If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Departments of Microbiology, Infectiology and Immunology, Centre Hospitalier Universitaire (CHU) Sainte-Justine Research Center, Université de Montréal, Montreal, Canada
Departments of Obstetrics and Gynecology, Mayo Clinic, Rochester, MNImmunology, Mayo Clinic, Rochester, MNDepartment of Obstetrics and Gynecology, Université de Montréal, Montreal, Canada
Preterm birth remains a leading obstetrical complication because of the incomplete understanding of its multifaceted etiology. It is known that immune alterations toward a proinflammatory profile are observed in women with preterm birth, but therapeutic interventions are still lacking because of scarcity of evidence in the integration of maternal and placental interrelated compartments.
Objective
This study aimed to obtain an integrated view of the maternal and placental contribution to preterm birth compared with normal term pregnancies for an in-depth understanding of the immune/inflammatory involvement, intending to identify novel strategies to mitigate the negative impact of inflammation.
Study Design
We prospectively recruited 79 women with preterm or term deliveries and collected placentas for RNA sequencing, histologic analyses, and to assess levels of inflammatory mediators. Blood samples were also collected to determine the circulating immune profiles by flow cytometry and to evaluate the circulating levels of inflammatory mediators.
Results
Placental transcriptomic analyses revealed 102 differentially expressed genes upregulated in preterm birth, including known and novel targets, which were highly enriched for inflammatory biological processes according to gene ontology analyses. Analysis of maternal immune cells revealed distinct profiles in preterm births vs term births, including an increased percentage of CD3− cells and monocyte subsets and decreased CD3+ cells along with Th17 subsets of CD4+ lymphocytes. Supporting our bioinformatic findings, we found increases in proinflammatory mediators in the plasma, placenta, and fetal membranes (primarily the amnion) of women with preterm birth, such as interleukin-6 and tumor necrosis factor-α. These findings were not distinct between spontaneous and iatrogenic preterm births except at a molecular level where spontaneous preterm birth presented with an elevated inflammatory profile compared with iatrogenic preterm birth. Analysis of placental histology revealed increased structural and inflammatory lesions in preterm vs term births. We found that genes upregulated in placentas with inflammatory lesions have enrichment of proinflammatory pathways.
Conclusion
This work sheds light on changes within the immune system in preterm birth on multiple levels and compartments to help identify pregnancies at high risk of preterm birth and to discover novel therapeutic targets for preterm birth.
Preterm birth (PTB) is a leading obstetrical complication, responsible for nearly 16% of all neonatal deaths worldwide, and is a major contributor to neonatal morbidity.
Global, regional, and national causes of under-5 mortality in 2000-15: an updated systematic analysis with implications for the sustainable development goals.
Our primary goal was to obtain an integrated view of the maternal and placental contribution to preterm birth to understand the compartment-specific contribution to the etiology of the syndrome.
Key findings
Placental transcriptomic analyses revealed differentially expressed genes that were highly enriched for inflammatory biological processes, which were supported through analyses of maternal immune cells, inflammatory mediators in maternal circulation and at the maternal–fetal interface, and histologic lesions in the maternal–fetal interface.
What does this add to what is known?
Distinct profiles in preterm vs term births shed light on changes within the immune system at multiple levels and compartments to help support previously used therapeutics and offer potential new targets to consider in clinical settings.
Inflammation is essential throughout pregnancy—from early fetal development to labor onset.
Inflammatory mechanisms are tightly regulated by local/systemic immune cells and mediators in uncomplicated pregnancies. However, evidence shows impaired inflammatory responses in gestational tissues and maternal circulation in pregnancy complications—notably PTB.
PTB can also occur without detectable microorganisms (ie, sterile inflammation), whereby endogenous danger signals derived from cellular stress or necrosis, known as damage-associated molecular patterns or alarmins, are often detected.
Cytokine profiling: variation in immune modulation with preterm birth vs. uncomplicated term birth identifies pivotal signals in pathogenesis of preterm birth.
Recent work has identified transcriptome changes in maternal peripheral blood, which could provide noninvasive ways for identifying women at risk of PTB.
These studies share the involvement of immune pathways in PTB; however, to understand the immune/inflammatory mechanisms involved, it is crucial to investigate the response of different maternal and fetal components to allow in-depth integration of PTB pathology.
Our study aimed to obtain an integrated view of the maternal and placental contribution to the PTB syndrome compared with normal pregnancies to allow an in-depth understanding of the immune/inflammatory contributions and to identify women who could benefit from targeted antiinflammatory therapies. We studied the placental transcriptome, identified maternal circulating immune cell and inflammatory profiles, investigated inflammatory mediators, and performed in-depth histologic analyses at the maternal–fetal interface. Through the integration of multiple compartments, we obtained a better understanding of the pathways underlying PTB, and importantly their specific localization, which could be leveraged for new therapeutic interventions and biomarkers for future studies.
