Maternal high-fat diet reversal improves placental hemodynamics in a nonhuman primate model of diet-induced obesity.

BACKGROUND: In a Japanese macaque model of diet-induced obesity, we have previously demonstrated that consumption of a high-fat, "Western-style" diet (WSD) is associated with placental dysfunction and adverse pregnancy outcomes, independent of an obese maternal phenotype. Specifically, we have reported decreased uterine placental blood flow and increased inflammation with maternal WSD consumption. We also previously investigated the use of a promising therapeutic intervention that mitigated the adverse placental effects of a WSD but had unexpected detrimental effects on fetal pancreatic development. Thus, the objective of the current study was to determine whether simple preconception diet reversal (REV) would improve placental function.
METHODS: Female Japanese macaques were divided into three groups: REV animals (n = 5) were switched from a chronic WSD (36% fat) to a low fat, CON diet (14% fat) prior to conception and throughout pregnancy. The CON (n = 6) and WSD (n = 6) cohorts were maintained on their respective diets throughout pregnancy. Maternal body weight and composition were regularly assessed and advanced noninvasive imaging was performed at midgestation (gestational day 90, G90, or 0.5 of gestation, where full term is G175), and G129, 1 day prior to C-section delivery at G130 (0.75 of gestation). Imaging studies comprised Doppler ultrasound (US), contrast-enhanced US, and dynamic contrast-enhanced magnetic resonance imaging to assess uteroplacental hemodynamics and maternal-side placental perfusion.
RESULTS: Dietary intervention resulted in significant maternal weight loss prior to pregnancy, and improved lean to fat mass ratio. By advanced imaging we demonstrated that a chronic WSD led to decreased blood flow velocity in the intervillous space, delayed blood flow transfer through the maternal spiral arteries, and reduced total placental blood flow compared to CON fed animals. Dietary reversal ameliorated these concerning derangements, restoring these hemodynamic parameters to CON levels.
CONCLUSIONS: Preconception dietary modification has beneficial effects on the maternal metabolic phenotype, and results in improved placental hemodynamics.

Introduction

Obesity is a major health issue in the United States, with over 35% of the population meeting criteria for obesity with a body mass index >30 kg/m2.[1] This epidemic has contributed significantly to the rising healthcare burden of coronary artery disease, diabetes and hypertension. Paralleling this national trend, obesity complicates up to 20% of pregnancies in the United States,[2] and is associated with multiple adverse pregnancy outcomes, including gestational hypertension, preeclampsia, gestational diabetes, and stillbirth.[2,3] The mechanisms underlying these increased risks are incompletely understood, but placental dysfunction has been implicated in many of these complications.[4-6] The placenta is the primary organ for nutrient exchange during pregnancy, and abnormal placental development has been associated with most adverse obstetric outcomes including abnormalities in fetal growth, preterm labor, preeclampsia, and stillbirth.[5,7,8] In addition, obese gravidae demonstrate impaired vascular function when compared with lean counterparts, which may contribute to their increased risk of preeclampsia and placental dysfunction.[9] Although the cause of obesity is multifactorial, poor diet is a major contributor, with the majority of Americans consuming a “Western-style diet” (WSD) high in calories and saturated fats.

We have previously demonstrated in our nonhuman primate (NHP) model of chronic excess nutrition that consumption of a WSD during pregnancy reduces uterine blood flow and increases the risk of stillbirth.[10,11] The deleterious effects of a WSD were independent of obesity, though the degree of placental dysfunction was greater in the animals that were obese and consuming a WSD. These results in the NHP support other studies in various animal models implicating obesity and overnutrition in placental dysfunction and poor obstetric outcomes.[12-14] In addition to the placenta, in utero WSD exposure also alters the normal development of key metabolic fetal organs in the early third trimester independent of maternal obesity, including the liver, brain, pancreas, and skeletal muscle.[15-18] In an effort to mitigate the deleterious effects of a WSD, our group previously investigated supplementation of the WSD with the polyphenol resveratrol as a potential therapeutic. Despite maternal and placental beneficial effects, we found an unexpected and concerning effect of resveratrol on fetal pancreatic development.[11] Thus alternative strategies are needed.

