The bone metastasis niche in breast cancer-potential overlap with the haematopoietic stem cell niche in vivo.

Tumour cell lines MDA-MB-231-GFP-IV [29] , T47D and MCF7 breast cancer cell lines (ATCC) were cultured in RPMI 1640 supplemented with 10% FBS (Life Technologies, Paisley, UK /Invitrogen) at 37 °C 5%CO 2, MDA-MB-231-NW1-Luc2 cells were cultured in DMEM (Life Technologies, Paisley, UK) + Pyruvate medium enriched with 100 U/mL penicillin/streptomycin and 10% FBS (Sigma Aldrich Co Ltd, Poole, UK). Prior to the injections, tumour cells were labelled either with the lipophilic membrane dye Vybrant-CM-DiI or Vybrant-DiD (Life Technologies Ltd, Paisley, UK) according to the manufacturer's instructions.

Animal models Six-week old and twelve-week old female

BALB/c nude mice (Charles River, UK) were used to assess the homing of breast cancer cells in bone and to establish any effects of modification of the niche. Mice were housed in a controlled environment with a 12 h light/dark cycle at 22 °C. They were provided with ad libitum access to 2018 Teklad Global 18% protein rodent diet containing 1.01% Calcium (Harlan Laboratories, UK) and water, and housed in groups of 5/6 in ventilated cages. The Research Ethics Committee for animal experimentation, of the University of Sheffield, UK reviewed and the Home Office approved all the work included in this manuscript. All in vivo experiments were performed in accordance with the UK Animals (Scientific Procedures) Act 1986 Home Office regulations under the authority of PPL 70/8964 and PPL 70/8799.

Bone homing studies

To assess the homing of breast cancer tumour cells to bone, 12-week old female BALB/c nude mice were injected intravenously (i.v.) with 1 × 10 5 MDA-MB-231-GFP-IV breast cancer cells labelled either with the membrane dye Vybrant-CM-DiI or Vybrant-DiD and culled on day 5. Hind limbs were dissected, muscles removed, femora and tibiae separated and processed for imaging as described below. To investigate whether different routes of administration affected the pattern of homing of breast cancer cells, 12-week old female BALB/c nude mice were injected either intravenously ( n = 1) or via the intracardiac (i.c) route ( n = 5) with 1 × 10 5 MDA-MB-231-GFP-IV breast cancer cells labelled with the membrane dye Vybrant-DiD and then culled on day 5 and long bones collected. To establish whether the oestrogen receptor (ER) status of the breast cancer cells affected their pattern of bone homing, 1 × 10 5 Vybrant-DiD labelled ER-ve MDA-MB-231-GFP-IV cells, ER+ve MCF-7 or T47D cells were injected i.c. in 12-week old female BALB/c nude mice ( n = 5/group). Long bones were collected 5 days after cell injection and processed for imaging as described below.

Modification of the HSC niche

To confirm that AMD3100 mobilises HSC/progenitor cells (PCs) from bone marrow niches into the circulation, 12-week old female BALB/c-Nude mice were treated with AMD3100 (5 mg/kg, i.p.) or Saline daily for 5 days. Three hours after the last drug administration, mice were culled and peripheral blood isolated by cardiac puncture using syringes containing 4% sodium citrate. Blood was centrifuged to remove the serum and red blood cells lysed with ammonium chloride solution (BioLegend Ltd, UK). Peripheral blood mononuclear cells were resuspended in IMDM supplemented with 2% FCS and 100 u/mL Penicillin/Streptomycin at a density of 2 × 10 6 cells/mL. Cell suspensions were diluted 1:10 with MethoCult GF 3434 medium (Stem Cell Technologies), seeded into 6-well plates and maintained in an humidified atmosphere at 37 °C with 5% CO 2 for 11 days to allow colony formation. Bright field images of colonies were taken using a Leica DMI4000B microscope. To assess whether the mobilisation of hematopoietic stem cells (HSCs) would affect homing of breast cancer cells in bone, 12-week old female BALB/c nude were injected i.p. with the CXCR4 antagonist AMD3100 (Sigma-Aldrich) 5 mg/kg (100 µl) or PBS daily for 5 days. 24 h after the last injection, animals were injected i.v with 1 × 10 5 MDA-MB-231-GFP-IV Vybrant-DiD -labelled cells and culled on day 10.

