IFN e VEGF-C

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

ORIGINAL ARTICLE

Decidual NK Cell-Derived Conditioned Medium (dNK-CM)


Mediates VEGF-C Secretion in Extravillous Cytotrophoblasts
Genevieve D. M. Eastabrook1,2, Yuxiang Hu1,2, Rusung Tan2,3, Jan P. Dutz2,4, Colin D. MacCalman1,2 ,
Peter von Dadelszen1,2
1
Department of Obstetrics and Gynaecology, University of British Columbia, Vancouver, BC, Canada;
2
The Child and Family Research Institute, University of British Columbia, Vancouver, BC, Canada;
3
Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, BC, Canada;
4
Department of Dermatology and Skin Science, University of British Columbia, Vancouver, BC, Canada

Keywords, Problem
Decidual natural killer cells, extravillous The regulatory mechanisms involved in VEGF-C secretion by tropho-
cytotrophoblasts, Janus kinase, p38, signal
blasts during placentation are poorly understood. We investigated
transducers and activators of transcription,
VEGF-C
whether or not decidual natural killer cell conditioned medium (dNK-
CM) stimulated VEGF-C secretion in the extravillous cytotrophoblast
Correspondence (EVT) cell line HTR8 ⁄ SVneo.
Genevieve Eastabrook, Department of
Obstetrics and Gynaecology, University of Method of Study
British Columbia, 2H30-4500 Oak Street, The effects of dNK-CM and recombinant IFN-c on VEGF-C induction by
Vancouver, BC V6H 3N1, Canada. HTR8 ⁄ SVneo were studied in the absence or presence of IFN-c or its
E-mail: geastabrook@cw.bc.ca receptor blocking antibodies, p38 inhibitor (SB202190), JAK inhibitor
 Deceased.
(JAK inhibitor-1, JI-1), and on STAT1 knockdown HTR8 ⁄ SVneo. VEGF-C
Submitted August 8, 2011;
was quantified by ELISA. FACS was used to investigate the phosphoryla-
accepted September 13, 2011. tions of Tyr701 or Ser727 of STAT1 on stimulated HTR8 ⁄ SVneo.

Citation Results
Eastabrook GDM, Hu Y, Tan R, Dutz JP, dNK-CM facilitated VEGF-C secretion by HTR8 ⁄ SVneo. IFN-c and IFN-
MacCalman CD, von Dadelszen P. Decidual NK cR1 or IFN-cR2 blocking antibodies reduced both dNK-CM- and IFN-c-
cell-derived conditioned medium (dNK-CM) induced VEGF-C secretion. Phosphorylations on Tyr701 or Ser727 of
mediates VEGF-C secretion in extravillous STAT1 were elevated upon stimulation. Secretion of VEGF-C was
cytotrophoblasts. Am J Reprod Immunol 2012;
reduced by treatment with SB202190, JI-1, or STAT1 knockdown by
67: 101–111
siRNA.
doi:10.1111/j.1600-0897.2011.01075.x
Conclusion
VEGF-C production by trophoblasts is regulated by soluble factors
secreted by dNK through p38 and JAK-STAT1 pathways.

and spiral artery remodeling. CTB expression of


Introduction
VEGF-A, VEGF-C, placental growth factor (PlGF),
During the first and second trimesters of pregnancy, and VEGFR-1 is accompanied by downregulation of
the invasive cytotrophoblasts (CTB) display a VEGFR-2 expression.1,2 These cells are able to
unique repertoire of substances that influence vascu- respond to the signals from endothelial cells and par-
logenesis and angiogenesis. Evidence indicates that ticipate in endothelial cell-guided tubule formation.3
VEGF-C, along with its receptors, plays a critical role At term, CTB express VEGF-A and PIGF but express
in trophoblast invasion, endovascular differentiation, little VEGF-C and VEGFR-2.1,2 They fail to respond
American Journal of Reproductive Immunology 67 (2012) 101–111
ª 2011 John Wiley & Sons A/S 101
EASTABROOK ET AL.

to signals from endothelial cells and do not partici- Phospho-Tyr701 STAT1 forms homodimers which
pate in endovascular remodeling, displaying translocate to the nucleus and bind to r-activating
restricted migratory and interactive properties.3 The sequences (GAS) in IFN-c-responsive promoters.20,21
regulation of VEGF-C expression in CTB is not fully The second phosphorylation site, Ser727 on STAT1,
understood, but merits further investigation both in is also activated by IFN-c.20,21 A variety of serine
normal and pathological placentation. kinases have been reported to phosphorylate Ser727,
Expression of both VEGF-C and VEGFR-3 is including phosphatidylinositol 3-kinase ⁄ AKT,19
decreased in placentas from pregnancies affected by Ca ⁄ calmodulin-dependent kinase II,22 protein
2+

