Atherosclerosis is a chronic inflammatory disease that has its origins in early life. Postnatal inflammation exacerbates atherosclerosis, but the possible effect of intrauterine inflammation is largely unexplored. Exposure to inflammation in utero is common, especially in infants born preterm, who have increased cardiovascular risk in adulthood. We hypothesised that exposure to inflammation before birth would accelerate the development of atherosclerosis, with the most severe atherosclerosis following exposure to both pre- and postnatal inflammation. Here we studied the effect of prenatal and postnatal inflammation on the development of atherosclerosis by combining established techniques for modelling histological chorioamnionitis and atherosclerosis using apolipoprotein E (ApoE) knockout mice. A single intra-amniotic (IA) injection of lipopolysaccharide (LPS) caused intrauterine inflammation, and increased atherosclerosis at 13 weeks of postnatal age. In mice exposed to postnatal LPS, chorioamnionitis modulated subsequent responses; atherosclerotic lesion size, number and severity were greatest for mice exposed to both intrauterine and postnatal inflammation, with a concomitant decrease in collagen content and increased inflammation of the atherosclerotic plaque. In conclusion, pre- and postnatal inflammation have additive and deleterious effects on the development of atherosclerosis in ApoE knockout mice. The findings are particularly relevant to preterm human infants, whose gestations are frequently complicated by chorioamnionitis and who are particularly susceptible to repeated postnatal infections. Human and mechanistic studies are warranted to guide preventative strategies.
Atherosclerosis is a chronic inflammatory process that underlies cardiovascular disease (CVD) . In adults, both acute and chronic inflammatory conditions, such as severe infections and inflammatory arthritis are associated with accelerated development of atherosclerosis and increased risk of cardiovascular morbidity and mortality [2–4]. Autopsy data indicate that atherosclerosis may be present in infancy and childhood [5,6].
Exposure to inflammation and/or infection in early life may increase the progression of atherosclerotic development and adult CVD risk . Although early pre-atherosclerotic lesions may regress during early life, they increase later in childhood, particularly in those with ongoing risk factors, such as inflammation .
Preterm birth and low birth weight are associated with increased CVD later in life  but the mechanisms underlying these epidemiological observations are poorly understood.
Histological chorioamnionitis (inflammation of the amniotic membrane and chorion) is common; it complicates up to 70% of preterm births  and 20% of term births . The effect of chorioamnionitis on atherosclerosis development in adulthood is unknown. Chorioamnionitis affects neonatal and postnatal immune function [11,12] and subsequently alters the risk of newborn and childhood infections , such as late-onset neonatal sepsis  and respiratory infections ; whether it impacts the development of cardiovascular risk due to modulation of postnatal immune function has not been reported.
Here we aimed to investigate the cardiovascular effects of prenatal and postnatal inflammation on atherosclerosis. We established an experimental model of intrauterine inflammation (histological chorioamnionitis) in ApoE−/− (apolipoprotein E) knockout mice. We compared the effects of prenatal (intra-amniotic (IA)) and postnatal (intraperitoneal (IP)) lipopolysaccharide (LPS), alone and in combination, on atherosclerosis development and severity. We hypothesised that combined intrauterine and postnatal inflammation would result in more severe atherosclerosis than either exposure alone.
Animals and experimental design
The protocols were approved by the Monash Medical Centre (MMC) Animal Ethics Committee A (approvals MMCA/2011/36 and MMCA/2011/37). Breeding colony and experimental ApoE−/− mice were housed in the MMC Specific Pathogen Free (SPF) animal facility and were kept under standard housing conditions (controlled light (12/12 h light/dark cycle), relative humidity (40–60%) and temperature (23 ± 1°C)). Two separate studies were conducted: a dose–response study for LPS-induced chorioamnionitis (i) and a follow-up atherosclerosis study (ii). Timed matings were set up overnight. Embryonic day (E)0.5 was defined as the morning when a vaginal plug was discovered.
Investigators administering prenatal IA or postnatal IP injections were blinded to treatment. Immunohistochemical staining was done in single batches for each antigen to minimise variation in staining. Each histological variable was measured by an individual investigator who was blinded to treatment.
