Drug-releasing mesenchymal cells strongly suppress B16 lung metastasis in a syngeneic murine model

Cells

The murine melanoma cell line B16 21] and MOLT-4 (human acute T-lymphoblastic leukemia) 22] were provided by Centro Substrati Cellulari, ISZLER (Brescia, Italy). B16 cells were
maintained in RPMI 1640 medium (Euroclone, UK) supplemented with 10 % fetal bovine
serum (FBS) (LONZA Walkersville MD USA). MOLT-4 cells were cultured in complete IMDM
+10 % FBS. The Mesenchymal stromal cell line SR4987 was established from a long-term
BM-derived cell culture of BDF/1 mice (ATCC CRL-2028) 23]. SR4987 cells were cultured in IMDM supplemented with 5 % FBS. SR4987 are positive
for stem cell marker Vimentin, CD44+, CD73+, CD105+, CD106+, Sca-1+, CD34+, and CD45+
and have the capacity to differentiate into osteoblasts and chondrocytes 23]. SR4987 cells were also transduced with green fluorescent protein (GFP) using lentiviral
vector ( pCCLsin.PPThPGK.GFPpre) as previously described 15].

Human lung-derived microvascular endothelial cells (L-MECs) were obtained from Dr.
Arnaldo Caruso (Laboratory of Microbiology, University of Brescia, Brescia, Italy).
L-MECs were routinely maintained in endothelial basal growth medium (EBM-GM) (EGM
bullet kit, LONZA Walkersville MD USA) on Collagen?+?Fibronectin coated T25 flasks.

Mouse lung stromal cells (mL-StCs) were freshly isolated from C57Bl6 mice by using
0.25 % collagenase D (Boehringer Mannheim Germany). The adherent mL-StCs were cultured
in IMDM +5 % FBS and analyzed for mesenchymal markers CD44, CD29 and CD90 (all purchased
from Abcam Cambridge UK, mouse MSCs panel) to confirm their mesenchymal origin.

PTX priming of SR4987 cells

PTX was purchased from Adipogen (Vinci-Biochem, Italy). The activity of PTX on SR4987
cells was determined in a 24 h (cytotoxicity test) and in a 7 day MTT assay (anti-proliferative
test) as previously described 24]. Priming of SR4987 with PTX (SR4987PTX) was carried out using PTX at 2,000 ng/ml
as previously described 15].

In vitro antitumor assay

The effect of conditioned medium (CM) from SR4987 and SR4987PTX cells on B16 proliferation
was evaluated by MTT assay 24]. The inhibitory concentrations (IC
₅₀
and IC
₉₀
) were determined according to the Reed and Muench formula 25]. The antitumor activity of CM from SR4987PTX cells was compared to that of pure PTX
and expressed as PTX equivalent concentration (PEC) 15]. The PTX released by a single cells was calculated dividing PEC for the number of
cells seeded and expressed as pg/cell. To establish the optimal dose of SR4987PTX
to inject, co-cultures with B16 cells were made by using transwell inserts (0.4 ?m
pore size) (Becton Dickinson, USA) 15], 16].

Rosette adherence assay and Transmission electron microscopy (TEM) analysis

To study the interaction between B16 cells and SR4987, a rosette test was performed
16]. B16 were mixed in a conical tube with SR4987 or SR4987PTX cells, in 500 ?l of IMDM?+?5 %
FBS (ratio B16/SR4987 5:1). After 24 h of incubation at 37 °C in air?+?5 % CO
₂
, 20 ?l of cells were collected by a micropipette and then transferred on a slide
to evaluate rosette formation under an inverted microscope. Rosette were then analyzed
by TEM as previously described 16].

Adhesion assay

Adhesion of SR4987 and SR4987PTX to L-MECs and to mL-StCs were performed following
a previously described procedure 16]. L-MECs and mL-StCs monolayers were stimulated by adding B16 derived CM (B16-CM),
tumor necrosis factor-alpha (TNF?; 10 ng/ml; Sigma Chemical) and their combination for 24 h. L-MECs and mL-StCs
were allowed to interact for 30’ with SR4987 and SR4987PTX (10
⁴
cells/well) in IMDM +0.2 % BSA. Each test was run in quadruplicate (see Additional
file 1).