intrauterine growth restriction (IUGR), suggesting kinase C isoform,23 and MAPK.24 In addition, p38 is
that abnormal levels of VEGF-C and VEGFR-3 may involved in LPS- and UV light-induced STAT1
contribute to the abnormal villous development.4 An Ser727 phosphorylation in macrophages25,26 and
imbalance between pro- and anti-angiogenic factors HeLa cells treated with IFN-c.27 STAT1 requires
is involved in the pathogenesis of pre-eclampsia. phosphorylation on Tyr701 and Ser727 to achieve
Soluble fms-like tyrosine kinase-1 (sFLt-1), a splice full transcriptional and anti-pathogenic activity.28
variant of the VEGFR-1 receptor lacking transmem- Our objective was to determine whether dNK-
brane and cytoplasmic domains, acts as a potent derived soluble factors were able to directly influ-
anti-angiogenic factor by binding to both VEGF and ence VEGF-C secretion by EVT, both in primary cell
PlGF, thus preventing their interaction with their cultures and in the EVT cell line HTR8 ⁄ SVneo. In
endogenous receptors.5 Recent evidence also sug- addition, we aimed to determine which factor(s)
gests that intracellular VEGF is decreased in the contained in dNK-CM contributed to this effect.
peripheral NK cells of women with pre-eclampsia as Here, we report that dNK-CM was able to induce
compared with normal pregnant controls.6 VEGF-C production in both primary CTB and the
During the first trimester, extravillous cytotropho- EVT cell line HTR8 ⁄ SVneo. IFN-c contained in the
blast (EVT) are in direct contact with dNK, a popula- CM was a major player in this process, as both IFN-c
tion of maternal lymphocytes with low cytotoxicity blocking antibody and IFN-cR1 and IFN-cR2 block-
as compared with peripheral blood NK cells. How- ing antibodies significantly reduced this effect. dNK-
ever, dNK are capable of secreting numerous cyto- CM and IFN-c phosphorylated both Tyr701 and
kines, chemokines, and growth factors, an Ser727 of STAT1. p38 MAPK inhibitor, JAK inhibi-
observation previously reported7–10 and also recently tor, and STAT1 siRNA decreased dNK-CM- or IFN-c-
described by our own group.11 These factors influ- mediated VEGF-C secretion.
ence EVT by regulating their invasion and vascular
growth.12–15 IFN-c is one of the critical cytokines
Materials and methods
produced by dNK. In mice, IFN-c regulates the
expression of >0.5% of the mouse genome, includ-
Preparation of dNK-CM
ing genes involved in variety of cell functions16,17
Furthermore, 90% of the IFN-c found in early to Following informed consent, placental tissues were
mid-gestation decidua basalis is derived from dNK collected from healthy women undergoing elective
cells.18 Specific examples of IFN-c-regulated genes terminations of pregnancy at gestational ages
include inducible nitric oxide synthase (iNOS), endo- between 6 and 10 weeks. Ethics approval was
thelial NOS (eNOS), and endothelin-1.16 IFN-c has obtained from the Children’s and Women’s Research
been shown to both enhance and reduce VEGF-C Ethics Board. dNK isolation and dNK-CM prepara-
secretion depending on the cell lines studied.18,19 tion were performed as described previously.12,13 In
However, there is a dearth of evidence pertaining to brief, specimens were thoroughly washed with PBS,
the role of decidual natural killer cells conditioned minced into fragments of approximately 1 mm3, and
medium (dNK-CM) or IFN-c in the regulation of digested for 1 hr at 37C in RPMI 1640 medium with
VEGF-C secretion by CTB. 1.5 mg ⁄ mL collagenase and 2 mg ⁄ mL hyaluronidase
IFN-c activates Janus kinases (JAKs) through (Sigma-Aldrich, Oakville, ON, Canada). Cell suspen-
binding to the cell surface receptors IFN-c receptor-1 sions were washed and plated in dishes with
(IFN-cR1) or IFN-c receptor-2 (IFN-cR2). This bind- 2 ng ⁄ mL IL-15 (R&D Systems, Minneapolis, MN,
ing causes the rapid phosphorylation of Tyr701 of USA), and after an overnight incubation at 37C,
the latent cytoplasmic transcription factor STAT1. non-adherent cells were harvested from the culture
American Journal of Reproductive Immunology 67 (2012) 101–111
102 ª 2011 John Wiley & Sons A/S
dNK-CM MEDIATES VEGF-C SECRETION IN EVT

dishes. The cell suspension obtained was then sepa- were then exposed to dNK-CM at 1:2–1:8 dilutions
rated by Ficoll-Hypaque gradient 1077 (Amersham for a total of 48 hr. At the end of incubation, super-
Biosciences, Baie d‘Urfe, QB, Canada). Subsequently, natants were collected and saved at )80C for
NK cells were enriched with human NK cell enrich- VEGF-C measurement. The concentration of VEGF-C
ment mixture following the manufacturer’s instruc- was measured with Quantikine ELISA kit (R&D
tions (StemCell Technologies, Vancouver, BC, Systems) following the manufacturer’s instructions.
Canada). On average, 6 · 105 enriched dNK per The results were expressed as normalized % using
gram of decidual tissue were isolated. Previously, we the following formula: normalized % = OD [assay
have confirmed the purity of this cell population group]*100 ⁄ OD [ctrl group], where ctrl group =
using flow cytometry.12 100%.
Decidual natural killer cells were cultured and To determine whether or not IFN-c contained in
maintained in vitro with DMEM ⁄ F12 supplemented the supernatants contributed to VEGF-C induction,
with 10% FBS and 10 ng ⁄ mL IL-15 for approxi- anti-IFN-c blocking antibody (mouse IgG2a; R&D
mately 1 month. The conditioned medium (CM) Systems) was pre-incubated with dNK-CM at a con-
from different passages of dNK and from different centration of 10 lg ⁄ mL for 1 hr. Antibody-treated or
subjects were collected, pooled, and stored at )80C control dNK-CM were then used to stimulate
in advance of the assays. The concentrations of HTR8 ⁄ SVneo cells. IFN-cR1 and IFN-cR2 blocking
VEGF-C and IFN-c contained in the dNK-CM were antibodies (Goat IgG; R&D Systems) were incubated
quantified. with HTR8 ⁄ SVneo at a concentration of 10 lg ⁄ mL
for 1 hr. dNK-CM was then used to stimulate anti-
IFN-cR1- or anti-IFN-cR2-treated or untreated
Culture of Primary CTB and HTR8 ⁄ SVneo cells
HTR8 ⁄ SVneo cells for a total of 48 hr. The experi-
The isolation and culture of primary CTB were ments were completed as above.
described previously.13 In brief, CTB-containing cell Recombinant IFN-c (R&D Systems), at a range of
suspension was obtained from villi by 0.1% trypsin 0.8–100 ng ⁄ mL, was tested on both CTB and
digestion, purified by 40% percoll density centrifuga- HTR8 ⁄ SVneo cells for VEGF-C induction. IFN-c
tion and further enriched by removing non-adherent blocking antibody or IFN-cR1 and IFN-cR2 antibod-
cells after overnight attachment. The CTB obtained ies were tested on HTR8 ⁄ SVneo cells as well. Other
contained approximately 80–90% cytokeratin posi- than for the initial titration, IFN-c at 20 ng ⁄ mL con-
tive cells. No further purification was carried out for centration was used throughout the study.
the initial assays, and the CTB were grown for To determine the involvement of p38 and JAK-
4–5 days before the assay was set up. The primary STAT1 pathways in the induction of VEGF-C by
CTB were only used for initial assays, to determine dNK-CM and IFN-c, HTR8 ⁄ SVneo cells were treated
whether they were capable of VEGF-C secretion with 3.1–12.5 lm of SB202190 (Biomol Interna-
after stimulation with either IFN-c or dNK-CM. tional, Plymouth Meeting, PA, USA) or 1.2–5 lm of
HTR8 ⁄ SVneo cells were a gift provided by Dr. JI-1 (EMD, Mississauga, ON, Canada) for 1 hr before
Charles H. Graham (Queen’s University, Kinston, exposure to dNK-CM or IFN-c. The experiments
ON, Canada).29 Cells were routinely maintained in were completed as above.
RPMI-1640 medium containing 5% FBS and were
passaged every 2–3 days. HTR8 ⁄ SVneo cells were
FACS Staining
used throughout the study.
Phosphorylation of Tyr701 and Ser727 of STAT1
were performed by flow cytometry as described with
Generation of VEGF-C-Containing Supernatants
minor modifications.30 HTR8 ⁄ SVneo cells maintained
from Trophoblasts and VEGF-C Measurement by
with 5% FBS containing PRMI1640 were harvested
ELISA
and divided into tubes, with 1 · 106 cells ⁄ mL per
HTR8 ⁄ SVneo and primary CTB were seeded into 24- tube. Cells were treated with either dNK-CM or IFN-c
well plates at 100,000 cells ⁄ well in DMEM ⁄ F12 sup- for different periods of time, or cells were pre-treated
plemented with 10% FBS. On the following day, the with anti-IFN-cR1 or IFN-cR2 antibodies at a con-
media were removed and replaced with fresh centration of 10 lg ⁄ mL for 1 hr followed by dNK-
DMEM ⁄ F12 supplemented with 10% FBS. Cells CM or IFN-c treatment for 1 hr. At the end of the
American Journal of Reproductive Immunology 67 (2012) 101–111
ª 2011 John Wiley & Sons A/S 103
EASTABROOK ET AL.

