<p> 1 2 1Supplementary material</p><p>2Tumour hypoxia promotes melanoma growth and metastasis via High Mobility Group Box-1 </p><p>3and M2-like macrophages</p><p>4</p><p>5Roman Huber1,†, Barbara Meier1, Atsushi Otsuka1,†, Gabriele Fenini1, Takashi Satoh1, Samuel</p><p>6Gehrke1, Daniel Widmer1, Mitchell P Levesque1, Joanna Mangana1, Katrin Kerl1, Christoffer</p><p>7Gebhardt3,4, Hiroko Fujii2, Chisa Nakashima2, Kenji Kabashima2, Yumi Nonomura2, Reinhard</p><p>8Dummer1, Emmanuel Contassot1,‡,*, and Lars E. French1,‡,*</p><p>9</p><p>101. Department of Dermatology, University Hospital Zürich, Zürich 8091, Switzerland</p><p>112. Department of Dermatology, Kyoto University Graduate School of Medicine, Kyoto, Japan</p><p>123. Skin cancer Unit, German Cancer Research Center (DKFZ), Heidelberg, Germany</p><p>134. Department of Dermatology, Venereology and Allergology, University Medical Center </p><p>14 Mannheim, Ruprecht-Karl University of Heidelberg, Mannheim, Germany</p><p>15</p><p>16Content:</p><p>17 - Supplementary methods</p><p>18 o Control of knock-down efficiency and stability</p><p>19 o In vitro cell proliferation and apoptosis</p><p>20 - Supplementary figures</p><p>21 o Fig S1. Assessment and validation of HIF1 and HMGB1 detection and </p><p>22 localization by immunofluorescence.</p><p>23 o Fig. S2. Selection of HMGB1 knock-downs and stability of silencing efficiency </p><p>24 over time in vitro and in vivo</p><p>25 o Fig. S3. Validation of B16 cells transduced with lamin-specific shRNA as control.</p><p>26 o Fig. S4. The in vitro growth properties of B16 cells transduced with shRNA </p><p>27 specific to lamin or HMGB1 are identical.</p><p>3 1 4 5 6 28Supplementary methods 29</p><p>30Control of knock-down efficiency and stability </p><p>31To determine the knock-down stability of the B16-F10 mouse melanoma cell-line transduced</p><p>32with shRNA specific to HMGB1, shHMGB1-B16 as well as shLamin-B16 were cultured at</p><p>3337°C in 5 % CO2 in cDMEM medium (DMEM supplemented with 1 % L-glutamine and 10 %</p><p>34fetal bovine serum) and 1x105 cells were lysed at day 0, 7, 14, 21 and 28 in 10mM Tris pH</p><p>357.5, 1% NP-40, 150mM NaCl, 5mM EDTA with protease inhibitors (Roche). The cells lysates</p><p>36were subjected to Western-blotting using a rabbit polyclonal anti--actin (Cell Signaling) or a</p><p>37rabbit polyclonal anti-HMGB1 antibody (Abcam). Secondary antibodies were coupled to</p><p>38alkaline phosphatise (AP). AP detection was performed using the NBT/ BCIP substrate kit</p><p>39(Promega, Madison, WC). Blots were scanned using the CanonScan 9950F scanner</p><p>40(Canon, Tokyo, Japan). </p><p>41</p><p>42In vitro cell proliferation and apoptosis</p><p>43To compare in vitro proliferation of B16-F10 stably expressing shRNA specific to HMGB1 or</p><p>5 44lamin, 1.5 ×10 cells were cultured at 37°C in 5 % CO2 in DMEM supplemented with 1 % L-</p><p>45glutamine and 10 % fetal bovine serum (cDMEM). At days 0, 1, 2 and 3 mitochondrial</p><p>46dehydrogenase activity of living cells was measured by incubation with 10 % MTT (Sigma-</p><p>47Aldrich) for 4 h at 37°C. Optical densities were measured by the SpectraMax190 plate</p><p>48reader (Molecular Devices).