CXCL12-CXCL4 heterodimerization prevents CXCL12-driven breast cancer cell migration

Journal Pre-proof CXCL12-CXCL4 heterodimerization prevents CXCL12-driven breast cancer cell migration

Khanh T.P. Nguyen, Lawrence J. Druhan, Belinda R. Avalos, Li Zhai, Lubica Rauova, Irina V. Nesmelova, Didier Dréau PII:

S0898-6568(19)30284-0

DOI:

https://doi.org/10.1016/j.cellsig.2019.109488

Reference:

CLS 109488

To appear in:

Cellular Signalling

Received date:

29 August 2019

Revised date:

24 November 2019

Accepted date:

26 November 2019

Please cite this article as: K.T.P. Nguyen, L.J. Druhan, B.R. Avalos, et al., CXCL12-CXCL4 heterodimerization prevents CXCL12-driven breast cancer cell migration, Cellular Signalling(2019), https://doi.org/10.1016/j.cellsig.2019.109488

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier.

Journal Pre-proof CXCL12-CXCL4 heterodimerization prevents CXCL12-driven breast cancer cell migration Authors: Khanh T.P. Nguyen1 , Lawrence J. Druhan2,3 , Belinda R. Avalos2,3 , Li Zhai4 , Lubica Rauova4,5, Irina V. Nesmelova3,6 , Didier Dréau1,3 Institutions: 1 Department of Biological Sciences, UNC Charlotte, Charlotte, NC 2

Department of Hematologic Oncology & Blood Disorders, Levine Cancer Institute, Atrium

of

Health, Charlotte NC CBES, UNC Charlotte, Charlotte NC

4

Department of Pediatrics, The Children’s Hospital of Philadelphia, PA

5

Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania,

-p

ro

3

Department of Physics and Optical Science, UNC Charlotte, Charlotte, NC

lP

6

re

Philadelphia, PA

na

This work was supported in part by grants from the Center for Biomedical Engineering and

ur

Science (CBES) at the University of North Carolina at Charlotte to IVN, DD, and BRA (Biological data collection and analysis) and by a research grant from the American Heart

Jo

Association (#10BGIA4170155) to IVN (NMR data collection and analysis). In addition, KTP Nguyen was supported in part by a Graduate School Summer Fellowship awarded by the University of North Carolina at Charlotte. *Address for correspondence and reprint requests to Dr. Didier Dréau, Department of Biological Sciences, UNC Charlotte 9201 University City Charlotte NC USA, [email protected] The online version of this article contains supplemental materials Abbreviations used in this article: NMR: Nuclear Magnetic Resonance

Journal Pre-proof Abstract: Despite improvements in cancer early detection and treatment, metastatic breast cancer remains deadly. Current therapeutic approaches have very limited efficacy in patients with triple negative breast cancer. Among the many mechanisms associated that contribute to cancer progression, signaling through the CXCL12-CXCR4 is an essential step in cancer cell migration. We previously demonstrated the formation of CXCL12-CXCL4 heterodimers (Carlson et al., 2013).

of

Here, we investigated whether CXCL12-CXCL4 heterodimers alter tumor cell migration.

ro

CXCL12 alone dose-dependently promoted the MDA-MB 231 cell migration (p<0.05), which

-p

could be prevented by blocking the CXCR4 receptor. The addition of CXCL4 inhibited the

re

CXCL12-induced cell migration (p<0.05). Using NMR spectroscopy, we identified the CXCL4CXCL12 binding interface. Moreover, we generated a CXCL4-derived peptide homolog of the

lP

binding interface that mimicked the activity of native CXCL4 protein. These results confirm the

na

formation of CXCL12-CXCL4 heterodimers and their inhibitory effects on the migration of breast tumors cells. These findings suggest that specific peptides mimicking heterodimerization

Jo

ur

of CXCL12 might prevent breast cancer cell migration.

Keywords: breast cancer cell migration, CXCL12, CXCR4, CXCL4, CXCR3, NMR

Journal Pre-proof 1. Introduction Chemokines are proteins that regulate various cellular processes in normal and pathophysiological conditions [1, 2]. In the tumor microenvironment, chemokines are key signaling molecules between the diverse cells constituting the tumor mass, including immune, endothelial, and tumor cells [3]. Through binding to G protein-coupled receptors, chemokines activate multiple secondary messengers modulating different intracellular signaling cascades

of

leading to alterations in cell trafficking and cell migration - key steps to immune responses,

ro

tumorigenesis and metastasis [4].

-p

Breast cancer progression, especially metastasis, relies on breast cancer cell migration

re

orchestrated mainly through CXCL12-CXCR4 chemokine-receptor signaling [5-8]. The role of CXCL12-CXCR4 signaling in cell migration has been clearly demonstrated in breast cancer cell

lP

models including in the triple negative breast cancer MDA-MB 231 cells [9-11]. CXCR4 tumor

na

cell expression associated with high constitutive secretion of CXCL12 in organs, including lungs, bone marrow, and liver generates a gradient that promotes cell migration and targeted

ur

metastasis [11]. In addition to cancer cell migration, the CXCL12-CXCR4 axis also plays a

Jo

critical role in maintaining cancer cell stemness and in the down regulation of the expression of tumor cell markers [12-14]. Moreover, clinically, the overexpression of CXCR4 on tumor cells is associated with the most aggressive form of breast cancer i.e., triple negative breast cancer [5, 15, 16]. The migration of breast cancer cells has been targeted to prevent cancer progression using CXCR4 chemokine receptor antagonists [17, 18], nanoparticles [19] and oncolytic virotherapy [20] to prevent cancer progression. However, the results of clinical trials using CXCR4 antagonist-based approaches have been mixed [21, 22], possibly because of the multiple forms of CXCL12 signaling through the CXCR4 receptor [22], and of the key role of CXCL12-

Journal Pre-proof CXCR4 signaling in normal immune cell trafficking [23-25]. Thus, new targeted approaches are needed to prevent breast cancer progression. Chemokines exist as monomers and can form homodimers, which either enhance or inhibit their respective monomer signaling [26-28]. In addition, many chemokines form heterodimers [29-36]. Using immuno-ligand blotting, von Hundelshausen et al. recently screened 45 human chemokines and identified over 200 distinct binary heterophilic chemokine interactions [37]. The

of

response from cells treated with chemokine mixtures differs from the cellular responses to

ro

individual chemokines or chemokines in various combinations, and can lead to synergistic

-p

enhancement or reduction [30, 33, 37]. For example, cell chemotaxis enhanced by monomeric

re

CXCL12 was inhibited by a mixture of CXCL4 and CXCL12 chemokines [37], suggesting that chemokine heterodimers may be responsible for the observed altered activity. This finding

lP

suggests that the CXCL4 chemokine may interfere with CXCL12-CXCR4 signaling, or,

na

alternatively, CXCL12 chemokine may interfere with CXCL4-mediated signaling. It has been shown that CXCL4 binds with greater affinity to the CXCR3B receptor isoform than to

ur

CXCR3A [38, 39] and promotes anti-tumorigenesis and apoptosis in tumor cells [38]. This

Jo

signaling contrasts with signaling induced by the activation of the CXCR3 receptor isoform CXCR3A by the chemokines CXCL9, CXCL10, and CXCL11 that promotes tumor growth, chemotactic migration, invasion, and metastasis [38, 40]. Whether CXCL12 affects CXCL4CXCR3 signaling is unclear. To unambiguously identify the activity of a chemokine heterodimer in the mixture of chemokines where competing homophilic and heterophilic interactions lead to an equilibrium of coexisting of species (monomers, homodimers, and heterodimers), non-dissociating CXCL4CCL5, CCL5-CCL17, and CXCL7-CXCL1 heterodimers have been generated and functionally

