Abstract
Heat shock proteins are involved in the response to stress including activation of the immune response. Elevated circulating heat shock proteins are associated with spontaneous preterm birth (SPTB). Intracellular heat shock proteins act as multifunctional molecular chaperones that regulate activity of nuclear hormone receptors. Since SPTB has a significant genetic predisposition, our objective was to identify genetic and transcriptomic evidence of heat shock proteins and nuclear hormone receptors that may affect risk for SPTB. We investigated all 97 genes encoding members of the heat shock protein families and all 49 genes encoding nuclear hormone receptors for their potential role in SPTB susceptibility. We used multiple genetic and genomic datasets including genome-wide association studies (GWASs), whole-exome sequencing (WES), and placental transcriptomics to identify SPTB predisposing factors from the mother, infant, and placenta. There were multiple associations of heat shock protein and nuclear hormone receptor genes with SPTB. Several orthogonal datasets supported roles for SEC63, HSPA1L, SACS, RORA, and AR in susceptibility to SPTB. We propose that suppression of specific heat shock proteins promotes maintenance of pregnancy, whereas activation of specific heat shock protein mediated signaling may disturb maternal–fetal tolerance and promote labor.
Similar content being viewed by others
Introduction
Heat shock proteins (HSPs) are evolutionarily highly conserved and present in all cell types in all organisms. They constitute a large family of proteins that are classified according to approximate molecular weights ranging from 10 to 100 kDa: HSP10 (i.e., approximately 10 kDa HSPs), HSP40, HSP60, HSP70, HSP90, and HSP110. Some HSPs are expressed constitutively under normal conditions, whereas others are stress-induced under adverse environmental conditions such as heat, hypoxia, oxidative stress, infection, and inflammation1,2,3. The role of HSPs depends on their localization. Intracellular HSPs act as molecular chaperones and, together with co-chaperones, contribute to the maintenance of cellular homeostasis. Intracellular HSPs stabilize proteins against aggregation, mediate folding of newly translated proteins, and assist with protein translocation across intracellular membranes1,4. Extracellular or circulating HSPs are involved in activation of innate and adaptive immune responses2,3. It is known that infection and inflammation are significant risk factors in preterm birth5. Moreover, HSPs have a role in maturation and inactivation of nuclear hormone receptors (NRs) such as glucocorticoid, androgen, estrogen, and progesterone receptors6,7.
The incidence of preterm birth (i.e., birth before 37 completed weeks of gestation) varies from about 5% in Scandinavian countries8 to up to 19% in Bangladesh9. In approximately 70% of preterm deliveries, labor starts spontaneously. There are no effective ways to either predict or prevent spontaneous preterm birth (SPTB). One reason for this is limited knowledge of the pathways that regulate the timing of birth. The mechanisms leading to onset of normal term delivery likely comprise a complex interplay among fetus, placenta, and mother. In SPTB, it is thought that several pathological processes affect one or more labor-initiating factors10. Addressing single risk factors independently does not prevent preterm birth, suggesting that multiple etiologies are part of a complex parturition-initiating mechanism11,12.
HSPs are an important part of the developmental program and are among the first proteins expressed by the zygote after fertilization13,14. HSPs are also expressed during early pregnancy in both the embryo and maternal side of the placenta (i.e. decidua). Moreover, HSPs are expressed during neurulation, organogenesis, and on throughout fetal maturation13,15. In addition, formation of extra-embryonic tissue and organs (i.e. placenta) requires controlled temporal and spatial patterns of HSP expression16. HSP27, HSP60, HSP70, and HSP90, at least, are expressed in normal human placenta, and these HSPs have been suggested to play a role in cell viability and function17. On the other hand, abnormal HSP levels have been associated with pregnancy complications like transient hypertension, preeclampsia, preterm prelabor rupture of membranes (PPROM), and SPTB18,19,20.
Maternal and fetal genomes are estimated to contribute to the variation in timing of birth by 25–40%21. Several large studies investigating the genetic background of SPTB have been conducted12,21,22, several maternal loci have been robustly associated with preterm birth23, as has at least one fetal locus24. A previous study identified rare, likely damaging genetic variants of HSPA1L (HSP70 family) in Finnish mothers from families with recurrent SPTB25. Furthermore, HSPA1L showed an association with SPTB in a large genome-wide association study (GWAS) in a European American population25.