</p><p>49Proliferation was measured using CFSE. Cells were incubated with PBS + 1 µM CFSE</p><p>50(Thermo Fisher Scientific, Waltham, MA) for 10 min at 37°C and seeded in 6-well plates. At</p><p>51days 0, 3, 6 and 8 cells were detached and CFSE-related fluorescence intensity was</p><p>52determined by flow cytometry. Acquisition was performed with a FACS Canto II (Becton-</p><p>53Dickinson) and sample analysis was done using the FACS DIVA software (Becton-</p><p>54Dickinson).</p><p>55To assess apoptosis in transfected cells by flow cytometry, cells were detached at day 0, 3, 6 7 2 8 9 10 56and 9 and stained with 1.0 µg/ml propidium iodide (Sigma-Aldrich) and Annexin V (Becton-</p><p>57Dickinson) for 15 min on ice. Acquisition was performed with a FACS Canto II (Becton-</p><p>58Dickinson) and analysed using the FACS DIVA software (Becton-Dickinson).</p><p>11 3 12 13 14 59Supplementary figures </p><p>60 a Metastatic melanoma cell line, hypoxic condition, control IgGs 61 DAPI Hif-1α HMGB1 DAPI HMGB1 Hif-1α HMGB1 Hif-1α</p><p>Metastatic melanoma cell line, normoxic conditions</p><p>Metastatic melanoma cell line, hypoxic conditions</p><p> b Metastatic melanoma cell line, normoxic conditions HMGB1 HIF1α HMGB1 HIF1α DAPI</p><p>100µm Metastatic melanoma cell line, hypoxic conditions</p><p>100µm</p><p>15 4 16 17 18 62Fig. S1. Hypoxia induces detectable HIF1 and HMGB1 relocalisation in human metastatic</p><p>63melanoma cell lines. (a) Immunofluorescence co-labelling with anti-HIF1 and anti-HMGB1</p><p>64antibodies of a metastatic melanoma cell line after 72 hrs in hypoxic conditions or left in normoxia. (b)</p><p>65Higher magnification of melanoma cells cultured in the same conditions as in (a). HIF1a is detectable</p><p>66only when cells are kept under low oxygen and is stabilized in both nucleus and cytosol whereas</p><p>67HMGB1 exhibit different localization from nuclear (normoxia) to cytosolic (hypoxia). (a) and (b) show 2</p><p>68independent metastatic melanoma cell lines and are representative of experiments repeated 3 times</p><p>69with each. 70</p><p>19 5 20 21 22 71 a 1.5 lamin shRNA 72 HMGB1 shRNA</p><p>1.0 WT-B16 73 T C   -</p><p>2 0.5</p><p>0.0 Sequence: 1 2 1 2 4 5</p><p>2.0 lamin shRNA HMGB1 shRNA 1.5 T C</p><p> 1.0 WT-B16  - 2 0.5</p><p>0.0 Clone: 1 2 3 4 5 6 1 3 5 7 8 9 11 14 15 16 17 19</p><p> b shLamin shHMGB1 WT-B16 cl1s2 cl17s5</p><p>HMGB1</p><p>-actin</p><p>0 7 4 1 8 0 7 4 1 8 D D 1 2 2 D D 1 2 2 D D D D D D</p><p> shLamin-B16 c DAPI HMGB1 DAPI HMGB1</p><p> shHMGB1-B16</p><p>23 6 24 25 26 74Fig. S2. Validation and selection of clones based on HMGB1 knock-down efficiency and</p><p>75stability. (a) B16 cells were transduced with 2 sequences of shRNA specific to lamin and 4 HMGB1-</p><p>76specific shRNA sequences. Quantitative PCR was performed on in vitro expanded</p><p>77transduced/selected cells (puromycin). B16 cells transduced with shRNA specific to lamin (sequence</p><p>782) or HMGB1 (sequence 5) were subsequently cloned by the limiting dilution method. Quantitative</p><p>79PCR was performed on in vitro expanded transduced/selected clones (puromycin). Results are</p><p>80expressed as 2-CT and standardized to wild-type B16 for which a 2-CT value of 1 was attributed</p><p>81(dashed line). (b) Cultures of B16 clones transduced