Unfortunately, the lack of noninvasive imaging modalities that facilitate the in vivo study of both normal and abnormal placental vascular perfusion impedes our understanding of placental vascular function and development, making the placenta the least understood human organ.[19] We have recently implemented advanced imaging methods to overcome the known limitations of standard Doppler ultrasound (D-US) to assess maternal-side placental blood flow. The first, contrast-enhanced ultrasound (CE-US), is a noninvasive technique that facilitates imaging of microvascular perfusion with the use of acoustic detection of gas-filled, lipid-encapsulated microbubble contrast agents.[20-22] The use of a contrast agent permits assessment of perfusion in small capillary networks that are difficult to assess solely with D-US. Utilizing this technology, we are able to calculate the microvascular flux rate constant, which is a measure of in-flow velocity, at the level of each individual maternal spiral artery. Although this is not an absolute measure of flow (volume/time), it is an estimate of impedance at the spiral artery. To measure total blood flow to the placenta, we have developed a Dynamic Contrast Enhanced Magnetic Resonance Imaging (DCE-MRI) method using a contrast agent that quantitatively measures blood flow to the placenta from all the spiral arteries filling the intervillous space.[23]

In this study, we investigated the impact of pre-pregnancy dietary reversal in NHPs on placental hemodynamics and blood flow kinetics using advanced imaging techniques. Since a WSD reduces uterine placental blood flow,[10] we hypothesized that switching NHPs from a chronic WSD to a standard chow diet prior to conception would alter intervillous blood flow hemodynamics and improve placental function.

Material and Methods

Non-human primate model

Female Japanese macaques (macaca fuscata, aged 6 – 15 years) were maintained on a WSD with 36% of calories from fat (Purina Mills, Inc., St. Louis, MO, USA) for at least 4 years (n=6). Control animals were maintained on a standard chow diet (CON, n=6) with 14% of calories from fat. The diet reversal cohort (REV, n=5) were maintained on a WSD for 4-7 years and then switched to a CON diet 3 months prior to the breeding season (Fig. 1A). All diets contained appropriate vitamin, mineral and protein content to support normal growth. Based on our prior published macaque studies,[11,24,25] a sample size of 5 or 6 animals per diet cohort was anticipated to provide adequate power for statistical analysis.

Figure 1

Study design and maternal metabolic parameters

A. Schematic representation of the study design indicating data collection time points with the proportion of gestation (0.5, 0.7 and 0.75) given in parentheses.

B. Body weight in all groups at preconception through the third trimester. C. Body weight in the diet reversal group at preconception (0=initiation of diet intervention), 3 months after dietary intervention, and third trimester (G130). D. Fat-to-lean mass ratio determined by dual-energy X-ray absorptiometry scan pre- and post-pregnancy. E. Cholesterol levels, F. Low density lipoprotein levels, and G. Triglyceride levels at G120. Data are presented as the mean ± standard error of the mean for each diet group (CON, n=6; WSD, n=6; REV, n=5). ap<0.001 REV vs. CON; bp<0.01 REV vs. CON, Kruskal-Wallis test; cp<0.013 months vs.0 months, Friedman test; dp<0.05 REV vs. CON; ep=0.007 WSD vs. CON, Kruskal-Wallis test.

G, gestational day; CON, control; WSD, Western-style diet; REV, diet reversal; trim, trimester; LDL, low density lipoprotein.

Animals were socially housed in groups of 8-10 females and 2 males per diet-determined cohorts in indoor/outdoor facilities with ad libitum access to food and water; thus animal selection was not randomized prior to the study. Animals bred naturally, and pregnancies were identified by monthly physicals during the breeding season. Gestational age was confirmed by ultrasonographic fetal biometry measurements in the early first trimester. All animal procedures were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee (IACUC) of the Oregon National Primate Research Center (ONPRC). The ONPRC abides by the Animal Welfare Act and Regulations enforced by the United States Department of Agriculture.

