Transcriptome signature identifies distinct cervical pathways induced in lipopolysaccharide-mediated preterm birth,

Transcriptome signature identifies distinct cervical pathways induced in... Abstract With half a million babies born preterm each year in the USA and about 15 million worldwide, preterm birth (PTB) remains a global health issue. Preterm birth is a primary cause of infant morbidity and mortality and can impact lives long past infancy. The fact that there are numerous, and many currently unidentified, etiologies of PTB has hindered development of tools for risk evaluation and preventative therapies. Infection is estimated to be involved in nearly 40% of PTBs of known etiology; therefore, understanding how infection-mediated inflammation alters the cervical milieu and leads to preterm tissue biomechanical changes are questions of interest. Using RNA-seq, we identified enrichment of components involved in inflammasome activation and unique proteases in the mouse cervix during lipopolysaccharide (LPS)-mediated PTB and not physiologically at term before labor. Despite transcriptional induction of inflammasome components, there was no evidence of functional activation based on assessment of mature IL1B and IL18 proteins. The increased transcription of proteases that target both elastic fibers and collagen and concentration of myeloid-derived cells capable of protease synthesis in the cervical stroma support the structural disruption of elastic fibers as a functional output of protease activity. The recent demonstration that elastic fibers contribute to the biomechanical function of the pregnant cervix suggests their protease-induced disruption in the infection model of LPS-mediated PTB and may contribute to premature loss of mechanical competency and preterm delivery. Collectively, the transcriptomics and ultrastructural data provide new insights into the distinct mechanisms of premature cervical remodeling in response to infection. Introduction Preterm birth (PTB) is defined as a delivery that occurs prior to 37 weeks of a 40 week gestation in humans. Preterm delivery is a primary cause of infant morbidity and mortality worldwide. Children who are born preterm and survive infancy are at increased risk of medical complications throughout their lives. The financial burden of medical care associated with preterm deliveries exceeds $26 million annually in the USA [1]. The 2015 PTB rate in the USA was 9.6% [2]. Although much research is underway to understand why PTBs occur, the rate peaked at 12.8% (2006) and has only very slowly decreased to its current level over the past 10 years [3]. The sustained high rates of PTB can be attributed to the variety of etiologies, some not yet discovered, of the syndrome [4]. In more than two thirds of cases, an exact cause of PTB cannot be pinpointed [5]. In the realm of PTB, a clinical space filled with unknowns, two facts drive the study outlined here. First, in the one-third of cases with known etiology, infection contributes to 40% of preterm deliveries, and growing evidence supports the role of sterile inflammation as an instigator of PTB [5–7]. Second, regardless of the timing or cause of labor initiation, changes in the cervix precede the onset of labor [8]. A commonly used model of infection-mediated inflammation is intrauterine (IU) administration of lipopolysaccharide (LPS) [9]. Previously published works highlight the distinct cervical features involved in LPS-mediated PTB in mice compared to cervical ripening at term [10,11]. These studies demonstrate that the physiological changes in the tissue that occur in response to a number of hormonal and nonhormonal cues at the end of pregnancy do not occur in LPS-mediated PTB. Rather, a unique set of features help define cervical remodeling in response to an infection or inflammation, more generally. Declining serum progesterone at the end of pregnancy, a process not seen in human, occurs at term and in response to LPS in the mouse [12, 13]. Interestingly, however, serum levels of estradiol do not increase to the same degree in LPS-mediated PTB as they do at term. This difference in the serum progesterone to estrogen ratio—P:E is 80 at term and 528 in LPS preterm—indicates distinct hormone environments in the two groups. Endogenous prostaglandins, specifically PGE2, PGF2a, PGD2, and 6-keto-PGF1a measured in cervical tissue, are increased in the cervix in LPS-mediated PTB and not increased at term. In fact, prostaglandins are essential for LPS-mediated preterm cervical compliance changes and dispensable for term cervical ripening [13]. Additionally, a distinct population of immune cells is present in the cervix during LPS-mediated PTB compared to term before labor. Whereas an influx of activated immune cells such as neutrophils and macrophages does not occur during a normal pregnancy until the postpartum tissue repair phase, these cells are increased in number or activated 6 h after IU LPS treatment on day 15 of mouse gestation, when compared to cells in the cervix of nontreated day 15 mice [10]. Given these features that are unique to an overall inflammatory cervical milieu in LPS-mediated PTB, we sought to better understand the molecular mechanisms underlying the processes by which the cervix responds to infection and inflammation and how these responses lead to preterm delivery. We utilized transcriptome data generated by RNA-seq to identify novel and exclusive pathways in the cervix in LPS-mediated PTB. We have identified enrichment of genes involved in inflammasome activation and upregulation of proteases in the cervix during LPS-mediated PTB exclusively. Subsequent assessment supports a protease-driven disruption of elastic fiber ultrastructure that appears to be achieved in the absence of active IL1B. Collectively, these studies provide insight into distinct features and mechanisms of LPS-mediated preterm cervical remodeling. Materials and methods Mice All animal studies were conducted in accordance with the standards of humane animal care as described in the NIH Guide for the Care and Use of Laboratory Animals. The research protocols were approved by the IACUC office at the University of Texas Southwestern Medical Center. Mice were housed under a 12 h-light/12 h-dark cycle at 22°C. Virgin C57B6/129sv 2–6-month-old female mice were caged with fertile males of the same strain for 6 h. The presence of a vaginal plug at the end of the 6 h was considered as day 0 of pregnancy, with birth of the pups generally occurring early morning day 19. Day 18 samples were collected between 5 and 7 pm before the onset of labor. Inflammation preterm labor model Intrauterine injection of LPS (day 15 LPS) was used to induce preterm labor as previously described [10]. Dosing of LPS was identical to previous studies from our lab [13, 14]. Briefly, mice on day 15 of pregnancy were anesthetized between 7:00 and 9:00 am, and 30 μL sterile water (sham) or 5 mg/mL LPS (Escherichia coli O55:B5 Sigma, St. Louis, MO) was injected IU. Cervical tissues were collected between 1 and 6 h after surgery and before the onset of labor, which occurs approximately 7 to 9 h after LPS administration. RNA sequencing RNA was isolated using an miRNeasy kit from Qiagen (Hilden, Germany). Quantity was determined using a Nanodrop and quality determined using a Bio-Rad Experion. RNAs with RNA Quality Index scores above 9 were used to make cDNA libraries. Eight total cDNA libraries from four different time points/treatment groups (day 15, day 15 sham, day 15 LPS, and day 18) were prepared, with duplicate libraries for biological replicates containing four cervices each. RNA (1.25 μg) from each cervix was pooled for 5 μg of starting RNA in each library. Library preparation was carried out as described previously [15]. One hundred base pair, paired end sequencing was carried out on an Illumina sequencer to a depth of 100 million reads. As indicated in Supplemental Table S1, the total reads for each library was in the range of 8–10 million reads with 94%–95% mappable raw reads. We developed a computational pipeline to determine the differentially expressed genes between the conditions, which included the following steps: reads were aligned to the mm9 genome using the spliced read aligner TopHat version v.2.0.4, transcriptome assembly was carried out using Cufflinks v.2.0.2 with default parameters, filtered transcripts were merged into distinct nonoverlapping sets using Cuffmerge, and Cuffdiff was used to calculate the differential expression genes between the conditions [16, 17]. The RNA-seq data were submitted to GEO with series number GSE98545. Transcriptome data analysis The differentially expressed genes extracted from the above analysis were then used in downstream analyses. Venn Diagrams were generated using Venn Diagram Plotter version 1.5.5228.29250 for the differentially expressed genes in different conditions. Gene Ontology (GO) Analyses and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were determined using DAVID, a web tool for functional annotation and gene enrichment analysis for the genes that are specifically expressed at day 15 LPS treatment as compared to day 18. Heatmaps were generated using Java TreeView for the significantly expressed genes in at least one condition to analyze the effect in the specified condition [18]. Cutoffs used were as follows: 0.5 ≤ fold change ≥ 1.5 and q ≤ 0.05. RNA isolation and quantitative polymerase chain reaction (PCR) Total RNA was isolated as previously described [19]. Complementary DNA synthesis was carried out using 0.5 μg total RNA and 5x iScript Reverse Transcription Supermix (Bio-Rad, Hercules, CA). Quantitative PCR primers used in this study and listed in Supplemental Table S2 were designed and purchased from Invitrogen. Target gene expression was normalized to the expression of housekeeping gene Ppib using the 2∧–ddCt relative gene expression method (User Bulletin no. 2; Applied BioSystems). IL1B enzyme-linked immunosorbent assay (ELISA) Flash-frozen cervical tissues were homogenized in 300 μL Abcam lysis buffer (cat no. 65658, AbCam, UK) and left on ice for 30 min before centrifugation at 4°C for 10 min. Protein concentration was determined (BCA assay, Thermo Scientific, Rockford, IL). IL1B protein was measured via Mouse IL-1B/IL-1F2 DuoSet ELISA (R&D Systems DY401-5) in technical replicates of 100 μL supernatant for day 15, day 15 sham, and day 18 samples while a 3x dilution (33 μL lysate + 66 μL lysis buffer) was used for day 15 LPS samples to maintain the measurements in assay range. Protein blotting for IL1B and IL18 Fifty microgram of whole tissue lysates was boiled for 10 min in Laemmli buffer with 5% β-mercaptoethanol and analyzed by Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on AnykD 15-well precast gels (Bio-Rad) along with protein standards (Precision Plus Protein Kaleidoscope, Bio-Rad) and positive control recombinant proteins (mouse IL1B, 25ng, cat no. 5204SF, Cell Signaling Technology, Danvers, MA; mouse IL18, 100ng, cat no. B0045, Medical & Biological Laboratories CO., LTD, Nagoya, Aichi Japan). Proteins were transferred onto a 0.22-μm nitrocellulose membrane (Bio-Rad) for 30 min at 4°C. Membranes were blocked at room temperature for 1 h in 5% nonfat dry milk in Tris-buffered saline, Tween 20 (TBST) (blocking grade blocker, nonfat dry milk, Bio-Rad). Blots were incubated with primary antibody (IL1B, 1:1000 dilution Cell Signaling Technology cat no. 12507; IL18 1:1000 dilution, Abcam Ab71495; α-tubulin,1:1000 dilution Millipore cat no. 05-829) in blocking solution overnight at 4°C, washed in TBST, and incubated with Horseradish peroxidase (HRP)-labeled secondary antibodies for 45 min at room temperature. The membrane was imaged with ECL Western Blotting Analysis System (GE Healthcare, Buckinghampshire, UK). Immunofluorescence for Elastin Studies were carried out as previously described [19]. Briefly, paraffin sections were deparaffinized in ethanol baths and rinsed in Phosphate-buffered saline (PBS). Antigen retrieval was carried out using 6M Guanidine HCl and iodoacetate. Sections were blocked for 30 min with normal goat serum (NGS) before primary antibody (1:250 dilution, Elastin Products Company cat #PR385, Owensville, MO) incubation in NGS overnight 4°C. Sections were washed in PBS then incubated with Alexa Fluor 546 goat anti-rabbit antibody (1:500, Invitrogen cat# A-11035, Carlsbad, CA) in NGS for 30 min at room temperature in the dark. Coverslips were mounted atop sections with 4΄,6-Diamidino-2-phenylindole dihydrochloride (DAPI) Prolong Gold (Life Technologies, Eugene, OR) and imaged using a Zeiss LSM880 microscope. Two-photon excited immunofluorescence Fifty micrometer thick frozen transverse sections were imaged on a Zeiss LSM880 Laser Scanning confocal microscope with multiphoton (NLO) laser configured with AxioObserver inverted microscope using an Achroplan 40x/0.8 W objective lens (Zeiss, Jena, Germany). A Coherent Chameleon Ultra Ti:Sapphire multiphoton tunable laser (Coherent Inc., Santa Clara, CA) tuned to 860 nm and laser output of 13.5% was focused on the subepithelial stroma of the endocervix. Z stacks of each frozen section were acquired by imaging at 4-μm intervals within the thickness of the 50-μm frozen section. Z-stack images were opened in ImageJ (version 1.51k; NIH, Bethesda, USA); six centrally located slices were compressed into Z-projections of maximum intensity. Brightness and contrast of each channel of the image were adjusted equally across experimental conditions. Transmission electron microscopy Transmission electron microscopy (TEM) was carried out as previously described [19]. Briefly, pregnant mice were perfused with heparinized saline, then glutaraldehyde and paraformaldehyde fixatives in sodium cacodylate buffer. Cervical tissue was removed and fixed in glutaraldehyde in sodium cacodylate buffer overnight at 4°C. The cervix was then sliced in transverse sections and processed as previously described [19]. A Tecnai G2 spirit transmission electron microscope at 120 kV and a side-mounted SIS Morada 11 megapixel CCD camera were used for image acquisition. Immunofluorescence staining for F4/80+ cells Immunofluorescence was carried out for F4/80+ cells in day 15, day 15 sham, day 15 LPS, and day 18 mouse 5-μm paraffin embedded cervix sections. Sections were deparaffinized in ethanol baths, rinsed in PBS, and treated with Proteinase K for 3 min at room temperature before being rinsed with PBS and blocked at room temperature for 30 min with NGS. Sections were incubated with F4/80 antibody (Serotec, Raleigh, NC, rat anti mouse F4/80) at 1:500 dilution overnight at 4°C. The next day, sections were washed in PBS then incubated with Alexa Fluor 488 goat anti-rat antibody (1:500, Invitrogen cat# A-11035, Carlsbad, CA) in NGS for 30 min at room temperature in the dark. Coverslips were mounted atop sections with DAPI Prolong Gold (Life Technologies, Eugene, OR) and imaged using a Zeiss LSM880 microscope and 40x lens. Statistical analysis Statistics were performed using Prism software (GraphPad Software). For comparison of multiple groups, one-way ANOVA was used followed by Tukey multiple comparisons test. The values were expressed as mean ± SEM and considered significant if P < 0.05. Number of animals used and data analysis are described in the figure legends. Results RNA-seq identifies unique transcriptomes in lipopolysaccharide-mediated preterm and term cervical ripening To elucidate the transcriptional pathways that guide cervical remodeling in response to inflammation, we utilized RNA Sequencing. This method of gene expression analysis was undertaken to identify novel and exclusive pathways induced in the cervix of the LPS-mediated PTB model versus mice at term before labor. To determine changes in gene expression during LPS-mediated PTB, we compared day 15 LPS to untreated day 15 (Figure 1A). To determine gene expression changes during term cervical ripening, we compared day 18 to day 15. To ensure the changes in gene expression observed upon LPS treatment were due to the LPS and not due to the survival surgery, sham to day 15 comparisons were also included. Overall