Materials and Methods
Ethics approval and patient data collection
Approval was obtained from the Centre Hospitalier Universitaire (CHU) Sainte-Justine (CHUSJ) Ethics Board (No: 2015-840). Women recruited (n=79) from the CHU Sainte-Justine Hospital had either an uncomplicated term birth (n=41, 16 laboring) or PTB (n=38, 22 laboring) and gave written informed consent at the time of their admission for delivery. Birth before 37 weeks of completed gestation was considered preterm. Preeclampsia, active infection (defined as clinical symptoms of infection and/or antibiotic treatment within the last 7 days), and congenital malformations were excluded. Demographic and obstetrical data were obtained through the evaluation of patients’ clinical files.
Tissue and blood collection
Placental biopsies (1 sample per quadrant, defined from umbilical cord), taken in the thickest portion of the placental villi, and fetal membrane biopsies were collected shortly after delivery. Placental samples were kept in RNA for 24 hours before being stored at −80°C or directly frozen and stored at −80°C. Placenta, fetal membrane, and umbilical cord biopsies were fixed in 4% paraformaldehyde (PFA) for 7 days, transferred to phosphate-buffered saline (PBS), and embedded into paraffin for hematoxylin and eosin (HE) staining. Blood samples were obtained from each patient within 24 hours before delivery, notably between 6 and 12 hours of delivery, with some collected between 3 hours (particularly patients that had cesarean deliveries) or between 12 and 24 hours. If a patient did not deliver within 24 hours of blood collection, an additional sample was collected, and the latter used for analysis. Fresh blood was immunolabeled with antibodies for immune cell subpopulations of interest (details below). The remaining blood was centrifuged at 2000 rpm (10 minutes, 4°C), and plasma was collected, aliquoted, and stored at −80°C.
Immune cell analysis by flow cytometry
Whole blood (100 μl) was immunolabeled for 15 minutes at room temperature using the following antibodies: FITC-CD3, APC-Cy7-CD4, PerCP-Cy5.5-CD8, PE-CD127, V500-CD19, PE-CF594-CD183, PE-Cy7-CD196, BV605-CD25, BV421-CD194, Alexa Fluor 700-CD14, and APC-CD56 (all from BD Biosciences, Mississauga, Canada). Red blood cells were lysed (0.06 M NH4Cl, 4 mM KHCO3, 0.052 mM EDTA), and peripheral blood mononuclear cells washed with PBS, PFA-fixed, and analyzed using a flow cytometer (LSRFortessa, BD Biosciences, Ontario, Canada). A minimum of 300,000 events were acquired per sample. Data analysis was performed using FlowJo (Tree Star Inc, Ashland, OR) (gating strategy is shown in Supplemental Figure 1).
Inflammatory mediator detection in the maternal plasma and at the maternal–fetal interface
A panel of cytokines/mediators (ie, interleukin [IL]-1β, IL-6, tumor necrosis factor [TNF]-α, monocyte chemoattractant protein-1 [MCP-1], S100A8, C-reactive protein [CRP], and IL-1 receptor antagonist [IL-1RA]) were analyzed in placental tissues and plasma via enzyme-linked immunosorbent assay (ELISA). In addition, ELISA analyses were performed for interferon-γ and IL-4 in the placenta only, whereas IL-8 and progesterone were analyzed in plasma only (all from R&D Systems, Minneapolis, MN; except for progesterone: Cayman Chemical Company, Ann Arbor, MI), as previously described.
Bulk RNA sequencing (RNA-seq) of placental tissues was run in 3 batches (adjusted for). Messenger RNA was extracted using the RNeasy Mini Kit (QIAGEN, Toronto, Canada) following manufacturer instructions and quantified (NanoDrop 8000 Spectrophotometer, Thermo Fisher Scientific, Missisauga, Canada), and the quality was assessed using the RNA integrity number (RIN ≥7; 2100 Bioanalyzer, Agilent Technologies, Santa Clara, CA). Libraries were created using the TruSeq RNA Library Prep Kit v2 (Illumina, San Diego, CA), and fragment sizes were validated on the 2100 Bioanalyzer (Agilent). Libraries were sequenced using the HiSeq 4000 System (Illumina) at the CHUSJ. RNA-seq data were analyzed in accordance with the Genome Analysis Toolkit (GATK) best practices. Alignment to the hg19 (GRCh37) genome reference was performed using the STAR aligner, and gene expression was measured with the HTSeq software using the Ensembl version 75 gene coordinates. Bioinformatic analyses were performed with RStudio, version 1.1.463 (RStudio, PBC, Boston, MA).