with lamin-specific shRNA (clone 1 of sequence</p><p>822) or HMGB1 (clone 17 of sequence 5) were harvested, lysed and subjected to western-blot analysis</p><p>83using anti-HMGB1 and anti--actin antibodies at the indicated time points. (c) B16 cells transduced</p><p>84with lamin-specific shRNA (clone 1 of sequence 2) or HMGB1 (clone 17 of sequence 5) were injected</p><p>85s.c. to C57BL/6 mice and the resulting tumours were dissected at day 13 and stained with an anti-</p><p>86HMGB1 antibody. Nuclei were visualized using DAPI. Pictures are representative of 7 tumours per</p><p>87group. 88</p><p>27 7 28 29 30 89 </p><p>400 shLamin cl1s2 )</p><p>3 wild-type m 300 m (</p><p> e m</p><p> u 200 l o v</p><p> r o</p><p> m 100 u T</p><p>0 0 6 8 9 10 11 12 13 Days after tumor inoculation 90</p><p>91</p><p>92</p><p>93</p><p>94</p><p>95Fig. S3. Exclusion of off-target effects upon transduction of B16 cells with shRNA to lamin.</p><p>96Wild type and lamin shRNA-expressing B16 cells (cl1s2, control) were injected s.c. into C57BL/6 mice</p><p>97and displayed comparable tumour growth in vivo (n=5 mice/group). Representative results of 3</p><p>98independent experiments are presented.</p><p>31 8 32 33 34 99</p><p>100 101 3 a b B16-F10 101 shLamin-B16 101 2 shHMGB1-B16 102 101 1 103 101 0 d e</p><p>104 t n u</p><p> o 109 c</p><p>105 s l l e C 106 108 0.5 shLamin c shHMGB1 0</p><p>7</p><p>107 y 0.4 10 a s D</p><p> e o</p><p> u t 0.3 l</p><p>108 a 6 d 10 v e</p><p> s i D l 0.2 a O</p><p>109 m r 5 o 0.1 10 n 0 2 4 6 8 10 12 14 110 0.0 Days</p><p>111</p><p>112 Day 0 Day 3 Day 6 Day 9 d 90 90 90 90 s</p><p>113 l l 80 e 80 80 80 c 70 e 70 70 70 v i t 114 i 60 s 60 60 60 o</p><p> p 50</p><p>50 50 50 I 15 15 15 P 15 /</p><p>115 V 10 n 10 10 10 i x</p><p> e 5 n 5 5 5 116 n A 0 0 0 0 117</p><p>118</p><p>119 e 0.4 shLamin shHMGB1 0</p><p>120 y</p><p> a 0.3 e D s</p><p> a o t e</p><p>121 l d e</p><p> r 0.2 e</p><p> s i H l a D</p><p>122 L m</p><p> r 0.1 o n 123 0.0</p><p>124</p><p>125Fig. S4. The in vitro growth properties of B16 cells transduced with shRNA to lamin and</p><p>126HMGB1 are identical. (a) B16 cells transduced with shRNA specific to lamin or HMGB1 were labelled</p><p>127with CFSE and collected after 72 and 144 h and analysed by flow cytometry. (b) B16 cells transduced</p><p>35 9 36 37 38 128with lamin- or HMGB1-specific shRNA were counted over a 15-day culture period. (c) Proliferation of</p><p>129B16 cells transduced with lamin shRNA (white histograms) or HMGB1 shRNA (black histograms) was</p><p>130assessed over a 3-day culture period using the MTT assay. (d) At day 0, 3, 6 and 9, B16 cells</p><p>131transduced with shRNA specific to lamin (shLamin) or HMGB1 (shHMGB1) were harvested, stained</p><p>132with Annexin V and propidium iodide (PI) and analyzed by flow cytometry. Annexin V +/PI+ cells were</p><p>133considered as late apoptotic. As a positive control, wild-type B16 cells (WT) were treated with the</p><p>134apoptosis inducer staurosporin. (e) Viability of B16 cells transduced with shRNA specific to lamin</p><p>135(white histograms) or HMGB1 (black histograms) was assessed over a 3-day culture period using an</p><p>136LDH release assay. The mean +/- SEM of 3 independent cultures is shown.</p><p>39 10 40</p>