Induction of intrauterine inflammation at E15.5
Mice were closely monitored daily for 3 days prior to and following surgery, for assessment of general health and well being, food and water intake. Surgery was performed on pregnant female mice at E15.5, for the IA injection of either LPS (to induce intrauterine inflammation) or saline (controls). All surgeries were performed under sterile conditions. Buprenorphine was subcutaneously (SC) injected to a maximum of 0.025 mg/kg, 30–40 min before surgery. Mice were anaesthetised using 3% isoflurane in 1 l/min oxygen via a nose-cone for the duration of surgery. Breathing rate was monitored throughout surgery. A midline incision (1–1.5 cm) was made in the maternal abdomen, the uterus was exposed using a cotton tip and a blunt probe, and the number of fetuses was noted. The uterus was rinsed with saline to keep it hydrated. IA injections were administered through the uterine wall to each sac using a 10-µl glass syringe and a 30-G needle. Care was taken to avoid contact between the needle and fetus by targeting pockets of fluid near the fetal forelimbs or behind the fetal back. Every amniotic sac within each dam received the same treatment. The uterus was then returned to the abdomen and isoflurane was reduced to 2%. The abdomen was sutured closed in two layers, using synthetic suture material (size 5-0 polypropylene) and a simple continuous suture technique. Saline was injected SC to a maximum of 5 ml for mouse hydration and the dam was allowed to recover. Povidone iodine was applied to the surgical site before returning the dam into its cage. Once the mouse had righted itself, was breathing regularly and with sufficient depth, the mouse was placed in its cage on a heat pad with mashed chow. Mice were closely monitored until they had completely recovered from the anaesthetic. Post-operative analgesia (buprenorphine, 0.025 mg/kg, SC) was administered at 12 and 24 h after surgery. Mouse body condition, behaviour and the wound were checked at least three times daily, for 3 days following the operation.
In the first study, during surgery, the mice were either left untreated (n=1 dam) or treated with 5 μl of sterile physiological saline (IA; n=3 dams), or a solution of 0.1, 1 or 2.5 μg LPS (Escherichia coli 055:B5; Sigma L4524, Sigma–Aldrich, St. Louis, MO) in 5 μl saline (IA; n=4 dams per dose), injected into each amniotic sac.
In the second study, either 1 μg LPS in 5 μl saline (the optimal dose to induce IA inflammation) or 5 μl saline alone was used for IA treatment. The dams were randomised to receive either LPS or saline, with all sacs within each litter receiving the same treatment.
Confirmation of placental inflammation at E17.5
In the first study, placental tissue was collected at E17.5 to confirm placental inflammation 2 days after IA LPS injection. Dams were anaesthetised using isoflurane, the uterus exteriorised and each fetus was removed, weighed and killed by decapitation. Once all fetuses and placentae had been collected, the dams were humanely killed, without recovery from anaesthesia, by cervical dislocation. Placental tissue from each fetus was then immersion-fixed in 4% paraformaldehyde (PFA) in preparation for histological assessment of intrauterine inflammation.
Induction of postnatal inflammation
In the second study, pregnant females were separated based on the IA injection they received and allowed to give birth to their litters undisturbed. At 4 weeks of age, pups were separated from the dam and placed in different same-sex, same-age and same-dam cages. Weaned pups were housed in the SPF animal facility and were fed a standard mouse diet. From 8 weeks of age, pups received weekly IP injections of either 50 µg LPS in 0.05 ml saline (1 mg/ml) or 0.05 ml of saline. LPS and saline IP injections were administered once every week for 5 weeks. At 13 weeks of age, the pups were humanely killed for tissue collection.
Blood was drawn from the inferior vena cava into a heparin-saline-flushed syringe for plasma lipid analysis as described below. The heart and the aortic branches were removed en bloc and rinsed with saline. Fat that surrounded the heart and aortic branches was removed. The heart was divided into two by a cross-sectional cut one-third proximally. Both parts were fixed in formalin and the upper part was used for aortic sinus sectioning. The brain was removed and weighed immediately.