Data were adjusted for batch effects and sex. Differential gene expression (adjusted P value <.05, log fold change [logFC]±1) analyses were achieved with the limma R package.
Gene set enrichment analysis (GSEA) and pathway analysis using Kyoto Encyclopedia of Genes and Genomes (KEGG) was performed in RStudio using clusterProfiler.
Five micrometer thick sections from the placenta (at least 3 samples) and fetal membranes were processed for HE staining and evaluated by a histopathologist blinded to pregnancy outcomes assessed according to the Amsterdam criteria, as previously described.
Images were captured using a slide scanner (Axioscan, ZEISS, Ontario, Canada).
Statistical analysis
Data are presented as mean±standard error of the mean. Data were analyzed using 1-way analysis of variance with Dunnett’s multiple comparisons, Sidak’s multiple comparisons, or Fisher exact test, as appropriate, and unpaired t tests and simple linear regressions using GraphPad Prism, version 9.1.2 (GraphPad Software, San Diego, CA). A probability value of <.05 was considered statistically significant.
Results
Population characteristics: demographic and obstetrical information
Seventy-nine women were included in this study (41 term, 38 preterm). Demographic and obstetrical characteristics are shown in Table 1. Maternal age differed between groups where term births were older (31.0±1.0 vs 33.7±0.67 years; P<.05). Both groups included predominantly White women with similar prepregnancy body mass indexes and family or personal history of hypertension and diabetes mellitus. PTBs were more likely to occur among primiparous women (65.5%±2.0% vs 35.5%±9.5%; P<.05), have lower gestational age at delivery (32.4±0.5 vs 39.8±0.1 weeks; P<.01), and lower birthweight (1666.5±88.0 g vs 3534.0±53.0 g in term births; P<.01), with 42% of small-for-gestational-age (SGA) neonates vs none in term births. Both groups had an equal proportion of labor inductions, cesarean delivery, and male/female newborns.
P<.0001. Statistical analysis by unpaired t test or Fisher exact test where appropriate.
HT current pregnancy (%)
0 (0)
2 (5.4)
DM current pregnancy (%)
4 (9.8)
8 (22)
Induction of labor (%)
13 (32)
14 (37)
Delivery by CD (%)
25 (60)
20 (52)
Sex (% of male)
21 (51)
20 (52)
Data presented as mean (range) or number (percentage).
BMI, body mass index; CD, cesarean delivery; DM, diabetes mellitus; GA, gestational age; HT, hypertension; SGA, small for gestational age (<10th percentile).
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
a P<.05
b P<.01
c P<.0001. Statistical analysis by unpaired t test or Fisher exact test where appropriate.
Bulk RNA-sequencing of the placenta revealed novel genes involved in inflammatory pathways associated with preterm birth
To deepen our understanding of the transcriptional profiles, we performed RNA-seq on the placentas of women with term and PTBs. Unbiased analyses included 39,193 genes, from which we extracted 331 differentially expressed genes (DEGs) (adjusted P<.05; logFC±1) between term and PTB conditions. Of these DEGs, 102 were upregulated and 229 were downregulated in PTB (Figure 1). Within the upregulated genes, many are highly involved in immune regulation (Table 2). An example of a downregulated gene is gamma-aminobutyric acid type A receptor subunit beta1 (GABRB1; logFC=−2.31; adjusted P=11.28×10−9), also observed to be negatively modulated in PTB by Paquette et al, supporting our findings.
A, We identified 331 DEGs between term and PTB, of which 229 were downregulated (blue) and 102 upregulated (red). B, The top 50 most differentially expressed genes (sorted by logFC) are shown as a row-scaled heatmap, where upregulated genes are in red and downregulated genes are in blue. Examples of C, highly upregulated and D, downregulated genes extracted from the heatmap identify potential future targets.
Triple asterisks denote P<.001 by moderated t-statistics in R. Cutoff criteria of P<.05 and logFC±1 were used for DEG analysis.
