- Email: [email protected]
Endogenous carbon monoxide production by menadione
Accepted Manuscript Endogenous carbon monoxide production by menadione Chioma U. Odozor, Nichole Peterson, Jessica Pudwell, Graeme N. Smith PII:
S0143-4004(18)30239-X
DOI:
10.1016/j.placenta.2018.09.007
Reference:
YPLAC 3879
To appear in:
Placenta
Received Date: 31 May 2018 Revised Date:
20 September 2018
Accepted Date: 25 September 2018
Please cite this article as: Odozor CU, Peterson N, Pudwell J, Smith GN, Endogenous carbon monoxide production by menadione, Placenta (2018), doi: https://doi.org/10.1016/j.placenta.2018.09.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: Endogenous carbon monoxide production by menadione
2
Author names and affiliations: Chioma U. Odozora, Nichole Petersona, Jessica Pudwellb, and Graeme
3
N. Smitha,b
4
a. Department of Biomedical and Molecular Sciences, Queen’s University, Kingston, Ontario, Canada,
5
K7L 3N6
6
b. Department of Obstetrics and Gynaecology, Queen’s University, Kingston, Ontario, Canada, K7L
7
3N6
8
Authors’ email addresses:
9
C.U. Odozor
[email protected]
10
N. Peterson
[email protected]
11
J. Pudwell
[email protected]
12
G.N. Smith
[email protected]
13
Corresponding Author:
14
Graeme N Smith MD, PhD, FRCSC
15
Professor and Head, Department of Obstetrics and Gynecology
16
Professor, Biomedical and Molecular Sciences and Imaging Services
17
Director, Clinician Investigator Program
18
Queen’s University
19
76 Stuart Street
20
Kingston, Ontario
21
K7L 2V7
22
Telephone: (613) 548-1372
M AN U
TE D
EP
AC C
SC
RI PT
1
ACCEPTED MANUSCRIPT
Fax: (613) 538-1330
24
Email: [email protected]
25
ABSTRACT
26
Introduction: Smokers have a significantly decreased risk of pre-eclampsia (PE), possibly attributed to
27
an increase in blood carbon monoxide (CO) concentrations. At physiological concentrations, CO has
28
been demonstrated to have placental vasodilatory and anti-inflammatory properties. Increasing
29
endogenous CO production may have therapeutic potential to either prevent or treat PE. Menadione
30
(MD), synthetic vitamin K3, increases CO in rat microsomes. Our objective was to investigate MD’s
31
ability to increase endogenous CO concentrations in pregnancy.
32
Methods: Three experiments were completed. First, in vitro CO production was measured using
33
isolated GD15 placentas. Second, non-pregnant normotensive mice received no, 1.5, 4.0 or 6.5 g/L MD
34
for 7 days. Lastly, pregnant normotensive mice received either no or 6.5 g/L MD in water from GD10.5
35
to GD17.5. Consumption was measured as average daily water intake per gram of body weight.
36
Maternal and fetal CO levels in the blood and tissue were quantified using headspace gas
37
chromatography.
38
Results: MD significantly increased CO production in isolated GD15 placentas. In both pregnant and
39
non-pregnant experiments, splenic CO, hepatic CO, and splenic mass were higher in treated mice
40
compared to controls (all p<0.05). Maternal %COHb and Hb in treated dams were not significantly
41
different compared to controls. The fetal:placental mass ratio was significantly lower in the treatment
42
group (p=0.002).
43
Discussion: Placental CO production was observed in GD15 placentas after co-incubation with MD.
44
MD administration increased CO in the liver and spleens of pregnant mice. Further investigation into
45
different doses of MD is required to identify one without demonstrable fetal/placental effects.
AC C
EP
TE D
M AN U
SC
RI PT
23
ACCEPTED MANUSCRIPT
KEYWORDS: Menadione; pregnancy; carbon monoxide; mouse placenta; pre-eclampsia
47
ABBREVIATIONS: AdsFlt-1: adenovirus soluble fms-like tyrosine kinase-1; CO: carbon monoxide;
48
COHb: carboxyhemoglobin; CPR: cytochrome p450 reductase; EGD: early gestational demise; LGD:
49
late gestational death; Hb: hemoglobin; HO: heme oxygenase; MD: menadione; NQO1:
50
NAD(P)H:quinone oxidoreductase 1; PE: pre-eclampsia; ROS: reactive oxygen species; SSA:
51
sulfosalicylic acid
52
INTRODUCTION:
53
Pre-eclampsia (PE) remains a leading cause of maternal-fetal morbidity and mortality worldwide.
54
Characterized by de novo hypertension and proteinuria, PE affects approximately 5-8% of pregnancies
55
[1]. Multiple strategies are being explored as potential targets for the treatment or prevention of PE. The
56
primary focus of investigation has been directed at the roles of immunology and maternal endothelial
57
dysfunction in the disease’s pathogenesis [2].
58
Smoking decreases the risk of developing PE by up to 33% [2–5]. This is not seen in smokeless tobacco
59
(snuff) users [3], suggesting that the presence of carbon monoxide (CO), a major combustion product,
60
may be responsible for the reduced risk [2,6]. Despite the recognition that CO is toxic at high
61
concentrations, numerous studies have shown its beneficial effects in many disease states at low
62
concentrations. Our focus is on the use of CO at therapeutic dosages in the treatment of PE. CO is a
63
gasotransmitter produced mainly from heme degradation by the enzyme heme oxygenase (HO). In
64
pregnancy, the rate of endogenous CO production increases by 30-40% due to increased heme
65
metabolism [7]. However, women with PE have significantly lower end-tidal breath CO levels than
66
healthy pregnant women, suggesting that decreased CO may be one of the contributors to PE
67
development [8].
AC C
EP
TE D
M AN U
SC
RI PT
46
ACCEPTED MANUSCRIPT
We previously demonstrated that the chronic exposure of 250 ppm CO to pregnant CD-1 mice increased
69
placental vessel branching, arterial size and placental perfusion [9,10], without demonstrable adverse
70
maternal or fetal effects [11]. Further, we demonstrated that in the adenovirus soluble fms-like tyrosine
71
kinase-1 (AdsFLT-1) mouse model of PE, the same dose normalized maternal blood pressure and renal
72
function compared to untreated mice [12]. Thus, the promotion of angiogenesis and vasodilation by
73
increased CO via alternative delivery methods may be beneficial in PE.
74
Menadione (MD), vitamin K3, is a synthetic naphthoquinone that is capable of either one or two
75
electron transfers, catalyzed by cytochrome p450 reductase (CPR) or NAD(P)H:quinone oxidoreductase
76
1 (NQO1), respectively [13]. Vukomanovic et al. (2014) demonstrated that MD activates HO-2 to
77
produce CO, resulting in a seven-fold increase in enzymatic activity [14]. Using human recombinant
78
enzymes and rat brain microsomes, they reported that HO’s coenzyme CPR increased MD-mediated CO
79
production, independent of HO [15]. In the present study, the ability of MD to increase CO production
80
in harvested placentas and pregnant mice was investigated. To provide a basis for the use of MD in
81
animal models of PE, we hypothesized that MD would increase CO production when incubated with
82
mouse placentas, and that its administration in vivo would increase endogenous CO production in female
83
mice.
84
METHODS
85
Animals
86
All experimental procedures were approved by the Queen’s University Animal Care Committee
87
(Protocol 2016-1691) in accordance with Canadian Council for Animal Care guidelines. All mice were
88
kept on a 12-hour light/dark cycle and received Purina LabDiet 5015 pellets and water ad libitum.
89
a) Incubation Study – Pregnant Mice with no Menadione Exposure
AC C
EP
TE D
M AN U
SC
RI PT
68
ACCEPTED MANUSCRIPT
Female, timed-pregnant gestational day (GD) 13 CD-1 mice (Charles River, USA) were housed until
91
GD15. GD1 corresponded with the day that a copulatory plug was observed in the vagina, as timed by
92
Charles River Laboratories.
93
b) Dosing Study - Non-Pregnant Mice
94
Virgin female CD-1 mice aged 4-8 weeks (Charles River, USA) were used in the non-pregnant mouse
95
experiments to first confirm the safety of daily oral MD administration. Non-pregnant mice were used
96
for the dose-finding study first to avoid undue harm.
97
c) Dosing Study - Pregnant Mice
98
Female CD-1 mice aged 4-8 weeks (Charles River, USA) were bred with adult male CD-1 mice
99
(Charles River, USA) for use in the pregnancy studies. GD0.5 corresponded with the day that a
M AN U
SC
RI PT
90
copulatory plug was observed in the vagina.
101
Menadione Preparation
102
i) Incubation Study
103
Menadione sodium bisulfite (Sigma Aldrich Inc., USA) was prepared by serial dilution from a 50 mM
104
stock solution dissolved in double distilled water (ddH2O) and stored at 4°C. Menadione preparations
105
were made in darkened conditions to avoid photodegradation.
106
ii) Dosing Studies
107
Menadione solutions were supplied to mice in 50 mL amber conical tubes (VWR International, USA).
108
Non-pregnant and pregnant mice were allocated to either the control or treatment groups. The control
109
group received regular drinking water. The treatment groups received drinking water for the first four
110
days before MD was added. Total water intake per cage and mouse weights were recorded to determine
111
the average daily water intake per gram of body weight in each cage. Each cage housed a maximum of 4
112
mice. Water volume was recorded and changed every 24 hours.
AC C
EP
TE D
100
ACCEPTED MANUSCRIPT
a) Non-Pregnant Mice
114
Non-pregnant mice in the treatment group received tap water from days 1 to 4, followed by 1.5 g/L, 4
115
g/L or 6.5 g/L of MD in drinking water from days 5 to 12.
