Treatment of ATP-induced
preeclampsia-like symptoms
with alkaline phosphatase in
pregnant rats
pter 8
Floor Spaans1 | Theo Borghuis2 | Pieter A. Klok2 | Paul de Vos1
Harry van Goor2 | Winston W. Bakker2 | Marijke M. Faas1
University of Groningen and University Medical Center
Groningen, Department of Pathology and Medical Biology,
1
Division of Medical Biology, 2Division of Pathology,
Groningen, the Netherlands.
In preparation
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Abstract
Introduction: Effective therapies for the severe pregnancy complication preeclampsia are
lacking.
High plasma ATP levels were found in women with preeclampsia and were suggested to
play a role in its pathogenesis, since ATP infusion in pregnant rats was shown to induce
preeclampsia-like symptoms. ATP may exert its effects by decreasing hemopexin (Hx) activity
or by increasing inflammation. We therefore tested whether therapeutic treatment of ATPinfused pregnant rats with alkaline phosphatase (AP), which hydrolyses ATP, decreased the
development of preeclampsia-like symptoms.
Methods: After infusion with ATP or saline (control) on day 14 of pregnancy, rats received
daily i.v. injections of AP (60 U/kg bw) or saline until day 20 of pregnancy. Plasma samples
were taken on day 13, 15, 17 and 20, and urine samples were obtained on day 17. At sacrifice
on day 20, kidneys and placentas with mesometrial triangle were collected. Plasma ATP
levels, hemopexin (Hx) activity, AP activity, urinary albumin excretion, foetal and placental
weight, trophoblast invasion and the inflammatory response in the kidney were analysed.
Results: ATP infusion increased plasma ATP levels, decreased Hx activity, induced
albuminuria and increased various glomerular leukocyte populations and increased numbers
of trophoblast cells in the mesometrial triangle. AP treatment of ATP-infused pregnant rats
reduced plasma ATP levels, decreased glomerular CD206+ macrophages and increased
glomerular iNOS+ macrophages and decreased trophoblast cells in the mesometrial triangle,
while Hx activity and urinary albumin excretion were unchanged.
Conclusions: AP treatment reduced plasma ATP levels and positively affected glomerular
inflammation. It did however not affect Hx activity and albuminuria. This limited effect of AP
in ATP-infused animals could be due to the lack of increase in total AP activity in AP treated
animals, or the still slightly increased ATP levels after AP treatment. This explorative study
shows that AP treatment may be a promising therapeutic against ATP-induced preeclampsia
symptoms.
144
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Introduction
Preeclampsia is a severe pregnancy complication, in which hypertension and proteinuria
develop in the second half of pregnancy. It occurs in approximately 3-5% of the pregnancies
[1] and unfortunately, effective treatment is unavailable, except for delivery of the child
and placenta. Although the exact pathogenesis is still unknown, poor placentation in the
first trimester is thought to be the main cause of the symptoms later in gestation. Poor
placentation is thought to lead to oxidative stress and inflammation in the placenta, which in
turn is thought to produce factors such as anti-angiogenic factors (such as sFlt-1 and sEng),
syncytiotrophoblast microparticles (STBMs), cytokines [2,3] and ATP [4]. These are suggested
to contribute to maternal systemic inflammation, endothelial dysfunction, and ultimately in
hypertension and proteinuria in the second half of pregnancy [2,3].
Hemopexin (Hx) is a free heme scavenger [5], but also has protease activity. This protease
activity can be inhibited by extracellular nucleotides like ATP [6]. It can also be reactivated
by ATP-hydrolysing enzymes like CD39 (ectonucleoside triphosphate diphosphohydrolase 1;
ENTPD1) and alkaline phosphatase (AP) [6]. During healthy human pregnancy, the plasma
Hx activity increases around week 15, and remains high throughout gestation [7,8]. The
increased Hx activity during pregnancy may play a role in the decreased responsiveness to
angiotensin II, displayed in healthy pregnant women [9,10]. In women with preeclampsia,
Hx activity was decreased [7,10]. This was suggested to be due to high ATP levels in these
preeclamptic women [7]. The decreased Hx activity during preeclampsia may play a role in
the pathophysiology of preeclampsia, since decreased Hx activity may result in increased
angiotensin II sensitivity [8].
We have previously shown that ATP infusion into pregnant rats decreased Hx activity and
induced preeclampsia-like symptoms, such as albuminuria, decreased foetal weight,
systemic inflammation and placental ischemia [11]. In the current study we hypothesized that
treatment of ATP-infused pregnant rats with AP, which hydrolyses ATP, reduces preeclampsialike signs. We therefore analysed the effect of AP treatment in ATP-infused and salineinfused control pregnant rats on plasma ATP levels, Hx activity, AP activity, urinary albumin
excretion, foetal and placental weight, trophoblast invasion in the mesometrial triangle and
the inflammatory response in the kidney.