Sections and slide preparation
The paraffin-embedded upper portion of the heart was sectioned at 5-μm until at least two leaflets of the aortic sinus were observed under a microscope. Subsequent sections were then cut and collected for analysis. Slides were washed three times with xylene for 5 min each and then rehydrated twice in 100 and 70% ethanol for 5 min each, prior to histological staining or immunohistochemical analysis.
Haematoxylin and Eosin staining
Slides were incubated with filtered Mayer’s Hematoxylin Solution (Amber Scientific) for 3 min and rinsed three times with distilled water. They were then counterstained with Eosin (1% Aqueous Solution, Amber Scientific) and dehydrated with 70 and 100% ethanol twice, and subsequently washed with xylene thrice. Finally, the slides were covered with coverslips using DPX and left for 24 h to dry. Dried slides were scanned and observed for lesion stage. Determination of lesion stage (I–V) was supported by Masson’s Trichrome staining, described below. Lesion size was measured by subtracting the area of aortic sinus with the lesion, from the total area of aortic sinus.
Masson’s Trichrome staining (aortic sinus)
Fibrous cap development was visualised by staining tissue with Masson’s Trichrome. Images were obtained and analysed using Fiji (ImageJ, version 2.0) to detect collagen (stained blue). Nuclei of cardiac muscle cells stained purple and were cropped out in Fiji. Image analysis was divided into two parts, aortic sinus area measurement and fibrous cap development. Total lesion area was obtained by subtracting the area of the image measured with lesion, from the area of the image without lesion.
Picrosirius Red analysis of collagen content
Collagen content was analysed by measuring area of collagen density using Picrosirius Red, as previously described . Tissue sections were prepared, and then stained for 30 min with Sirius Red (0.1% of Sirius Red in saturated aqueous picric acid). Sections were then mounted for observation under light microscopy. Area of staining density was obtained by subtracting the area of the image with collagen staining, from the total area of aortic sinus.
Analysis of lesion stage
Atherosclerosis stage was determined based on the criteria provided by The Committee on Vascular Lesions, American Heart Association . In short, stage II plaques are characterised by the presence of isolated foam cells. In Stage III, groups of foam cells are detected. Stage IV is characterised by the formation of a lipid core, whereas in stage V, the lipid core is accompanied by a fibrous cap.
Histological assessment of intrauterine inflammation at E17.5
CD45 immunohistochemistry was performed on PFA-fixed paraffin-embedded sections of mouse placenta removed at E17.5 days of gestation. Paraffin wax was removed and sections were rehydrated in xylene and a series of ethanol and distilled water. Antigen retrieval was performed using citrate buffer (10 mM Sodium Citrate, pH 6.0) for 20 min at 98°C and left to cool for 40 min. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 15 min. Non-specific binding was blocked with 4% BSA in Tris-Buffered Saline (TBS) for 30 min at room temperature (RT). Slides were incubated with rat α-mouse CD45 primary antibody (1:100 in Dako antibody diluent; BD Pharmigen #550539) overnight at 4°C. After washing in TBS, the slides were incubated with secondary antibody biotinylated rabbit anti-rat IgG (1:200 in Dako antibody diluent, VECTOR, BA-4000) for 1 h at RT. Streptavidin-HRP reagent (1:200) was added for 45 min at RT. Sections were stained using DAB for 3 min (diluted 1:10 in milliQ water, MP, Cat# 980681) and then counterstained with Mayer’s Haemotoxylin. Sections were then dehydrated in a series of ethanol and xylene solutions and the slides were coverslipped with DPX.
From each placenta, two 5-μm sections of placenta were used, with each section separated by at least 50 μm. Five random fields of view (FOV) from each section were used for the analysis. CD45-labelled cells were identified by brown immunostaining, and counted by eye within the fetal zone in each FOV. The counts were averaged for each animal.