116
b) Pregnant Mice
117
The experimental period for the pregnant mice started on GD6.5 to correspond with the same twelve-day
118
timeline for the non-pregnant mice experiment. Pregnant mice in the treatment group received tap water
119
from GD6.5 to GD9.5, followed by 6.5 g/L of MD in drinking water from GD10.5 to GD17.5.
120
Determination of %COHb and Hb in Dosing Studies
121
a) Non-Pregnant Mice
122
Blood collection occurred on days 1, 6, and 9 in non-pregnant mice. 50 uL of blood was collected from
123
the submandibular vein into 1.5 mL polypropylene microcentrifuge tubes containing 8 uL of EDTA. 300
124
uL of saline (0.9% NaCl) was given i.p. following each collection. 100 – 500 uL of blood was collected
125
from the retro-orbital sinus on Day 12 to allow for larger blood volume collection, as permitted by the
126
ethics committee as a terminal procedure.
127
b) Pregnant Mice
128
Blood collection occurred on GD6.5, 11.5, and 14.5 in pregnant mice from the submandibular vein,
129
followed by a saline injection as detailed above. 100 – 500 uL of blood was collected from the retro-
130
orbital sinus on GD17.5.
131
Hemoglobin (Hb) and %COHb levels in all mice were measured as previously described [11]. Briefly,
132
Hb was measured in duplicate using the HemoCue Hb 201+ System (Radiometer, Sweden) following
133
blood collection. A Hamilton repeating dispenser attached to a 10 µL airtight Hamilton syringe
134
(Hamilton, USA) was used to dispense 1 µL of blood into each GC vial containing 20 µL of 2%
135
sulfosalicylic acid (SSA). Vials for %COHb quantification were made in triplicate, with another vial
AC C
EP
TE D
M AN U
SC
RI PT
113
ACCEPTED MANUSCRIPT
containing only 2% SSA as a blank. Following 60 minutes of incubation on ice, CO levels were
137
determined using gas chromatography (Peak Performer 1 Analyzer, Peak Laboratories, Mountain View,
138
CA).
139
Animal Sacrifice and Perfusion
140
i) Incubation Study
141
Mice were anesthetized on GD15 using sodium pentobarbital (110 mg/kg, intraperitoneally) (Ceva Santé
142
Animale, France). GD15 corresponds with the time at which hypertension was observed in AdsFLT-1
143
mice [12]. Mice were systemically perfused with Krebs’ buffer (pH = 7.40) by inserting a 21-gauge
144
butterfly needle (#367344, BD Vacutainer, USA) into the left ventricle and piercing the right atrium to
145
allow for blood egress until the liver was visibly blanched. Perfusion was the final assurance of death.
146
1 L of Krebs’ buffer was prepared by dissolving final concentrations of the following solutions in
147
ddH2O: 120 mM NaCl (Bioshop Canada Inc., Burlington), 5.6 mM KCl (Fisher Scientific, USA), 1.2
148
mM MgSO47H2O (BDH Inc, Toronto), 1.2 mM NaH2PO4H2O (Fisher Scientific, USA), 2.5 mM
149
CaCl22H2O (Sigma Aldrich Inc., USA), 11 mM D-glucose (BDH Inc, Toronto), and 25 mM NaHCO3
150
(Sigma Aldrich, Inc., USA). The solution was bubbled with 5% CO2/balance O2 (Praxair Air, Kingston)
151
for 10 minutes and pH was recorded using a benchtop pH metre (Accumet AB15, Fisher Scientific,
152
USA).
153
ii) Dosing Studies
154
Animals were sacrificed (as detailed above) at the end of the twelve-day experimental period,
155
corresponding to day 12 in non-pregnant mice and GD17.5 in pregnant mice. Gravity perfusion with
156
Krebs' buffer (pH = 7.40) was performed prior to tissue CO determination. In perfused pregnant mice,
157
the uterine horns were excised, and viable fetuses and placentas were counted and weighed. Fetuses
158
from unperfused mice were decapitated to collect blood with pipettes containing 5 uL of EDTA. Early
AC C
EP
TE D
M AN U
SC
RI PT
136
ACCEPTED MANUSCRIPT
gestational demises (EGDs) and late gestational deaths (LGDs) were also noted, as previously described
160
[11].
161
Menadione Incubation with Whole Placentas
162
For the incubation study, GD15 placentas, unexposed to MD prior to harvest, were harvested from the
163
uterine horns, weighed and immediately placed on ice. 2 mL amber gas chromatography (GC) vials
164
(Chromatographic Specialties, Inc.) with open-top caps were sealed with 8 mm blue chromatherm septa
165
(Chromatographic Specialties Inc., C13302). For each sample vial, one weighed halved or whole
166
placenta (~0.01 – 0.1 g) was added to 90 uL of buffer and 10 uL of 0, 0.01, 0.05, 0.1, 0.5 or 1 mM MD
167
for initial dosing studies and 0, 1, 5, 10 or 50 mM MD for extended dosing studies. At minimum,
168
samples were made in duplicates. Triplicate vials containing either 100 uL of buffer alone, or in
169
combination with one placenta each, were made to establish vehicle and tissue blanks, respectively.
170
Menadione Incubation with Sonicated Placentas
171
Three weighed GD15 placentas, unexposed to MD prior to harvest, were randomly selected from each
172
litter, homogenized with a glass tissue grinder in 4 equivalent volumes of ice-cold ddH2O and sonicated
173
on ice at 3 watts for 5 seconds using the Sonic Dismembrator Model 100 (Fisher Scientific, USA). GC
174
vials containing 5 – 10 ul of 20% w/v tissue sonicate were prepared in triplicate, equivalent to 1 – 2 mg
175
of tissue per vial. All vials were made up to 100 uL total volume using Krebs’ buffer. Tissue blanks and
176
sample vials contained sonicated tissue, while the latter also contained 10 uL of either 0.01, 0.05, 0.1,
177
0.5, 1, 5, 10 or 50 mM MD as appropriate for the given study.
178
Tissue CO Determination
179
i) Incubation Study
AC C
EP
TE D
M AN U
SC
RI PT
159
ACCEPTED MANUSCRIPT
Vials were incubated for 30 minutes in a 37° C water bath, then placed on dry ice to stop any chemical
181
reactions. CO content was analyzed using headspace GC as previously described [16], and expressed as
182
pmol CO/g tissue.
183
ii) Dosing Studies
184
Perfused organs were harvested and immediately placed on ice. For all comparisons, the organ mass was
185
normalized to body weight. Livers, spleens and right kidneys were collected from non-pregnant mice.
186
The same organs were harvested from pregnant mice, in addition to three pooled placentas. 80 to 2000
187
mg of each tissue were homogenized and sonicated as detailed above. 10 uL of each 20% w/v tissue
188
sonicate were prepared in triplicates, equivalent to 2 mg of fresh tissue per vial. Each vial contained 5
189
uL of 30% SSA, with sample vials containing 10 uL of sonicated tissue. All vials were made up to 40 uL
190
total volume using ddH2O. CO content in the sonicates was then quantified after 1 hour of incubation in
191
ice using headspace gas chromatography, and expressed as pmol CO/mg tissue as described by Vreman
192
et al. (2005) [16].
193
Statistical Analysis
194
Data are presented as means ± standard deviation (SD) and analyzed using SPSS v24 (SPSS Inc,
195
Chicago, IL, USA) and GraphPad Prism 6. A p value of <0.05 was considered statistically significant for
196
all analyses.
197
i) Incubation Study
198
CO production in whole and sonicated placentas was modelled using a linear mixed effects model to
199
control for mouse of origin (random effect) for the placental sample and dose (Y= Mouse + Dose +
200
Error). Four models were generated; (1) sonicated placenta initial dosing, (2) whole placenta initial
201
dosing, (3) sonicated placenta extended dosing, and (4) whole placenta extended dosing.
202
ii) Dosing Studies
AC C
EP
TE D
M AN U
SC
RI PT
180
ACCEPTED MANUSCRIPT
Linear regression analyses with generalized estimating equations were used to report parameter
204
estimates ± standard error per unit change in dose in both non-pregnant and pregnant mice, taking into
205
account within-cage variation, as mice were housed together during the study. Each mouse was
206
considered an experimental unit, while the cage and/or day (as appropriate) were the within-subject
207
factors. A generalized estimating equation with binomial distribution was used to compare the
208
proportion of live pups to EGDs and LGDs between the two groups. Fetal and placental data were
209
analyzed as per-litter average values. The main outcome parameters were tissue CO, organ mass,
210
fetal:placental mass and change in maternal weight, %COHb and Hb. Cage exposure to MD and the
211
treatment day were covariates in blood analyses. The number of mice in a cage was a covariate for
212
fetal:placental mass comparisons. The number of implantation sites and the number of mice per cage
213
were covariates for maternal weight gain and litter survival comparisons. Repeated measures ANOVAs
214
were used to compare average daily water intake between the control and treatment groups in both non-
215
pregnant and pregnant groups. The analysis was completed at the cage level, with average water intake
216
calculated based on total water intake divided by number of mice in the cage. Non-dosing days and
217
dosing days were analysed separately.
218
RESULTS
219
1. Carbon monoxide production in isolated GD15 mouse placentas
220
CO production for the initial concentration-response studies in sonicated placentas (n=4) and intact
221
placentas (n=8) can be seen in Figure 1A. CO production for the extended concentration-response
222
studies in sonicated placentas (n=4) and intact placentas (n=4) can be seen in Figure 1B. In both studies,
223
increasing doses of MD resulted in a concentration-dependent increase in CO production, with higher
224
CO values produced using sonicated tissue. Parameter estimates ± standard error for the dose variable in
225
each of the 4 models were 30065±4466, 694±100, 2919±740, and 31±14 pmol CO/g tissue, respectively.