Table 2Top differentially expressed genes between term and preterm deliveries
Gene name
Description
LogFC
AveExpr
Adjusted P value
RPL3P4
Ribosomal protein L3 pseudogene 4
3.43
−0.87
1.28×10−2
ENSG00000267954
Not previously described
2.76
1.45
2.09×10−3
ADAM9
ADAM metallopeptidase domain 9
2.75
1.35
6.95×10−3
PPP1R18
Protein phosphatase 1 regulatory subunit 18
2.44
0.65
1.03×10−3
KCNQ1
Potassium voltage-gated channel subfamily Q member 1
2.24
−2.79
2.24×10−2
HLA-E
Major histocompatibility complex class 1, E
1.99
2.94
1.66×10−2
CPNE7
Copine 7
1.91
−3.04
2.48×10−3
PARP4
Poly (ADP-ribose) polymerase family member 4
1.89
−3.32
7.40×10−3
DPEP1
Dipeptidase 1
1.89
−2.65
6.10×10−5
CCL20
C-C motif chemokine ligand 20
1.89
−3.00
5.82×10−3
ENSG00000263620
Not previously described
−3.39
0.79
5.50×10−5
BZW1P2
Basic leucine zipper and W2 domains 1 pseudogene 2
−3.00
1.06
1.33×10−4
TCAF2C
TRPM8 channel associated factor 2C
−2.99
−1.66
7.70×10−3
OPRK1
Opioid receptor kappa 1
−2.78
−0.77
6.30×10−3
ENSG00000270149
Not previously described
−2.70
2.06
1.97×10−4
H3P6
H3 histone pseudogene 6
−2.64
1.91
2.57×10−3
FHDC1
FH2 domain containing 1
−2.38
5.15
5.97×10−4
MUC20P1
Mucin 20, cell surface-associated pseudogene 1
−2.35
−1.48
1.31×10−3
GABRB1
Gamma-aminobutyric acid type A receptor subunit beta1
−2.31
−1.02
1.28×10−5
RSC1A1
Regulator of solute carriers 1
−2.28
−2.03
4.41×10−3
Top 10 most upregulated differentially expressed genes (DEGs) in PTB (above) and top 10 most downregulated DEGs in PTB (below) sorted by logFC (P<.05). Gene name is represented by Ensembl Gene ID when the gene has not been previously defined. LogFC represents log2fold changes between preterm and term births, and AveExpr represents average log2fold change within sequencing.
ADP, adenosine diphosphate; AveExpr, average of the log-expression values; FH2, formin homology 2; LogFC, log fold change; TRPM8, transient receptor potential cation channel subfamily M (melastatin) member 8.
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
To explore biological specializations and decipher enrichments, we used gene ontology (GO) (false discovery rate [FDR] P<.05) and found that upregulated genes had processes related to inflammatory pathways, whereas downregulated genes had no statistically enriched biological processes (Figure 2, A). We further removed redundancy using Cytoscape and found that the main pathways affected in PTB are processes related to lymphocyte chemotaxis, neutrophil migration, and antibacterial humoral response (Supplemental Figure 2). To strengthen our findings, we investigated gene set enrichment (GSEA) and pathway enrichment (KEGG), and both supported the enrichment of immune and inflammatory processes in PTB (Figure 2, B–D). We identified dysregulated genes within these enriched pathways (eg, upregulated 4-1BBL in the cytokine–cytokine receptor interactions identified through KEGG analysis) (Supplemental Figure 3). Following this work, which highlighted immune-related pathways in PTB, we further investigated which components of the immune system might be involved.
A, Biological processes using upregulated DEGs obtained from gene ontology (GO) analysis (P<.05; logFC±1) showed an overrepresentation of immune-related and inflammatory pathways in PTB (FDR<0.05). B, Gene set enrichment analysis (GSEA; P<.05; minimum gene set=3, maximum gene set=800; 10,000 permutations) of all DEGs supported the GO analysis with regard to activation of immune and inflammatory pathways and suppression of housekeeping processes. C, Pathway enrichment analysis (KEGG; P<.05; minimum gene set=3, maximum gene set=800; 10,000 permutations) additionally found enrichment for immune pathways, where D, an example of the enrichment of genes within the cytokine–cytokine receptor interaction pathways can be observed.
DEG, differentially expressed gene; FDR, false discovery rate; IL, interleukin; KEGG, Kyoto Encyclopedia of Genes and Genomes; logFC, log fold change; PTB, preterm birth; TNF, tumor necrosis factor.
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
The maternal circulating immune cell profile is altered in preterm as compared with term deliveries
Of the circulating immune cell subpopulations studied, significant changes were observed for both CD3+ and CD3− cells in our patient groups (Figure 3). CD3− cells were increased in PTB (52.9%±13.0% vs 40.9%±14.0%; P<.001) (Figure 3, A), particularly the monocyte subtype (46.2%±12.9% vs 34.3%±18.9%; P<.01) (Figure 3, B and E). Other CD3− subsets such as B-lymphocytes and natural killer (NK) cells remained unchanged (data not shown). Conversely, CD3+ cells were decreased in PTB (46.8%±13.0% vs 58.7%±14.0%; P<.001) (Figure 3, C), particularly the Th17 subset (4.1%±2.1% vs 5.6%±2.8%; P<.05) (Figure 3, D). Other CD3+ subsets such as CD8+, CD4+, Treg, NK-T, and T-helper cells remained unchanged (data not shown).
