Life Sciences 299 (2022) 120541

Contents lists available at ScienceDirect

Life Sciences
journal homepage: www.elsevier.com/locate/lifescie

Overexpression of multiple epidermal growth factor like domains 11

rescues anoikis survival through tumor cells-platelet interaction in triple
negative breast Cancer cells
Ching-Po Huang a, Yi-Fang Tsai b, c, Yen-Shu Lin b, c, Chun-Yu Liu b, d, e, Tzu-Ting Huang b, f,
Chi-Cheng Huang b, g, Jen-Hwey Chiu b, h, i, *, 1, Ling-Ming Tseng b, c, 1
a
Faculty of Medicine, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan, ROC
b
Comprehensive Breast Health Center & Division of General Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan, ROC
c
Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan, ROC
d
School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan, ROC
e
Division of Transfusion Medicine, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan, ROC
f
Women's Malignancies Branch, Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA
g
Department of Public Health, College of Public Health, National Taiwan University, Taiwan, ROC
h
Department of Surgery, Cheng-Hsin General Hospital, Taiwan, ROC
i
Institute of Traditional Medicine, School of Medicine, National Yang Ming Chiao Tung University, Taiwan, ROC

A R T I C L E I N F O A B S T R A C T

Keywords: Aims: The aim of this study was to test the hypothesis that overexpression of the Multiple Epidermal Growth
MEGF11 Factor Like Domains 11 (MEGF11) gene in TNBC cells increases tumor cell survival against anoikis via inter­
TNBC action with platelets.
Anoikis
Main methods: The role of MEGF11 was studied in human TNBC MDA-MB-231 and MDA-MB-468 cells by
Circulation tumor cell
Platelet
overexpressing MEGF11 (o/e MEGF11) using non-attached culture. Mouse wild type 4 T1 and Δmegf11-4 T1
cells were implanted into the mammary fat pad of BALB/c mice. Circulating tumor cells were isolated and
cultured. Plasma platelets were added to these cell lines to carry out a platelet binding assay by confocal mi­
croscopy. Anoikis was observed by Live/Dead staining and a quantitative PBMC-specific cytotoxic assay (with/
without platelets) was carried out to measure calcein release. The protein levels of MEGF11, Akt, and caspase 3
were assessed by Western blotting.
Key results: Reduced number of circulating Δmegf11 4 T1 cells, and 4 T1-platelet clusters and reduced p-Akt
expression, accompanied by increased specific calcein release, were observed in the Δmegf11 4T1group
compared to the 4 T1 wild type group. There was significant increased cancer-platelet binding using the o/e
MEGF11 MDA-MB-231/468 lines. Cell survival, caspase 3 activation and PBMC-specific cytotoxicity in the o/e
MEGF11 MDA-MB-468 cells, but not in the MDA-MB-231 cells, could be rescued by platelet binding (+),
compared to the platelet (− ) group.
Significance: We conclude that MEGF11 overexpression in TNBC cells rescues tumor cells from anoikis via
interaction with platelets in mouse 4 T1 cells and human MDA-MB-468 cells.

Abbreviations: ARRIVE, Animal Research: Reporting of In Vivo Experiments; CTC, circulating tumor cells; HER2, human epidermal growth factor receptor 2;
MEGF11, Multiple epidermal growth factor-like domains protein 11; PBMC, Peripheral blood mononuclear cells; PD-L1, programmed cell death ligand 1; PEAR1,
Platelet endothelial aggregation receptor 1, MEGF12; PRP, platelet-rich plasma; RFP, Red Fluorescent Protein; TNBC, triple negative breast cancer.
* Corresponding author at: Comprehensive Breast Health Center, Taipei Veterans General Hospital, No. 201, Sec.II, Shih-pei Rd, Taipei 112, Taiwan, ROC.
E-mail addresses: [email protected] (C.-P. Huang), [email protected] (Y.-F. Tsai), [email protected] (Y.-S. Lin), [email protected] (C.-Y. Liu), tzu-
[email protected] (T.-T. Huang), [email protected] (C.-C. Huang), [email protected] (J.-H. Chiu), [email protected] (L.-M. Tseng).
1
Tseng, LM and Chiu, JH contributed equally in this manuscript.

https://doi.org/10.1016/j.lfs.2022.120541
Received 1 January 2022; Received in revised form 27 March 2022; Accepted 4 April 2022
Available online 7 April 2022
0024-3205/© 2022 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
C.-P. Huang et al. Life Sciences 299 (2022) 120541