AC C
EP
TE D
M AN U
SC
RI PT
203
ACCEPTED MANUSCRIPT
2. Dose exploration in non-pregnant mice
227
Shown in Supplemental Figure S1-A, both hepatic and splenic mass (n=4-5 mice/dose/organ) increased
228
with MD dose in non-pregnant mice (kidney mass: +0.180 ± 0.083 mg/g body weight, p=0.03; spleen
229
mass: +0.364 ± 0.128 mg/g body weight, p=0.004). As seen in Supplemental Figure S1-B, spleen CO
230
also increased with MD dose (+3.278 ± 0.635 pmol CO/ g tissue, p<0.001). No significant changes were
231
observed for kidney CO, liver CO, and liver mass, respectively.
232
Increasing cage MD exposure had a positive effect on %COHb (+0.014 ± 0.0064 %COHb, p=0.03). No
233
significant difference in %COHb was observed when comparing days of collection (data not shown). In
234
contrast, no significant difference in Hb was observed with increasing MD dose, while the day of blood
235
collection had a negative effect (-0.890 ± 0.313 g/L; p=0.004). In total 3 cages received no MD and 2
236
cages each received, 1.5, 4.0, or 6.5 g/L MD. No differences in average water intake per cage were
237
observed between dose groups during either the non-dosing period (days 1-4) or dosing period (days 5-
238
12).
239
3. The effect of high dose MD on maternal weight gain and daily water intake
240
High dose MD for seven days did not significantly affect daily water consumption in pregnant, treated
241
dams (n= 7 cages) compared to control dams (n=6 cages) (p=0.05). Dams in the MD treatment group
242
(n=14 mice) tended to gain less weight from GD0.5 to GD17.5 than dams in the control group (n=13
243
mice) (-0.748 ± 0.285; p=0.009).
244
4. Maternal – fetal outcomes after high dose MD administration
245
Carbon monoxide production was observed in all tissue samples from treated mice (n= 9 mice, 5 cages)
246
(Fig. 2A), with significantly higher tissue CO in the spleen (+2.354 ± 0.432 pmol CO/g tissue, p<0.001)
247
and liver (+0.236 ± 0.0995 pmol CO/g tissue, p=0.017) compared to the control (n=8 mice, 5 cages).
248
Splenomegaly was observed in spleens of most treated mice (+0.325 ± 0.113 mg/g body weight,
AC C
EP
TE D
M AN U
SC
RI PT
226
ACCEPTED MANUSCRIPT
p=0.004), however placental and renal CO, and renal and hepatic mass did not differ between treatment
250
and controls (Fig. 2B). Maternal %COHb and Hb were not significantly increased by MD exposure.
251
However, in both groups Hb decreased (p<0.001) and %COHb increased (p<0.001) at later collection
252
days.
253
The numbers of EGDs and LGDs were not significantly different between treated (n=17 litters) and
254
control (n=16 litters) dams. There was a non-significant increase in fetal %COHb and Hb in treated mice
255
(Fig. 3A, 3B). Fetal mass was lower in treated mice (p<0.001) resulting in a lower ratio of fetal:placental
256
mass (10.22 ± 1.13, n=16 litters vs. 8.80 ± 1.45, n=17 litters, p=0.002) (Fig. 3C,3D). However, no
257
difference in placental mass was observed (Fig. 3E).
258
DISCUSSION
SC
M AN U
259
RI PT
249
The results presented in this paper suggest that in vivo MD increases CO production in maternal tissues, particularly in the liver and spleen, but not in the kidney, placenta or circulation. However, in
261
vitro exposure to MD increased CO production by placental tissue in our study. It has been shown that
262
endogenous CO production occurs in human chorionic villi [17], and that MD undergoes redox cycling
263
in human placental microsomes [18]. This is the first study to demonstrate increased CO production by
264
MD in vitro in placental tissue. We demonstrated that MD increases CO production in isolated GD15
265
mouse placentas at 50 mM MD. This concentration is equivalent to approximately 13.8 g/L MD in vivo.
266
We expected that sonication would result in a higher percent increase in CO production at increasing
267
concentrations of MD, due to the greater surface area and therefore increased availability of heme to
268
interact as a substrate. Sonication prior to CO quantification produced higher levels of CO compared to
269
MD incubation with whole placentas. However, we were unable to report a statistically significant
270
increase in placental CO production in our in vivo studies, despite observing a trend towards increased
271
production.
AC C
EP
TE D
260
ACCEPTED MANUSCRIPT
272
High doses of MD were used to determine possible adverse effects in dams and fetuses. Given that higher %COHb levels were reported with inhalational CO without adverse maternal and/or fetal
274
effects [11] and that we did not observe significant changes in %COHb attributable to MD dose, the
275
adverse maternal and fetal effects seen in this study are not likely due to a %COHb change but rather a
276
direct effect of the MD. While we did not observe a statistically significant difference in water intake,
277
there did appear to be a trend towards decreased intake in the treatment group during the dosing days
278
that was not observed on the non-dosing days. This was likely a result of taste aversion to high dose
279
MD, possibly leading to the decrease in maternal weight gain in treated mice. Normally, maternal
280
weight gain during gestation increases as the fetus develops [19]. However, treated mice did not
281
demonstrate the same growth compared to the controls. As fetal mass was significantly different
282
between the two experimental groups, it is also possible that decreases in fetal mass contributed to the
283
lower maternal weight gain in treated mice.
M AN U
SC
RI PT
273
The fetal:placental mass ratio is often used as a proxy measure for placental efficiency,
285
indicative of the placenta’s ability to maximize nutrient delivery to the developing fetus. Low placental
286
efficiency may be caused by low fetal weight, high placental weight or a combination of both [20]. The
287
former is true in this case, as only fetal mass was significantly lower on average in treated mice.
288
However, absence of a change in placental mass does not disprove that there may be structural or
289
functional differences between the placentas of control and treated mice. Lo et al. (2017) looked at
290
fetotoxic effects and placental function after exposure to high doses of alcohol [21]. While fetal mass
291
was significantly lower in the alcohol-exposed group, there was no significant difference in placental
292
mass. However, placental perfusion and oxygenation were reduced. Moreover, the slight improvement
293
in placental function later in gestation was speculated to be compensatory but unable to maintain normal
294
fetal growth, causing the decrease in fetal mass. It could be that in our study, high dose MD led to early
AC C
EP
TE D
284
ACCEPTED MANUSCRIPT
changes in placental function, which may have had an impact on fetal mass and survival. However, a
296
compensation for these changes may have led to increases in placental size. Histological analyses of the
297
collected tissues would lend some insight on whether high dose MD affects normal placental
298
development and/or the degree of hypoxia.
299
RI PT
295
Similar to the %COHb levels observed in smokers, we hypothesized that high dose MD could significantly increase %COHb to 5-10%, and allow for vasodilation and increased vessel branching as
301
an adaptive response to increased CO production [9]. However, %COHb never rose above 1% in our
302
study. Differences in CO delivery, as well as the location of MD’s targets may account for this. In our
303
previous studies, CO was delivered exogenously, while MD-mediated CO production is an endogenous
304
process. It may be that the location of the enzymes upon which MD acts limits its ability to increase CO
305
production in the blood. Both HO and CPR are membrane-bound enzymes located on the endoplasmic
306
reticulum [22]. Erythrocytes lack membrane bound organelles, and therefore would not be expected to
307
degrade heme enzymatically without transport to the spleen. However, macrophages in the blood may
308
still be a possible source of HO and CPR [22], thus increasing CO in the blood of treated mice, albeit to
309
a lesser degree than in tissues.
M AN U
TE D
High dose MD increased CO in the tissues of treated mice, although not to the same extent as
EP
310
SC
300
with direct CO delivery [11]. As the spleen is the primary site of heme degradation, and its open
312
circulation made it difficult to completely perfuse, it was expected that its higher substrate availability
313
would allow for more CO to be released. Likewise, the liver is also a major site of heme metabolism,
314
which could have contributed to its ability to produce higher amounts of CO at 6.5 g/L MD.
315
AC C
311
In our study, we were unable to report a statistically significant difference in fetal %COHb and
316
Hb in treated mice, however we did observe a trend towards increased values. It is possible that
317
increases in %COHb may be causing some degree of hypoxia, which could stimulate compensatory
ACCEPTED MANUSCRIPT
erythropoietin production to also increase Hb levels [23]. Consequently, fetal tissues should be collected
319
to quantify CO, hypoxia and erythropoietin levels after maternal exposure to various MD doses. Given
320
that the fat-soluble vitamin K has limited transfer across the placenta, it is not expected that MD would
321
cause a significant disruption in the fetus’ normal adaptive responses to angiogenesis and alterations in
322
flow from the maternal side of the placenta.
323
RI PT
318
This study demonstrates that MD increases endogenous CO production in healthy female mice and lays a framework for future experimentation with MD and CO production in various animal models
325
of PE. The administration of 6.5 g/L MD had negative impacts on fetal survival and growth. It is
326
possible that physiological and biochemical differences between placentas from normotensive and PE
327
pregnancies may have an impact on both the extent and location of MD-mediated CO production.
328
However, as mice were normotensive, PE-relevant physiological endpoints could not be attained.
329
Therefore, further studies must be done to determine whether lower doses of MD mitigate PE-like
330
symptoms in a PE mouse model, without severe consequence to the fetus, before it can be deemed a
331
viable therapeutic option in the treatment of PE.
332
Acknowledgements
333
The authors would like to acknowledge Dr. Kanji Nakatsu and his lab for initial help with
334
troubleshooting gas chromatography experiments. This research was funded by the Canadian Institutes
335
of Health Research (CIHR) Catalyst Grant: Catalyzing Innovation in Preterm Birth Research (Grant No.
336
RN297868 - 371228).
337
Conflicts of interest: None.