Histological assessment of CD45 and CD163 in the plaques
White blood cells in aortic sinus lesions were identified by CD45 and CD163 immunohistochemistry, followed by Haematoxylin staining, to visualise the nucleus and foam cells (macrophages). Slides were deparaffinised and washed with 1× TBS thrice. CD45 labelling was performed as described above. For CD163 labelling, slides were rehydrated using xylene and ethanol, and washed with distilled water. Antigen retrieval was performed by heating slides in citrate buffer (0.01 M, pH = 6.0) for 15 min with a subsequent cooling period of 10 min, then washing them with PBS thrice. Slides were blocked with 1.5% normal goat serum for 1 h at RT. Primary antibody (rabbit anti mouse CD163, EPR19518, Abcam ab182422) was added at 1:250 and incubated at 4°C overnight. After thorough washing, slides were incubated in 3% hydrogen peroxide in PBS to reduce endogenous peroxidase activity. After washing in PBS, biotinylated secondary antibody (VECTASTAIN ABC kit VEPK-4001 anti-Rabbit IgG) was added at 1:250 and incubated for 1 h at RT. After washing, ABC reagent was added and incubated for 30 min at RT. After washing, DAB was added until the colour changed (approximately 6 min). Samples were washed in running water and rinsed in distilled water and counterstained with Harris’ Haematoxylin for 4 min. After rinsing in distilled water, the slides were incubated for 30 s in Scott’s tap water and rinsed again. Slides were dehydrated through ethanol and covered with coverslips. The total number of CD45- and CD163-positive cells in the tissue area of aortic sinus were counted.
Plasma was collected by centrifuging whole blood at 1200 rpm for 8 min. Plasma total cholesterol and triglycerides were measured using a Cholesterol E resp Triglyceride E test according to the manufacturer’s instructions (Wako, 439-17501 resp 432-40201). High-density lipoprotein (HDL) cholesterol was first isolated by adding 1/10 Dextran solution to the plasma (1:1 1 M MgCl2 and Dextraslip 50 (Sigma, D8787-1G) 20 g/l (2% solution)). The sample was centrifuged at maximum speed (30 min, 4°C) and cholesterol levels of the supernatant were subsequently measured using the Cholesterol E test as described above.
Data were analysed using one-way analysis of covariance (ANCOVA) with sex and pregnancy as covariates. Differences in values between sexes were analysed using an unpaired Student’s t test or one-way ANOVA with Bonferroni post-test where indicated. The findings were considered statistically significant at P≤0.05. Data were checked for normality using normality tests in GraphPad or SPSS. Data were analysed using IBM® SPSS® Statistics version 20 (IBM Corporation, 2011) and presented as indicated. Data were illustrated using GraphPad Prism® version 7.02 (Prism 7 for Windows/Mac, v7.02, U.S.A.).
Induction of chorioamnionitis in ApoE mice
Histological chorioamnionitis can be modelled in animals by inducing localised intrauterine inflammation by the IA injection of LPS. A chorioamnionitic model in an atherosclerosis-prone mouse (e.g. ApoE−/−) has not been previously reported. The effects of intrauterine inflammation induced vary by IA LPS dose and gestational age . Therefore, we first established the optimal dose of IA LPS to induce chorioamnionitis in ApoE−/− mice, comparing 0.1, 1 and 2.5 μg LPS injected (IA) at E15.5. Placental CD45+ cells were stained (Figure 1A) and counted (Figure 1B), confirming a dose-dependent increase in intrauterine inflammation. The highest LPS dose (2.5 μg) resulted in the greatest placental inflammation, but also in fetal death in 50% of mice. Therefore, we chose a dose of 1 μg of LPS as this induced significant intrauterine inflammation but did not cause preterm delivery or fetal death.
Successful induction of chorioamnionitis in ApoE−/− mice
Pre- and postnatal inflammation does not affect survival, body weight and plasma lipids but decreases brain weight
Forty-eight pregnant ApoE−/− dams received an IA injection of LPS or saline. A total of 133 pups were weaned at 4 weeks and received IP injections of LPS or saline, resulting in four treatment groups: Sal/Sal; LPS/Sal; Sal/LPS; LPS/LPS (Figure 2A). Survival of mice exposed to prenatal and/or postnatal LPS treatment did not differ from that of saline-treated animals (data not shown). Body weight at week 13 was comparable in all groups (Figure 2B) and was not affected by prenatal and/or postnatal inflammation.