Maternal metabolic parameters

Maternal body weight was measured prior to pregnancy and at regular intervals throughout pregnancy. Body composition for each animal was determined by dual-energy X-ray absorptiometry (DEXA) scans (Hologic QDR Discovery A; Hologic, Inc., Bedford, MA, USA) prior to pregnancy and in the postpartum period. Total body scans were performed in supine position under Telazol sedation (3-5 mg/kg intramuscular; Fort Dodge Animal Health, Pfizer, NY, USA), and QDR software used to calculate body composition. Maternal whole blood was collected from the femoral artery at gestational day (G)120 (0.7 of gestation) after overnight food withdrawal, and plasma and serum were isolated by centrifugation. Serum cholesterol, low-density lipoprotein (LDL), and triglyceride measurements were assayed by the Endocrine Technology and Support Core Laboratory at the ONPRC using a Horiba ABX Pentra 400 Clinical Chemistry System (Horiba Medical, Irvine, CA, USA).

Imaging studies

Placental imaging was performed at G90 (0.5 of gestation) and G129 (0.75 of gestation, the early third trimester). Animals were initially sedated with 3-5 mg/kg of intramuscular Telazol, intubated and maintained under anesthesia with 1-2% isoflurane gas for the duration of the examination. Intravenous Flumazenil (APP Pharmaceuticals, Schaumburg, IL, USA) at 0.01 mg/kg was given following isoflurane administration to facilitate hemodynamic stabilization.[26] Animals were placed in dorsal recumbent position and underwent physiologic vital sign monitoring for the duration of the procedure. All animals experienced the same sedation protocol and remained hemodynamically stable throughout the imaging procedures. For each imaging study, the ultrasonographer (AEF) and MRI operator (MCS) were blinded to animal experimental group.

Doppler ultrasound

Image-directed pulsed and color Doppler equipment (GE Voluson 730 Expert, Kretztechnik, Zipf, Austria) with a 5- to 9-MHz sector probe was used. The lowest high-pass filter level (100 Hz), and an angle of 15° or less between the vessel and Doppler beam was considered acceptable. Uterine artery blood flow velocity waveforms were obtained from the proximal portion of the uterine artery as described previously.[10,11,27,28] The cross-sectional area (CSA) of the vessel was calculated as CSA = π(d/2)[2], where d=diameter. The calculated uterine artery volume blood flow (cQuta), which estimates blood flow through the uterine artery, was calculated as the velocity-time integral × CSA × fetal heart rate and normalized by maternal weight in kilograms. Pulsatility index (PI) was calculated from waveforms obtained from the uterine artery and the umbilical artery in a free loop of umbilical cord.

Contrast-enhanced ultrasound

Contrast-enhanced ultrasound was performed with a Sequoia System (Siemens Medical Systems, Mountain View, CA, USA) using a 15L8 transducer at 7MHz with a 0.18 resting mechanical index (MI), and a 55dB dynamic range as previously described.[21,22] In brief, microbubble contrast reagent composed of lipid-shelled octofluoropropane was prepared in 0.9% normal saline to a concentration of 5% (Definity, Lantheus Medical Imaging, Billerica, MA, USA), and administered by continuous intravenous infusion at 60 ml/hr in the left upper extremity. Individual spiral arteries supplying a placental cotyledon were identified and centered directly beneath the ultrasound probe. A 2-second acoustic pulse (MI 1.9) was applied to destroy the microbubbles in the acoustic beam path, and microbubble re-entry into the intervillous space (IVS) was recorded at 1 frame per 75 milliseconds (ms) for rapidly re-filling vessels, or 1 frame per 125 ms for slower re-filling vessels until the area of interest reached signal saturation. Three sequential replicate recordings were taken for each identified region of interest (ROI). Analysis was performed using a CEUS analysis program. Blood flow kinetics within the IVS were assessed with manually specified ROI encompassing individual cotyledons. A kinetic curve was generated by fitting the data to the function y=A(1-e−βt), where y is the video intensity (VI) at the pulsing interval t (ms), A is the VI plateau and β is the flux rate constant (FRC, s−1). The FRC represents the mean microbubble transfer rate within the IVS and therefore reflects the blood velocity.[20] The spiral artery transit time (SAT) was defined as the time from microbubble destruction to observed initial cotyledon filling, or the time required for blood to flow through the spiral artery to the IVS (ms).