Methods
Animals
The Institutional Animal Care and Use Committee of the University of Groningen approved all
animal experiments. Female Wistar outbred rats (about 200 g) were kept in a temperature
and light-controlled room (lights on from 7:30 AM till 7:30 PM) with free access to food
and water. Until selection for experiments, vaginal smears were taken daily, and rats were
145
8
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
rendered pregnant by housing them on pro-oestrus with fertile males for one night. When
spermatozoa were detected in the smear the next day, this day was designated as day 0
of pregnancy. A cannula was inserted into the right jugular vein under isoflurane/oxygen
anaesthesia on day 0 or 1 of pregnancy as described previously [12]. The cannula allows
stress free infusions and blood sampling. Rats were subdivided in four treatment groups: 1)
saline infusion + saline treatment (saline group; n=7), 2) saline infusion + AP treatment (AP
group; n=7), 3) ATP infusion + saline treatment (ATP group; n=11) and 4) ATP infusion + AP
treatment (ATP + AP group; n=8).
Experimental design
Rats were infused with 3000 μg/kg bw ATP in 2.0 ml saline or with 2.0 ml saline alone on day
14 of pregnancy for one hour as previously described [11]. During the infusion (30 minutes
after the start), the rats received a single i.p. injection of alkaline phosphatase (AP) of 60
units/kg bw (bovine intestinal AP; AM-Pharma, Bunnik, the Netherlands) in 300 µl saline,
or 300 µl saline alone as a control. After this first injection, rats received i.v. injections (via
the cannula) of AP (60 units/kg bw) or saline every 24 hours, until sacrifice (day 20). EDTA
blood samples (300 µl) were obtained via the jugular vein cannula on day 13, 15 and 17 of
pregnancy for measurement of plasma ATP levels. Rats were sacrificed by aortic puncture
under anaesthesia (isoflurane/oxygen) on day 20 of pregnancy, at which an EDTA (for Hx
measurement) and heparin (for measurement of AP activity) blood sample were obtained.
Foetuses and placentas were isolated and weighted, and the numbers of resorptions were
noted. Kidneys and placenta’s with mesometrial triangle were obtained and snap frozen
(kidney) or fixed in zinc-buffer for 24 hours (placenta’s with mesometrial triangle) as
described before [13] to evaluate the presence of immune cells in the kidney, and trophoblast
invasion in the mesometrial triangle.
Plasma ATP levels
Plasma from EDTA blood samples obtained on day 13, 15 and 17 of pregnancy was diluted
(1:3, total volume of 100 µl) with dilution buffer (25 mM Tris/H3PO4 pH=7.8 with 4.5 mM
EDTA) in a 96-wells plate. D-Luciferine solution of 87 μg/mL luciferin (Beetle E 1602,
Promega, USA) supplemented with MgCl2 (10.0 mM) and luciferase (500,000 RLU/ml in PBS
(Quantilum r-luciferase; Promega, Madison, Wisconsin, USA) was added (100 µl per well).
Instantly, the relative light units (RLU) of each sample were measured for three times on a
luminometer (Thermo Luminoskan Ascent, Thermo Fisher Scientific, Waltham, MA, USA). A
standard curve calculated from samples with concentrations between 10−3 M till 10−10 M
ATP in dilution buffer was used to calculate the plasma ATP concentration from the average
of three measurements.
146
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Plasma Hx activity
Plasma from EDTA blood samples obtained on day 20 was used for analysis of the protease
activity of Hx. The protease activity of Hx using the ‘glomerular apyrase stripping assay’ on
kidney cryostat sections, and semi-quantitatively scored as described previously [11,14]. In
short, this assay is based on the ability of active Hx to decrease the expression of extracellular
matrix molecules like CD39. Rat cryostat sections (4 µm) were fixed in acetone (10 min.) and
incubated with EDTA plasma samples from day 20 (100 µl, diluted 1:8 in PBS) for 60 min.
After incubation the sections were histochemically stained for CD39 expression according to
standard methods [14]. Reaction product was evaluated in a double-blind fashion. Intensity
of CD39 expression was semi-quantitatively scored by an independent observer using 7
categories, with 0 signifying the highest staining intensity and 7 demonstrating absence of
staining. The assay was performed in duplo, and the average score was taken.
Plasma total and placental AP activity
Total and placental AP activities were measured using a chromogenic substrate assay for
AP activity. Heparinized plasma samples from day 20 of pregnancy were diluted (1:40) with
buffer in a 96-wells plate. Buffer with substrate (Tris 0.1 M pH 9.8 with MgCl2 2 mM and
PNPP substrate 1.25 mM) was added to the samples. Substrate breakdown (generation of
4-nitrophenol) by AP was analysed by measuring the absorbance at 405 nm using a Varioskan
spectrophotometer (Thermo Fisher Scientific) at one minute intervals for 30 minutes in total.
The slope of the absorbance curve over 30 minutes resembled the AP activity. Four isoforms
of AP exist: tissue non-specific (or liver-bone-kidney) AP, intestinal AP, germ cell AP and
placental AP [15]. Placental AP is the only heat stable AP isoform [16]. Therefore, placental AP
activity was measured by heating part (30 µl) of the plasma to 65°C for 1 hour and analysing
AP activity according to the same protocol.
