Lipopolysaccharide-Induced Chorioamnionitis Induces Fetal Intestinal Injury and Affects Gut Colonization in Rats

Background: Chorioamnionitis is associated with an increased risk of multiple adverse outcomes in offspring, especially neonatal necrotizing enterocolitis (NEC), which is one of the serious gastrointestinal diseases in neonates. However, the underlying mechanism remains undened. We hypothesize that lipopolysaccharide (LPS)-induced chorioamnionitis causes intestinal injury in offspring, thereby affecting the composition of the intestinal microbiome. Methods: Pregnant Sprague Dawley rats were received intraperitoneal injections with 700 μg/kg Lipopolysaccharide (LPS) or saline at 15 days of gestation. Pups were allowed to deliver naturally, and euthanized at days 0,3 and 7 after birth. Intestinal tissue and faeces samples from offspring were collected to evaluate the effects of intrauterine infection on intestinal ora colonization and intestinal mucosal development. Results: Signicant intestinal injury of the offspring induced by prenatal LPS exposure was observed at day 0 and 3 after birth.. In addition, prenatal LPS exposure also induced signicant changes in the intestinal microbiome of the offspring with a signicant increase in Proteobacteria (Escherichia-Shigella) and a decrease in Firmicutes at 7 days after birth . Conclusions: Thus, our ndings suggest that LPS-induced chorioamnionitis induces intestinal injury in offspring and subsequently leads to NEC-like changes in the composition of the intestinal microbiome. ,and


Background
Chorioamnionitis is used to refer to the intrauterine infection/in ammation, which includes amniotic uid, placenta (decidua, chorion and amniotic membrane) and fetal infection caused by pathogenic microorganisms invading the amniotic cavity [1]. Intrauterine infection is linked with adverse neonatal outcomes including neonatal early-onset sepsis, neonatal necrotizing enterocolitis(NEC), patent ductus arteriosus, neonatal respiratory distress syndrome, bronchopulmonary dysplasia, cerebral palsy ,and hearing impairment [2][3][4][5][6][7]. Among them, intestinal injury, especially NEC is a serious gastrointestinal disease during the neonatal period, and the mortality rate is as high as 20%-30% [8]. Several studies have reported that intrauterine infection is associated with an increased risk of NEC [6,9].However, the mechanisms linking intrauterine infections to neonatal intestinal injury remain unclear,and how such intestinal injury of the newborns develops into NEC has not been elucidated.
Several animal studies have demonstrated that intrauterine infections prevent fetal intestinal development during pregnancy [10],leading to the disruption of the tight junctional protein zonula occludens protein 1 (ZO-1) ,the increase of intestinal fatty acid-binding protein in serum [10][11][12] the loss of epithelial cell integrity, which further impaires epithelial differentiation. Research have shown that intrauterine infections cause direct injury to the placenta but indirect injury to the fetal intestine [13] based on the ndings that lipopolysaccharide(LPS) cannot pass through the placenta. Furthermore, placental microbiome is unable to be detected in a model of LPS-induced intrauterine infection. It is widely recognized that fetal gastrointestinal tract, amniotic uid and placenta are sterile [14],and microorganisms quickly colonized in newborns only after delivery,even if there is still controversy [15,16]. Therefore, whether the fetal intestinal injury caused by intrauterine infections have an in uence on the initial colonization of gastrointestinal microbes in infants is of great signi cance. Studies have proved that immature or damage of the intestinal epithelial barrier caused by premature birth or antibiotic interference is associated with abnormal colonization of the intestinal ora [14,17].Intestinal ora disorders are associated with the onset of NEC, presenting a bloom of Proteobacteria, speci cally Enterobacteriacae prior to NEC [18][19][20]. These results suggest that aberrant gut colonization and intestinal injury may have an important impact on the development of NEC.
We hypothesize that LPS-induced chorioamnionitis causes intestinal injury and subsequent affects the composition of the intestinal microbiome.To better understand the association between intrauterine infection and subsequent neonatal intestinal injury, and to explore the potential effects on the composition of the intestinal microbiome, we investigated the changes of the intestinal tissue structure, the intestinal epithelial barrier,and the composition of intestinal microbiome after establishing a LPSinduced rat model for intrauterine infection on.