gene expression patterns between LPS preterm and term were quite distinct (Figure 1B). The number of differentially expressed genes exclusive to day 15 LPS vs day 15 (2918 genes) was large, compared to day 18 vs day 15 (475 genes) (Figure 1C). Relatively few genes (267) were common between the two groups and differentially expressed as compared to day 15. A similar number of genes were up- and downregulated in the day 15 LPS group compared to gestation day 15, while the majority of differentially expressed genes were upregulated in the day 18 group (Figure 1D). In agreement with previous reports using quantitative PCR (qPCR), gene expressions of canonical parturition genes including connexin 26 (gap junction protein, beta 2 [Gjb2] (also connexin 26)), hyaluronan synthase 2 (Has2), and steroid 5 alpha reductase 1 (Srd5a1), as measured using RNA-seq, are upregulated at term before labor and not changing in a statistically significant manner upon LPS treatment on day 15 (Supplemental Table S3) [10]. Reciprocally, proinflammatory genes previously reported via qPCR to be upregulated upon LPS treatment on day 15 and not changing at term before labor, in particular cyclooxygenase 2 (Ptgs2), prostaglandin E synthase (Ptges), interleukin 6 (Il6), interleukin 1 alpha (Il1a), and C-X-C motif chemokine ligand 2 (Cxcl2), are identified as upregulated via RNA-seq in day 15 LPS and not on day 18, both compared to day 15 (Supplemental Table S4) [10]. Figure 1. View largeDownload slide RNA-seq analyses of the pregnant mouse cervix demonstrate unique transcriptome profiles in term and LPS-mediated preterm cervical remodeling. (A) Schematic indicating time points (day 15 and day 18) and treatment groups (day 15 sham and day 15 LPS) used in RNA-seq. Two libraries of four cervices each were constructed for each of the four conditions. (B) Heatmap of all coding transcripts in day 15 LPS vs day 15 (LPS preterm) and day 18 vs day 15 (term). Fold changes (from 0.012 to 463) comparing to day 15 to day 15 LPS and day 18 are shown. (C) Venn diagram with the total number of statistically significantly and differentially expressed transcripts in each group (day 15 sham, day 15 LPS, and day 18) compared to day 15. (D) Table with numbers of statistically significant upregulated and significantly downregulated transcripts in each condition (day 15 sham, day 15 LPS, and day 18) compared to day 15. Figure 1. View largeDownload slide RNA-seq analyses of the pregnant mouse cervix demonstrate unique transcriptome profiles in term and LPS-mediated preterm cervical remodeling. (A) Schematic indicating time points (day 15 and day 18) and treatment groups (day 15 sham and day 15 LPS) used in RNA-seq. Two libraries of four cervices each were constructed for each of the four conditions. (B) Heatmap of all coding transcripts in day 15 LPS vs day 15 (LPS preterm) and day 18 vs day 15 (term). Fold changes (from 0.012 to 463) comparing to day 15 to day 15 LPS and day 18 are shown. (C) Venn diagram with the total number of statistically significantly and differentially expressed transcripts in each group (day 15 sham, day 15 LPS, and day 18) compared to day 15. (D) Table with numbers of statistically significant upregulated and significantly downregulated transcripts in each condition (day 15 sham, day 15 LPS, and day 18) compared to day 15. Gene Ontology Analyses identify exclusive pathways in lipopolysaccharide-mediated preterm birth To gain insight into potential pathways driving LPS-mediated premature cervical changes and term cervical ripening, GO Analyses were performed. The most enriched processes from the group of genes upregulated in LPS preterm exclusively included immune, defense, and inflammatory responses and response to wounding (Figure 2A). Analyses of genes upregulated exclusively at term before labor identified the biological processes lipid catabolic process, female pregnancy, epidermis development, and ectoderm development as the most enriched. Biological processes from genes upregulated both in LPS preterm and term before labor include positive regulation of developmental process, keratinization, positive regulation of biological process, and positive regulation of cellular processes (Figure 2A). Further in-depth comparisons of GO pathways based on Molecular Function, Biological Processes, and KEGG Pathways specific to the LPS preterm group identify a common theme consistent with NFkB activation, pathogen sensing, inflammasome components, and downstream endpoints of inflammasome activation (Figure 2B). Given the critical role of inflammasome activation in host defense against pathogens, sterile insults, and host-derived molecules via activation of downstream proinflammatory events, we focused our studies on inflammasome activation in the cervix as a novel pathway specific to LPS-mediated PTB. Figure 2. View largeDownload slide GO and KEGG pathway analyses of term and LPS-mediated preterm cervical remodeling. (A) GO analyses demonstrate enrichment of unique biological processes in term and LPS-mediated PTB. The most statistically significant (as indicated by log10P-value) processes with genes expressed exclusively in each of the groups and those commonly expressed between them (common) are presented. Percentages indicate the percentage of genes in the process represented in the data set. (B) GO analyses (nucleotide binding related molecular functions and cell death biological processes) and KEGG analysis (PRRs related pathways) of genes expressed in LPS preterm exclusively. Components of inflammasome activation fall within several of the indicated pathways and processes. Figure 2. View largeDownload slide GO and KEGG pathway analyses of term and LPS-mediated preterm cervical remodeling. (A) GO analyses demonstrate enrichment of unique biological processes in term and LPS-mediated PTB. The most statistically significant (as indicated by log10P-value) processes with genes expressed exclusively in each of the groups and those commonly expressed between them (common) are presented. Percentages indicate the percentage of genes in the process represented in the data set. (B) GO analyses (nucleotide binding related molecular functions and cell death biological processes) and KEGG analysis (PRRs related pathways) of genes expressed in LPS preterm exclusively. Components of inflammasome activation fall within several of the indicated pathways and processes. Components of the inflammasome are upregulated in lipopolysaccharide-mediated PTB Genes encoding components of the inflammasome activation pathway were identified in the RNA-seq data set as upregulated in LPS preterm and not changing at term before labor (Figure 3A). These include the pattern recognition receptors (PRRs) Nod 2, Nrlc4, Nrlc5, Nrlp3, activator protein Gbp5, executioner proteins Casp1 and Casp4, and cytokine Il1b. Gene expression analysis by quantitative PCR corroborates RNA-seq results for the majority of genes and shows an increase in gene expression in inflammation preterm (day 15 LPS vs day 15) and no change at term before labor (day 18 vs day 15) for Gbp5, Nlrc5, and Nod2 (Figure 3B). Transcripts of Casp4 were significantly induced 2, 4, and 6 h after LPS treatment, compared to untreated day 15 and not at term (Figure 3C). Using quantitative PCR, transcripts encoding Nlrp3, Nlrc4, and Casp1 were not significantly induced in the LPS-mediated PTB group or at term before labor (day 18). Figure 3. View largeDownload slide Components of the inflammasome are upregulated in LPS-mediated PTB exclusively. (A) Heatmap of select inflammasome-related genes and their expression pattern in day 15 LPS cervices vs day 15 cervices (LPS preterm) and in day 18 cervices vs day 15 cervices (term). Fold changes (from 0 to 35.5) comparing to day 15 to day 15 LPS and day 18 are shown. (B) qPCR results of selected inflammasome genes showing relative gene expression on day 15, day 15 sham, day 15 LPS, and day 18. The values expressed are mean ± SEM. (n = 5–7 cervices per group, one-way ANOVA *P < 0.05 relative to day 15). (C) qPCR results for Casp1 and Casp4 showing relative gene expression on day 15, day 15 sham, day 15 after 1, 2, 4, or 6 h LPS treatment, and day 18. The values expressed are mean ± SEM. (n = 5–6 cervices per group, one way ANOVA *P < 0.05 relative to day 15). Figure 3. View largeDownload slide Components of the inflammasome are upregulated in LPS-mediated PTB exclusively. (A) Heatmap of select inflammasome-related genes and their expression pattern in day 15 LPS cervices vs day 15 cervices (LPS preterm) and in day 18 cervices vs day 15 cervices (term). Fold changes (from 0 to 35.5) comparing to day 15 to day 15 LPS and day 18 are shown. (B) qPCR results of selected inflammasome genes showing relative gene expression on day 15, day 15 sham, day 15 LPS, and day 18. The values expressed are mean ± SEM. (n = 5–7 cervices per group, one-way ANOVA *P < 0.05 relative to day 15). (C) qPCR results for Casp1 and Casp4 showing relative gene expression on day 15, day 15 sham, day 15 after 1, 2, 4, or 6 h LPS treatment, and day 18. The values expressed are mean ± SEM. (n = 5–6 cervices per group, one way ANOVA *P < 0.05 relative to day 15). A functional consequence of canonical and noncanonical inflammasome complex formation and activation is the secretion of the cytokines IL1B and IL18 and induction of pyroptosis. To investigate a functional role for inflammasome activation in inflammation preterm, transcripts of Il1b were analyzed using qPCR (Figure 4A). Il1b was significantly upregulated 2, 4, and 6 h after LPS treatment on day 15, compared to untreated day 15 samples. IL1B protein levels were measured by ELISA in whole cervical tissue lysates from untreated day 15 mice, day 15 sham mice, day 15 LPS mice, and day 18 mice (Figure 4B). IL1B was measureable in each of the day 15 LPS samples tested and amounts were significantly increased compared to nontreated day 15 (Figure 3B). To determine if the increase in IL1B protein was the mature and biologically active 17kDa form, protein blotting was undertaken using 50 μg whole cervical tissue lysate taken from mice exposed to LPS for 0, 1, 2, 4, and 6 h (Figure 4C). Although the time-dependent exposure to LPS increased synthesis of pro-IL1B, mature IL1B protein was not detectable at any time point after treatment. The time-dependent increase in pro-IL1B abundance did not achieve statistical significance, despite the marked temporal increase visualized in all four experiments (Figure 4D). The protein expression of mature IL18 was also evaluated in cervical tissue of mice treated with LPS (Figure 4E). Similar to IL1B, no mature IL18 protein was detectable. Figure 4. View largeDownload slide Profile of IL1B and IL18 as outputs of the inflammasome in LPS-mediated PTB. (A) Relative gene expression of Il1b in day 15 (NT), day 15 sham (sham), 1, 2, 4, and 6 h after LPS treatment. The values expressed are mean ± SEM. (n = 5–7 cervices per group, one way ANOVA *P < 0.05 relative to day 15). (B) ELISA results for IL1B in whole tissue lysates from the cervix of day 15, day 15 sham, day 15 LPS, and day 18 mice. Three separate experiments were performed with a total of 13–15 total cervices per group. One-way ANOVA *P < 0.05 relative to day 15. (C) Representative immunoblot for pro (31 kDa) and mature (17 kDa) forms of IL1B in mouse cervix tissues on day 15 without treatment (NT) and 1, 2, 4, and 6 h after LPS treatment. As a positive control, 25 ng recombinant mature mouse IL1B (mIL1B) was run in the first lane. A time-dependent increase in pro-IL1B is observed with no detectable expression of the mature form. Four separate experiments were run with two samples pooled per treatment group in each experiment. α-tubulin was used as a loading control. (D) Quantification of average pro-IL1B abundance from four independent blotting experiments. Pro-IL1B abundance is expressed as a ratio compared to alpha-tubulin and normalized to untreated day 15. (E) Representative immunoblot for mouse IL18 in mouse cervix tissues on day 15 without treatment (NT) and 1, 2, 4, and 6 h after LPS treatment with 100 ng positive control recombinant mouse IL18 (18 kDa) run in the first lane. No expression of active form of IL18 was found in any sample. Four independent experiments were run with two samples pooled per treatment group in each experiment. α-tubulin was used as a loading control. Figure 4. View largeDownload slide Profile of IL1B and IL18 as outputs of the inflammasome in LPS-mediated PTB. (A) Relative gene expression of Il1b in day 15 (NT), day 15 sham (sham), 1, 2, 4, and 6 h after LPS treatment. The values expressed are mean ± SEM. (n = 5–7 cervices per group, one way ANOVA *P < 0.05 relative to day 15). (B) ELISA results for IL1B in whole tissue lysates from the cervix of day 15, day 15 sham, day 15 LPS, and day 18 mice. Three separate experiments were performed with a total of 13–15 total cervices per group. One-way ANOVA *P < 0.05 relative to day 15. (C) Representative immunoblot for pro (31 kDa) and mature (17 kDa) forms of IL1B in mouse cervix tissues on day 15 without treatment (NT) and 1, 2, 4, and 6 h after LPS treatment. As a positive control, 25 ng recombinant mature mouse IL1B (mIL1B) was run in the first lane. A time-dependent increase in pro-IL1B is observed with no detectable expression of the mature form. Four separate experiments were run with two samples pooled per treatment group in each experiment. α-tubulin was used as a loading control. (D) Quantification of average pro-IL1B abundance from four independent blotting experiments. Pro-IL1B abundance is expressed as a ratio compared to alpha-tubulin and normalized to untreated day 15. (E) Representative immunoblot for mouse IL18 in mouse cervix tissues on day 15 without treatment (NT) and 1, 2, 4, and 6 h after LPS treatment with 100 ng positive control recombinant mouse IL18 (18 kDa) run in the first lane. No expression of active form of IL18 was found in any sample. Four independent experiments were run with two samples pooled per treatment group in each experiment. α-tubulin was used as a loading control. Expression of proteases in the cervix is upregulated in lipopolysaccharide-mediated PTB Another group of genes identified as upregulated in LPS-mediated premature cervical remodeling is proteases (Figure 5). RNA-seq data set analysis identified two GO terms—proteolysis and peptidase activity—as significantly enriched in genes exclusive to day 15 LPS vs day 15 (Figure 5A). The data set contains a large number of genes (107 and 75 respectively) within these pathways (Figure 5A). These genes, including matrix metalloproteinase (Mmp), a disintegrin and metalloproteinase with thrombospondin type 1 motifs (Adamts), and cathepsin (Cts) proteases, target and modify components of the extracellular matrix (ECM). Known functions of these proteases include postsynthesis processing of mature ECM molecules including collagen, degradation during physiologic turnover processes (i.e., pregnancy), and pathologic breakdown due to inflammation (i.e., arthritis) [20–22]. The heatmap in Figure 5B depicts fold change values of genes encoding ADAMTS, MMP, and CTS proteases identified as differentially regulated in LPS preterm or term cervical ripening. Gene expression analyses using RT-qPCR for select genes Mmp13, Mmp12, Mmp8, Ctsl, Ctsc, Adamts1, Adamts5, and Adamts15 identified significantly increased expression of Mmp13, Adamts1, and Adamts5 in cervices from day 15 LPS-treated mice (Figure 5C). While there was a trend toward an increase in Ctsl gene expression that did not achieve statistical significance, subsequent studies shown in Supplemental Figure S1 using time course samples 1, 2, 4, and 6 h after LPS treatment indicate an increase in Ctsl expression 6 h after LPS treatment on day 15. We have previously reported increased expression of Mmp8, Adamts1, and Adamts4 in the cervix of day 15 LPS-treated mice [10]. Both Adamts1 and Adamts4 have been shown to be upregulated in term ripening at gestation day 18, although the changes seen in these quantitative PCR experiments for Adamts1 do not reach statistical significance on day 18 (Figure 