Journal Pre-proof assessed [36, 37, 41]. The CXCL4-CCL5 and CCL5-CCL17 heterodimers induced cellular responses with higher potency and efficacy than combinations of the individual chemokines [37, 41]. In contrast, CXCL7-CXCL1 heterodimers activated CXCR2-driven cellular responses similarly to either CXCL1 or CXCL7 chemokines alone [36]. These studies support the concept of active chemokine heterodimers in vivo and suggest that the heterodimer mechanism may differ and be chemokine-specific. These data also highlight the need to functionally assess other

of

chemokine heterodimers in biological conditions, in particular in cancer [3]. Previously, we have

ro

shown that CXCL12 and CXCL4 form heterodimers using electrospray ionisation mass

-p

spectrometry (ESI-MS) experiments and co-immunoprecipitation of CXCL12-CXCL4

re

heterodimers from human platelets [35]. While the regulation of breast cancer cell migration through monomeric CXCL12 and CXCR4 is well established in patients, animal and in vitro cell

lP

models, whether CXCL12-CXCL4 heterodimers influence breast cancer cell migration is

na

unknown. Here we determined the effects of CXCL12-CXCL4 heterodimers on breast cancer

Jo

model [42].

ur

cell migration in the well-established triple negative MDA-MB 231 breast cancer cell in vitro

Journal Pre-proof 2. Materials and methods: 2.1.MDA-MB 231 cells and CXCL12-driven migration using wound-healing assays MDA-MB 231 (ATCC, Manassas, VA) were cultured in DMEM/F12 media (Corning) supplemented with 10% FBS (Atlanta Biologics), L-glutamine, Amphotericin B and Gentamycin (Corning). Briefly, following an overnight coating with Collagen type I (12ug/cm2 , BD Biosciences) at 37o C with 5% CO 2 and >85% humidity and washes of the unbound Collagen I

of

with sterile PBS, 96-well tissue culture plates (Greiner) were seeded (40,000 cells/well) with

ro

MDA-MB 231 cell suspension in culture media. Cells were grown to confluency and then

-p

incubated overnight with fresh media without FBS (0%). Cells were identified through the

re

addition of the vital nuclear dye Hoechst (Promega). The confluent MDA-MB 231 cell monolayers were then scratched using a sterile pipet tip and the wells washed to remove non-

lP

adherent cells. Thereafter, cells were incubated with concentrations of CXCL12 (0-100nM;

na

Shenandoah Biotechnology Inc), CXCL4 (0-200nM; Shenandoah Biotechnology Inc) chemokines and/or peptides diluted with 0% FBS media as described. Overlapping

ur

microphotographs encompassing the entire area of each scratch/wound were taken at the start of

Jo

the treatment (0 hr.) and following a 24-hour incubation (24 hrs.) using an IX71 Olympus microscope equipped with a DP70 camera and the associated software (Olympus). Overlapping microphotographs were stitched together, and the area of the wound was determined using ImageJ software (NIH). After normalization to the area measured at time 0, results were expressed as percentage of wound healing.

Journal Pre-proof 2.2.CXCR4 and CXCR3 receptor expression and inhibition The presence of CXCR4 and CXCR3b receptors was determined by flow-cytometry as previously described [43, 44]. Briefly, MDA-MB 231 cells were grown and treated with vehicle (media as above); the CXCR4 inhibitor AMD3100 (MedChem Express) or the CXCR3 inhibitor AMG487 (MedChem Express) as above and harvested using trypsin and cell dissociation solution (Corning Inc.). After wash and blocking in BSA 1% solution, cells were incubated with

of

either rat anti-human CXCR4 antibody (R&D Systems, Minneapolis, MN) or mouse anti-human

ro

CXCR3b antibody (ProteinTech Inc., Rosemont, IL) for 2 hrs. Following more washes, cells

-p

were incubated with FITC-conjugated anti-rat or FITC-conjugated anti-mouse antibody, as

re

appropriate. After additional washes, the presence of FITC was determined by flow-cytometry (Fortessa FACS flow-cytometer; BD Biosciences). Data were analyzed using FlowJo software

lP

(Ashland, OR).

na

To investigate whether migration was critically dependent of either the ligand-receptor CXCL12CXCR4 or the ligand-receptor CXCL4-CXCR3 signaling pathways, wound healing assays were

ur

conducted in the presence of 20nM of CXCR4 inhibitor AMD3100 or 5nM of the CXCR3

Jo

inhibitor AMG487. The optimal concentrations of AMD3100 and AMG487 were defined in preliminary experiments and derived from previous work [11, 17, 45]. Wound healing assays with CXCL12 and/or CXCL4 following pre-treatment with either AMD3100 and/or AMG487 were conducted and evaluated as detailed above. 2.3. Uniformly

15

NMR spectroscopy

N-enriched CXCL4 was expressed and purified as described previously [22].

Briefly, wild-type (WT) human PF4 (CXCL4) in the pT7-7 was expressed in BL21DE30 pLysS [46].

15

N-CXCL4 was isolated from the supernatant of the bacterial lysate grown in M9 media

Journal Pre-proof (3% KH2 PO4, 12.8% Na2 HPO 4 *7H2 O, 0.5% NaCl, 1% 15 NH4 Cl) by affinity chromatography using a HiTrap Heparin high-performance (HP) affinity column. Eluted proteins were purified further by fast protein liquid chromatography (FPLC) using a Resource RPC FPLC column. Protein purity was assessed by 12% NuPAGE Bis-Tris gel electrophoresis (Thermofisher). Protein concentrations were determined according to the manufacturer's instructions using the bicinchoninic acid assay (Pierce) with BSA as standard.

15

N- CXCL4 was dissolved in a H2 O/D2 O (95%/5%) mixture containing 20

ro

further purification.

of

CXCL12 was purchased from Shenandoah Biotechnology Inc. (Warwick, PA) and used without

-p

mM NaCl at the concentration of 1 mg/ml. pH was adjusted to 5.0 by adding microliter

re

increments of 0.1M HCl. A series of two-dimensional 1 H- 15 N HSQC (heteronuclear single quantum coherence) spectra were collected by titrating unlabeled CXCL12 to

15

N-CXCL4

lP

solution at 1:1 and 2:1 molar ratio and chemical shift changes of 15 N-CXCL4 caused by the 15

N and 1 H were positioned at

na

interactions with CXCL12 were monitored. Carrier frequencies for

116.5 and 5.2 ppm, respectively. All NMR experiments were carried out at 313 K on a Bruker

ur

Advance-III 950 MHz spectrometer at David H. Murdock Research Institute (Kannapolis, NC).

Jo

Raw data were converted and processed using NMRPipe [47] and analyzed using NMRview [48] software. CXCL4 resonance assignments have been reported previously [31, 49]. The spectra collected in this work were identical to those published previously [31], allowing NMR chemical shift assignments. 2.4.

CXCL4-derived peptide

A CXCL4-derived peptide matching the putative binding interface with CXCL12 was derived from our NMR studies and synthetized AHITSLEVIKAG (Pepmic Co, Suzhou, China). The peptide at increasing concentrations (0-500nM) was used in combination with 50nM CXCL12 to

Journal Pre-proof investigate the effects of peptide/CXCL12 on MDA-MB 231 cell migration. These experiments were repeated in the presence of AMG487 (5nM). 2.5.

Alterations in RNA expression

MDA-MB 231 cells were cultured as described above in the presence of either vehicle (Ctrl), CXCL12 (100 nM), CXCL4 (100 nM) or the combination CXCL4 (100 nM)-CXCL12 (100 nM) for 3 hrs. The cells were then harvested and total RNA isolated through a one-step lysis (Isol-

of

RNA, ThermoThermo Fisher Scientific, Waltham, MA USA). RNA expression was determined

ro

using Affymetrix protocols, reagents and platform using RNA expression microarrays with

-p

coverage of the Human Genome U133 Set plus 6,500 additional genes for analysis of over

re

47,000 transcripts (GeneChip™ Human Genome U133 Plus 2.0 Array; ThermoScientific). The features of the microarray used have been detailed earlier (for more details,

lP

https://www.thermofisher.com/order/catalog/product/900466). The microarray experiments were

na

performed and analyzed by the Molecular Biology Core Facility at The Levine Cancer Institute, Atrium Health (Charlotte NC). RNA expression changes defined as average RNA expression of

ur

cells treated with CXCL12 (100nM) 15% higher or 15% lower (i.e., >115% or <85%) compared

Jo

to the average of control cells (no treatment) are presented. An organized heat map of grouping genes with related functions to highlight changes associated with treatments is provided. Additionally, a dendrogram highlighting the proximity of expression was generated using open source versions of Cluster 3.0 and TreeView software. 2.6.