Figure 3Immune cells found in maternal blood shortly antepartum using FACS
A, An increase in CD3− cells in PTB can be seen as a percentage of live cells in the bar graph and through the shifts of CD3 expression in representative participants in the histogram. Monocytes were also increased in PTB as a percentage of CD3− cells and through the shift in CD14 expression seen in B, the histogram and E, the bivariate contour overlay. C, A decrease in CD3+ cells in PTB can be seen as a percentage of live cells in the bar graph and through the shifts of CD3 expression in representative participants seen in the histogram. Th17 cells were also decreased in PTB as a percentage of CD4+ lymphocytes and through the D, shift in CD196 expression and J, the bivariate contour overlay. Results are presented as mean±standard error of the mean.
Single asterisk denotes P<.05; double asterisks denote P<.01; triple asterisks denote P<.001 by unpaired t tests used for statistical analysis.
Proinflammatory mediators are elevated within the maternal circulation and at the maternal–fetal interface
A detailed analysis of the immune mediators in the maternal circulation revealed important differences in PTB. None of the classical cytokines studied showed significant changes (namely IL-1β, IL-6, TNF-α, IL-10, MCP-1, S100A8, IL-8, and IL-1Ra; data not shown). However, when looking at other inflammation-related mediators, we found that progesterone was decreased (89±10 vs 297±32 ng/mL; P<.0001; none of the patients were treated therapeutically with progesterone) (Figure 4, A), whereas CRP was increased (7.2±1.2 vs 2.4±5.0 ug/mL; P<.001) (Figure 4, B) in the maternal circulation in PTB.
Figure 4Maternal circulation and maternal–fetal interface immune mediator levels
Women with PTB showed (A) a decrease in progesterone and (B) an increase in CRP in maternal circulation before delivery. Within the placenta, PTB had significantly increased levels of (C) IL-6 and (D) TNF-α. Fetal membranes were analyzed separately (amnion and chorion) and there was (E) an increase in MCP-1 in both compartments. Only the amnion showed increases in (F) CRP and (G) S100A8 and a decrease in (H) IL-6. Data presented as mean±standard error of the mean.
Single asterisk denotes P<.05; double asterisks denote P<.01; triple asterisks denote P<.001; four asterisks denote P<.0001. Unpaired t tests were used for statistical analysis.
Analysis of the immune mediators at the maternal–fetal interface revealed increased levels of proinflammatory IL-6 (1.50±0.20 vs 0.90±0.08 pg/mg; P<.01) (Figure 4, C) and TNF-α (2.50±0.38 vs 1.60±0.20 pg/mg; P<.05) (Figure 4, D) in PTB, which was correlated to gestational age for IL-6 only (Supplemental Figure 4). Within fetal membranes, we found that both amnion and chorion showed elevated levels of MCP-1 in PTB (155.2±34.0 vs 43.5±8.9 pg/mg; 198.5±35.6 vs 71.0±9.7 pg/mg; P<.05, respectively) (Figure 4, E). In addition, specific mediators were found altered only in the fetal-facing amnion such as increased CRP (59.0±10.7 vs 24.0±4.4 ng/mg; P<.05) (Figure 4, F) and S100A8 (1.50±0.33 vs 0.20±0.04 ng/mg; P<.05) (Figure 4, G), alongside a decrease in IL-6 (124.1±41.7 vs 427.7±113.4 pg/mg; P<.01) (Figure 4, H) in PTB.
Histologic analysis revealed an increase in placental lesions in preterm births
Placentas from PTBs had elevated structural lesions compared with term deliveries (60.5% vs 14.6%; P<.0001). These included maternal/fetal vascular malperfusion, accelerated villous maturation, increased syncytial knots, and fibrin (Figure 5; Supplemental Table). In addition, there were more inflammatory lesions in PTB placentas (31.6% vs 12.2%; P=.05), such as deciduitis (Figure 5, B). Conversely, term placentas were largely found to be lesion-free (7.9% vs 73.0%; P<.0001) (Figure 5, B).