5C). The known targets of Adamts1 and Adamts4 include proteoglycans [10, 23]. The upregulation of proteases that target the main structural components collagen and elastic fibers in response to pathological inflammation but not in the normal catabolic processes led to assessment of the cervical ECM structure in the day 15 LPS-treated mice. Figure 5. View largeDownload slide Protease genes and pathways are upregulated in the cervix during LPS-mediated PTB exclusively. (A) GO biological process proteolysis and molecular function peptidase activity are enriched in the gene set exclusive to LPS-mediated PTB. Count refers to the number of genes from the data set that fall into the GO pathways. (B) Heatmap of selected proteases and their expression pattern in day 15 LPS cervices vs day 15 cervices (LPS preterm) and in day 18 cervices vs day 15 cervices (term). (C) qPCR results of selected protease genes showing relative gene expression on day 15, day 15 sham, day 15 LPS, and day 18. The values expressed are mean ± SEM. (n = 4–6 cervices per group, one-way ANOVA *P < 0.05 relative to day 15). Figure 5. View largeDownload slide Protease genes and pathways are upregulated in the cervix during LPS-mediated PTB exclusively. (A) GO biological process proteolysis and molecular function peptidase activity are enriched in the gene set exclusive to LPS-mediated PTB. Count refers to the number of genes from the data set that fall into the GO pathways. (B) Heatmap of selected proteases and their expression pattern in day 15 LPS cervices vs day 15 cervices (LPS preterm) and in day 18 cervices vs day 15 cervices (term). (C) qPCR results of selected protease genes showing relative gene expression on day 15, day 15 sham, day 15 LPS, and day 18. The values expressed are mean ± SEM. (n = 4–6 cervices per group, one-way ANOVA *P < 0.05 relative to day 15). Elastic fiber ultrastructure is disrupted in lipopolysaccharide-mediated preterm birth Recently published studies from our lab in the mouse cervix highlight the contribution of elastic fiber structural organization to the biomechanical function of the cervix [19]. Furthermore, these studies document an increased elastic fiber density in the stroma adjacent to the epithelia, termed the subepithelial stroma, as compared to the midstroma region. Elastic fibers are composed of tropoelastin protein crosslinked in and around a microfibrillar scaffold consisting of fibrillin 1 and 2, plus a number of other proteins [19]. To visualize potential alterations in cervical elastin morphology in the LPS-mediated PTB, an antibody generated against tropoelastin was utilized for immunofluorescence staining. Tropoelastin staining in day 15 LPS appeared similarly localized in the subepithelial stroma, and long elastin strands were evident to a similar degree as compared to sham day 15 and untreated day 15 controls (Figure 6). Images of separate DAPI and tropoelastin channels are shown in Supplemental Figure S2. In addition, two-photon excited fluorescence (TPEF) was utilized as a second approach to assess elastic fiber structure. Similar to the tropoelastin immunofluorescence results, elastin fiber structure in TPEF images was indistinguishable between day 15 LPS and day 15 control groups (Figure 6). Subsequent analysis was carried out to evaluate elastic fiber ultrastructure by TEM, which allows for assessment of both elastin and the microfibrillar scaffold of the fiber (Figure 7). Elastic fibers in the subepithelial stroma region of the cervix were evaluated in day 15, day 15 sham, day 15 LPS, and day 18 mice. Overall, elastic fibers in the day 15 LPS group appeared disrupted with less darkly stained elastin integrated into the scaffold, resulting in increased visibility of the microfibrillar scaffold component of the elastic fiber (Figure 7). Cervical elastic fibers from sham animals were similar in structure and density to untreated day 15. No such disruption of elastic fiber ultrastructure occurs at term before labor. Comparing day 18 to day 15, elastic fibers appear intact with the darkly stained elastin covering the microfibrillar scaffold. In contrast to the LPS model, disrupted elastic fibers were a fraction of otherwise normal term fibers in the mifepristone-PTB model (Supplemental Figure S3). Figure 6. View largeDownload slide Immunofluorescence staining for tropoelastin and TPEF show similar cervical elastin morphology upon LPS treatment on day 15. (Top row) Immunofluorescence imaging of tropoelastin protein (red) and nuclei by DAPI (blue) in cervical sections from day 15, day 15 sham, and day 15 LPS mice. Tropoelastin is most concentrated in the subepithelial stroma region, directly below the superimposed white dotted line between the epithelium and stroma, and can be visualized in long-fiber structures, presumed to be tropoelastin integrated into elastic fibers. The overall impression from this experiment is that elastin appears similar in location and morphology in all three groups. Five to six images in each cervical section for 3–5 mice per group. Images shown here are representative of each group. Signal intensity was optimized for day 15, and the same settings were used for day 15 sham and day 15 LPS. (Bottom row) Representative Z-projected images of autofluorescence of elastic fibers imaged using TPEF. Images were taken in the subepithelium of the endo cervix of day 15, day 15 sham, and day 15 LPS-treated mice. Images were obtained from 5–7 regions per tissue and from 3–4 mice per treatment group. No major change in elastic fiber localization or morphology was seen among the groups. Scale bar: 20 μm. Figure 6. View largeDownload slide Immunofluorescence staining for tropoelastin and TPEF show similar cervical elastin morphology upon LPS treatment on day 15. (Top row) Immunofluorescence imaging of tropoelastin protein (red) and nuclei by DAPI (blue) in cervical sections from day 15, day 15 sham, and day 15 LPS mice. Tropoelastin is most concentrated in the subepithelial stroma region, directly below the superimposed white dotted line between the epithelium and stroma, and can be visualized in long-fiber structures, presumed to be tropoelastin integrated into elastic fibers. The overall impression from this experiment is that elastin appears similar in location and morphology in all three groups. Five to six images in each cervical section for 3–5 mice per group. Images shown here are representative of each group. Signal intensity was optimized for day 15, and the same settings were used for day 15 sham and day 15 LPS. (Bottom row) Representative Z-projected images of autofluorescence of elastic fibers imaged using TPEF. Images were taken in the subepithelium of the endo cervix of day 15, day 15 sham, and day 15 LPS-treated mice. Images were obtained from 5–7 regions per tissue and from 3–4 mice per treatment group. No major change in elastic fiber localization or morphology was seen among the groups. Scale bar: 20 μm. Figure 7. View largeDownload slide Elastic fiber ultrastructure is disrupted in the cervix of LPS-treated mice before onset of PTB. TEM analysis of elastic fibers in cervices from day 15, day 15 sham, day 15 IU LPS, and day 18 mice. Elastic fiber ultrastructure (black arrows) is abnormal in day 15 LPS compared to day 15 and sham. (n = 3–4 mice per group). A total of 11–62 elastic fibers were imaged per sample. Scale bar: 1000 nm. Figure 7. View largeDownload slide Elastic fiber ultrastructure is disrupted in the cervix of LPS-treated mice before onset of PTB. TEM analysis of elastic fibers in cervices from day 15, day 15 sham, day 15 IU LPS, and day 18 mice. Elastic fiber ultrastructure (black arrows) is abnormal in day 15 LPS compared to day 15 and sham. (n = 3–4 mice per group). A total of 11–62 elastic fibers were imaged per sample. Scale bar: 1000 nm. Localization of myeloid cells in cervical stroma during lipopolysaccharide-mediated preterm birth Recently published work from our lab has demonstrated that 7/4 + cells cluster in the subepithelial stroma region of the mouse cervix on day 15 upon LPS treatment, compared to untreated day 15 samples [14]. We sought to investigate the localization of other immune cells that are capable of producing and secreting proteases that have been identified as upregulated in response to LPS and potentially targeting the elastic fibers identified as disrupted in these tissues. Immunofluorescence was carried out in day 15, day 15 sham, day 15 LPS, and day 18 samples to investigate the localization of immune cells, namely F4/80+ macrophages. F4/80+ cells are equally distributed throughout the mid and subepithelial stroma in day 15, day 15 sham, day 15 LPS, and day 18 samples (Figure 8). Consistent with previous analysis of cervical F4/80+ macrophages in the LPS model by flow cytometry, there was no apparent increase in macrophage numbers in the LPS group as compared to other groups [10]. Figure 8. View largeDownload slide F4/80+ cells are present in the mid and subepithelial cervical stroma. The localization of F4/80+ cells (green) was determined using immunofluorescence techniques on paraffin sections from cervices of mice on day 15, day 15 sham, day 15 LPS, and day 18 mice. N = 3 animals per group, 3–8 images taken in each the sub-epithelial stroma and the mid-stroma. Representative images are shown. Figure 8. View largeDownload slide F4/80+ cells are present in the mid and subepithelial cervical stroma. The localization of F4/80+ cells (green) was determined using immunofluorescence techniques on paraffin sections from cervices of mice on day 15, day 15 sham, day 15 LPS, and day 18 mice. N = 3 animals per group, 3–8 images taken in each the sub-epithelial stroma and the mid-stroma. Representative images are shown. Discussion Elucidation of the distinct pathways that drive LPS-mediated preterm cervical remodeling but not cervical ripening at term will advance the development of clinically relevant therapies to prevent premature changes in the cervix that lead to preterm deliveries. This study builds upon previous findings that demonstrate distinct features of the cervical remodeling process in response to inflammation are not present at term before labor [10, 13, 19]. The distinct transcriptome signatures uncovered in this study add to the field's understanding of the differences between LPS-mediated and term cervical remodeling. In particular, LPS does not alter expression of many genes regulated in term cervical ripening, while pathways related to and downstream of inflammatory responses were most dramatically induced upon LPS treatment. While much valuable information remains to be extracted from the RNA-seq data sets, in the present study, we have leveraged these data to identify transcriptional pathways that may direct features of LPS-mediated preterm cervical remodeling—including proinflammatory responses and regionalized disorganization of collagen fibers in the stroma—that have previously been identified [13, 14]. Specifically, we have explored pathways that include components of an activated inflammasome and protease action in the ECM. These pathways in the cervix are exclusive to LPS-mediated PTB. Recent studies implicate inflammasomes as mediators of both sterile inflammation leading to spontaneous labor at term and pathological inflammation leading to preterm delivery both in women with antiphospholipid syndrome and women with acute histologic chorioamnionitis [6, 24, 25]. Endogenous danger signals, termed damage-associated molecular patterns (DAMPs), resulting from cellular stresses can trigger sterile inflammation. Pathogens, which are recognized by PRRs, and subsequent DAMP induction can induce pathological inflammation. Components of inflammasomes are expressed in the chorioamnion of women at term, and there is evidence of increased caspase-1 activation and mature IL1B in tissues from women at term and in labor compared to nonlaboring term tissues [24]. In response to pathogens recognized by a number of PRRs, human fetal membrane explants can mount an impressive immune response (i.e., secrete cytokines), indicating these tissues express machinery needed for inflammasome activation [6, 24, 26]. Chorioamnion membranes from women with acute histologic chorioamnionitis at term show increased PRR gene expression, protein abundance, and downstream inflammasome activation as measured via caspase-dependent IL1B production in vivo [6]. In contrast to evidence that inflammasome-mediated increases in caspase 1 activation and mature IL1B play a role in fetal membrane signaling for the initiation of labor both at term and with pathogen-mediated inflammation, our data in the pregnant mouse cervix demonstrate the transcriptional activation of inflammasome components and caspase 4 in response to LPS but no evidence for inflammasome-dependent production of biologically active IL1B or IL18. These findings are consistent with previous reports that caspase-mediated generation of bioactive IL1B requires induced expression of pro-IL1B by a toll-like receptor ligand such as LPS followed by a secondary signal by DAMPs to induce pro-IL1B processing and secretion [27]. While caspase 1 mRNA expression was not altered with LPS treatment, the induction of caspase 4 is potentially sufficient as work in other infection and inflammatory disease states demonstrates that caspase-4 can directly sense LPS and induce a noncanonical inflammasome activation reaction [28]. Biologically active IL1B can also be generated by caspase-1/4 independent pathways that include caspase-8, neutrophil elastase, cathepsin G, granzyme A, and matrix metalloproteases [29–31]. While the potential role of inflammasome-mediated production of bioactive IL1B in premature ripening induced by ascending pathogens that are not represented by the LPS PTB model (e.g., gram + bacteria and/or viruses) warrants further study, an important observation from the current findings is the evidence that protease-mediated disruption of cervical elastic fibers and premature ripening can be achieved in the absence of bioactive IL1B. Because multiple inflammation-related pathways are induced in the cervix exclusively during LPS-mediated PTB and not at term, we anticipate that other inflammatory pathways, including NFkB activation pathways, are sufficient to accomplish the premature changes in cervical ripening in LPS-mediated PTB. Future studies will investigate the inflammatory pathways that are required to achieve premature remodeling. The cervical transcriptional signature of LPS-treated mice includes robust upregulation of proteases, including Mmp13, Mmp8, and Ctsl, that target the major structural proteins of the cervix, namely collagen and components of the elastic fiber (Figure 5; Supplemental Figure S1). The process by which infection and inflammation induce upregulation of these particular collagen and elastin targeting proteases in the cervix we hypothesized would be similar to the pathophysiology of the proinflammatory disease osteoarthritis in which IL1B has been shown to directly upregulate proteases such as collagenase Mmp13 [32]. While transcriptional upregulation of the Il1b transcript precedes the upregulation of proteases such as Mmp13 and Ctsl (Figure 4; Supplemental Figure S1), the observed disruption of elastic fiber ultrastructure in the absence of active IL1B suggests an IL1B independent pathway of protease activation. Cathepsin L, a protease targeting both elastin and collagen, is a lysosomal and secreted enzyme that has yet to be appreciated in the context of cervical remodeling [33–35]. Its expression is exclusive to day 15 LPS cervices and is not upregulated at term before labor, indicating further studies are warranted to investigate its potential role in the cervical elastic fiber disruption seen in LPS-mediated PTB. While neither transcripts encoding neutrophil elastase, Elane, or macrophage elastase, Mmp12 were elevated in cervical tissue with LPS treatment, the protein synthesis and packaging of these proteases into granules occurs prior to infiltration of these immune cells into tissue, and thus transcriptional changes would not be anticipated. Future investigations to demonstrate