Animals and sample collection
All animal studies were approved by the animal ethics committee of Guangxi Medical University(Guangxi,China) and in accordance with ARRIVE guidelines.The Sprague Dawley rats purchased from the Animal Experimental Center of Guangxi Medical University were used.All rats were housed with a 12-hour light cycle and free to get food and water during both housing and experimental time. At the gestation of day 15, pregnancy rats were randomly assigned to two groups, receiving a single 700 µg/kg intraperitoneal injection of LPS (Escherichia coli 055:B5;Sigma-Aldrich,St.Louis,MO) or the equivalent volume of saline for sham controls,respectively. All pregnancy rats were allowed to delivered naturally. Pups were allowed to stay together with their mothers after birth and allowed to breast-feed ad libitum. Pups were weighed and then euthanized by CO 2 inhalation at days 0,3 and 7 after birth. The terminal ileum was collected and xed overnight in 10% formalin solution for immunohistochemical stainings and histology stainings. Feces were collected in a sterile container and stored at -80 ℃ until processing.

Page 4/15
The formalin-xed terminal ileum tissues were embedded in para n and cut into 4 µm thick sections.
Intestinal ileum sections were stained with hematoxylin and eosin(H&E). Morphology changes in the intestinal ileum were assessed microscopically by a single investigator in a blinded manner [21]. The injury scores was as follows: Grade 0, normal mucosal villi; Grade 1, the subepithelial Gruenhagen's spaces developed with capillary congestion; Grade 2, extension of the subepithelial space with moderate lifting of epithelial layer from the lamina propria; Grade 3, massive epithelial lifting down the sides of villi;Grade 4, denuded villi with lamina propria and dilated capillaries exposed; and Grade 5, digestion and disintegration of lamina propria; hemorrhage and ulceration.

Immunohistochemistry
For staining of ZO-1 expressing cells, the ileum sections were boiled under high pressure in sodium-citrate buffer (pH 6.0) for 10 minutes for antigen retrieval. Subsequently, endogenous peroxidase activity was blocked by incubating with 3% H 2 O 2 for 10 minutes. Nonspeci c binding sites was blocked by normal goat serum for 30 minutes at room temperature. Thereafter, the slides were incubated with anti-ZO-1 antibodies for overnight at 4 ℃. After washing, goat anti-rabbit biotin-conjugated secondary antibody were added to the sections and incubated for 10 minutes at room temperature. ZO-1 antibodies were recognized with streptavidin-biotin method (Zsbio ,Peking, China) and visualized with nickel-DAB. The nuclei were counterstained with hematoxylin. Slides were observed using a light microscope and scaned using the cellSens imaging system(Olympus,Tokyo,Japan). Stain-positive average optical density(AOD) were measured in 5 high power elds using Image J(1.52b software,National Institutes of Health, Bethesda, MD, USA). The average values of AOD in 5 high power elds at the end of the ileum of each animal were used for the analyses.

Fecal microbiome analysis:
Deoxyribonucleic acid(DNA)was isolated from feces samples using the E.Z.N.A.Stool DNA Kit (Omega, Inc, USA) according to the manufacturer's instructions. The extracted DNA samples was stored at -20℃ until analysed. The V3-V4 hypervariable regions of the bacterial 16S ribosomal ribonucleic acid (rRNA) gene were ampli ed using the following primers: 338F(5'-ACTCCTACGGGAGGCAGCAG-3') and 806R: (5'-GGACTACHVGGGTWTCTAAT-3').Polymerase chain reaction(PCR) was performed using the following 20 µl reactions: 10 ng of template DNA, 0.8 µl of forward primer(5 µM), 0.8 µl of reverse primer(5 µM), 2 µl of dNTPs (2.5 mM), 4 µl of 5 × FastPfu Buffer, 0.4 µl of FastPfu Polymerase, and deionized ultrapure water to adjust the volume. The PCR programme included an initial denaturation at 95 ° C for 3 minutes, then 27 cycles of 95 °C for 30 seconds,55 °C for 30 seconds, 72 °C for 45seconds, and the nal extension at 72˚C for 10 minutes. There were 3 replicates per sample, and the PCR products of the same sample were mixed and detected by 2% agarose gel electrophoresis, followed by puri cation using AxyPrep DNA Gel Extraction Kit(Axygen,USA) according to the manufacturer's protocols. The puri ed PCR products In order to obtain high-quality sequences ,FLASH and Trimmomatic were used to quality lter and optimize the raw sequences according to the ltering criteria as previously described [22] .The highquality sequences with 97% similarity were clustered to the same operational taxonomic units (OTUs) by Usearch(vsesion 7.0 http://drive5.com/uparse/). The classi cation of the representative sequences of each OTU was analyzed by RDP Classi er(http://rdp.cme.msu.edu/) with the con dence threshold of 70% based on the 16S rRNA database of Silva (Release128 http://www.arb-silva.de). OTU abundance information was normalized using a standard of sequence number corresponding to the sample with the least sequences.To compare the species richness and evenness between samples, we used Mothur 1.30.1 to calculate Alpha-diversity index consisted of Chao, Sobs and Shannon diversity indices.

