Articles in PresS. Am J Physiol Regul Integr Comp Physiol (May 13, 2004). 10.1152/ajpregu.00222.2004
Contractile properties of human placental anchoring villi
Anne E. Farleya , Charles H. Grahamb, Graeme N. Smitha, b
Departments of aObstetrics & Gynecology, bAnatomy & Cell Biology, Queen’s
University, Kingston, Ontario, Canada, K7L 3N6.
Short Title: Contraction of anchoring villi
All correspondence and reprint requests should be addressed to:
Dr. Graeme N. Smith
Department of Obstetrics and Gynaecology
Kingston General Hospital
76 Stuart St.
Kingston, Ontario
Canada, K7L 2V7
E-mail: [email protected]
Tel: (613) 548-2405
Fax number: 613-548-1330
Or
Dr. Charles H. Graham
Department of Anatomy and Cell Biology
Botterell Hall, Room 859
Queen’s University
Kingston, Ontario
Canada K7L 3N6
E-mail: [email protected]
Tel: (613) 533-2852
Fax: (613) 533-2566
Copyright © 2004 by the American Physiological Society.
R-00222-2004.R1 2
Abstract: The presence of myofibroblasts arranged in parallel to the longitudinal axes
of anchoring villi of the placenta has previously been described. Furthermore, it has been
suggested that intraplacental blood volume, and hence fetal-maternal oxygen-nutrient
exchange, may in part be regulated through the longitudinal contraction of anchoring
villi. We demonstrate here that anchoring villi have the ability to contract and relax
longitudinally. Anchoring villi from normal term human placentae were dissected and
suspended from force displacement transducers in order to determine their longitudinal
contractility in response to potassium chloride (KCl), NZ-Nitro -L– arginine methyl ester
hydrochloride (L-NAME) and the nitric oxide donors sodium nitroprusside (SNP) and
glyceryl trinitrate (GTN). Treatment with both KCl and L-NAME resulted in up to a
62% and 74% increase, respectively, in longitudinal contraction over resting tone. In
contrast, both SNP and GTN caused a dose-dependent relaxation of pre-contracted villi.
Immunohistochemistry of longitudinal sections of villi confirmed the presence of alpha
actin-containing cells in the extravascular space. Histological staining with H & E
confirm that the tissue used in these experiments were anchoring villi. These findings
suggest that the contraction of anchoring villi may be an important mechanism whereby
the placenta may regulate intraplacental volume.
Keywords: placenta, anchoring villi, blood flow, smooth muscle, contraction
R-00222-2004.R1 3
Introduction
The anchoring villi (figure 1) of the human placenta are tree-like structures made up of a
primary trunk attached to the chorionic plate, secondary branches, and tertiary, very fine
‘anchoring’ branches attached to the basal plate (2). Histochemically and ultrastructurally, smooth muscle cells and myofibroblasts have been identified in human
anchoring villi (11). They also showed that a strip of tissue dissected from an anchoring
villus was capable of lengthening and shortening in response to calcium chloride,
potassium chloride and barium chloride (11). It has been shown that myosin, the major
contractile protein, is present in large amounts within the “extravascular contractile
system” (5) of the anchoring villi (10). The innermost layer of longitudinally oriented
extravascular stromal cells of the villi are immunoreactive to all five of the characteristic
cytoskeletal proteins for smooth muscle cells (vimentin, desmin, alpha and gamma actin
and myosin) (4). In contrast to smooth muscle cells of the tunica media of blood vessels,
cells in the extravascular contractile system of anchoring villi also express
dipeptidylpeptidase IV (a known marker for the extravascular contractile system in stem
villi (8)). While the evidence for the presence of cells capable of contracting along the
longitudinal axes of anchoring villi is strong, little research has been carried out
demonstrating its physiological significance. The purpose of this study was to determine
whether anchoring villi are capable of contraction/relaxation in an isolated tissue bath and
to characterize the underlying mechanisms involved.
R-00222-2004.R1 4
Figure 1:
Chorionic Vein & Artery
Chorionic
Plate
Primary
Trunk
Secondary
Branch
Basal
Plate
Anchoring or
Tertiary
Branches
Figure 1:
Left: drawing of an anchoring villus (large arrow heads point to anchors or tertiary
branches attached to the basal plate.