activity of proteases that target elastic fibers are warranted. While proteases that target collagen and elastic fibers are upregulated in the cervix specifically in LPS-mediated PTB, protease members of the ADAMTS family including Adamts1 and Adamts4 are upregulated both at term and in LPS-mediated PTB. These proteases are known to target proteoglycans and likely contribute to the normal cervical ECM turnover required for physiological remodeling [23, 36]. In the current study, a potential functional outcome of increased cervical protease gene expression in LPS-mediated PTB is the observed disruption of elastic fiber ultrastructure. In contrast, disrupted elastic fibers were not evident in TEM images from term before labor and were a minor subset of elastic fibers in mifepristone-induced preterm cervical ripening (Supplemental Figure S3).We have previously reported that the density of elastic fibers is greatest in the subepithelial region of the cervical stroma and that elastic fibers along with collagen fibers dictate the mechanical strength of the cervix [37]. Recent studies from our group focused on understanding the structural changes in cervical collagen upon LPS treatment demonstrate a preferential disorganization of collagen fiber structure in the subepithelial stroma, as determined by second harmonic generation and subsequent image analysis [14]. Collectively, these findings suggest that LPS-driven activation of proteases leads to a pathological disruption of the cervical ECM, which allows for mechanical weakening of the cervix and subsequent PTB. We suggest that this model, which Mont Liggins once proposed as the normal physiology of term cervical ripening, is a key and distinct mechanism of preterm cervical remodeling in a mouse model of inflammation-mediated PTB [38]. Based on insights gained from the cervical transcriptome signature of LPS-mediated PTB, we propose a model in which an IL1B-independent pathway upregulates proteases, likely produced by immune cells in addition to other cells of the cervix, that target and degrade ECM components, namely elastic fibers and collagen in the subepithelial stroma, leading to disrupted ECM structure, decreased biomechanical integrity, and leading to PTB (Figure 7) [14]. These data presented here, including cervical transcriptome pathways induced in LPS-mediated PTB exclusively and altered elastic fiber ultrastructure, connect to form a plausible pathway linking infection to pathological ECM structural changes and subsequent loss of biomechanical integrity leading to premature delivery. While LPS treatment provides insight into responses anticipated from gram-negative bacteria, we expect that the described pathway will be conserved in response to other vaginal bacteria such as Ureaplasma, Mycoplasma, and Gardnerella species, which have been linked to spontaneous preterm deliveries [39]. Further mining of this data set in addition to other infection and inflammation models may help elucidate additional pathways for investigation. The clinical implications of this work extend beyond an increased understanding of the distinct pathways that mediate PTB in the face of infection-induced inflammation. PTB is the most robust risk factor for future PTBs [5]. The pathological disruption of elastic fibers in LPS-mediated preterm cervical remodeling may have an impact on cervical function in future pregnancies. If elastic fibers, whose synthesis is thought to be limited to times of development, are unable to repair appropriately following an inflammation-mediated PTB, cervical competency may be compromised long before the parturition process begins in subsequent pregnancies. Future experiments to determine the long-term impact of pathologic elastic fiber disruption and the ability of disrupted elastic fibers to functionally recover postpartum will be critical to test this hypothesis. Supplementary data Supplementary data are available at BIOLRE online. Supplementary Table 1. Antibody Information. Supplemental Figure S1. Timecourse of Tnf, Adamts4, Ctsl, and Mmp13 gene upregulation in response to intrauterine LPS. Gene expression analysis by qPCR demonstrates that Tnf transcription is significantly induced in the mouse cervix 2 h post-LPS compared to untreated day 15 (NT). Gene expression of proteases Adamts4 is significantly upregulated at 2 and 4 h time points, Ctsl by 6 h, and Mmp13 at 4 and 6 h post-LPS compared to untreated day 15 (NT) samples. (n = 4–6 samples per group, P < 0.05 compared to day 15, bars represent mean ± SEM. One-way ANOVA). Supplemental Figure S2. Immunofluorescence staining for tropoelastin. (A) Immunofluorescence imaging of tropoelastin protein (red) and nuclei by DAPI (blue) in cervical sections from day 15, day 15 sham, and day 15 LPS mice. Elastin is most concentrated in the subepithelial stroma region, directly below the superimposed white dotted line between the epithelium and stroma, and can be visualized in long fiber structures, presumed to be elastin integrated into elastic fibers. Signal intensity was optimized for day 15, and the same settings were used for day 15 sham and day 15 LPS. Supplemental Figure S3. Representative images of elastic fibers from cervices of day 15 mice treated with mifepristone. Transmission electron microscopy identified highly variable elastic fiber ultrastructure on day 15, 12 h after mifepristone treatment [10]. Representative images of the variable elastic fibers are shown here. Cervices from three different day 15 mifepristone-treated animals and 20–32 fibers per animal were imaged. No major changes in elastic fibers or collagen was observed among the groups. Scale bar: 1000 nm. Supplemental Table S1. Mapping data from RNA-seq libraries. Supplemental Table S2. Forward (F) and reverse (R) primer sequences. Supplemental Table S3. Fold change of canonical parturition genes measured by RNA-seq. Supplemental Table S4. Fold change of proinflammatory genes, measured by RNA-seq. Acknowledgments We would like to thank the UTSW Electron Microscopy Core for help with tissue processing and sample preparation for TEM and the UTSW Live Cell Imaging Core Facility for guidance in imaging. We thank Dr Ann Word for helpful discussions regarding experimental results. Footnotes † Grant Support: NIH R21 HD075228, NIH R01 HD084695, and NIH TL1TR001104. ‡ Accession Number: GSE98545 References 1. Behrman RE, Butler AS (eds.). Preterm Birth: Causes, Consequences, and Prevention . Washington (DC): The National Academies Press; 2007: 31– 52. 2. Martin JA, Hamilton BE, Osterman MJ, Driscoll AK, Mathews TJ. Births: final Data for 2015. Natl Vital Stat Rep  2017; 66: 1. Google Scholar PubMed  3. Martin JA, Hamilton BE, Osterman MJ, Curtin SC, Matthews TJ. Births: final data for 2013. Natl Vital Stat Rep  2015; 64: 1– 65. 4. Romero R, Espinoza J, Kusanovic JP, Gotsch F, Hassan S, Erez O, Chaiworapongsa T, Mazor M. The preterm parturition syndrome. BJOG  2006; 3( 113 Suppl): 17– 42. Google Scholar CrossRef Search ADS   5. Ferrero DM, Larson J, Jacobsson B, Di Renzo GC, Norman JE, Martin JN Jr., D’Alton M, Castelazo E, Howson CP, Sengpiel V, Bottai M, Mayo JA et al.   Cross-country individual participant analysis of 4.1 million singleton births in 5 countries with very high human development index confirms known associations but provides no biologic explanation for 2/3 of all preterm births. PLoS One  2016; 11: e0162506. Google Scholar CrossRef Search ADS PubMed  6. Gomez-Lopez N, Romero R, Xu Y, Plazyo O, Unkel R, Leng Y, Than NG, Chaiworapongsa T, Panaitescu B, Dong Z, Tarca AL, Abrahams VM et al.   A role for the inflammasome in spontaneous preterm labor with acute histologic chorioamnionitis. Reprod Sci  2017; 24: 1382– 1401. Google Scholar CrossRef Search ADS PubMed  7. Romero R, Miranda J, Chaiworapongsa T, Korzeniewski SJ, Chaemsaithong P, Gotsch F, Dong Z, Ahmed AI, Yoon BH, Hassan SS, Kim CJ, Yeo L. Prevalence and clinical significance of sterile intra-amniotic inflammation in patients with preterm labor and intact membranes. Am J Reprod Immunol  2014; 72: 458– 474. Google Scholar CrossRef Search ADS PubMed  8. Mahendroo M, Hoffman B, Cunningham G. Physiology of Labor. In: Cunningham FG, Leveno KJ, Bloom SL, Spong CY, Dashe JS, Hoffman BL, Casey BM, Shefield JS (eds.), Williams Obstetrics , 24 ed. New York: McGraw-Hill Education; 2014: 408– 432. 9. Elovitz MA, Wang Z, Chien EK, Rychlik DF, Phillippe M. A new model for inflammation-induced preterm birth: the role of platelet-activating factor and Toll-like receptor-4. Am J Pathol  2003; 163: 2103– 2111. Google Scholar CrossRef Search ADS PubMed  10. Holt R, Timmons BC, Akgul Y, Akins ML, Mahendroo M. The molecular mechanisms of cervical ripening differ between term and preterm birth. Endocrinology  2011; 152: 1036– 1046. Google Scholar CrossRef Search ADS PubMed  11. Gonzalez JM, Dong Z, Romero R, Girardi G. Cervical remodeling/ripening at term and preterm delivery: the same mechanism initiated by different mediators and different effector cells. PLoS One  2011; 6: e26877. Google Scholar CrossRef Search ADS PubMed  12. McCormack JT, Greenwald GS. Progesterone and oestradiol-17beta concentrations in the peripheral plasma during pregnancy in the mouse. J Endocrinol  1974; 62: 101– 107. Google Scholar CrossRef Search ADS PubMed  13. Timmons BC, Reese J, Socrate S, Ehinger N, Paria BC, Milne GL, Akins ML, Auchus RJ, McIntire D, House M, Mahendroo M. Prostaglandins are essential for cervical ripening in LPS-mediated preterm birth but not term or antiprogestin-driven preterm ripening. Endocrinology  2014; 155: 287– 298. Google Scholar CrossRef Search ADS PubMed  14. Nallasamy S, Akins M, Tetreault B, Luby-Phelps K, Mahendroo M. Distinct reorganization of collagen architecture in lipopolysaccharide - mediated premature cervical remodeling. Biol Reprod  2017. 15. Zhong S, Joung JG, Zheng Y, Chen YR, Liu B, Shao Y, Xiang JZ, Fei Z, Giovannoni JJ. High-throughput illumina strand-specific RNA sequencing library preparation. Cold Spring Harb Protoc  2011; 2011: 940– 949. Google Scholar CrossRef Search ADS PubMed  16. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol  2013; 14: R36. Google Scholar CrossRef Search ADS PubMed  17. Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, van Baren MJ, Salzberg SL, Wold BJ, Pachter L. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol  2010; 28: 511– 515. Google Scholar CrossRef Search ADS PubMed  18. Dennis G Jr., Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol  2003; 4: P3. Google Scholar CrossRef Search ADS PubMed  19. Nallasamy S, Yoshida K, Akins M, Myers K, Iozzo R, Mahendroo M. Steroid hormones are key modulators of tissue mechanical function via regulation of collagen and elastic fibers. Endocrinology  2017; 158: 950– 962. Google Scholar CrossRef Search ADS PubMed  20. Colige A, Vandenberghe I, Thiry M, Lambert CA, Van Beeumen J, Li SW, Prockop DJ, Lapiere CM, Nusgens BV. Cloning and characterization of ADAMTS-14, a novel ADAMTS displaying high homology with ADAMTS-2 and ADAMTS-3. J Biol Chem  2002; 277: 5756– 5766. Google Scholar CrossRef Search ADS PubMed  21. Rajabi MR, Solomon S, Poole AR. Biochemical evidence of collagenase-mediated collagenolysis as a mechanism of cervical dilatation at parturition in the guinea pig. Biol Reprod  1991; 45: 764– 772. Google Scholar CrossRef Search ADS PubMed  22. Asquith DL, Miller AM, Reilly J, Kerr S, Welsh P, Sattar N, McInnes IB. Simultaneous activation of the liver X receptors (LXRalpha and LXRbeta) drives murine collagen-induced arthritis disease pathology. Ann Rheum Dis  2011; 70: 2225– 2228. Google Scholar CrossRef Search ADS PubMed  23. Sandy JD, Westling J, Kenagy RD, Iruela-Arispe ML, Verscharen C, Rodriguez-Mazaneque JC, Zimmermann DR, Lemire JM, Fischer JW, Wight TN, Clowes AW. Versican V1 proteolysis in human aorta in vivo occurs at the Glu 441 -Ala 442 bond, a site that is cleaved by recombinant ADAMTS-1 and ADAMTS-4. J Biol Chem  2001; 276: 13372– 13378. Google Scholar CrossRef Search ADS PubMed  24. Romero R, Xu Y, Plazyo O, Chaemsaithong P, Chaiworapongsa T, Unkel R, Than NG, Chiang PJ, Dong Z, Xu Z, Tarca AL, Abrahams VM et al.   A role for the inflammasome in spontaneous labor at term. Am J Reprod Immunol  2016. 25. Mulla MJ, Salmon JE, Chamley LW, Brosens JJ, Boeras CM, Kavathas PB, Abrahams VM. A role for uric acid and the Nalp3 inflammasome in antiphospholipid antibody-induced IL-1beta production by human first trimester trophoblast. PLoS One  2013; 8: e65237. Google Scholar CrossRef Search ADS PubMed  26. Hoang M, Potter JA, Gysler SM, Han CS, Guller S, Norwitz ER, Abrahams VM. Human fetal membranes generate distinct cytokine profiles in response to bacterial toll-like receptor and Nod-like receptor agonists. Biol Reprod  2014; 90: 39. Google Scholar CrossRef Search ADS PubMed  27. Dinarello CA. Interleukin-1 in the pathogenesis and treatment of inflammatory diseases. Blood  2011; 117: 3720– 3732. Google Scholar CrossRef Search ADS PubMed  28. Shi J, Zhao Y, Wang Y, Gao W, Ding J, Li P, Hu L, Shao F. Inflammatory caspases are innate immune receptors for intracellular LPS. Nature  2014; 514: 187– 192. Google Scholar CrossRef Search ADS PubMed  29. Shenderov K, Riteau N, Yip R, Mayer-Barber KD, Oland S, Hieny S, Fitzgerald P, Oberst A, Dillon CP, Green DR, Cerundolo V, Sher A. Cutting edge: endoplasmic reticulum stress licenses macrophages to produce mature IL-1 in response to TLR4 stimulation through a Caspase-8- and TRIF-dependent pathway. J Immunol  2014; 192: 2029– 2033. Google Scholar CrossRef Search ADS PubMed  30. Hazuda DJ, Strickler J, Kueppers F, Simon PL, Young PR. Processing of precursor interleukin 1 beta and inflammatory disease. J Biol Chem  1990; 265: 6318– 6322. Google Scholar PubMed  31. Herzog C, Haun RS, Kaushal V, Mayeux PR, Shah SV, Kaushal GP. Meprin A and meprin alpha generate biologically functional IL-1beta from pro-IL-1beta. Biochem Biophys Res Commun  2009; 379: 904– 908. Google Scholar CrossRef Search ADS PubMed  32. Mengshol JA, Vincenti MP, Coon CI, Barchowsky A, Brinckerhoff CE. Interleukin-1 induction of collagenase 3 (matrix metalloproteinase 13) gene expression in chondrocytes requires p38, c-jun N-terminal kinase, and nuclear factor kappaB: Differential regulation of collagenase 1 and collagenase 3. Arthritis Rheum  2000; 43: 801– 811. Google Scholar CrossRef Search ADS PubMed  33. Hashimoto Y, Kondo C, Katunuma N. An active 32-kDa Cathepsin L is secreted directly from HT 1080 fibrosarcoma cells and not via lysosomal exocytosis. PLoS One  2015; 10: e0145067. Google Scholar CrossRef Search ADS PubMed  34. Kirschke H, Kembhavi AA, Bohley P, Barrett AJ. Action of rat liver cathepsin L on collagen and other substrates. Biochem J  1982; 201: 367– 372. Google Scholar CrossRef Search ADS PubMed  35. Mason RW, Johnson DA, Barrett AJ, Chapman HA. Elastinolytic activity of human cathepsin L. Biochem J  1986; 233: 925– 927. Google Scholar CrossRef Search ADS PubMed  36. Kelwick R, Desanlis I, Wheeler GN, Edwards DR. The ADAMTS (A disintegrin and metalloproteinase with thrombospondin motifs) family. Genome Biol  2015; 16: 113. Google Scholar CrossRef Search ADS PubMed  37. Nallasamy S, Yoshida K, Akins M, Myers K, Iozzo R, Mahendroo M. Steroid hormones are key modulators of tissue mechanical function via regulation of collagen and elastic fibers. Endocrinology  2017; 158: 950– 962. Google Scholar CrossRef Search ADS PubMed  38. Liggins M. Cervical Ripening as an inflammatory reaction. In: Ellwood DAA, Anne B (ed.), The Cervix in Pregnancy and Labour: Clinical and Biochemical Investigations . Churchill Livingstone: The University of Michigan; 1981: 950– 962. 39. Goldenberg RL, Hauth JC, Andrews WW. Intrauterine infection and preterm delivery. N Engl J Med  2000; 342: 1500– 1507. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2017. Published by Oxford University Press on behalf of Society for the Study of Reproduction. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Biology of Reproduction Oxford University Press

Transcriptome signature identifies distinct cervical pathways induced in lipopolysaccharide-mediated preterm birth,

Loading next page...