Statistical analysis
Statistical analysis was performed using SPSS 17.0(SPSS, Inc., Chicago, IL, USA).The data were shown as mean ± SEM or median and interquartile range (IQR) based on the distribution. The Mann-Whitney U tests or unpaired student T tests were performed to compare the differences between the two groups. The P value less than 0.05 indicated statistically signi cant difference.

The effect of LPS-induced intrauterine infection on survival of female and neonatal rats
In order to explore the appropriate dose of LPS for establishing an animal model of intrauterine infection, we observed maternal mortality, abnormal pregnancy rates, normal delivery rates and neonatal mortality by intraperitoneal injection of different doses of LPS in pregnant rats (Fig. 1A,n = 40 pregnant rats). We found that the increased doses of LPS did not cause premature delivery, but the fetus died in utero and were absorbed by the mothers or were directly responsible for the mother's death. All newborns were delivered at term. The normal delivery rates decreased with the increased LPS doses of 0,0.3,0.5,0.7,0.9 mg/kg,and the normal delivery rates were 100%, 50%, 33%, 25%and 0%. The abnormal pregnancy rates, which were the rates at which the fetus died in utero and were absorbed by the mothers, also varied with the dose of LPS. When LPS was administered at 0 ,0.3, 0.5, 0.7 mg/kg, the abnormal pregnancy rates were 0%, 50%, 67%, 75%, and the maternal mortality were all at 0%.However, when the dose was raised to 0.9 mg/kg, the maternal mortality rate rose by 25% with an abnormal pregnancy rate of 75%. As a result, we chosed a dose of 0.7 mg/kg LPS for subsequent experimental studies to maximize intestinal changes and adequate litters while minimizing the risk of maternal mortality. In order to ensure that pregnant rats have a similar response to chorioamnionitis by intraperitoneal injection of LPS at a dose of 0.7 mg/kg, we collected uterus specimens for pathological examination. Pregnant rats exposed to LPS had obvious in ammatory cell in ltration in the uterus (Fig. 1B).
During the experimental period of 7 days,we occasionally found the phenomenon that mothers ate their pups and apart from that, none of the pups die. The activities, feeding and hair color of all pups included in the LPS group and the sham group were normal after birth. There was no signi cant difference of birth weight and body weight gains between the LPS group and the sham group (Fig. 1C,n = 136 pups).