Baseline characteristics of ApoE−/− pups exposed to pre- and/or postnatal inflammation
Males had higher body weights than females, irrespective of treatment (Figure 2B). Plasma lipids (total cholesterol, triglycerides and HDL cholesterol) did not differ between groups (Supplementary Figure S1A) but males had significantly higher total cholesterol, triglyceride and HDL levels compared with females (Supplementary Figure S1B). There was evidence of an interactive effect of pre- and postnatal inflammation reducing brain growth (Figure 2C, P<0.01), confirming previous murine prenatal inflammation models [19,20].
Chorioamnionitis and postnatal inflammation act synergistically to increase atherosclerosis lesion size in mice
Pre- and postnatal LPS increased the size of atherosclerotic lesions in the aortic sinus (Figure 3A). In mice receiving saline prenatally, postnatal treatment with LPS (Sal/LPS) resulted in an increased lesion size compared with Sal/Sal treated animals (mean area (±SD) density 11.8 ± 0.9 vs. 7.6 ± 0.6% respectively; Figure 3B, P=0.01). Similarly, in mice that received prenatal LPS, larger lesions developed in the LPS/LPS group compared with the LPS/Sal group (19.3 ± 1 vs. 9.5 ± 0.6% respectively; Figure 3B). Prenatal LPS increased lesion size compared with prenatal saline, although the effect was less marked (9.5 ± 0.6% in LPS/Sal vs. 7.6 ± 0.6% in Sal/Sal). Overall, the combination of both pre- and postnatal LPS (LPS/LPS) resulted in the largest atherosclerotic lesions (P<0.005 compared with Sal/Sal). Lesion size was higher in female compared with male mice (Figure 3C), with the greatest difference in the LPS/LPS treated group (P=0.05).
Chorioamnionitis with subsequent postnatal inflammation significantly increases atherosclerotic lesion size
Pre- and postnatal inflammation increase lesion numbers and the severity of atherosclerotic plaques
Pre- and postnatal inflammation were associated with an increased number of lesions in aortic sinus sections (Figure 4A). There were more lesions in the aortic sinus of the Sal/LPS group (3.5 ± 0.3, P<0.005) and LPS/Sal group (2.5 ± 0.3, P=0.01) compared with the Sal/Sal treated animals (1.3 ± 0.2, Figure 4A). The LPS/LPS group also had more lesions than the Sal/Sal group (2.8 ± 0.3 vs. 1.3 ± 0.2, P<0.005, Figure 4A). There was no difference in lesion number between the LPS/LPS and the LPS/Sal group. Female mice in the LPS/Sal and Sal/LPS groups had a higher number of lesions than male mice (Figure 4B, P=0.05).
Pre- and postnatal inflammation interact to increase atherosclerotic lesion number and severity in ApoE−/− mice
The severity of atherosclerotic lesions (Figure 4C) in the aortic sinus was greatest following exposure to both pre- and postnatal LPS, with more stage IV and V lesions in the LPS/LPS group than in the other groups (Figure 4D). Female mice had more severe lesion staging than male mice (Supplementary Figure S1C).
Collagen deposition in the aortic sinus lesions was consistent with lesion staging; collagen density was significantly reduced by pre- and postnatal inflammation (Figure 4E, P=0.05). Females generally had lower aortic sinus collagen content than males (Supplementary Figure S1D).
Inflammation in aortic sinus sections was measured by counting cells positive for CD45, a general leucocyte marker, and CD163, a macrophage marker. There was an increased abundance of CD45+ cells in the aortic sinus of LPS/LPS mice compared with all other groups (Figure 4F,G, P<0.005). Sal/LPS mice also showed increased CD45 staining, although less pronounced than the LPS/LPS group. There were fewer CD163-expressing macrophages in pre- and/or postnatal LPS treatment groups, with the lowest number of CD163+ macrophages in the LPS/LPS group (Figure 4H, P<0.005). The proportion of CD163+ relative to CD45+ cells (CD163/CD45) was reduced after both pre- and postnatal inflammation (Figure 4I). Furthermore, female mice had fewer CD163+ cells in the aortic sinus than males in both groups treated with postnatal LPS (Supplementary Figure S1E).
