Evaluating the Efficiency of Hyaluronic Acid for Tumor Targeting via

Subscriber access provided by Stockholm University Library

Article

EVALUATING THE EFFICIENCY OF HYALURONIC ACID FOR TUMOR TARGETING via CD44 Alice Spadea, Julio Manuel Rios de la Rosa, Annalisa Tirella, Marianne B. Ashford, Kaye J. Williams, Ian J. Stratford, Nicola Tirelli, and Manal Mehibel Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/ acs.molpharmaceut.9b00083 • Publication Date (Web): 23 Apr 2019 Downloaded from http://pubs.acs.org on April 24, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

EVALUATING THE EFFICIENCY OF HYALURONIC ACID FOR TUMOUR TARGETING via CD44

Alice Spadea1,2,3, Julio Manuel Rios de la Rosa1,2,4, Annalisa Tirella1,2, Marianne B. Ashford2,6, Kaye J. Williams1,3, Ian J. Stratford1,2,3*, Nicola Tirelli1,2,5* and Manal Mehibel1,7

1 Division

of Pharmacy and Optometry, Faculty of Biology, Medicine and Health, Stopford Building, University of Manchester and Manchester Academic Health Science Centre,, Manchester, M13 9PT, UK. 2 North

West Centre of Advanced Drug Delivery (NoWCADD), Division of Pharmacy & Optometry, School of Health Sciences, Faculty of Biology, Medicine and Health, Stopford Building, Manchester, M13 9PT, UK. 3 Manchester

Cancer Research Centre, The University of Manchester, 555 Wilmslow Road, Manchester, M20 4GJ UK. BiOncoTech Therapeutics S.L., Science 2 Business Foundation, C/ Santiago Grisolia 2 Tres Cantos, Madrid 28760, Spain. 4

5 Laboratory

of Polymers and Biomaterials, Fondazione Istituto Italiano di Tecnologia, 16163, Genova, Italy. 6 Pharmaceutical

Sciences, Innovative Medicines Biotech Unit, AstraZeneca, Macclesfield, SK10

2NA, UK. 7

Department of Radiation Oncology, Stanford University , Stanford, CA , USA.

*Corresponding authors: Nicola Tirelli, Laboratory of Polymers and Biomaterials, Fondazione Istituto Italiano di Tecnologia, 16163, Genova, Italy, e-mail, [email protected] and Ian Stratford Division of Pharmacy and Optometry, Faculty of Biology, Medicine and Health, Stopford Building, University of Manchester, UK, e-mail [email protected]

1

ACS Paragon Plus Environment

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT The development of delivery systems capable of tumor targeting represent a promising strategy to overcome issues related to non-specific effects of conventional anticancer therapies. Currently, one of the most investigated agents for cancer targeting is hyaluronic acid (HA), since its receptor, CD44, is overexpressed in many cancers. However, most of the studies on CD44/HA interaction have been so far performed in cell-free or genetically modified systems, thus leaving some uncertainty regarding which cell-related factors influence HA binding and internalization (collectively called “uptake”) into CD44 expressing cells. To address this, the expression of CD44 (both standard and variants, designated CD44s and CD44v respectively) was evaluated in human dermal fibroblasts (HDF) and a large panel of cancer cell lines, including breast, prostate, head and neck, pancreatic, ovarian, colorectal, thyroid and endometrial cancer. Results showed that CD44 isoform profiles and expression levels vary across the cancer cell lines and HDF, and are not consistent within the cell origin. Using composite information of CD44 expression, HA binding and internalization we found that the expression of CD44v can negatively influence the uptake of HA and, instead, when cells primarily expressed CD44s a positive correlation was observed between expression and uptake. In other words, CD44shigh cells bound and internalized more HA compared to CD44slow cells. Moreover, CD44shigh HDF were less efficient in uptaking HA compared to CD44shigh cancer cells. The experiments described here are a first step towards understanding the interplay between CD44 expression, its functionality and the underlying mechanism(s) for HA uptake. The results show that factors other than the amount of CD44 receptor can play a role in the interaction with HA and this represents an important advance with respect to the design of HAbased carriers and the selection of tumors to treat according to their CD44 expression profile. Keywords: CD44, hyaluronic acid, tumor targeting, drug delivery, cancer.

2


Page 2 of 28

Page 3 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


GRAPHICAL ABSTRACT

3



INTRODUCTION The success of cancer chemotherapy is limited by the toxic effects often seen in normal tissues. Therefore, the development of a delivery system capable of specifically targeting tumors, while minimizing exposure to normal tissues, is a promising strategy [1]. Hyaluronic acid (HA) is a non-sulfated linear glycosaminoglycans and is one of the main components of the extracellular matrix [2]. In recent years, HA has been widely explored as a tumor targeting agent [3] and its potential and recent applications in this field have been recently reviewed by Cadete et al. [4] and Rios de la Rosa et al. [5]. HA presents many advantages: it is biocompatible, biodegradable and hydrophilic [2, 6, 7]. Further, it is anionic and has functional groups which allow complexation with cationic polymers, lipids and other therapeutic entities [8-14], preserving an overall negative surface charge of obtained polycomplexes. HA can interact with membrane receptors with a common structural domain, named the “link module”, which contains around 100 amino acids and is involved in the ligand binding [15]. Among them we find the cluster of differentiation 44 (CD44), the receptor for hyaluronic acid mediated motility (RHAMM), the lymphatic vessel endothelial-1 (LYVE-1) and the hyaluronan receptor for endocytosis (HARE). The targeting capacities associated with HA are as a consequence of its most studied receptor, CD44 [16, 17], being found overexpressed in many tumors (including breast, pancreatic, colorectal, gastric and prostate) [18-22]. CD44 constitutes a family of glycoproteins encoded by a single gene; however, variants exist that differ in size (from about 85 to about 250 kDa) due to alternative splicing of specific exons (exon 6 to 15, also called v1-v10) in the extracellular domains of the receptor and posttranslational modifications, such as N-glycosylation and O-glycosylation. The smallest or standard CD44 isoform (from now on referred as CD44s) lacks variant exons. It is physiologically expressed on the membrane of most vertebrate cells and is involved in many physiological cellular processes, including regulation of cell-cell and cell-matrix interactions, regulation of cell growth, survival, differentiation and motility. CD44 molecules that include variant exons (CD44v), whilst expressed on many cells during embryonic development, are only expressed on some epithelial and hematopoietic cells, lymphocytes during maturation and inflammation-driven activation and in some leukemias and carcinomas [23-27]. Overexpression of CD44, including both CD44s and CD44v (panCD44) is considered a reliable diagnostic and prognostic marker for cancer [28, 29]. In addition the ability of CD44 to bind and internalize its ligands makes it an adeal candidate for active targeting in cancer therapy. Exploiting CD44/HA interactions, a number of HA-based formulations have been generated using HA as a tumor targeting moiety, either in its soluble form or to decorate the surface of 4


Page 4 of 28

Page 5 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


nanocarriers [8-10, 12, 13, 30-35]. HA conjugation has been shown to increase drug solubility, stability in biological fluids, increase blood circulation (leading to increased passive targeting) and targeting tumor cells (active targeting) [36]. However, the success of HA-based carriers is sometimes limited by the still poor understanding of cellular/molecular mechanisms that influence binding and internalization (callectively called “uptake”) through CD44. In general, post-translational modifications (glycosylation, acylation, sialylation, etc.) and variant isoform expression are known to affect the interaction between CD44 and HA, but it is still not completely clear whether they influence the receptor directly or the structure of the modified CD44 favors its oligomerization on lipid rafts and, consequently, modifies HA uptake [37-40]. The molecular weight of HA is another well-known variable that can influence its affinity with the receptor. This aspect has been widely studied and, as a rule, it appears that high molecular weight (≥ 30 kDa, HMW) HA, free in solution, grafted to liposomes or coated in nanoparticles (NPs), can create stronger binding to CD44 compared to a low molecular weight (< 30 kDa, LMW) HA, [33, 41-44]. This is likely to be because longer HA chains can have multiple interactions with CD44 molecules, leading to an increase in affinity, but also favoring receptor clustering and enhancing HA internalization into cells [40, 44]. However, most of these studies on CD44/HA interactions were performed using genetically modified or cell free systems [41, 42, 45]. In living cells there will be many more factors that influence HA binding to CD44, which include different isoforms coexisting within the same cell population [46], as well as the receptor status itself. Therefore, there is a real need to appreciate what cell-based parameters will determine delivery to the tumor cell surface and potential internalization of any payload. In this work, we have used one single molecular weight HA (𝑀𝑤 = 180 kDa) and studied 26 human tumor cell lines plus a dermal fibroblast (HDF) cell line as healthy, non-malignant control, to determine their CD44 expression profile and interaction with HA (i.e. binding and internalization). We observed that in cancer cells, high expression of CD44s (CD44shigh) could be correlated with enhanced HA interactions. In contrast, cancer cells with high expression of CD44v (CD44vhigh) or CD44shigh non-maliganat cells showed much less interaction with HA. We have attempted to establish correlations between these parameters in order to aid in our understanding of how they relate to each other and, importantly, how these findings can be used to improve the efficency of drug delivery and for selecting appropriate tumors to treat with HA-based delivery systems.