treatment, treated and untreated cells were fixed by 100% transfection efficiency was chosen to transfect
adding 37% formaldehyde solution directly into the target siRNA. ON-TARGETplus STAT1 siRNA and
cell suspension to obtain a final concentration of non-targeting control siRNA were purchased from
1.5% formaldehyde. Cells were incubated for Dharmacon (Lafayette, CO, USA). STAT1 siRNA (but
20 min and pellets were then collected by centrifu- not control siRNA) at a concentration as low as
gation. Cells were permeabilized by 100% ice-cold 12.5 nm showed >95% STAT1 mRNA reduction as
methanol for 10 min. Following permeabilization, detected by real-time PCR (data not shown). After
cells were either stained immediately or saved at assay optimization, HTR8 ⁄ SVneo were transfected
)20C for future staining. with 25 or 50 nM STAT1 siRNA or control siRNA.
Cells were washed with PBS containing 1% BSA Twenty-four hours after transfection, the transfected
(staining solution) before staining procedures. Cells cells were exposed to dNK-CM or IFN-c for an addi-
were incubated with 1:100 diluted rabbit anti- tional 48 hr. The culture supernatants were then
human Tyr701 or Ser727 of STAT1 (Cell Signalling collected and VEGF-C quantification was performed
Technology (CST), Pickering, ON, Canada) for 2 hr as described above.
at room temperature, then washed with staining
solution three times. FITC-conjugated goat anti-rab-
Statistics
bit IgG (BD PharMingen, Mississauga, ON, Canada)
at 1:100 dilution was added to cells and incubated Because our data was not normally distributed,
for 30 min. Cells were then washed and subjected to groups were compared using non-parametric Krus-
FACS analysis in a FACScan flow cytometer (BD kal–Wallis anova, with Dunn’s post-test, as appropri-
Biosciences, Mississauga, ON, Canada). ate. graphpad prism 4.0 software (GraphPad Software
Inc., San Diego, CA, USA) was used for all analyses.
Statistical significance was set at P < 0.05.
Transfection of STAT1 siRNA into HTR8 ⁄ SVneo
cells
Results
X-tremeGENE siRNA transfection reagent (Roche,
Mississauga, ON, Canada) was used for transfection.
dNK-CM Increases CTB and HTR8 ⁄ SVneo
The cytotoxicity of the reagent was detected by its
Secretion of VEGF-C
serial dilutions, and transfection efficiency was deter-
mined by fluorescein-conjugated control siRNA Cytotrophoblasts and HTR8 ⁄ SVneo grown in
(CST) monitored under fluorescent microscope. The DMEM ⁄ F12 with 10% FBS produced basal levels of
condition which resulted in no cytotoxicity but with VEGF-C. dNK-CM significantly increased VEGF-C

(a) (b)

Fig. 1 dNK-CM facilitated VEGF-C induction in both cytotrophoblasts (CTB) (a) and HTR8 ⁄ SVneo (b). CTB and HTR8 ⁄ SVneo cells were stimulated
with serial dilutions of dNK-CM. The culture supernatants were collected and VEGF-C concentrations were measured. The results were expressed
as normalized % with following formula: normalized % = OD [assay group]*100 ⁄ OD [control medium group], where control medium group = 100%.
The dNK-CM significantly increased VEGF-C secretions from both CTB and HTR8 ⁄ SVneo. The data shown (median) were from one representative
experiment. Dotted line and P-value: Kruskal–Wallis non-parametric ANOVA. Dunn’s multiple post hoc test: *P < 0.05 compared with control (a, b).