Genomics alterations and overall survival in breast cancer patients

Cancer genome data were accessed from the cBioPortal (www.cbioportal.org) for Cancer Genomics [50, 51]. Twelve breast cancer data sets were used to depict the genetic alterations (mutations; amplifications, or deep deletions) in CXCL12, CXCL4 chemokine and their

Journal Pre-proof receptors CXCR4 and CXCR3. In addition, the same data sets were used to evaluate differences in overall survival (OS) between all patients and patients with genetic alterations affecting either the CXCL12-CXCR4 pathway or the CXCL4-CXCR3 pathway through log-rank analysis. The number of patients analyzed, patients deceased, and median overall survival (OS in months) are reported. Detailed information regarding published datasets and associated publications are presented along with the analysis. Statistical analyses

of

2.7.

ro

All data are presented as mean  SEM. Differences between chemokine conditions were assessed

-p

using one-way ANOVA and Tukey post-hoc tests. For RNA expression gene comparison of the

re

microarray data, p values were adjusted to account for multiple comparisons. For patient cohort analyses, data were presented using Kaplan-Meier plots and log-rank tests were used to compare

Jo

ur

na

lP

overall survival. Statistical significance was set at p < 0.05.

Journal Pre-proof 3. Results 3.1.

CXCL12 promotes and CXCL4 inhibits MDA-MB 231 breast cancer cell

migration CXCL12 stimulates the migration of MDA-MB 231 breast cancer cells uniquely through CXCL12-CXCR4 signaling [52-54]. Therefore, we first confirmed that CXCL12 stimulates the migration of MDA-MB 231 breast cancer cells using wound-healing assays. As expected, the

of

CXCL12-driven migration of MDA-MB 231 cells was dose-dependent and increased as the

ro

concentration of CXCL12 increased in the in vitro migration assays (p<0.05; Fig. 1A). In

-p

particular, when the tumor cells were treated with 100nM of CXCL12, a two-fold migration

re

increase was observed as reported previously [55]. The presence of CXCR4 receptor on the cell membrane of MDA-MB 231 cells was confirmed by flow-cytometry (Fig 1B). The activation of

lP

the CXCL12-CXCR4 signaling pathway leading to MDA-MB 231 cell migration was verified by

na

inhibition of cell migration in the presence of the CXCR4 antagonist AMD3100 (20nM, p<0.05; Fig. 1C). Incubation with AMD3100 alone at concentrations ranging from 5nM to 25nM had no

ur

effect on MDA-MB 231 cell migration (data not shown).

Jo

Next, in similar assays, we assessed the effect of CXCL4 on MDA-MB 231 cell migration. No effect was observed in the presence of 50nM of CXCL4 (Fig 1D). However, the addition of CXCL4 at a higher concentration (100nM) led to ~2-fold decrease of MDA-MB 231 cells migration (p<0.05; Fig 1D). CXCR3B receptors were present on MDA-MB 231 cells (Fig 1E), and their inhibition using the CXCR3 inhibitor AMG487 (5nM) prevented CXCL4-CXCR3B activation and the associated inhibition (Fig 1F). 3.2.The CXCL12 and CXCL4 chemokine mixture inhibits MDA-MB 231 breast cancer cell migration through inhibition of CXCL12-CXCR4 signaling

Journal Pre-proof CXCL12 and CXCL4 readily form heterodimers in vitro [35, 37]. To assess the biological activity of the CXCL12-CXCL4 chemokine mixture, combinations of CXCL12 and CXCL4 at different ratios were tested on MDA-MB 231 cell migration in wound-healing assays (Fig 2A). In these experiments, the concentration of CXCL12 was kept constant at 100nM, whereas the concentration of CXCL4 varied (0 to 200nM, Fig 2B-C). Unlike CXCL12 alone, the incubation with CXCL12 and CXCL4 mixtures significantly and dose-dependently reduced the MDA-MB

of

231 cell migration (Fig. 2BC, p<0.01). The effect observed with CXCL12-CXCL4 mixtures is

ro

enhanced resulting in a greater than 2.5 times reduction compared to CXCL4 alone, suggesting

-p

that the response is synergistic rather than additive. These data show that CXCL4 effectively

re

inhibits the CXCL12-induced MDA-MB 231 cell migration, and the effect is more potent than when it is added alone, suggesting that the CXCL12-CXCL4 interaction could be, at least in part,

lP

responsible for the observed alteration in migration.

na

3.3.The CXCL12-CXCL4 chemokine mixture prevents MDA-MB 231 breast cancer cell

signaling

ur

migration through CXCR4 receptor signaling and does not affect the CXCR3

Jo

We tested whether the synergistic effect of the CXCL12-CXCL4 mixture on MDA-MB-231 cell migration is CXCR4 receptor mediated. The addition of the CXCR4 inhibitor AMD3100 (20nM) blocked the migration of MDA-MB-231 cells following incubation with either 50 or 100nM CXCL12 (p<0.05; Fig 1C) as well as with the mixture of CXCL12 (100nM) and CXCL4 (50 or 100nM) (Fig 3A) confirming that CXCL12-driven migration of MDA-MB 231 cells occurs via the activation of the CXCR4 receptor. As CXCL4-CXCR3 signaling can inhibits cell migration [56], we tested whether the CXCR3 receptor inhibitor AMG487 could reverse the effect of CXCL4 on MDA-MB 231 cells migration

Journal Pre-proof induced by CXCL12 (Fig 1A & 2B). At the concentration used, AMG487 had no effect on the 2fold increase in MDA-MB 231 cell migration induced by CXCL12 (100nM; p<0.001; Fig 3A&B). Moreover, regardless of the presence of AMG487, the CXCL12-CXCL4 mixture significantly prevented MDA-MB 231 cell migration (p<0.001; Fig 3B). The observation that 5nM of AMG487 blocks the CXCR3 associated signaling (see Fig 1F) suggests that the inhibition observed in the presence of the CXCL12- CXCL4 mixture is likely due to inhibition of

of

the CXCL12-CXCR4 signaling.

ro

3.4.The CXCL12-CXCL4 chemokine mixture alters the gene expression profile

-p

associated with CXCL12 stimulation towards a profile closely mimicking control

re

conditions

lP

To gain further insight into the changes in RNA expression and pathways altered following treatment with individual chemokines (CXCL12 and CXCL4) or the CXCL12-CXCL4

na

chemokine mixture, gene expression was analyzed in MDA-MB 231 cells incubated with the vehicle alone, individual chemokines, or the chemokine mixture. No drastic change was

ur

observed (Fig. 4). RNA transcripts up-regulated and downregulated in MDA-MB 231 cells

Jo

incubated in the presence of CXCL12 alone contrasted with RNA expression observed in MDAMB 231 cells treated with vehicle alone, CXCL4 alone, or the CXCL12 + CXCL4 mixture (Fig 4). Following a 3-hrs incubation with 100nM of CXCL12, 101 transcripts were expressed at least 15% above or below the controls. Moreover, 29 and 21 of the transcripts had expression significantly increased and decreased, respectively, compared to control conditions (Fig. 4, P<0.05). Notably, CXCL12-induced upregulation of four transcripts, MYOZ3; KNG1; TMEM237 and AGRN involved in calcineurin, kininogen, WNT, and epidermal signaling,

Journal Pre-proof respectively, was abrogated in the presence of the CXCL12-CXCL4 mixture (Fig. 4, P<0.05). Similarly, the CXCL12-induced downregulation of three transcripts, the GTP-binding protein GTPBP2, the von Willebrand factor A domain containing protein involved in the formation of filamentous networks in the extracellular matrices MATN2, and the sodium-potassium-chloride cotransporter SLC12A1, respectively, was abrogated in the presence of the CXCL12-CXCL4 mixture (Fig. 4, P<0.05). Interestingly, mutations in MYOZ3, KNG1, GTPBP2 and/or MATN2

of

have been associated with a decrease in overall survival in breast cancer patients (p<0.01; data

ro

not shown) according to queries using the cBioportal for cancer genomics (cBioportal.org as

-p

detailed in the supplemental material: similar analysis using the same cohort of 8772 patients:

re

Suppl. Fig 2S).

lP

3.5.Binding interface between CXCL4 and CXCL12 chemokines defined by NMR spectroscopy

na

To define the binding interface between CXCL12 and CXCL4, NMR titration experiments

ur

were performed where unlabeled CXCL12 was added into solution of uniformly

15

N-labeled

CXCL4 at two molar ratios 1:1 and 1:2 (Suppl. Fig 1S). The interactions between CXCL12 and N-labeled CXCL4 were assessed through evaluation of chemical shift changes in

Jo

15

15

N-CXCL4

monitored by recording [1 H- 15 N]-HSQC spectra upon addition of CXCL12 (Fig. 5A-B). The [1 H15

N]-HSQC spectrum of CXCL4 alone shows multiple peaks of varying intensities for multiple

amino acid residues suggesting (1) the co-existence of monomer, dimer, and tetramer states in a slow exchange on the NMR experimental time scale [57, 58], and (2) the asymmetry of the tetramer [59]. Consequently, some amino acid residues could not be unambiguously assigned. Nevertheless, the residues that could be definitively assigned were well distributed throughout the protein, enabling the mapping of the CXCL4-CXCL12 interaction interface.