Comparison between young and mature animals

To compare the seeding of breast cancer cells in the bone microenvironment of young and mature animals 6- and 12-week old female BALB/c nude mice ( n = 8/group) were injected i.v. with 1 × 10 5 MDA-MB-231-GFP-IV cells labelled with the membrane dye Vybrant-CM-DiI. Animals were culled 5 days after tumour cell injection.

2.6 µCT analysis

Microcomputed tomography (µCT) analysis was carried out on the resected long bones ex vivo using a Skyscan 1272 X-ray-computed microtomograph (Skyscan). Image acquisition was performed using a voltage of 50 kV, a current of 200 µA, a medium camera resolution of 2016 × 1344, an aluminium filter of 0.5 mm and a pixel size of 4.3 µm. Images were captured of proximal tibias every 0.7° through a 180° rotation. Acquired images were reconstructed and analyzed using NRecon and CTAn software (Skyscan).

Two-photon microscopy

Animals were culled, hind limbs dissected, muscles removed, femora and tibiae separated and collected snap frozen in liquid nitrogen and stored at −80 °C. Cryopreserved long bones were then embedded into Cryo-M-Bed (Instrument Co. Ltd, Huntingdon, UK) and the bone marrow exposed using a Bright OTF Cryostat and a 3020 microtome (Bright Instrument Co. Ltd, Huntingdon, UK). A stack area of exposed bone marrow (2104 µm × 2525 µm 70 µm depth) was captured using a Zeiss LSM510 NLO upright 2-photon/confocal microscope (Carl Zeiss Inl, Cambridge, UK). DiD-labelled and CM-Dil-labelled breast cancer cells were visualised using a 633 nm HeNe and 543 HeNe lasers, respectively. The bone structure was detected with a Chameleon 2-photon laser at 900 nm (Coherent, Santa Clara, CA.). The Volocity 3D Image Analysis software 6.01 (PerkinElmer, Cambridge, UK) was used to quantify the abundance of Vybrant-CM-DiI or Vybrant-DiD events, their distance to the nearest bone surface and to other DTCs [27] . The quantification of dye-labelled tumour cells and their spatial distribution were performed in two different regions of interest (ROIs, Fig. 2 B) from which the cortical bone was excluded.

Confocal microscopy

Long bones were fixed in 4%PFA, decalcified in 0.5 M EDTA, embedded in gelatin and frozen at −80 °C [27] . Immunofluorescence was performed on 30 µm thick cryosections (three non-consecutive levels) for each bone. Tissue sections were permeabilized for 30 min in 0.3% Triton X-100, where appropriate Streptavidin/Biotin blocking (Vector Labs) was performed. Cryosections were incubated for 1 h with primary antibody against murine Endomucin (2 µg/mL, Santa Cruz Biotechnology), Osteopontin (R&D Systems) or human CD29 and CD59 (10 µg/mL, BioLegend). Tissue sections were washed with PBS and incubated with either a fluorescently-conjugated (5 µg/mL) or biotinylated secondary (7.5 µg/mL) antibody followed by fluorescently-conjugated streptavidin. Tissue sections were washed with PBS and incubated with either a fluorescently-conjugated or biotinylated secondary antibody followed by fluorescently-conjugated streptavidin. All incubations were carried out at ambient temperature and ProLong Gold Antifade reagent (Thermo Fisher) to mount the glass coverslips. Images of the immunofluorescence staining were captured with Nikon A1 (Nikon Instruments Europe) or Zeiss LSM 880 AiryScan (Carl Zeiss Microscopy GmBH) confocal systems.

Statistical analysis

Statistical analyses were performed using GraphPad Prism software (Version 6.0 and 7.0). Student T-tests and Two-way ANOVA and Tukey's post hoc test were used as indicated in each figure legend. A p -value of p < 0.05 was considered significant.

Appendix Supplementary materials Image, application 1 Image, application 2