Transgenic mice
We generated adipocyte-specific knockout Hsp47 mice by crossing Hsp47 Loxp/Loxp mice with Adipoq-cre mice (Jackson Lab, 010803). Asporin knock-out mice are from Dr. Paula Hurley at Vanderbilt University. Tail tip samples were collected for DNA isolation and genotyping PCR between 14 and 21 days after born. Genotyping PCR was performed using the 2X Mater Mixer and primers listed in the Supplemental Table. Female C57BL/6 mice, or Hsp47 adipocyte-specifically knockout (Hsp47 Adi-KO) or asporin KO female litter mates were randomly mixed and grouped for the xenograft experiments. All the mice were housed under 12/12 light/darkness cycles with free access to diet and water in the Division of Laboratory Animal Resources at the University of Kentucky. All procedures were performed with Division of Laboratory Animal Resources guidelines at the University of Kentucky.
Human and mouse cell lines and cell line classification
The human breast cancer cell lines MDA-MB 231 and human embryonic kidney cell line HEK293T were obtained from American Type Culture Collection (ATCC). Mouse breast cancer cell lines EO771 was obtained from ATCC. MDA-MB 231 cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM)/F12 (Sigma Aldrich, D8437) with 10% fetal bovine serum (FBS) (Sigma Aldrich, F2442), 10 units/ml of penicillin and 0.1 mg/ml of streptomycin (Sigma, P4333). HEK293 FT cells were maintained in DMEM (Sigma Aldrich, D6429) with 10% FBS (Sigma Aldrich), 10 units/ml of penicillin, and 0.1 mg/ml of streptomycin (Invitrogen). EO771 cells were maintained in RPMI-1640 (Sigma Aldrich, R8758) with 10% FBS (Sigma Aldrich), 10 mM HEPES, 10 units/ml of penicillin, and 0.1 mg/ml of streptomycin (Invitrogen). Mouse 3T3-L1 cells were maintained in DMEM (Sigma Aldrich) with 10% Newborn calf serum (NCS) (Sigma, N4762), 10 units/ml of penicillin, and 0.1 mg/ml of streptomycin (Invitrogen). All the cells were cultured at 5% CO2 and 95% O2 at 37 °C. All cells were treated with Plasmocin™ (Invivo Gen, ant-mpt-1) to eliminate and prevent mycoplasma contamination.
Kaplan-meier survival analysis and association analysis
To address the clinical relevance of enhanced Asporin expression, we assessed the association between asporin mRNA levels and TNBC patient survival using Breast Cancer Gene-Expression Miner with multiple published datasets [24]. Tumor samples were equally grouped into low and high asporin expression based on the mRNA levels. The Cox proportional hazard (log-rank) test assessed significant differences in recurrent-free survival ratio. The correlation between Hsp47 expression and asporin was analyzed using the public microarray dataset generated from human breast cancer tissues.
DNA constructs, transfection, and viral transduction
SERPINH1 cDNA was cloned into pCDH1 plasmid and generated expression vector pCDH1-Hsp47 [23]. Flag-tagged mouse asporin full cDNA and truncated asporin cDNA (Mutant 1, Mutant 2) were sub-cloned from pCMV3-Asporin-FLAG (Sino Biological, Supplemental Table) into pCDH1 plasmid and generated expression vector pCDH1-Asporin (WT)-Flag, pCDH1-Asporin (MT1)-Flag and pCDH1-Asporin (MT2)-Flag. Mouse SERPINH1 Knockdown plasmids shHsp47 (Target Sequence: CGAACACTCCAAGATCAACTT, TRCN0000008532) were purchased from Sigma. HEK293 cells were transfected with pCDH1-Hsp47 and pCDH1-Asporin (WT)-Flag/pCDH1-Asporin (MT1)-Flag/pCDH1-Asporin (MT2)-Flag using FuGENE® HD Transfection Reagent (Promega, E2312). Cell lysis was collected for immunoprecipitation assay 48 h after transfection.