Journal Pre-proof Overall, addition of CXCL12 to the

15

N-labeled CXCL4 solution caused major chemical shift

changes of many amino acid peaks in the [1 H- 15 N]-HSQC spectrum of CXCL4, while some of amino acids remained minimally chemically altered (Figures 5A and Suppl. Figure 1S-A). Some cross-peaks merely shifted such as residues A32, E28, and T25. For other residues, the number of cross-peaks representing these residues was reduced as observed for G33 and G48, for which only two peaks remained following the CXCL12 addition to the CXCL4 solution. The latter

of

findings indicate a significant shift in the monomer-dimer-tetramer states CXCL4 equilibrium.

ro

Several new peaks were observed in the [1 H- 15 N]-HSQC spectrum of CXCL4 in the presence of

-p

CXCL12, possibly representing a new CXCL12-CXCL4 heterodimer state. While many CXCL4

re

resonances were affected by the addition of CXCL12, some resonances remained minimally affected, i.e., K66, T44, or I63, supporting the specificity of the interactions between CXCL4

lP

and CXCL12. The amino acid residues in CXCL4 demonstrating chemical shift changes above

na

(red) and below (blue) averages are highlighted on the CXCL4 monomer three-dimensional structure (Fig 5B, Suppl. Fig 2S-A-B), and on the amino acid sequence (Fig. 5C). Notably, the

ur

amino acid residues with large chemical shift changes are clustered on the first beta strand

Jo

suggesting that these residues may form the contact interface with CXCL12 (Fig 5B-C). This contact interface concurs with our previously proposed computational CXC chemokine model of CXCL12-CXCL4 heterodimers [35]. 3.6.The CXCL4-CXCL12 binding interface derived peptide AHITSLEVIKAG inhibits the CXCL12-stimulated MDA-MB 231 breast cancer cell migration To further confirm that CXCL4-CXCL12 interactions inhibit CXCL12-driven migration, we synthesized the CXCL4-derived peptide AHITSLEVIKAG mimicking the CXCL12 binding site of the CXCL4 chemokine (Fig 5C). This peptide dose-dependently decreased the CXCL12-

Journal Pre-proof induced migration of MDA-MB 231 cells (Fig. 6A; p<0.05). At a concentration of 500nM, addition of the peptide fully prevented the stimulatory effect of CXCL12. Additionally, activity of the peptide on CXCL12-CXCR4 driven MDA-MB migration was not significantly altered by

Jo

ur

na

lP

re

-p

ro

of

the presence of the CXCR3 inhibitor AMG 487 (Fig 6B).

Journal Pre-proof Discussion Chemokine signaling is attracting increasing interest in cancer biology as chemokines critically regulate cellular functions, including cell migration, and are highly expressed within the tumor microenvironment [60]. Most chemokines exist in a monomer-dimer equilibrium under physiological conditions and structural analyses demonstrated that chemokines can form homodimers [26, 58, 61, 62]. Moreover, monomers and homodimers can significantly modulate

of

signaling through their cognate receptors [26, 27, 63]. For some chemokines such as CXCL8, the

ro

dimerization is essential for receptor binding [64]. More recently, the formation of heterodimers

-p

has been demonstrated [30, 31, 35]. Here we assessed the biological relevance of CXCL12-

re

CXCL4 heterodimers in the CXCL12-CXCR4 driven migration, an essential step in breast cancer progression [9, 24]. Our observations in the MDA-MB 231 breast cancer cell migration in

lP

vitro model indicate that a mixture of CXCL12 and CXCL4, likely through the formation of

na

CXCL12-CXCL4 heterodimers, inhibits of the CXCL12-CXCR4 signaling independently of the CXCL4-CXCR3 signaling. Moreover, signaling through CXCL12-CXCR4 was associated with

ur

changes in multiple transcripts encoding proteins associated with cytoskeleton and membrane

Jo

reshaping, which were abrogated in the presence of the CXCL12-CXCL4 mixture. Using NMR spectroscopy, we identified a putative region of CXCL4 interacting with CXCL12 to form CXCL12-CXCL4 heterodimers and demonstrated that a CXCL4 peptide mimicking the putative CXCL4-CXCL12 binding sequence led to inhibition of MDA-MB 231 cell migration, similar to that observed with the CXCL12-CXCL4 mixture. CXCL12-driven tumor cell migration through CXCR4 signaling plays a critical role in breast cancer development [65]. Indeed, the CXCL12-CXCR4 signaling has been shown to be involved progression of more than 20 different types of human cancers that over-express CXCR4,

Journal Pre-proof including breast cancer [66]. CXCR4-expressing tumor cells have been shown to exhibit increased proliferation, migration, invasiveness, and ultimate metastasis to distant organs in part through their stem-like phenotype [67]. The significantly worse OS of patients with breast tumors containing genomic alterations, including amplification of CXCL12 and mutation of CXCR4 compared to patients without those genomic alterations (Fig 2S) suggests that CXCL12CXCR4 signaling is clinically relevant.

of

The MDA-MB 231 breast cancer cell line in vitro model has been used extensively to

ro

investigate cancer cell migration [9, 68] and the signaling events that promote cancer progression

-p

[10, 69]. In vitro analyses of breast cancer cells have demonstrated that following CXCL12-

re

CXCR4 activation induces signaling via the Rac1 protein, a member of the small GTPases superfamily that regulates the actin dynamics [70], as well as PI3K, and the Raf/MEK/ERK [71]

lP

and/or the JAK2/STAT3 pathway [72]. Notably, our RNA expression data indicate that CXCL12

na

promotes alterations in cytoskeleton-related transcripts and that the CXCL12-CXCL4 mixture inhibits the CXCL12-driven RNA expression. Regardless of the CXCL12 signaling cascade

ur

involved, the cytoskeleton changes induced by actin polymerization through the activation of

Jo

Rac1 and RhoA are critical for the formation of lamellopodia and stress fiber contraction, both important steps in tumor cell migration [69, 70]. Our data also demonstrate that incubation with CXCL4 at a concentration of 100nM reduced the migration of MDA-MB 231 cells by 30%, possibly through CXCR3B activation. Although activation of the CXCR3A receptor activation leads to pro-migratory responses and tumor metastasis, activation of the CXCR3B receptor isoform mediates tumor growth inhibition and induces apoptosis [73]. Importantly, the inhibitory effects of the CXCL12-CXCL4 mixture were observed with CXCR3 signaling inhibition, supporting the inhibition of CXCL12-CXCR4

Journal Pre-proof signaling by the CXCL12-CXCL4 mixture, possibly through CXCL12-CXCL4 heterodimers. Additionally, the mixture of a CXCL4-derived peptide mimicking the CXCL12-CXCL4 binding interface also suppressed CXCL12-driven tumor cell migration, further supporting the hypothesis that CXCL12-CXCL4 heterophilic interactions prevent tumor migration. The observation that the CXCL12-CXCR4 pathway (Fig 2S) may be altered by the formation of CXCL12-CXCL4 heterodimers likely has clinical relevance given the inflammatory milieu of

of

the tumor microenvironment resulting from the presence of multiple chemokines [74, 75].

ro

CXCL12 is secreted by tumor and immune cells [11, 76], while CXCL4 is also released from

-p

platelets at high micromolar concentrations [35], rendering the formation of CXCL12-CXCL4

re

more likely. Our data also highlight the therapeutic potential of CXCL4-mimicking peptides in preventing CXCL12-CXCR4 signaling and thus limiting breast tumor cell migrations. Although

lP

limited as suggested by our data (Fig 6), the effects of CXCL4-mimicking peptide(s) on the

na

CXCR3B signaling pathway should be clarified. Furthermore, as the CXCL12-CXCR4 and the CXCL4-CXCR3A/CXCR3B signaling mechanisms are essential for normal physiological

ur

functions including immune responses, further investigations are needed to validate the benefits

Jo

the treatment with CXCL4-mimicking peptide(s) in in vivo immuno-competent models of breast cancer progression.