5



Page 6 of 28

EXPERIMENTAL SECTION General cell culture The panel of cancer cell lines and cell culture media used in the study is detailed in Table 1. All materials herein described, unless stated, were purchased from Sigma-Aldrich, UK. Cells, were grown in T75 flasks (Falcon, Runcorn, UK) with the respective media (supplemented with 2 mM L-glutamine (Sigma-Aldrich, UK) and 10% (v/v) heat-inactivated foetal bovine serum (FBS) (Gibco, UK). For LNCaP cells, an untreated FBS (GE Hyclone, UK) was used to supplement the media. Cells were incubated in a humidified 5% CO2 (v/v) atmosphere at 37˚C and routinely trypsinized with a solution of 0.5% (w/v) trypsin and 0.2% (w/v) ethylenediaminetetraacetic acid (EDTA) (Sigma-Aldrich, UK). Cells were also regularly tested for mycoplasma contamination and were generally used when between passage 10-20 and, in any case, never used over passage 30 for the experiments, except HDF, which were used up to passage 10. Table 1: Characteristics of human cell lines used in this study. Cell Line Disease Tissue MDA-MB-231 Breast adenocarcinoma Mammary gland/breast; derived from metastatic site: pleural effusion MDA-MB-468 Breast adenocarcinoma Mammary gland/breast; derived from metastatic site: pleural effusion MCF7 Breast adenocarcinoma Mammary gland, breast; derived from metastatic site: pleural effusion T-47D Ductal carcinoma Mammary gland; derived from metastatic site: pleural effusion ISHIKAWA Endometrial Uterus adenocarcinoma HeLa Cervical Cervix adenocarcinoma Ca Ski Epidermoid cervical Cervix; derived from metastatic carcinoma site: small intestine SiHa Grade II cervical Cervix squamous cell carcinoma HT 29 Colorectal Colon adenocarcinoma HTC 116 Colorectal carcinoma Colon 8505C Thyroid carcinoma Thyroid FTC133 Follicular thyroid Thyroid (lymph node metastasis) carcinoma

6


Growth Media DMEM DMEM DMEM RPMI EMEM EMEM RPMI EMEM RPMI RPMI RPMI RPMI

Page 7 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


LNCaP

A2780

Prostate; derived from metastatic site: left supraclavicular lymph node Prostate carcinoma Prostate; derived from metastatic site: brain Grade IV prostate Prostate; derived from metastatic adenocarcinoma site: bone Glioblastoma; Brain astrocytoma; classified as grade IV as of 2007 Grade 3, stage III, Ovary primary malignant adenocarcinoma; clear cell carcinoma Grade 3, stage IIIC, Ovary; derived from metastatic malignant papillary site: ascites serous adenocarcinoma Grade 3, stage IIIC, Ovary primary malignant adenocarcinoma; endometrioid carcinoma Ovarian Ovary adenocarcinoma Ovarian Ovary; derived from metastatic adenocarcinoma site: ascites Ovarian Ovary adenocarcinoma Ovarian carcinoma Ovary

DETROIT 562

Pharyngeal carcinoma

FaDu

Squamous carcinoma Pancreas carcinoma Pancreas DMEM Human dermal Primary human dermal DMEM fibroblasts fibroblasts isolated from adult skin

DU145 PC3 U87 TOV-21G

OV-90 TOV-112D

CAOV-3 SK-OV-3 OVCAR-3

MIA PaCa-2 HDF

Prostate carcinoma

RPMI EMEM RPMI DMEM RPMI

RPMI RPMI

DMEM RPMI RPMI RPMI

Pharynx: derived from metastatic RPMI site: pleural effusion cell Pharynx EMEM

CD44 expression analysis Western blot. Cells were grown in 10 cm dishes until ~70% confluence, and then lysed with lysis buffer prepared as follows: 50 mM Tris-HCl pH 7.5, 150 mM NaCl, 50 mM NaF, 1% (v/v) Triton-X100, 0.5% (v/v) IGEPAL and 1X Protease inhibitor cocktail (Roche, UK). The lysates were sonicated on ice and centrifuged at 13,000 g for 10 minutes at 4˚C. The supernatants were collected and the protein concentration in cell lysates was determined using the Pierce Bicinchoninic acid (BCA) assay (Sigma-Aldrich, UK). Equal amount of proteins were loaded in Sample Buffer (250 mM Tris-HCl pH 6.8, 37.5% (v/v) glycerol (Invitrogen, UK), 5% (w/v) SDS, 5% (v/v) β-mercaptoethanol and Bromophenol Blue until an intense blue colour was obtained). Proteins were separated by 7% polyacrylamide SDS-PAGE using a Mini-

7



Protean II Cell gel equipment (BioRad, Hertfordshire, UK) and transferred to a polyvinylidene fluoride (PVDF) membrane (0.2 μm, BioRad, UK) using the Trans-Blot turbo transfer starter system (BioRad, UK). After blocking with 0.1% (v/v) Tween-20 in PBS (PBS-T) supplemented with 5% (w/v) skimmed dried milk, the membranes were incubated overnight at 4˚C with a mouse monoclonal anti-CD44 antibody (clone 156-3C11, Cell Signaling, Danvers, USA) diluted 1:1000 in TBS-T (0.1% (v/v) Tween-20 in Tris-buffered saline). This clone was selected as widely used in literature and considered suitable to investigate CD44 expression [13, 47, 48]. The peroxidase-conjugated (HPR) anti-mouse secondary antibody (SigmaAldrich, UK) was diluted 1:2500 in 5% (w/v) skimmed dried milk in PBS-T. β-actin was used as a loading control (1:5000 in PBS-T, Sigma-Aldrich, UK). The protein bands were finally detected using the Clarity Western ECL Blotting Substrate kit (BioRad, UK) and developed with ChemiDoc™ MP system (BioRad, UK). Densitometry analysis was carried out using ImageJ software (http://rsb.info.nih.gov/ij). Flow cytometry. Cells were grown in T75 flasks until 70% confluence, and then harvested using Accutase solution (Sigma-Aldrich, UK). Briefly, 5 105 cells were transferred in each sample tube and washed with PBS. Two protocols were used for either analyze membrane bound (immediate immunostaining) or the intracellular and the membrane bound CD44 (cells fixed and permeabilized first, then immunostained). Protocols are described below. Fixation/Permeabilization. Cells were collected by centrifugation, resuspended in 0.5 mL of 4% (w/v) formaldehyde in PBS, and incubated at 37˚C for 10 minutes. The tubes were then transferred and cooled on ice for 1 minute and cells were permeabilized by slowly adding icecold methanol absolute, while gently vortexing, to a final concentration of 90% (v/v) methanol. Cells were incubated on ice for a further 30 minutes. Immunostaining. Cells were washed with the incubation buffer (0.5% (w/v) of Bovine Serum Albumin (BSA) in PBS), then incubated for 1 hour on ice with 0.1 mL of primary antibody in incubation buffer (antibodies used and respective dilutions are listed in Table 2). The antibody clones were selected as already widely employed for the detection of panCD44 and CD44v3, v4, v5 and v6 expression [49-51]. Cells were washed and incubated with the fluorochromeconjugated secondary antibody (Table 2) for 1 hour on ice. Cells were washed, resuspended in 0.5 mL of PBS and analyzed using a BD LSRFortessa cytometer (BD Bioscience, San Jose CA, USA) equipped with the FACSDiva software (v8.0.1). Gating in FSC-H/FSC-A and FSCA/SSC-A windows was performed in order to acquire 10,000 viable and single cells, respectively. The data were analyzed using Summit software (Dako, Colorado, USA).