Histological analyses of intestinal ileum indicated that prenatal LPS exposure induced intestinal injury in neonatal rats
To determine whether prenatal LPS exposure would induce neonatal intestinal injury and has an effect on intestinal development, we collected distal ileum from different ages of pups for pathological histological analyses.We found that the overall structure of intestinal mucous from the sham groups were normal at any age,the intestinal villis were well-shaped and neatly arranged, the intestinal mucosal epithelial cells were not necrotic and exfoliated, and no in ammatory cells in ltration were observed in the intestinal mucosal layer (Fig. 2A). However, the intestinal structure of pups exposed to LPS at rst days of life were disordered. Compared with the sham group, we could nd shortened and irregular villis, edema of the submucosa, and the in ltration of a large number of in ammatory cells in the submucosa. At 3 days after birth, a large number of mucosal epithelial cells were denatured and swollen, submucosal edema was further aggravated and a large number of in ammatory cells were in ltrated. However, on the 7th day, the intestinal mucosa structure was better than before, villis structure were orderly arranged, no in ammatory cell in ltration was observed, but edema of some mucosal epithelial cells was still visible (Fig. 2A).
The results indicated that prenatal LPS exposure de nitely induced fetal intestinal injury. Subsequently,we further assessed the injury using the intestinal injury score as previously described [21]. The injury scores of the pups exposed to LPS were signi cantly higher than that of the sham groups at d0 and d3 (Fig. 2B)( P = 0 .045, n = 9 pups in the sham group and 7 pups in the LPS group at d0; P = 0.038, n = 7 pups in the sham group and 9 pups in the LPS group at d3).Although the difference did not reach signi cance, the injury scores of LPS group was still higher than the sham group at 7 days after birth (P = 0.909; LPS n = 7, Sham n = 6).

Immunohistochemical analysis suggested that prenatal LPS exposure caused the loss of intestinal wall integrity in neonatal rats
We further evaluated intestinal wall integrity by staining epithelial tight junction protein ZO-1, which plays an important role in maintaining intestinal wall barrier integrity by connecting intestinal epithelial cells [23,24].As shown in Fig. 2C, the intestinal epithelium ZO-1 protein in the sham groups showed a strong brown staining and was evenly distributed on the top of the junction of ileal epithelial cells. However, the ZO-1 protein staining in the ileum tissue of the LPS group was weaker and unevenly distributed, showing discontinuous spots or short bands.
Compared to controls,intestinal ZO-1 expression in pups with prenatal LPS exposure decreased signi cantly at d0 and d3 (Fig. 2D)( P = 0 .005 n = 10 pups in the sham group and 7 pups in the LPS group at d0; P = 0.003, n = 8 pups in the sham group and 10 pups in the LPS group at d3). At 7 days, we observed that the expression level of ZO-1 in the LPS group was slightly lower than that in the sham group, but there was no statistical difference(P = 0.327; Sham n = 8, LPS n = 7).

LPS-induced intestinal injury changes the composition of the intestinal microbiome
We next investigated the fecal microbiota to determine if prenatal LPS exposure would affect gut colonization in rats. As shown in Fig. 3A, the Shannon diversity index in two groups gradually increased with the increase of age, and the diversity index of the sham group at d7 was signi cantly higher than that of pups at 3 days after birth(P < 0.05), however, no signi cant difference was found in the diversity index between different ages in the LPS groups. Compared to sham controls,there were no signi cant differences in diversity index at any ages.
We further investigated the microbiota composition in different groups at different ages at the phyla and genus levels. At the phylum level, the composition of the microbiota in 20 fecal specimens of the two groups of pups revealed that the phyla of Firmicutes,Proteobacteria were the most abundant (Fig. 3B). The relative abundance of Bacteroidetes in the LPS group(0.045%) was signi cantly higher as compared to the sham group(0.006%) at 3 days of life (P = 0.045) ( Fig. 3C). Interestingly, we found that the relative abundance of the Firmicutes in the sham group increased signi cantly from 49.52-91.61% (P = 0.012) and the relative abundance of the proteobacteria decreased signi cantly from 49.96-7.93% (P = 0.012) with the increasing age, however, no signi cant differences were observed in the composition of the microbiota at the phylum level among different ages in the LPS groups.Furthermore, we found that the proportion of Proteobacteria and Bacteroidetes in the LPS group signi cantly increased (P = 0.012 and 0.044)and the proportion of Firmicutes signi cantly decreased(P = 0.012) when compared with that in the Sham group at 7 days after birth (Fig. 3D).
As shown in Fig. 4, at the genus level, the population of Actinomyces and Enterococcus in the LPS group at 3 days of life was lower than the Sham group (P = 0.036 and 0.011), and the relative abundance of Bacteroides was higher in the LPS group compared to the Sham groups (P = 0.025)( Figure 4B).At 7 days of life, fecal samples from the LPS group had higher levels of Escherichia-Shigella(P = 0.012)and Bacteroides (P = 0.044)compared to Sham group. Meanwhile, the proportion of Lactobacillus, Rodentibacter and Veillonella signi cant decreased in LPS group as compared to Sham group(P = 0.012,0.036 and 0.021) ( Figure 4C).