338 339 340
AC C
EP
TE D
M AN U
SC
324
ACCEPTED MANUSCRIPT
REFERENCES:
342
[1] J.M. Roberts, M. Druzin, P. August, R.R. Gaiser, G. Bakris, J.P. Granger, J.R. Barton, A. Jeyabalan,
343
I. Bernstein, D.D. Johnson, S.A. Karumanchi, C.Y. Spong, M.D. Lindheiner, E. Tsingas, M.Y. Owens,
344
J.N. Martin Jr, G.R. Saade, B.M. Sibai, ACOG Guidelines: Hypertension in pregnancy, 2012.
345
[2] C.C. Venditti, G.N. Smith, Involvement of the heme oxygenase system in the development of
346
preeclampsia and as a possible therapeutic target., Women’s Health (Lond. Engl). 10 (2014) 623–43.
347
[3] L.D. Longo, The biological effects of carbon monoxide on the pregnant woman, fetus, and newborn
348
infant., Am. J. Obstet. Gynecol. 129 (1977) 69–103.
349
[4] D. Kreiser, M. Baum, D.S. Seidman, A. Fanaroff, D. Shah, I. Hendler, D.K. Stevenson, E. Schiff,
350
M.L. Druzin, End tidal carbon monoxide levels are lower in women with gestational hypertension and
351
pre-eclampsia., J. Perinatol. 24 (2004) 213–217.
352
[5] A.K. Wikström, O. Stephansson, S. Cnattingius, Tobacco use during pregnancy and preeclampsia
353
risk: effects of cigarette smoking and snuff., Hypertension. 55 (2010) 1254–1259.
354
[6] S.M. Engel, T.M. Janevic, C.R. Stein, D.A. Savitz, Maternal smoking, preeclampsia, and infant
355
health outcomes in New York City, 1995-2003, Am. J. Epidemiol. 169 (2009) 33–40.
356
[7] E.C. Miller, H. Cao, S.W. Wen, Q. Yang, J. Lafleche, M. Walker, The risk of adverse pregnancy
357
outcomes is increased in preeclamptic women who smoke compared with nonpreeclamptic women who
358
do not smoke, Am. J. Obstet. Gynecol. 203 (2010) 334.e1-334.e8.
AC C
EP
TE D
M AN U
SC
RI PT
341
ACCEPTED MANUSCRIPT
[8] S.A. Bainbridge, E.H. Sidle, G.N. Smith, Direct placental effects of cigarette smoke protect women
360
from pre-eclampsia: The specific roles of carbon monoxide and antioxidant systems in the placenta,
361
Med. Hypotheses. 64 (2005) 17–27.
362
[9] C.C. Venditti, R. Casselman, M.S.Q. Murphy, S.L. Adamson, J.G. Sled, G.N. Smith, Chronic carbon
363
monoxide inhalation during pregnancy augments uterine artery blood flow and uteroplacental vascular
364
growth in mice., Am. J. Physiol. Regul. Integr. Comp. Physiol. 305 (2013) R939-48.
365
[10] S.A. Bainbridge, A.E. Farley, B.E. McLaughlin, C.H. Graham, G.S. Marks, K. Nakatsu, J.F. Brien,
366
G.N. Smith, Carbon monoxide decreases perfusion pressure in isolated human placenta, Placenta. 23
367
(2002) 563–569.
368
[11] C.C. Venditti, R. Casselman, G.N. Smith, Effects of chronic carbon monoxide exposure on fetal
369
growth and development in mice, BMC Pregnancy Childbirth. 11 (2011) 101.
370
[12] C.C. Venditti, R. Casselman, I. Young, S.A. Karumanchi, G.N. Smith, Carbon monoxide prevents
371
hypertension and proteinuria in an adenovirus sFlt-1 preeclampsia-like mouse model, PLoS One. 9
372
(2014) 1–7.
373
[13] J.P. Gray, S. Karandrea, D.Z. Burgos, A.A. Jaiswal, E.A. Heart, NAD(P)H-dependent quinone
374
oxidoreductase 1 (NQO1) and cytochrome P450 oxidoreductase (CYP450OR) differentially regulate
375
menadione-mediated alterations in redox status, survival and metabolism in pancreatic beta cells,
376
Toxicol. Lett. 262 (2016) 1–11.
377
[14] D. Vukomanovic, M.N. Rahman, Y. Bilokin, A.G. Golub, J.F. Brien, W.A. Szarek, Z. Jia, K.
378
Nakatsu, In vitro activation of heme oxygenase-2 by menadione and its analogs., Med. Gas Res. 4
379
(2014) 4.
AC C
EP
TE D
M AN U
SC
RI PT
359
ACCEPTED MANUSCRIPT
[15] D. Vukomanovic, M. Rahman, Z. Jia, K. Nakatsu, Drug-enhanced carbon monoxide production
381
from heme by cytochrome P450 reductase, Med. Gas Res. 7 (2017) 37-44.
382
[16] H.J. Vreman, R.J. Wong, T. Kadotani, D.K. Stevenson, Determination of carbon monoxide (CO) in
383
rodent tissue: Effect of heme administration and environmental CO exposure, Anal. Biochem. 341
384
(2005) 280–289.
385
[17] B.E. McLaughlin, G.E. Lash, C.H. Graham, G.N. Smith, H.J. Vreman, D.K. Stevenson, G.S.
386
Marks, K. Nakatsu, J.F. Brien, Endogenous carbon monoxide formation by chorionic villi of term,
387
Placenta. (2001) 886–888.
388
[18] J.Z. Byczkowski, A.P. Kulkarni, NADPH-dependent drug redox cycling and lipid peroxidation in
389
microsomes from human term placenta, Int. J. Biochem. 21 (1989) 183–190.
390
[19] B.A. Croy, A.T. Yamada, F.J. DeMayo, S.L. Adamson, The guide to investigation of mouse
391
pregnancy, 2015.
392
[20] A.L. Fowden, A.N. Sferruzzi-Perri, P.M. Coan, M. Constancia, G.J. Burton, Placental efficiency
393
and adaptation: endocrine regulation, J. Physiol. 587 (2009) 3459–3472.
394
[21] J.O. Lo, M.C. Schabel, V.H.J. Roberts, X. Wang, K.S. Lewandowski, K.A. Grant, A.E. Frias, C.D.
395
Kroenke, First trimester alcohol exposure alters placental perfusion and fetal oxygen availability
396
affecting fetal growth and development in a non-human primate model, Am. J. Obstet. Gynecol. 216
397
(2017) 302.e1–302.e8.
398
[22] Y. Gottlieb, M. Truman, L.A. Cohen, Y. Leichtmann-Bardoogo, E.G. Meyron-Holtz, Endoplasmic
399
reticulum anchored heme-oxygenase 1 faces the cytosol, Haematologica. 97 (2012) 1489–1493.
AC C
EP
TE D
M AN U
SC
RI PT
380
ACCEPTED MANUSCRIPT
400
[23] P.W. Soothill, W. Morafa, G.A. Ayida, C.H. Rodeck, Maternal smoking and fetal
401
carboxyhaemoglobin and blood gas levels, Br. J. Obstet. Gynaecol. 103 (1996) 78–82.
402
RI PT
403 404 405
SC
406 407
M AN U
408 409 410 411
415 416 417 418 419 420 421 422
EP
414
AC C
413
TE D
412
ACCEPTED MANUSCRIPT
FIGURE LEGENDS
424
Figure 1. Comparison of MD-mediated CO production in whole and sonicated placentas.
425
Legend: (A) Isolated GD15 mouse placentas were incubated for 30 minutes with increasing
426
concentrations of up to 1 mM MD at 37 ° C. (B) As an extended concentration-response study, isolated
427
GD15 mouse placentas were incubated for 30 minutes with increasing concentrations of up to 50 mM
428
MD at 37 ° C. Data points are expressed as means ± standard deviation.
429
Figure 2. The effects of high dose MD on CO production in pregnant mice.
430
Legend: (A) CO production in maternal organs at a high dose of MD (6.5 g/L MD). CO production in
431
the spleen is shown on the right y-axis, and separated from the other three tissues by the vertical dotted
432
line. (B) Relative organ mass normalized by gram of body weight. Liver mass is shown on the right y-
433
axis, and separated from the other two tissues by the vertical dotted line. (C) Maternal %COHb in
434
control and treated dams during gestation. Data points are expressed as means ± standard deviation.
435
Figure 3. The effects of high dose MD on fetal CO and placental efficiency.
436
Legend: (A, B) Fetal %COHb and Hb in pooled blood from unperfused GD17.5 pups from control (n=6
437
litters) and MD-treated (n=7 litters) dams. (C) Comparison of fetal:placental mass ratios in control
438
(n=16 litters) and MD-treated (n=17 litters) mice. (D) Comparison of fetal mass in control (n=16 litters)
439
and MD-treated (n=17 litters) mice. (E) Comparison of placental mass in control (n=16 litters) and MD-
440
treated (n=17 litters) mice. Data points are expressed as means ± standard deviation. ** p<0.01,
441
***p<0.001.
442 443
AC C
EP
TE D
M AN U
SC
RI PT
423
ACCEPTED MANUSCRIPT
Author Contributions
445
Chioma U. Odozor: I declare that I participated in the study design, data collection, laboratory work,
446
statistical analysis and writing of the manuscript. I have seen and approved the final version. I have no
447
conflicts of interest.
448
Nichole Peterson: I declare that I participated in the data collection and laboratory work. I have seen
449
and approved the final version. I have no conflicts of interest.
450
Jessica Pudwell: I declare that I participated in the statistical analysis. I have seen and approved the
451
final version. I have no conflicts of interest.
452
Graeme N. Smith I declare that I participated supervising all aspects of the study and contributed to all
453
versions of the manuscript. I have seen and approved the final version. I have no conflicts of interest.
