Objective
Our goal was to define mechanisms that protect murine pregnancies deficient in spiral arterial remodeling from hypertension, hypoxia, and intrauterine growth restriction.
Study Design
Microultrasound analyses were conducted on virgin, gestation day 2, 4, 7, 9, 10, 12, 14, 16, 18, and postpartum BALB/c (wild type) mice and BALB/c- Rag2 −/− /Il2rg −/− mice, an immunodeficient strain lacking spiral arterial remodeling.
Results
Rag2 −/− /Il2rg −/− dams had normal spiral arterial flow velocities, greatly elevated uterine artery flow velocities between gestational day 10-16 and smaller areas of placental flow from gestational day 14 to term than controls. Maternal heart weight and output increased transiently. Conceptus alterations included higher flow velocities in the umbilical-placental circulation that became normal before term and bradycardia persistent to term.
Conclusion
Transient changes in maternal heart weight and function accompanied by fetal circulatory changes successfully compensate for deficient spiral arterial modification in mice. Similar compensations may contribute to the elevated risk for cardiovascular diseases seen in women and their children who experience preeclamptic pregnancies.
Normal pregnancy significantly changes reproductive tract and systemic circulations to accommodate growth and metabolic demands of developing conceptuses. In species with hemochorial placentation, such as humans, uterine vascular changes include increased permeability, angiogenesis, and structural remodeling of most spiral arteries (SA) to reduce vasoactivity and increase capacity. In humans, inadequate modification of decidual and myometrial SA is often accompanied by preeclampsia (PE) and/or intrauterine growth restriction (IUGR). SA remodeling is progressive and involves loss of vascular smooth muscle coat and elastic lamina, mural invasion by trophoblast cells that deposit nonvasoactive fibrinoid and transient replacement of vascular endothelial cells by intravascular trophoblast. In humans, SA remodeling normally occurs during first and second trimester with approximately two-thirds of these vessels dilating 5-10 fold. SA remodeling slows maternal blood flow, minimizes turbulence, and optimizes exchange time with the fetal circulation. Decidual lymphocytes, mainly uterine natural killer (uNK) cells, contribute to early stages of decidual SA transformation through cytokine secretion as do mechanical flow properties such as shear forces and pressure.
The systemic circulatory changes of early normal pregnancy include increases in cardiac output, blood volume, and glomerular filtration rate that result in an overall maternal state of high blood flow with low vascular resistance. Inadequate or excessive cardiovascular adaptations before week 20 of pregnancy predict complications. Cardiac disease occurs in 1-4% of pregnancies in women with no preexisting heart abnormalities. Further, women who experience adverse outcomes such as PE, carry increased risks for cardiovascular and metabolic diseases into later life. Thus, identifying links between uterine vascular remodeling and systemic cardiovascular abnormalities is critical for appropriate management of human gestational complications. Early in normal human and mouse pregnancies, mean arterial blood pressure decreases slightly or is stable. This indicates that gains in uterine flow are more attributable to morphologic remodeling of uterine arteries than to alterations in their smooth muscle tone.
Mouse models have promoted understanding of mechanisms regulating gestational changes to the cardiovascular system. Comparisons of microultrasound measurements of placental growth and blood flow velocity between normal and mutant mice (ie, Nos3 −/− ) revealed significant similarities to human pregnancies. The reports on pregnancies in mutant mice examined by microultrasound do not include reports on animals without SA modification. Lack of SA modification is a characteristic of mice deficient in uNK cells. It is a phenotype readily reversed by transplantation of marrow from mice genetically deficient for T and B cell differentiation (ie, NK + T − B − marrow). SA modification is histologically detected between gestation day 9-10 of 19-20 day mouse pregnancy, a time-point recognized for allantoic fusion with the chorionic plate to open placental circulation. In this study, we examine hemodynamic features of wild-type BALB/c +/+ (WT; ie, NK + T + B + ) and alymphoid (deficient in NK, T, and B cells [NK − T − B − ]) BALB/c- Rag2 −/− /Il2rg −/− mice before, across, and after pregnancy using microultrasound. Unlike most patients with impaired SA remodeling, Rag2 −/− /Il2rg −/− dams with unmodified SA have normotensive pregnancies without detectable elevation in hypoxia of placental, fetal, or maternal tissue. They also show no elevation in fetal loss and their offspring are not growth impaired. Although these interspecies outcome differences strongly implicate lymphocyte-based immunity in progression of human gestational complications accompanying incomplete SA remodeling, they do not explain the physiologic mechanisms that provide normoxic placental, fetal, and maternal tissues in the absence of maternal hypertension. We now report specific cardiovascular differences between pregnant Rag2 −/− /Il2rg −/− and WT females. These differences include greater uterine arterial blood velocity and increased heart weight and performance, especially over the second half of pregnancy. These maternal adaptations are accompanied by lower fetal heart rates and reduced placental vascular space.