 
/lp/ou_press/transcriptome-signature-identifies-distinct-cervical-pathways-induced-R2iG0C10TS
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of Society for the Study of Reproduction. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0006-3363
eISSN
1529-7268
D.O.I.
10.1093/biolre/iox180
Publisher site
See Article on Publisher Site

Abstract

Abstract With half a million babies born preterm each year in the USA and about 15 million worldwide, preterm birth (PTB) remains a global health issue. Preterm birth is a primary cause of infant morbidity and mortality and can impact lives long past infancy. The fact that there are numerous, and many currently unidentified, etiologies of PTB has hindered development of tools for risk evaluation and preventative therapies. Infection is estimated to be involved in nearly 40% of PTBs of known etiology; therefore, understanding how infection-mediated inflammation alters the cervical milieu and leads to preterm tissue biomechanical changes are questions of interest. Using RNA-seq, we identified enrichment of components involved in inflammasome activation and unique proteases in the mouse cervix during lipopolysaccharide (LPS)-mediated PTB and not physiologically at term before labor. Despite transcriptional induction of inflammasome components, there was no evidence of functional activation based on assessment of mature IL1B and IL18 proteins. The increased transcription of proteases that target both elastic fibers and collagen and concentration of myeloid-derived cells capable of protease synthesis in the cervical stroma support the structural disruption of elastic fibers as a functional output of protease activity. The recent demonstration that elastic fibers contribute to the biomechanical function of the pregnant cervix suggests their protease-induced disruption in the infection model of LPS-mediated PTB and may contribute to premature loss of mechanical competency and preterm delivery. Collectively, the transcriptomics and ultrastructural data provide new insights into the distinct mechanisms of premature cervical remodeling in response to infection. Introduction Preterm birth (PTB) is defined as a delivery that occurs prior to 37 weeks of a 40 week gestation in humans. Preterm delivery is a primary cause of infant morbidity and mortality worldwide. Children who are born preterm and survive infancy are at increased risk of medical complications throughout their lives. The financial burden of medical care associated with preterm deliveries exceeds $26 million annually in the USA [1]. The 2015 PTB rate in the USA was 9.6% [2]. Although much research is underway to understand why PTBs occur, the rate peaked at 12.8% (2006) and has only very slowly decreased to its current level over the past 10 years [3]. The sustained high rates of PTB can be attributed to the variety of etiologies, some not yet discovered, of the syndrome [4]. In more than two thirds of cases, an exact cause of PTB cannot be pinpointed [5]. In the realm of PTB, a clinical space filled with unknowns, two facts drive the study outlined here. First, in the one-third of cases with known etiology, infection contributes to 40% of preterm deliveries, and growing evidence supports the role of sterile inflammation as an instigator of PTB [5–7]. Second, regardless of the timing or cause of labor initiation, changes in the cervix precede the onset of labor [8]. A commonly used model of infection-mediated inflammation is intrauterine (IU) administration of lipopolysaccharide (LPS) [9]. Previously published works highlight the distinct cervical features involved in LPS-mediated PTB in mice compared to cervical ripening at term [10,11]. These studies demonstrate that the physiological changes in the tissue that occur in response to a number of hormonal and nonhormonal cues at the end of pregnancy do not occur in LPS-mediated PTB. Rather, a unique set of features help define cervical remodeling in response to an infection or inflammation, more generally. Declining serum progesterone at the end of pregnancy, a process not seen in human, occurs at term and in response to LPS in the mouse [12, 13]. Interestingly, however, serum levels of estradiol do not increase to the same degree in LPS-mediated PTB as they do at term. This difference in the serum progesterone to estrogen ratio—P:E is 80 at term and 528 in LPS preterm—indicates distinct hormone environments in the two groups. Endogenous prostaglandins, specifically PGE2, PGF2a, PGD2, and 6-keto-PGF1a measured in cervical tissue, are increased in the cervix in LPS-mediated PTB and not increased at term. In fact, prostaglandins are essential for LPS-mediated preterm cervical compliance changes and dispensable for term cervical ripening [13]. Additionally, a distinct population of immune cells is present in the cervix during LPS-mediated PTB compared to term before labor. Whereas an influx of activated immune cells such as neutrophils and macrophages does not occur during a normal pregnancy until the postpartum tissue repair phase, these cells are increased in number or activated 6 h after IU LPS treatment on day 15 of mouse gestation, when compared to cells in the cervix of nontreated day 15 mice [10]. Given these features that are unique to an overall inflammatory cervical milieu in LPS-mediated PTB, we sought to better understand the molecular mechanisms underlying the processes by which the cervix responds to infection and inflammation and how these responses lead to preterm delivery. We utilized transcriptome data generated by RNA-seq to identify novel and exclusive pathways in the cervix in LPS-mediated PTB. We have identified enrichment of genes involved in inflammasome activation and upregulation of proteases in the cervix during LPS-mediated PTB exclusively. Subsequent assessment supports a protease-driven disruption of elastic fiber ultrastructure that appears to be achieved in the absence of active IL1B. Collectively, these studies provide insight into distinct features and mechanisms of LPS-mediated preterm cervical remodeling. Materials and methods Mice All animal studies were conducted in accordance with the standards of humane animal care as described in the NIH Guide for the Care and Use of Laboratory Animals. The research protocols were approved by the IACUC office at the University of Texas Southwestern Medical Center. Mice were housed under a 12 h-light/12 h-dark cycle at 22°C. Virgin C57B6/129sv 2–6-month-old female mice were caged with fertile males of the same strain for 6 h. The presence of a vaginal plug at the end of the 6 h was considered as day 0 of pregnancy, with birth of the pups generally occurring early morning day 19. Day 18 samples were collected between 5 and 7 pm before the onset of labor. Inflammation preterm labor model Intrauterine injection of LPS (day 15 LPS) was used to induce preterm labor as previously described [10]. Dosing of LPS was identical to previous studies from our lab [13, 14]. Briefly, mice on day 15 of pregnancy were anesthetized between 7:00 and 9:00 am, and 30 μL sterile water (sham) or 5 mg/mL LPS (Escherichia coli O55:B5 Sigma, St. Louis, MO) was injected IU. Cervical tissues were collected between 1 and 6 h after surgery and before the onset of labor, which occurs approximately 7 to 9 h after LPS administration. RNA sequencing RNA was isolated using an miRNeasy kit from Qiagen (Hilden, Germany). Quantity was determined using a Nanodrop and quality determined using a Bio-Rad Experion. RNAs with RNA Quality Index scores above 9 were used to make cDNA libraries. Eight total cDNA libraries from four different time points/treatment groups (day 15, day 15 sham, day 15 LPS, and day 18) were prepared, with duplicate libraries for biological replicates containing four cervices each. RNA (1.25 μg) from each cervix was pooled for 5 μg of starting RNA in each library. Library preparation was carried out as described previously [15]. One hundred base pair, paired end sequencing was carried out on an Illumina sequencer to a depth of 100 million reads. As indicated in Supplemental Table S1, the total reads for each library was in the range of 8–10 million reads with 94%–95% mappable raw reads. We developed a computational pipeline to determine the differentially expressed genes between the conditions, which included the following steps: reads were aligned to the mm9 genome using the spliced read aligner TopHat version v.2.0.4, transcriptome assembly was carried out using Cufflinks v.2.0.2 with default parameters, filtered transcripts were merged into distinct nonoverlapping sets using Cuffmerge, and Cuffdiff was used to calculate the differential expression genes between the conditions [16, 17]. The RNA-seq data were submitted to GEO with series number GSE98545. Transcriptome data analysis The differentially expressed genes extracted from the above analysis were then used in downstream analyses. Venn Diagrams were generated using Venn Diagram Plotter version 1.5.5228.29250 for the differentially expressed genes in different conditions. Gene Ontology (GO) Analyses and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were determined using DAVID, a web tool for functional annotation and gene enrichment analysis for the genes that are specifically expressed at day 15 LPS treatment as compared to day 18. Heatmaps were generated using Java TreeView for the significantly expressed genes in at least one condition to analyze the effect in the specified condition [18]. Cutoffs used were as follows: 0.5 ≤ fold change ≥ 1.5 and q ≤ 0.05. RNA isolation and quantitative polymerase chain reaction (PCR) Total RNA was isolated as previously described [19]. Complementary DNA synthesis was carried out using 0.5 μg total RNA and 5x iScript Reverse Transcription Supermix (Bio-Rad, Hercules, CA). Quantitative PCR primers used in this study and listed in Supplemental Table S2 were designed and purchased from Invitrogen. Target gene expression was normalized to the expression of housekeeping gene Ppib using the 2∧–ddCt relative gene expression method (User Bulletin no. 2; Applied BioSystems). IL1B enzyme-linked immunosorbent assay (ELISA) Flash-frozen cervical tissues were homogenized in 300 μL Abcam lysis buffer (cat no. 65658, AbCam, UK) and left on ice for 30 min before centrifugation at 4°C for 10 min. Protein concentration was determined (BCA assay, Thermo Scientific, Rockford, IL). IL1B protein was measured via Mouse IL-1B/IL-1F2 DuoSet ELISA (R&D Systems DY401-5) in technical replicates of 100 μL supernatant for day 15, day 15 sham, and day 18 samples while a 3x dilution (33 μL lysate + 66 μL lysis buffer) was used for day 15 LPS samples to maintain the measurements in assay range. Protein blotting for IL1B and IL18 Fifty microgram of whole tissue lysates was boiled for 10 min in Laemmli buffer with 5% β-mercaptoethanol and analyzed by Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on AnykD 15-well precast gels (Bio-Rad) along with protein standards (Precision Plus Protein Kaleidoscope, Bio-Rad) and positive control recombinant proteins (mouse IL1B, 25ng, cat no. 5204SF, Cell Signaling Technology, Danvers, MA; mouse IL18, 100ng, cat no. B0045, Medical & Biological Laboratories CO., LTD, Nagoya, Aichi Japan). Proteins were transferred onto a 0.22-μm nitrocellulose membrane (Bio-Rad) for 30 min at 4°C. Membranes were blocked at room temperature for 1 h in 5% nonfat dry milk in Tris-buffered saline, Tween 20 (TBST) (blocking grade blocker, nonfat dry milk, Bio-Rad). Blots were incubated with primary antibody (IL1B, 1:1000 dilution Cell Signaling Technology cat no. 12507; IL18 1:1000 dilution, Abcam Ab71495; α-tubulin,1:1000 dilution Millipore cat no. 05-829) in blocking solution overnight at 4°C, washed in TBST, and incubated with Horseradish peroxidase (HRP)-labeled secondary antibodies for 45 min at room temperature. The membrane was imaged with ECL Western Blotting Analysis System (GE Healthcare, Buckinghampshire, UK). Immunofluorescence for Elastin Studies were carried out as previously described [19]. Briefly, paraffin sections were deparaffinized in ethanol baths and rinsed in Phosphate-buffered saline (PBS). Antigen retrieval was carried out using 6M Guanidine HCl and iodoacetate. Sections were blocked for 30 min with normal goat serum (NGS) before primary antibody (1:250 dilution, Elastin Products Company cat #PR385, Owensville, MO) incubation in NGS overnight 4°C. Sections were washed in PBS then incubated with Alexa Fluor 546 goat anti-rabbit antibody (1:500, Invitrogen cat# A-11035, Carlsbad, CA) in NGS for 30 min at room temperature in the dark. Coverslips were mounted atop sections with 4΄,6-Diamidino-2-phenylindole dihydrochloride (DAPI) Prolong Gold (Life Technologies, Eugene, OR) and imaged using a Zeiss LSM880 microscope. Two-photon excited immunofluorescence Fifty micrometer thick frozen transverse sections were imaged on a Zeiss LSM880 Laser Scanning confocal microscope with multiphoton (NLO) laser configured with AxioObserver inverted microscope using an Achroplan 40x/0.8 W objective lens (Zeiss, Jena, Germany). A Coherent Chameleon Ultra Ti:Sapphire multiphoton tunable laser (Coherent Inc., Santa Clara, CA) tuned to 860 nm and laser output of 13.5% was focused on the subepithelial stroma of the endocervix. Z stacks of each frozen section were acquired by imaging at 4-μm intervals within the thickness of the 50-μm frozen section. Z-stack images were opened in ImageJ (version 1.51k; NIH, Bethesda, USA); six centrally located slices were compressed into Z-projections of maximum intensity. Brightness and contrast of each channel of the image were adjusted equally across experimental conditions. Transmission electron microscopy Transmission electron microscopy (TEM) was carried out as previously described [19]. Briefly, pregnant mice were perfused with heparinized saline, then glutaraldehyde and paraformaldehyde fixatives in sodium cacodylate buffer. Cervical tissue was removed and fixed in glutaraldehyde in sodium cacodylate buffer overnight at 4°C. The cervix was then sliced in transverse sections and processed as previously described [19]. A Tecnai G2 spirit transmission electron microscope at 120 kV and a side-mounted SIS Morada 11 megapixel CCD camera were used for image acquisition. Immunofluorescence staining for F4/80+ cells Immunofluorescence was carried out for F4/80+ cells in day 15, day 15 sham, day 15 LPS, and day 18 mouse 5-μm paraffin embedded cervix sections. Sections were deparaffinized in ethanol baths, rinsed in PBS, and treated with Proteinase K for 3 min at room temperature before being rinsed with PBS and blocked at room temperature for 30 min with NGS. Sections were incubated with F4/80 antibody (Serotec, Raleigh, NC, rat anti mouse F4/80) at 1:500 dilution overnight at 4°C. The next day, sections were washed in PBS then incubated with Alexa Fluor 488 goat anti-rat antibody (1:500, Invitrogen cat# A-11035, Carlsbad, CA) in NGS for 30 min at room temperature in the dark. Coverslips were mounted atop sections with DAPI Prolong Gold (Life Technologies, Eugene, OR) and imaged using a Zeiss LSM880 microscope and 40x lens. Statistical analysis Statistics were performed using Prism software (GraphPad Software). For comparison of multiple groups, one-way ANOVA was used followed by Tukey multiple comparisons test. The values were expressed as mean ± SEM and considered significant if P < 0.05. Number of animals used and data analysis are described in the figure legends. Results RNA-seq identifies unique transcriptomes in lipopolysaccharide-mediated preterm and term cervical ripening To elucidate the transcriptional pathways that guide cervical remodeling in response to inflammation, we utilized RNA Sequencing. This method of gene expression analysis was undertaken to identify novel and exclusive pathways induced in the cervix of the LPS-mediated PTB model versus mice at term before labor. To determine changes in gene expression during LPS-mediated PTB, we compared day 15 LPS to untreated day 15 (Figure 1A). To determine gene expression changes during term cervical ripening, we compared day 18 to day 15. To ensure the changes in gene