American Journal of Reproductive Immunology 67 (2012) 101–111


104 ª 2011 John Wiley & Sons A/S
dNK-CM MEDIATES VEGF-C SECRETION IN EVT

secretion by both CTB (Fig. 1a) and the EVT cell line
HTR8 ⁄ SVneo, seeded on bare plates (Fig. 1b) in a
dose-dependent manner. Previously, it has been
shown that HTR8 ⁄ SVneo cells produce VEGF-C
when seeded on Matrigel.13 dNK-CM at 1:2 dilution
was used for the remainder of the study, as dNK-CM
at this dilution induced maximal VEGF-C secretion
among the tested concentrations. HTR8 ⁄ SVneo cells
were chosen for completion of the entire study.
dNK-CM harvested from short-term culture of dNK
(48 hr after dNK collection from decidua) was simi-
larly capable of VEGF-C induction. However, long-
term dNK-CM from long-term culture was used
throughout the study.
Previously, it has been reported that dNK express
VEGF-C.31 Our data confirmed VEGF-C mRNA
expression as detected by real-time PCR and VEGF-C
secretion by dNK (data not shown). While the VEGF-
C concentration was <5 pg ⁄ mL in dNK-CM, it was
>300 pg ⁄ mL in the supernatants from untreated
HTR8 ⁄ SVneo. Thus, the impact of VEGF-C existing in
dNK-CM as compared with the quantity produced by Fig. 2 dNK-CM-mediated VEGF-C induction on HTR8 ⁄ SVneo was
blocked by both IFN-c and IFN-cR1 or IFN-cR2 neutralizing antibodies.
HTR8 ⁄ SVneo was negligible. VEGF-C concentration
dNK-CM was pre-treated with IFN-c neutralizing antibody and then
was even greater after dNK-CM stimulation. Previous
used for stimulation on HTR8 ⁄ SVneo. Alternately, cells were pre-trea-
evidence has shown that IFN-c increased18 or ted with IFN-cR1 or IFN-cR2 blocking antibodies and were then
decreased VEGF-C secretion19 in different cell lines. exposed to dNK-CM. Data demonstrated that both IFN-c neutralizing
Our next step was to determine whether or not IFN-c antibody and IFN-cR1 or IFN-cR2 antibodies partially, but significantly,
contained in dNK-CM contributed to VEGF-C induc- reduced dNK-CM-mediated VEGF-C secretion. The data shown (median)
tion by HTR8 ⁄ SVneo exposed to dNK-CM. were from one representative experiment. Dotted line and P-value:
Kruskal–Wallis non-parametric ANOVA. Dunn’s multiple post hoc test:
*P < 0.05 compared with control; **P < 0.01 compared with control.
IFN-c Neutralizing Antibody and IFN-cR1 or
IFN-cR2 Neutralizing Antibodies Reduce
dNK-CM-Mediated VEGF-C Secretion from the antibodies tested above (Fig. 3c) but not by
HTR8 ⁄ SVneo cells mouse IgG2a or goat IgG isotype controls.
dNK-CM was pre-treated with anti-IFN-c blocking
antibody for 1 hr before addition to HTR8 ⁄ SVneo IFN-c Can Activate STAT1 Tyr701 and Ser727
cells. Alternately, HTR8 ⁄ SVneo cells were pre-treated
with anti-IFN-cR1 or IFN-cR2 blocking antibodies IFN-c-induced STAT1 Tyr701 activation acts through
and then exposed to dNK-CM. Our results demon- JAK.20,21 STAT1 Ser727 activation may or may not
strated that both IFN-c blocking antibody and be dependent on p38 MAPK.27,32,33 Other serine
IFN-cR1 or IFN-cR2 blocking antibodies were able to kinases are also proposed to be involved.22–24,34 We
partially, but significantly, reduce VEGF-C secretion; sought to determine whether dNK-CM or IFN-c was
mouse IgG2a or goat IgG isotype controls had no able to increase phosphorylations of Tyr701 and
effect (Fig. 2). As measured by conventional sand- Ser727 of STAT1 on HTR8 ⁄ SVneo.
wich ELISA, the dNK-CM in our system contained
approximately 3.5 ng ⁄ mL IFN-c. Recombinant IFN-c dNK-CM and IFN-c Mediate Phosphorylations of
in the range of 0.8–100 ng ⁄ mL was able to increase Tyr701 and Ser727 of STAT1 on HTR8 ⁄ SVneo
VEGF-C secretion in both CTB (Fig. 3a) and
HTR8 ⁄ SVneo (Fig. 3b); the VEGF-C-inducing ability Phosphorylations of Tyr701 and Ser727 of STAT1
of IFN-c on HTR8 ⁄ SVneo cells was ameliorated by on HTR8 ⁄ SVneo occurred very quickly following

American Journal of Reproductive Immunology 67 (2012) 101–111


ª 2011 John Wiley & Sons A/S 105
EASTABROOK ET AL.

(a) (b)

(c)

Fig. 3 Recombinant IFN-c enhanced VEGF-C secretions from both cytotrophoblasts (a) and HTR8 ⁄ SVneo (b). IFN-c-mediated VEGF-C induction on
HTR8 ⁄ SVneo was blocked by both IFN-c and IFN-cR1 or IFN-cR2 neutralizing antibodies (c). Cells were stimulated with various concentrations of
IFN-c and the supernatants were collected for VEGF-C measurement. For antibody blocking assays, the methods were the same as using dNK-CM
as a stimulator as described above. The data (median) showed that VEGF-C induction was increased by IFN-c (a, b) and was blocked by antibodies
tested (c). Dotted line and P-value: Kruskal–Wallis non-parametric ANOVA. Dunn’s multiple post hoc test: *P < 0.05 compared with control (a, b), or
IFN-c (c); **P < 0.01 compared with IFN-c (c).