Immunoblotting analysis
Total tissue protein was isolated from adipocyte tissue using Minute™ Total Protein Extraction Kit (Invent, AT-022). Deposited ECM protein was isolated from culture cells using Compartment Protein Extraction Kit (Sigma, 2145). Culture cells were lysed in 2% SDS in PBS buffer containing phosphatase and protease inhibitor cocktails (EMD Millipore, 539131). Conditioned medium was collected at 24 h after being replaced with plan medium and precipitated by Acetone. Protein concentration was measured using Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, 23225). Equal amounts of protein lysates or cell conditional medium (normalized to cell numbers) were subjected to SDS gel electrophoresis, immunoblotted with primary antibodies at 4 °C overnight, and DyLight 680/800-conjugated secondary antibodies for 2 h at room temperature. The secondary antibodies used in the study were: DyLight 680 conjugated goat anti-rabbit IgG secondary antibody (Thermo Fisher Scientific, 35569), DyLight 800 conjugated goat anti-mouse IgG secondary antibody (Thermo Fisher Scientific, SA5-35521), DyLight 680 conjugated donkey anti-goat IgG secondary antibody (Thermo Fisher Scientific, SA5-10090). The primary antibodies used in this study were listed in Supplemental Table.
Reverse transcription-polymerase chain reaction (RT-PCR)
Total RNA was extracted from adipocyte tissue of control or Hsp47 Adi-KO mice using TRIzol reagent (Invitrogen, 15-596-018). cDNA was synthesized using the SuperScript™ III First-Strand Synthesis System (Invitrogen, 18-080-051) from 1.0 µg RNA samples. RT-PCRs were performed using Taq DNA polymerase on the C1000 Touch Thermo Cycler (Bio-Rad). Thermal cycling was conducted at 95 °C for 60 s, followed by 40 cycles of amplification at 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 15 s. PCR products were performed agarose gel electrophoresis using 2% Agarose in TAE buffer (40 mM Tris-acetate, 1 mM EDTA). Information of primers used for amplification SERPINH1, β-Tubulin, and 18 S rRNA is in Supplemental Table.
Special diet feeding
3-week old males and females carrying different genotypes (Adipo-Cre/HSP47Loxp/Loxp, Adipo-Cre/HSP47Loxp/+, Adipo-Cre/HSP47+/+) were fed with a low-fat diet (LFD, 10 kcal% fat, Research Diets, D12450B) or high-fat diet (HFD, 60 kcal% fat, Research Diets, D12492). This special diet was renewed every week, and the changes in food weight were weighed for measuring the food intake of the mice. The body weights of mice with LFD were measured every week during the 12-week feeding period, and the body weights of mice with HFD were measured every week during the 20-week feeding period.
Glucose measurement and glucose tolerance tests
HFD-fed mice with different genotypes fasted for 6 h in the morning for glucose measurements. The tail vein was lanced with the lancet and a drop of blood from the tail was read as the baseline glucose level. For the glucose tolerance test, 6 h fasted mice were intraperitoneally injected with Glucose (2 g/kg body weight, 10 ml/kg volume). Blood glucose was measured at 15 min, 30 min, 60 min, 90 min, and 120 min after injection.
Serum insulin measurement
HFD-fed mice with different genotypes fasted for 6 h in the morning for insulin level measure. The tail vein was lanced with the lancet. Round 50 µl blood sample was collected and centrifuged for at least 15 min at 2200–2500 RPM within one hour of collection. Then the serum was transferred to new EP tubes for insulin levels measurement by the Mouse Ultrasensitive Insulin ELISA kit (ALPCO, 80-INSMSU-E01).
Calorimetry experiments
HFD-fed mice with different genotypes were examined the energy expenditure and energy balance via the TSE calorimetry chambers and the EchoMRI. Calorimetry will be performed according to the SOP as described in the 2016–2337 Protocol for The Center of Research in Obesity and Cardiovascular Diseases (COCVD) Mouse Metabolism Core of the University of Kentucky.