Migration assay

Transwell supports were used to test spontaneous migration and chemotaxis of SR4987
and SR4987PTX. Migration assay was performed as previously described 26]. Test samples consisting in: mouse SDF-1 (1-100 ng/ml) (RD system), TNF? (1-50 ng/ml),
B16-CM (1:1-1:100 dilutions) mL-StCs-CM (1:1-1:100 dilutions), CM derived from mL-StCs
primed for 24 h either with B16-CM (1:1) (mL-StCs-B16-CM), TNF? 10 ng/ml (mL-StCs-TNF?-CM)
or B16-CM?+?TNF? (mL-StCs-B16-TNF?-CM), were placed in the lower compartment. The
blocking effect of monoclonal antibodies (mAbs) anti mouse SDF-1 (R D System), as
well control mouse IgG, (0.01-1ug/ml) was tested by adding mAbs to medium containg
SDF-1 or to CMs. To block SDF-1 receptors CXCR4/CXCR7, SR4987 and SR4987PTX (5×10
⁵
) were pre-incubated for 1 h at 37 °C in IMDM?+?0.2 % BSA in presence of different
concentration of AMD3100 (0.1-10 ?M) (purchased from Sigma). Each determination was
done in duplicate (see Additional file 1).

Immunoassay for mouse SDF-1

CM samples recovered from 48 h cultured SR4987, SR4987PTX, B16 cells, or from mL-StCs
stimulated or not with B16-CM, TNF? and B16-CM?+?TNF? were analyzed for the presence
of SDF-1 using an Elisa kit (RD System). The value of SDF-1 detected in the CM were
normalized for the same number of cells counted at the end of incubation.

Flow Cytometry and PCR

The expression of Sca-1, CXCR4 and CXCR7 on SR4987 was determind by flow cytometry
(FC) using the following antibodies: anti-Sca-1 FITC, anti-CXCR4 (ab1670) and anti-CXCR7
(all purchased from Abcam Cambridge UK). 20,000 events were acquired for each analysis
using Epics “XL-MLC” (Beckman Coulter, USA) and histogram elaboration was performed
with EXPO 32 software.

For the analysis of mRNA levels 1000 ng of total RNA isolated using the RNeasy kit
(Qiagen) was reverse-transcribed using iScript cDNA Synthesis Kit (Bio-Rad Laboratories).
Triplicate polymerase chain reactions were carried out on an CFX 96 Touch Real Time
PCR Detection System (Bio-Rad Laboratories). Relative gene expression was calculated
by a comparative method (2
^-??Ct
) using GAPDH as an housekeeping gene. Primers sequences were designed using Primer3
software.

In vivo experiments and Immunohistochemistry

Six-to eight-week old male C57BL/6 mice were purchased from Charles River (Calco,
Milan, Italy). All the animal experiments were performed at the Animal Facility of
Regina Elena National Cancer Institute in Rome Italy following directive 2010/63/EU.
Healthy mice (2 each group) receiving i.v. injection of saline, 10
⁶
SR4987GFP and 10
⁶
SR4987GFP-PTX. Fluorescent cells were used to confirm their capacity to arrest into
the lung. Mice were sacrificed at 6 h, 24 h and 48 h after treatments and lungs were
minced and digested by collagenase (for 2 h at 37 °C) to obtain a cell suspension
that was rapidly analyzed under Fluorescent microscopy. To study the anti-metastatic
activity of SR4987PTX, mice were injected i.v. with 2.5 × 10
⁵
B16 melanoma cells /0.2 ml. On day 5, after B16 injection, mice (12 each group) were
treated i.v. with 0.2 ml saline (controls), PTX 10 mg/Kg, SR4987 (10
⁵
/0.2 ml) and SR4987PTX at 10
⁵
and 5×10
⁴
/0.2 ml. Treatments were repeated on day 10 and 15. On day 21 after B16 injection,
mice were euthanized, lungs removed and fixed in Bouin’s solution. Lung nodules were
counted under a dissecting stereomicroscope. Metastasis were also investigated by
histology using Hematoxylin and Eosin (HE) and by Fontana-Masson staining to detect
micrometastasis 26]. To investigate the capacity of SR4987 and SR4987PTX to home lung B16 nodules, mice
(3 each group) were injected i.v. with 10
⁶
B16 cells. Upon 14 days mice were i.v. treated with saline (controls), PTX (10 mg/kg),
SR4987 (10
⁵
) and SR4987PTX (10
⁵
) and sacrificed 48 h after treatments. The lungs were processed by immunohistochemistry
for Sca-1?+?cells detection by using rabbit monoclonal anti-Sca1/Ly6A/E antibody (1:150,
Abcam) and polyclonal anti-SDF-1 (Santa Cruz Biotechnology Inc USA) for SDF-1 expression
according to standard protocols 26], 27]. The number of Sca-1
⁺
cells was evaluated at high power field (HPF), 400X magnification. At least 10 fields
for each case were randomly evaluated with variability less than 5 %.

Statistical analysis

Data are expressed as the mean?±?standard deviation (SD). Differences between mean
values were evaluated according to Student’s t-test performed by GRAPHPADINSTAT program
(GraphPad Software Inc., San Diego, CA, USA). p values??0.05 were considered statistically
significant.