Figure 5Histologic analysis of placentas/fetal membranes from term and preterm deliveries
A, Representative images of normal and lesioned placenta/fetal membranes including (1) normal villi and (2) normal fetal membrane from term placenta. The most frequent inflammatory lesions included (3) maternal inflammatory response and (4) deciduitis, and the most frequent structural defects included (5) excess syncytial knots and (6) excess fibrin/avascular villi. Scale bar: 100 um. B, The percentages of structural and/or inflammatory lesions in term and preterm patients demonstrate that preterm deliveries had more structural and inflammatory lesions and fewer lesion-free placentas compared with term deliveries.
PTB, preterm birth.
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
Considering that PTB is a syndrome with varied etiologies, we first addressed if the differences we observed were detected between spontaneous and iatrogenic deliveries. Within our preterm subset (n=38), we had 18 spontaneous (sPTB) and 20 iatrogenic deliveries (iPTB; 7 were induced but only 3 delivered vaginally). No difference was observed in the maternal immune cells or in plasma/placenta inflammatory mediators between the subgroups of PTB (data not shown). However, chorionic membranes showed an elevation in CRP and TNF-α in sPTB (Figure 6, A and B). In addition, from our transcriptomic data (above), we observed that clustering was equally distributed (Figure 6, C), but there were differences between the genes from sPTB and iPTB placentas; 310 DEGs were found between sPTB and iPTB, of which 125 were upregulated and 185 were downregulated in sPTB (Figure 6, D). Importantly, only the upregulated DEGs in sPTB showed overrepresentation for inflammatory pathways (Figure 6, E).
Figure 6Gene expression based on spontaneous vs iatrogenic preterm birth
Boxplots show an increase in (A) proinflammatory TNF-α (7.3±1.7 vs 2.6±0.8; P<.05) and (B) CRP (79.4±19.28 vs 25.2±9.6) in sPTB vs iPTB. (C) Multidimensional scaling plot shows clustering within PTB patients (sPTB [n=18, red] vs iPTB [n= 20, blue; out of which 7 were induced but only 3 had vaginal deliveries]). (D) Row-scaled heatmap of the top 50 DEGs (P<.05) in order of logFC between sPTB and iPTB shows the pattern of gene expression. (E) Gene ontology pathways from all the upregulated DEGs (red) and the downregulated DEGs (blue) between sPTB and iPTB indicate that immune and inflammatory pathways are overrepresented in sPTB.
In addition, we investigated if differences in gene expression were associated with the presence of lesions within the placenta. We looked at gene expression changes between patients in our entire cohort (Figure 7, A), specifically within the PTB or term group (Figure 7, B and C). We found that placentas with inflammatory lesions had significant DEGs, and these showed an overrepresentation in immune pathways (Figure 7, D). On the other hand, placentas with structural lesions showed no DEGs compared with placentas without lesions.
Figure 7Gene expression based on placental lesions
Placentas were classified as having either inflammatory (In, red), structural (Str, green), or no (No, blue) lesions. Multidimensional scaling plots show that placental lesions are not clustering individually. Plots show the clustering of placental lesions within (A) all patients (n=79), (B) those with preterm births (n=38), and (C) those with term birth (n=41). Only the differentially expressed genes found between placentas with inflammatory lesions and placentas without lesions within all the patients (group A) showed upregulated genes, which had an overrepresentation of inflammatory pathways (red), and (D) downregulated genes had an enrichment of biological pathways (blue).
dim, dimension; logFC, log fold change.
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
We investigated the transcriptional changes in the placenta concerning preterm vs term births, alongside the immune/inflammatory profiling within the maternal circulation and the maternal–fetal interface. We found 102 upregulated DEGs in PTB, which were enriched for inflammatory pathways according to GO, GSEA, and KEGG analyses. In the maternal circulation, we observed increased CD3− cells and monocytes and decreased CD3+ and Th17 cells, both involved in the proinflammatory response. Furthermore, CRP was elevated, whereas progesterone was decreased. Similarly, a proinflammatory bias was observed at the maternal–fetal interface, with elevated levels of IL-6 and TNF-α in the placenta along with MCP-1, CRP, and S100A8 increased predominantly in the fetal-facing amnion. Interestingly, strong differences were observed in the transcriptional profiles and proteins between sPTB and iPTB, suggesting that changes are associated with the etiology of PTB.
Results in the context of what is known
Within the 102 upregulated DEGs in PTB, inflammatory pathways were the most significantly enriched using GO, GSEA, and KEGG, which is in line with other reports.
strengthening its potential involvement. Many other genes have important roles in inflammatory processes, such as ADAM metallopeptidase domain 9 (ADAM9) and C-C motif chemokine ligand 20 (CCL20), which are involved in the recruitment of Th17 and Tregs to sites of inflammation.
An immune paradox: how can the same chemokine axis regulate both immune tolerance and activation?: CCR6/CCL20: a chemokine axis balancing immunological tolerance and inflammation in autoimmune disease.