1. Introduction 11 in platelet adhesion-induced Akt activation and anoikis resistance.

Accordingly, the aim of this study was to test the hypothesis that
Triple-negative breast cancer (TNBC) does not express either estro­ MEGF11 gene overexpression in TNBC cells increases tumor cells sur­
gen receptors (ER), or progesterone receptors (PR) and they also do not vival by suppressing anoikis via an interaction with platelets.
overexpress human epidermal growth factor receptor 2 (HER2). TNBC is
characterized by a poor prognosis, being more common in younger 2. Methods
patients, and having high resistance to chemotherapy, hormone therapy
and target therapy [1]. Several biomarker, such as programmed cell 2.1. Cell lines and reagents
death ligand 1 (PD-L1) and poly (ADP-ribose) polymerase (PARP), have
been identified and targeted as part of immunotherapy and target Human TNBC cell lines MDA-MB-231 (RRID:CVCL_0062) and MDA-
therapy [2], but there remains unmet needs in terms of TNBC treatment. MB-468 (RRID:CVCL_0419) and the mouse mammary tumor 4 T1 cell
There is evidence to suggest that poor prognosis in breast cancer line were obtained from the American Type Culture Collection (ATCC,
might be related to an increase in number of circulating tumor cells Manassas, VA) and were maintained using F12 MEM, high glucose MEM,
(CTCs) present in the patients [3,4]. The survival of tumor cells in the and RPMI, all supplemented with 10% FBS, 2 mM L-glutamine and
circulation requires three characteristics, namely, resistance to anoikis, penicillin/streptomycin, respectively. All cell lines were cultured at
the ability to escape immune surveillance, and being able to survive 37 ◦ C in a humidified atmosphere containing 5% CO2.
blood shearing forces. Resistance to anoikis requires activation of
several signaling pathways related to anchorage-independent growth 2.2. Gene manipulations of the TNBC cell lines
and the epithelial–mesenchymal transition, including Akt, TrkB, Src, as
well as a number of other signaling pathways [5]. Several hypothesis Lentiviral infection was carried out to allow short hairpin RNA
related to how CTCs evade immune surveillance have been proposed, (shRNA) silencing of the MEGF11 gene (Δ megf11) in the mouse 4 T1 cell
such as adherence to platelets, adherence to myeloid-derived suppressor line; this followed the protocol of the Academia Sinica, ROC. In addition,
cells [6,7], and mediation via PD-L1 [8]. Nonetheless, the role of Mul­ the MEGF11 over-expression cell line (o/e MEGF11) MDA-MB-231/468
tiple epidermal growth factor-like domains protein 11 (MEGF11) in the had already been established and has been described previously [11].
survival of CTCs remains to be elucidated. Infection efficiency was validated by Western blot analysis (supple­
PEAR1 (Platelet endothelial aggregation receptor 1, also known as mentary reference 2). Red Fluorescent Protein (RFP) was cloned into
MEGF12) is a paralog of MEGF11. Previous studies have demonstrated TNBC cells for the platelet binding assay. The study designs using mouse
that MEGF12, when present on platelets, results in persistent PI3K and 4 T1 cell line (Fig. 1) and human MDA-MB-231/MDA-MB-468 lines
Akt activation [9] and that the ligand of MEGF12 is High Affinity (Fig. 2) were summarized.
Immunoglobulin E Receptor Subunit α (FcεR1α), which is found on
platelets [10]. Recently, our results have suggested that knocked down
2.3. Selection of circulating mouse mammary breast cancer 4 T1 cells
of MEGF11 in TNBC cell lines decreases cell proliferation via the down-
regulation of the Akt-mTOR and NF-κB signaling pathways. In addition,
All protocols that involved experimental mice were approved by the
overexpression of MEGF11 in TNBC cell lines has been found to signif­
Institutional Animal Committee of Taipei Veterans General Hospital
icantly increase Akt activation and increase the levels of pro-
(No. 2018-029) and followed the ARRIVE (Animal Research: Reporting
inflammatory cytokines and chemokines. Finally, in a clinical context,
of In Vivo Experiments) guidelines. The circulating murine mammary
overexpression of MEGF11 has been shown to increase the risk of
breast cancer 4 T1 cells (4 T1-CTC) was obtained from 4 T1 bearing mice
recurrence in TNBC patients [11].
and selected using 6-thioguanine (60 μM) (A4882, Sigma-Aldrich, MO,
Platelets are known to influence cancer progression via several
USA) as described previously [11,17] (supplementary reference 3).
mechanisms, including the elevation of proliferative signals, an ability
to increase cell survival, an ability to support angiogenesis, by affecting
the epithelial–mesenchymal transition, by changing the status of cancer 2.4. Platelet isolation
stem cells, and by helping cells to escape immune surveillance [7,12].
Based on the premise that platelets are strongly associated with cancer, The animal experiments in this study followed the ARRIVE (Animal
doxorubicin-loaded platelets have been demonstrated to suppress tumor Research: Reporting of In Vivo Experiments) guidelines and were
growth in vivo [13], and this has been proposed as a novel treatment for approved by the Institutional Animal Committee of Taipei Veterans
lymphoma. Furthermore, miR-24 is reported to be transferred to the General Hospital (No. 2020-047). After the approval of the Human
tumor site by platelet-derived microparticles infiltration, and this Institutional Review Board of Taipei Veterans General Hospital (# 2020-
eventually induces tumor cell apoptosis and an inhibition of cancer 08-007AC), human platelets were isolated from healthy volunteers.
progression in vivo [14]. Isolation of platelet in mouse and human plasma was performed as
There is consensus that distant metastasis results in poor prognosis of described previously [18]. In brief, 5 mL of whole blood was added to an
TNBC [1] and our previous works have demonstrated that MEGF11 is
overexpressed in tumors with subsequent recurrence compared to those
without recurrence. Besides, knocked down of MEGF11 in a mouse 4 T1
mammary cancer cell line seems to significantly reduce the number of
CTCs in vivo [11]. Recent investigations demonstrated that CTCs cluster
with neutrophils results in an increase in metastasis potential [15,16].
Nevertheless, information concerning the relationship between CTCs
cluster with platelets and CTCs survival remains clear. Due to the fact
that MEGF11 and MEGF12 share similarity in terms of both their
structure (supplementary reference 1) and their signaling pathways, we
hypothesize that TNBC cells overexpressing MEGF11, on adhering to
platelets, will increase Akt activation, and that this will eventually lead
to anoikis resistance among CTCs. Therefore, we became interested in
investigating several questions. Firstly, whether MEGF11 has an impact
on platelet binding to TNBC cells and, secondly, what is the role of MEGF Fig. 1. Study designs and experimental flow chart in mouse 4 T1 cell line.