exposure to either dNK-CM or IFN-c. After 10 min but was seen to a lesser extent on dNK-CM-induced
treatment with dNK-CM or IFN-c, both Tyr701 and Ser727 phosphorylation (Fig. 5b).
Ser727 phosphorylations were increased signifi- To confirm the involvement of p38 and JAK-
cantly. The signals remained at high level at the STAT1 pathways in VEGF-C induction, SB202190,
observation times 1hr (Fig. 4a–d) and 6 hr (data JI-1, and STAT1 siRNA were introduced to the
not shown). The high basal levels of phosphoryla- study.
tion of Tyr701 and Ser727 on HTR8 ⁄ SVneo were
observed in the medium control containing 10%
SB202190, JI-1, and STAT1 siRNA Reduce dNK-
FBS.
CM- or IFN-c-Mediated VEGF-C Secretions on
Decidual natural killer cells conditioned medium
HTR8 ⁄ SVneo
or IFN-c-induced phosphorylations of Tyr701 and
Ser727 on HTR8 ⁄ SVneo were reduced by IFN-cR1 HTR8 ⁄ SVneo were pre-treated with SB202190 at
and IFN-cR2 blocking antibodies (Fig. 5a–d). This concentrations ranging from 3.1 to 12.5 lm, and JI-1
effect was especially significant for IFN-c-induced at concentrations ranging from 1.2 to 5 lm for 1 h.
Tyr701 and Ser727 phosphorylations (Fig. 5c,d), and The untreated and treated HTR8 ⁄ SVneo cells were
dNK-CM-induced Tyr701 phosphorylation (Fig. 5a), then exposed to dNK-CM or IFN-c. The results
American Journal of Reproductive Immunology 67 (2012) 101–111
106 ª 2011 John Wiley & Sons A/S
dNK-CM MEDIATES VEGF-C SECRETION IN EVT

(a) (b)

(c) (d)

Fig. 4 Both dNK-CM and IFN-c activated STAT1 Tyr701 and Ser727 phosphorylations (a–d). dNK-CM- or IFN-c-treated HTR8 ⁄ SVneo were fixed, per-
meabilized, and then conducted for FACS staining. The cells were incubated with anti-human Tyr701 or Ser727 of STAT1 followed by incubating
with FITC-conjugated secondary Ab. The results showed the high basal level phosphorylations of Tyr701 and Ser727 on control cells; both dNK-CM
(a, b) and IFN-c (c, d) were able to enhance Tyr701 and Ser727 phosphorylations.

showed that SB202190 and JI-1 decreased both IFN-c- but significantly, reduced in STAT1 siRNA transfect-
and dNK-CM-induced VEGF-C secretion in a dose- ed cells but not in control siRNA transfected cells
dependent manner (Fig. 6a,b). For SB202190, this (Fig. 7).
effect was statistically significant only at the highest
concentrations of the inhibitor when stimulated with
Discussion
either IFN-c or dNK-CM. Likewise, for JI-1, the
inhibitory effect was statistically significant for VEGF-C plays important roles in angiogenesis and
IFN-c-induced VEGF-C secretion at the two high- lymphangiogenesis. Endothelial cells are a major
est concentrations (5 and 2.5 lm, respectively). source of VEGF-C. Invasive trophoblasts1–3 and
Although a dose-dependent inhibition of dNK-CM- dNK31 also produce VEGF-C, findings which have
induced VEGF-C secretion was observed, this did not been previously reported and have been confirmed
reach statistical significance. in this study. However, we have found that VEGF-C
To study whether or not STAT1 knockdown would is found at very low, to undetectable, levels in our
affect VEGF-C secretion, HTR8 ⁄ SVneo were trans- dNK-CM.11
fected with STAT1 siRNA or control siRNA at various Trophoblasts may acquire their endothelial pheno-
concentrations. Twenty-four hours after transfection, type through their production of VEGF-C and other
cells were exposed to IFN-c or dNK-CM. The data angiogenic factors.2 Previously, we reported that
demonstrated that VEGF-C secretion was partially, dNK-CM facilitated the formation of capillary tube

American Journal of Reproductive Immunology 67 (2012) 101–111


ª 2011 John Wiley & Sons A/S 107
EASTABROOK ET AL.

(a) (b)

(c) (d)

Fig. 5 dNK-CM- and IFN-c-mediated Tyr701 and Ser727 phosphorylations on HTR8 ⁄ SVneo were blocked by IFN-cR1 or IFN-cR2 antibodies (a–d).
HTR8 ⁄ SVneo were pre-treated with anti-IFN-cR1 or IFN-cR2 antibodies, followed by either dNK-CM or IFN-c stimulation. At the end of the treat-
ment, the cells were fixed, permeabilized, and then stained for FACS analysis. IFN-cR1 and IFN-cR2 antibodies showed greatest inhibition of dNK-
CM-induced (a) and IFN-c-induced Tyr701 (c) phosphorylation, and to a lesser extent, dNK-CM-induced (b) and IFN-c-induced (d) phosphorylation.

and network structures on thick-layer Matrigel both recombinant IFN-c to promote VEGF-C secretion
in primary CTB and in HTR8 ⁄ SVneo, accompanied was completely blocked by IFN-c blocking antibody
by a higher level of VEGF-C secretion as compared or IFN-cR1 and IFN-cR2 blocking antibodies
to VEGF-A.13 In this study, we found that dNK-CM (Fig. 3c). Thus, the data indicate that in addition to
was able to facilitate VEGF-C secretion by both pri- IFN-c, other soluble factors contained in the dNK-
mary CTB and the EVT cell line, HTR8 ⁄ SVneo CM contribute to the stimulation of VEGF-C secre-
(Fig. 1a,b) in the absence of Matrigel. tion. For example, we observed that TNF-a in
IFN-c contained in the dNK-CM appeared to con- dNK-CM increased VEGF-C production, and its effect
tribute significantly to VEGF-C secretion by CTB and was reduced by TNF-a blocking antibody. Recombi-
HTR8 ⁄ SVneo, as both IFN-c neutralizing antibody nant TNF-a also had the ability to induce VEGF-C
and IFN-cR1 or IFN-cR2 antibodies were able to secretion by HTR8 ⁄ SVneo (data not shown).
partially, but significantly, block dNK-CM-induced Both Tyr701 and Ser727 phosphorylations of
VEGF-C secretion (Fig. 2). dNK-CM contained approxi- STAT1 are required for full STAT1 transcriptional
mately 3.5 ng ⁄ mL IFN-c. However, the VEGF-C stimu- and anti-pathogenic activity.20,21 Both Tyr701 and
latory ability of dNK-CM was greater than that of Ser727 of STAT1 were quickly phosphorylated upon
the equivalent concentration of recombinant IFN-c exposure to either dNK-CM or IFN-c in the study.
alone (Figs 1 and 3). Furthermore, the ability of After 10 min of exposure, elevated phosphorylations