Inhibiting chemokine signaling is extensively investigated [77]. In particular, inhibitions of signaling through CXCL1-CXCR2 [78] and CXCL12-CXCR4 [79] have shown promises in in vivo murine cancer models. However, so far clinical trials using various molecules directly targeting the CXC chemokine signaling have not been successful [80-82], likely because of the pleiotropic nature of the chemokine signaling. Nevertheless, as CXCL12 uniquely signal through the CXCR4 receptor [54], new approaches are being considered to modulate the CXCL12-

Journal Pre-proof CXCR4 signaling pathway targeting the ligand CXCL12, rather than the CXCR4 receptor, to prevent cancer progression [54, 79, 83]. Importantly, our observations provide evidence of the potential of CXCL4 peptides in the prevention of the CXCL12-CXCR4 signaling, an essential step in breast tumor progression. Whether targeting the CXCL12 ligand through the administration of optimized CXCL4 peptide(s) will prevent the CXCL12-CXCR4 signaling driven cancer progression remains to be ascertained in relevant in vivo breast tumor models.

of

Collectively, our data indicate that CXCL12-CXCL4 heterophilic interactions alter the

ro

CXCL12-CXCR4 signaling associated with breast tumor cell migration. These results warrant

-p

further studies of the molecular mechanisms by which chemokine heterodimers modulate

re

specific signaling events, in particular those associated with tumor cell migration. Inhibition of the CXCL12-driven tumor cell migration induced by a peptide mimicking the binding sequence

lP

of the CXCL12-CXCL4 chemokine heterodimers opens new avenues for chemokine targeted

Jo

ur

na

approaches to prevent cancer progression.

Journal Pre-proof Acknowledgments: The authors would like to acknowledge the help of undergraduate students and the Atrium Health Molecular Biology Core facility staff and Director Dr. N. Steuerwald (Charlotte NC) for conduction the microarray experiments. We thank Dr. Kevin Knagge at the David H. Murdock Institute for technical assistance with NMR experiments. NMR data collection and analysis was supported by a research grant from the American Heart Association (#10BGIA4170155) to IVN. Biological analyses were supported by seed grants from the Center

of

for Biomedical Engineering and Science at the University of North Carolina at Charlotte to IVN,

-p

ro

DD, and BRA.

re

Conflict of interest: The authors declare no conflict of interest.

lP

Author Contributions: KTPN and DD designed, conducted and analyzed the biological experiments. LJD and BRA participated in the analyses of the microarray studies. KTPN and

na

IVN designed, conducted and analyzed the biophysical investigations, aided by LZ and LR. DD,

ur

KTPN, and IVN completed the initial draft of the manuscript and all authors participated in the

Jo

review and revisions of the manuscript.

References: [1] D. Raman, T. Sobolik-Delmaire, A. Richmond, Chemokines in health and disease, Exp Cell Res 317(5) (2011) 575-89. [2] K. Chen, Z. Bao, P. Tang, W. Gong, T. Yoshimura, J.M. Wang, Chemokines in homeostasis and diseases, Cellular & Molecular Immunology 15(4) (2018) 324-334. [3] G. D'Agostino, V. Cecchinato, M. Uguccioni, Chemokine Heterocomplexes and Cancer: A Novel Chapter to Be Written in Tumor Immunity, Frontiers in immunology 9 (2018) 2185. [4] A. Ben-Baruch, The multifaceted roles of chemokines in malignancy, Cancer Metastasis Rev 25(3) (2006) 357-71. [5] A.L. Guembarovski, R.L. Guembarovski, B.K.B. Hirata, G.A.F. Vitiello, K.M. Suzuki, M.T. Enokida, M.A.E. Watanabe, E.M.V. Reiche, CXCL12 chemokine and CXCR4 receptor: association with susceptibility and prognostic markers in triple negative breast cancer, Mol Biol Rep (2018).

Journal Pre-proof

Jo

ur

na

lP

re

-p

ro

of

[6] R. Dayer, S. Babashah, S. Jamshidi, M. Sadeghizadeh, Upregulation of CXC chemokine receptor 4-CXC chemokine ligand 12 axis ininvasive breast carcinoma: A potent biomarker predicting lymph node metastasis, J Cancer Res Ther 14(2) (2018) 345-350. [7] S. Mukherjee, A. Manna, P. Bhattacharjee, M. Mazumdar, S. Saha, S. Chakraborty, D. Guha, A. Adhikary, D. Jana, M. Gorain, S.A. Mukherjee, G.C. Kundu, D.K. Sarkar, T. Das, Nonmigratory tumorigenic intrinsic cancer stem cells ensure breast cancer metastasis by generation of CXCR4(+) migrating cancer stem cells, Oncogene 35(37) (2016) 4937-48. [8] W. Wu, L. Qian, X. Chen, B. Ding, Prognostic significance of CXCL12, CXCR4, and CXCR7 in patients with breast cancer, Int J Clin Exp Pathol 8(10) (2015) 13217-24. [9] S.Y.N. Jamaludin, I. Azimi, F.M. Davis, A.A. Peters, T.J. Gonda, E.W. Thompson, S.J. Roberts-Thomson, G.R. Monteith, Assessment of CXC ligand 12-mediated calcium signalling and its regulators in basal-like breast cancer cells, Oncol Lett 15(4) (2018) 4289-4295. [10] Y. Sun, X. Mao, C. Fan, C. Liu, A. Guo, S. Guan, Q. Jin, B. Li, F. Yao, F. Jin, CXCL12CXCR4 axis promotes the natural selection of breast cancer cell metastasis, Tumour Biol 35(8) (2014) 7765-73. [11] P.J. Boimel, T. Smirnova, Z.N. Zhou, J. Wyckoff, H. Park, S.J. Coniglio, B.Z. Qian, E.R. Stanley, D. Cox, J.W. Pollard, W.J. Muller, J. Condeelis, J.E. Segall, Contribution of CXCL12 secretion to invasion of breast cancer cells, Breast Cancer Res 14(1) (2012) R23. [12] S. Inaguma, M. Riku, H. Ito, T. Tsunoda, H. Ikeda, K. Kasai, GLI1 orchestrates CXCR4/CXCR7 signaling to enhance migration and metastasis of breast cancer cells, Oncotarget 6(32) (2015) 33648-57. [13] K. Nobutani, Y. Shimono, K. Mizutani, Y. Ueda, T. Suzuki, M. Kitayama, A. Minami, K. Momose, K. Miyawaki, K. Akashi, T. Azuma, Y. Takai, Downregulation of CXCR4 in Metastasized Breast Cancer Cells and Implication in Their Dormancy, PLoS One 10(6) (2015) e0130032. [14] M.P. Ablett, C.S. O'Brien, A.H. Sims, G. Farnie, R.B. Clarke, A differential role for CXCR4 in the regulation of normal versus malignant breast stem cell activity, Oncotarget 5(3) (2014) 599-612. [15] K.A. Norton, A.S. Popel, N.B. Pandey, Heterogeneity of chemokine cell-surface receptor expression in triple-negative breast cancer, American journal of cancer research 5(4) (2015) 1295-307. [16] Q.D. Chu, L. Panu, N.T. Holm, B.D. Li, L.W. Johnson, S. Zhang, High chemokine receptor CXCR4 level in triple negative breast cancer specimens predicts poor clinical outcome, J Surg Res 159(2) (2010) 689-95. [17] K.X. Zhou, L.H. Xie, X. Peng, Q.M. Guo, Q.Y. Wu, W.H. Wang, G.L. Zhang, J.F. Wu, G.J. Zhang, C.W. Du, CXCR4 antagonist AMD3100 enhances the response of MDA-MB-231 triplenegative breast cancer cells to ionizing radiation, Cancer Lett 418 (2018) 196-203. [18] C. Xu, H. Zhao, H. Chen, Q. Yao, CXCR4 in breast cancer: oncogenic role and therapeutic targeting, Drug design, development and therapy 9 (2015) 4953-64. [19] C. Chittasupho, P. Kewsuwan, T. Murakami, CXCR4-targeted Nanoparticles Reduce Cell Viability, Induce Apoptosis and Inhibit SDF-1alpha Induced BT-549-Luc Cell Migration In Vitro, Curr Drug Deliv 14(8) (2017) 1060-1070. [20] M. Gil, M. Seshadri, M.P. Komorowski, S.I. Abrams, D. Kozbor, Targeting CXCL12/CXCR4 signaling with oncolytic virotherapy disrupts tumor vasculature and inhibits breast cancer metastases, Proc Natl Acad Sci U S A 110(14) (2013) E1291-300.

