Triple 0.1206 0.0006 0.4638 Taxol+Cyclo Triple 0.2577 0.9322 0.0795 1 Data truncated at 70 days. 2 Data analyzed from day 70 forwards.