American Journal of Reproductive Immunology 67 (2012) 101–111


108 ª 2011 John Wiley & Sons A/S
dNK-CM MEDIATES VEGF-C SECRETION IN EVT

(a) (b)

Fig. 6 Both dNK-CM- and IFN-c-mediated VEGF-C inductions were reduced by SB202190 (a) and JI-1 (b) treatment. HTR8 ⁄ SVneo were pre-treated
with SB202190 or JI-1, and then exposed to either dNK-CM or IFN-c for VEGF-C inductions. The results showed that both SB202190 and Janus
kinase inhibitor-1 decreased dNK-CM- or IFN-c-induced VEGF-C secretions. Dotted line and P-value: Kruskal–Wallis non-parametric ANOVA. Dunn’s
multiple post hoc test with SB202190 (a): *P < 0.05 compared with control; **P < 0.01 compared with ctrl; ***P < 0.05 compared with IFN-c;
#
P < 0.05 compared with CM. Dunn’s multiple post hoc test with JI-1 (b): *P < 0.05 compared with control; **P < 0.01 compared with control;
***P < 0.01 compared with IFN-c; #P < 0.05 compared with IFN-c.

Fig. 7 Both dNK-CM- and IFN-c-induced VEGF-C secretions were reduced by siRNA-mediated STAT1 knockdown on HTR8 ⁄ SVneo. Cells were pre-
transfected with STAT1 siRNA and non-targeting control siRNA and were then stimulated with either dNK-CM or IFN-c for VEGF-C inductions. The
results demonstrated that STAT1 siRNA, but not non-targeting control siRNA, partially but significantly blocked dNK-CM- and IFN-c-mediated VEGF-C
inductions. The data shown (median) were from one representative experiment. Dotted lines and P-values: Kruskal–Wallis non-parametric ANOVA.

were observed (Fig. 4). High levels of phosphoryla- and Ser727 phosphorylation were detected using
tion were maintained up to 6 hr after treatment FACS after both dNK-CM and IFN-c treatments
(data not shown). HTR8 ⁄ SVneo showed a high basal (Fig. 4a–d). High basal levels of Tyr701 and Ser727
level of Tyr701 or Ser727 phosphorylation under phosphorylation explain the basal level VEGF-C
control conditions (DMEM supplemented with 10% secretion from HTR8 ⁄ SVneo, indicating the existence
FBS). Clearly, detectable elevations of both Tyr701 of VEGF-C-inducing factors in FBS. Elevated Tyr701

American Journal of Reproductive Immunology 67 (2012) 101–111


ª 2011 John Wiley & Sons A/S 109
EASTABROOK ET AL.