Xenograft studies
Control, Hsp47 Adi-KO and Asporin KO female mice were weaned at three-week-old and fed with HFD for twelve weeks. The wnt tumor tissue was harvested from mice and digested with collagenase, and the tumor organoids were collected for xenograft experiments. Mice were randomly grouped and injected with 1 × 105 EO771 cells or Wnt organoids at the 4th mammary fat pad. Tumors were measured with a caliper every 2 days to analyze tumor growth. At the experimental endpoint, tumors were harvested and fixed with a 10% Formalin solution for the paraffin-embedded section or frozen in OCT.
Five-week-old female C57BL/6 mice were fed with HFD for four weeks and randomly grouped and daily injected with vehicle control or 10 mg/kg Col003 by intraperitoneal injection for eight weeks. Mice were injected with 1 × 105 EO771 cells at the 4th mammary fat pad. Tumors were measured with a caliper every 2 days and when tumors reached 60 ~ 70 mm3, mice were treated with vehicle control or 10 mg/kg Col003 daily. At the experimental endpoint, tumors were harvested and fixed with a 10% Formalin solution for the paraffin-embedded section or frozen in OCT for SHG imaging.
Second harmonic generation (SHG) imaging
SHG images were collected for adipose and tumor tissue sections through a Zeiss 20× water-dipping objective at 880 nm excitation. A cube filter set containing emission filters SHG (425 nm) was used in on the Zeiss 880 multiphoton microscope equipped with an Insight X3 extended wavelength laser (Spectra Physics, Santa Clara, CA, United States). Images were quantified with ImageJ.
Masson’s trichrome staining
Masson’s trichrome staining was performed following the manufacturer’s protocol (Polysciences, 25088). Mammary gland sections from 8-week-old control or Hsp47 Adi-KO female mice were deparaffinized and rehydrated through 100% alcohol, 95% alcohol, and 70% alcohol to water. Samples were re-fixed in Bouin’s solution at 60 °C for 60 min, stained in Weigert’s working hematoxylin for 10 min, and then stained in Biebrich scarlet-acid fuchsin solution for 5 min. Sections were incubated in phosphomolybdic/phosphotungstic acid solution for 10 min, and then were transferred to an aniline blue solution and incubated for 5 min. Then sections were incubated with 1% acetic acid for 1 min and washed in distilled water. Images were taken with a Nikon Eclipse 80i microscope.
Immunohistochemical staining and H&E staining
Xenograft tumor sections were deparaffinized and rehydrated through 100% alcohol, 95% alcohol, and 70% alcohol to PBS solution. Endogenous peroxidase was blocked by incubation with 3% H2O2 for 20 min. At the antigen retrieval step, slides were steamed in citrate sodium buffer for 30 min. Slides were blocked with Avidin/Biotin Blocking Kit (Vector Laboratories, SP-2001) and 5% goat serum, incubated with primary antibodies (anti-Hsp47, anti-Ki67, anti-Active Caspase 3, anti-F4/80, Supplemental Table) at 4 °C overnight, and then the sections were incubated with Biotinylated Goat Anti-Mouse IgG Antibody (Vector Laboratories, BA-9200) or Biotinylated Goat Anti-Rabbit IgG Antibody (Vector Laboratories, BA-1000) at room temperature for 60 min. After incubated with Streptavidin, Horseradish Peroxidase (Vector Laboratories, SA-5704) at room temperature for 30 min, staining signaling was accomplished after incubated with substrate diaminobenzidine (DAB, Vector Laboratories, SK-4100) and counterstained with hematoxylin. Images were taken with Nikon Eclipse 80i microscope.
Adipocyte differentiation induction
3T3-L1 Cells were seeded in a 6-well plate at a density of 6 × 105 cells/well as described in published paper [25]. After 24 h, the culture medium was renewed when cells reached 100% confluence. After 48 h, cell differentiation was induced by changing the medium to differentiation medium I: DMEM containing 10% FBS, 10 units/ml of penicillin, and 0.1 mg/ml of streptomycin, 0.5 mM IBMX, 0.25 µM dexamethasone, and 1 µg/ml insulin, 2 µM rosiglitazone. After 48 h, the medium was changed to differentiation medium II: DMEM containing 10% FBS, 10 units/ml of penicillin and 0.1 mg/ml of streptomycin, and 1 µg/ml insulin. After 48 h, the medium was changed to DMEM containing 10% FBS and 10 units/ml of penicillin and 0.1 mg/ml of streptomycin.