Right: H&E staining of an anchoring villus (large arrow heads point to anchors or
tertiary branches attached to the basal plate).
R-00222-2004.R1 5
Materials and Methods
Term human placentae from uncomplicated pregnancies were obtained immediately
following caesarean section or vaginal delivery at the Kingston General Hospital,
Kingston, Ontario, Canada. Anchoring villi were isolated by gentle dissection from the
chorionic and basal plates. Villi were rinsed in Kreb’s solution to remove excess blood.
The villi were tied at both ends by silk suture, suspended from a force displacement
transducer, and placed in an isolated tissue bath in 15mL of standard Kreb’s solution at
37qC. Tissues baths were bubbled with 95% or 21% O2, 5% CO2, balance N2. In all
experiments, the force-displacement transducer (GRASS FTO3D) coupled to a
POWERLAB® (ADInstruments, Colorado, USA), was calibrated to a maximum of 2 gm
of tension. Placental arterial rings of approximately 3mm in length were used as positive
controls for contraction. The arterial ring was suspended from 2 stainless steel hooks
inserted through the lumen of the ring. Based on preliminary studies, the tissue was set to
1.5 gm resting tension and allowed to equilibrate in the tissue baths for one hour with
rinses of fresh oxygenated Kreb’s solution every 15 minutes.
Potassium Chloride Dose Response
After equilibration, the resting tension was adjusted to 1.5 gm and KCl 0.12 mol/L was
added to each bath to precontract the villi (n=6). The villi were allowed to equilibrate for
15 minutes and then rinsed three times in 15-minute increments for a total of 1hour. KCl
0.015 molL/L, 0.03 mol/L and 0.06 mol/L was added to each bath and allowed to
equilibrate for 15 minutes.
L-NAME dose response
R-00222-2004.R1 6
Villi (n=17) were suspended as previously described but not pre-contracted with KCl
treatment. L-NAME (L-NAME, Sigma – Aldrich, Canada) was added in increasing
concentrations (25-200 Pmol/L) as an accumulating dose to the tissue baths.
Sodium nitroprusside/glyceryl trinitrate dose response
The villi were maximally pre-contracted with KCl 0.06 mol/L. Sodium nitroprusside
(SNP; Sigma - Aldrich, Canada) in 5 increasing concentrations (0.7 µmol/L-700 µmol/L),
as an accumulating dose without rinsing, was added to the baths in 15-minute increments
(n = 6). In a separate set of experiments, glyceryl trinitrate (GTN; Tridil®, Sabex,
Boucherville, QC) in 5 increasing concentrations (1 mmol/L-15mmol/L), as an
accumulating dose, was added to the baths in 15-minute increments (n = 13). After the
final concentrations were added and allowed to equilibrate, the villi were rinsed, allowed
to equilibrate and again KCl 0.06mol/L was added to the bath to confirm tissue viability
by contraction.
Tissue Preparation
At the conclusion of the experiments the villi were weighed and processed for histology
and immunostaining for alpha-actin as previously described (12). Paraffin embedded
tissue was cut into 5 µm thick serial sections. For alpha-actin immunohistochemistry,
primary polyclonal antibody anti-alpha-actin (A2066; Sigma-Aldrich, Canada) was
applied to the sections overnight at a dilution of 1:50 in PBS buffer. Antigen was detected
using a biotinylated anti-rabbit IgG secondary antibody (1:200 dilution) and avidinbiotin-peroxidase (Vector Laboratories, Burlington, Ontario, Canada) followed by
precipitation of diaminobenzidine chloride (Dako, Mississauga, Ontario, Canada). Fresh
R-00222-2004.R1 7
tissue was processed and stained with H & E for histological identification of an
anchoring villus (Figure 1).
Statistical analysis:
Data were analyzed with repeated measures one-way ANOVA and post-hoc NewmanKeuls test for multiple comparisons between group means using GraphPad Prism 3.0
(GraphPad Software Inc., San Diego, Cal.). Results were statistically significant if
p<0.05. All results are stated as means ± standard deviation (SD).