Dynamic Contrast Enhanced Magnetic Resonance Imaging (DCE-MRI)

MRI studies were performed on a nonhuman primate-dedicated 3T Siemens TIM-Trio scanner (Erlangen, Germany) using a circularly-polarized (CP) transmit, 15-channel receive radiofrequency (RF) “extremity” coil (QED, Cleveland, OH) as previously detailed.[23] Following localization of the placenta and acquisition of T2-weighted half-Fourier acquisition single-shot turbo spin-echo (HASTE) anatomic images in the coronal and axial planes, axial 2D multislice spoiled gradient echo (SPGR) images spanning the entire uterus, were acquired at six in-phase echo times with monopolar readout gradients. Subsequently, 3D SPGR images were acquired in the coronal plane to allow estimation of T1 (longitudinal relaxation time) with the variable flip angle (VFA) method.[29] Immediately after acquisition of VFA data, 150 volumes of 3D SPGR images were acquired for DCE-MRI with field of view and resolution matched to the VFA images. Ten baseline images were acquired prior to intravenous injection of a standard dose of 0.1 mmol/kg of gadoteridol CR (Prohance, Bracco Diagnostics Inc, Princeton, NJ) at a rate of 30 mL/min using a syringe pump (Harvard Apparatus, Holliston, MA).[30,31] Anatomic and multiecho imaging was performed during expiratory breath holding, achieved by temporarily suspending ventilation, while DCE-MRI data were acquired during ventilated free breathing. Physiological monitoring of pulse rate, arterial blood oxygen saturation, and respiration rate and end-tidal CO2 partial pressure was performed throughout each MRI study.

Placental tissue collection

Cesarean section was performed at G130 (0.75 of gestation) in the ONPRC Surgical Services Unit. This time point is equivalent to the early third trimester and was chosen to facilitate optimal sample collection for multiple concurrent fetal experiments. The fetal weight and total weight of the two placental lobes was recorded. Individual placental cotyledons were weighed and their volume measured by saline displacement. Two full thickness (fetal membranes to maternal decidua) biopsy samples were obtained from each placental cotyledon, fixed in formalin and paraffin-embedded for stereological analysis. Villous tissue was collected from each individual placental cotyledon and flash frozen in liquid N2 for later analysis.

Placental stereology

Each fixed placental tissue sample was serially sectioned into two series of eight sections (5μm thickness) with 100μm of tissue discarded between the two series. Two sections per tissue sample (4 sections per cotyledon) were stained using standard hematoxylin and eosin (H&E) methods. Digital images of stained sections were obtained with a virtual microscopy slide scanner (Olympus, Center Valley, PA, USA). Placental tissue structure was assessed by stereological analysis using newCAST™ software (Visiopharm, Hoersholm, Denmark). Systematic uniform random sampling was employed to identify tissue areas for analysis within each slide, and a grid of randomly superimposed geometric point probes were utilized to count placental structures. Blinded stereology analysis was performed at 10× magnification, and volume fractions of the villi and IVS were determined.[32] Tissue shrinkage factor in fixed and processed tissues can be calculated using red blood cells as a size marker. This was not calculated in the current study, however all tissue sections were processed with standardized methods and fixation duration to reduce variability in data end points due to shrinkage.

Placental inflammation

Frozen placental villous tissue samples were homogenized in lysis buffer (20mM Tris with pH 7.5, 1mM EDTA, 1mM EGTA, 20mM sodium fluoride, 0.15M sodium chloride, 0.5% Nonidet P-40, 0.5% Triton X-100, 200μM sodium orthovanadate, 2μM leupeptin, 5.8μM pepstatin, 200μM 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride, and 5μM N-tosyl-L-lysine chloromethyl ketone) and centrifuged for 5 minutes at 20,000 × g to remove cell debris. Inflammatory markers were assessed with a commercially available antibody array (Abcam, Cambridge, MA, USA) using pooled placental samples from each dietary group. Enzyme-linked immunosorbent assay (ELISA) kits [TNFα, Life Technologies, Carlsbad, CA, USA; IL-12, Invitrogen Life Technologies, Grand Island, NY, USA; [regulated on activation, normal T-cell expressed and secreted (RANTES)], Thermo Scientific, Rockford, IL, USA; IL-1ß, Invitrogen Life Technologies, Grand Island, NY, USA] were used to assess selected inflammatory markers. Assays were performed according to the manufacturers’ instructions and expression levels derived from known standards. Reported data was corrected for protein content determined by BCA assay (Thermo Scientific, Rockford, IL, USA).