8
Kidney AP activity
To evaluate the endogenous renal AP activity, rat cryostat sections (4 µm) were stained for
AP activity using a histochemical enzyme assay according to Gomori method described by
Van Goor [17]. Rat cryostat kidney sections were dried, fixed with ice-cold acetone (10 min.)
and dried again. All sections were then incubated in calciumchloride solution with 48.5mM
diethylbarbituric acid sodium salt, 2.46mM magnesiumsulfate and 12mM β-glycerophosphate
(pH ±9.0) for 60 min. After 5 min. incubation in calciumchloride (0.07 M; pH 9.2), they were
incubated in Cerium incubation-medium (0.3M Glycine, 3mM cesiumchloride and 9 mM
sodiumcitrate; pH9.5) for 30 min. Sections were subsequently incubated with 0.1M Glycine/
NaOH (pH8.5) for 5 min., with 0.1M Glycine/NaOH (pH8.5) with 0.3% H2O2 for 15 min., and
then with 0.1M Tris/HCl (pH7.6) for 5 min. Finally sections were incubated with 0.1M Tris/HCl
(pH7.6) with 15mM Imidazole and 1.4mM 3,3′-Diaminobenzidine (at 60°C for 10 min.) and
haematoxylin. Sections were washed in tap water and covered with glycerin/gelatin. All of
147
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
the incubation steps, except for the last visualization step with 3,3′-Diaminobenzidine, were
carried out at room temperature.
Semi-quantitative analysis of AP activity in the kidney
Intensity of staining for AP activity was different in the tubules and glomeruli between the
groups and was therefore separately scored. Staining of tubular and glomerular AP intensity
was semi-quantitatively scored by an independent observer using 7 categories, with 0
signifying absence of staining and 7 demonstrating the highest staining intensity. Placental
AP staining intensity was semi-quantitatively scored in the same way.
Urinary albumin excretion
Rats were housed in metabolic cages for 24 hours on day 17 of pregnancy to obtain 24 hour
urine samples. Urinary albumin was analysed in the urine samples (diluted 1:5) using a rat
albumin ELISA Nephrat kit (Exocell, Philadelphia, PA, USA), according to the manufactures
instructions.
Immunohistochemical staining of kidney tissue
Rat cryostat kidney sections (4 µm) were stained for the presence of total macrophages
(mouse-anti-rat CD68, 1:100 diluted, clone ED1, AbD Serotec), iNOS+ macrophages (rabbitanti-rat iNOS, 1:2000 diluted; Abcam, Cambridge, UK), CD206+ macrophages (rabbit-antirat CD206, 1:1000 diluted; Abcam), granulocytes (mouse-anti-rat His48, 1:50 diluted, BD
Biosciences) and T lymphocytes (mouse-anti-rat CD3, 1:100 diluted, Abcam). All sections
were fixed with ice-cold acetone (10 min.). Only the sections stained for iNOS were preincubated with a mixture of 0.5% saponin, 4% BSA, 1% normal goat serum and 10% normal
rat serum in PBS (30 min.) and washed with PBS afterwards. Before incubation with primary
antibodies (with 1% normal rat serum (60 min.)), sections stained for CD68 were incubated
with 10% normal goat serum (30 min.) and sections stained for CD206, iNOS, His48 and CD3
with 2% BSA with 1% ELK in PBS (20 min.). After washing with PBS, sections were blocked
with 3% H2O2 in methanol and with a Biotin blocking kit (Dako, Heverlee, Belgium). After
washing with PBS, biotin-conjugated goat-anti-mouse (for CD68, CD3 and His48, Southern
Biotech, Birmingham, AL, USA) and biotin-conjugated goat-anti-rabbit (for iNOS and CD206,
Dako) were added as a second step (30 min.). After washing with PBS, peroxidase conjugated
streptavidin (Dako) was added (30 min.) and the staining was subsequently visualized by
3-amino-9-ehtyl-carbazole and haematoxylin. All of the incubation steps were carried out at
room temperature.
After staining, all slides were scanned with the Aperio TMAscanner (Aperio, Vista, USA).
Control sections (sections incubated without the first antibody) were consistently negative.
148
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Analysis of kidney sections
Kidney sections stained for total macrophages, iNOS+ and CD206+ macrophages,
granulocytes and lymphocytes were quantitatively scored in a double-blind manner by an
independent observer. Positive cells were counted in glomeruli of the total kidney section,
and results are expressed as number of positive cells per glomerulus. The M1/M2 ratio was
calculated by dividing the iNOS+ macrophages (M1) by the CD206+ macrophages (M2).
Immunohistochemical staining of placental sections: selection of placental sections
After 24 hour fixation with zinc-buffer, placental tissue was dehydrated with 60 min.
incubations in 70, 96, 100% alcohol and in xylol, and afterwards overnight incubated in
paraffin and embedded in paraffin. Subsequently, the entire placenta with mesometrial
triangle was cut into 4 µm sections. In order to stain the same location within the mesometrial
triangle, only sections containing the maternal channel (the large centrally located artery in
the rat placenta) were used for immunohistochemical staining [13].