R-00222-2004.R1 8
Results
The villi exhibited contraction in a dose-dependent manner to increasing concentrations
of KCl, with significance obtained at KCl concentrations of 0.015 mol/L and 0.06 mol/L
(figure 2, 3, p<0.05, n=6). KCl increased tension from 20% ± 12.0 to 62.3% ± 34.5 of
maximal contraction. The cumulative L-NAME administration caused contraction of the
tissue from its resting state and the results were statistically significant (figure 4, p<0.05,
n=17). L-NAME was capable of increasing tension from 63.5% ± 19.1 to 74.3% ± 21
above resting state. SNP produced a dose-dependent relaxation of the pre-contracted villi
over the cumulative concentration range that was statistically significant between groups
(figure 5, 6, p<0.05, n=6). SNP caused a decrease in maximal tension of 4.8% ± 5.5 to
58.1% ± 23.5. GTN induced relaxation of 20.2% ± 44.0 to 73.4% ± 51.0 (figure 7,
p<0.05, n=13). Immunostaining of anchoring villi demonstrated the presence of alphaactin throughout the longitudinal axis of the extravascular villous tissue of the primary
villus trunk (figure 8).
R-00222-2004.R1 9
Discussion:
Our study demonstrates that human chorionic villi are capable of contracting
along their longitudinal axes. Specifically, the results revealed that KCl, an agent that
depolarizes the cell and causes the release of intracellular calcium via a receptorindependent mechanism, was able to reversibly contract the villi. Furthermore,
pharmacological blockade of nitric oxide (NO) production with the NOS antagonist LNAME also induced contraction of the villi, thus indicating that NO-mediated signalling
plays a role in regulating villous tone. In support of this concept are the results of
experiments in which two classes of NO mimetic agents, namely SNP and GTN, were
able to induce relaxation of KCl pre-contracted villi. The relatively high dose of LNAME required to produce the observed effect likely reflects the layers of tissue needed
to penetrate to the villi to inhibit NO production and result in longitudinal contraction.
Smooth muscle-like spindle-shaped cells were first observed in the chorionic plate
of the term human placenta in 1906 (11). However, it was not until 1916 that Iizuka
confirmed their smooth muscle cell identity (11). Also noted at that time was that these
muscle cells were denser at the center of the placenta where the umbilical cord inserted
and more sparse towards the periphery. In 1922, Naujoks proposed the hypothesis that
the human placenta is capable of contraction (11) and Krantz & Arey later demonstrated
that these cells in the placenta have all the morphological characteristics of smooth
muscle (11). More recently, Graf et al. confirmed the presence of myofibroblasts in stem
villi by demonstrating the immunolocalisation of vimentin, desmin, alpha actin, gamma
actin and myosin in these cells (5). Those investigators also proposed that the change in
villus length is an active process (5). Using a similar approach, we confirmed previous
findings in which alpha-actin was shown to be present in longitudinally oriented cells
R-00222-2004.R1 10
within the extravascular stroma of the primary trunk of anchoring villi (4;6). Graf et al.
also identified a system of elastic-collagen fibres running longitudinally along the fetal
vessels of the placenta, and confirmed that all three elements of elastic tissue were
present in human placenta (7). Thus, the demonstration of contractile elements in
anchoring villi supports the hypothesis that the placenta has the ability to regulate its own
intravillous volume. This may in fact be a combination of longitudinal
contraction/relaxation of both the extravascular contractile system of anchoring villi and
the smooth muscle cells of the tunica media of blood vessels resulting in an overall
shortening/lengthening of the anchoring villi within the placenta. The
contraction/relaxation of the villi may impact on both maternal perfusion and fetal villous
blood flow and thus contribute to maternal-fetal perfusion matching. For example,
increase in fetal oxygen/nutrient demands may result in increased maternal/intravillous
placental flow as a function of relaxation of the anchoring villi. A failure of this effect in
regulating placental perfusion may play a role in adverse obstetrical outcomes of
placental origin such intrauterine growth restriction. Similarly, as pre-eclampsia is
thought to be related to abnormalities of placental perfusion (perhaps regionally), this
may be secondary to an inherent defect in the placenta itself or contribute to poor
maternal perfusion because of impaired spiral artery invasion.