expression observed upon LPS treatment were due to the LPS and not due to the survival surgery, sham to day 15 comparisons were also included. Overall gene expression patterns between LPS preterm and term were quite distinct (Figure 1B). The number of differentially expressed genes exclusive to day 15 LPS vs day 15 (2918 genes) was large, compared to day 18 vs day 15 (475 genes) (Figure 1C). Relatively few genes (267) were common between the two groups and differentially expressed as compared to day 15. A similar number of genes were up- and downregulated in the day 15 LPS group compared to gestation day 15, while the majority of differentially expressed genes were upregulated in the day 18 group (Figure 1D). In agreement with previous reports using quantitative PCR (qPCR), gene expressions of canonical parturition genes including connexin 26 (gap junction protein, beta 2 [Gjb2] (also connexin 26)), hyaluronan synthase 2 (Has2), and steroid 5 alpha reductase 1 (Srd5a1), as measured using RNA-seq, are upregulated at term before labor and not changing in a statistically significant manner upon LPS treatment on day 15 (Supplemental Table S3) [10]. Reciprocally, proinflammatory genes previously reported via qPCR to be upregulated upon LPS treatment on day 15 and not changing at term before labor, in particular cyclooxygenase 2 (Ptgs2), prostaglandin E synthase (Ptges), interleukin 6 (Il6), interleukin 1 alpha (Il1a), and C-X-C motif chemokine ligand 2 (Cxcl2), are identified as upregulated via RNA-seq in day 15 LPS and not on day 18, both compared to day 15 (Supplemental Table S4) [10]. Figure 1. View largeDownload slide RNA-seq analyses of the pregnant mouse cervix demonstrate unique transcriptome profiles in term and LPS-mediated preterm cervical remodeling. (A) Schematic indicating time points (day 15 and day 18) and treatment groups (day 15 sham and day 15 LPS) used in RNA-seq. Two libraries of four cervices each were constructed for each of the four conditions. (B) Heatmap of all coding transcripts in day 15 LPS vs day 15 (LPS preterm) and day 18 vs day 15 (term). Fold changes (from 0.012 to 463) comparing to day 15 to day 15 LPS and day 18 are shown. (C) Venn diagram with the total number of statistically significantly and differentially expressed transcripts in each group (day 15 sham, day 15 LPS, and day 18) compared to day 15. (D) Table with numbers of statistically significant upregulated and significantly downregulated transcripts in each condition (day 15 sham, day 15 LPS, and day 18) compared to day 15. Figure 1. View largeDownload slide RNA-seq analyses of the pregnant mouse cervix demonstrate unique transcriptome profiles in term and LPS-mediated preterm cervical remodeling. (A) Schematic indicating time points (day 15 and day 18) and treatment groups (day 15 sham and day 15 LPS) used in RNA-seq. Two libraries of four cervices each were constructed for each of the four conditions. (B) Heatmap of all coding transcripts in day 15 LPS vs day 15 (LPS preterm) and day 18 vs day 15 (term). Fold changes (from 0.012 to 463) comparing to day 15 to day 15 LPS and day 18 are shown. (C) Venn diagram with the total number of statistically significantly and differentially expressed transcripts in each group (day 15 sham, day 15 LPS, and day 18) compared to day 15. (D) Table with numbers of statistically significant upregulated and significantly downregulated transcripts in each condition (day 15 sham, day 15 LPS, and day 18) compared to day 15. Gene Ontology Analyses identify exclusive pathways in lipopolysaccharide-mediated preterm birth To gain insight into potential pathways driving LPS-mediated premature cervical changes and term cervical ripening, GO Analyses were performed. The most enriched processes from the group of genes upregulated in LPS preterm exclusively included immune, defense, and inflammatory responses and response to wounding (Figure 2A). Analyses of genes upregulated exclusively at term before labor identified the biological processes lipid catabolic process, female pregnancy, epidermis development, and ectoderm development as the most enriched. Biological processes from genes upregulated both in LPS preterm and term before labor include positive regulation of developmental process, keratinization, positive regulation of biological process, and positive regulation of cellular processes (Figure 2A). Further in-depth comparisons of GO pathways based on Molecular Function, Biological Processes, and KEGG Pathways specific to the LPS preterm group identify a common theme consistent with NFkB activation, pathogen sensing, inflammasome components, and downstream endpoints of inflammasome activation (Figure 2B). Given the critical role of inflammasome activation in host defense against pathogens, sterile insults, and host-derived molecules via activation of downstream proinflammatory events, we focused our studies on inflammasome activation in the cervix as a novel pathway specific to LPS-mediated PTB. Figure 2. View largeDownload slide GO and KEGG pathway analyses of term and LPS-mediated preterm cervical remodeling. (A) GO analyses demonstrate enrichment of unique biological processes in term and LPS-mediated PTB. The most statistically significant (as indicated by log10P-value) processes with genes expressed exclusively in each of the groups and those commonly expressed between them (common) are presented. Percentages indicate the percentage of genes in the process represented in the data set. (B) GO analyses (nucleotide binding related molecular functions and cell death biological processes) and KEGG analysis (PRRs related pathways) of genes expressed in LPS preterm exclusively. Components of inflammasome activation fall within several of the indicated pathways and processes. Figure 2. View largeDownload slide GO and KEGG pathway analyses of term and LPS-mediated preterm cervical remodeling. (A) GO analyses demonstrate enrichment of unique biological processes in term and LPS-mediated PTB. The most statistically significant (as indicated by log10P-value) processes with genes expressed exclusively in each of the groups and those commonly expressed between them (common) are presented. Percentages indicate the percentage of genes in the process represented in the data set. (B) GO analyses (nucleotide binding related molecular functions and cell death biological processes) and KEGG analysis (PRRs related pathways) of genes expressed in LPS preterm exclusively. Components of inflammasome activation fall within several of the indicated pathways and processes. Components of the inflammasome are upregulated in lipopolysaccharide-mediated PTB Genes encoding components of the inflammasome activation pathway were identified in the RNA-seq data set as upregulated in LPS preterm and not changing at term before labor (Figure 3A). These include the pattern recognition receptors (PRRs) Nod 2, Nrlc4, Nrlc5, Nrlp3, activator protein Gbp5, executioner proteins Casp1 and Casp4, and cytokine Il1b. Gene expression analysis by quantitative PCR corroborates RNA-seq results for the majority of genes and shows an increase in gene expression in inflammation preterm (day 15 LPS vs day 15) and no change at term before labor (day 18 vs day 15) for Gbp5, Nlrc5, and Nod2 (Figure 3B). Transcripts of Casp4 were significantly induced 2, 4, and 6 h after LPS treatment, compared to untreated day 15 and not at term (Figure 3C). Using quantitative PCR, transcripts encoding Nlrp3, Nlrc4, and Casp1 were not significantly induced in the LPS-mediated PTB group or at term before labor (day 18). Figure 3. View largeDownload slide Components of the inflammasome are upregulated in LPS-mediated PTB exclusively. (A) Heatmap of select inflammasome-related genes and their expression pattern in day 15 LPS cervices vs day 15 cervices (LPS preterm) and in day 18 cervices vs day 15 cervices (term). Fold changes (from 0 to 35.5) comparing to day 15 to day 15 LPS and day 18 are shown. (B) qPCR results of selected inflammasome genes showing relative gene expression on day 15, day 15 sham, day 15 LPS, and day 18. The values expressed are mean ± SEM. (n = 5–7 cervices per group, one-way ANOVA *P < 0.05 relative to day 15). (C) qPCR results for Casp1 and Casp4 showing relative gene expression on day 15, day 15 sham, day 15 after 1, 2, 4, or 6 h LPS treatment, and day 18. The values expressed are mean ± SEM. (n = 5–6 cervices per group, one way ANOVA *P < 0.05 relative to day 15). Figure 3. View largeDownload slide Components of the inflammasome are upregulated in LPS-mediated PTB exclusively. (A) Heatmap of select inflammasome-related genes and their expression pattern in day 15 LPS cervices vs day 15 cervices (LPS preterm) and in day 18 cervices vs day 15 cervices (term). Fold changes (from 0 to 35.5) comparing to day 15 to day 15 LPS and day 18 are shown. (B) qPCR results of selected inflammasome genes showing relative gene expression on day 15, day 15 sham, day 15 LPS, and day 18. The values expressed are mean ± SEM. (n = 5–7 cervices per group, one-way ANOVA *P < 0.05 relative to day 15). (C) qPCR results for Casp1 and Casp4 showing relative gene expression on day 15, day 15 sham, day 15 after 1, 2, 4, or 6 h LPS treatment, and day 18. The values expressed are mean ± SEM. (n = 5–6 cervices per group, one way ANOVA *P < 0.05 relative to day 15). A functional consequence of canonical and noncanonical inflammasome complex formation and activation is the secretion of the cytokines IL1B and IL18 and induction of pyroptosis. To investigate a functional role for inflammasome activation in inflammation preterm, transcripts of Il1b were analyzed using qPCR (Figure 4A). Il1b was significantly upregulated 2, 4, and 6 h after LPS treatment on day 15, compared to untreated day 15 samples. IL1B protein levels were measured by ELISA in whole cervical tissue lysates from untreated day 15 mice, day 15 sham mice, day 15 LPS mice, and day 18 mice (Figure 4B). IL1B was measureable in each of the day 15 LPS samples tested and amounts were significantly increased compared to nontreated day 15 (Figure 3B). To determine if the increase in IL1B protein was the mature and biologically active 17kDa form, protein blotting was undertaken using 50 μg whole cervical tissue lysate taken from mice exposed to LPS for 0, 1, 2, 4, and 6 h (Figure 4C). Although the time-dependent exposure to LPS increased synthesis of pro-IL1B, mature IL1B protein was not detectable at any time point after treatment. The time-dependent increase in pro-IL1B abundance did not achieve statistical significance, despite the marked temporal increase visualized in all four experiments (Figure 4D). The protein expression of mature IL18 was also evaluated in cervical tissue of mice treated with LPS (Figure 4E). Similar to IL1B, no mature IL18 protein was detectable. Figure 4. View largeDownload slide Profile of IL1B and IL18 as outputs of the inflammasome in LPS-mediated PTB. (A) Relative gene expression of Il1b in day 15 (NT), day 15 sham (sham), 1, 2, 4, and 6 h after LPS treatment. The values expressed are mean ± SEM. (n = 5–7 cervices per group, one way ANOVA *P < 0.05 relative to day 15). (B) ELISA results for IL1B in whole tissue lysates from the cervix of day 15, day 15 sham, day 15 LPS, and day 18 mice. Three separate experiments were performed with a total of 13–15 total cervices per group. One-way ANOVA *P < 0.05 relative to day 15. (C) Representative immunoblot for pro (31 kDa) and mature (17 kDa) forms of IL1B in mouse cervix tissues on day 15 without treatment (NT) and 1, 2, 4, and 6 h after LPS treatment. As a positive control, 25 ng recombinant mature mouse IL1B (mIL1B) was run in the first lane. A time-dependent increase in pro-IL1B is observed with no detectable expression of the mature form. Four separate experiments were run with two samples pooled per treatment group in each experiment. α-tubulin was used as a loading control. (D) Quantification of average pro-IL1B abundance from four independent blotting experiments. Pro-IL1B abundance is expressed as a ratio compared to alpha-tubulin and normalized to untreated day 15. (E) Representative immunoblot for mouse IL18 in mouse cervix tissues on day 15 without treatment (NT) and 1, 2, 4, and 6 h after LPS treatment with 100 ng positive control recombinant mouse IL18 (18 kDa) run in the first lane. No expression of active form of IL18 was found in any sample. Four independent experiments were run with two samples pooled per treatment group in each experiment. α-tubulin was used as a loading control. Figure 4. View largeDownload slide Profile of IL1B and IL18 as outputs of the inflammasome in LPS-mediated PTB. (A) Relative gene expression of Il1b in day 15 (NT), day 15 sham (sham), 1, 2, 4, and 6 h after LPS treatment. The values expressed are mean ± SEM. (n = 5–7 cervices per group, one way ANOVA *P < 0.05 relative to day 15). (B) ELISA results for IL1B in whole tissue lysates from the cervix of day 15, day 15 sham, day 15 LPS, and day 18 mice. Three separate experiments were performed with a total of 13–15 total cervices per group. One-way ANOVA *P < 0.05 relative to day 15. (C) Representative immunoblot for pro (31 kDa) and mature (17 kDa) forms of IL1B in mouse cervix tissues on day 15 without treatment (NT) and 1, 2, 4, and 6 h after LPS treatment. As a positive control, 25 ng recombinant mature mouse IL1B (mIL1B) was run in the first lane. A time-dependent increase in pro-IL1B is observed with no detectable expression of the mature form. Four separate experiments were run with two samples pooled per treatment group in each experiment. α-tubulin was used as a loading control. (D) Quantification of average pro-IL1B abundance from four independent blotting experiments. Pro-IL1B abundance is expressed as a ratio compared to alpha-tubulin and normalized to untreated day 15. (E) Representative immunoblot for mouse IL18 in mouse cervix tissues on day 15 without treatment (NT) and 1, 2, 4, and 6 h after LPS treatment with 100 ng positive control recombinant mouse IL18 (18 kDa) run in the first lane. No expression of active form of IL18 was found in any sample. Four independent experiments were run with two samples pooled per treatment group in each experiment. α-tubulin was used as a loading control. Expression of proteases in the cervix is upregulated in lipopolysaccharide-mediated PTB Another group of genes identified as upregulated in LPS-mediated premature cervical remodeling is proteases (Figure 5). RNA-seq data set analysis identified two GO terms—proteolysis and peptidase activity—as significantly enriched in genes exclusive to day 15 LPS vs day 15 (Figure 5A). The data set contains a large number of genes (107 and 75 respectively) within these pathways (Figure 5A). These genes, including matrix metalloproteinase (Mmp), a disintegrin and metalloproteinase with thrombospondin type 1 motifs (Adamts), and cathepsin (Cts) proteases, target and modify components of the extracellular matrix (ECM). Known functions of these proteases include postsynthesis processing of mature ECM molecules including collagen, degradation during physiologic turnover processes (i.e., pregnancy), and pathologic breakdown due to inflammation (i.e., arthritis) [20–22]. The heatmap in Figure 5B depicts fold change values of genes encoding ADAMTS, MMP, and CTS proteases identified as differentially regulated in LPS preterm or term cervical ripening. Gene expression analyses using RT-qPCR for select genes Mmp13, Mmp12, Mmp8, Ctsl, Ctsc, Adamts1, Adamts5, and Adamts15 identified significantly increased expression of Mmp13, Adamts1, and Adamts5 in cervices from day 15 LPS-treated mice (Figure 5C). While there was a trend toward an increase in Ctsl gene expression that did not achieve statistical significance, subsequent studies shown in Supplemental Figure S1 using time course samples 1, 2, 4, and 6 h after LPS treatment indicate an increase in Ctsl expression 6 h after LPS treatment on day 15. We have previously reported increased expression of Mmp8, Adamts1, and Adamts4 in the cervix of day 15 LPS-treated mice [10]. Both Adamts1 and Adamts4 have been shown to be upregulated in term ripening at gestation day 