2
C.-P. Huang et al. Life Sciences 299 (2022) 120541

Fig. 2. Study designs and experimental flow chart in human MDA-MB-231/MDA-MB-468 lines.

equivalent volume of PBS containing heparin in a 15-mL polypropylene different intervals of incubation, tumor-platelet aggregates underwent
tube. Mixture was centrifuged at 100 ×g for 15 min at room temperature either Western blot analysis or the anoikis assay using Live (Calcein
(RT) to collect platelet-rich plasma (PRP). The PRP was transferred to a -AM) / Dead (Ethidium homodimer-1) staining. Finally aggregates were
new 15-mL polypropylene tube and centrifuged at 800 ×g for 20 min at observed by fluorescence microscopy [19].
RT to collect platelets as a pellet, which was then resuspended in
Tyrode's buffer (135 nM NaCl, 5 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2, 2.7. Western blotting analysis
20 mM HEPES, pH 7.4, 5 mM Glucose and 0.5% (w/w) bovine serum
albumin) to give a concentration of 4 × 108 platelets/mL. In both mouse Cultured cells were lysed in a buffer containing 150 mM KCl, 10 mM
and human PRP fractions, thrombin (1 μg/mL) were used in each ex­ Tris pH 7.4, 1% Triton X-100 plus phosphatase inhibitor and protease
periments. For the platelet binding assay, the PRP was stained with 2 μM inhibitors cocktail (Complete Mini; Roche, Mannheim, Germany). The
calcein-AM (L3224, Invitrogen Detection Technologies, Eugene, OR) for protein concentrations in cell homogenates were measured using the
45 min at RT in the dark before the second centrifugation. Bradford's method [20]. In total, 30 μg of protein was separated by 10%
SDS-PAGE and then transferred to a PVDF membrane. Next, the mem­
2.5. Platelet binding assay brane was blocked with 5% bovine serum albumin and then probed with
the various specific primary antibodies, namely against MEGF11
Human MDA-MB-231 and MDA-MB-468 cells were labeled with red (GTX49102, Genetex), p-Akt (Ser473, #9271S, Cell Signaling Technol­
fluorescent protein (RPF) by lentiviral infection. The efficiency of RFP ogy, Beverly, MA, USA), t-Akt (#9272S, Cell Signaling Technology,
transfection was evaluated by fluorescent microscopy. The platelet- Beverly, MA, USA), caspase 3 (full length- 35, cleaved-17, #9662, Cell
tumor binding assay followed a previous protocol with some modifica­ Signaling Technology, Beverly, MA, USA), and β-Actin (GTX629630,
tion [18]. In brief, 50 μL of RFP labeled MDA-MB-468, or MDA-MB-231 Genetex). Western Blot image denoting the area cropped was shown in
cells (2 × 106 cells/mL) in Tyrode's buffer was added to a 1.5-mL tube. supplementary reference 4.
The PRP was then labeled with calcein AM according to the manufac­
turer's instructions (L3224, Invitrogen Detection Technologies, Eugene, 2.8. Peripheral blood mononuclear cells (PBMC) isolation
OR). An equivalent volume of calcein AM (+)-platelets suspension (4 ×
108 platelets/mL) in Tyrode's buffer was added to the RFP-labeled A volume of 2 mL of whole blood was added to an equivalent volume
cancer cells with final tumor/platelet ratio of 1:200. The platelet-cell of PBS containing heparin in a 15-mL polypropylene tube. Next, 3 mL of
mixture was then incubated at 37 ◦ C for 30 min, which was followed Ficoll-Paque PREMIUM (density: 1.084 for mouse blood and 1.073 for
by centrifugation at 100 ×g for 5 min to pull down the cell-platelet human blood)) (17–5446-02, GE Healthcare Bio-Sciences, Sweden)
aggregates. The pellet (the cell-platelet aggregates) was fixed with 4% gradient medium was added to another 15-mL polypropylene tube, and
paraformaldehyde for 30 min at RT and then observed by either fluo­ the diluted blood sample were gently layered onto the Ficoll-Paque
rescent microscopy or confocal microscopy. Three random low and high medium. The samples were then centrifuge at 400 ×g for 30 min. The
power fields (100×, or 400×) images were captured per slide by a isolated buffy coat cells were transferred to a new 15-mL polypropylene
blinded observer. tube containing 6 mL of PBS. The mixture was centrifuged again at 400
×g for 10 min, after which the supernatant was discarded and the
2.6. Anoikis induction remaining PBMC cells were resuspended in Tyrode's buffer at a con­
centration of 1 × 107 cells/mL.
To induce anoikis, TNBC cells were prevented from adhering to an
anoikis plate according to the manufacture's instruction (CBA081m Cell 2.9. PBMC cytotoxic Calcein release assay
Biolabs, San Diego, CA). In brief, 100 μL/well of TNBC cells (2 × 104/
well) with or without platelets were plated onto anoikis plate and The PBMC cytotoxic Calcein release assay was adapted from the
cultured at 37 ◦ C in a humidified atmosphere containing 5% CO2. After method of Simona Neri and colleagues [21]. Mouse 4 T1 cells (WT,