SC
M AN U
TE D EP AC C
RI PT
444
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
HIGHLIGHTS: Menadione (MD) increased carbon monoxide (CO) in tissues of pregnant mice
•
6.5 g/L MD did not significantly increase maternal %COHb
•
6.5 g/L MD during gestation decreased fetal mass without changes to placental mass
AC C
EP
TE D
M AN U
SC
RI PT
•
S0143-4004(18)30239-X
DOI:
10.1016/j.placenta.2018.09.007
Reference:
YPLAC 3879
To appear in:
Placenta
Received Date: 31 May 2018 Revised Date:
20 September 2018
Accepted Date: 25 September 2018
Please cite this article as: Odozor CU, Peterson N, Pudwell J, Smith GN, Endogenous carbon monoxide production by menadione, Placenta (2018), doi: https://doi.org/10.1016/j.placenta.2018.09.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: Endogenous carbon monoxide production by menadione
2
Author names and affiliations: Chioma U. Odozora, Nichole Petersona, Jessica Pudwellb, and Graeme
3
N. Smitha,b
4
a. Department of Biomedical and Molecular Sciences, Queen’s University, Kingston, Ontario, Canada,
5
K7L 3N6
6
b. Department of Obstetrics and Gynaecology, Queen’s University, Kingston, Ontario, Canada, K7L
7
3N6
8
Authors’ email addresses:
9
C.U. Odozor
[email protected]
10
N. Peterson
[email protected]
11
J. Pudwell
[email protected]
12
G.N. Smith
[email protected]
13
Corresponding Author:
14
Graeme N Smith MD, PhD, FRCSC
15
Professor and Head, Department of Obstetrics and Gynecology
16
Professor, Biomedical and Molecular Sciences and Imaging Services
17
Director, Clinician Investigator Program
18
Queen’s University
19
76 Stuart Street
20
Kingston, Ontario
21
K7L 2V7
22
Telephone: (613) 548-1372
M AN U
TE D
EP
AC C
SC
RI PT
1
ACCEPTED MANUSCRIPT
Fax: (613) 538-1330
24
Email: [email protected]
25
ABSTRACT
26
Introduction: Smokers have a significantly decreased risk of pre-eclampsia (PE), possibly attributed to
27
an increase in blood carbon monoxide (CO) concentrations. At physiological concentrations, CO has
28
been demonstrated to have placental vasodilatory and anti-inflammatory properties. Increasing
29
endogenous CO production may have therapeutic potential to either prevent or treat PE. Menadione
30
(MD), synthetic vitamin K3, increases CO in rat microsomes. Our objective was to investigate MD’s
31
ability to increase endogenous CO concentrations in pregnancy.
32
Methods: Three experiments were completed. First, in vitro CO production was measured using
33
isolated GD15 placentas. Second, non-pregnant normotensive mice received no, 1.5, 4.0 or 6.5 g/L MD
34
for 7 days. Lastly, pregnant normotensive mice received either no or 6.5 g/L MD in water from GD10.5
35
to GD17.5. Consumption was measured as average daily water intake per gram of body weight.
36
Maternal and fetal CO levels in the blood and tissue were quantified using headspace gas
37
chromatography.
38
Results: MD significantly increased CO production in isolated GD15 placentas. In both pregnant and
39
non-pregnant experiments, splenic CO, hepatic CO, and splenic mass were higher in treated mice
40
compared to controls (all p<0.05). Maternal %COHb and Hb in treated dams were not significantly
41
different compared to controls. The fetal:placental mass ratio was significantly lower in the treatment
42
group (p=0.002).
43
Discussion: Placental CO production was observed in GD15 placentas after co-incubation with MD.
44
MD administration increased CO in the liver and spleens of pregnant mice. Further investigation into
45
different doses of MD is required to identify one without demonstrable fetal/placental effects.
AC C
EP
TE D
M AN U
SC
RI PT
23
ACCEPTED MANUSCRIPT
KEYWORDS: Menadione; pregnancy; carbon monoxide; mouse placenta; pre-eclampsia
47
ABBREVIATIONS: AdsFlt-1: adenovirus soluble fms-like tyrosine kinase-1; CO: carbon monoxide;
48
COHb: carboxyhemoglobin; CPR: cytochrome p450 reductase; EGD: early gestational demise; LGD:
49
late gestational death; Hb: hemoglobin; HO: heme oxygenase; MD: menadione; NQO1:
50
NAD(P)H:quinone oxidoreductase 1; PE: pre-eclampsia; ROS: reactive oxygen species; SSA:
51
sulfosalicylic acid
52
INTRODUCTION:
53
Pre-eclampsia (PE) remains a leading cause of maternal-fetal morbidity and mortality worldwide.
54
Characterized by de novo hypertension and proteinuria, PE affects approximately 5-8% of pregnancies
55
[1]. Multiple strategies are being explored as potential targets for the treatment or prevention of PE. The
56
primary focus of investigation has been directed at the roles of immunology and maternal endothelial
57
dysfunction in the disease’s pathogenesis [2].
58
Smoking decreases the risk of developing PE by up to 33% [2–5]. This is not seen in smokeless tobacco
59
(snuff) users [3], suggesting that the presence of carbon monoxide (CO), a major combustion product,
60
may be responsible for the reduced risk [2,6]. Despite the recognition that CO is toxic at high
61
concentrations, numerous studies have shown its beneficial effects in many disease states at low
62
concentrations. Our focus is on the use of CO at therapeutic dosages in the treatment of PE. CO is a
63
gasotransmitter produced mainly from heme degradation by the enzyme heme oxygenase (HO). In
64
pregnancy, the rate of endogenous CO production increases by 30-40% due to increased heme
65
metabolism [7]. However, women with PE have significantly lower end-tidal breath CO levels than
66
healthy pregnant women, suggesting that decreased CO may be one of the contributors to PE
67
development [8].
AC C
EP
TE D
M AN U
SC
RI PT
46
ACCEPTED MANUSCRIPT
We previously demonstrated that the chronic exposure of 250 ppm CO to pregnant CD-1 mice increased
69
placental vessel branching, arterial size and placental perfusion [9,10], without demonstrable adverse
70
maternal or fetal effects [11]. Further, we demonstrated that in the adenovirus soluble fms-like tyrosine
71
kinase-1 (AdsFLT-1) mouse model of PE, the same dose normalized maternal blood pressure and renal
72
function compared to untreated mice [12]. Thus, the promotion of angiogenesis and vasodilation by
73
increased CO via alternative delivery methods may be beneficial in PE.
74
Menadione (MD), vitamin K3, is a synthetic naphthoquinone that is capable of either one or two
75
electron transfers, catalyzed by cytochrome p450 reductase (CPR) or NAD(P)H:quinone oxidoreductase
76
1 (NQO1), respectively [13]. Vukomanovic et al. (2014) demonstrated that MD activates HO-2 to
77
produce CO, resulting in a seven-fold increase in enzymatic activity [14]. Using human recombinant
78
enzymes and rat brain microsomes, they reported that HO’s coenzyme CPR increased MD-mediated CO
79
production, independent of HO [15]. In the present study, the ability of MD to increase CO production
80
in harvested placentas and pregnant mice was investigated. To provide a basis for the use of MD in
81
animal models of PE, we hypothesized that MD would increase CO production when incubated with
82
mouse placentas, and that its administration in vivo would increase endogenous CO production in female
83
mice.
84
METHODS
85
Animals
86
All experimental procedures were approved by the Queen’s University Animal Care Committee
87
(Protocol 2016-1691) in accordance with Canadian Council for Animal Care guidelines. All mice were
88
kept on a 12-hour light/dark cycle and received Purina LabDiet 5015 pellets and water ad libitum.
89
a) Incubation Study – Pregnant Mice with no Menadione Exposure
AC C
EP
TE D
M AN U
SC
RI PT
68
ACCEPTED MANUSCRIPT
Female, timed-pregnant gestational day (GD) 13 CD-1 mice (Charles River, USA) were housed until
91
GD15. GD1 corresponded with the day that a copulatory plug was observed in the vagina, as timed by
92
Charles River Laboratories.
93
b) Dosing Study - Non-Pregnant Mice
94
Virgin female CD-1 mice aged 4-8 weeks (Charles River, USA) were used in the non-pregnant mouse
95
experiments to first confirm the safety of daily oral MD administration. Non-pregnant mice were used
96
for the dose-finding study first to avoid undue harm.
97
c) Dosing Study - Pregnant Mice
98
Female CD-1 mice aged 4-8 weeks (Charles River, USA) were bred with adult male CD-1 mice
99
(Charles River, USA) for use in the pregnancy studies. GD0.5 corresponded with the day that a
M AN U
SC
RI PT
90
copulatory plug was observed in the vagina.
101
Menadione Preparation
102
i) Incubation Study
103
Menadione sodium bisulfite (Sigma Aldrich Inc., USA) was prepared by serial dilution from a 50 mM
104
stock solution dissolved in double distilled water (ddH2O) and stored at 4°C. Menadione preparations
105
were made in darkened conditions to avoid photodegradation.
106
ii) Dosing Studies
107
Menadione solutions were supplied to mice in 50 mL amber conical tubes (VWR International, USA).
108
Non-pregnant and pregnant mice were allocated to either the control or treatment groups. The control
109
group received regular drinking water. The treatment groups received drinking water for the first four
110
days before MD was added. Total water intake per cage and mouse weights were recorded to determine
111
the average daily water intake per gram of body weight in each cage. Each cage housed a maximum of 4
112
mice. Water volume was recorded and changed every 24 hours.
AC C
EP
TE D
100
ACCEPTED MANUSCRIPT
a) Non-Pregnant Mice
114
Non-pregnant mice in the treatment group received tap water from days 1 to 4, followed by 1.5 g/L, 4
115
g/L or 6.5 g/L of MD in drinking water from days 5 to 12.
116
b) Pregnant Mice
117
The experimental period for the pregnant mice started on GD6.5 to correspond with the same twelve-day
118
timeline for the non-pregnant mice experiment. Pregnant mice in the treatment group received tap water
119
from GD6.5 to GD9.5, followed by 6.5 g/L of MD in drinking water from GD10.5 to GD17.5.