8


Page 8 of 28

Page 9 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Immunofluorescence. Detroit 562, Ishikawa, PC3, DU145, FaDu, MDA-MB-468 and HT29 cells were seeded at a density of 2105 cells/well in sterile coverslips placed into 6-well plates. MIA PaCa-2 and MDA-MB-231 were seeded at a density of 1105 cell/well, HDF cells were seeded at a density of 0.8x105 cell/well, and LNCaP were seeded at a density of 9105 cell/well. Cells were allowed to attach overnight at 37˚C, 5% CO2 (v/v), then fixed by addition of 1 mL/well of cold absolute methanol and incubation for 10 minutes at -20˚C. Cells were washed and the unspecific sites were blocked by incubation in blocking buffer (1% (w/v) BSA, 10% (v/v) secondary antibody host serum (donkey serum, Sigma-Aldrich, UK) in PBS) for 30 minutes at room temperature. Primary antibody (anti-CD44, clone 156-3C11, Cell Signalling, USA) was added to each coverslip diluted 1:400 in blocking buffer and left overnight at 4˚C. The fluorochrome-conjugated secondary antibody (donkey anti-mouse, Alexa Fluor568conjugate, Life technologies, UK) was made up 1:1000 in 1% BSA (w/v) in PBS and 0.1 mL was added to the samples for 1 hour at 37˚C. 0.1 mL of DAPI (1 μg/mL) in PBS were added to each samples to stain the nuclei and incubated for 10 minutes at room temperature. Images were collected on an Olympus BX51 upright microscope using a 20X/0.50 UPlanFLN or 60X/0.65-1.25 UPlanFLN objective and captured using a Coolsnap EZ camera (Photometrics) through MetaVue Software (Molecular Devices). Specific band pass filter sets for DAPI and Texas red were used to prevent bleed through from one channel to the other. Images were then processed and analysed using ImageJ software (http://rsb.info.nih.gov/ij). Immunohistochemistry. Tumors for histological analysis came from an in house bank of formalin fixed human tumour xenografts. These were originally derived following the harvesting of subcutaneously implanted tumors (Ishikawa, Detroit 562 and FaDu) or tumors that were grown orthotopically (MDA-MB-231). Tumor blocks were sectioned using a microtome (Leica RM2155) with a thickness of 5 μm and collected onto adhesion microscope slides. Sections were de-waxed in xylene then in sequence in graded Ethanol solutions of 100% (v/v), 90% (v/v), 70% (v/v) ethanol and then washed in dH2O. Antigen retrieval was performed in a microwave using 10 mM pH 6 sodium citrate buffer and maintaining them at a sub-boiling temperature for 15 min. Endogenous peroxidase activity was blocked with Peroxo-Bloc Peroxidase Inhibitor (Life Technologies, UK). To prevent non-specific antibody binding sections were blocked using 10% secondary antibody host serum (v/v) in PBST for 1 hour. 0.1 mL of primary antibody antiCD44 (clone EPR1013Y, Rabbit monoclonal, Millipore) diluted 1:100 in blocking buffer was added to each section and incubated overnight at 4˚C in a humidified chamber. Sections were

9



Page 10 of 28

then incubated with a secondary anti-Rabbit biotinylated antibody (Vector laboratories) diluted 1:200 in PBS for 1 hour. ABC reagent (VECTASTAIN ABC Kit, Vector Laboratories, California, USA) was prepared accordingly to manufacturer instructions and added to the sections. Sections were incubated with DAB reagent (DAKO, Cambridgeshire, UK) until brown. The reaction was stopped by placing the slides in dH2O. Slides were counterstained with haematoxylin, washed with tap water and dipped for one minute in alkaline Scott’s tap water. Finally, they were dehydrated through increasing ethanol concentrations and washed in xylene. Cover slips were mounted and images were acquired on a 3D-Histech Pannoramic-250 microscope slide-scanner using a 0.30 Plan Achromat objective (Zeiss) and the color camer for brightfield. Snapshots of the slide-scans were taken using the Case Viewer software (3DHistech, 1.15.4).

Table 2. Antibodies used for Flow cytometry Antibodies Species

Dilution

Supplier

Anti-CD44 (156-3C11)

Mouse

1:100

Cell Signaling, USA

Anti-CD44v3 (VFF-327)

Mouse

1:100

AbD Serotec, BioRad, UK

Anti-CD44v4 (VFF-11)

Mouse

1:50


Anti-CD44v5 (VFF-8)

Mouse

1:100


Anti-CD44v6 (2F10)

Mouse

1:100


Mouse IgG2a Isotype control

Mouse

1:100

R&D systems, Abingdon, UK

Anti-Mouse IgG/FITC

Rabbit

1:100

Dako, UK

Hyaluronic acid uptake (binding and internalization) HA internalization analysis via flow cytometry. Detroit 562, HT29, MDA-MB-468, MIA PaCa-2, PC3 cells were seeded at a density of 1.7105 cells/well in 12-well plates (Falcon, Runcorn, UK), Ishikawa, DU145, FaDu cells were seeded at a density of 1105 cells/well, HDF and MDA-MB-231 cells were seeded at a density of 0.5105 cells/well and left to attach overnight at 37˚C, 5% (v/v) CO2. HA covalently labelled with Lissamine™ Rhodamine B ethylenediamine (HA-RhoB; 𝑀𝑤 = 180 kDa), prepared as previously reported [13], was firstly dissolved in dH2O at a concentration of 0.25 mg/mL and then diluted to a final concentration of 0.125 mg/mL mixing equal volumes of a 0.25 mg/mL HA-Rho solution with an equal volume of 2X complete medium (DMEM, MEM or RPMI powder) (Sigma-Aldrich, UK). 2X complete media was prepared, according to manufacturer instructions, from powder in deionized water followed by addition of HEPES 50 mM as previously described [12, 13]. The 10


Page 11 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


pH was adjusted to 7 by adding adequate volumes of HCl 0.1 M. The solution was then filtered using 0.22 µm filters (Corning, UK) and supplemented with 20% (v/v) FBS. Cells were incubated for 0, 1, 4 or 24 hours with 1 mL/well of 0.125 mg/mL HA-RhoB/media. After treatment, cells were washed twice with PBS and detached using 0.5 mL per well of Trypsin/EDTA solution for 10 minutes at 37˚C. The reaction was stopped with fresh media and trypsin was discarded after centrifugation at 1,500 rpm for 5 minutes. The pellets were washed twice with PBS and then resuspended in 0.5 mL of PBS. HA-RhoB internalization was analyzed using a BD LSRFortessa cytometer (BD Bioscience, San Jose CA, USA) equipped with the FACSDiva software (v8.0.1). HA binding and internalization analysis via cell lysate analysis. Cells were seeded using the same protocol detailed in the FC section. Cells were incubated with HA-RhoB/media at a concentration of 0.125 mg/mL for 1 hour. HA-RhoB containing media was then removed, cells were washed twice with PBS and then 0.12 mL of lysis buffer was added to each well. After 30 minutes of shaking at room temperature, 0.1 mL of cell lysate from each well was transferred in a black 96-well plate. The total HA-RhoB uptake (membrane bound and internalized) was measured using the fluorescence of cell lysates (Ex: 540/25, Em: 620/40 nm; Synergy2 Biotek plate reader with Gen5 software (Optical position: top 50%. Light source: Xenon flash)). A calibration curve, obtained using concentrations of HA-RhoB dispersed in lysis buffer between 0 to 31.25 µg/mL, was used to quantify the amount of HA-RhoB, which was normalized against the total cell protein content measured using the BCA assay for each well. The total amount of proteins per well were converted to cell number per well using cell number/protein concentration calibration curves produced for every cell line. The uptake was finally expressed as μg of HA-RhoB per cell. Laser scanning confocal microscopy (LSCM). Detroit, Ishikawa, MIA PaCa-2, PC3, DU145, FaDu, MDA-MB-231 were seeded at a density of 25,000 cells/well in an Ibidi 8 well µ-Slide (Ibidi, UK). MDA-MB-468, HT29 cells were seeded at a density of 42,500 cells/well, HDFs were seeded at a density of 10,000 cells/well and LNCaP were seeded at a density of 100,000 cells/well. Cells were let attach overnight at 37˚C, 5% (v/v) CO2 then a volume of 0.3 mL per well of HA-RhoB (0.125 mg/mL) media was added to each well and incubated for 1 hour at 37°C, 5% (v/v) CO2. After HA-RhoB treatment, cells were washed with PBS and incubated for 10 minutes with 0.3 mL of 1 μg/mL solution of Hoechst (Invitrogen, UK) in cell culture media at 37°C, 5% (v/v) CO2, to counterstain nuclei. CellMask™ (5 μg/mL) in cell culture media was added to the cells and incubated for 10 minutes to highlight the membrane phospholipids. MIA PaCa-2 cells were in parallel treated with 200 nM LysoTracker green 11