Journal Pre-proof

Jo

ur

na

lP

re

-p

ro

of

[21] I. Pusic, J.F. DiPersio, Update on clinical experience with AMD3100, an SDF-1/CXCL12CXCR4 inhibitor, in mobilization of hematopoietic stem and progenitor cells, Curr Opin Hematol 17(4) (2010) 319-26. [22] D.M. Ramsey, S.R. McAlpine, Halting metastasis through CXCR4 inhibition, Bioorg Med Chem Lett 23(1) (2013) 20-5. [23] Q. Liu, Z. Li, J.L. Gao, W. Wan, S. Ganesan, D.H. McDermott, P.M. Murphy, CXCR4 antagonist AMD3100 redistributes leukocytes from primary immune organs to secondary immune organs, lung, and blood in mice, Eur J Immunol 45(6) (2015) 1855-67. [24] J. Teixido, M. Martinez-Moreno, M. Diaz-Martinez, S. Sevilla-Movilla, The good and bad faces of the CXCR4 chemokine receptor, Int J Biochem Cell Biol 95 (2018) 121-131. [25] J. Redondo-Munoz, A. Garcia-Pardo, J. Teixido, Molecular Players in Hematologic Tumor Cell Trafficking, Frontiers in immunology 10 (2019) 156. [26] C.A. Herring, C.M. Singer, E.A. Ermakova, B.I. Khairutdinov, Y.F. Zuev, D.J. Jacobs, I.V. Nesmelova, Dynamics and thermodynamic properties of CXCL7 chemokine, Proteins 83(11) (2015) 1987-2007. [27] P. Ray, S.A. Lewin, L.A. Mihalko, S.C. Lesher-Perez, S. Takayama, K.E. Luker, G.D. Luker, Secreted CXCL12 (SDF-1) forms dimers under physiological conditions, Biochem J 442(2) (2012) 433-42. [28] L.J. Drury, J.J. Ziarek, S. Gravel, C.T. Veldkamp, T. Takekoshi, S.T. Hwang, N. Heveker, B.F. Volkman, M.B. Dwinell, Monomeric and dimeric CXCL12 inhibit metastasis through distinct CXCR4 interactions and signaling pathways, Proc Natl Acad Sci U S A 108(43) (2011) 17655-60. [29] A.Z. Dudek, I. Nesmelova, K. Mayo, C.M. Verfaillie, S. Pitchford, A. Slungaard, Platelet factor 4 promotes adhesion of hematopoietic progenitor cells and binds IL-8: novel mechanisms for modulation of hematopoiesis, Blood 101(12) (2003) 4687-94. [30] P. von Hundelshausen, R.R. Koenen, M. Sack, S.F. Mause, W. Adriaens, A.E. Proudfoot, T.M. Hackeng, C. Weber, Heterophilic interactions of platelet factor 4 and RANTES promote monocyte arrest on endothelium, Blood 105(3) (2005) 924-30. [31] I.V. Nesmelova, Y. Sham, A.Z. Dudek, L.I. van Eijk, G. Wu, A. Slungaard, F. Mortari, A.W. Griffioen, K.H. Mayo, Platelet factor 4 and interleukin-8 CXC chemokine heterodimer formation modulates function at the quaternary structural level, J Biol Chem 280(6) (2005) 494858. [32] I.V. Nesmelova, Y. Sham, J. Gao, K.H. Mayo, CXC and CC chemokines form mixed heterodimers: association free energies from molecular dynamics simulations and experimental correlations, J Biol Chem 283(35) (2008) 24155-66. [33] R.R. Koenen, P. von Hundelshausen, I.V. Nesmelova, A. Zernecke, E.A. Liehn, A. Sarabi, B.K. Kramp, A.M. Piccinini, S.R. Paludan, M.A. Kowalska, A.J. Kungl, T.M. Hackeng, K.H. Mayo, C. Weber, Disrupting functional interactions between platelet chemokines inhibits atherosclerosis in hyperlipidemic mice, Nat Med 15(1) (2009) 97-103. [34] Y. Iida, B. Xu, H. Xuan, K.J. Glover, H. Tanaka, X. Hu, N. Fujimura, W. Wang, J.R. Schultz, C.R. Turner, R.L. Dalman, Peptide inhibitor of CXCL4-CCL5 heterodimer formation, MKEY, inhibits experimental aortic aneurysm initiation and progression, Arterioscler Thromb Vasc Biol 33(4) (2013) 718-26. [35] J. Carlson, S.A. Baxter, D. Dréau, I.V. Nesmelova, The heterodimerization of plateletderived chemokines, Biochim Biophys Acta 1834(1) (2013) 158-68.

Journal Pre-proof

Jo

ur

na

lP

re

-p

ro

of

[36] A.J. Brown, P.R. Joseph, K.V. Sawant, K. Rajarathnam, Chemokine CXCL7 Heterodimers: Structural Insights, CXCR2 Receptor Function, and Glycosaminoglycan Interactions, Int J Mol Sci 18(4) (2017). [37] P. von Hundelshausen, S.M. Agten, V. Eckardt, X. Blanchet, M.M. Schmitt, H. Ippel, C. Neideck, K. Bidzhekov, J. Leberzammer, K. Wichapong, A. Faussner, M. Drechsler, J. Grommes, J.P. van Geffen, H. Li, A. Ortega-Gomez, R.T. Megens, R. Naumann, I. Dijkgraaf, G.A. Nicolaes, Y. Doring, O. Soehnlein, E. Lutgens, J.W. Heemskerk, R.R. Koenen, K.H. Mayo, T.M. Hackeng, C. Weber, Chemokine interactome mapping enables tailored intervention in acute and chronic inflammation, Science translational medicine 9(384) (2017). [38] C. Billottet, C. Quemener, A. Bikfalvi, CXCR3, a double-edged sword in tumor progression and angiogenesis, Biochim Biophys Acta 1836(2) (2013) 287-95. [39] A.K. Singh, R.K. Arya, A.K. Trivedi, S. Sanyal, R. Baral, O. Dormond, D.M. Briscoe, D. Datta, Chemokine receptor trio: CXCR3, CXCR4 and CXCR7 crosstalk via CXCL11 and CXCL12, Cytokine Growth Factor Rev 24(1) (2013) 41-9. [40] H. Bronger, A. Karge, T. Dreyer, D. Zech, S. Kraeft, S. Avril, M. Kiechle, M. Schmitt, Induction of cathepsin B by the CXCR3 chemokines CXCL9 and CXCL10 in human breast cancer cells, Oncol Lett 13(6) (2017) 4224-4230. [41] S.M. Agten, R.R. Koenen, H. Ippel, V. Eckardt, P. von Hundelshausen, K.H. Mayo, C. Weber, T.M. Hackeng, Probing Functional Heteromeric Chemokine Protein-Protein Interactions through Conformation-Assisted Oxime Ligation, Angew Chem Int Ed Engl 55(48) (2016) 14963-14966. [42] C. Belloc, H. Lu, C. Soria, R. Fridman, Y. Legrand, S. Menashi, The effect of platelets on invasiveness and protease production of human mammary tumor cells, Int J Cancer 60(3) (1995) 413-7. [43] S.L. Rego, R.S. Helms, D. Dreau, Breast tumor cell TACE-shed MCSF promotes proangiogenic macrophages through NF-kappaB signaling, Angiogenesis 17(3) (2014) 573-85. [44] L.J. Moore, L.D. Roy, R. Zhou, P. Grover, S.T. Wu, J.M. Curry, L.M. Dillon, P.M. Puri, M. Yazdanifar, R. Puri, P. Mukherjee, D. Dreau, Antibody-Guided In Vivo Imaging for Early Detection of Mammary Gland Tumors, Transl Oncol 9(4) (2016) 295-305. [45] G. Zhu, H.H. Yan, Y. Pang, J. Jian, B.R. Achyut, X. Liang, J.M. Weiss, R.H. Wiltrout, M.C. Hollander, L. Yang, CXCR3 as a molecular target in breast cancer metastasis: inhibition of tumor cell migration and promotion of host anti-tumor immunity, Oncotarget 6(41) (2015) 43408-19. [46] K.S. Park, S. Rifat, H. Eck, K. Adachi, S. Surrey, M. Poncz, Biologic and biochemic properties of recombinant platelet factor 4 demonstrate identity with the native protein, Blood 75(6) (1990) 1290-5. [47] F. Delaglio, S. Grzesiek, G.W. Vuister, G. Zhu, J. Pfeifer, A. Bax, NMRPipe: a multidimensional spectral processing system based on UNIX pipes, J Biomol NMR 6(3) (1995) 277-93. [48] B.A. Johnson, R.A. Blevins, NMR View: A computer program for the visualization and analysis of NMR data, Journal of biomolecular NMR 4(5) (1994) 603-14. [49] K.H. Mayo, V. Roongta, E. Ilyina, R. Milius, S. Barker, C. Quinlan, G. La Rosa, T.J. Daly, NMR solution structure of the 32-kDa platelet factor 4 ELR-motif N-terminal chimera: a symmetric tetramer, Biochemistry 34(36) (1995) 11399-409. [50] E. Cerami, J. Gao, U. Dogrusoz, B.E. Gross, S.O. Sumer, B.A. Aksoy, A. Jacobsen, C.J. Byrne, M.L. Heuer, E. Larsson, Y. Antipin, B. Reva, A.P. Goldberg, C. Sander, N. Schultz, The