120
Determination of %COHb and Hb in Dosing Studies
121
a) Non-Pregnant Mice
122
Blood collection occurred on days 1, 6, and 9 in non-pregnant mice. 50 uL of blood was collected from
123
the submandibular vein into 1.5 mL polypropylene microcentrifuge tubes containing 8 uL of EDTA. 300
124
uL of saline (0.9% NaCl) was given i.p. following each collection. 100 – 500 uL of blood was collected
125
from the retro-orbital sinus on Day 12 to allow for larger blood volume collection, as permitted by the
126
ethics committee as a terminal procedure.
127
b) Pregnant Mice
128
Blood collection occurred on GD6.5, 11.5, and 14.5 in pregnant mice from the submandibular vein,
129
followed by a saline injection as detailed above. 100 – 500 uL of blood was collected from the retro-
130
orbital sinus on GD17.5.
131
Hemoglobin (Hb) and %COHb levels in all mice were measured as previously described [11]. Briefly,
132
Hb was measured in duplicate using the HemoCue Hb 201+ System (Radiometer, Sweden) following
133
blood collection. A Hamilton repeating dispenser attached to a 10 µL airtight Hamilton syringe
134
(Hamilton, USA) was used to dispense 1 µL of blood into each GC vial containing 20 µL of 2%
135
sulfosalicylic acid (SSA). Vials for %COHb quantification were made in triplicate, with another vial
AC C
EP
TE D
M AN U
SC
RI PT
113
ACCEPTED MANUSCRIPT
containing only 2% SSA as a blank. Following 60 minutes of incubation on ice, CO levels were
137
determined using gas chromatography (Peak Performer 1 Analyzer, Peak Laboratories, Mountain View,
138
CA).
139
Animal Sacrifice and Perfusion
140
i) Incubation Study
141
Mice were anesthetized on GD15 using sodium pentobarbital (110 mg/kg, intraperitoneally) (Ceva Santé
142
Animale, France). GD15 corresponds with the time at which hypertension was observed in AdsFLT-1
143
mice [12]. Mice were systemically perfused with Krebs’ buffer (pH = 7.40) by inserting a 21-gauge
144
butterfly needle (#367344, BD Vacutainer, USA) into the left ventricle and piercing the right atrium to
145
allow for blood egress until the liver was visibly blanched. Perfusion was the final assurance of death.
146
1 L of Krebs’ buffer was prepared by dissolving final concentrations of the following solutions in
147
ddH2O: 120 mM NaCl (Bioshop Canada Inc., Burlington), 5.6 mM KCl (Fisher Scientific, USA), 1.2
148
mM MgSO47H2O (BDH Inc, Toronto), 1.2 mM NaH2PO4H2O (Fisher Scientific, USA), 2.5 mM
149
CaCl22H2O (Sigma Aldrich Inc., USA), 11 mM D-glucose (BDH Inc, Toronto), and 25 mM NaHCO3
150
(Sigma Aldrich, Inc., USA). The solution was bubbled with 5% CO2/balance O2 (Praxair Air, Kingston)
151
for 10 minutes and pH was recorded using a benchtop pH metre (Accumet AB15, Fisher Scientific,
152
USA).
153
ii) Dosing Studies
154
Animals were sacrificed (as detailed above) at the end of the twelve-day experimental period,
155
corresponding to day 12 in non-pregnant mice and GD17.5 in pregnant mice. Gravity perfusion with
156
Krebs' buffer (pH = 7.40) was performed prior to tissue CO determination. In perfused pregnant mice,
157
the uterine horns were excised, and viable fetuses and placentas were counted and weighed. Fetuses
158
from unperfused mice were decapitated to collect blood with pipettes containing 5 uL of EDTA. Early
AC C
EP
TE D
M AN U
SC
RI PT
136
ACCEPTED MANUSCRIPT
gestational demises (EGDs) and late gestational deaths (LGDs) were also noted, as previously described
160
[11].
161
Menadione Incubation with Whole Placentas
162
For the incubation study, GD15 placentas, unexposed to MD prior to harvest, were harvested from the
163
uterine horns, weighed and immediately placed on ice. 2 mL amber gas chromatography (GC) vials
164
(Chromatographic Specialties, Inc.) with open-top caps were sealed with 8 mm blue chromatherm septa
165
(Chromatographic Specialties Inc., C13302). For each sample vial, one weighed halved or whole
166
placenta (~0.01 – 0.1 g) was added to 90 uL of buffer and 10 uL of 0, 0.01, 0.05, 0.1, 0.5 or 1 mM MD
167
for initial dosing studies and 0, 1, 5, 10 or 50 mM MD for extended dosing studies. At minimum,
168
samples were made in duplicates. Triplicate vials containing either 100 uL of buffer alone, or in
169
combination with one placenta each, were made to establish vehicle and tissue blanks, respectively.
170
Menadione Incubation with Sonicated Placentas
171
Three weighed GD15 placentas, unexposed to MD prior to harvest, were randomly selected from each
172
litter, homogenized with a glass tissue grinder in 4 equivalent volumes of ice-cold ddH2O and sonicated
173
on ice at 3 watts for 5 seconds using the Sonic Dismembrator Model 100 (Fisher Scientific, USA). GC
174
vials containing 5 – 10 ul of 20% w/v tissue sonicate were prepared in triplicate, equivalent to 1 – 2 mg
175
of tissue per vial. All vials were made up to 100 uL total volume using Krebs’ buffer. Tissue blanks and
176
sample vials contained sonicated tissue, while the latter also contained 10 uL of either 0.01, 0.05, 0.1,
177
0.5, 1, 5, 10 or 50 mM MD as appropriate for the given study.
178
Tissue CO Determination
179
i) Incubation Study
AC C
EP
TE D
M AN U
SC
RI PT
159
ACCEPTED MANUSCRIPT
Vials were incubated for 30 minutes in a 37° C water bath, then placed on dry ice to stop any chemical
181
reactions. CO content was analyzed using headspace GC as previously described [16], and expressed as
182
pmol CO/g tissue.
183
ii) Dosing Studies
184
Perfused organs were harvested and immediately placed on ice. For all comparisons, the organ mass was
185
normalized to body weight. Livers, spleens and right kidneys were collected from non-pregnant mice.
186
The same organs were harvested from pregnant mice, in addition to three pooled placentas. 80 to 2000
187
mg of each tissue were homogenized and sonicated as detailed above. 10 uL of each 20% w/v tissue
188
sonicate were prepared in triplicates, equivalent to 2 mg of fresh tissue per vial. Each vial contained 5
189
uL of 30% SSA, with sample vials containing 10 uL of sonicated tissue. All vials were made up to 40 uL
190
total volume using ddH2O. CO content in the sonicates was then quantified after 1 hour of incubation in
191
ice using headspace gas chromatography, and expressed as pmol CO/mg tissue as described by Vreman
192
et al. (2005) [16].
193
Statistical Analysis
194
Data are presented as means ± standard deviation (SD) and analyzed using SPSS v24 (SPSS Inc,
195
Chicago, IL, USA) and GraphPad Prism 6. A p value of <0.05 was considered statistically significant for
196
all analyses.
197
i) Incubation Study
198
CO production in whole and sonicated placentas was modelled using a linear mixed effects model to
199
control for mouse of origin (random effect) for the placental sample and dose (Y= Mouse + Dose +
200
Error). Four models were generated; (1) sonicated placenta initial dosing, (2) whole placenta initial
201
dosing, (3) sonicated placenta extended dosing, and (4) whole placenta extended dosing.
202
ii) Dosing Studies
AC C
EP
TE D
M AN U
SC
RI PT
180
ACCEPTED MANUSCRIPT
Linear regression analyses with generalized estimating equations were used to report parameter
204
estimates ± standard error per unit change in dose in both non-pregnant and pregnant mice, taking into
205
account within-cage variation, as mice were housed together during the study. Each mouse was
206
considered an experimental unit, while the cage and/or day (as appropriate) were the within-subject
207
factors. A generalized estimating equation with binomial distribution was used to compare the
208
proportion of live pups to EGDs and LGDs between the two groups. Fetal and placental data were
209
analyzed as per-litter average values. The main outcome parameters were tissue CO, organ mass,
210
fetal:placental mass and change in maternal weight, %COHb and Hb. Cage exposure to MD and the
211
treatment day were covariates in blood analyses. The number of mice in a cage was a covariate for
212
fetal:placental mass comparisons. The number of implantation sites and the number of mice per cage
213
were covariates for maternal weight gain and litter survival comparisons. Repeated measures ANOVAs
214
were used to compare average daily water intake between the control and treatment groups in both non-
215
pregnant and pregnant groups. The analysis was completed at the cage level, with average water intake
216
calculated based on total water intake divided by number of mice in the cage. Non-dosing days and
217
dosing days were analysed separately.
218
RESULTS
219
1. Carbon monoxide production in isolated GD15 mouse placentas
220
CO production for the initial concentration-response studies in sonicated placentas (n=4) and intact
221
placentas (n=8) can be seen in Figure 1A. CO production for the extended concentration-response
222
studies in sonicated placentas (n=4) and intact placentas (n=4) can be seen in Figure 1B. In both studies,
223
increasing doses of MD resulted in a concentration-dependent increase in CO production, with higher
224
CO values produced using sonicated tissue. Parameter estimates ± standard error for the dose variable in
225
each of the 4 models were 30065±4466, 694±100, 2919±740, and 31±14 pmol CO/g tissue, respectively.