Statistical analysis

Data are expressed as mean ± standard error of the mean (SEM), and analysis performed using statistical analysis software (GraphPad Prism, La Jolla, CA). Differences in mean body weight at multiple time points within the REV group were compared using a Friedman test followed by Dunn’s multiple comparisons test. Differences between the three diet groups (CON, WSD, REV) were analyzed using a Kruskal-Wallis test followed by Dunn’s multiple comparisons test. Within each diet group, differences between G90 and G129 were compared using the Mann-Whitney test. A P-value <0.05 was considered statistically significant.

Results

Diet reversal improves the maternal phenotype

In the pre-conception period, the REV dams were significantly heavier than the CON dams (Fig. 1B, p<0.001). Following dietary intervention, the REV dams experienced a 20% weight loss (Fig. 1C), and were comparable in weight to the WSD dams at the time of conception (Fig. 1B). Appropriate weight gain was observed throughout pregnancy in all diet groups. By the third trimester, the REV dams remained comparable in weight to the WSD dams, although were significantly heavier than CON animals (p<0.01). Prior to pregnancy, the mean ratio of fat to lean mass was higher in REV dams compared to CON dams (Fig. 1D, p<0.01). Following pregnancy, the fat to lean mass in the REV cohort had decreased and was comparable to both WSD and CON animals.

Maternal serum cholesterol and LDL levels at 0.7 of gestation were comparable between diet groups (Fig. 1E-F). Maternal serum triglyceride levels were increased two-fold with WSD compared to CON (Fig. 1G) and reduced to control levels in the diet reversal cohort.

Diet reversal mitigates perturbed placental perfusion observed with WSD consumption

Total blood flow to both placental lobes was quantified by DCE-MRI at G90 and G129 (0.5 and 0.75 of gestation respectively). We observed a modest increase in total blood flow from mid-gestation to the early third trimester in all groups (Fig. 2A) but significantly diminished flow in the WSD animals at both time points compared to CON (p<0.05). Placental blood flow was restored to CON levels in the REV cohort (Fig. 2).

Figure 2

Dynamic Contrast-Enhanced MRI

A. Total placental blood flow quantified by DCE-MRI at gestational days 90 and 129 in CON (n=6), WSD (n=6) and REV (n=5) animals. Data are presented as mean ± standard error of the mean. ap<0.05 vs. G90 CON; bp<0.05 vs. G129 CON. Placental blood flow (measured in ml blood/ml placenta/minute) as quantified by DCE-MRI for an imaging section through the placenta in a single representative animal from each of our three groups: CON (B), WSD (C), and REV (D).

Additionally, uteroplacental hemodynamics were investigated using Doppler ultrasound at G90 and G129. The uterine artery PI, an estimate of vascular impedance, was stable throughout gestation in all diet groups, and was comparable between groups with no differences between CON, WSD and REV groups at either G90 or G129 (Fig. 3A). The mean umbilical artery PI did not change significantly from G90 to G129 in either the CON or REV cohort (Fig. 3B). However, we did detect a 42% decrease in PI from G90 to G129 in our WSD cohort. The umbilical artery PI did not differ between diet groups at either gestational time point. The calculated uterine artery volume blood flow (cQUta), normalized to maternal weight, did not change significantly from G90 to G129 in any of our cohorts (Fig. 3C). We were also unable to detect any significant differences in cQUta between our dietary groups at either G90 or G129 (Fig. 3A).

Figure 3

Doppler ultrasound

A. Uterine artery PI. B. Umbilical artery PI. C. Calculated uterine artery volume blood flow (cQuta) corrected for maternal body weight. ap<0.05 G129 vs. G90, Mann-Whitney test.

Open bars, G90; closed bars, G129; PI, pulsatility index.

Interrogation of spiral artery perfusion was achieved using Contrast-enhanced ultrasound. Analysis of kinetic curves allowed for the direct determination of the flux rate constant (FRC) and spiral artery transit time (SAT). Multiple cotyledons were analyzed in each animal study, and data reported as pooled values for each diet group at the two gestational time points. The total number of cotyledons analyzed per diet group at each time point were: CON G90 (n=28), CON G129 (n=26), WSD G90 (n=19), WSD G129 (n=28), REV G90 (n=23), REV G129 (n=25).