Page 12 of 28

(Invitrogen, UK) in cell culture media for 15 min at 37˚C, 5% (v/v) CO2. Cells were then washed and media without phenol red was added. Confocal acquisitions were performed on live cells using a Leica TCS SP5 AOBS inverted confocal equipped with an immersion oil objective (63X/1.40/HCX PL Apo, Leica). For the acquisitions, the confocal settings were set as follows: pinhole 1 airy unit, scan speed 400Hz unidirectional, format 20482048 (1X zoom, pixel size 240 nm), format 10241024 (4X zoom, pixel size 89 nm). To eliminate any possible crosstalk between channels, images were collected with a sequential scan using the following laser lines and mirror settings: 405(100%) 410-483nm; 488(25%) 495-550nm; 561(100%) nm. Images

were

processed

to

enhance

brightness

using

the

ImageJ

software

(http://rsb.info.nih.gov/ij). Scoring method. Using the quantifications obatined from different techniques employed for the investigation of both CD44 expression and HA uptake, cells could be ranked in a decreasing order. For example, the relative density obtained with WB (panCD44/β-actin) for MDA-MB468 cells was 2.3, the highest value measured between all the cell lines, and, for instance, PC3 cells value was 0.9. A percentage of 100% was assigned to the cell line with the highest quantification and the percentages of the other cells were calculated proportionally as reported in Tables from 2SI to 9SI (in this case, MDA-MB-468 cells were assigned 100% and PC3 39.8%). To every percentage range, it corresponded a score from 9 to 0, as shown in Table 1SI (therfore MDA-MB-468 had a score of 9 and PC3 a score of 3). We identified four categories (panCD44, CD44s, CD44v, HA uptake) to study the CD44/HA uptake correlation. Scores obtained from the different techniques were summed up for each category, i.e. the final score of the “panCD44” cathegory is given by the sum of scores obtained from quantification of panCD44 with WB plus FC (showed in bold Table 11SI, i.e. MDA-MB-468 had a final panCD44 score of 17 given by the sum of panCD44 WB, which was 9, and panCD44 FC, which was 8). Cells were finally ranked in a decreasing order according to HA uptake (sum of scores from binding and internalization) (see Table 3). RESULTS Evaluation of CD44 expression The expression of CD44 was analyzed via western blotting (WB) in a panel of 26 cancer cell lines (Figure 1A), using human dermal fibroblasts as a non-tumor control. The antibody used (clone 156-3C11) binds an epitope on the amino-terminal region of the receptor (exons 1-5), recognizing all CD44 isoforms, hence bands of different molecular weight correspond to different isoforms of the receptor (panCD44) [52]. This analysis revealed a substantial

12


Page 13 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


heterogeneity in the expression of CD44, which can be easily recognizable between: CD44s (band at a MW of ~90 kDa) and CD44v (band at a MW >90 kDa). Accordingly to this classification, cells were categorized into three groups: 1) CD44shigh, where the standard form of the receptor is highly expressed; these cell lines are highlighted with a red bar in the top panel of Figure 1A; 2) CD44vhigh, showing expression of the increased molecular weight variants of the receptor (CD44v) independently of the amount of CD44s, these cell lines are highlighted with a blue bar. 3) CD44low, where CD44 (in both forms) is poorly expressed; these cells are highlighted with a black bar. In order to interrogate the role of CD44 for controlling HA uptake, we carried out further experiments choosing at least three representative cell lines from each of the groups characterized as CD44shigh, CD44vhigh and CD44low, and these are highlighted with arrows in Figure 1A: MDA-MB-231, DU145, PC3 and the non-tumoral HDF for CD44shigh; MDA-MB-468, MIA PaCa-2, Detroit 562 and HT29 for CD44vhigh and Ishikawa, LNCaP and FaDu for CD44low. Immunofluorescence (IF; Figure 1B) generally confirmed that the CD44 expression in the previously named CD44low group was lower than for the CD44shigh and CD44vhigh groups; unsurprisingly, CD44vhigh and CD44shigh were virtually indistinguishable when stained with an antibody for panCD44. It is noteworthy that there were strong differences between the three lines in the CD44low group. In particular, LNCaP appeared almost entirely devoid of CD44 expression, Ishikawa cells showed a heterogeneous population with cells both lacking and presenting CD44, while FaDu appeared to present levels of the receptor not much lower than in the cells of the other two groups. Flow cytometry (FC) was then employed for a second quantitative analysis of CD44 expression (Figure 1C). Firstly, it is noteworthy that panCD44 was detectable in Ishikawa and FaDu cells, but not in LNCaP cells, therefore confirming the qualitative IF observations. All the other cell lines were ~100% positive for panCD44, although with different median fluorescence intensities (MFI) (Figure 1C and 1SI C). Using FC was possible to investigate the expression of some CD44v relevant in cancer [53-57]. The expression of v3, v5 and v6 CD44 variants in MDA-MB-468 and Detroit 562 was confirmed, but these variants were not detected in MIA PaCa-2 cells. HT29 cells were positive for CD44s and only partially for variants v3 and v5 (Figure 1SI C). In order to explore CD44 expression in a physiologically relevant environment, immunohistochemistry (IHC) analysis of tumor xenografts was performed (Figure 2). A high and quite homogenous panCD44 expression was found in MDA-MB-231 and Detroit 562 tumors. On the other hand, Ishikawa and FaDu xenografts showed a heterogeneous expression 13



of the panCD44. FaDu tumors showed a higher intensity in staining compared to Ishikawa’s, with both tumors exhibiting an increase of CD44 positive cells towards specific micro-regions.

Figure 1. A. CD44 expression via WB for cancer cell lines and HDF. A primary antibody against panCD44 was used, and β-actin was employed as a loading control (n=3). B. Assessment of membrane bound CD44 via confocal immunofluorescence. Cells were fixed, and stained first with a primary antibody against panCD44 and then with an Alexa Fluor568-conjugated secondary antibody. Scale bar 20 μm. C. CD44 expression via FC: CD44 is expressed as the median fluorescence intensity (MFI) of living cells after incubation with primary antibodies

14


Page 14 of 28

Page 15 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


against panCD44 or single variants (v3, v4, v5 and v6) and then with a FITC-labelled secondary antibody. Values are the mean (± SD) of three independent experiments and are reported as the ratio between MFI of the sample and the MFI of the isotype control (fold change). In the legend, the CD44shigh types are underlined in red, the CD44vhigh in blue, while the CD44low are not underlined.

Figure 2. Representative images of CD44 staining in FaDu subcutaneous tumors (n=4), Ishikawa subcutaneous tumours (n=4), Detroit 562 subcutaneous tumours (n=2), MDA-MB-231 orthotopic tumours (n=3). Scale bar, 200 μm. Scale bar inner images, 2000 μm.

CD44 expression: Western blot vs Flow cytometry As we want to explore the use of CD44 for tumor targeting therapies, we focused our interest on the expression of CD44 on the extracellular part of cell membrane, which is relevant for binding HA-based delivery systems. For this purpose, FC was initially carried out in living cells (Figure 1C): data were compared with those obtained from WB quantifications (Figure 1ASI). It should be noted that WB technique will detect CD44 receptors localized both extraand intra-cellularly (analysis of whole cell lysate), which is different from FC on living cells where extra-cellular receptors only are detected. Interestingly, it was found that the two techniques ranked the cells differently in terms of CD44 expression (Figure 3A). In particular we noticed that, when the dataset are compared it was possible to classify the cells into two distinctive groups: CD44fc/wbhigh and CD44fc/wblow (groups divided by the dotted line in Figure 3A). These two groups displayed internal congruence between the two techniques used, with a proportional increase being observed between FC MFI and WB band intensity in each different region in the graph. To investigate whether the distinction between the two groups, CD44fc/wbhigh and CD44fc/wblow, is influenced by the presence of intracellular receptors, we performed FC in permeabilized cells. Figure 3B shows that little difference is seen when compared to the live cell analysis, which strongly suggests that the ranking of the different cell lines (CD44fc/wbhigh and CD44fc/wblow) could not be ascribed to the intracellular amount of the receptor.

15



Figure 3. Correlation between panCD44 expression as evaluated via FC (vertical axis) and WB on cell lysates (horizontal axis), please note that the same antibody was used. A. MFI obtained using living cells normalized to HT29; the symbols are put together in two groups (CD44fc/wbhigh – dark grey; CD44fc/wblow- light grey) that internally show proportionality between MFI and WB results. The names of cell lines are tagged with bold boxes if the cells belonged to the CD44shigh group, with single or double boxes if they belonged to the CD44vhigh one. B. MFI obtained using fixed and permeabilized cells normalized to HT29.