3
C.-P. Huang et al. Life Sciences 299 (2022) 120541

Δmegf11) and human MDA-MB-231 and MDA-MB-468 (WT, scramble, group is compared to the vehicle or no treatment group.
o/e MEGF11) cells were labeled with calcein AM according to the
manufacturer's instructions (L3224, Invitrogen Detection Technologies, 3. Results
Eugene, OR). Next, in brief, 50 μL/well of MDA-MB-468, or MDA-MB-
231 cells (2 × 106 cells/mL) together with 50 μL/well of Tyrode's To investigate the difference between mouse 4T1 cells and 4T1
buffer or platelets (4 × 108 platelets/mL) were mixed with 50 μL/well of circulating tumor cells (4T1-CTC) in terms of Akt signaling and the
Tyrode's buffer alone or PBMC (1 × 107 cells/mL) in order to give a final anoikis assay, Western blotting of MEGF11 and its related signaling Akt
target effector ratio of 1:5 and then the mixtures were added to wells of a protein were performed. The results showed that there were no signifi­
96-well plate. The control used to access the maximum release (Tyrode's cant differences between 4T1 and 4T1-CTC phenotypes in terms of
buffer with 2% Triton X100) was also added to the same plate. After 4 h morphology (Fig. 3A), MEGF11 expression, Akt signaling (Fig. 3B) and
of incubation at 37 ◦ C in a humidified atmosphere containing 5% CO2, the anoikis assay (Fig. 3C).
the media in each well were harvested and centrifuged at 800 ×g for 20 After mouse platelets were labeled with Calcein-AM (green color),
min to prevent contamination with suspended cells, platelets or other 4T1-platelet binding was observed by confocal microscopy, the results
debris. After this, the supernatant was transferred to another 96-well demonstrated that there were more 4T1-platelet clusters in WT group
plate for fluorescence detection (Ext./Exm.: 494/517 nm). The specific than Δmegf11 4T1 group (Fig. 4A). Under fluorescent microscopy ex­
percentage of lysis was calculated as [(test release − spontaneous amination (Fig. 4B), there was a significant decreased 4T1-platelet
release)/(maximum release − spontaneous release)] × 100%. cluster in Δmegf11 4T1 group compared to the WT group (Fig. 4C).
Western blot analysis showed an increased p-Akt expression in WT-
platelet (+) than WT-platelet (− ) and a decreased p-Akt expression
2.10. Statistic analysis
(Fig. 4D, E) in Δmegf11 4T1 group compared to the WT group. Besides,
there was an increase in specific calcein release (PBMC-specific cyto­
All results are expressed as means ± SEM. The differences between
toxic assay, Fig. 4F) by the Δmegf11 4T1group compared to the 4T1 WT
two groups were analyzed using the Mann-Whitney U test or the Stu­
group. These findings suggest that MEGF11 plays a crucial role in
dent's t-test, while differences between groups at each time point were
platelet binding to 4T1 cells.
identified by one-way ANOVA, followed by Dunnet's post hoc test. A p-
value <0.05 is considered statistically significant when the experimental

Fig. 3. MEGF11 expression in the mouse 4T1 mammary cancer cell line and 4T1 circulation tumor cells. Using the spontaneously occurring mouse mammary tumor
4T1 cell line as an in vivo metastatic model, circulating 4T1 cells (A) selected with 6-thioguanine were observed under light microscopy (200×) as described in
Methods. MEGF11, p-Akt, t-Akt, and β-Actin expression (B) was analyzed by Western blot. For anoikis assay, 4T1 and 4T-CTC cells were seeded onto adherent and
anoikis (non-adherent) plate for 24 h and 48 h, followed by stained with Calcein-AM (green) and Ethidium homodimer-1 (live and dead stain) solutions (C) and the
cell death rate (%) were quantified. n = 4 in each group. Bar indicates 50 μm. (For interpretation of the references to color in this figure legend, the reader is referred
to the web version of this article.)

4
C.-P. Huang et al. Life Sciences 299 (2022) 120541

Fig. 4. The role of MEGF11 on platelet-cancer cell interaction in mouse 4T1 cell. MEGF11 gene was knocked down in mouse 4T1 cells. After 4T1 cells (WT or
Δmegf11) were co-cultured with or without calcein-AM labeled platelets (green color) for 1 h, they were observed under a confocal microscope (A, bar = 50 μm) and
a light microscope (B, bar = 50 μm). The platelet-cancer cell clusters (B) were quantified (C, n = 5 in each group) and the Akt signaling (p-Akt and t-Akt) was
analyzed by Western blot (D) and quantified (E, n = 5 in each group). Peripheral blood mononuclear cell (PBMC)-specific lysis (F, n = 3 in each group) on 4T1 cancer
cells were quantified by calcein-AM release assay as described in Methods. *, p < 0.05 compared to the WT group; #, p < 0.05 compared to the platelet (− ) group by
Mann-Whitney U test. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3.1. The role of MEGF11 in platelet binding to human TNBC cells 3.3. The binding of platelets did not affect MDA-MB-231 anoikis
resistance
To understand the roles of MEGF11 in platelet binding to human
TNBC cell lines, gene manipulation of MEGF11 in MDA-MB-231 and The MDA-MB-231 cell lines were cultured with and without platelet
MDA-MB-468, including wild type, scramble and o/e MEGF11, was treatment on the anoikis plates and analyzed using Western blotting,
carried out and the resulting cell lines were labeled with red fluorescent Live/Dead stain and PBMC cytotoxic assay. Although there was a sig­
protein (RFP). Then, these cell lines were co-cultured with calcein AM- nificant increased cancer-platelet binding in the o/e MEGF11 MDA-MB-
labeled platelets in order to measure the effects of MEGF11 on platelet 231 line (Fig. 7A, B) group, there were no significant changes in cell
binding. When MDA-MB-468 (Fig. 5A) and MDA-MB-231 (Fig. 5D) were survival as measured by the anoikis assay (Fig. 7C), by caspase 3 acti­
observed using confocal microscopy, cancer cells (red color) – platelets vation (Fig. 7D, E) and by PBMC-specific cytotoxicity (Fig. 7F) in o/e
(green color) binding was obvious. By Western blot analysis (Fig. 5B, E), MEGF11 MDA-MB-231 group when the platelet (+) group and the
the results showed that there was a significant increase in Akt activation platelet (− ) group were compared.
in MDA-MB-468 (Fig. 5C) and MDA-MB-231 (Fig. 5F) cells, but no dif­
ference between platelet (− ) and platelet (+) groups. 4. Discussion