The FRC increased by 92% from G90 to G129 in CON dams (Fig. 4A, p<0.05). This effect was blunted in the WSD group, with a FRC at G90 that was comparable to CON dams, but no observed increase across gestation. Diet reversal resulted in a 69% increase in FRC compared to WSD at G90 (p<0.05), and a 2-fold increase in FRC at G129 (p<0.005).

Figure 4

Contrast-Enhanced Ultrasound

A. Perfusion rate of the intervillous space (PRIVS, s−1). B. Time to initial IVS perfusion (TIVSP, ms). ap<0.05 vs. G90, Mann-Whitney test; bp<0.001 vs. CON at G129; cp<0.05 vs. WSD at G90; dp<0.01 vs. WSD at G129; ep<0.01 vs. CON at G129; fp<0.001 vs. WSD at G129, Kruskal-Wallis test.

PRIVS, perfusion rate of the intervillous space; TIVSP, time from microbubble destruction to initial cotyledon filling.

The SAT was stable across gestation from G90 to G129 in all diet cohorts (Fig.4B). However, WSD resulted in prolongation of the SAT at G129 compared to CON dams. Diet reversal restored the SAT to CON levels at G129, with a 58% reduction in SAT compared to WSD dams.

Gross fetal and placental structural measures were not affected by WSD or diet reversal

The weights of fetal offspring taken at the time of necropsy (G130) did not differ between groups (Table 1). There was no difference in overall placental weight or volume between groups, and the mean number of placental cotyledons and mean cotyledon volume were similar. The feto:placental mass ratio, a measure of placental efficiency, did not differ between diet groups. Overall, mean villous and IVS volume were similar between diet groups (Table 1). Additionally, there were no differences in the absolute volume of fetal capillary or trophoblast in a subset of analyzed samples (n=3 animals/diet cohort, Table 1).

Table 1

Fetal and placental structural measurements.

Measurements	CON (n=6)	WSD (n=6)	REV (n=5)
Fetal weight (g)	340.8 ± 8.2	351.4 ± 8.0	367.2 ± 11.5
Placental weight (g)	106.3 ± 7.0	112.0 ±6.4	110.0 ± 5.3
Feto:placental weight ratio	3.1 ± 0.17	3.0 ± 0.17	3.4 ± 0.14
Placental volume (cm3)	96.0 ± 4.1	99.2 ± 4.1	96.2 ± 4.7
Mean cotyledon number	11.3 ± 0.9	11.5 ± 0.9	12.3 ± 1.4
Mean cotyledon volume (cm3)	9.5 ± 8.2	9.6 ± 8.2	8.5 ± 0.7
Intervillous space
Volume fraction (%)	36 ± 0.7	37 ± 0.7	36 ± 0.7
Absolute volume (mm3)	0.073 ± 0.003	0.075 ± 0.003	0.073 ± 0.004
Villous
Volume fraction (%)	64 ± 0.7	63 ± 0.7	63 ± 0.8
Absolute volume (mm3)	0.129 ± 0.0005	0.126 ± 0.0006	0.127 ± 0.0005
Fetal capillary absolute volume (mm3)	0.093 ± 0.018	0.100 ± 0.019	0.108 ± 0.034
Trophoblast absolute volume (mm3)	0.246 ± 0.038	0.207 ± 0.030	0.229 ± 0.009

Data are presented as mean ± standard error of the mean. CON, control; WSD, Western-style diet; REV, diet reversal.

A WSD resulted in overall upregulation of multiple placental inflammatory markers, including TNFα, IL-12, IL-1β and RANTES (Fig. 5). Diet reversal reduced placental inflammation to mean levels comparable to controls. Selected confirmatory comparisons by ELISA revealed similar trends that did not reach statistical significance (Fig. 5).

Figure 5

Placental inflammation

A. Protein expression levels of multiple inflammatory markers were assessed by antibody array. Protein expression was normalized to control levels. Inflammation was increased in the WSD cohort compared to controls, and decreased in the diet reversal cohort (*,

CON, control; WSD, Western-style diet; REV, diet reversal.