18, although the changes seen in these quantitative PCR experiments for Adamts1 do not reach statistical significance on day 18 (Figure 5C). The known targets of Adamts1 and Adamts4 include proteoglycans [10, 23]. The upregulation of proteases that target the main structural components collagen and elastic fibers in response to pathological inflammation but not in the normal catabolic processes led to assessment of the cervical ECM structure in the day 15 LPS-treated mice. Figure 5. View largeDownload slide Protease genes and pathways are upregulated in the cervix during LPS-mediated PTB exclusively. (A) GO biological process proteolysis and molecular function peptidase activity are enriched in the gene set exclusive to LPS-mediated PTB. Count refers to the number of genes from the data set that fall into the GO pathways. (B) Heatmap of selected proteases and their expression pattern in day 15 LPS cervices vs day 15 cervices (LPS preterm) and in day 18 cervices vs day 15 cervices (term). (C) qPCR results of selected protease genes showing relative gene expression on day 15, day 15 sham, day 15 LPS, and day 18. The values expressed are mean ± SEM. (n = 4–6 cervices per group, one-way ANOVA *P < 0.05 relative to day 15). Figure 5. View largeDownload slide Protease genes and pathways are upregulated in the cervix during LPS-mediated PTB exclusively. (A) GO biological process proteolysis and molecular function peptidase activity are enriched in the gene set exclusive to LPS-mediated PTB. Count refers to the number of genes from the data set that fall into the GO pathways. (B) Heatmap of selected proteases and their expression pattern in day 15 LPS cervices vs day 15 cervices (LPS preterm) and in day 18 cervices vs day 15 cervices (term). (C) qPCR results of selected protease genes showing relative gene expression on day 15, day 15 sham, day 15 LPS, and day 18. The values expressed are mean ± SEM. (n = 4–6 cervices per group, one-way ANOVA *P < 0.05 relative to day 15). Elastic fiber ultrastructure is disrupted in lipopolysaccharide-mediated preterm birth Recently published studies from our lab in the mouse cervix highlight the contribution of elastic fiber structural organization to the biomechanical function of the cervix [19]. Furthermore, these studies document an increased elastic fiber density in the stroma adjacent to the epithelia, termed the subepithelial stroma, as compared to the midstroma region. Elastic fibers are composed of tropoelastin protein crosslinked in and around a microfibrillar scaffold consisting of fibrillin 1 and 2, plus a number of other proteins [19]. To visualize potential alterations in cervical elastin morphology in the LPS-mediated PTB, an antibody generated against tropoelastin was utilized for immunofluorescence staining. Tropoelastin staining in day 15 LPS appeared similarly localized in the subepithelial stroma, and long elastin strands were evident to a similar degree as compared to sham day 15 and untreated day 15 controls (Figure 6). Images of separate DAPI and tropoelastin channels are shown in Supplemental Figure S2. In addition, two-photon excited fluorescence (TPEF) was utilized as a second approach to assess elastic fiber structure. Similar to the tropoelastin immunofluorescence results, elastin fiber structure in TPEF images was indistinguishable between day 15 LPS and day 15 control groups (Figure 6). Subsequent analysis was carried out to evaluate elastic fiber ultrastructure by TEM, which allows for assessment of both elastin and the microfibrillar scaffold of the fiber (Figure 7). Elastic fibers in the subepithelial stroma region of the cervix were evaluated in day 15, day 15 sham, day 15 LPS, and day 18 mice. Overall, elastic fibers in the day 15 LPS group appeared disrupted with less darkly stained elastin integrated into the scaffold, resulting in increased visibility of the microfibrillar scaffold component of the elastic fiber (Figure 7). Cervical elastic fibers from sham animals were similar in structure and density to untreated day 15. No such disruption of elastic fiber ultrastructure occurs at term before labor. Comparing day 18 to day 15, elastic fibers appear intact with the darkly stained elastin covering the microfibrillar scaffold. In contrast to the LPS model, disrupted elastic fibers were a fraction of otherwise normal term fibers in the mifepristone-PTB model (Supplemental Figure S3). Figure 6. View largeDownload slide Immunofluorescence staining for tropoelastin and TPEF show similar cervical elastin morphology upon LPS treatment on day 15. (Top row) Immunofluorescence imaging of tropoelastin protein (red) and nuclei by DAPI (blue) in cervical sections from day 15, day 15 sham, and day 15 LPS mice. Tropoelastin is most concentrated in the subepithelial stroma region, directly below the superimposed white dotted line between the epithelium and stroma, and can be visualized in long-fiber structures, presumed to be tropoelastin integrated into elastic fibers. The overall impression from this experiment is that elastin appears similar in location and morphology in all three groups. Five to six images in each cervical section for 3–5 mice per group. Images shown here are representative of each group. Signal intensity was optimized for day 15, and the same settings were used for day 15 sham and day 15 LPS. (Bottom row) Representative Z-projected images of autofluorescence of elastic fibers imaged using TPEF. Images were taken in the subepithelium of the endo cervix of day 15, day 15 sham, and day 15 LPS-treated mice. Images were obtained from 5–7 regions per tissue and from 3–4 mice per treatment group. No major change in elastic fiber localization or morphology was seen among the groups. Scale bar: 20 μm. Figure 6. View largeDownload slide Immunofluorescence staining for tropoelastin and TPEF show similar cervical elastin morphology upon LPS treatment on day 15. (Top row) Immunofluorescence imaging of tropoelastin protein (red) and nuclei by DAPI (blue) in cervical sections from day 15, day 15 sham, and day 15 LPS mice. Tropoelastin is most concentrated in the subepithelial stroma region, directly below the superimposed white dotted line between the epithelium and stroma, and can be visualized in long-fiber structures, presumed to be tropoelastin integrated into elastic fibers. The overall impression from this experiment is that elastin appears similar in location and morphology in all three groups. Five to six images in each cervical section for 3–5 mice per group. Images shown here are representative of each group. Signal intensity was optimized for day 15, and the same settings were used for day 15 sham and day 15 LPS. (Bottom row) Representative Z-projected images of autofluorescence of elastic fibers imaged using TPEF. Images were taken in the subepithelium of the endo cervix of day 15, day 15 sham, and day 15 LPS-treated mice. Images were obtained from 5–7 regions per tissue and from 3–4 mice per treatment group. No major change in elastic fiber localization or morphology was seen among the groups. Scale bar: 20 μm. Figure 7. View largeDownload slide Elastic fiber ultrastructure is disrupted in the cervix of LPS-treated mice before onset of PTB. TEM analysis of elastic fibers in cervices from day 15, day 15 sham, day 15 IU LPS, and day 18 mice. Elastic fiber ultrastructure (black arrows) is abnormal in day 15 LPS compared to day 15 and sham. (n = 3–4 mice per group). A total of 11–62 elastic fibers were imaged per sample. Scale bar: 1000 nm. Figure 7. View largeDownload slide Elastic fiber ultrastructure is disrupted in the cervix of LPS-treated mice before onset of PTB. TEM analysis of elastic fibers in cervices from day 15, day 15 sham, day 15 IU LPS, and day 18 mice. Elastic fiber ultrastructure (black arrows) is abnormal in day 15 LPS compared to day 15 and sham. (n = 3–4 mice per group). A total of 11–62 elastic fibers were imaged per sample. Scale bar: 1000 nm. Localization of myeloid cells in cervical stroma during lipopolysaccharide-mediated preterm birth Recently published work from our lab has demonstrated that 7/4 + cells cluster in the subepithelial stroma region of the mouse cervix on day 15 upon LPS treatment, compared to untreated day 15 samples [14]. We sought to investigate the localization of other immune cells that are capable of producing and secreting proteases that have been identified as upregulated in response to LPS and potentially targeting the elastic fibers identified as disrupted in these tissues. Immunofluorescence was carried out in day 15, day 15 sham, day 15 LPS, and day 18 samples to investigate the localization of immune cells, namely F4/80+ macrophages. F4/80+ cells are equally distributed throughout the mid and subepithelial stroma in day 15, day 15 sham, day 15 LPS, and day 18 samples (Figure 8). Consistent with previous analysis of cervical F4/80+ macrophages in the LPS model by flow cytometry, there was no apparent increase in macrophage numbers in the LPS group as compared to other groups [10]. Figure 8. View largeDownload slide F4/80+ cells are present in the mid and subepithelial cervical stroma. The localization of F4/80+ cells (green) was determined using immunofluorescence techniques on paraffin sections from cervices of mice on day 15, day 15 sham, day 15 LPS, and day 18 mice. N = 3 animals per group, 3–8 images taken in each the sub-epithelial stroma and the mid-stroma. Representative images are shown. Figure 8. View largeDownload slide F4/80+ cells are present in the mid and subepithelial cervical stroma. The localization of F4/80+ cells (green) was determined using immunofluorescence techniques on paraffin sections from cervices of mice on day 15, day 15 sham, day 15 LPS, and day 18 mice. N = 3 animals per group, 3–8 images taken in each the sub-epithelial stroma and the mid-stroma. Representative images are shown. Discussion Elucidation of the distinct pathways that drive LPS-mediated preterm cervical remodeling but not cervical ripening at term will advance the development of clinically relevant therapies to prevent premature changes in the cervix that lead to preterm deliveries. This study builds upon previous findings that demonstrate distinct features of the cervical remodeling process in response to inflammation are not present at term before labor [10, 13, 19]. The distinct transcriptome signatures uncovered in this study add to the field's understanding of the differences between LPS-mediated and term cervical remodeling. In particular, LPS does not alter expression of many genes regulated in term cervical ripening, while pathways related to and downstream of inflammatory responses were most dramatically induced upon LPS treatment. While much valuable information remains to be extracted from the RNA-seq data sets, in the present study, we have leveraged these data to identify transcriptional pathways that may direct features of LPS-mediated preterm cervical remodeling—including proinflammatory responses and regionalized disorganization of collagen fibers in the stroma—that have previously been identified [13, 14]. Specifically, we have explored pathways that include components of an activated inflammasome and protease action in the ECM. These pathways in the cervix are exclusive to LPS-mediated PTB. Recent studies implicate inflammasomes as mediators of both sterile inflammation leading to spontaneous labor at term and pathological inflammation leading to preterm delivery both in women with antiphospholipid syndrome and women with acute histologic chorioamnionitis [6, 24, 25]. Endogenous danger signals, termed damage-associated molecular patterns (DAMPs), resulting from cellular stresses can trigger sterile inflammation. Pathogens, which are recognized by PRRs, and subsequent DAMP induction can induce pathological inflammation. Components of inflammasomes are expressed in the chorioamnion of women at term, and there is evidence of increased caspase-1 activation and mature IL1B in tissues from women at term and in labor compared to nonlaboring term tissues [24]. In response to pathogens recognized by a number of PRRs, human fetal membrane explants can mount an impressive immune response (i.e., secrete cytokines), indicating these tissues express machinery needed for inflammasome activation [6, 24, 26]. Chorioamnion membranes from women with acute histologic chorioamnionitis at term show increased PRR gene expression, protein abundance, and downstream inflammasome activation as measured via caspase-dependent IL1B production in vivo [6]. In contrast to evidence that inflammasome-mediated increases in caspase 1 activation and mature IL1B play a role in fetal membrane signaling for the initiation of labor both at term and with pathogen-mediated inflammation, our data in the pregnant mouse cervix demonstrate the transcriptional activation of inflammasome components and caspase 4 in response to LPS but no evidence for inflammasome-dependent production of biologically active IL1B or IL18. These findings are consistent with previous reports that caspase-mediated generation of bioactive IL1B requires induced expression of pro-IL1B by a toll-like receptor ligand such as LPS followed by a secondary signal by DAMPs to induce pro-IL1B processing and secretion [27]. While caspase 1 mRNA expression was not altered with LPS treatment, the induction of caspase 4 is potentially sufficient as work in other infection and inflammatory disease states demonstrates that caspase-4 can directly sense LPS and induce a noncanonical inflammasome activation reaction [28]. Biologically active IL1B can also be generated by caspase-1/4 independent pathways that include caspase-8, neutrophil elastase, cathepsin G, granzyme A, and matrix metalloproteases [29–31]. While the potential role of inflammasome-mediated production of bioactive IL1B in premature ripening induced by ascending pathogens that are not represented by the LPS PTB model (e.g., gram + bacteria and/or viruses) warrants further study, an important observation from the current findings is the evidence that protease-mediated disruption of cervical elastic fibers and premature ripening can be achieved in the absence of bioactive IL1B. Because multiple inflammation-related pathways are induced in the cervix exclusively during LPS-mediated PTB and not at term, we anticipate that other inflammatory pathways, including NFkB activation pathways, are sufficient to accomplish the premature changes in cervical ripening in LPS-mediated PTB. Future studies will investigate the inflammatory pathways that are required to achieve premature remodeling. The cervical transcriptional signature of LPS-treated mice includes robust upregulation of proteases, including Mmp13, Mmp8, and Ctsl, that target the major structural proteins of the cervix, namely collagen and components of the elastic fiber (Figure 5; Supplemental Figure S1). The process by which infection and inflammation induce upregulation of these particular collagen and elastin targeting proteases in the cervix we hypothesized would be similar to the pathophysiology of the proinflammatory disease osteoarthritis in which IL1B has been shown to directly upregulate proteases such as collagenase Mmp13 [32]. While transcriptional upregulation of the Il1b transcript precedes the upregulation of proteases such as Mmp13 and Ctsl (Figure 4; Supplemental Figure S1), the observed disruption of elastic fiber ultrastructure in the absence of active IL1B suggests an IL1B independent pathway of protease activation. Cathepsin L, a protease targeting both elastin and collagen, is a lysosomal and secreted enzyme that has yet to be appreciated in the context of cervical remodeling [33–35]. Its expression is exclusive to day 15 LPS cervices and is not upregulated at term before labor, indicating further studies are warranted to investigate its potential role in the cervical elastic fiber disruption seen in LPS-mediated PTB. While neither transcripts encoding neutrophil elastase, Elane, or macrophage elastase, Mmp12 were elevated in cervical tissue with LPS treatment, the protein synthesis and packaging of these proteases into granules occurs prior to infiltration of these