3.2. The effect of platelet binding on MDA-MB-468 anoikis resistance Our previous studies have suggested that MEGF11 plays an impor­
tant role in the mechanisms associated with TNBC recurrence via cyto­
To understand the protective effect of cancer-platelet binding via kine (especially IL-17A) and chemokine cascades [11]; this favours a
MEGF11 overexpression on anoikis resistance, the three lines of MDA- microenvironment that is suitable for tumor metastasis [19]. There is
MB-468 were cultured with or without platelet treatment on anoikis evidence that the role of MEGF11 expression in tumor recurrence and
plates and then the cells were analyzed by Western blotting, by Live/ overall survival is minimal in hormone receptor or HER-2 positive breast
Dead staining and by PBMC cytotoxic assay. The results showed that cancer patients, even though MEGF11 protein is also expressed in these
there was a significant increase in cell-platelet binding when using the breast subtypes (supplementary reference 5). In this article, we have
o/e MEGF11 MDA-MB-468 line (Fig. 6A, B). Cell survival as measured demonstrated that MEGF11 has a role in TNBC metastasis and that the
by the anoikis assay (Fig. 6C), by caspase 3 activation (Fig. 6D, E) and by protein seems to work through cancer-platelet interaction via the Akt
PBMC-specific cytotoxicity (Fig. 6F) of the o/e MEGF11 MDA-MB-468 signaling pathway.
cells was found to be rescued when the platelet binding (+) group was Although the mouse 4T1 breast cancer line is a commonly used
compared to the platelet (− ) group. model for the study cancer metastasis [22], information related to cir­
culation tumor cells (CTCs) and the 4T1 metastasis model needs further

5
C.-P. Huang et al. Life Sciences 299 (2022) 120541

Fig. 5. The role of MEGF11 on platelet-cancer cell binding in human triple negative cancer cell lines MDA-MB-468 and MDA-MB-231. MEGF11 gene was over­
expressed in human triple negative cancer cell (TNBC) lines MDA-MB-468 and MDA-MB-231 (WT, scramble or o/e MEGF11) and they were labeled with red
fluorescent protein (RFP). These RFP-labeled cancer cells (red color) were co-cultured with or without calcein-AM labeled platelets (green color) for 1 h and were
observed under a confocal microscope (A, D, bar = 100 μm). The lower panels were the magnified field obtained from the upper panels. The Akt signaling (p-Akt and
t-Akt) was analyzed by Western blot (B, E) and quantified (C, F, n = 3 in each group). *, p < 0.05 compared to the WT group by Mann-Whitney U test. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

elucidation. In this study, circulating 4T1 cells were isolated from the prognosis of TNBC patients where there is overexpression of MEGF11
peripheral blood of tumor-bearing mice; the cells were selected through [11].
the use of a 6-thioguanine-containing culture [23]. Our results demon­ Platelet endothelial aggregation receptor 1 (PEAR1, also known as
strate that there was no significant change in MEGF11 protein level MEGF12) is a paralog of MEGF11 [26]. Previous studies have indicated
between the 4T1 and 4T1-CTC lines. After a 4T1 cell line overexpressing that MEGF12, when present on activated platelets, is able to produce
MEGF11 was established by lentivirus infection, it was found that there persistent PI3K and Akt activation [9]. Given the similarity in sequence
was no significant difference in MEGF11 protein expression between the and signaling pathways between MEGF11 and MEGF12, we hypothesize
4T1 wild type and the o/e megf11 4T1 line (Supplementary reference 6). that MEGF11 may play a crucial role in TNBC's platelet-tumor interac­
This finding suggests that 4T1 wild type is already a cell line that tion scenario. Our findings reveal that overexpression of MEGF11 in­
overexpresses MEGF11. creases the platelet binding ability of human TNBC cell lines and that
Intercellular interactions in the bloodstream are important to the knock-down of MEGF11 in 4T1 brings about a decrease in Akt activa­
survival of circulating tumor cells (CTCs).There is evidence showing that tion. Although there is structural and functional similarity between
CTCs cluster with white blood cells, especially neutrophils, and that this MEGF11 and MEGF12 (supplementary reference 1), the expression of
results in an increase in metastasis via an acceleration of their cell cycle MEGF12 in human TNBC MDA-MB 231/468 lines is different from that
[16]. In addition, breast cancer oligoclonal CTC-clusters have been of MEGF11, suggesting the diverse expression pattern between these
shown to have a 23 fold to 50 fold higher metastatic potential as two proteins in such cell lines (supplementary reference 7). In addition,
compared with single CTCs [15]. Furthermore, on binding with neu­ we have found that MEGF11 provides survival benefits when adhered to
trophils, cancer cells are also able to interact with platelets and this platelets by activating the Akt signaling pathway, by enhancing anoikis
seems to facilitate metastasis via a number of mechanisms including resistance and by protecting tumor cells against PBMC cytotoxicity.
increased cell survival, a support of angiogenesis, an elevation in pro­ Nevertheless, these effects seem to be cell line-dependent.
liferative signaling, alterations to the status of the cancer stem cells and There is consensus that TNBC cells present as different subtypes
providing help with escaping immune surveillance [7,12,24]. Recent clinically [27–29]. Due to significant difference in signaling pathways,
evidence demonstrates that platelet reduces anoikis and promotes the complete pathological response to various neoadjuvant chemother­
metastasis by activating YAP1 signaling in mice [25]. Our results show apies can range from 18% to 41% for the different subgroups of TNBC
that overexpression of MEGF11 in MDA-MB-231 and MDA-MB-468 cells [30]. There are also TNBC subtypes associated with the various cell
increases the ability of these cells to bind. This finding seems to partially culture system, including MDA-MB-468 (Basal like 1), MDA-MB-231
explain the decrease in the number of CTCs found in vivo when the Δ (Mesenchymal stem cell like), BT-549 (Mesenchymal), MDA-MB-453
megf11 4T1 cell line is used and this, in turn, explains the poorer clinical (Luminal androgen receptor), and DU4475 (Immune modulatory)