Discussion

Utilizing advanced imaging to overcome known limitations in assessment of placental function in vivo, in this study we demonstrate perturbations of placental blood flow hemodynamics that are mitigated by dietary reversal in our WSD model. Specifically, by DCE-MRI we demonstrated decreased total maternal-side placental blood flow. The advantage of DCE-MRI is that it measures tissue perfusion of the placenta directly by measuring blood from each individual spiral artery through the entire volume of the placenta. Importantly, this method overcomes utilizing calculated uterine artery volumetric blood flow (in a large maternal vessel proximal to the placenta) as a proxy for placental microvascular tissue perfusion. DCE-MRI confirms, in this smaller cohort, the prior observations of disrupted placental hemodynamics on the maternal side of the placenta with WSD.[10,11]

The CEUS measures of spiral artery hemodynamics provide additional insight by providing a method to interrogate transit time through the spiral arteries and in-flow velocity of blood flow to the placenta. Spiral artery transit time was increased with WSD which could be secondary to the decreased total blood flow measured by DCE-MRI. Similarly, decreased in-flow velocity was also apparent in the WSD cohort and improved with REV. While the mechanisms for disruption of these parameters at the level of spiral artery in-flow to the placenta are not known several possibilities exist and merit further investigation. First, these measures could be affected by the decrease in total perfusion in the placenta measured by DCE-MRI. Second, we could speculate that a chronic WSD and dietary reversal may result in structural differences at the level of the spiral artery (vessel caliber or tortuosity) and the IVS, which could affect the velocity of blood transfer detected by CEUS. One technical limitation of CEUS when used to interrogate the placenta is the inability to assess fetal-side placental blood flow which is an important consideration for nutrient uptake and gaseous exchange at the materno-fetal interface. The close apposition of the two circulations makes the fetal vasculature susceptible to altered maternal flow parameters[33,34], however our stereological analysis did not reveal differences in the villous volume or the proportion of fetal capillary and trophoblast within the placental tissue. This does not exclude the possibility that remodeling of maternal spiral arteries is adversely impacted by WSD exposure, and that changes in the placental villous architecture may be too subtle to detect using 2 dimensional stereological assessment. Third, chronic inflammation within the spiral artery, at the level of the trophoblast, or endothelial cell activation of the spiral arteries could affect blood flow as has been reported in other blood vessels in this model.[35] We demonstrate some evidence of increased inflammation in the current study, and have previously demonstrated increased expression of placental pro-inflammatory mediators with a WSD in the NHP.[10,11]

The Japanese macaque model of diet-induced obesity has been used for more than a decade and our collaborative group have investigated multiple aspects of maternal WSD consumption in relation to fetal and juvenile outcomes. We have demonstrated fetal liver lipotoxicity[15] and hepatic inflammation,[36] decreased oxidative metabolism in fetal muscle,[16] impaired islet vascularization[18] and increased pancreatic inflammation,[17] in addition to our reports of perturbed placental function and development.[10,11] The current study demonstrated positive effects of a dietary intervention on maternal body weight, composition and circulating triglycerides. This observation is clinically pertinent, as human studies have shown that a lower pre-pregnancy weight is associated with decreased adverse pregnancy outcomes.[8,37] Additionally, although the complex relationships between maternal obesity, overnutrition, pregnancy outcomes and offspring health are not well understood, maternal obesity and WSD consumption are causal factors for long-term risk of obesity and metabolic dysfunction in offspring.[38,39] Importantly, we have reported that impairments in fetal pancreatic development are mitigated by diet reversal,[18] reducing offspring susceptibility to type II diabetes in later life. Similarly, placental hemodynamic function was improved with dietary reversal despite persistence of a maternal obesity compared to controls.

In conclusion, here we show that diet reversal improves placental function in our nonhuman primate model of diet-induced obesity utilizing advanced imaging. Preconception diet reversal is a feasible and inexpensive intervention that addresses a modifiable risk factor for poor pregnancy outcomes, with suggested improvements in metabolic programming of the offspring.[18] Perhaps the greatest barrier to effective use is that only the minority of women present to an obstetric provider for preconception counseling, and the majority seek initial care in the middle of the first trimester, at which time rapid and significant weight loss is not recommended. Therefore, all women of reproductive age who present in any capacity to primary care should be counseled about the importance of maintaining a healthy diet and lifestyle, and the implications of poor diet and obesity on pregnancy outcomes. Furthermore, continued innovation in functional in vivo imaging will aid our understanding of placental development, and facilitate effective clinical interventions during at-risk pregnancies.