Journal Pre-proof

Jo

ur

na

lP

re

-p

ro

of

cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data, Cancer Discov 2(5) (2012) 401-4. [51] J. Gao, B.A. Aksoy, U. Dogrusoz, G. Dresdner, B. Gross, S.O. Sumer, Y. Sun, A. Jacobsen, R. Sinha, E. Larsson, E. Cerami, C. Sander, N. Schultz, Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal, Sci Signal 6(269) (2013) pl1. [52] H. Tamamura, A. Hori, N. Kanzaki, K. Hiramatsu, M. Mizumoto, H. Nakashima, N. Yamamoto, A. Otaka, N. Fujii, T140 analogs as CXCR4 antagonists identified as anti-metastatic agents in the treatment of breast cancer, FEBS Lett 550(1-3) (2003) 79-83. [53] A.Z. Fernandis, A. Prasad, H. Band, R. Klosel, R.K. Ganju, Regulation of CXCR4-mediated chemotaxis and chemoinvasion of breast cancer cells, Oncogene 23(1) (2004) 157-67. [54] P.M. Murphy, 10 - Chemokines and Chemokine Receptors, in: R.R. Rich, T.A. Fleisher, W.T. Shearer, H.W. Schroeder, A.J. Frew, C.M. Weyand (Eds.), Clinical Immunology (Fifth Edition), Content Repository, London, 2019, pp. 157-170.e1. [55] E. Ridolfi, E. Matteucci, P. Maroni, M.A. Desiderio, Inhibitory effect of HGF on invasiveness of aggressive MDA-MB231 breast carcinoma cells, and role of HDACs, Br J Cancer 99(10) (2008) 1623-34. [56] Q. Wu, R. Dhir, A. Wells, Altered CXCR3 isoform expression regulates prostate cancer cell migration and invasion, Mol Cancer 11 (2012) 3. [57] M.J. Chen, K.H. Mayo, Human platelet factor 4 subunit association/dissociation thermodynamics and kinetics, Biochemistry 30(26) (1991) 6402-11. [58] K.H. Mayo, M.J. Chen, Human platelet factor 4 monomer-dimer-tetramer equilibria investigated by 1H NMR spectroscopy, Biochemistry 28(24) (1989) 9469-78. [59] R. St Charles, D.A. Walz, B.F. Edwards, The three-dimensional structure of bovine platelet factor 4 at 3.0-A resolution, J Biol Chem 264(4) (1989) 2092-9. [60] N. Nagarsheth, M.S. Wicha, W. Zou, Chemokines in the cancer microenvironment and their relevance in cancer immunotherapy, Nature reviews. Immunology 17(9) (2017) 559-572. [61] K. Rajarathnam, G.N. Prado, H. Fernando, I. Clark-Lewis, J. Navarro, Probing receptor binding activity of interleukin-8 dimer using a disulfide trap, Biochemistry 45(25) (2006) 78828. [62] C.T. Veldkamp, F.C. Peterson, A.J. Pelzek, B.F. Volkman, The monomer-dimer equilibrium of stromal cell-derived factor-1 (CXCL 12) is altered by pH, phosphate, sulfate, and heparin, Protein Sci 14(4) (2005) 1071-81. [63] C.K. Arnatt, Y. Zhang, Bivalent ligands targeting chemokine receptor dimerization: molecular design and functional studies, Curr Top Med Chem 14(13) (2014) 1606-18. [64] H. Fernando, C. Chin, J. Rosgen, K. Rajarathnam, Dimer dissociation is essential for interleukin-8 (IL-8) binding to CXCR1 receptor, J Biol Chem 279(35) (2004) 36175-8. [65] J. Wang, R. Loberg, R.S. Taichman, The pivotal role of CXCL12 (SDF-1)/CXCR4 axis in bone metastasis, Cancer metastasis reviews 25(4) (2006) 573-87. [66] S. Chatterjee, B. Behnam Azad, S. Nimmagadda, The intricate role of CXCR4 in cancer, Adv Cancer Res 124 (2014) 31-82. [67] S. Singh, U.P. Singh, W.E. Grizzle, J.W. Lillard, Jr., CXCL12-CXCR4 interactions modulate prostate cancer cell migration, metalloproteinase expression and invasion, Laboratory investigation; a journal of technical methods and pathology 84(12) (2004) 1666-76. [68] E.A. Ramos, M. Grochoski, K. Braun-Prado, G.G. Seniski, I.J. Cavalli, E.M. Ribeiro, A.A. Camargo, F.F. Costa, G. Klassen, Epigenetic changes of CXCR4 and its ligand CXCL12 as prognostic factors for sporadic breast cancer, PLoS ONE 6(12) (2011) e29461.