6
C.-P. Huang et al. Life Sciences 299 (2022) 120541

Fig. 6. The role of MEGF11 on platelet-cancer cell interaction in human triple negative cancer cell MDA-MB-468 line. MEGF11 gene was overexpressed in MDA-MB-
468 cells (WT, scramble or o/e MEGF11) and they were co-cultured with or without calcein-AM labeled platelets (green color) for 1 h and were observed under a
confocal microscope (A, D, bar = 50 μm). The platelet-cancer cell binding (platelet count per cancer cell) were quantified (B, n = 3 in each group). After cancer cells
were cultured in non-adherent plate for 24 h, the anoikis rate (C, n = 3 in each group) was determined by live and dead stain as described in methods, while the
caspase 3 signaling (pro-caspase 3 and cleaved caspase 3) was analyzed by Western blot (D) and quantified (E, n = 5 in each group). Peripheral blood mononuclear
cell (PBMC)-specific lysis (F, n = 5) on MDA-MB-468 cells was quantified by calcein-AM release assay as described in Methods. *, p < 0.05 compared to the WT group;
#, p < 0.05 compared to the platelet (− ) group by Mann-Whitney U test. (For interpretation of the references to color in this figure legend, the reader is referred to
the web version of this article.)

[29]. Recent evidence has demonstrated that there are different re­ 5. Conclusion
sponses to berberine treatment among these TNBC subtypes; these
include an increase in the S + G2/M fraction for the MDA-MB-231 and We conclude that MEGF11 overexpression in TNBC cells increases
MDA-MB-453 lines, and an increased in the G0/G1 fraction for the MDA- tumor cell survival against anoikis via an interaction with platelets using
MB-468 and BT-549 cell lines [31]. Our results show that platelet mouse 4T1 cells and human MDA-MB-468 cells. These novel findings
binding provides a significant protective effect against anoikis with the could lead to the development of a novel therapeutic strategy for the
o/e MEGF11 MDA-MB-468 cell line, but not with the o/e MEGF11 MDA- prevention of TMBC metastasis.
MB-231 cell line. We propose that this effect might be due to three
possibilities. Firstly, there might be differences in the level of over- Credit authorship contribution statement
expression of MEGF11 in the different cell lines. Secondly, the
different cell line might have important differences in other character­ Chiu, JH formed the idea. Tseng, LM and Chiu, JH contributed
istics. Thirdly, that both of the above effects are important. When equally in this manuscript. Tseng, LM supervised the experiments.
compared with their wild type, the level of MEGF11 over-expression in Huang, CP wrote the manuscript. Huang, TT and Liu, CY constructed the
human TNBC cell lines is one where MDA-MB-468 shows greater over-expression MEGF11 plasmid. Tsai, YF, Lin YS, Huang CC, and
expression than MDA-MB-231 and the mouse 4T1 cell line is a MEGF11 Huang, CP performed the experiments and analyzed the data.
constitutive expressed cell line (supplementary reference 8). The
platelet protective effect in these three lines is 4 T1 > MDA-MB-468 > Ethics approval and consent to participate
MDA-MB-231. Although we cannot exclude the 2nd and the 3rd possi­
bilities, the first possibility would seem to make an important contri­ Study protocols involving experimental mice followed ARRIVE
bution to the protective effect against anoikis of platelets during cancer (Animal Research: Reporting of In Vivo Experiments) guidelines and
metastasis. were approved by the Institutional Animal Committee of Taipei Veter­
In summary, MEGF11 overexpression in TNBC cells not only up- ans General Hospital (No. 2020-047).The human study for platelet
regulates cytokines and chemokines expression, increases circulation isolation and cancer-platelet biding assays were approved by the Insti­
tumor cells, but also increases the resistance of tumor cell against tutional Review Board of Taipei Veterans General Hospital (# 2020-08-
anoikis via an interaction with platelets in both mouse 4 T1 cells and 007AC).
human MDA-MB-468 cells, but not in MDA-MB-231 cells.
Consent of publication

Not applicable.