In our study, 11 genes related to neutrophils, which help establish proper local inflammation, were upregulated in the placentas from PTBs. These included dipeptidase 1 (DPEP1), serum amyloid A1 and 2 (SAA1/2), platelet factor 4 (PF4), proplatelet basic protein (PPBP), proteinase 3 (PRTN3), and C-C chemokines,
Morphologic and histochemical evidence for the occurrence of collagenolysis and for the role of neutrophilic polymorphonuclear leukocytes during cervical dilation.
the latter which are also found upregulated in the enriched proinflammatory pathways (Table 2), but the actual number and function of neutrophils in PTB remains to be further elucidated. Potential markers were also identified at the protein level, such as changes in IL-6 and TNF-α, alongside increased CRP, all suggested to be associated with increased risks of neonatal complications.
Cytokine profiling: variation in immune modulation with preterm birth vs. uncomplicated term birth identifies pivotal signals in pathogenesis of preterm birth.
In addition, we observed decreased progesterone, which normally increases during pregnancy and is strongly antiinflammatory, suggesting possible involvement in premature labor.
In terms of immune cell profiles in the maternal circulation, we found decreased CD3+ cells and their Th17 subset, which contrasts with growing evidence that some pregnancy complications are linked to increasing Th17 cells in the periphery.
Monocytes are a major source of cytokines in the inflammatory phase of early pregnancy; however, they are known to change throughout pregnancy, consistent with “activation.”
Although circulating monocytes were increased in PTB, future work is needed to decipher their activation status in the placenta.
Clinical implications
Our study aimed to find new therapeutic targets for clinical use and can also be used to support currently used therapeutics. For example, our observed decrease in progesterone supports its use in high-risk women.
It will be interesting to further study the components of identified dysregulated signaling pathways to identify novel therapeutic targets (eg, 4-1BBL is enriched in PTB and has been studied as an anticancer target) (Supplemental Figure 3).
Other genes such as DPEP1 and transmembrane and immunoglobulin domain containing 3 (TMIGD3), involved in neutrophil recruitment and NFκB inhibition, respectively, are to the best of our knowledge newly associated with PTB and will be interesting to investigate in future work.
Although a diagnosis of high-risk pregnancy has to occur before delivery, an in-depth analysis of the placenta could help to identify subgroups with and without inflammation to adapt therapeutic strategies and mitigate the impact of inflammation postnatally on infant development.
Research implications
Our study has generated a large transcriptional repertoire pertaining to the changes occurring at the maternal–fetal interface in both term and preterm labor. We identified genes and pathways dysregulated in PTB, and these specific signatures could be used to identify women who would benefit from antiinflammatory therapeutic interventions alongside continuing to integrate findings from multiple interrelated compartments to elucidate their involvement in preterm labor.
Strengths and limitations
Our study had equivalent population groups, and we combined many techniques and several compartments, which strengthened our findings. By design, term pregnancies were used as controls because there is no ethical way to obtain gestational age–matched tissues from healthy pregnancies. However, we previously reported that the inflammatory changes associated with term physiological labor were minimal when compared with those occurring in pathologic pregnancies such as those with PTB.
Given that the placenta grows throughout gestation, it might cause changes in gene expression levels; however, we did not find correlations with gestational age (unless mentioned), which is supported by other studies, showing that PTB cannot be divided into gestational age clusters.
To our surprise, we did not observe alterations in classical cytokines, other than IL-6 and TNF-α, in PTB vs term pregnancies, which could be because we excluded women with active infections and therefore might have detected primarily sterile inflammatory processes rather than pathogen-associated inflammation. Another limitation to our study is the fact that some patients were in labor at the time of blood collection, both in the PTB and term group. However, no differences in circulating cytokines were observed in the subgroups of labor and no-labor (data not shown). Furthermore, our PTB group had an elevated proportion of SGA neonates, most of which had iPTB, which could have contributed to the differences observed.
Conclusions
Our work sheds light on changes in placental transcriptomes, immune cell populations, immune mediators in maternal circulation and at the maternal–fetal interface, and histologic lesions in the placenta in PTB, and indicates a striking contrast that points to aberrant inflammation in PTB. We additionally identified pathways enriched in spontaneous vs iatrogenic PTBs and identified transcriptomic changes concerning placental lesions. A better understanding of these changes will be beneficial to identifying pregnancies at high risk of PTB to develop novel therapeutic targets and subsequently promote neonatal health.
Acknowledgments
We thank all the patients who participated in this study and Sophie Perreault, the research nurse, for patient recruitment.