NRs have also been associated with SPTB. A recent study identified the glucocorticoid receptor (GR) signaling pathway as a candidate for SPTB risk25. The GR signaling pathway has crucial roles in glucose metabolism, growth development, and immune function, and may interact with progesterone, a key hormone required for normal pregnancy26,27.
Because HSPs have many roles during pregnancy and some have been linked to pregnancy complications including SPTB25, we sought to evaluate the importance of HSP coding genes, as well as HSP-regulated NRs, in relation to SPTB. We used multiple available data sources, such as GWAS, whole-exome sequencing (WES), and placental transcriptomic data to mine for evidence of HSP and NR gene involvement in susceptibility to SPTB.
Results
HSPs have been linked to preterm birth25 and other pregnancy complications18,28. We investigated all 97 genes encoding members of the HSP families and all 49 genes encoding NRs, the targets of HSPs, for their role in SPTB susceptibility (Table S1 and S2, respectively). They included all HSPs and NRs in the UniProt database29 at the time of analysis. Pathway analysis showed HSPs and NRs were the most enriched in “Protein processing in endoplasmic reticulum” and “Estrogen signaling” KEGG pathways. Since estrogen is critical for maintenance of pregnancy and initiation of labor30, a role for HSPs and NRs in pregnancy and labor is plausible.
A broad array of SPTB datasets, including GWAS, WES, and placental transcriptomics data, were screened for evidence suggesting associations with HSP and NR family members (Fig. 1). We first queried five GWAS datasets [23andMe (mothers only), Nordic metadata with Finnish, Danish, and Norwegian subsets, and a Northern Finnish dataset] to examine common variant associations using both “mother as affected” (giving birth preterm) and “child as affected” (born preterm). Secondly, we examined maternal and fetal exomes belonging to the Northern Finnish population set and in maternal exomes belonging to the Danish population set for potentially damaging, rare variants in HSP and NR genes. Thirdly, we sought changes in transcription of HSP genes in placentas from premature deliveries and spontaneous (STB) and elective (ETB) term controls. Findings are listed in Table 1, and main results are shown in Fig. 2.
Summary of the datasets used in datamining. A genome-wide association study (GWAS) datamining included five different datasets [23andMe; Nordic metadata with Finnish, Danish and The Norwegian Mother, Father and Child Cohort Study (MoBa) subsets; and Northern Finnish dataset]. Whole exome sequencing (WES) was from two different datasets; the Northern Finnish and Danish population sets. Datamining of two transcriptomics datasets was conducted. Samples of the transcriptomics were from placentas from spontaneous preterm deliveries (SPTB) and spontaneous (STB) and elective (ETB) term controls. * Danish WES; n = 93 sister pairs and two families with three affected siblings. 1Zhang G et al. (2017) Genetic Associations with Gestational Duration and Spontaneous Preterm Birth23. 2Unpublished data. 3Zhang G et al. (2015) Assessing the Causal Relationship of Maternal Height on Birth Size and Gestational Age at Birth: A Mendelian Randomization Analysis31. 4Tiensuu H et al. (2019) Risk of spontaneous preterm birth and fetal growth associates with fetal SLIT232. 5Huusko JM et al. (2018) Whole exome sequencing reveals HSPA1L as a genetic risk factor for spontaneous preterm birth25. 6Ackerman WE et al. (2016) Comprehensive RNA profiling of villous trophoblast and decidua basalis in pregnancies complicated by preterm birth following intra-amniotic infection33. 7Brockway HM et al. (2019) Unique transcriptomic landscapes identified in idiopathic spontaneous and infection related preterm births compared to normal term births34.
Top findings of HSPs and NRs associated with SPTB. Several datasets supported roles for SEC63, HSPA1L, SACS, RORA, and AR in susceptibility to SPTB. These genes are underlined in the figure.
Several HSP and NR genes have suggestive association signals in GWAS datasets
In the five GWAS datasets we sought significant (p < 5 × 10−8) and suggestive (p < 1 × 10−5) associations between SPTB and 100-kb windows surrounding each of 97 HSP and 49 NR genes. For comparing HSP and NR variants with exome data and transcriptomics, we extracted all GWAS loci with p < 1 × 10−4 to construct a large preliminary gene set of potential importance in SPTB. The GWAS datasets supported roles for multiple HSP and NR genes in SPTB susceptibility (Fig. 2, Table 1).