Flow cytometry analysis
To analyze adipocyte tissue infiltrated macrophage (Adipocyte tissue macrophages, ATMs), macrophage markers expression in adipocyte tissues were examined by flow cytometry analysis. Immune cells were isolated from adipocyte tissue as previously described [26]. In general, subcutaneous white adipocyte tissues (WAT) were collected from HFD-fed mice and digested by collagenase II digest solution. Stromal vascular cells were incubated with Fc-block (anti-CD16/32) and stained with cell surface markers (Supplemental Table) for 30 min at 4 °C protected from light to identify ATMs. For controls, additional stromal vascular cells are stained with each antibody individually (single stained controls) and the antibody cocktail minus one antibody, fluorescence minus one (FMO controls). Single stained controls and FMO controls were used to set up the compensation matrix and sorting gates. Add an additional 400 ul PBS and kept on ice for FACS analysis. FACS analysis was done with Becton Dickinson LSR II and data were analyzed by FlowJo software.
Decellularization and ex vivo implantation
Adult female mammary gland tissues were collected and sliced into 1–2 mm thick pieces, then placed in a sterile 6-well plate and incubated in rinse solution (deionized water with 5x penicillin/streptomycin) for 1 h at 4 °C [27]. Mammary gland tissues were removed from the rinse solution and incubated in 4% sodium deoxycholate (SDC solution) with 1x Pen/Srep for 24 h (change after 12 h) at room temperature to lyse cells. After rinsing 5 times with rinse solution for 1 h, mammary gland tissues were incubated in 8 mM 3-((3-cholamidopropyl)-dimethylammonium)-1-propanesulfonate (CHAPS solution) for 24 h (change after 12 h). Following rinsing 5 times with rinse solution, tissues were incubated in DNase solution (DNase 100 U/mL, with 5 × penicillin/streptomycin) 2 h at room temperature to lyse DNA. Mammary gland tissues were incubated with isopropanol for 24 h (change after 12 h, 4 °C), and then washed with rinse solution (20 min, three times) and stored at − 80 °C. Decellularized mammary gland tissue was cut into 5 mm square cubes and placed in a p-HEMA-coated 24-well plate. 0.2 × 106 MDA-MB 231-Luc cells were seeded on the top of the decellularized mammary gland tissue and incubated at 37 °C, 5% CO2 for 2 h to allow cells to attach. DMEM/F12 medium (1×B27, 20 ng/ml bFGF, 20 ng/ml EGF, 0.5 µg/ml hydrocodison, 5 µg/ml Insulin, 100 µg/ml Gentamicin (Life Technologies, Grand Island, NY, USA)) was carefully added to each well and incubated at 37 °C, 5% CO2 for the desired time. Luminescent intensity of MDA-MB 231-Luc cells attached and grew in the decellularized mammary tissue was measured using an in vivo imaging system (IVIS) on day 5 after seeding.
IF staining
Frozen tissue sections were fixed with 4% paraformaldehyde and permeabilized with 0.5% Triton X-100 for 20 min. After being blocked with 10% goat serum at room temperature for 60 min, tissue sections were incubated overnight with primary antibody (anti-Collagen IV) at 4 °C. After being washed three times, tissue sections were incubated in the dark with fluorescent conjugated secondary antibody for 1 h at room temperature. Stained samples were covered with DAPI-containing anti-fade mounting media (Vector labs, H1200-10) and imaged with a Nikon upright epi fluorescence microscope.