Using a model of histological chorioamnionitis in ApoE−/− mice, we showed that pre- and postnatal inflammation result in more severe atherosclerosis in adulthood, characterised by increased atherosclerotic plaque lesion size, number and severity. Furthermore, a combination of pre- and postnatal inflammation induced the most severe atherosclerotic lesions, with the lowest collagen density and greatest inflammatory infiltrate. This increase in atherosclerotic severity was not accompanied by alterations in plasma lipids, suggesting that chorioamnionitis with subsequent postnatal inflammation exacerbates atherosclerosis development via immune mechanisms. Whereas the effect of postnatal LPS on atherosclerosis development is consistent with previous reports [3,21,22], our observation of an exacerbation of atherosclerosis by a prenatal pro-inflammatory insult is novel and has important implications.
The developing cardiovascular system is known to be affected by chorioamnionitis [6,9,13,23–25], but there are few data on the possible effects of chorioamnionitis on atherosclerosis development per se. In a small prospective human study, histologically diagnosed chorioamnionitis was associated with increased cord blood inflammation, but not with differences in aortic intima–media thickness in early life . A retrospective study demonstrated inflammatory infiltrates and reduced smooth muscle actin in the umbilical artery from infants exposed to severe chorioamnionitis .
The role of inflammation in atherosclerosis is widely recognised. Postnatal pro-inflammatory insults, such as infections, are robustly associated with atherosclerosis and CVD [2,26,27]. The CANTOS trial showed that in humans at increased risk for cardiovascular events, reducing systemic inflammation by inhibition of interleukin-1β reduced CVD risk independently from plasma lipid levels . Here we showed that pre- and postnatal inflammation increased inflammatory infiltrates in the plaque, with increased CD45+ cell infiltrates in mice exposed to pre- and postnatal LPS. Conversely, the number of CD163+ cells was reduced following pre- and postnatal inflammation. CD163 has been long recognised as a marker for alternatively activated macrophages , which have anti-atherogenic properties. Recent studies however suggest that CD163 can also be pro-atherogenic, promoting angiogenesis, vessel leakiness and inflammation . Thus, further analysis of macrophage phenotypes in the plaque is warranted. The increased lesion size, number, staging and decreased collagen content consistently indicate a worsening of atherosclerosis, which is generally associated with a pro-inflammatory phenotype .
The population of ex-preterm infants is increasing. There are approximately 20 million children in Europe and North America who were born preterm. This burgeoning population of ex-preterm individuals may contribute to the observed increase in cardiovascular risk in childhood , adolescence  and adulthood , and the increasing prevalence of morbidities including hypertension  and stroke . The mechanisms underlying these epidemiological observations are poorly understood. Chorioamnionitis is particularly common in pregnancies resulting in preterm delivery [9,35], but also occurs in up to 20% of those progressing to term . Preterm infants are particularly susceptible to infections throughout childhood . The human preterm population is therefore likely exposed to increased chorioamnionitis and postnatal infections, analogous to the pre- and postnatal inflammatory stimuli used in our experiment.
A recent paradigm shift in immunology is that the innate immune system can ‘remember’ previous exposures, resulting in an increased response on subsequent stimulation. This ‘trained innate immunity’ or ‘innate immune memory’ can be demonstrated by a second stimulation that may be related or unrelated to the first . We recently hypothesised that the association between infections and atherosclerosis could be explained by trained immunity . It is plausible that this mechanism underlies our observation of the combined effect of intrauterine and postnatal inflammation and warrants further investigation. Recent work from Yellowhair et al.  in rats showed that IA LPS injection after 1 h of umbilical artery ligation primes the immune system and induces an increased peripheral lymphocyte response upon ex vivo stimulation. Although the experimental intervention and species used is different from ours, this study confirms the capacity of prenatal pro-inflammatory insults to alter subsequent immune function.