and Ser727 phosphorylation induced by dNK-CM or


Acknowledgements
IFN-c were reduced by IFN-cR1 or IFN-cR2 blocking
antibodies (Fig. 5a–d), indicating specific ligand stim- We thank the CARE Program, BC Women’s Hospital
ulation by IFN-c. The reduction in both dNK-CM- and Health Centre, for their assistance in gaining
induced Tyr701 (Fig. 5a) and IFN-c-induced Tyr701 access to reproductive tissue, and Dr. Geoff Ham-
(Fig. 5c) phosphorylation was seen to a greater mond and Dr. Peter Leung for access to laboratory
extent than the reduction in dNK-CM-induced equipment. The study was funded by a pilot grant
(Fig. 5b), and particularly IFN-c-induced Ser727 from the Infection and Immunity Program, Child
phosphorylation (Fig. 5d). dNK-CM-induced Ser727 and Family Research Institute (CFRI). YH is funded
activation may be acting through alternative means through grants from the Canadian Institutes for
aside from IFN-c receptors. TNF-a, another dNK-pro- Health Research, from whom PvD receives salary
duced pro-inflammatory cytokine, has also been pro- support. PvD receives salary support from CFRI.
posed in mediating p38 activation.35,36 PvD, JPD, and RT receive salary support from the
It had been shown that Tyr701 was activated by Michael Smith Foundation for Health Research.
JAK, with the most likely candidate being JAK2.20,21
JI-1 resulted in significant inhibition of VEGF-C
References
secretion from HTR8 ⁄ SVneo (Fig. 6b), indicating the
involvement of JAK kinase in the process. STAT1 1 Zhou Y, McMaster M, Woo K, Janatpour M, Perry J, Karpanen T,
Alitalo K, Damsky C, Fisher SJ: Vascular endothelial growth factor
Ser727 activation may or may not rely on
ligands and receptors that regulate human cytotrophoblat survival
p38.27,32,33 The involvement of other kinases has are dysregulated in severe preeclampsia and hemolysis, elevated
been proposed.22,23,34 The p38 MAPK inhibitor liver enzymes, and low platelets syndrome. Am J Pathol 2002;
SB202190 was able to significantly inhibit VEGF-C 160:1405–1423.
secretion from HTR8 ⁄ SVneo (Fig. 6a), implying that 2 Zhou Y, Bellingard V, Feng KT, McMaster M, Fisher SJ: Human
p38 MAPK participated in dNK-CM- or IFN-c-medi- cytotrophoblast promote endothelial survival and vascular
remodeling through secretion of Ang2, PLGF, and VEGF-C. Dev Biol
ated VEGF-C responses. However, whether or not 2003; 263:114–125.
p38 mediates STAT1 Ser727 phosphorylation has yet 3 Kalkunte S, Lai Z, Tewari N, Chichester C, Romero R, Padbury J,
to be determined. Sharma S: In vitro and in vivo evidence for lack of endovascular
HTR8 ⁄ SVneo cells transfected with STAT1 siRNA remodeling by third trimester trophoblasts. Placenta 2008; 29:871–
878.
at concentration as low as 12.5 nm were able to
4 Dunk C, Ahmed A: Expression of VEGF-C and activation of its
knock down STAT1 mRNA by >95%. This effect was receptors VEGFR-2 and VEGFR-3 in trophoblast. Histol Histolpathol
not seen with control siRNA (data not shown). 2001; 16:359–375.
VEGF-C secretion from HTR8 ⁄ SVneo cells was also 5 Silasi M, Cohen B, Karumanchi SA, Rana S: Abnormal
reduced by STAT1 siRNA (Fig. 7). However, as much placentation, angiogenic factors, and the pathogenesis of
preeclampsia. Obstet Gynecol Clin North Am 2010; 37:239–253.
as 50 nm STAT1 siRNA was not able to block VEGF-
6 Molvarec A, Ito M, Shima T, Yoneda S, Toldi G, Stenczer B,
C secretion completely. Other STAT1-independent, Vásárhelyi B, Rigó Jr J, Saito S: Decreased proportion of peripheral
IFN-c-induced transcription factors such as STAT3 blood vascular endothelial growth factor-expressing T and natural
and STAT5,37,38 NF-kB39 and AP-140,41 are also likely killer cells in preeclampsia. Am J Obstet Gynecol 2010; 203:567.e1–8.
activated, as previously reported in other systems. 7 Lash GE, Schiessl B, Kirkley M, Innes BA, Cooper A, Searle RF,
In summary, we have shown that maternally Robson SC, Bulmer JN: Expression of angiogenic growth factors by
uterine natural killer cells during early pregnancy. J Leukoc Biol
derived dNK cells are capable of influencing fetal- 2006; 80:572–580.
derived trophoblast secretion of angiogenic factors, 8 Saito S, Nishikawa K, Morii T, Enomoto M, Narita N, Motoyoshi K,
including VEGF-C. As a result, dNK may influence Ichijo M: Cytokine production by CD16)CD56bright natural killer
trophoblast endovascular differentiation and uterine cells in the human early pregnancy decidua. Int Immunol 1993;
5:559–563.
artery remodeling. IFN-c secreted by dNK plays an
9 Li XF, Charnock-Jones DS, Zhang E, Hiby S, Malik S, Day K,
important role in VEGF-C induction in EVT. Both Licence D, Bowen JM, Gardner L, King A, Loke YW, Smith SK:
the p38 MAPK and the conventional JAK-STAT1 Angiogenic growth factor messenger ribonucleic acids in
pathways are involved in dNK-CM- and IFN-c-medi- uterine natural killer cells. J Clin Endocrinol Metab 2001;
ated VEGF-C secretions by HTR8 ⁄ SVneo. Other 86:1823–1834.
STAT1-independent, IFN-c-induced transcription fac- 10 Hanna J, Goldman-Wohl D, Hamani Y, Avraham I, Greenfield C,
Natanson-Yaron S, Prus D, Cohen-Daniel L, Arnon TI, Manaster I,
tors may also contribute and warrant further investi- Gazit R, Yutkin V, Benharroch D, Porgador A, Keshet E, Yagel S,
gation.
American Journal of Reproductive Immunology 67 (2012) 101–111
110 ª 2011 John Wiley & Sons A/S
dNK-CM MEDIATES VEGF-C SECRETION IN EVT