immune cells into tissue, and thus transcriptional changes would not be anticipated. Future investigations to demonstrate activity of proteases that target elastic fibers are warranted. While proteases that target collagen and elastic fibers are upregulated in the cervix specifically in LPS-mediated PTB, protease members of the ADAMTS family including Adamts1 and Adamts4 are upregulated both at term and in LPS-mediated PTB. These proteases are known to target proteoglycans and likely contribute to the normal cervical ECM turnover required for physiological remodeling [23, 36]. In the current study, a potential functional outcome of increased cervical protease gene expression in LPS-mediated PTB is the observed disruption of elastic fiber ultrastructure. In contrast, disrupted elastic fibers were not evident in TEM images from term before labor and were a minor subset of elastic fibers in mifepristone-induced preterm cervical ripening (Supplemental Figure S3).We have previously reported that the density of elastic fibers is greatest in the subepithelial region of the cervical stroma and that elastic fibers along with collagen fibers dictate the mechanical strength of the cervix [37]. Recent studies from our group focused on understanding the structural changes in cervical collagen upon LPS treatment demonstrate a preferential disorganization of collagen fiber structure in the subepithelial stroma, as determined by second harmonic generation and subsequent image analysis [14]. Collectively, these findings suggest that LPS-driven activation of proteases leads to a pathological disruption of the cervical ECM, which allows for mechanical weakening of the cervix and subsequent PTB. We suggest that this model, which Mont Liggins once proposed as the normal physiology of term cervical ripening, is a key and distinct mechanism of preterm cervical remodeling in a mouse model of inflammation-mediated PTB [38]. Based on insights gained from the cervical transcriptome signature of LPS-mediated PTB, we propose a model in which an IL1B-independent pathway upregulates proteases, likely produced by immune cells in addition to other cells of the cervix, that target and degrade ECM components, namely elastic fibers and collagen in the subepithelial stroma, leading to disrupted ECM structure, decreased biomechanical integrity, and leading to PTB (Figure 7) [14]. These data presented here, including cervical transcriptome pathways induced in LPS-mediated PTB exclusively and altered elastic fiber ultrastructure, connect to form a plausible pathway linking infection to pathological ECM structural changes and subsequent loss of biomechanical integrity leading to premature delivery. While LPS treatment provides insight into responses anticipated from gram-negative bacteria, we expect that the described pathway will be conserved in response to other vaginal bacteria such as Ureaplasma, Mycoplasma, and Gardnerella species, which have been linked to spontaneous preterm deliveries [39]. Further mining of this data set in addition to other infection and inflammation models may help elucidate additional pathways for investigation. The clinical implications of this work extend beyond an increased understanding of the distinct pathways that mediate PTB in the face of infection-induced inflammation. PTB is the most robust risk factor for future PTBs [5]. The pathological disruption of elastic fibers in LPS-mediated preterm cervical remodeling may have an impact on cervical function in future pregnancies. If elastic fibers, whose synthesis is thought to be limited to times of development, are unable to repair appropriately following an inflammation-mediated PTB, cervical competency may be compromised long before the parturition process begins in subsequent pregnancies. Future experiments to determine the long-term impact of pathologic elastic fiber disruption and the ability of disrupted elastic fibers to functionally recover postpartum will be critical to test this hypothesis. Supplementary data Supplementary data are available at BIOLRE online. Supplementary Table 1. Antibody Information. Supplemental Figure S1. Timecourse of Tnf, Adamts4, Ctsl, and Mmp13 gene upregulation in response to intrauterine LPS. Gene expression analysis by qPCR demonstrates that Tnf transcription is significantly induced in the mouse cervix 2 h post-LPS compared to untreated day 15 (NT). Gene expression of proteases Adamts4 is significantly upregulated at 2 and 4 h time points, Ctsl by 6 h, and Mmp13 at 4 and 6 h post-LPS compared to untreated day 15 (NT) samples. (n = 4–6 samples per group, P < 0.05 compared to day 15, bars represent mean ± SEM. One-way ANOVA). Supplemental Figure S2. Immunofluorescence staining for tropoelastin. (A) Immunofluorescence imaging of tropoelastin protein (red) and nuclei by DAPI (blue) in cervical sections from day 15, day 15 sham, and day 15 LPS mice. Elastin is most concentrated in the subepithelial stroma region, directly below the superimposed white dotted line between the epithelium and stroma, and can be visualized in long fiber structures, presumed to be elastin integrated into elastic fibers. Signal intensity was optimized for day 15, and the same settings were used for day 15 sham and day 15 LPS. Supplemental Figure S3. Representative images of elastic fibers from cervices of day 15 mice treated with mifepristone. Transmission electron microscopy identified highly variable elastic fiber ultrastructure on day 15, 12 h after mifepristone treatment [10]. Representative images of the variable elastic fibers are shown here. Cervices from three different day 15 mifepristone-treated animals and 20–32 fibers per animal were imaged. No major changes in elastic fibers or collagen was observed among the groups. Scale bar: 1000 nm. Supplemental Table S1. Mapping data from RNA-seq libraries. Supplemental Table S2. Forward (F) and reverse (R) primer sequences. Supplemental Table S3. Fold change of canonical parturition genes measured by RNA-seq. Supplemental Table S4. Fold change of proinflammatory genes, measured by RNA-seq. Acknowledgments We would like to thank the UTSW Electron Microscopy Core for help with tissue processing and sample preparation for TEM and the UTSW Live Cell Imaging Core Facility for guidance in imaging. We thank Dr Ann Word for helpful discussions regarding experimental results. Footnotes † Grant Support: NIH R21 HD075228, NIH R01 HD084695, and NIH TL1TR001104. ‡ Accession Number: GSE98545 References 1. Behrman RE, Butler AS (eds.). Preterm Birth: Causes, Consequences, and Prevention . Washington (DC): The National Academies Press; 2007: 31– 52. 2. Martin JA, Hamilton BE, Osterman MJ, Driscoll AK, Mathews TJ. Births: final Data for 2015. Natl Vital Stat Rep  2017; 66: 1. Google Scholar PubMed  3. Martin JA, Hamilton BE, Osterman MJ, Curtin SC, Matthews TJ. Births: final data for 2013. Natl Vital Stat Rep  2015; 64: 1– 65. 4. Romero R, Espinoza J, Kusanovic JP, Gotsch F, Hassan S, Erez O, Chaiworapongsa T, Mazor M. The preterm parturition syndrome. BJOG  2006; 3( 113 Suppl): 17– 42. Google Scholar CrossRef Search ADS   5. Ferrero DM, Larson J, Jacobsson B, Di Renzo GC, Norman JE, Martin JN Jr., D’Alton M, Castelazo E, Howson CP, Sengpiel V, Bottai M, Mayo JA et al.   Cross-country individual participant analysis of 4.1 million singleton births in 5 countries with very high human development index confirms known associations but provides no biologic explanation for 2/3 of all preterm births. PLoS One  2016; 11: e0162506. Google Scholar CrossRef Search ADS PubMed  6. Gomez-Lopez N, Romero R, Xu Y, Plazyo O, Unkel R, Leng Y, Than NG, Chaiworapongsa T, Panaitescu B, Dong Z, Tarca AL, Abrahams VM et al.   A role for the inflammasome in spontaneous preterm labor with acute histologic chorioamnionitis. Reprod Sci  2017; 24: 1382– 1401. Google Scholar CrossRef Search ADS PubMed  7. Romero R, Miranda J, Chaiworapongsa T, Korzeniewski SJ, Chaemsaithong P, Gotsch F, Dong Z, Ahmed AI, Yoon BH, Hassan SS, Kim CJ, Yeo L. Prevalence and clinical significance of sterile intra-amniotic inflammation in patients with preterm labor and intact membranes. Am J Reprod Immunol  2014; 72: 458– 474. Google Scholar CrossRef Search ADS PubMed  8. Mahendroo M, Hoffman B, Cunningham G. Physiology of Labor. In: Cunningham FG, Leveno KJ, Bloom SL, Spong CY, Dashe JS, Hoffman BL, Casey BM, Shefield JS (eds.), Williams Obstetrics , 24 ed. New York: McGraw-Hill Education; 2014: 408– 432. 9. Elovitz MA, Wang Z, Chien EK, Rychlik DF, Phillippe M. A new model for inflammation-induced preterm birth: the role of platelet-activating factor and Toll-like receptor-4. Am J Pathol  2003; 163: 2103– 2111. Google Scholar CrossRef Search ADS PubMed  10. Holt R, Timmons BC, Akgul Y, Akins ML, Mahendroo M. The molecular mechanisms of cervical ripening differ between term and preterm birth. Endocrinology  2011; 152: 1036– 1046. Google Scholar CrossRef Search ADS PubMed  11. Gonzalez JM, Dong Z, Romero R, Girardi G. Cervical remodeling/ripening at term and preterm delivery: the same mechanism initiated by different mediators and different effector cells. PLoS One  2011; 6: e26877. Google Scholar CrossRef Search ADS PubMed  12. McCormack JT, Greenwald GS. Progesterone and oestradiol-17beta concentrations in the peripheral plasma during pregnancy in the mouse. J Endocrinol  1974; 62: 101– 107. Google Scholar CrossRef Search ADS PubMed  13. Timmons BC, Reese J, Socrate S, Ehinger N, Paria BC, Milne GL, Akins ML, Auchus RJ, McIntire D, House M, Mahendroo M. Prostaglandins are essential for cervical ripening in LPS-mediated preterm birth but not term or antiprogestin-driven preterm ripening. Endocrinology  2014; 155: 287– 298. Google Scholar CrossRef Search ADS PubMed  14. Nallasamy S, Akins M, Tetreault B, Luby-Phelps K, Mahendroo M. Distinct reorganization of collagen architecture in lipopolysaccharide - mediated premature cervical remodeling. Biol Reprod  2017. 15. Zhong S, Joung JG, Zheng Y, Chen YR, Liu B, Shao Y, Xiang JZ, Fei Z, Giovannoni JJ. High-throughput illumina strand-specific RNA sequencing library preparation. Cold Spring Harb Protoc  2011; 2011: 940– 949. Google Scholar CrossRef Search ADS PubMed  16. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol  2013; 14: R36. Google Scholar CrossRef Search ADS PubMed  17. Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, van Baren MJ, Salzberg SL, Wold BJ, Pachter L. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol  2010; 28: 511– 515. Google Scholar CrossRef Search ADS PubMed  18. Dennis G Jr., Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol  2003; 4: P3. Google Scholar CrossRef Search ADS PubMed  19. Nallasamy S, Yoshida K, Akins M, Myers K, Iozzo R, Mahendroo M. Steroid hormones are key modulators of tissue mechanical function via regulation of collagen and elastic fibers. Endocrinology  2017; 158: 950– 962. Google Scholar CrossRef Search ADS PubMed  20. Colige A, Vandenberghe I, Thiry M, Lambert CA, Van Beeumen J, Li SW, Prockop DJ, Lapiere CM, Nusgens BV. Cloning and characterization of ADAMTS-14, a novel ADAMTS displaying high homology with ADAMTS-2 and ADAMTS-3. J Biol Chem  2002; 277: 5756– 5766. Google Scholar CrossRef Search ADS PubMed  21. Rajabi MR, Solomon S, Poole AR. Biochemical evidence of collagenase-mediated collagenolysis as a mechanism of cervical dilatation at parturition in the guinea pig. Biol Reprod  1991; 45: 764– 772. Google Scholar CrossRef Search ADS PubMed  22. Asquith DL, Miller AM, Reilly J, Kerr S, Welsh P, Sattar N, McInnes IB. Simultaneous activation of the liver X receptors (LXRalpha and LXRbeta) drives murine collagen-induced arthritis disease pathology. Ann Rheum Dis  2011; 70: 2225– 2228. Google Scholar CrossRef Search ADS PubMed  23. Sandy JD, Westling J, Kenagy RD, Iruela-Arispe ML, Verscharen C, Rodriguez-Mazaneque JC, Zimmermann DR, Lemire JM, Fischer JW, Wight TN, Clowes AW. Versican V1 proteolysis in human aorta in vivo occurs at the Glu 441 -Ala 442 bond, a site that is cleaved by recombinant ADAMTS-1 and ADAMTS-4. J Biol Chem  2001; 276: 13372– 13378. Google Scholar CrossRef Search ADS PubMed  24. Romero R, Xu Y, Plazyo O, Chaemsaithong P, Chaiworapongsa T, Unkel R, Than NG, Chiang PJ, Dong Z, Xu Z, Tarca AL, Abrahams VM et al.   A role for the inflammasome in spontaneous labor at term. Am J Reprod Immunol  2016. 25. Mulla MJ, Salmon JE, Chamley LW, Brosens JJ, Boeras CM, Kavathas PB, Abrahams VM. A role for uric acid and the Nalp3 inflammasome in antiphospholipid antibody-induced IL-1beta production by human first trimester trophoblast. PLoS One  2013; 8: e65237. Google Scholar CrossRef Search ADS PubMed  26. Hoang M, Potter JA, Gysler SM, Han CS, Guller S, Norwitz ER, Abrahams VM. Human fetal membranes generate distinct cytokine profiles in response to bacterial toll-like receptor and Nod-like receptor agonists. Biol Reprod  2014; 90: 39. Google Scholar CrossRef Search ADS PubMed  27. Dinarello CA. Interleukin-1 in the pathogenesis and treatment of inflammatory diseases. Blood  2011; 117: 3720– 3732. Google Scholar CrossRef Search ADS PubMed  28. Shi J, Zhao Y, Wang Y, Gao W, Ding J, Li P, Hu L, Shao F. Inflammatory caspases are innate immune receptors for intracellular LPS. Nature  2014; 514: 187– 192. Google Scholar CrossRef Search ADS PubMed  29. Shenderov K, Riteau N, Yip R, Mayer-Barber KD, Oland S, Hieny S, Fitzgerald P, Oberst A, Dillon CP, Green DR, Cerundolo V, Sher A. Cutting edge: endoplasmic reticulum stress licenses macrophages to produce mature IL-1 in response to TLR4 stimulation through a Caspase-8- and TRIF-dependent pathway. J Immunol  2014; 192: 2029– 2033. Google Scholar CrossRef Search ADS PubMed  30. Hazuda DJ, Strickler J, Kueppers F, Simon PL, Young PR. Processing of precursor interleukin 1 beta and inflammatory disease. J Biol Chem  1990; 265: 6318– 6322. Google Scholar PubMed  31. Herzog C, Haun RS, Kaushal V, Mayeux PR, Shah SV, Kaushal GP. Meprin A and meprin alpha generate biologically functional IL-1beta from pro-IL-1beta. Biochem Biophys Res Commun  2009; 379: 904– 908. Google Scholar CrossRef Search ADS PubMed  32. Mengshol JA, Vincenti MP, Coon CI, Barchowsky A, Brinckerhoff CE. Interleukin-1 induction of collagenase 3 (matrix metalloproteinase 13) gene expression in chondrocytes requires p38, c-jun N-terminal kinase, and nuclear factor kappaB: Differential regulation of collagenase 1 and collagenase 3. Arthritis Rheum  2000; 43: 801– 811. Google Scholar CrossRef Search ADS PubMed  33. Hashimoto Y, Kondo C, Katunuma N. An active 32-kDa Cathepsin L is secreted directly from HT 1080 fibrosarcoma cells and not via lysosomal exocytosis. PLoS One  2015; 10: e0145067. Google Scholar CrossRef Search ADS PubMed  34. Kirschke H, Kembhavi AA, Bohley P, Barrett AJ. Action of rat liver cathepsin L on collagen and other substrates. Biochem J  1982; 201: 367– 372. Google Scholar CrossRef Search ADS PubMed  35. Mason RW, Johnson DA, Barrett AJ, Chapman HA. Elastinolytic activity of human cathepsin L. Biochem J  1986; 233: 925– 927. Google Scholar CrossRef Search ADS PubMed  36. Kelwick R, Desanlis I, Wheeler GN, Edwards DR. The ADAMTS (A disintegrin and metalloproteinase with thrombospondin motifs) family. Genome Biol  2015; 16: 113. Google Scholar CrossRef Search ADS PubMed  37. Nallasamy S, Yoshida K, Akins M, Myers K, Iozzo R, Mahendroo M. Steroid hormones are key modulators of tissue mechanical function via regulation of collagen and elastic fibers. Endocrinology  2017; 158: 950– 962. Google Scholar CrossRef Search ADS PubMed  38. Liggins M. Cervical Ripening as an inflammatory reaction. In: Ellwood DAA, Anne B (ed.), The Cervix in Pregnancy and Labour: Clinical and Biochemical Investigations . Churchill Livingstone: The University of Michigan; 1981: 950– 962. 39. Goldenberg RL, Hauth JC, Andrews WW. Intrauterine infection and preterm delivery. N Engl J Med  2000; 342: 1500– 1507. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2017. Published by Oxford University Press on behalf of Society for the Study of Reproduction. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Biology of ReproductionOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off