The mechanisms controlling villous contractions are not fully characterized. As
our findings indicate, locally produced NO may be involved in this process. Carbon
monoxide (CO), a molecule produced during the breakdown of heme through a reaction
catalyzed by the enzyme heme oxygenase (HO), may also play a role in this regulation.
There is evidence that CO is capable of inducing vascular relaxation through a
mechanism that, in a similar manner to NO, involves activation of soluble guanylyl
R-00222-2004.R1 11
cyclase and generation of cGMP (1;9). While in the present study both SNP and GTN
induced villus relaxation, the concentrations required to induce significant relaxation
were much lower for SNP than for GTN. One possible explanation for this is that GTN is
a pro-drug that requires biotransformation while SNP releases NO more readily in
solution. It is unknown if the contractile elements of anchoring villi are more/less
sensitive to vasodilators/constrictors compared to vascular smooth muscle. In Bobier et al
(3) rings of chorionic plate vessels (CPV) were pre-contracted with KCl 0.12M and submaximally contracted (30-50% of max) with KCl 0.03M, the same concentrations of KCl
achieved similar results in the anchoring villi. Relaxation was achieved with SNP in the
CPV rings using concentration ranges of 10-12 to 10-1 M , whereas, SNP in concentrations
that were 6 magnitudes higher ( 0.7x10-6 to 0.7x10-3) were required to cause relaxation in
the tissue. The observation of greater drug concentrations required in the present study
likely relates to the need to diffuse across a larger tissue mass to get to the site of action.
In summary, our findings demonstrate that anchoring villi are capable of
contraction/relaxation. This may represent a novel mechanism by which the placenta
controls its own volume and thereby maintains adequate nutrient and oxygen delivery as
well as participating in maternal-fetal perfusion matching.
Acknowledgements:
Our thanks to Dr. Kanji Nakatsu for assistance with interpretation of the data and Dr.
Peter Kaufman for his interest in this work. This work was supported by a grant from the
Canadian Institutes of Health Research (grant # MOP64305) and the Heart and Stroke
R-00222-2004.R1 12
Foundation of Ontario. Dr. Graeme N. Smith is the recipient of a New Investigator
Award from the Canadian Institutes of Health Research.
% Increase in
Tension
120
100
80
60
40
20
0
b
ab
a
0.015
0.03
0.06
Concentration of KCl (M)
Figure 2: KCl dose response (n = 6). Group means with different letters are statistically
significant from each other (p < 0.05). Bars represent the mean percent change (± SD) in
contractile force.
Chart Window
M
0.030
0.015
5
4
3
WASHES
2
1
0.12
Villi 1 (g)
0.060
05/09/2002 12:29:17.100 PM
1:06:40
1:23:20
1:40:00
1:56:40
2:13:20
2:30:00
Figure 3: Powerlab® chart tracing of KCl induced contraction.
2:46:40
R-00222-2004.R1 13
% Increase
in Tension
120
a
a
25
50
b
b
150
200
b
80
40
0
100
Concentration of L-NAME (mmol/L)
Figure 4: The effect of L – NAME on resting anchoring villi (n = 17). Group means with
different letters are statistically significant from each other (p < 0.05).Bars represent the
mean percent change (± SD) in tension from resting state.
Chart Window
22/08/2002 12:56:39.460 PM
2.6
Villi 2 SNP (g)
2.4
2.2
2.0
0.06M
KCL
1.8
1.6
:23:20
1:40:00
1:56:40
2:13:20
2:30:00
2:46:40
3:03:20
3:20:00
R-00222-2004.R1 14
Figure 5: Powerlab® chart tracing of KCl induced contraction (arrow at start) and SNP
induced relaxation (short arrows point to additions of SNP 0.7 µmol/L-700 µmol/L)
% Decrease in Force
120
c
c
b
80
40
a
0
0.7uM
7uM
70uM
700uM
Concentration of SNP (uM)
Figure 6: Effect of SNP on pre – contracted anchoring villi (n = 6). Sodium
Nitroprusside (SNP) was added in cumulative increasing concentrations. Group means
with different letters are statistically significant from each other (p < 0.05). Bars
represent the mean percent change (± SD) in contractile force compared to maximal
contraction.
d
% Relaxation
160
bc
120
a
a
ab
80
40
0
1
3
5
10
15
Concentration of GTN (mmol/L)
Figure 7: The effect of GTN on pre-contracted anchoring villi (n = 13). Glyceryl
Trinitrate (GTN) was added in cumulative increasing concentrations. Group means with
different letters are statistically significant from each other (p< 0.05). Bars represent the
R-00222-2004.R1 15
mean percent change (± SD) in contractile force compared to maximal
contraction.