AC C
EP
TE D
M AN U
SC
RI PT
203
ACCEPTED MANUSCRIPT
2. Dose exploration in non-pregnant mice
227
Shown in Supplemental Figure S1-A, both hepatic and splenic mass (n=4-5 mice/dose/organ) increased
228
with MD dose in non-pregnant mice (kidney mass: +0.180 ± 0.083 mg/g body weight, p=0.03; spleen
229
mass: +0.364 ± 0.128 mg/g body weight, p=0.004). As seen in Supplemental Figure S1-B, spleen CO
230
also increased with MD dose (+3.278 ± 0.635 pmol CO/ g tissue, p<0.001). No significant changes were
231
observed for kidney CO, liver CO, and liver mass, respectively.
232
Increasing cage MD exposure had a positive effect on %COHb (+0.014 ± 0.0064 %COHb, p=0.03). No
233
significant difference in %COHb was observed when comparing days of collection (data not shown). In
234
contrast, no significant difference in Hb was observed with increasing MD dose, while the day of blood
235
collection had a negative effect (-0.890 ± 0.313 g/L; p=0.004). In total 3 cages received no MD and 2
236
cages each received, 1.5, 4.0, or 6.5 g/L MD. No differences in average water intake per cage were
237
observed between dose groups during either the non-dosing period (days 1-4) or dosing period (days 5-
238
12).
239
3. The effect of high dose MD on maternal weight gain and daily water intake
240
High dose MD for seven days did not significantly affect daily water consumption in pregnant, treated
241
dams (n= 7 cages) compared to control dams (n=6 cages) (p=0.05). Dams in the MD treatment group
242
(n=14 mice) tended to gain less weight from GD0.5 to GD17.5 than dams in the control group (n=13
243
mice) (-0.748 ± 0.285; p=0.009).
244
4. Maternal – fetal outcomes after high dose MD administration
245
Carbon monoxide production was observed in all tissue samples from treated mice (n= 9 mice, 5 cages)
246
(Fig. 2A), with significantly higher tissue CO in the spleen (+2.354 ± 0.432 pmol CO/g tissue, p<0.001)
247
and liver (+0.236 ± 0.0995 pmol CO/g tissue, p=0.017) compared to the control (n=8 mice, 5 cages).
248
Splenomegaly was observed in spleens of most treated mice (+0.325 ± 0.113 mg/g body weight,
AC C
EP
TE D
M AN U
SC
RI PT
226
ACCEPTED MANUSCRIPT
p=0.004), however placental and renal CO, and renal and hepatic mass did not differ between treatment
250
and controls (Fig. 2B). Maternal %COHb and Hb were not significantly increased by MD exposure.
251
However, in both groups Hb decreased (p<0.001) and %COHb increased (p<0.001) at later collection
252
days.
253
The numbers of EGDs and LGDs were not significantly different between treated (n=17 litters) and
254
control (n=16 litters) dams. There was a non-significant increase in fetal %COHb and Hb in treated mice
255
(Fig. 3A, 3B). Fetal mass was lower in treated mice (p<0.001) resulting in a lower ratio of fetal:placental
256
mass (10.22 ± 1.13, n=16 litters vs. 8.80 ± 1.45, n=17 litters, p=0.002) (Fig. 3C,3D). However, no
257
difference in placental mass was observed (Fig. 3E).
258
DISCUSSION
SC
M AN U
259
RI PT
249
The results presented in this paper suggest that in vivo MD increases CO production in maternal tissues, particularly in the liver and spleen, but not in the kidney, placenta or circulation. However, in
261
vitro exposure to MD increased CO production by placental tissue in our study. It has been shown that
262
endogenous CO production occurs in human chorionic villi [17], and that MD undergoes redox cycling
263
in human placental microsomes [18]. This is the first study to demonstrate increased CO production by
264
MD in vitro in placental tissue. We demonstrated that MD increases CO production in isolated GD15
265
mouse placentas at 50 mM MD. This concentration is equivalent to approximately 13.8 g/L MD in vivo.
266
We expected that sonication would result in a higher percent increase in CO production at increasing
267
concentrations of MD, due to the greater surface area and therefore increased availability of heme to
268
interact as a substrate. Sonication prior to CO quantification produced higher levels of CO compared to
269
MD incubation with whole placentas. However, we were unable to report a statistically significant
270
increase in placental CO production in our in vivo studies, despite observing a trend towards increased
271
production.
AC C
EP
TE D
260
ACCEPTED MANUSCRIPT
272
High doses of MD were used to determine possible adverse effects in dams and fetuses. Given that higher %COHb levels were reported with inhalational CO without adverse maternal and/or fetal
274
effects [11] and that we did not observe significant changes in %COHb attributable to MD dose, the
275
adverse maternal and fetal effects seen in this study are not likely due to a %COHb change but rather a
276
direct effect of the MD. While we did not observe a statistically significant difference in water intake,
277
there did appear to be a trend towards decreased intake in the treatment group during the dosing days
278
that was not observed on the non-dosing days. This was likely a result of taste aversion to high dose
279
MD, possibly leading to the decrease in maternal weight gain in treated mice. Normally, maternal
280
weight gain during gestation increases as the fetus develops [19]. However, treated mice did not
281
demonstrate the same growth compared to the controls. As fetal mass was significantly different
282
between the two experimental groups, it is also possible that decreases in fetal mass contributed to the
283
lower maternal weight gain in treated mice.
M AN U
SC
RI PT
273
The fetal:placental mass ratio is often used as a proxy measure for placental efficiency,
285
indicative of the placenta’s ability to maximize nutrient delivery to the developing fetus. Low placental
286
efficiency may be caused by low fetal weight, high placental weight or a combination of both [20]. The
287
former is true in this case, as only fetal mass was significantly lower on average in treated mice.
288
However, absence of a change in placental mass does not disprove that there may be structural or
289
functional differences between the placentas of control and treated mice. Lo et al. (2017) looked at
290
fetotoxic effects and placental function after exposure to high doses of alcohol [21]. While fetal mass
291
was significantly lower in the alcohol-exposed group, there was no significant difference in placental
292
mass. However, placental perfusion and oxygenation were reduced. Moreover, the slight improvement
293
in placental function later in gestation was speculated to be compensatory but unable to maintain normal
294
fetal growth, causing the decrease in fetal mass. It could be that in our study, high dose MD led to early
AC C
EP
TE D
284
ACCEPTED MANUSCRIPT
changes in placental function, which may have had an impact on fetal mass and survival. However, a
296
compensation for these changes may have led to increases in placental size. Histological analyses of the
297
collected tissues would lend some insight on whether high dose MD affects normal placental
298
development and/or the degree of hypoxia.
299
RI PT
295
Similar to the %COHb levels observed in smokers, we hypothesized that high dose MD could significantly increase %COHb to 5-10%, and allow for vasodilation and increased vessel branching as
301
an adaptive response to increased CO production [9]. However, %COHb never rose above 1% in our
302
study. Differences in CO delivery, as well as the location of MD’s targets may account for this. In our
303
previous studies, CO was delivered exogenously, while MD-mediated CO production is an endogenous
304
process. It may be that the location of the enzymes upon which MD acts limits its ability to increase CO
305
production in the blood. Both HO and CPR are membrane-bound enzymes located on the endoplasmic
306
reticulum [22]. Erythrocytes lack membrane bound organelles, and therefore would not be expected to
307
degrade heme enzymatically without transport to the spleen. However, macrophages in the blood may
308
still be a possible source of HO and CPR [22], thus increasing CO in the blood of treated mice, albeit to
309
a lesser degree than in tissues.
M AN U
TE D
High dose MD increased CO in the tissues of treated mice, although not to the same extent as
EP
310
SC
300
with direct CO delivery [11]. As the spleen is the primary site of heme degradation, and its open
312
circulation made it difficult to completely perfuse, it was expected that its higher substrate availability
313
would allow for more CO to be released. Likewise, the liver is also a major site of heme metabolism,
314
which could have contributed to its ability to produce higher amounts of CO at 6.5 g/L MD.
315
AC C
311
In our study, we were unable to report a statistically significant difference in fetal %COHb and
316
Hb in treated mice, however we did observe a trend towards increased values. It is possible that
317
increases in %COHb may be causing some degree of hypoxia, which could stimulate compensatory
ACCEPTED MANUSCRIPT
erythropoietin production to also increase Hb levels [23]. Consequently, fetal tissues should be collected
319
to quantify CO, hypoxia and erythropoietin levels after maternal exposure to various MD doses. Given
320
that the fat-soluble vitamin K has limited transfer across the placenta, it is not expected that MD would
321
cause a significant disruption in the fetus’ normal adaptive responses to angiogenesis and alterations in
322
flow from the maternal side of the placenta.
323
RI PT
318
This study demonstrates that MD increases endogenous CO production in healthy female mice and lays a framework for future experimentation with MD and CO production in various animal models
325
of PE. The administration of 6.5 g/L MD had negative impacts on fetal survival and growth. It is
326
possible that physiological and biochemical differences between placentas from normotensive and PE
327
pregnancies may have an impact on both the extent and location of MD-mediated CO production.
328
However, as mice were normotensive, PE-relevant physiological endpoints could not be attained.
329
Therefore, further studies must be done to determine whether lower doses of MD mitigate PE-like
330
symptoms in a PE mouse model, without severe consequence to the fetus, before it can be deemed a
331
viable therapeutic option in the treatment of PE.
332
Acknowledgements
333
The authors would like to acknowledge Dr. Kanji Nakatsu and his lab for initial help with
334
troubleshooting gas chromatography experiments. This research was funded by the Canadian Institutes
335
of Health Research (CIHR) Catalyst Grant: Catalyzing Innovation in Preterm Birth Research (Grant No.
336
RN297868 - 371228).
337
Conflicts of interest: None.
338 339 340
AC C
EP
TE D
M AN U
SC
324
ACCEPTED MANUSCRIPT
REFERENCES:
342
[1] J.M. Roberts, M. Druzin, P. August, R.R. Gaiser, G. Bakris, J.P. Granger, J.R. Barton, A. Jeyabalan,
343
I. Bernstein, D.D. Johnson, S.A. Karumanchi, C.Y. Spong, M.D. Lindheiner, E. Tsingas, M.Y. Owens,
344
J.N. Martin Jr, G.R. Saade, B.M. Sibai, ACOG Guidelines: Hypertension in pregnancy, 2012.