Journal Pre-proof

Jo

ur

na

lP

re

-p

ro

of

[69] A.P. Struckhoff, M.K. Rana, S.S. Kher, M.E. Burow, J.L. Hagan, L. Del Valle, R.A. Worthylake, PDZ-RhoGEF is essential for CXCR4-driven breast tumor cell motility through spatial regulation of RhoA, J Cell Sci 126(Pt 19) (2013) 4514-26. [70] A.B. LHarvey Lodish, S Lawrence Zipursky, Paul Matsudaira, David Baltimore, and James Darnell., Molecular cell Biology, seventh ed., W. H. Freeman and Company, New York, 2000. [71] S.C. Mills, P.H. Goh, J. Kudatsih, S. Ncube, R. Gurung, W. Maxwell, A. Mueller, Cell migration towards CXCL12 in leukemic cells compared to breast cancer cells, Cell Signal 28(4) (2016) 316-24. [72] X. Liu, Q. Xiao, X. Bai, Z. Yu, M. Sun, H. Zhao, X. Mi, E. Wang, W. Yao, F. Jin, L. Zhao, J. Ren, M. Wei, Activation of STAT3 is involved in malignancy mediated by CXCL12-CXCR4 signaling in human breast cancer, Oncol Rep 32(6) (2014) 2760-8. [73] H. Zhou, J. Wu, T. Wang, X. Zhang, D. Liu, CXCL10/CXCR3 axis promotes the invasion of gastric cancer via PI3K/AKT pathway-dependent MMPs production, Biomed Pharmacother 82 (2016) 479-88. [74] J.M. Ponert, S. Schwarz, R. Haschemi, J. Muller, B. Potzsch, G. Bendas, M. Schlesinger, The mechanisms how heparin affects the tumor cell induced VEGF and chemokine release from platelets to attenuate the early metastatic niche formation, PLoS One 13(1) (2018) e0191303. [75] J. Pasquier, N. Abu-Kaoud, H. Abdesselem, A. Madani, J. Hoarau-Vechot, H.A. Thawadi, F. Vidal, B. Couderc, G. Favre, A. Rafii, SDF-1alpha concentration dependent modulation of RhoA and Rac1 modifies breast cancer and stromal cells interaction, BMC Cancer 15 (2015) 569. [76] T.G. Pestell, X. Jiao, M. Kumar, A.R. Peck, M. Prisco, S. Deng, Z. Li, A. Ertel, M.C. Casimiro, X. Ju, A. Di Rocco, G. Di Sante, S. Katiyar, A. Shupp, M.P. Lisanti, P. Jain, K. Wu, H. Rui, D.C. Hooper, Z. Yu, A.R. Goldman, D.W. Speicher, L. Laury-Kleintop, R.G. Pestell, Stromal cyclin D1 promotes heterotypic immune signaling and breast cancer growth, Oncotarget 8(47) (2017) 81754-81775. [77] K.H. Susek, M. Karvouni, E. Alici, A. Lundqvist, The Role of CXC Chemokine Receptors 1-4 on Immune Cells in the Tumor Microenvironment, Frontiers in immunology 9 (2018) 2159. [78] N. Wang, W. Liu, Y. Zheng, S. Wang, B. Yang, M. Li, J. Song, F. Zhang, X. Zhang, Q. Wang, Z. Wang, CXCL1 derived from tumor-associated macrophages promotes breast cancer metastasis via activating NF-kappaB/SOX4 signaling, Cell death & disease 9(9) (2018) 880. [79] S. Pillozzi, A. Bernini, O. Spiga, B. Lelli, G. Petroni, L. Bracci, N. Niccolai, A. Arcangeli, Peptides and small molecules blocking the CXCR4/CXCL12 axis overcome bone marrowinduced chemoresistance in acute leukemias, Oncol Rep 41(1) (2019) 312-324. [80] S. Scala, Molecular Pathways: Targeting the CXCR4-CXCL12 Axis--Untapped Potential in the Tumor Microenvironment, Clin Cancer Res 21(19) (2015) 4278-85. [81] J.P. Morton, O.J. Sansom, CXCR2 inhibition in pancreatic cancer: opportunities for immunotherapy?, Immunotherapy 9(1) (2017) 9-12. [82] K. Pilatova, K. Greplova, R. Demlova, B. Bencsikova, G.L. Klement, L. ZdrazilovaDubska, Role of platelet chemokines, PF-4 and CTAP-III, in cancer biology, Journal of hematology & oncology 6 (2013) 42. [83] A.M.E. Walenkamp, C. Lapa, K. Herrmann, H.J. Wester, CXCR4 Ligands: The Next Big Hit?, J Nucl Med 58(Suppl 2) (2017) 77s-82s.

Jo

ur

na

lP

re

-p

ro

of

Journal Pre-proof

Journal Pre-proof Figure legends: Fig 1. MDA-MB 231 breast cancer cell migration is promoted by CXCL12 through CXCR4 receptor signaling and inhibited by high concentrations of CXCL4 signaling through CXCR3 receptor signaling. (A) CXCL12 (50 or 100 nM) significantly increased MDA-MB231 cell migration (p<0.01) compared to control conditions. (B) MDA-MB 231 cells express membrane CXCR4 receptors: dark and light gray represent isotype control and CXCR4, respectively. (C) The CXCR4 receptor antagonist AMD3100 (20nM) prevented the CXCL12driven MDA-MB-231 cell migration (p = n.s.). (D) 100nM but not 50nM of CXCL4

of

significantly decreased MDA-MB-231 cell migration (p<0.05). (E) MDA-MB 231 cells express

ro

membrane CXCR3b receptors: dark and light gray represent isotype control and CXCR3b, respectively. (F) Notably, AMG487 (5nM) prevented CXCL4 (100nM) associated cell migration

-p

inhibition. Data presented as mean ± SEM, N>3 independent repeats. n.s. not significant;

re

*p<0.05; **p<0.01.

lP

Fig 2. CXCL12-driven MDA-MB 231 breast cancer cell migration is dose-dependently inhibited by concurrent increase of CXCL4 concentrations. (A) Representative

na

microphotographs of the MDA-MB-231 cell monolayer wound at 0 and 24hrs following the different chemokine treatments. (B) Quantification of the MDA-MB-231 cell migration

ur

following treatment with CXCL12 (100nM) or the combination of CXCL12 (100nM) + increasing concentrations (0-100 nM) of CXCL4. Data presented as mean ± SEM, N>4

Jo

independent repeats. **p<0.01, ***p<0.001, and ****p<0.0001.

Fig 3. CXCL12-driven MDA-MB 231 breast cancer cell migration is inhibited by concurrent increase in CXCL4 concentrations through CXCL12-CXCR4 signaling. Quantification of the MDA-MB-231 cell migration following treatment with CXCL12 (100nM) or the combination CXCL12 (100nM) + increasing concentrations (0-100 nM) of CXCL4 in the presence of the CXCR4 inhibitor AMD3100 (20nM) (A) or of the CXCR3 inhibitor AMG487 (5nM) (B). Data presented as mean ± SEM, N>4 independent repeats. **p<0.01, ***p<0.001, and ****p<0.0001.

Journal Pre-proof Fig 4. The CXCL12-CXCL4 chemokine mixture prevents CXCL12-induced changes in MDA-MB 231 breast cancer cell RNA expression. (A) Clustering of the RNA expression of MDA-MB-231 cells treated with control vehicle (N 1 , N 2 ), CXCL12 (100nM) alone (samples 121 , 122 ), the combination CXCL12 (100nM)-CXCL4 (100nM) (samples 12-41 , 12-42 ) and CXCL4 (100nM) alone (samples 41 , 42 ) tested. (B) The heat-map of RNA with higher (red) or lower (green) expression clustered by function, highlighting the variations in MDA-MB-231 RNA expression following chemokine treatments. Average sample signal for each condition and

of

transcript are depicted. Data presented as mean, N≥2 independent repeats. Fig 5. NMR analyses of CXCL12-CXCL4 heterodimers. (A) Chemical shift changes of I51,

ro

T25 and T44 of 15 N-CXCL4 following the addition of unlabeled CXCL12 as determined by

-p

NMR. (B) The CXCL4-CXCL12 heterodimer structure as determined by NMR along with NMRPipe [47] and analyzed using NMRview [48] software. Residues experiencing chemical

re

shift changes above average are shown in red, whereas those that did not show any chemical shift changes are shown in blue on the CXCL4 monomer and the CXCL12 monomer is shown in

lP

green. (C) The amino acid sequence of the first beta sheet of CXCL4 highlighting (in red) the

sequence.

na

amino acids with chemical shift in the interface sequence and the CXCL4-derived peptide

ur

Fig 6. The CXCL4-based peptide AHITSLEVIKAG mimics CXCL4 inhibition of CXCL12driven cell migration (A) CXCL4-derived peptide dose-dependently reduced the CXCL12-

Jo

driven migration of MDA-MB-231 breast cancer cells. (B) In the presence of the 5nM of the CXCR3 inhibitor AMG487, CXCL12-driven migration of MDA-MB-231cell was not significantly decreased. Data presented as mean ± SEM, N≥3 independent repeats.

Journal Pre-proof Author Contributions: KTPN and DD designed, conducted and analyzed the biological experiments. LJD and BRA participated in the analyses of the microarray studies. KTPN and IVN designed, conducted and analyzed the biophysical investigations, aided by LZ and LR. DD, KTPN, and IVN completed the initial draft of the manuscript and all authors participated in the

Jo

ur

na

lP

re

-p

ro

of

review and revisions of the manuscript.

Journal Pre-proof

Highlights • The heterodimer CXCL12-CXCL4 inhibits CXCR4-driven MDA-MB-231 breast cancer cell migration. • The CXCL4-CXCL12 binding interface was identified using NMR spectroscopy.

Jo

ur

na

lP

re

-p

ro

of

• The CXCL4 peptide sequence mimicking the CXCL4-CXCL12 binding interface also prevents CXCR4-driven MDA-MB-231 breast cancer cell migration.