A
C
B
D
Figure 8: Immunostaining with anti-alpha actin and DAB peroxidase substrate (dark
brown). Bars represent 100 microns. A) Positive arterial ring shows concentrated
staining in the media of the vessel, B) Control arterial ring, no anti-alpha actin staining,
C) Longitudinal section of the primary villus trunk of an anchoring villus positive for
alpha actin shows staining in the extravascular contractile system, D) Boxed area of slide
C shows detailed staining of longitudinal cells in the extravascular contractile system.
R-00222-2004.R1 16
Reference List
1. Bainbridge SA, Farley A. E., McLaughlin BE, Graham CH, Marks GS, Nakatsu K,
Brien JF, and G. N. Smith. Carbon Monoxide Decreases Perfusion Pressure in
Isolated Human Placenta. Placenta 23: 563-569, 2002.
2. Benirschke K and Kaufmann P. Pathology of the Human Placenta. New York,
Springer-Verlag. 2000.
3. Bobier HS, G. N. Smith, Pancham SR, and Brien JF. Comparative study of sodium
nitroprusside induced vasodilation of human placental veins from premature and
full-term normotensive and preeclamptic pregnancy. Can J Physiol Pharmacol 73:
1118-1122, 1995.
4. Demir R, Kosanke G, Kohnen G, Kertschanska S, and Kaufmann P. Classification
of human placental stem villi: Review of structural and functional aspects. Microsc
Res Tech 38: 29-41, 1997.
5. Graf R, Langer J, Schonfelder G, Oney T, Hartel-Schenk S, Reutter W, and
Schmidt HHHW. The extravascular contractile system in the human placenta. Anat
Embryol 190: 541-548, 1994.
6. Graf R, Matejevic D, Schuppan D, Neudeck H, Shakibaei M, and Vetter K.
Molecular anatomy of the perivascular sheath in human placental stem villi: the
R-00222-2004.R1 17
contractile apparatus and its association to the extraacellular matrix. Cell Tissue Res
290: 601-607, 1997.
7. Graf R, Neudeck H, Gossrau R, and Vetter K. Elastic fibres are an essential
component of human placental stem villous stroma and an integrated part of the
perivascular contractile sheath. Cell Tissue Res 283: 133-141, 1996.
8. Graf R, Schonfelder G, Muhlberger M, and Gutsmann M. The perivascular
contractile sheath of the human placental stem villi: Its isolation and
characterization. Placenta 16: 57-66, 1995.
9. Hussain, A. S., G. S. Marks, J. F. Brien, and K. Nakatsu. The soluble guanylyl
cyclase inhibitor 1H-[1,2,4]oxadiazolo[4,3- alpha]quinoxalin-1-one (ODQ) inhibits
relaxation of rabbit aortic rings induced by carbon monoxide, nitric oxide, and
glyceryl trinitrate. Can J Physiol Pharmacol 75: 1034-1037, 1997.
10. Huszar, G. and P. Bailey. Isolation and characterization of myosin in the human
term placenta. Am J Obstet Gynecol 135: 707-712, 1979.
11. Krantz, K. and J. Parker. Contractile Properties of Smooth Muscle In The Human
Placenta. Clinical Obstetrics Gynecology 6: 26-38, 1963.
12. Lash GE, McLaughlin BE, MacDonald-Goodfellow SK, G. N. Smith, Brien JF,
Marks GS, Nakatsu K, and Graham CH. Relationship between tissue damage and
R-00222-2004.R1 18
heme oxygenase expression in chorionic villi of term human placenta. Am J Physiol
Heart Circ Physiol 284: H160-H167, 2003.