A, All viable, single cells were divided into CD3+ and CD3− populations according to the CD3 (FITC) marker (center). From the CD3− gate, A, NK cells (CD3−/CD56+) were gated using the CD56 (APC) marker (bottom), B, B-lymphocytes (CD3−/CD19+) were gated using the CD19 (V500) marker (bottom), and C, monocytes (CD3−/CD14+) were gated using the CD14 (Alexa Fluor 700) marker (bottom). From the CD3+ gate, B, NK-T cells (CD3+/CD56+) were gated using the CD56 (APC) marker (center), A, CD8+ cells were gated using the CD8 (PerCP.Cy5.5) marker (top), and A, CD4+ cells were gated using the CD4 (APC-Cy7) marker (top). From the CD4+ gate, B, Treg cells (CD3+/CD4+/CD25+) were gated using the CD25 (BV605) marker (top), C, CCR4+ cells were gated using the CCR4 (BV421) marker (top), and C, CXCR3+ cells were gated using the CXCR3 (PE-CF594) marker (bottom). D, From CCR4+ cells, Th cell subsets were gated on the basis of their expression of CD4 (APC-Cy7) and high (Th17: CD3+/CD4+/CD194+/CD196high) or low (Th2: (CD3+/CD4+/CD194+/CD196low) expression of CCR6 (PE-Cy7) (top). Likewise, from CXCR3+ cells, Th cell subsets were gated on the basis of their expression of CD4 (APC-Cy7) and high (ThTh17: CD3+/CD4+/CD183+/CD196high) or low (Th1: CD3+/CD4+/CD183+/CD196low) expression of CCR6 (PE-Cy7) (center). Data were analyzed using FlowJo software.
NK, natural killer.
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
We clustered all of the upregulated pathways using Cytoscape to remove redundancy (biological processes, medium network specificity, pathway P<.05). Three major pathways were upregulated in PTB, namely lymphocyte chemotaxis (white), neutrophil migration (gray), and antibacterial humoral response (black).
PTB, preterm birth.
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
Genes upregulated (red) and downregulated (green) in KEGG pathway enrichments provide potential targets for future work looking to regulate these pathways.
KEGG, Kyoto Encyclopedia of Genes and Genomes.
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
A, IL-6 levels in term and PTB placentas show correlations with GA (T: R2=0.104; P<.05; PTB: R2=0.220; P<.01). B, Both term and PTB showed no significant correlations with GA for their expression of placental TNF-α (T: R2=0.023; P=.35; PTB: R2=0.006; P=.65). Statistical analysis by simple linear regression.
All placental lesions were analyzed by an expert pathologist according to the international Amsterdam criteria. Data presented as abnormality (percentage).
PTB, preterm birth.
Couture. Compartment-specific proinflammatory changes in preterm birth. Am J Obstet Gynecol 2023.
Global, regional, and national causes of under-5 mortality in 2000-15: an updated systematic analysis with implications for the sustainable development goals.
Cytokine profiling: variation in immune modulation with preterm birth vs. uncomplicated term birth identifies pivotal signals in pathogenesis of preterm birth.
An immune paradox: how can the same chemokine axis regulate both immune tolerance and activation?: CCR6/CCL20: a chemokine axis balancing immunological tolerance and inflammation in autoimmune disease.
Morphologic and histochemical evidence for the occurrence of collagenolysis and for the role of neutrophilic polymorphonuclear leukocytes during cervical dilation.
This work was supported by grants from the Canadian Institutes of Health Research to S.G. and scholarships from the Quebec Research Network in Perinatal Determinants of Children Health and the Centre de recherche en reproduction et fertilité to C.C. Funding sources were not involved in any part of the study design, analysis, interpretation, and writing of this work.
Part of this work was presented at the 69th annual meeting of the Society for Reproductive Investigation, Denver, CO, March 14–19, 2022.
Cite this article as: Couture C, Brien ME, Boufaied I, et al. Proinflammatory changes in the maternal circulation, maternal–fetal interface, and placental transcriptome in preterm birth. Am J Obstet Gynecol 2023;228:332.e1-17.
00677-9&id=gr1.jpg){kind=link}
00677-9&id=gr2.jpg){kind=link}
00677-9&id=gr3.jpg){kind=link}
00677-9&id=gr4.jpg){kind=link}
00677-9&id=gr5.jpg){kind=link}
00677-9&id=gr6.jpg){kind=link}
00677-9&id=gr7.jpg){kind=link}
00677-9&id=fx1.jpg){kind=link}
00677-9&id=fx2.jpg){kind=link}
00677-9&id=fx3.jpg){kind=link}
00677-9&id=fx4.jpg){kind=link}