7
C.-P. Huang et al. Life Sciences 299 (2022) 120541

Fig. 7. The role of MEGF11 on platelet-cancer cell interaction in human triple negative cancer cell MDA-MB-231 line. MEGF11 gene was overexpressed in MDA-MB-
231 cells (WT, scramble or o/e MEGF11) and they were co-cultured with or without calcein-AM labeled platelets (green color) for 1 h and were observed under a
confocal microscope (A, D, bar = 50 μm). The platelet-cancer cell binding (platelet count per cancer cell) were quantified (B, n = 3 in each group). After cancer cells
were cultured in non-adherent plate for 24 h, the anoikis rate (C, n = 3 in each group) was determined by live and dead stain as described in methods, while the
caspase 3 signaling (pro-caspase 3 and cleaved caspase 3) was analyzed by Western blot (D) and quantified (E, n = 5 in each group). Peripheral blood mononuclear
cell (PBMC)-specific lysis (F, n = 5 in each group) on MDA-MB-468 cells was quantified by calcein-AM release assay as described in Methods. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version of this article.)

Data availability Foundation.

All data generated or analyzed during this study are included in this Appendix A. Supplementary data
published article (and its Supplementary Information files).
Supplementary data to this article can be found online at https://doi.
Declaration of competing interest org/10.1016/j.lfs.2022.120541.

We declare that we have no conflicts of interest, including financial References

and non-financial interests such as the following items.
[1] F. Podo, L.M. Buydens, H. Degani, R. Hilhorst, E. Klipp, I.S. Gribbestad, et al.,
Unpaid membership in a government or non-governmental Triple-negative breast cancer: present challenges and new perspectives, Mol.
organization. Oncol. 4 (2010) 209–229.
Unpaid membership in an advocacy or lobbying organization. [2] D.B. Page, H. Bear, S. Prabhakaran, M.E. Gatti-Mays, A. Thomas, E. Cobain, et al.,
Two may be better than one: PD-1/PD-L1 blockade combination approaches in
Unpaid advisory position in a commercial organization. metastatic breast cancer, NPJ Breast Cancer 5 (2019) 34.
Writing or consulting for an educational company. [3] M. Cristofanilli, G.T. Budd, M.J. Ellis, A. Stopeck, J. Matera, M.C. Miller, et al.,
Acting as an expert witness. Circulating tumor cells, disease progression, and survival in metastatic breast
cancer, N. Engl. J. Med. 351 (2004) 781–791.
[4] Y. Zhang, Y. Lv, Y. Niu, H. Su, A. Feng, Role of circulating tumor cell (CTC)
Data availability monitoring in evaluating prognosis of triple-negative breast cancer patients in
China, Med. Sci. Monit. 23 (2017) 3071–3079.
[5] P. Paoli, E. Giannoni, P. Chiarugi, Anoikis molecular pathways and its role in
Data will be made available on request.
cancer progression, Biochim. Biophys. Acta 1833 (2013) 3481–3498.
[6] Q. Liu, Q. Liao, Y. Zhao, Myeloid-derived suppressor cells (MDSC) facilitate distant
Acknowledgements metastasis of malignancies by shielding circulating tumor cells (CTC) from immune
surveillance, Med. Hypotheses 87 (2016) 34–39.
[7] X.L. Lou, J. Sun, S.Q. Gong, X.F. Yu, R. Gong, H. Deng, Interaction between
We are in debt to Chou, SS, and Hsu, TH for their technique supports. circulating cancer cells and platelets: clinical implication, Chin. J. Cancer Res. 27
This work was supported by the biobank from the Division of Experi­ (2015) 450–460.
mental Surgery, Department of Surgery, Taipei Veterans General Hos­ [8] M. Mazel, W. Jacot, K. Pantel, K. Bartkowiak, D. Topart, L. Cayrefourcq, et al.,
Frequent expression of PD-L1 on circulating breast cancer cells, Mol. Oncol. 9
pital. We also thank Dr. Ralph Kirby for his English editing on this (2015) 1773–1782.
manuscript. This work was supported by grants from the Ministry of [9] A. Kauskot, M. Di Michele, S. Loyen, K. Freson, P. Verhamme, M.F. Hoylaerts,
Health and Welfare (Center of Excellence for Cancer Research at Taipei A novel mechanism of sustained platelet alphaIIbbeta3 activation via PEAR1,
Blood 119 (2012) 4056–4065.
Veterans General Hospital phase III, MOHW110-TDU-B-212-010001),
Taipei Veterans General hospital (V110C-168), and Melissa Lee Cancer