Materials and Methods
Mice
This study used 8-10 week old mice (39 BALB/c +/+ ; Jackson Laboratory, Bar Harbor, ME and 44 in-house bred BALB/c -Rag2 −/− /Il2rg −/− ). Timed matings were prepared using syngeneic males; copulation plug detection was dated gestational day 0. An ultrasound examination was performed once per female either before mating (nonpregnant, NP), at gestational day 2, 4, 7, 9, 10, 12, 14, 16, 18, or at 48-hour postpartum (PP). Females were then euthanized, weighed, and their organs were immediately dissected, weighed, and processed for other purposes. Wet weights were used to calculate organ to body weight (BW) ratios. Pregnancies were confirmed at gestational day 2-4 by postmortem embryo flushing and at gestational day 7-18, by viable implantation site visualization. Mean litter sizes in this study were not significantly different between genotypes. All procedures were conducted under Animal Utilization Protocols approved by the Animal Care Committee, Queen’s University.
Ultrasound procedures
Anesthetized (inhaled isoflurane) mice, after fur removal (Nair; Church & Dwight Co., Inc, Princeton, NJ), were taped (Transpore; 3M, Maplewood, MN) onto the instrument platform. A thick layer of prewarmed coupling gel (Ecogel 100; ECO-MED Pharmaceutical, Mississauga, Ontario, Canada) was applied over the area to be imaged using a 40-MHz transducer probe (Vevo 770; VisualSonics Inc, Toronto, Ontario, Canada). Body temperature was maintained at 36-37°C by the warmed platform. Maternal heart and respiration rates (ECG/Tm/Resp) were collected with a physiological controller unit.
For maternal uterine artery (UtA) power Doppler scans, bladder was first identified, then the probe was moved to locate the uterine artery arising from the common iliac. SA flows, characterized by pulsed waveforms, were imaged at the mesometrial decidual edge in 2-4 implantation sites of each gestational day 10-18 pregnancy. Peak systolic velocities of fetal umbilical arteries (UmA) were recorded before these arteries entered the placenta and branched into chorionic plate arteries (CPA) that run across the placental surface and branch into the major intraplacental arteries (IPA) in developing primary villi. After UmA study, peak systolic velocity of CPA and IPA flows were recorded in 3-5 locations/implantation sites. Maternal left renal artery (RA) flow velocity waveform was recorded to assess systemic blood velocity. For maternal data, 3-6 different females of each genotype were imaged per time-point (ie, no repeated study was made on any animal). For conceptus measurements, 6-15 implantation sites were averaged per presented time-point. Waveforms were analyzed offline.
Motion modulation (M-mode) images were constructed to evaluate maternal heart chamber dimensions at various times throughout the cardiac cycle. Three-dimensional (3-D) power Doppler data were used to estimate placental volume and structure. The Doppler beam angle was set at 27° to acquire consistent data. Doppler velocity waveforms from UtA, SA, UmA, RA, IPA, and CPA were captured using brightness mode (B-mode) imaging with the following settings: pulse repetition frequency, 10 kHz; wall filter, 100 Hz; display window, 2000 ms; sound speed, 1540 m/s; Doppler gain, 5.00 dB. The display range of the 3-D Doppler measurements of placental volume and vascular density was 19 dB (minimum)-30 dB (maximum).
Data management and analysis
Instrument software packages provided the analyses. Peak systolic velocity, mean velocity, and end diastolic velocity were measured from Doppler images. Doppler indices calculated to interpret the data were:
pulsatility index (PI) = (peak systolic velocity – end diastolic velocity)/mean velocity;
resistance index (RI) = (peak systolic velocity – end diastolic velocity)/(peak systolic velocity);
S/D ratio = peak systolic velocity/end diastolic velocity.
Cardiac measurements were obtained from M-mode data ( Supplementary Figure 4 ). Data are presented as mean (± standard deviation [SD]). Comparisons were performed by Student t test or 1-way analysis of variance (ANOVA) where appropriate for SPSS 13 analysis (SPSS Inc, Chicago, IL). A P value of less than .05 was considered significant.