Immunohistochemical staining of placental tissue for cytokeratin
Placental sections were stained for the presence of trophoblast cells (mouse-anti-human
cytokeratin, clone MNF116, 1:200 diluted, Dako, Heverlee, Belgium) according to methods
described by Pijnenborg et al. [13]. In short, sections of placentae were pre-heated for 60
min. at 60 °C and deparaffinised. After incubation with blocking solution (containing 2%
BSA, 1% ELK and 0,1% Tween-80 in PBS, for 20 min.) sections were incubated with primary
antibodies (at 4 °C; overnight). Sections were washed with PBS and thereafter all sections
were incubated with blocking solution (20 min.) followed by incubation with biotinconjugated rabbit-anti-mouse (Dako, 30 min.). Subsequently, sections were washed with
PBS, after which alkaline phosphatase conjugated streptavidin (Dako) was added (30 min.).
Sections were visualized by nitro-blue tetrazolium and 5-bromo-4-chloro-3’-indolyphosphate
(NBT-BCIP) and counterstained with periodic acid-Schiff (PAS) staining and haematoxylin. All
of the incubation steps except incubation with the primary antibody were carried out at room
temperature.
After staining, all slides were scanned with the Aperio TMAscanner (Aperio). Control sections
(sections incubated without the first antibody) were consistently negative.
Analysis of placental sections
All analyses were performed with the Aperio Imagescope program (Aperio Vista, USA).
Trophoblast invasion was analysed by calculating the surface area invaded by trophoblast
cells and the total surface area of the mesometrial triangle; percentage of surface area of the
mesometrial triangle invaded by trophoblast cells was calculated. The amount of positively
cytokeratin stained pixels as well as the total amount of pixels (reflecting the total amount
of tissue) in the mesometrial triangle were calculated using the ‘Positive pixel count V9’
149
8
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
algorithm and percentage positive area was calculated.
Statistical analysis
Differences were compared using Mann Whitney U tests. To evaluate the effect of ATP
infusion and/or AP treatment on the longitudinal data of the plasma ATP levels, Friedman
repeated measures tests were performed followed by Dunn’s post-tests comparing day 13
pre-infusion values with day 15 or 17 values separately within each experimental group.
Data are presented as medians with interquartile range. Differences were considered to be
significant if p<0.05 and a statistical trend if p<0.1.
Results
ATP infusion increased ATP levels and decreased plasma Hx activity
AP is able to hydrolyse ATP. Therefore, we analysed the effect of AP treatment on plasma ATP
levels on day 13, 15 and 17 of pregnancy. Mean ATP levels on day 13 were 0,23 (±0,11) µM.
Compared with day 13 of pregnancy (pre-infusion values), plasma ATP levels increased on
day 15 and 17 of pregnancy in the ATP infused rats (p<0.05; Figure 1A). Even though to a much
lesser extent, treatment with AP in ATP infused animals also increased ATP levels on day 17
compared to day 13 (p<0.05; Figure 1A). Unfortunately, due to the sampling method at day
20 (aortic puncture), plasma ATP levels on that day were extremely high in all groups (0.01
Figure 1. Plasma ATP levels and
Hx activity. A) Deviation in plasma
ATP levels on day 15 and 17 of
pregnancy compared with day 13 in
ATP (solid lines) or saline infused
(dashed lines) and AP (squares)
or saline treated (circles) pregnant
rats. *p<0.05; significantly different
compared with pre-infusion (day
13) values in the same experimental
group, Friedman followed by Dunns
post-tests. B) Plasma Hx activity
on day 20 in ATP (black bars) or
saline infused (white bars) and
AP (dashed bars) or saline treated
(open bars) pregnant rats. *p<0.05;
significantly decreased compared
with saline infused animals, MannWhitney U test.
150
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
mM). This is probably due to activation of circulating WBC/RBC or platelets [18], due to the
sampling procedure, i.e. aortic puncture under isoflurane/oxygen anaesthesia.
Decreased plasma ATP levels may result in increased Hx activity. Therefore, we assessed the
effect of AP treatment on plasma Hx activity on day 20 of pregnancy. Since we could only
obtain limited amounts of plasma on day 13, 15 and 17, we were only able to measure Hx
activity on day 20 of pregnancy, as during sacrifice larger amounts of blood/plasma were
collected. Plasma Hx activity decreased in ATP infused animals (p<0.05; Figure 1B). However,
no significant effect of AP treatment was observed in either saline or ATP infused rats (Figure
1B).
No changes in plasma AP activity
Plasma AP activity was measured at day 20 of pregnancy. Since AP activity is measured in
heparinized plasma, which could only be obtained on day 20, we were only able to measure
AP activity on this time point. ATP infusion or AP treatment had no effect on total AP activity
(Figure 2A). AP treatment (both in saline and ATP infused animals) appeared to reduce
placental AP activity (Figure 2B), however, this was not significant.