8
C.-P. Huang et al. Life Sciences 299 (2022) 120541

[10] Y. Sun, C. Vandenbriele, A. Kauskot, P. Verhamme, M.F. Hoylaerts, G.J. Wright, [21] S. Neri, E. Mariani, A. Meneghetti, L. Cattini, A. Facchini, Calcein-acetyoxymethyl
A human platelet receptor protein microarray identifies the high affinity cytotoxicity assay: standardization of a method allowing additional analyses on
immunoglobulin E receptor subunit α (FcεR1α) as an activating platelet recovered effector cells and supernatants, Clin. Diagn. Lab. Immunol. 8 (2001)
endothelium aggregation receptor 1 (PEAR1) ligand, Mol. Cell. Proteomics 14 1131–1135.
(2015) 1265–1274. [22] L.H. Lai, Q.H. Fu, Y. Liu, K. Jiang, Q.M. Guo, Q.Y. Chen, et al., Piperine suppresses
[11] J.H. Chiu, L.M. Tseng, T.T. Huang, C.Y. Liu, J.Y. Wang, C.P. Huang, et al., MEGF11 tumor growth and metastasis in vitro and in vivo in a 4T1 murine breast cancer
is related to tumour recurrence in triple negative breast cancer via chemokine model, Acta Pharmacol. Sin. 33 (2012) 523–530.
upregulation, Sci. Rep. 10 (2020) 8060. [23] S. Huang, P.J. Houghton, Targeting mTOR signaling for cancer therapy, Curr. Opin.
[12] A.T. Franco, A. Corken, J. Ware, Platelets at the interface of thrombosis, Pharmacol. 3 (2003) 371–377.
inflammation, and cancer, Blood 126 (2015) 582–588. [24] T. Placke, M. Orgel, M. Schaller, G. Jung, H.G. Rammensee, H.G. Kopp, et al.,
[13] P. Xu, H. Zuo, B. Chen, R. Wang, A. Ahmed, Y. Hu, et al., Doxorubicin-loaded Platelet-derived MHC class I confers a pseudonormal phenotype to cancer cells that
platelets as a smart drug delivery system: an improved therapy for lymphoma, Sci. subverts the antitumor reactivity of natural killer immune cells, Cancer Res. 72
Rep. 7 (2017) 42632. (2012) 440–448.
[14] J.V. Michael, J.G.T. Wurtzel, G.F. Mao, A.K. Rao, M.A. Kolpakov, A. Sabri, et al., [25] M. Haemmerle, M.L. Taylor, T. Gutschner, S. Pradeep, M.S. Cho, J. Sheng, et al.,
Platelet microparticles infiltrating solid tumors transfer miRNAs that suppress Platelets reduce anoikis and promote metastasis by activating YAP1 signaling, Nat.
tumor growth, Blood 130 (2017) 567–580. Commun. 8 (2017) 310.
[15] N. Aceto, A. Bardia, D.T. Miyamoto, M.C. Donaldson, B.S. Wittner, J.A. Spencer, et [26] A.V. Krivtsov, F.N. Rozov, M.V. Zinovyeva, P.J. Hendrikx, Y. Jiang, J.W. Visser, et
al., Circulating tumor cell clusters are oligoclonal precursors of breast cancer al., Jedi–a novel transmembrane protein expressed in early hematopoietic cells,
metastasis, Cell 158 (2014) 1110–1122. J. Cell. Biochem. 101 (2007) 767–784.
[16] B.M. Szczerba, F. Castro-Giner, M. Vetter, I. Krol, S. Gkountela, J. Landin, et al., [27] M.D. Burstein, A. Tsimelzon, G.M. Poage, K.R. Covington, A. Contreras, S.A. Fuqua,
Neutrophils escort circulating tumour cells to enable cell cycle progression, Nature et al., Comprehensive genomic analysis identifies novel subtypes and targets of
566 (2019) 553–557. triple-negative breast cancer, Clin. Cancer Res. 21 (2015) 1688–1698.
[17] B.A. Pulaski, S. Ostrand-Rosenberg, Mouse 4T1 breast tumor model, Curr. Protoc. [28] B.D. Lehmann, J.A. Bauer, X. Chen, M.E. Sanders, A.B. Chakravarthy, Y. Shyr, et
Immunol. 39 (20) (2000), pp. 2.1-.2.16. al., Identification of human triple-negative breast cancer subtypes and preclinical
[18] A.J. Gareau, C. Brien, S. Gebremeskel, R.S. Liwski, B. Johnston, M. Bezuhly, models for selection of targeted therapies, J. Clin. Invest. 121 (2011) 2750–2767.
Ticagrelor inhibits platelet-tumor cell interactions and metastasis in human and [29] L.M. Tseng, J.H. Chiu, C.Y. Liu, Y.F. Tsai, Y.L. Wang, C.W. Yang, et al.,
murine breast cancer, Clin. Exp. Metastasis 35 (2018) 25–35. A comparison of the molecular subtypes of triple-negative breast cancer among
[19] Y.F. Tsai, C.C. Huang, Y.S. Lin, C.Y. Hsu, C.P. Huang, C.Y. Liu, et al., Interleukin non-asian and taiwanese women, Breast Cancer Res. Treat. 163 (2017) 241–254.
17A promotes cell migration, enhances anoikis resistance, and creates a [30] B.D. Lehmann, B. Jovanovic, X. Chen, M.V. Estrada, K.N. Johnson, Y. Shyr, et al.,
microenvironment suitable for triple negative breast cancer tumor metastasis, Refinement of triple-negative breast cancer molecular subtypes: implications for
Cancer Immunol. Immunother. 70 (2021) 2339–2351. neoadjuvant chemotherapy selection, PLoS One 11 (2016), e0157368.
[20] M.M. Bradford, A rapid and sensitive method for the quantitation of microgram [31] Y.S. Lin, Y.C. Chiu, Y.H. Tsai, Y.F. Tsai, J.Y. Wang, L.M. Tseng, et al., Different
quantities of protein utilizing the principle of protein-dye binding, Anal. Biochem. mechanisms involved in the berberine-induced antiproliferation effects in triple-
72 (1976) 248–254. negative breast cancer cell lines, J. Cell. Biochem. 120 (2019) 13531–13544.