Binding and internalization of HA in cells Fluorescently labeled HA (HA-RhoB) was used to investigate the binding and internalization in the panel of cell lines with distinctive CD44 profiles. An initial screening was performed at an early time point (1 hour incubation with HA-RhoB) using laser scanning confocal microscopy (LSCM), with results shown in Figure 4A. The internalization of HA-RhoB was clearly visible in all cells after 1 hour, including the CD44-negative LNCaP cells. As expected on the basis of literature evidence [58-61], HA-RhoB appeared to show significant colocalisation with late endosomes/lysosomes and this was observed already after 1 hour incubation in MIA PaCa-2 cells (Figure 4B). To further investigate internalization solely, all the cell lines were incubated with HA-RhoB for different times, i.e.1, 4 and 24 hours, and internalization was measured as MFI with FC (Figure 4C). Prior to FC analysis cells were trypsinized to remove cell surface bound HA-RhoB (Figure 2SI) as previously reported [62]. Thus, the fluorescence measured in the trypsinized cells was attributable to internalized HARhoB only. Interestingly, although the cell lines differed from each other in terms of amount of HA-RhoB internalized, they seemed to follow similar kinetics, as is suggested from the double logarithmic scale graph (Figure 4C, right panel), with no cell line reaching a plateau at the latest time point (24 hours). Interestingly, the CD44low Ishikawa cells and the non-tumoral HDF (CD44shigh) were the least efficient in HA internalization.

16


Page 16 of 28

Page 17 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Internalization of HA-RhoB in cancer cell lines and HDF (confocal microscopy, no fixation). A. HARhoB internalization. Cells were treated for 1 hour with HA-RhoB (125 μg/mL) at 37˚C. Nuclei: Hoechst stain; plasma membrane: CellMask Green. Scale bar, 10 μm. B. HA-RhoB intracellular localization. MIA PaCa-2 cells were treated for 1 hour with HA-RhoB (125 μg/mL). Lysosomes: Lysotracker Red. Scale bar, 10 μm. C. Internalization kinetics. Cells were treated for 0, 1, 4 or 24 hours with HA-RhoB (125 μg/mL) at 37˚C. Graphs are reported as a linear (left) or a double logarithmic scale (right) study of HA internalization as a function of the time. Values are the mean (± SD) of three independent experiments and are reported as the ratio between MFI of the sample and the MFI of the untreated cells (fold change).

The relationship between HA binding and internalization and cellular CD44 expression To evaluate the interplay between the receptor expression on the cell surface and the binding and internalization (uptake) of HA we evaluated both the HA-RhoB bound and internalized by cells at an early incubation time (1 hour) through analysis of cell lysate and the material internalized at a late time point (24 hours) through FC. The lysed cells were not pre-treated 17



with trypsin and thus it was possible to measure the sum of membrane-bound and internalized HA-RhoB. We compared the panCD44 expression obtained with WB (Figure 5, panels A and B) or FC (Figure 5, panels C and D) with HA-RhoB in cell lysates after 1 hour incubation (Figure 5, panels A and C) or with MFI of intracellular HA-RhoB after 24 hours incubation (Figure 5, panels B and D). Four areas were identified within each panel separating the five most efficient “internalizer” cell lines from the other six cell lines (horizontal dotted line) and the five highest panCD44 expressing cells from the other six (vertical dotted line). Interestingly, some cells appeared to be more efficient in uptake at 1 hour than in internalizing at 24 hours (Ishikawa, DU145, MDA-MB-231, box highlighted in dark grey), whereas other cell lines (LNCaP and MIA PaCa-2), had the opposite behavior: they were better in internalizing at 24 hours than in uptake at 1 hour (box highlighted in light grey). The remaining cell lines did not show differences between binding and internalization and both analysis methods gave consistent results. In order to gain further insight into the relationship between CD44 expression and HA uptake, we assigned a score (from 0 to 9 as described in the experimental section) to the cells depending on the amount of panCD44 (CD44s plus CD44v expression) and HA uptake (HA binding plus HA internalization). In Table 3, the cells were ranked from the best to least efficient in total uptake and divided into two groups: high and low HA uptake, respectively the first five cell lines and the last six cell lines. When this ranking was applied to CD44 expression, it appeared that a similar classification was observable for both the panCD44 expression (WB+FC) and CD44s (only WB). Inspection of Table 3 further indicates that, despite their high panCD44 expression, MDA-MB-468, Detroit 562 and HDF are ranked as low “uptakers” (Table 11SI). We also went on to rank cells on the basis of HA binding only or internalization only (Tables 12SI and 13SI). Although the ranking was slightly different, it was interesting to note that the MDA-MB-468, Detroit 562 and HDF cells maintained their position in the lower part of the table, hence confirming their relatively low binding or internalization of HA compared to their high levels of panCD44 expressed.

18


Page 18 of 28

Page 19 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Correlation graphs between HA uptake vs panCD44 expression. A. Correlation between panCD44 expression evaluated via WB and total HA uptake (binding + internalization) evaluated via analysis of cell lysates after 1 hour treatment with HA-RhoB (125 μg/mL) at 37˚C. B. Correlation between panCD44 (WB) and HA internalization evaluated via FC after 24 hours treatment with HA-RhoB. C. Correlation between panCD44 MFI obtained with FC and total HA uptake evaluated via analysis of cell lysates after 1 hour treatment with HA-RhoB. B. Correlation between panCD44 MFI (FC) and HA internalization after 24 hours treatment with HA-RhoB.

Table 3: CD44 expression vs HA uptake scoring table.

19



DISCUSSION The interaction between CD44 and HA has been widely studied and it has been found to be regulated at different levels in both physiological and pathological conditions [4, 8-12, 63, 64]. The molecular mechanism of CD44 activation is still unclear, although there is evidence that removal of sialic acid from some specific sugar chains in the CD44-HA binding domain activates binding [65, 66]. Further, palmitoylation of specific residues can be important for receptor association with lipid rafts and, consequently, for internalization of HA, but it has no effect on binding [37]. So we can deduce that: firstly, binding and internalization are not always sequential, but are often separate events; and secondly, the “affinity” between HA and CD44 seems to be important for binding, but it is the “avidity”, which is in turn influenced by the density of the receptor on the cell membrane and its clustering in lipid rafts, that mainly affects internalization. It has also been shown that, beside post-translational modifications, the binding of HMW HA to CD44 [44] and expression of variant exons can also enhance CD44 clustering and local density [40]. There are enough evidences to believe that other HA receptors have just a marginal involvement in HA uptake. In fact, little is known about RHAMM mechanism of HA endocytosis, but more is known about its role in the signal transduction, i.e. it can associate with CD44 and co-signal through the ERK1/2 pathway [67]. It has been reported that in RHAMM expressing mesenchimal adherent cells anti-RHAMM antibodies did not block HA uptake. In the same cells, RHAMM overexpression led to higher HA uptake, but this was still CD44-dependent, although RHAMM was found to regulate CD44 expression [68]. In head and neck cancer cells, RHAMM and LYVE-1 were also found not play a significant role in HAconjugates uptake [69]. LYVE-1 is mainly expressed on endothelial cells of lymphatic vessels and HA bound to it is recognised by CD44 receptor expressed on leucocytes to facilitate their migration into lymphatic vessels. The mechanism of HA internalization is thought to be similar to the CD44-mediated one although, also in this case, it is not very well established [61]. HARE is found in the liver (sinusoid hepatocytes), in the spleen (venous sinuses of the red pulp) and in the in lymph nodes (medullary sinuses). It is, therefore, mainly involved in systemic HA turnover [70]. Starting from this knowledge base and with the aim of validating the HA-based systems as potential tumor targeting therapies, we studied the interplay between HA and CD44, as main receptor for HA endocytosis, in a panel of cancer cell lines and in HDF (as a nonmalignant control cell line). To achieve this, we evaluated the expression of CD44 (panCD44 and single variants v3, v4, v5 and v6) using WB, FC and IF. We found that the different tumor cell types, not only expressed different amount of the receptor, but also different isoforms. At 20


