The crosstalk between ECM-educated macrophages and cancer cells

Marques-Magalhães Â.1,2,3, Pinto M.L.4, Castro F.1,2, Bidarra S.1,2, Canão P.5, Barrias C.1,2,3, Carneiro F.1,5,6, Cardoso A.P.1,2*, Gonçalves R.1,2,3*, Oliveira M.J.1,2,3,7*

1 i3S - Institute for Research and Innovation in Health, University of Porto; 2 INEB, Institute of Biomedical Engineering, University of Porto; 3 Institute for Biomedical Sciences Abel Salazar (ICBAS), University of Porto; 4 CNC - Center for Neuroscience and Cell Biology, University of Coimbra; 5 Department of Pathology, Centro Hospitalar Universitário São João, Porto; 6 IPATIMUP – Institute of Molecular Pathology and Immunology of the University of Porto; 7 Department of Pathology and Oncology, Faculty of Medicine, University of Porto; * Equally contributing last authors

For additional information, please contact: [email protected] INTRODUCTION ML Pinto et al., Biomaterials, 2017

A B C D

Figure 1. Macrophages differentiated in tumor dECM present a more anti-inflammatory M2-like phenotype. (A) Representative images of Scanning Electron Microscopy from normal and tumor repopulated ECM. (B) mRNA expression of pro-(CD80, CCR7, and TNF) and anti-inflammatory (MRC1, CD163, TGFb1, CCL18, VCAN and MMP1) evaluation by qRT-PCR. (C) Secretion levels of CCL18 evaluated by ELISA in conditioned media of normal and tumor repopulated ECM. (D) Quantification of matrigel invasion assays where RKO cells were incubated in the upper compartment, while recombinant human CCL18 was incubated in the lower compartment, in the presence or not of the anti-CCL18 neutralizing antibody. Abbreviations: dECM – decellularized extracellular matrix

MAIN GOAL Identify the tumor ECM cues that regulate CCL18 production by macrophages and CCL18 cellular effects on colorectal cancer cells

 Marques-Magalhães, unpublished MATERIALS & METHODS

Extracellular Matrix Biochemical Analysis

A) Proteomics-based approach: Mass Spectrometry (LC-MS/MS)

DECELLULARIZATION PROTOCOL Day 1

- 3x 4with mm PBSdiameter 1x, RT, biopsy 165 cylinder rpm, 1h punch each - Hypotonic Buffer A, RT, 165 rpm, overnight ⤸

Day 2 ⤸ - 3x SDSwith 0.1% PBS, RT,1x, 165 RT, rpm, 165 24hrpm, 1h each

⤸ Day 3

- 3xDNAse with PBStreatment 1x, RT,, 37 165℃, rpm,165 rpm, 1h each3h

⤸ ⤸ B) Educated-guess approach: Hyaluronidase Treatment HA: hyaluronic acid

Confocal RESULTS & DISCUSSION

A) Proteomics-based approach: Identification of the ECM components that induce macrophage-CCL18 expression

1. A. Naba (ZipTips) protocol was the most suitable to obtain 2. Fifty-four candidates were found differentially expressed between an ECM enrichment normal and tumor human decellularized matrices

Human matrices decellularization (dECM) 54 potential candidates Only in Tumor Only in Normal AEBP1 MEST ABCA8 FASN ATP2B4 NLRX1 ACADM HLA-DRB1 BASP1 OCIAD1 ADH5 HMGCL CD63 PRSS8 CAVIN1 LONP1 NORMAL dECM TUMOR dECM CERS6 SEC62 CCT7 RPL22 DAVIDB Bioinformatic Analysis CLIC1 SLC25A1 ELAVL1 SAMHD1 ONTOLOGY TERMS CLUSTERING CSRP2 SLC25A17 ETFB SORBS1 DYNLL1 SLC44A1 EFEMP2 THBS4 Integrin Cell-matrix FAP THEM6 Binding Adhesion FBLN1 VMP1 GOLT1B WISP2 COP9 Plasma H2AFV ZYX Signalosome Membrane Extracellular Increased in Tumor Increased in Normal Signal Transduction Space EMP1 COL4A5 OGN Cell CXCL12 DCXR PDHB Adhesion PXN EPB41L2 SYNCRIP GPA33 WNT2B GPX1 ZG16 HBD INDIVIDUAL TERMS COP9 Mitochondrion Signalosome Figure 2. Venn Diagram of normal and tumor dECM. Number of identified in normal and Figure 3. Fifty-four candidates were found differentially expressed between normal and tumor human tumor samples (n=1) after LC-MS/MS analysis with Proteome Discoverer 2.4 software. MS/MS decellularized matrices. (A) ECM proteins differentially expressed in normal and tumor dECM (n=3) with spectra were searched against UniProt databases. ECM total, ECM core matrisome and ECM A. Naba (ZipTips) protocol. (B) Bioinformatic analysis was performed with Database for Annotation, associated-matrisome classification was based on Naba A, et al., Matrix Biol. 2016. Visualization and Integrated Discovery (DAVID) software, version 6.8. RESULTS & DISCUSSION B) Educated-guess approach: Evaluation of the hyaluronic acid (HA) effect on macrophage-CCL18 expression 1. Hyaluronidase (HAase) treatment efficiently reduced dECM-hyaluronic acid DAPI HABP Merge

Figure 4. Hyaluronidase (HAase) treatment efficiently reduced dECM-hyaluronic acid. Tumor decellularized rd Negative Control matrices without HAase exhibited HABP staining and no DAPI (3 row). HAase treatment decreased HABP staining, showing that HAase is successfully removing the HA from the tumor matrices (4th and 5th rows). As positive and negative staining controls, native tumor tissues from colon were used (1st and 2nd rows). Tumor Native Tissue Tumor Positive Control

Ongoing: Evaluate the impact of HA on macrophage-CCL18 expression HAase no no dECM Tumor Tumor dECM

RT-qPCR ELISA HAase dECM Tumor Tumor dECM 1.5 mg/ml 1.5 mg/ml HAase dECM Tumor Tumor dECM 1.8 mg/ml 1.8 mg/ml hyaluronic acid-binding protein (HABP) CONCLUSION & FUTURE PERSPECTIVES

Molecules associated with A B CCL18-mediated invasion Therapy Relevant ECM components Therapy

N=68

Figure 5. CCL18 is highly expressed at the invasive front of more advanced CRC cases. (A) Sixty- eight colorectal cancer (CRC) cases were immunohistochemically stained with anti-CCL18 antibody and scored accordingly to the staining at the invasive front: absent/few positive cells (score 1), less than 40% (score 2), between 40 and 80% positivity (score 3) and more than 80% (score 4). Arrows indicate positive areas (scale bar = 100 mm). (B) The association of CCL18 expression with tumor stages I and II vs. III and IV was assessed using a two-sided Fisher's exact test (*P = 0.033) Through this strategy, we expect to identify the ECM components responsible for and multivariate ordinal regression (*P = 0.035). macrophage-CCL18 production over tumor decellularized matrices and CCL18 cellular ML Pinto et al., Biomaterials, 2017 effects on colorectal cancer cell invasion.

We believe that this approach will contribute for the development of novel therapeutic ↑ number of colorectal cancer cases to 150 strategies to modulate the expression of this immunosuppressive and pro-invasive (ongoing) chemokine at the TME.

ACKNOWLEDGMENTS