Decreased tubular and glomerular AP activity after AP treatment
Kidney sections were stained for AP activity according to standard methods. In the kidney,
AP activity was mainly found in glomeruli and tubuli (Figure 3A-H). In the tubules, AP activity
appeared to be specifically present in the proximal tubuli (in the brush border membrane)
and was absent in de distal tubules (Figure 3A, C, E and G). As AP activity patterns were
different between the groups in the tubuli and glomeruli, these were scored separately. No
significant effect of ATP infusion on tubular AP activity was observed (Figure 3I). However, AP
8
Figure 2. Total and placental plasma AP activity. Total (A) and placental (B) AP activity in units per litre (U/L) on day
20 of pregnancy in ATP (black bars) or saline infused (white bars) and AP (dashed bars) or saline treated (open bars)
pregnant rats. Figure 2. Total and placental plasma AP activity. Total (A) and placental (B) AP activity in units per litre
(U/L) on day 20 of pregnancy in ATP (black bars) or saline infused (white bars) and AP (dashed bars) or saline treated
(open bars) pregnant rats.
151
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Figure 3. Renal AP activity. Representative photomicrographs of tubules (A+C+E+G) and glomeruli (B+D+F+H) from
saline infused (A+B), AP treated (C+D), ATP infused (E+F) and ATP infused and AP treated animals (G+H), displaying AP
activity as black reaction product, are shown. AP activity in proximal tubules is shown by red arrows, while absence of
AP activity in distal tubules is shown by yellow arrows. Tubular (I) and glomerular (J) AP activity on day 20 of pregnancy
in ATP (black bars) or saline infused (white bars) and AP (dashed bars) or saline treated (open bars) pregnant rats.
a:p<0.05; decreased in AP treated animals compared with saline treated animals that received the same infusion,
Mann-Whitney U test.
treatment decreased AP activity in saline infused animals, but not in ATP infused animals
(p<0.05; Figure 3I). In the glomeruli, ATP infusion also had no significant effect on AP activity
(Figure 3J). Glomerular AP activity was unaffected by AP treatment in saline-infused animals,
however, AP treatment in ATP-infused animals decreased glomerular AP activity (p<0.05;
Figure 3J).
Urinary albumin excretion is increased in ATP infused animals
We next assess the effect of AP treatment on 24 hour urinary albumin excretion. Urinary
albumin excretion increased in rats infused with ATP compared with saline infusion (p<0.05;
152
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Figure
4.
excretion.
Urinary
Urinary
albumin
albumin
excretion in 24h urine samples
obtained on day 17 of pregnancy
in ATP (black bars) or saline
infused (white bars) and AP
(dashed bars) or saline treated
(open
bars)
pregnant
rats.
*p<0.05; significantly increased
compared with saline infused
animals, Mann-Whitney U test.
Figure 4) on day 17 of pregnancy. However, AP treatment did not change the urinary albumin
excretion in ATP or in saline infused rats (Figure 4).
AP treatment decreased glomerular CD206+ macrophage, increased iNOS+ macrophage
numbers and increased M1/M2 ratio
Total macrophages: CD68+ macrophages were found in the glomeruli and occasionally
in the tubulointerstitium, typically around vessels. No clear effect of ATP infusion or AP
treatment on the location of CD68+ macrophages was detected. ATP infusion increased
glomerular macrophage numbers (p<0.05; Figure 5A). No effect of AP treatment on glomerular
macrophage numbers was observed in either saline or ATP infused animals (Figure 5A).
CD206+ macrophages: CD206+ macrophages were found in the glomeruli, on the glomerularinterstitial border and sporadically interstitially. No changes in the location of these cells
were observed after ATP infusion or AP treatment. Infusion of ATP had no effect on glomerular
CD206+ macrophage infiltration (Figure 5B), while treatment with AP decreased glomerular
CD206+ macrophages in ATP, but not in saline infused animals (p<0.05; Figure 5B).
iNOS+ macrophages: Small numbers of iNOS+ macrophages were found in our kidney
samples, and these were mainly located in the glomeruli, without any effect of ATP infusion
or AP treatment on their location. Decreased numbers of glomerular iNOS+ macrophages
were observed after ATP infusion (p<0.05; Figure 5C). Treatment of ATP infused rats with AP
increased glomerular iNOS+ macrophage numbers compared with saline treated ATP infused
rats (p<0.05; Figure 5C), while AP treatment in saline infused animals had no effect on iNOS+
macrophage numbers.
M1/M2 ratio: As differences in iNOS+ and CD206+ macrophages were found after ATP
infusion, the M1 (iNOS+)/ M2 (CD206+) ratio was calculated. After ATP infusion the M1/M2
ratio was decreased compared with saline infused animals (p<0.05; Figure 5D), while after AP
treatment, in ATP infused animals only, the M1/M2 ratio was increased compared with saline
treatment (p<0.05; Figure 5D).
153
8
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Figure 5. Glomerular leukocyte infiltration. Glomerular counts of total macrophages (CD68, A), CD206-positive
macrophages (B), iNOS-positive macrophages (C), the M1/M2 ratio (D), granulocytes (His48, E) and lymphocytes
(CD3, F) in ATP (black bars) or saline infused (white bars) and AP (dashed bars) or saline treated (open bars) pregnant
rats. *:p<0.05; different compared with saline infused animals that received saline treatment, Mann-Whitney U test.
a:p<0.05; b:p<0.10; different in AP treated animals compared with saline treated animals that received the same
infusion, Mann-Whitney U test.