ECM extraction and mass spectrometry analysis
To enrich the ECM proteins of adipocyte tissues, we isolated ECM proteins from adipocyte tissues using the Compartment Protein Extraction Kit (Sigma, 2145) [28]. In general, WAT(100 mg) from mouse mammary glands were homogenized in 500 µl of Buffer C containing protease inhibitors using a tissue homogenizer until the tissue is completely disrupted. After sequential extraction of intracellular soluble proteins (cytosolic proteins, nuclear proteins, membrane proteins, the cytoskeletal proteins), the ECM-enriched pellets were suspended by adding the appropriate volume of 8 M urea with 10mM DTT. The ECM-enriched samples were alkylated by iodoacetamide, deglycosylated by PNGaseF, as well as digestion by Lys-C and trypsin. The digested samples were quenched with formic acid, concentrated and desalted with Ziptip before injected for MS analysis with a 150 min gradient adapted from literature. MS datasets were searched with Proteome Discoverer (Thermo Scientific) and Mascot version 1.3 (Matrix Science). Mascot search parameters were 10 ppm mass tolerance for precursor ions and 0.8 Da for fragment ions; two missed cleavages of trypsin with dynamic modifications including carbamidomethylation of cdegested ysteine, oxidized methionine, deamidation of asparagine, pyro-glutamic acid modification at N-terminal glutamine, and hydroxylation of lysine and proline.
Immunoprecipitation assay
Hsp47 expression plasmid pCDH1-Hsp47 and asporin expression plasmid CDH1-Asporin (WT)-Flag/pCDH1-Asporin (MT1)-Flag/pCDH1-Asporin (MT2)-Flag transfected HEK293 cells were lysed with ice cold hypotonic gentle lysis buffer (10 mM Tris-HCl [pH 7.5], 10 mM NaCl, 2mMEDTA, 0.5% Triton X-100, phosphatase and protease inhibitor cocktails (EMD Millipore, 539131) and incubated on ice for 10 min. 5 M NaCl was added in the cell lysis to 150mM concentration. The protein complexes were pulled down with anti-Flag M2 affinity gel (Sigma Aldrich, A2220), and the immunoprecipitated protein was eluted for immunoblotting analysis.
DSP experiments
Formalin-fixed paraffin embedded tissue sections were analyzed using Nanostring’s GeoMx Digital Spatial Profiling platform. All steps were carried out per Nanostring’s demonstrated protocol following manufacturer’s instructions. Slides underwent deparaffinization and antigen retrieval, followed by overnight incubation with Mouse Immune Cell Profiling Protein Core (Nanostring #GMX-PROCO-NCT-MICP-12), as well as anti-pan cytokeratin-AF532 and anti-CD45-AF594 (Nanostring #GMX-PRO-MORPH-MST-12) to visualize tissue sections and aid in region of interest selection. After incubation, slides were post-fixed, and nuclei were stained with Syto 13 before loading slides on the GeoMx instrument for whole slide fluorescent imaging. Regions of interest were selected from each sample and the nucleotide barcodes for panel antibodies bound in the region of interest were cleaved via UV exposure, then aspirated and transferred to a 96 well plate. The sample plate was processed per manufacturer’s instructions and analyzed on the Nanostring Sprint platform. The raw DSP data was normalized by the expression of the housekeeping protein Histone H3 within each ROI, then log2-transformed for downstream statistical analysis. Protein expression profiles of Hsp47 KO vs. Hsp47 pos groups within the tumor-adjacent (T-A) adipose tissue were visualized by heatmap. The linear mixed model with a fixed effect on Hsp47 genotype and a random intercept was applied to compare the Hsp47 KO vs. Hsp47 pos subgroups. The log2-fold change and p-value estimated from the linear mixed model were used to generate volcano plots within tumor and T-A adipose tissues, respectively, where the dots in red represent the proteins with p-value less than 0.05. Boxplots of proteins of interests were generated to manifest the direction of change between the Hsp47 KO vs. Hsp47 pos subgroups.
Quantification and statistical analysis
All experiments were repeated at least three times. Results were reported as mean ± standard error of the mean (S.E.M), the significance of difference was assessed by independent Student’s t-test or one-way analysis of variance (ANOVA) with SigmaPlot (Systat Software) or SPSS Statistics (IBM, Statistics). P < 0.05 is considered statistical significance, and P < 0.01 represented sufficient statistical significance. All reported P values were from 2-tailed tests.