Mandelboim O: Decidual NK cells regulate key developmental 27 Goh KC, Haque SJ, Williams BR: p38 MAP kinase is required for
processes at the human fetal-maternal interface. Nat Med 2006; STAT1 serine phosphorylation and transcriptional activation
12:1065–1074. induced by interferons. EMBO J 1999; 18:5601–5608.
11 Eastabrook G, Hu Y, Dutz J, Tan R, MacCalman C, von Dadelszen P: 28 Wen Z, Zhong Z, Darnell Jr JE: Maximal activation of transcription
Cytokine stimulation of decidual NK cells increases IFN-c secretion by STAT1 and STAT3 requires both tyrosine and serine
but does not alter the secretion of other soluble products important phosphorylation. Cell 1995; 82:241–250.
in trophoblast migration, invasion, and placental angiogenesis. 29 Graham CH, Hawley TS, Hawley RG, MacDougall JR, Kerbel RS,
Pregnancy Hypertens 2010; 1(Suppl 1):S52. Khoo N, Lala PK: Establishment and characterization of first
12 Hu Y, Dutz JP, MacCalman CD, Yong P, Tan R, von Dadelszen P: trimester human trophoblast cells with extended lifespan. Exp Cell
Decidual natural killer cells alter in vitro first trimester extravillous Res 1993; 206:204–211.
cytotrophoblast migration; a role for interferon-gamma. J Immunol 30 Krutzik PO, Nolan GP: Intracellular phospho-protein staining
2006; 177:8522–8530. techniques for flow cytometry: monitoring single cell signaling
13 Hu Y, Eastabrook G, Tan R, MacCalman CD, Dutz JP, von events. Cytometry A 2003; 55:61–70.
Dadelszen P: Decidual NK cell-derived conditioned medium 31 Kalkunte SS, Mselle TF, Norris WE, Wira CR, Sentman CL, Sharma
enhances capillary tube and network organization in an S: Vascular endothelial growth factor C facilitates immune
extravillous cytotrophoblast cell line. Placenta 2010; 31: tolerance and endovascular activity of human uterine NK cells at
213–221. the maternal-fetal interface. J Immunol 2009; 182:4085–4092.
14 Parham P: NK cells and trophoblasts: partners in pregnancy. J Exp 32 Ramsauer K, Sadzak I, Porras A, Pilz A, Nebrada AR, Decker T,
Med 2004; 200:951–955. Kovair P: p38 MAPK enhances STAT1-dependent transcription
15 Boehm U, Klamp T, Groot M, Howard JC: Cellular responses to independently of Ser-727 phosphorylation. PNAS 2002;
interferon-gamma. Annu Rev Immunol 1997; 15:749–795. 99:12859–12864.
16 Boehm U, Guethlein L, Klamp T, Ozbek K, Schaub A, Futterer A, 33 Huang H, Rose JL, Hoyt DG: p38 mitogen-activated protein kinase
Pfeffer K, Howard JC: Two families of GTPases dominate the mediates synergistic induction of inducible nitric-oxide synthase by
complex cellular response to IFN-gamma. J Immunol 1998; lipopolysaccharide and interferon-c through signal transducer and
161:6715–6723. activator of transcription 1 Ser727 phosphorylation in murine
17 Ashkar AA, Croy BA: Interferon-gamma contributes to the aortic endothelial cells. Mol Pharmacol 2004; 66:302–311.
normalcy of murine pregnancy. Biol Reprod 1999; 61:493–502. 34 Deb DK, Sassano A, Lekmine F, Majchrzak B, Verma A,
18 Kommineni VK, Nagineni CN, Detrick B, Hooks JJ: Inflammatory Kambhampati S, Uddin S, Rahman A, Fish EN, Platanias LC:
mediators induce VEGF-A and VEGF-C secretion by human retinal Activation of protein kinase C delta by IFN-gamma. J Immunol
cells. FASEB J 2007; 21:711–719. 2003; 171:267–273.
19 Chen J, Zhou J, Wang J: Effect of interferon-gamma on the 35 Suzuki M, Tetsuka T, Yoshida S, Watanabe N, Kobayashi M, Matsui
expression of vascular endothelial growth factor C on Hep-2 N, Okamoto T: The role of p38 mitogen-activated protein kinase in
laryngeal carcinoma cell lines. Lin Chung Er Bi Yan Hou Tou Jing IL-6 and IL-8 production from TNF-a or IL-1b-stimulated
Wai Ke Za Zhi 2008; 22:108–111. rheumatoid synovial fibroblasts. FEBS Lett 2000; 465:23–27.
20 Gough DJ, Levy DE, Johnstone RW, Clarke CJ: IFNc signaling-does 36 Yin H, Gao L, Shen B, Chao L, Chao J: Kallistatin inhibits
it mean JAK-STAT? Cytokine Growth Factor Rev 2008; 19:383–394. vascular inflammation by antagonizing tumor necrosis factor-
21 O’Shea JJ: Jaks, STATs, cytokine signal transduction, and alpha-induced nuclear factor kappaB activation. Hypertension 2010;
immunoregulation: are we there yet? Immunity 1997; 7:1–11. 56:260–267.
22 Nguyen H, Ramana CV, Bayes J, Stark GR: Roles of 37 Meinke A, Barahmand-Pour F, Wohrl S, Stoiber D, Decker T:
phosphatidylinositol 3-kinase in interferon-gamma-dependent Activation of different STAT5 isoforms contributes to cell-type-
phosphorylation of STAT1 on serine 727 and activation of gene restricted signaling in response to interferons. Mol Cell Biol 1996;
expression. J Biol Chem 2001; 276:33361–33368. 16:6937–6944.
23 Nair JS, DaFonseca CJ, Tjernberg A, Sun W, Darnell Jr JE, Chait 38 Ramana CV, Kumar A, Enelow R: STAT1-independent induction of
BT, Zhang JJ: Requirement of Ca2+and CaMKII for Stat1 Ser727 SOCS-3 by interferon-gamma is mediated by sustained activation of
phosphorylation in response to IFN-gamma. Proc Natl Acad Sci U S A STAT3 in mouse embryonic fibroblast. Biochem Biophys Res Commun
2002; 99:5971–5976. 2005; 327:727–733.
24 David M, Petricoin III E, Benjamin C, Pine R, Weber MJ, Larner 39 Deb A, Haque SJ, Mogensen T, Silverman RH, Williams BR: RNA-
AC: Requirement for MAK kinase (ERK2) activity in interferon dependent protein kinase PKR is required for activation of NF-
alpha- and interferon beta-stimulated gene expression through kappa B by IFN-gamma in a STAT1-independent pathway.
STAT proteins. Science 1995; 269:1721–1723. J Immunol 2001; 166:6170–6180.
25 Kovarik P, Stoiber D, Eyers PA, Menghini R, Neininger A, Gaestel 40 Gough DJ, Sabapathy K, Ko EY, Arthur HA, Schreiber RD, Trapani
M, Cohen P, Decker T: Stress-induced phosphorylation of STAT1 at JA, Clark CJ: A novel c-Jun-dependent signal transduction
Ser727 requires p38 mitogen-activated protein kinase whereas IFN- pathway necessary for the transcriptional activation of interferon
gamma uses a different signaling pathway. Proc Natl Acad Sci U S A gamma response genes. J Biol Chem 2007; 282:938–946.
1999; 96:13956–13961. 41 Clarke CJ, Apostolidis V, Hii LL, Gough DJ, Trapani JA, Johnstone
26 Kovarik P, Stoiber D, Novy M, Decker T: STAT1 combines signals RW: Critical role of the transcription factor AP-1 for the
derived from IFN-gamma and LPS receptors during macrophage constitutive and interferon-induced expression of IFI 16. J Cell
activation. EMBO J 1998; 17:3660–3668. Biochem 2003; 89:80–93.

American Journal of Reproductive Immunology 67 (2012) 101–111


ª 2011 John Wiley & Sons A/S 111

You might also like