Some strengths and limitations are acknowledged. We established an animal model of chorioamnionitis in ApoE−/− mice, a transgenic strain that is widely used to study atherosclerosis . However, ApoE−/− mice are not ideal for studies on inflammation since ApoE is expressed on several immune cells . Furthermore, female ApoE−/− mice are known to develop more severe atherosclerosis than males , as we observed, whereas in humans opposite sex differences exist . Notwithstanding, the effects of intrauterine and postnatal inflammation on atherosclerosis was evident in both sexes. We chose not to feed our mice a high fat ‘Western’ diet, which would have exacerbated atherosclerosis further, reasoning that it may obscure differences due to the experimental treatments that were the focus of our investigation. Further human studies on the effects of pre- and postnatal inflammation on atherosclerosis development are warranted.
Histological chorioamnionitis did not affect fetal or postnatal survival, or body weight in our experiment. These observations are consistent with the outcomes of human  and experimental animal data . In recent human cohorts, rates of small for gestational age births were not increased by histological chorioamnionitis [46,47]. Although small reductions in fetal growth have been reported for at least one large historical cohort , this study did not present growth parameters contingent on precise gestational age, which makes interpretation difficult. Experimental animal models where LPS is injected systemically to the mother or within the uterus but outside the amniotic sacs report effects on fetal growth [49,50], but differences with methodology of the current study and with the pathogenesis of chorioamnionitis in humans make direct comparison difficult. Importantly, postnatal growth to 2 years of age is not altered in children born after histological chorioamnionitis . Thus, we consider our findings implicate intrauterine inflammation per se as responsible for accelerating atherogenesis, independently from established associations between fetal growth  or maternal hypercholesterolaemia  and atherosclerosis.
In conclusion, we established a murine model of experimental histological chorioamnionitis in ApoE−/− mice, which is associated with increased postnatal atherosclerosis in offspring. Pre- and postnatal inflammation had additive and deleterious effects on the development of atherosclerosis in this model. Human and mechanistic studies are indicated to guide preventative and treatment strategies to reduce the burden of CVD.
Exposure to inflammation in utero is common, especially in infants born preterm, who have increased cardiovascular risk in adulthood.
Pre- and postnatal inflammation have additive and deleterious effects on the development of atherosclerosis in ApoE−/− mice.
The findings are particularly relevant to human preterm infants, whose pregnancies are frequently complicated by chorioamnionitis and are particularly susceptible to repeated postnatal infections.
We would like to thank Louise O’Connor and Michelle Scurr for CD163 staining, Nikeh Shariatian for performing plasma assays and Christian Dees for CD45 placental staining.
The authors declare that there are no competing interests associated with the manuscript.
This work was supported by the National Health and Medical Research Council of Australia: Program [grant number 606789], Centre for Research Excellence [grant number 1057514], Research Fellowships [grant numbers 1043294 (to T.J.M.), 1064629 (to D.P.B.)]; the National Heart Foundation Australia [grant number G12M6422]; the Victorian Government’s Operational Infrastructure Support Program; the Rubicon grant from the Dutch Scientific Organisation (NWO) [grant number 452173113 (to S.B.)]; the Post-graduate Biomedical Scholarship of the National Heart Foundation (Australia) [grant number PB12M6953]; and an Honorary Future Fellow of the National Heart Foundation (Australia) [grant number 100026 (to D.P.B.)].
Conceptualisation: D.P.B. and T.J.M. Methodology: T.J.M., H.L. and M.J.W. Validation: T.J.M., H.L. and M.J.W. Formal analysis: S.B., A.P.L. and M.U.N. Investigation: A.P.L, M.U.N., L.K.W., H.L., T.R.M. and M.J.W. Resources: T.J.M. and D.P.B. Writing – original draft preparation: S.B. Writing – review and editing: D.P.B., T.J.M., M.J.W., S.P., H.L., A.P.L., M.U.N. and L.K.W. Visualisation: S.B. Supervision: M.M.C., D.P.B. and T.J.M. Funding acquisition: T.J.M. and D.P.B.
These authors contributed equally to this work.