Granulocytes: His48+ granulocytes were found in the glomeruli and interstitially and this
was not influenced by ATP infusion or AP treatment. Glomerular His48+ granulocyte numbers
were higher in ATP infused animals compared with saline infused rats (p<0.05; Figure 5E).
AP treatment tended to increase the number of glomerular His48+ granulocytes in saline-
154
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Figure 6. Foetal and placental weight. Average foetal (A) and placental (B) weight per nest in ATP (black bars) or saline
infused (white bars) and AP (dashed bars) or saline treated (open bars) pregnant rats.
infused control animals, but not in ATP infused animals (p<0.10; Figure 5E).
T lymphocytes: CD3+ T lymphocytes were mainly found within the parietal epithelial layer
and sometimes in the tubulointerstitium. ATP infusion or AP treatment did not affect the site
of these CD3+ cells. CD3+ glomerular T lymphocyte numbers were not different after either
ATP infusion or AP treatment (Figure 5F).
No changes in foetal or placental weight by either ATP infusion or AP treatment
ATP infusion did not induce changes in the average pup weight on day 20 of pregnancy
(Figure 6A). In addition, AP treatment had no effect on foetal weight, both in saline or in
ATP infused animals (Figure 6A). Placental weight was unaffected by both ATP infusion and/
or AP treatment (Figure 6B). Moreover, three of the ATP infused and saline treated animals
and two of the saline infused and AP treated animals displayed single resorptions, while no
resorptions were observed in saline infused or ATP infused and AP treated animals (data not
shown).
AP treatment decreased the percentage of trophoblast cells in the mesometrial triangle
Next, we checked the effect of AP treatment on trophoblast invasion in the mesometrial
triangle. We studied this on day 20 of pregnancy, since we treated the rats with AP until this
day of pregnancy, therefore, placentas with mesometrial triangle were obtained at this time
point.
On day 20 of pregnancy, trophoblast cells were present throughout the mesometrial triangle,
the total invaded area of the mesometrial triangle was similar in ATP infused compared with
saline infused rats (Figure 7A), while also no effect of AP treatment was observed (Figure 7A).
In contrast, ATP infused animals tended to show increased percentage of cytokeratin-positive
tissue as compared with saline infused controls (p<0.10; Figure 7B). In addition, AP treatment
significantly decreased the percentage of cytokeratin-positive tissue in the mesometrial
triangle exclusively in ATP infused animals (p<0.05; Figure 7B).
155
8
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Figure 7. Trophoblast invasion in the mesometrial triangle. Invading capacity of trophoblast cells was assessed by
analysing the percentage of total area of the mesometrial triangle invaded by trophoblast cells (A) and the percentage
of cytokeratin-positive tissue (B) in ATP (black bars) or saline infused (white bars) and AP (dashed bars) or saline
treated (open bars) pregnant rats. #:p<0.05; increased compared with saline infused animals that received saline
treatment, Mann-Whitney U test. a:p<0.05; decreased in AP treated animals compared with saline treated animals that
received the same infusion, Mann-Whitney U test.
Discussion
In the current study, we tested whether treatment with AP, which hydrolyses ATP and increases
Hx activity, decreases ATP-induced preeclampsia-like signs in pregnant rats. We confirmed
that ATP infusion in pregnant rats increased plasma ATP levels, decreased Hx activity,
increased albuminuria on day 17 and affected glomerular numbers of inflammatory cells.
Moreover, in the mesometrial triangle, ATP infusion increased the amount of cytokeratinpositive tissue on day 20 of pregnancy. Although AP treatment of ATP infused pregnant rats
reduced plasma ATP levels and affected glomerular inflammation, Hx activity and urinary
albumin excretion were unchanged after AP treatment. The amount of trophoblast cells in the
mesometrial triangle on day 20 was decreased by AP treatment. The limited effect of AP in
ATP infused rats could be due to the lack of increase in total AP activity in AP treated animals,
or the fact that AP did not decrease ATP levels to values seen in normal pregnancy.
ATP infusion in pregnant rats increased plasma ATP levels on day 15 and 17, and decreased
plasma Hx activity on day 20 of pregnancy. This is in accordance with previous studies,
however, in the present study, the increase in ATP levels appeared earlier in pregnancy [11].
As ATP is hydrolysed by extracellular enzymes like CD39 and AP in the circulation within
seconds [19], it seems unlikely that the increased ATP levels at days 17 and 20 are due to the
ATP infused on day 14. Sources of ATP release at days 17 and 20 may be activated immune
cells [4] or the placenta, which has been shown to be hypoxic after ATP treatment [11]. The
decreased Hx activity in these ATP treated animals is most likely the result of the increased
ATP levels, since ATP is the natural inhibitor of Hx [6].