Five HSP genes (DNAJB8, DNAJB14, DNAJC6, DNAJA3, and SEC63) exhibited potentially significant associations with mothers giving birth preterm (p < 4 × 10−9–1 × 10−4) in the 23andMe dataset (Table S3). A previously reported, genome-wide significant signal (p < 1 × 10−8) was detected upstream of DNAJB8 in an intron of EEFSEC, and the association replicated in the Nordic dataset (p < 1 × 10−4) (Table S4)23. DNAJB14 had a suggestive association (p < 1 × 10−5) that replicated in the maternal Nordic Danish data (p < 1 × 10−4, Tables S3 and S4). Six other genes (DNAJA1, DNAJC1, DNAJC2, DNAJC11, DNAJC17, and MKKS) were detected at the alpha level of p < 1 × 10−4 in the maternal Nordic datasets (Table S4). Most of the maternal HSP SNP associations in the Nordic data overlapped with those in the 23andMe data.
Among NR genes, suggestive maternal associations (p < 1 × 10−5) with SPTB were observed in NR2F2 and THRA, and PPARG and RORA had variants with p < 1 × 10−4 (Table S5, 23andMe dataset). In the Nordic datasets, variants in RORA, ESR1, and NR2F6 had p values < 1 × 10−4 (Table S6). No associations were seen in the Northern Finnish maternal data.
In the GWAS datasets in which the infant was treated as affected, only one variant in HSPA12B showed a suggestive association (p < 1 × 10−5) in the Northern Finnish data. Variants in HSP90AA1, DNAJC5B, DNAJC12, and CCT3 had p values of < 1 × 10−4 in the Nordic Finnish, MoBa, or meta-analysis datasets (Tables S7 and S8). In the maternal GWAS data, significant and suggestive association signals were detected for DNAJB8/EEFSEC and DNAJB14, respectively; these associations were replicated in independent GWAS datasets. In addition, many variants with p < 1 × 10−4 were shared across different GWASs in both maternal and fetal data. After GWAS, we investigated the presence of potentially damaging variants in the HSP and NR genes from maternal and fetal exomes.
Potentially damaging variants of HSP and NR genes identified in whole exome sequence
The presence of rare [minor allele frequency (MAF) < 1%] and common (MAF 1–10%), potentially damaging variants [category 1–3 in accordance with classification of the American College of Medical Genetics (ACMG)35] in the HSP and NR genes was investigated in maternal and fetal exomes. Among affected Northern Finnish individuals (either giving birth preterm or born preterm), we found 15 HSP genes and ten NR genes with potentially damaging heterozygous variants in multiple individuals or single individuals with multiple, potentially damaging variants in a single gene. We found 18 HSP and 12 NR genes with rare, possibly damaging variants that were shared by affected Danish maternal sibling pairs. Genes CCT7, HSPA1L, HSPA5, HYOU1, SACS, SEC63, AR, and NR1H4 had rare, potentially damaging variants in both Finnish and Danish affected exomes. We previously reported the presence of rare, damaging variants in Heat Shock Protein 70-kDa-like 1 (HSPA1L) in Finnish and Danish families with recurrent SPTBs25. One of the variants, rs34620296, also showed a trend toward significance in the 23andMe GWAS data (p = 1 × 10−3; MAF in cases, 0.0025 and MAF in controls, 0.0010), and was previously shown to reduce chaperone activity36 and affect decidualization25.
Sacsin (SACS), a HSP gene associated with spastic ataxia (OMIM: #270,550)37,38, had four potentially damaging (ACMG category 1–3) missense variants in the Finnish exomes (rs192610957, rs144267558, rs116907814, and rs17325713) and two missense variants in the Danish exomes (rs147099630 and chr13:23915410G > A, p.H119Y). According to the Combined Annotation Dependent Depletion (CADD, v1.6)39 score (> 30), these variants are among the top 1% of deleterious variants in the human genome. Rs192610957 and rs116907814 were enriched in the Finnish population compared to the general European population (MAF 0.007 vs. 0.0005, and 0.011 vs. 0.0009, respectively) (http://www.sisuproject.fi/; https://gnomad.broadinstitute.org/).
In both Finnish and Danish families, HSPA5 variant rs56136100 was shared by two affected mothers within a family. Rs56136100 is a non-conservative, missense variant (p.Glu557Gly) that is predicted to be damaging by multiple in silico tools (SIFT v6.2.140, PolyPhen-2 v241, and MutationTaster v242) and has a CADD v1.6 score of 33. This missense variant, could potentially affect the physiochemical properties of HSPA5, as the sequence variant causes a change from an acidic (Glu) to a hydrophobic (Gly) amino acid.