Discussion
Chorioamnionitis is a serious problem as it is often clinically silent,and the diagnosis is based on pathological examination of the placenta or obvious clinical manifestations which was considered to re ect the more serious side of the continuum. This prenatal in ammation is associated with an increased risk of several complications in neonate including NEC [2][3][4][5][6][7]. However, the exact mechanism is not yet clear. Our study clearly shows that maternal prenatal LPS exposure indeed in uences the normal development of intestine in utero and postnatal young rats.Intrauterine infection alone,not combined with prematurity, signi cantly reduced the expression of the tight junction protein ZO-1, which plays a key role in maintaining the integrity of paracellular intestinal barrier. Furthermore, our study also showed that intrauterine in ammation affected the colonization of intestinal microbiome in young rats with signi cantly increased the abundance of Enterobacteriaceae, Escherichia-Shigella, and signi cantly decreased the abundance of Lactobacillus and other bene cial bacteria, which is similar to the alteration of intestinal microbiome in human infants who develop into NEC [18][19][20].
The intestinal epithelium not only separates internal organs from the harmful environment of gut lumen, but also protects against incursion of toxins and foreign microorganisms [25,26]. ZO-1 is one of the main transmembrane proteins that compose the tight junction of the intestinal epithelium,and plays an important role in maintaining the integrity of the intestinal mucosal barrier and intestinal permeability [27]. Disruption of the intestinal barrier is associated with downregulation of tight junction proteins and has been observed in many intestinal diseases such as NEC [28,29]. It is well known that intrauterine infection is associated with an increased risk of NEC [6].Animal studies have shown that intrauterine infections induced by prenatal LPS exposure induced fetal intestinal injury [10,13,30], however the detailed changes taking place at epithelial barrier during intrauterine infections is still not revealed. Our study shows that intrauterine infections induced by prenatal LPS exposure had a disrupted villous mucosal structure at birth, and this effect lasted for 3 days after birth and gradually recovered by 7 days, but the injury score still higher than sham group(P > 0.05) (Fig. 2B). Furthermore,the expression of tight junction protein ZO-1 at different ages between the two groups was consistent with the results of intestinal histological examinations (Fig. 2C Figure2D). These results support the ndings in previous resports that prenatal in ammation exposure does disrupt the normal development of the intestinal mucosal barriers of offspring [10,11,31].However, Fricke et.al [13]have reported that intestinal injury of the offspring induced by prenatal in ammation exposure persists into adulthood. The differences of animal models and the injury scores possibly contributes to the different results found in our model. In addition, the prior animal studies were focused on prematurity, whose intestinal mucosal barrier was immature [11][12][13]. Therefore, it was not clear whether in ammation or prematurity or even their combination induced intestinal injury.Our study demonstrated that intrauterine infection alone,not combined with prematurity, impaired the normal development of the intestinal mucosal barrier in offspring,and we speculate that such effect may initiate during gestation and last for one week or even longer in human infants.
Gut injury associated with abnormal colonization of the intestinal ora [17] ,since intestinal epithelium is not only the energy source of intestinal ora, but also the important habitat of intestinal ora [32].However, the effects of gut injury induced by intrauterine infection on colonization of the intestinal ora of offspring are not known. Our data show that there were no signi cant differences in diversity index at any ages of life when compared to sham controls,which is consistent with the results of previous study on human infants [33]. While Firmicutes and Proteobacteria dominated in all the samples, there is an increase of the Firmicutes with corresponding a decrease of Proteobacteria in sham groups, which is similar to the change of intestinal microbiome in full term infants [34]. However, no obvious change was found in the intestinal microbiome in pups whose mothers exposure to LPS during the period of our observation. Interestingly,when compared to the sham group,we found that pups whose mothers exposure to LPS had lower Firmicutes (Lactobacillus)and higher Proteobacteria (Escherichia-Shigella)in their intestinal microbiome at 7 days of samples. Using a mice models, Elgin et al reported [35]that pups exposure to maternal in ammation did not induce alterations in the composition of the microbiome. However, Puri et al reported [33] that the relative abundance of family Mycoplasmataceae(phylum Tenericutes), genus Prevotella (phylum Bacteroidetes) and genus Sneathia(phylum Fusobacteria) was higher in preterm infants with chorioamnionitis,they found that aberrant intestinal colonization induced by chorioamnionitis was associated with later sepsis. It is possible that the different results of microbial composition with maternal in ammation between our study and those of others are different research species, different experimental designs and different environments.
Several studies have reported [18][19][20], that dysbiosis in early colonizing organisms often prior to the onset of NEC in human infants, which is characterized by an increase in Proteobacteria, especially Enterobacteriaceae. It is well known that intrauterine infection is associated with an increased risk of NEC [6]. We found that intrauterine infections induced both intestinal injury and aberrant intestinal colonization in their offspring. Historically,the gastrointestinal tract of infants is generally considered to be sterile before birth [14], and the intestinal microbiome begins to colonize gradually after birth, even if there is still controversy [15,16]. Moreover, recent animal study have found that the placenta of pregnant mice injected with LPS had no detectable microbiome [13]. Our study revealed that intestinal injury induced by intrauterine infections was present at birth, suggesting that intestinal injury may begin in utero. However, pups whose mothers exposure to LPS did not show abnormal changes in their intestinal microbiome until 7 days after birth. Based on these ndings, we speculate that intrauterine infection may hinder the normal intestinal development of the fetus and cause intestinal injury to the fetus. Intestinal injury induced by intrauterine infection may impair the the ability of the intestine to resist the invasion of bacteria and infection of normal sterile tissue.This results in NEC-like changes in intestinal colonization, characterized by an explosion in Proteobacteria (Escherichia-Shigella) and a decrease in Firmicutes.Bacteria pass through injured intestinal epithelial barrier, leading to tissue destruction, which develops into NEC [36]. Our data provide possible explanation for the increased risk of NEC in infants exposed to intrauterine infection.
There are some shortcomings in our study. First, the sample size in this study is limited, and the results can be confused by differences between models and shams factors such as gestational age or experimental operation. Second, our current studies have not observed either the long-term effects of intrauterine infection on the intestinal microbiota of the offspring or the effects of intrauterine infection on the development of intrauterine intestinal mucosa in the fetus. Therefore, studies with much larger scale with much longer follow-up are necessary to address these issues.