345
[2] C.C. Venditti, G.N. Smith, Involvement of the heme oxygenase system in the development of
346
preeclampsia and as a possible therapeutic target., Women’s Health (Lond. Engl). 10 (2014) 623–43.
347
[3] L.D. Longo, The biological effects of carbon monoxide on the pregnant woman, fetus, and newborn
348
infant., Am. J. Obstet. Gynecol. 129 (1977) 69–103.
349
[4] D. Kreiser, M. Baum, D.S. Seidman, A. Fanaroff, D. Shah, I. Hendler, D.K. Stevenson, E. Schiff,
350
M.L. Druzin, End tidal carbon monoxide levels are lower in women with gestational hypertension and
351
pre-eclampsia., J. Perinatol. 24 (2004) 213–217.
352
[5] A.K. Wikström, O. Stephansson, S. Cnattingius, Tobacco use during pregnancy and preeclampsia
353
risk: effects of cigarette smoking and snuff., Hypertension. 55 (2010) 1254–1259.
354
[6] S.M. Engel, T.M. Janevic, C.R. Stein, D.A. Savitz, Maternal smoking, preeclampsia, and infant
355
health outcomes in New York City, 1995-2003, Am. J. Epidemiol. 169 (2009) 33–40.
356
[7] E.C. Miller, H. Cao, S.W. Wen, Q. Yang, J. Lafleche, M. Walker, The risk of adverse pregnancy
357
outcomes is increased in preeclamptic women who smoke compared with nonpreeclamptic women who
358
do not smoke, Am. J. Obstet. Gynecol. 203 (2010) 334.e1-334.e8.
AC C
EP
TE D
M AN U
SC
RI PT
341
ACCEPTED MANUSCRIPT
[8] S.A. Bainbridge, E.H. Sidle, G.N. Smith, Direct placental effects of cigarette smoke protect women
360
from pre-eclampsia: The specific roles of carbon monoxide and antioxidant systems in the placenta,
361
Med. Hypotheses. 64 (2005) 17–27.
362
[9] C.C. Venditti, R. Casselman, M.S.Q. Murphy, S.L. Adamson, J.G. Sled, G.N. Smith, Chronic carbon
363
monoxide inhalation during pregnancy augments uterine artery blood flow and uteroplacental vascular
364
growth in mice., Am. J. Physiol. Regul. Integr. Comp. Physiol. 305 (2013) R939-48.
365
[10] S.A. Bainbridge, A.E. Farley, B.E. McLaughlin, C.H. Graham, G.S. Marks, K. Nakatsu, J.F. Brien,
366
G.N. Smith, Carbon monoxide decreases perfusion pressure in isolated human placenta, Placenta. 23
367
(2002) 563–569.
368
[11] C.C. Venditti, R. Casselman, G.N. Smith, Effects of chronic carbon monoxide exposure on fetal
369
growth and development in mice, BMC Pregnancy Childbirth. 11 (2011) 101.
370
[12] C.C. Venditti, R. Casselman, I. Young, S.A. Karumanchi, G.N. Smith, Carbon monoxide prevents
371
hypertension and proteinuria in an adenovirus sFlt-1 preeclampsia-like mouse model, PLoS One. 9
372
(2014) 1–7.
373
[13] J.P. Gray, S. Karandrea, D.Z. Burgos, A.A. Jaiswal, E.A. Heart, NAD(P)H-dependent quinone
374
oxidoreductase 1 (NQO1) and cytochrome P450 oxidoreductase (CYP450OR) differentially regulate
375
menadione-mediated alterations in redox status, survival and metabolism in pancreatic beta cells,
376
Toxicol. Lett. 262 (2016) 1–11.
377
[14] D. Vukomanovic, M.N. Rahman, Y. Bilokin, A.G. Golub, J.F. Brien, W.A. Szarek, Z. Jia, K.
378
Nakatsu, In vitro activation of heme oxygenase-2 by menadione and its analogs., Med. Gas Res. 4
379
(2014) 4.
AC C
EP
TE D
M AN U
SC
RI PT
359
ACCEPTED MANUSCRIPT
[15] D. Vukomanovic, M. Rahman, Z. Jia, K. Nakatsu, Drug-enhanced carbon monoxide production
381
from heme by cytochrome P450 reductase, Med. Gas Res. 7 (2017) 37-44.
382
[16] H.J. Vreman, R.J. Wong, T. Kadotani, D.K. Stevenson, Determination of carbon monoxide (CO) in
383
rodent tissue: Effect of heme administration and environmental CO exposure, Anal. Biochem. 341
384
(2005) 280–289.
385
[17] B.E. McLaughlin, G.E. Lash, C.H. Graham, G.N. Smith, H.J. Vreman, D.K. Stevenson, G.S.
386
Marks, K. Nakatsu, J.F. Brien, Endogenous carbon monoxide formation by chorionic villi of term,
387
Placenta. (2001) 886–888.
388
[18] J.Z. Byczkowski, A.P. Kulkarni, NADPH-dependent drug redox cycling and lipid peroxidation in
389
microsomes from human term placenta, Int. J. Biochem. 21 (1989) 183–190.
390
[19] B.A. Croy, A.T. Yamada, F.J. DeMayo, S.L. Adamson, The guide to investigation of mouse
391
pregnancy, 2015.
392
[20] A.L. Fowden, A.N. Sferruzzi-Perri, P.M. Coan, M. Constancia, G.J. Burton, Placental efficiency
393
and adaptation: endocrine regulation, J. Physiol. 587 (2009) 3459–3472.
394
[21] J.O. Lo, M.C. Schabel, V.H.J. Roberts, X. Wang, K.S. Lewandowski, K.A. Grant, A.E. Frias, C.D.
395
Kroenke, First trimester alcohol exposure alters placental perfusion and fetal oxygen availability
396
affecting fetal growth and development in a non-human primate model, Am. J. Obstet. Gynecol. 216
397
(2017) 302.e1–302.e8.
398
[22] Y. Gottlieb, M. Truman, L.A. Cohen, Y. Leichtmann-Bardoogo, E.G. Meyron-Holtz, Endoplasmic
399
reticulum anchored heme-oxygenase 1 faces the cytosol, Haematologica. 97 (2012) 1489–1493.
AC C
EP
TE D
M AN U
SC
RI PT
380
ACCEPTED MANUSCRIPT
400
[23] P.W. Soothill, W. Morafa, G.A. Ayida, C.H. Rodeck, Maternal smoking and fetal
401
carboxyhaemoglobin and blood gas levels, Br. J. Obstet. Gynaecol. 103 (1996) 78–82.
402
RI PT
403 404 405
SC
406 407
M AN U
408 409 410 411
415 416 417 418 419 420 421 422
EP
414
AC C
413
TE D
412
ACCEPTED MANUSCRIPT
FIGURE LEGENDS
424
Figure 1. Comparison of MD-mediated CO production in whole and sonicated placentas.
425
Legend: (A) Isolated GD15 mouse placentas were incubated for 30 minutes with increasing
426
concentrations of up to 1 mM MD at 37 ° C. (B) As an extended concentration-response study, isolated
427
GD15 mouse placentas were incubated for 30 minutes with increasing concentrations of up to 50 mM
428
MD at 37 ° C. Data points are expressed as means ± standard deviation.
429
Figure 2. The effects of high dose MD on CO production in pregnant mice.
430
Legend: (A) CO production in maternal organs at a high dose of MD (6.5 g/L MD). CO production in
431
the spleen is shown on the right y-axis, and separated from the other three tissues by the vertical dotted
432
line. (B) Relative organ mass normalized by gram of body weight. Liver mass is shown on the right y-
433
axis, and separated from the other two tissues by the vertical dotted line. (C) Maternal %COHb in
434
control and treated dams during gestation. Data points are expressed as means ± standard deviation.
435
Figure 3. The effects of high dose MD on fetal CO and placental efficiency.
436
Legend: (A, B) Fetal %COHb and Hb in pooled blood from unperfused GD17.5 pups from control (n=6
437
litters) and MD-treated (n=7 litters) dams. (C) Comparison of fetal:placental mass ratios in control
438
(n=16 litters) and MD-treated (n=17 litters) mice. (D) Comparison of fetal mass in control (n=16 litters)
439
and MD-treated (n=17 litters) mice. (E) Comparison of placental mass in control (n=16 litters) and MD-
440
treated (n=17 litters) mice. Data points are expressed as means ± standard deviation. ** p<0.01,
441
***p<0.001.
442 443
AC C
EP
TE D
M AN U
SC
RI PT
423
ACCEPTED MANUSCRIPT
Author Contributions
445
Chioma U. Odozor: I declare that I participated in the study design, data collection, laboratory work,
446
statistical analysis and writing of the manuscript. I have seen and approved the final version. I have no
447
conflicts of interest.
448
Nichole Peterson: I declare that I participated in the data collection and laboratory work. I have seen
449
and approved the final version. I have no conflicts of interest.
450
Jessica Pudwell: I declare that I participated in the statistical analysis. I have seen and approved the
451
final version. I have no conflicts of interest.
452
Graeme N. Smith I declare that I participated supervising all aspects of the study and contributed to all
453
versions of the manuscript. I have seen and approved the final version. I have no conflicts of interest.
SC
M AN U
TE D EP AC C
RI PT
444
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
HIGHLIGHTS: Menadione (MD) increased carbon monoxide (CO) in tissues of pregnant mice
•
6.5 g/L MD did not significantly increase maternal %COHb
•
6.5 g/L MD during gestation decreased fetal mass without changes to placental mass
AC C
EP
TE D
M AN U
SC
RI PT
•