The androgen (dihydrotestosterone) receptor, AR, a NR gene associated with androgen insensitivity syndrome (OMIM: #300,068), harbored five potentially damaging variants: rs137852593 and rs5031002 (Finnish exomes), and rs201934623; chrX:66766114C > A, p.P186T; and chrX:66900678A > G, p.R628G (Danish exomes). In addition, several individuals who belonged to the Finnish families had longer androgen receptor (AR) exon-1 CAGn repeats, together with HSPA1L rs34620296, which was previously associated with SPTB25. Furthermore, in our previous study, longer AR CAGn repeats were overrepresented in preterm infants compared to term controls43.
In the progesterone receptor (PGR), two variants, rs11571145 and rs11571222, were shared by Northern Finnish affected mothers in two different families. According to RegulomeDB (www.regulomedb.org), rs11571145 (p.Pro186Leu in the PGR-B isoform) is annotated as likely to affect binding to DNA. This variant is also predicted to be damaging and disease causing by the in silico tools SIFT v6.2.140 and MutationTaster v242. Moreover, in the family with rs11571145, two affected mothers also shared common missense variants rs1042838 and rs3740753, which showed nominal significance (p = 0.007 and p = 0.008, respectively) for gestational age (GA) in the 23andMe data as well as a trend (p = 0.074, effect (eff) = − 1.213 and p = 0.046, eff = − 1.362, respectively) in the maternal Nordic metadataset. Variants rs11571145, rs1042838, and rs3740753 are missense variants (Pro22Leu, Val496Leu, and Ser180Thr, respectively, in the PGR-A isoform, and Pro186Leu, Val660Leu, and Ser344Thr, respectively, in the PGR-B isoform). The role of different PGR isoforms in pregnancy is well supported44,45. Especially, the expression of low affinity variant of PGR, PGR-A increases towards term labor46.
The WES datasets also supported roles for HSP and NR genes in SPTB susceptibility (Fig. 2, Table 1). Genes CCT7, HSPA1L, HSPA5, HYOU1, SACS, SEC63, AR, and NR1H4 had rare, potentially damaging variants in both Finnish and Danish exomes. Both genetic variation in HSP genes and differences in their expression are associated with SPTB18,19,20. Thus, after datamining the genetic datasets, we investigated changes in HSP and NR mRNA expression in the placenta by investigating transcriptomics datasets.
Placental transcriptomics identify differences in HSP and NR gene expression during SPTB
Expression levels of HSPs change during pregnancy, especially in complicated pregnancies28, suggesting that changes in HSP expression might have a role in pregnancy complications. To investigate whether HSP and NR expression levels change in SPTB, we examined RNA levels of 97 HSP genes and 49 NR genes in placentas from premature deliveries and term controls from Northern Finland. We compared expression levels in three groups: spontaneous preterm birth (SPTB, n = 6), spontaneous term birth (STB, n = 6), and elective term birth (ETB, n = 6). 15 HSP genes (Table 2) and seven NR genes (Table 3) were significantly up- or downregulated (p < 0.05) in groupwise comparisons.
These results imply that expression levels of multiple HSP and NR genes change in preterm birth. The most robust changes in placental gene expression were HSPA1, DNAJC30, HSPD1, and NR6A1, which exhibited congruent, significant (p < 0.05) differences in comparisons of SPTB vs. spontaneous term placentas and SPTB vs. elective term placentas. This suggests that mRNA expression changes in these genes are associated with prematurity, rather than spontaneous labor.
Top genes in placental villous and decidual cells from discovery and validation SPTB vs. elective term birth datasets
We also compared RNA expression levels in maternal placental tissues (decidua basalis) and fetal placental tissues (villous tissue) in spontaneous preterm births (SPTB, n = 5) and elective term births (ETB, n = 5)33,34. DNAJB7, AR, and ESRRA were upregulated in SPTB vs. ETB in the decidua, whereas HSPA7, NR6A1 and RORA were downregulated in SPTB compared to ETB in villous tissue (Table 4, p < 0.05, t test).
Some of these findings were mirrored in the validation datasets. NR6A1 was among the top findings in placental transcriptomics data that originated in northern Finland, whereas modest association signals (p < 0.0001) were detected for the region encompassing RORA in 23andMe and Nordic GWAS datasets.