Page 20 of 28

Page 21 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a first examination, WB and FC analysis appeared to provide a similar qualitative picture about the CD44 fingerprint, however differences were noted in the expression of variants: FC analysis confirmed the expression of CD44 v3, v5 and v6 in MDA-MB-468 and Detroit 562. A small percentage of HT29 cells were positive for variants v3 and v5 (Figure 1SI). It is also of note that our analysis confirms literature reports that MIA PaCa-2, PC3 and MDA-MB-231 lack expression of CD44v and that MDA-MB-468 and Detroit 562 cells express high levels of CD44v [19, 46, 71, 72]. It should be noted that variants v7, v8, v9 and v10 were not investigated in this work, therefore we cannot exclude that these cell lines might express these isoforms as reported for HT29 and MDA-MB-468 [21, 50, 73]. A recent study reporting a detailed analysis using RNA sequencing of many cancer cell lines showed that the expression of CD44 is almost never limited to a single isoform in these cells, although there could be be a predominant isoform [46]. According to this latter study, for example DU145 cells, presented a mixed population with 85% CD44s, 5.5% CD44v10, 5% CD44v8-v10 and 3% CD44v3-v10, while no variants were detected from us at protein level using FC or WB (Figure 1S) (DU145 were designated by us as CD44shigh). CD44 is a plasma membrane receptor, but it can be localized intracellularly as a consequence of its own turnover. Interestingly, it was previously highlighted that HA can influence the receptor half-life in chondrocytes, enhancing receptor internalization when removed from the pericellular matrix with hyaluronidases [74]. The two orthogonal techniques we have used to detect CD44 (WB and FC) can give us different information about the status of the receptor. Using either one of the two techniques in isolation ranked the cell lines differently. The basis for this difference is not due to: firstly, specificity for the detection of the receptor, because the same antibody was employed for all cell types using both techniques; secondly, the presence of CD44 variants, because these two groups did not correlate with the previous classification into CD44shigh and CD44vhigh (see the random distribution of the specific boxes, Figure 3A and B); or thirdly, intracellular localization of the receptors. We are inclined to ascribe the difference between CD44fc/wbhigh and CD44fc/wblow to surface features (e.g. glycoprotein conformation, interactions with other receptors, higher packing density in lipid rafts, etc.) that might decrease the FC detectability of panCD44 in CD44fc/wblow cells. Indeed, it has been reported that certain antibodies may not recognize CD44 when it is glycosylated in specific regions, detection being restored when samples were treated with glycosidases [72]. In other words, the ranking/grouping obtained in this panel of cell lines likely reflects the detectability of the receptor on cell surfaces (CD44fc/wbhigh, CD44fc/wblow).

21



CD44 expression in vivo correlated with our in vitro results for the MDA-MB-231, Detroit 562, FaDu and Ishikawa tumor xenografts that we examined. In the FaDu and Ishikawa tumors there is an indication of micro-regional variation in expression with the strongly CD44positive cells being concentrated in specific areas within the tumors. Previous studies showed that increased levels of CD44 correlated with the hypoxic/necrotic areas in breast cancer tumors [75] suggesting that micro-environmental factors in vivo can also contribute to CD44 expression. Having extensively investigated the panCD44 and single variants expression in a selected panel of cancer cell lines and in HDF, we aimed to explore if this expression was linked to HA uptake. We showed that cells presented a different behaviour between binding at early time points and internalization at 24 hours, with some cells more efficient in binding whereas others were more efficient in internalization. At first sight, there was no clear correlation between binding/internalization and the amount or nature of the CD44 isoforms expressed. Perhaps this is not surprising when it is considered that many diverse factors can play a role in HA internalization; such as, CD44 activation state (which can be affected by external stimuli, posttranslational modifications, variants expression, receptor clustering in lipid rafts [37-40]), other internalization mechanisms (i.e. macropinocytosis [76]) or an association with other receptors that might be involved in HA turnover [39, 77-79]. However, in order to determine whether our data could be used in a predictive manner to relate CD44 expression with HA uptake, we devised a new scoring method that showed we could indeed relate CD44 expression to uptake of HA (Table 3). But this was only observable in cancer cell lines that predominantly expressed CD44s (CD44shigh). In fact, MDA-MB-468 and Detroit 562, which were the cell lines that expressed the highest amount of CD44v, were weak “uptakers”. HDF also showed poor HA uptake, despite having CD44shigh expression. This being a non-cancerous cell line, it is likely that the receptor present in these cells will be less active compared to the same receptor present in cancer cells. This difference between normal and malignant cells has been observed previously. For example, Yang et al. recently demonstrated that HA-coated NPs were internalized by panCD44high breast cancer cells, but not in panCD44high normal cells, thus demonstrating the existence of two different CD44 activation states in breast cancer and normal cells [80]. Similarly, liposomes conjugated with 4.8 kDa or 12 kDa HA were taken up by CD44 expressing Mia PaCa-2, but not by CD44low expressing normal primary pancreatic mesenchymal cells [43]. Bringing together different techniques to assess the expression of CD44 and the binding and internalization of HA allows us to draw conclusions regarding the importance of different 22


Page 22 of 28

Page 23 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


parameters on overall uptake. It would appear that expression of CD44s is important whereas expression of CD44v is less so. Previous work in our own lab has shown a positive correlation between the binding of HA (soluble or coated on nanoparticles) and panCD44 expression levels in differently polarized THP-1 macrophages (M1 > M0 ≥ M2). However, M2 showed higher uptake despite the lower panCD44 expression. An explanation might be that, as M2 macrophages expressed CD44v6, an increase in receptor clustering led to higher uptake [13]. Other studies have also shown that inducing the expression of CD44v can increase uptake of HA [40]. Our results are supported by other studies where it was reported that lymphoma cells presenting a low molecular weight CD44 were able to bind HA-coated surfaces, while the same cells transfected with high molecular weight CD44 were not, and had reduced local tumor formation and metastatic capacity [81]. Similar results were obtained in melanoma cells [82]. However, almost every published study on CD44s or CD44v/HA interaction, the CD44v expression was induced, genetically modified in a specific cell type or using artificial surfaces, while, here, we compared HA uptake in distinct cell types with different native CD44 profiles (wild-type). It is now well known that CD44v (mainly v6) has a role in cell signaling acting as coreceptors for tyrosine kinase receptors such as c-MET and VEGFR-2 (reviewed in [83]). This is because the variant exons contain some specific post-translational modifications (i.e. exon v3 contains a heparan sulphate modification), which can bind other growth factors or coreceptors activating pathways that promote tumorigenic functions. In particular, in this recent work it was reported that among a population of breast cancer cells, a subgroup of cells expressing CD44v showed increased capacity of lung metastasis than those expressing CD44s. Interestingly, CD44v, but not CD44s, was found to respond to osteopontin, which enhanced invasion and lung metastasis [84]. Therefore, it is possible to speculate that some cells that express CD44v might interact with other ligands and use the receptor mainly for signaling and matrix remodelling purposes rather than HA turnover. This can be the case of Detroit 562 and MDA-MB-468 cells, which we found positive to CD44v and poor in HA internalization compared to other CD44s expressing cancer cells. To conclude, in this work we have studied in detail the expression of CD44 and single variants on a range of cancer cell lines and investigated the link between this and binding and internalization. A new scoring method helped to sum up all the different techniques used for the analysis of these two different aspects revealing a correlation between CD44 and HA uptake, except when cancer cell lines expressed high amounts of CD44v. It was also confirmed 23



Page 24 of 28

that HA better targets CD44s expressed on cancer cells than the same receptor expressed on non-cancer cells. We have, therefore, not only provided strong evidence for a potential promising role of HA in targeting tumor cells, but these findings can also represent important knowledge to predict response to drug HA-based delivery systems, according to the tumor’s CD44 expression profile. ACKNOWLEDGEMENT The Bioimaging Facility microscopes used in this study were purchased with grants from BBSRC, Wellcome and the University of Manchester Strategic Research Fund. Special thanks goes to Peter March, Roger Meadows and Steve Marsden for their help with the microscopy. AS was funded via a PhD studentship from Cancer Research UK / Manchester Cancer Research Center / AstraZeneca. JMRdlR was supported by an EPSRC PhD studentship (part of the North-West

Nanoscience

(NoWNano)

Doctoral

Training

Centre,

EPSRC

Grant

EP/G03737X/1). MM was a recipient of Manchester Pharmacy School Research Fellowship, as well as funding from the Wellcome Trust and Breast Cancer Now. The work was carried out under the auspicies of the North West Centre for Advanced Drug Delivery (NoWCADD), a jointly funded venture between AstraZeneca and the University of Manchester. SUPPORTING INFORMATION: CD44 expression in cancer cells and human dermal fibroblasts, quantifications of WB and FC analysis (Figure 1SI); Trypsin cleaves surface bound-HA, confocal microscopy imaging (Figure 2SI); CD44/HA uptake correlation: scoring method (from Table 1SI to Table 13SI).