156
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
Surprisingly, in AP treated animals, we observed no changes in total plasma AP activity
compared with saline treated rats at the end of pregnancy. Since we showed that AP treatment
decreased renal AP activity, we suggest that, as a compensatory mechanism, endogenous AP
production was downregulated. This assumption is supported by the absence of placental
plasma AP activity after AP treatment. How AP activity is regulated is unknown, but it has been
found that ATP stimulation of osteogenic ligament cells induces AP production [20]. Therefore
decreased ATP levels (due to AP treatment) probably downregulate AP activity. Alternatively,
increased plasma levels of exogenous AP may have downregulated endogenous AP activity.
In addition, inorganic phosphate, formed during dephosphorylation, is an inhibitor of
enzyme activity [21], suggesting that lower AP activity is due to decreased enzyme activity.
Unfortunately, we were only able to measure AP activity at day 20 of pregnancy. It might well
be that AP activity is increased at earlier time points. This suggestion is in line with the fact
that AP treatment decreased ATP levels.
Despite the fact that ATP levels were decreased following AP treatment both at day 15 and
day 17, we did not observe an effect of AP treatment on urinary albumin excretion on day 17.
It seems that the albuminuria is induced by a process which is not inhibited by AP treatment.
Since we started AP treatment, by i.p. AP injections, 30 minutes after the start of the infusion,
the infused ATP is not hydrolysed in the first 30 minutes of the infusion. The infused ATP
may have induced a process (directly by the infused ATP or indirectly via other mechanisms)
leading to kidney damage and albuminuria in this half hour. This process can apparently not
by inhibited by AP treatment. This has to be tested in future studies by giving AP before ATP
infusion, so that ATP levels do not increase during the infusion.
In line with previous data ([11] and Chapter 4) ATP infusion increased glomerular total
macrophages and granulocytes, and decreased iNOS+ macrophages, on day 20 of pregnancy.
Infiltration of these cells after ATP infusion is most likely due to secondary increased ATP
levels from day 17 onwards, since we have previously shown no effect of ATP infusion into
pregnant rats on glomerular numbers of inflammatory cells on day 15 and 17 of pregnancy
(Chapter 4). Although AP treatment did not affect the total numbers of macrophages, it
decreased CD206+ macrophages and increased iNOS+ macrophages on day 20 of pregnancy.
iNOS and CD206 were chosen as markers for M1 and M2 macrophages respectively [22].
Although the exact role of M1 and M2 macrophages in the kidney during pregnancy remains
to be studied, it has however been shown that iNOS+ macrophages predominate in the first
48 hours after renal ischemia/reperfusion injury, while CD206+ macrophages prevail at
later stages, in which they may induce tissue repair [23]. Therefore, the relative increased
CD206+ macrophage numbers after ATP infusion in the kidney may be a sign of repair of
renal damage. This suggestion may be in line with previously observed decrease of urinary
albumin excretion on day 20 as compared with day 17 after ATP infusion [11]. AP treatment,
157
8
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
however, decreased CD206+ macrophages and increased iNOS+ macrophages. This may be
a direct effect of AP, since similar trends were also observed in saline infused pregnant rats.
AP treatment had no effect on the average pup and placental weight, neither in saline, nor
in ATP treated animals. This indicates that there are no adverse effects of AP treatment
on pregnancy outcome. We found that ATP infusion increased the number of trophoblast
cells in the mesometrial triangle, which was prevented by AP treatment. In the mesometrial
triangle, from day 17 onwards trophoblast cell numbers decrease, likely due to cell death
and regression of the mesometrial triangle itself as delivery approaches [13]. We previously
showed that ATP inhibits trophoblast invasion on day 17 (Chapter 5). This suggested that
trophoblast invasion may have been delayed by ATP infusion. As a result of this delayed
trophoblast invasion, the trophoblast cell death at the end of pregnancy may also be
delayed, resulting in increased numbers of trophoblast cells in the mesometrial triangle at
day 20 of pregnancy. As AP treatment of ATP infused rats decreased trophoblast numbers
in the mesometrial triangle at day 20, it may be suggested that AP treatment inhibited the
decreased trophoblast invasion at day 17, resulting in decreased numbers of trophoblast
cells at day 20. This suggestion needs to be verified in future studies.
In the current study we showed that AP treatment reduced several ATP-induced effects: it
decreased ATP levels, decreased glomerular CD206+ macrophages and increased glomerular
iNOS+ macrophages and the M1/M2 ratio. It also affected trophoblast cell numbers in the
mesometrial triangle at the end of pregnancy. However, AP treatment in the current dose,
was unable to increase Hx activity. This may be due to the still slightly increased ATP levels
after AP treatment. AP treatment was also not able to decrease urinary albumin excretion.
Therefore, treatment with AP could possibly be further optimized to increase plasma AP
activity, for instance with higher doses or more frequent injections. Moreover, in further
experiments with AP treatment as potential therapeutic for preeclampsia, the effect of AP
treatment on earlier stages of placental development, i.e. at days 15 and 17, should be
assessed. However, in this primary study, AP treatment may be a promising therapeutic
against ATP-induced preeclampsia symptoms.
References
[1] Duley L. The global impact of pre-eclampsia and eclampsia. Semin Perinatol 2009; 33:130-137.
[2] Redman CW, Sargent IL. Placental stress and pre-eclampsia: a revised view. Placenta 2009; 30 Suppl A:S38-42.