Conclusion
Our ndings demonstrate that intrauterine infection does induce intestinal injury to the offspring, and the effect may initiate in utero. The intrauterine infection induces signi cant alterations of the composition of intestinal microbiota after intestinal injury,which is similar to the microbiota in human infants who develop into NEC. Our results provide support from animal evidence for understanding the mechanism why infants exposed to intrauterine infections before birth have a higher incidence of NEC.

Consent for publication
Not applicable.

Availability of data and materials
The data and materials presented in this study are available from the corresponding author.

Competing interests
The authors declare that they have no competing interests. Authors' Contributions QH designed the study, performed experiments, collected and analyzed data, wrote the rst draft of the manuscript. SL performed experiments.YZ performed experiments and collected data.BW conceptualized the study, edited and revised the manuscript.FB performed experiments and collected data.YC designed the study, interpreted data, and nalized the manuscript. All authors have approved the version of the submitted manuscript.   Shannon diversity index of samples in different ages from two groups (3A). Microbiota composition at phylum level in fecal samples from LPS groups and sham groups at different ages(3B). Comparison of the relative abundance of microbial communities between the LPS groups and sham groups on d3, d7. C,Sham group;T, LPS group (3C, 4D). Asterisk indicates signi cantly differences(* P <0.05).

Figure 4
Microbiota composition at genus level in fecal samples from LPS groups and sham groups at different ages(4A). Comparison of the relative abundance of microbial communities between the LPS groups and sham groups on d3, d7. C,Sham group; T, LPS group(4B, 4C). Asterisk indicates signi cantly differences(* P <0.05).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. BMCTheARRIVEguidelines2.0authorchecklist.docx