24


Page 25 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Tibbitt, M.W., J.E. Dahlman, and R. Langer, Emerging Frontiers in Drug Delivery. J Am Chem Soc, 2016. 138(3): p. 704-17. Toole, B.P., Hyaluronan: From extracellular glue to pericellular cue. Nature Reviews Cancer, 2004. 4(7): p. 528-539. Huang, G. and H. Huang, Application of hyaluronic acid as carriers in drug delivery. Drug Deliv, 2018. 25(1): p. 766-772. Cadete, A. and M.J. Alonso, Targeting cancer with hyaluronic acid-based nanocarriers: recent advances and translational perspectives. Nanomedicine (Lond), 2016. 11(17): p. 234157. Rios de la Rosa, J.M., A. Tirella, and N. Tirelli, Receptor-Targeted Drug Delivery and the (Many) Problems We Know of: The Case of CD44 and Hyaluronic Acid. Adv Healthc Mater, 2018. Ossipov, D.A., Nanostructured hyaluronic acid-based materials for active delivery to cancer. Expert Opinion on Drug Delivery, 2010. 7(6): p. 681-703. Jin, Y.-J., T. Ubonvan, and D.-D. Kim, Hyaluronic Acid in Drug Delivery Systems. Journal of Pharmaceutical Investigation, 2010. 40(spec): p. 33-43. Almalik, A., P.J. Day, and N. Tirelli, HA-Coated Chitosan Nanoparticles for CD44-Mediated Nucleic Acid Delivery. Macromolecular Bioscience, 2013. 13(12): p. 1671-1680. Almalik, A., et al., Hyaluronic acid-coated chitosan nanoparticles: Molecular weightdependent effects on morphology and hyaluronic acid presentation. Journal of Controlled Release, 2013. 172(3): p. 1142-1150. Almalik, A., et al., Hyaluronic acid (HA) presentation as a tool to modulate and control the receptor-mediated uptake of HA-coated nanoparticles. Biomaterials, 2013. 34(21): p. 53695380. de la Fuente, M., B. Seijo, and M.J. Alonso, Novel hyaluronic acid-chitosan nanoparticles for ocular gene therapy. Invest Ophthalmol Vis Sci, 2008. 49(5): p. 2016-24. Lallana, E., et al., Chitosan/Hyaluronic Acid Nanoparticles: Rational Design Revisited for RNA Delivery. Mol Pharm, 2017. 14(7): p. 2422-2436. Rios de la Rosa, J.M., et al., The CD44-Mediated Uptake of Hyaluronic Acid-Based Carriers in Macrophages. Adv Healthc Mater, 2017. 6(4). Chen, B., R.J. Miller, and P.K. Dhal, Hyaluronic acid-based drug conjugates: state-of-the-art and perspectives. J Biomed Nanotechnol, 2014. 10(1): p. 4-16. Day, A.J., The structure and regulation of hyaluronan-binding proteins. Biochem Soc Trans, 1999. 27(2): p. 115-21. Aruffo, A., et al., Cd44 Is the Principal Cell-Surface Receptor for Hyaluronate. Cell, 1990. 61(7): p. 1303-1313. Culty, M., et al., The Hyaluronate Receptor Is a Member of the Cd44 (H-Cam) Family of Cell-Surface Glycoproteins. Journal of Cell Biology, 1990. 111(6): p. 2765-2774. Liu, A.Y., Expression of Cd44 in Prostate-Cancer Cells. Cancer Letters, 1994. 76(1): p. 6369. Olsson, E., et al., CD44 isoforms are heterogeneously expressed in breast cancer and correlate with tumor subtypes and cancer stem cell markers. BMC Cancer, 2011. 11: p. 418. Li, X.P., et al., Expression of CD44 in pancreatic cancer and its significance. International Journal of Clinical and Experimental Pathology, 2015. 8(6): p. 6724-6731. Kopp, R., et al., Frequent expression of the high molecular, 673-bp CD44v3,v8-10 variant in colorectal adenomas and carcinomas. Int J Mol Med, 2009. 24(5): p. 677-683. Chen, Y., et al., The prognostic value of CD44 expression in gastric cancer: a meta-analysis. Biomed Pharmacother, 2014. 68(6): p. 693-7. Misra, S., et al., Interactions between hyaluronan and its receptors (CD44, RHAMM) regulate the activities of inflammation and cancer. Frontiers in Immunology, 2015. 6. Zoller, M., CD44: can a cancer-initiating cell profit from an abundantly expressed molecule? Nature Reviews Cancer, 2011. 11(4): p. 254-267.

25



25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

Ponta, H., L. Sherman, and P.A. Herrlich, CD44: from adhesion molecules to signalling regulators. Nat Rev Mol Cell Biol, 2003. 4(1): p. 33-45. Ruiz, P., C. Schwarzler, and U. Gunthert, Cd44 Isoforms during Differentiation and Development. Bioessays, 1995. 17(1): p. 17-24. Goodison, S., et al., Rapid analysis of distinctive CD44 RNA splicing preferences that characterize colonic tumors. Cancer Research, 1997. 57(15): p. 3140-3144. Todaro, M., et al., Colon cancer stem cells dictate tumor growth and resist cell death by production of interleukin-4. Cell Stem Cell, 2007. 1(4): p. 389-402. Guo, W. and P.S. Frenette, Alternative CD44 splicing in intestinal stem cells and tumorigenesis. Oncogene, 2014. 33(5): p. 537-8. Cohen, Z.R., et al., Localized RNAi Therapeutics of Chemoresistant Grade IV Glioma Using Hyaluronan-Grafted Lipid-Based Nanoparticles. Acs Nano, 2015. 9(2): p. 1581-1591. Ganesh, S., et al., In vivo biodistribution of siRNA and cisplatin administered using CD44targeted hyaluronic acid nanoparticles. Journal of Controlled Release, 2013. 172(3): p. 699706. Landesman-Milo, D., et al., Hyaluronan grafted lipid-based nanoparticles as RNAi carriers for cancer cells. Cancer Letters, 2013. 334(2): p. 221-227. Qhattal, H.S. and X. Liu, Characterization of CD44-mediated cancer cell uptake and intracellular distribution of hyaluronan-grafted liposomes. Mol Pharm, 2011. 8(4): p. 123346. Rosato, A., et al., HYTAD1-p20: A new paclitaxel-hyaluronic acid hydrosoluble bioconjugate for treatment of superficial bladder cancer. Urologic Oncology-Seminars and Original Investigations, 2006. 24(3): p. 207-215. Zaki, N.M., A. Nasti, and N. Tirelli, Nanocarriers for Cytoplasmic Delivery: Cellular Uptake and Intracellular Fate of Chitosan and Hyaluronic Acid-Coated Chitosan Nanoparticles in a Phagocytic Cell Model. Macromolecular Bioscience, 2011. 11(12): p. 1747-1760. Zhang, L. and N. Zhang, How nanotechnology can enhance docetaxel therapy. Int J Nanomedicine, 2013. 8: p. 2927-41. Thankamony, S.P. and W. Knudson, Acylation of CD44 and its association with lipid rafts are required for receptor and hyaluronan endocytosis. J Biol Chem, 2006. 281(45): p. 346019. Skelton, T.P., et al., Glycosylation provides both stimulatory and inhibitory effects on cell surface and soluble CD44 binding to hyaluronan. J Cell Biol, 1998. 140(2): p. 431-46. Lesley, J., et al., Variant cell lines selected for alterations in the function of the hyaluronan receptor CD44 show differences in glycosylation. J Exp Med, 1995. 182(2): p. 431-7. Sleeman, J., et al., Regulated clustering of variant CD44 proteins increases their hyaluronate binding capacity. Journal of Cell Biology, 1996. 135(4): p. 1139-1150. Wolny, P.M., et al., Analysis of CD44-hyaluronan interactions in an artificial membrane system: insights into the distinct binding properties of high and low molecular weight hyaluronan. J Biol Chem, 2010. 285(39): p. 30170-80. Mizrahy, S., et al., Hyaluronan-coated nanoparticles: the influence of the molecular weight on CD44-hyaluronan interactions and on the immune response. J Control Release, 2011. 156(2): p. 231-8. Dalla Pozza, E., et al., Targeting gemcitabine containing liposomes to CD44 expressing pancreatic adenocarcinoma cells causes an increase in the antitumoral activity. Biochimica Et Biophysica Acta-Biomembranes, 2013. 1828(5): p. 1396-1404. Yang, C., et al., The high and low molecular weight forms of hyaluronan have distinct effects on CD44 clustering. J Biol Chem, 2012. 287(51): p. 43094-107. Banerji, S., et al., Structures of the Cd44-hyaluronan complex provide insight into a fundamental carbohydrate-protein interaction. Nat Struct Mol Biol, 2007. 14(3): p. 234-9. Birzele, F., et al., CD44 Isoform Status Predicts Response to Treatment with Anti-CD44 Antibody in Cancer Patients. Clinical Cancer Research, 2015. 21(12): p. 2753-2762. Miyazaki, H., et al., CD44 exerts a functional role during EMT induction in cisplatinresistant head and neck cancer cells. Oncotarget, 2018. 9(11): p. 10029-10041.