[3] Roberts JM, Hubel CA. The two stage model of preeclampsia: variations on the theme. Placenta 2009; 30 Suppl
A:S32-37.
[4] Spaans F, Vos PD, Bakker WW, van Goor H, Faas MM. Danger Signals From ATP and Adenosine in Pregnancy and
Preeclampsia. Hypertension 2014; 63:1154-1160.
[5] Tolosano E, Fagoonee S, Morello N, Vinchi F, Fiorito V. Heme scavenging and the other facets of hemopexin. Antioxid
Redox Signal 2010; 12:305-320.
158
CHAPTER 8 | HEMOPEXIN ACTIVITY AND EXTRACELLULAR ATP IN THE PATHOGENESIS OF PREECLAMPSIA | FLOOR SPAANS
[6] Bakker WW, Borghuis T, Harmsen MC, van den Berg A, Kema IP, Niezen KE, Kapojos JJ. Protease activity of plasma
hemopexin. Kidney Int 2005; 68:603-610.
[7] Bakker WW, Donker RB, Timmer A, van Pampus MG, van Son WJ, Aarnoudse JG, van Goor H, Niezen-Koning KE, Navis
G, Borghuis T, Jongman RM, Faas MM. Plasma hemopexin activity in pregnancy and preeclampsia. Hypertens Pregnancy
2007; 26:227-239.
[8] Bakker WW, Henning RH, van Son WJ, van Pampus MG, Aarnoudse JG, Niezen-Koning KE, Borghuis T, Jongman RM,
van Goor H, Poelstra K, Navis G, Faas MM. Vascular contraction and preeclampsia: downregulation of the Angiotensin
receptor 1 by hemopexin in vitro. Hypertension 2009; 53:959-964.
[9] Bakker WW, Spaans F, El Bakkali L, Borghuis T, van Goor H, van Dijk E, Buijtink J, Faas MM. Plasma Hemopexin as a
Potential Regulator of Vascular Responsiveness to Angiotensin II. Reprod Sci 2012; 20:234-237.
[10] Gant NF, Daley GL, Chand S, Whalley PJ, MacDonald PC. A study of angiotensin II pressor response throughout
primigravid pregnancy. J Clin Invest 1973; 52:2682-2689.
[11] Faas MM, van der Schaaf G, Borghuis T, Jongman RM, van Pampus MG, de Vos P, van Goor H, Bakker WW. Extracellular
ATP induces albuminuria in pregnant rats. Nephrol Dial Transplant 2010; 25:2468-2478.
[12] Faas MM, Broekema M, Moes H, van der Schaaf G, Heineman MJ, de Vos P. Altered monocyte function in experimental
preeclampsia in the rat. Am J Obstet Gynecol 2004; 191:1192-1198.
[13] Vercruysse L, Caluwaerts S, Luyten C, Pijnenborg R. Interstitial trophoblast invasion in the decidua and mesometrial
triangle during the last third of pregnancy in the rat. Placenta 2006; 27:22-33.
[14] Cheung PK, Stulp B, Immenschuh S, Borghuis T, Baller JF, Bakker WW. Is 100KF an isoform of hemopexin?
Immunochemical characterization of the vasoactive plasma factor 100KF. J Am Soc Nephrol 1999; 10:1700-1708.
[15] Fishman WH. Alkaline phosphatase isozymes: recent progress. Clin Biochem 1990; 23:99-104.
[16] Aleem FA. Biochemical studies of the placental alkaline phosphatase. Clin Chim Acta 1971; 33:125-134.
[17] van Goor H, Gerrits PO, Hardonk MJ. Enzyme histochemical demonstration of alkaline phosphatase activity in
plastic-embedded tissues using a Gomori-based cerium-DAB technique. J Histochem Cytochem 1989; 37:399-403.
[18] Mills DC, Robb IA, Roberts GC. The release of nucleotides, 5-hydroxytryptamine and enzymes from human blood
platelets during aggregation. J Physiol 1968; 195:715-729.
[19] Fitz JG. Regulation of cellular ATP release. Trans Am Clin Climatol Assoc 2007; 118:199-208.
[20] Sawada T, Kishiya M, Kanemaru K, Seya K, Yokoyama T, Ueyama K, Motomura S, Toh S, Furukawa K. Possible role of
extracellular nucleotides in ectopic ossification of human spinal ligaments. J Pharmacol Sci 2008; 106:152-161.
[21] Borosky GL, Lin S. Computational modeling of the catalytic mechanism of human placental alkaline phosphatase
(PLAP). J Chem Inf Model 2011; 51:2538-2548.
[22] van Putten SM, Ploeger DT, Popa ER, Bank RA. Macrophage phenotypes in the collagen-induced foreign body
reaction in rats. Acta Biomater 2013; 9:6502-6510.
[23] Lee S, Huen S, Nishio H, Nishio S, Lee HK, Choi BS, Ruhrberg C, Cantley LG. Distinct macrophage phenotypes
contribute to kidney injury and repair. J Am Soc Nephrol 2011; 22:317-326.
159
8
Chapt