Results
Maternal circulations in WT and Rag2 −/− /Il2rg −/− pregnancies
1. RA flow patterns of WT and Rag2 −/− /Il2rg −/− mice before, during, and after pregnancy . To evaluate the systemic circulation, maternal RA velocity was studied in WT and Rag2 −/− /Il2rg −/− mice (summarized data, Figure 1 , A ; full data, Supplementary Figure 1 ). RA flow velocity was similar in both strains before mating, suggesting similar homeostatic states. For both strains, RA blood velocity increased at gestational day 2 ( P < .05), remained statistically stable to gestational day 7 and dropped to a nadir at gestational day 10, the first day of placental circulation. Velocities then rebounded with a slightly delayed time course in Rag2 −/− /Il2rg −/− (peak at gestational day 12 in WT; gestational day 14 in Rag2 −/− /Il2rg −/− [ P < .05, compared with other gestational days]). RA PIs rose after conceptus implantation and peaked at gestational day 9-12 in WT ( P < .05, compared with NP). In Rag2 −/− /Il2rg −/− , the peak RA PI was delayed to gestational day 14-18 ( Figure 1 , B). At matched time points, there was no statistical difference in RA PI between the 2 strains ( Figure 1 , B).
2. Uterine arterial flow patterns of WT and Rag2 −/− /Il2rg −/− mice before, during, and after pregnancy . Uterine blood delivery is bidirectional from ovarian and uterine arteries. Because WT and Rag2 −/− /Il2rg −/− mice had similar velocities in the RA, a vessel arising close to or upstream of the ovarian artery, ovarian arterial study was not conducted. Uterine arterial velocity was estimated from vessels arising from the internal iliac ( Figure 2 , A and B) . Mean uterine arterial velocities were similar between WT and Rag2 −/− /Il2rg −/− before pregnancy, at preimplantation (gestational day 2), periimplantation (gestational day 4), and term ( Figure 2 , C). For WT mice, Doppler-detected uterine arterial flow velocity was not different from that of the virgin uterus up to gestational day 9. A statistically significant gain in flow velocity was present from gestational day 9 to PP ( Figure 2 , C and Supplementary Figure 2 , A). The higher flow velocity seen in WT mice at gestational day 9 was stable to gestational day 12, when SA modification is complete histologically. WT uterine arterial velocity was then characterized by a second statistically distinct gain at gestational day 14. This velocity rate continued until parturition ( Supplementary Figure 2 , A). Uterine arterial flow velocity of Rag2 −/− /Il2rg −/− matched WT mice to gestational day 4 and at gestational day 18 and postpartum. It increased earlier than WT (gestational day 7), rose more steeply (doubling that of WT by gestational day 14) then declined ( Figure 2 , C and Supplementary Figure 2 , B).
Calculated resistance indices of WT uterine arteries (PI, RI, S/D ratio) had downward postconception trends to gestational day 14 (37% decrease of PI) then gradually returned to prepregnancy levels ( Figure 2 , D and Supplementary Figure 2 , C and D). In virgin Rag2 −/− /Il2rg −/− females, the uterine arterial PI was similar to that of WT but, after mating, increased to gestational day 9, declined to gestational day 12 then was stable to postpartum. Rag2 −/− /Il2rg −/− uterine arterial PI ( Figure 2 , D) and RI ( Supplementary Figure 2 , C) significantly exceeded WT on gestational days 9, 10, 14, and 16. These data reveal anomalous Rag2 −/− /Il2rg −/− uterine arterial flow velocities over mid to late gestation that normalized peripartum (gestational day 19).
3. Spiral arterial flow patterns of WT and Rag2 −/− /Il2rg −/− mice during pregnancy . Murine decidual SA development begins after implantation. The earliest SA waveform our system identified was at gestational day 10 when WT mean velocity was 18 mm/s ( Figure 2 , E-G). This increased to term with significant elevation at gestational day 16-18 when velocities achieved 32-35 mm/s ( Figure 2 , G and Supplementary Figure 2 , E). Gestational day 10 Rag2 −/− /Il2rg −/− SA flow velocity was 23 mm/s. Velocity increased at gestational day 14-16 ( P < .05) and at gestational day 18 dropped to baseline ( Figure 2 , G and Supplementary Figure 2 , F). There were no cross strain differences in mean velocities of SA ( Figure 2 , G) but RIs (PI, RI, and S/D ratios) at gestational days 14, 16, and 18 were higher for Rag2 −/− /Il2rg −/− than WT ( Figure 2 , H and Supplementary Figure 2 , G and H).