26


Page 26 of 28

Page 27 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65.

66. 67. 68. 69.

Ssadh, H.A., et al., Measurements of heterotypic associations between cluster of differentiation CD74 and CD44 in human breast cancer-derived cells. Oncotarget, 2017. 8(54): p. 92143-92156. Paulis, Y.W., et al., CD44 enhances tumor aggressiveness by promoting tumor cell plasticity. Oncotarget, 2015. 6(23): p. 19634-46. Shirure, V.S., et al., CD44 variant isoforms expressed by breast cancer cells are functional Eselectin ligands under flow conditions. Am J Physiol Cell Physiol, 2015. 308(1): p. C68-78. Tjhay, F., et al., CD44 variant 6 is correlated with peritoneal dissemination and poor prognosis in patients with advanced epithelial ovarian cancer. Cancer Sci, 2015. 106(10): p. 1421-8. Denning, S.M., CD44 and CD45R Cluster report. In Leucocyte Typing V. White cell differentiation antigens. Oxford University Press. , 1995. 2,AS10: p. 1713 – 1719. Wang, S.J., V.B. Wreesmann, and L.Y. Bourguignon, Association of CD44 V3-containing isoforms with tumor cell growth, migration, matrix metalloproteinase expression, and lymph node metastasis in head and neck cancer. Head Neck, 2007. 29(6): p. 550-8. Ni, J., et al., CD44 variant 6 is associated with prostate cancer metastasis and chemo/radioresistance. Prostate, 2014. 74(6): p. 602-17. Wu, X.J., et al., Clinical significance of CD44s, CD44v3 and CD44v6 in breast cancer. J Int Med Res, 2015. 43(2): p. 173-9. Todaro, M., et al., CD44v6 is a marker of constitutive and reprogrammed cancer stem cells driving colon cancer metastasis. Cell Stem Cell, 2014. 14(3): p. 342-56. Lackner, C., et al., Soluble CD44 v5 and v6 in serum of patients with breast cancer. Correlation with expression of CD44 v5 and v6 variants in primary tumors and location of distant metastasis. Breast Cancer Res Treat, 1998. 47(1): p. 29-40. Cyphert, J.M., C.S. Trempus, and S. Garantziotis, Size Matters: Molecular Weight Specificity of Hyaluronan Effects in Cell Biology. Int J Cell Biol, 2015. 2015: p. 563818. Erickson, M. and R. Stern, Chain gangs: new aspects of hyaluronan metabolism. Biochem Res Int, 2012. 2012: p. 893947. Knudson, W., G. Chow, and C.B. Knudson, CD44-mediated uptake and degradation of hyaluronan. Matrix Biol, 2002. 21(1): p. 15-23. Racine, R. and M.E. Mummert, Hyaluronan Endocytosis: Mechanisms of Uptake and Biological Functions. Molecular Regulation of Endocytosis, 2012: p. 377-390. Hua, Q., C.B. Knudson, and W. Knudson, Internalization of hyaluronan by chondrocytes occurs via receptor-mediated endocytosis. J Cell Sci, 1993. 106 ( Pt 1): p. 365-75. Gennari, A., et al., Mannosylation Allows for Synergic (CD44/C-Type Lectin) Uptake of Hyaluronic Acid Nanoparticles in Dendritic Cells, but Only upon Correct Ligand Presentation. Adv Healthc Mater, 2016. 5(8): p. 966-76. Contreras-Ruiz, L., et al., Ocular tolerance to a topical formulation of hyaluronic acid and chitosan-based nanoparticles. Cornea, 2010. 29(5): p. 550-8. Gee, K., M. Kozlowski, and A. Kumar, Tumor necrosis factor-alpha induces functionally active hyaluronan-adhesive CD44 by activating sialidase through p38 mitogen-activated protein kinase in lipopolysaccharide-stimulated human monocytic cells. J Biol Chem, 2003. 278(39): p. 37275-87. Faller, C.E. and O. Guvench, Terminal sialic acids on CD44 N-glycans can block hyaluronan binding by forming competing intramolecular contacts with arginine sidechains. Proteins, 2014. 82(11): p. 3079-89. Hatano, H., et al., Overexpression of receptor for hyaluronan-mediated motility (RHAMM) in MC3T3-E1 cells induces proliferation and differentiation through phosphorylation of ERK1/2. J Bone Miner Metab, 2012. 30(3): p. 293-303. Veiseh, M., et al., Uncovering the dual role of RHAMM as an HA receptor and a regulator of CD44 expression in RHAMM-expressing mesenchymal progenitor cells. Front Cell Dev Biol, 2015. 3: p. 63. Cai, S., et al., Cellular uptake and internalization of hyaluronan-based doxorubicin and cisplatin conjugates. J Drug Target, 2014. 22(7): p. 648-57.

27



70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84.

Harris, E.N., et al., Expression, processing, and glycosaminoglycan binding activity of the recombinant human 315-kDa hyaluronic acid receptor for endocytosis (HARE). J Biol Chem, 2007. 282(5): p. 2785-97. Zhao, S., et al., CD44 Expression Level and Isoform Contributes to Pancreatic Cancer Cell Plasticity, Invasiveness, and Response to Therapy. Clinical Cancer Research, 2016. 22(22): p. 5592-5604. Matsuki, H., et al., Monoclonal antibodies with defined recognition sequences in the stem region of CD44: detection of differential glycosylation of CD44 between tumor and stromal cells in tissue. Cancer Res, 2003. 63(23): p. 8278-83. Mackay, C.R., et al., Expression and Modulation of Cd44 Variant Isoforms in Humans. Journal of Cell Biology, 1994. 124(1-2): p. 71-82. Aguiar, D.J., W. Knudson, and C.B. Knudson, Internalization of the hyaluronan receptor CD44 by chondrocytes. Exp Cell Res, 1999. 252(2): p. 292-302. Marangoni, E., et al., CD44 targeting reduces tumour growth and prevents postchemotherapy relapse of human breast cancers xenografts. Br J Cancer, 2009. 100(6): p. 91822. Greyner, H.J., et al., Inducible macropinocytosis of hyaluronan in B16-F10 melanoma cells. Matrix Biol, 2010. 29(6): p. 503-10. Lesley, J., et al., The role of the CD44 cytoplasmic and transmembrane domains in constitutive and inducible hyaluronan binding. Eur J Immunol, 2000. 30(1): p. 245-53. Lesley, J., et al., Hyaluronan binding by cell surface CD44. J Biol Chem, 2000. 275(35): p. 26967-75. Lesley, J. and R. Hyman, CD44 can be activated to function as an hyaluronic acid receptor in normal murine T cells. Eur J Immunol, 1992. 22(10): p. 2719-23. Yang, C.X., et al., Selective killing of breast cancer cells expressing activated CD44 using CD44 ligand-coated nanoparticles in vitro and in vivo. Oncotarget, 2015. 6(17): p. 1528315296. Sy, M.S., Y.J. Guo, and I. Stamenkovic, Distinct Effects of 2 Cd44 Isoforms on TumorGrowth Invivo. Journal of Experimental Medicine, 1991. 174(4): p. 859-866. Bartolazzi, A., et al., Interaction between Cd44 and Hyaluronate Is Directly Implicated in the Regulation of Tumor-Development. Journal of Experimental Medicine, 1994. 180(1): p. 5366. Orian-Rousseau, V., CD44 acts as a signaling platform controlling tumor progression and metastasis. Frontiers in Immunology, 2015. 6. Hu, J., et al., A CD44v(+) subpopulation of breast cancer stem-like cells with enhanced lung metastasis capacity. Cell Death Dis, 2017. 8(3): p. e2679.

28


Page 28 of 28

Evaluating the Efficiency of Hyaluronic Acid for Tumor Targeting via

Recommend Documents