Umbilical and placental circulations in WT and Rag2 −/− /Il2rg −/− pregnancies
Placental circulation follows fusion of the allantois (umbilical cord progenitor) with the chorion. Development of umbilical arterial waveforms without end diastolic velocity were clearly detected by gestational day 10 ( Figure 3 , A-C). Patterns for umbilical arterial peak systolic velocity were similar in WT and Rag2 −/− /Il2rg −/− ; velocities increased at gestational day 12 ( P < .05, compared with gestational day 10 for each strain), gestational day 14 ( P < .05, compared with gestational day 10 or 12) and were maintained until term ( Figure 3 , D). During gestational day 12-16 Rag2 −/− /Il2rg −/− velocities were greater ( P < .05) than WT.
Heart rates (estimated from umbilical artery pulses) were higher in WT than Rag2 −/− /Il2rg −/− fetuses from midpregnancy to birth ( P < .05, Figure 3 , E). WT fetal heart rates increased significantly at gestational day 12 and gestational day 14, then were stable at 156-167 beats/min until birth. Rag2 −/− /Il2rg −/− fetal heart rates rose significantly only at gestational day 12 to 98 beats/min ( P < .05) and remained stable until term. These data indicate Rag2 −/− /Il2rg −/− fetuses differ from WT and suggest they use heart enlargement and/or greater stroke volume rather than increased beating rate to support midpregnancy placental flow. To address placental circulation, chorionic plate, and IPA velocities were assessed ( Figure 3 , F and G). From gestational day 10-18, these waveforms had progressively increasing end diastolic velocities indicating they were branches from umbilical not SA. In both strains, chorionic plate and IPA velocities increased during pregnancy. Compared with WT conceptuses, Rag2 −/− /Il2rg −/− had higher chorionic plate (gestational day 10-16) and IPA (gestational day 10-14) velocities.
Analyses of WT and Rag2 −/− /Il2rg −/− placental structure
Mouse placenta is recognized by its echo image density. Placental size and vascularity (percent placental space with detectable flow) were calculated from 3-D power Doppler images ( Figure 4 , A) . WT placental volume was 7.0 mm 3 at gestational day 10. This increased rapidly to gestational day 14 ( P < .05 compared with gestational day 10), then more slowly to term ( Figure 4 , C). The Rag2 −/− /Il2rg −/− placental growth pattern was similar but smaller volumes were estimated at gestational day 10, 12, and 16 ( P < .05, Figure 4 , C). For both strains, placental vascularities were similar at gestational day 10 and 12, sharply increased at gestational day 14 ( P < .05, compared with gestational day 10, 12), then were stable until term ( Figure 4 , D). During gestational day 12-18 Rag2 −/− /Il2rg −/− placentas had significantly less functional vascular space than WT ( P < .05, Figure 4 , D).
Two-dimensional (2-D) measurements of maximum placental thickness and diameter revealed additional placental differences ( Figure 4 , B). Rag2 −/− /Il2rg −/− were thinner than WT placentae at gestational day 14-16 and narrower than WT placentae at gestational day 10, 16, and 18 ( Figure 4 , E and F). Geometric shapes of growing placentae were compared as ratios of measured placental diameter and thickness. Ratios for Rag2 −/− /Il2rg −/− placentae decreased linearly from gestational day 10 to gestational day 18 but those for WT placentae formed a reversed parabolic curve with its lowest points at gestational day 12, 14 ( Figure 4 , G). These data indicate that WT placental growth is not linear with gestational age and that, in addition to their unusual hemodynamic patterns, Rag2 −/− /Il2rg −/− use different placental growth strategies along each axis.
Maternal body and organ mass in WT and Rag2 −/− /Il2rg −/− mice before, during, and after pregnancy
As published, Rag2 −/− /Il2rg −/− females are larger than age-matched WT before mating and over pregnancy (Δ7.9-12.8 g; P < .05; Figure 5 , A ; Supplementary Table 1 ). This may be due to advancement of adipocytye maturation caused by a lack of lymphocyte-derived interferon γ. Pregnancy weight gain was significant in WT females by gestational day 12 but by gestational day 10 in Rag2 −/− /Il2rg −/− females ( P < .05 compared with corresponding NP). By peripartum gestational day 18, WT dams had gained 10.4 g and Rag2 −/− /Il2rg −/− females 15.0 g (46% and 50% of prepregnancy weights, respectively).