Confirmation of SPTB associations in orthogonal datasets
We datamined multiple GWAS, WES, and transcriptomics datasets from mothers, infants, and placenta. HSPs and NRs that are potentially associated with SPTB were found in all these datasets. Multiple HSPs and NRs (Table 1, Fig. 2) have potential roles in SPTB susceptibility in at least one of the datasets. Finally, we compared the results from different analyses. Several datasets supported roles for SEC63, HSPA1L, SACS, RORA, and AR associations with spontaneous preterm birth (Table 1, Fig. 2).
Discussion
Previous studies have suggested a role for certain HSPs in pregnancy complications, which led us to look for HSPs and NRs that might affect SPTB risk. Analysis of GWAS, WES, and transcriptomics data from mothers, infants, and placentas, revealed HSP and NR gene associations with SPTB in each (Table 1). More significantly, we identified several HSPs and NRs with associations with SPTB in multiple, orthogonal datasets – notably SEC63, SACS, RORA, AR, and PGR. Previous studies have suggested that SPTB to be attributable to multiple pathological processes11. Thus, varying pathways leading to SPTB could partly explain the variations in the results among the datasets. Maternal, and also to some extent fetal, genomes affect the susceptibility to preterm birth and duration of pregnancy in general23,47,48,,48. We analyzed several maternal and fetal GWAS and WES datasets to identify HSP and NR genes associated with preterm birth. DNAJB8, DNAJB14, SEC63, and RORA showed associations in at least two genomic datasets.
Previous studies have indicated that in addition to changes in levels of HSP expression, changes in the distribution/relative concentration of different HSPs could lead to pregnancy complications. For example, early changes in the ratio of circulating HSP60 to HSP70 have been shown to predict miscarriage14. Additionally, there are well-defined temporal and spatial patterns of HSP expression in the human placenta13. Changes in HSP expression could affect placental pathology and cause pregnancy complications like preterm birth. Consequently, we also searched for SPTB-associated changes in HSP mRNA expression in the placenta. A comparison of SPTB and placentas from spontaneous and elective term pregnancies (Table 2) indicated that multiple HSPs differed in expression.
Elevated circulating HSP concentrations have previously been associated with increased risk of pregnancy complications such as preeclampsia and preterm delivery3,28,49. Circulating HSPA1A (of Hsp70 family) levels are elevated in patients at high risk for preterm delivery50. In our current study, HSPA1A mRNA expression was upregulated in SPTB placentas compared to placentas from term pregnancies. Circulating HSPA1A levels are downregulated in women with a normal pregnancy compared to in nonpregnant women28,50,51. Extracellular HSPA1A may be removed by innate immune mechanisms as part of tolerogenic changes in the immune system and, as a result, may promote the maintenance of immunological tolerance to the fetus. The ability of extracellular HSPA1A to elicit immune responses might be harmful in pregnancy and could lead to maternal immune rejection of the fetus28. By disturbing this tolerance, upregulation of HSPA1A during pregnancy could increase the risk of preterm labor. In contrast to HSPA1A, some HSPs, like Hsp60, are present in the peripheral circulation of healthy nonpregnant and pregnant individuals52,53. Protein levels of some HSPs increase along with advancing gestational age, which may reflect their involvement in initiation of labor54. Suppression of HSP production during pregnancy could be an important mechanism for maintaining pregnancy. However, it is also possible that elevated HSP levels are a consequence of harmful conditions such as preterm labor and a sign of the body’s attempt to maintain homeostasis.
HSPs are essential to the maturation and inactivation of NRs. In this study, we found rare, potentially damaging variants located within the exons of PGR and AR receptor genes from families with multiple SPTB. Moreover, in 23andMe and Nordic GWAS datasets RORA associated with SPTB. Progesterone has an essential role in the maintenance of pregnancy55. Progesterone withdrawal has been noted to result in parturition in some animals, but plasma levels of progesterone in humans remain high until the placenta is removed. On the other hand, PGRs are potential regulators of timing of birth56. Progesterone and PGRs may have a role in anti-inflammatory responses in the myometrium, and impaired function of PGRs may lead to initiation of labor. Moreover, AR has important roles during pregnancy57 and has previously been linked to pregnancy complications. For example, longer AR CAGn repeats are overrepresented in women with recurrent spontaneous abortions58 and in SPTB infants43, and increased AR expression has been observed in the placentas of preeclamptic women59. Additionally, higher AR ligand levels result in myometrial contractions, cervical dilatation and in preterm birth57. RORA, on the other hand, regulates genes involved in inflammatory response and circadian rhythm, for instance60,61,62. Changes in circadian rhythm have been shown to associate with placental detachment and SPTB61,62.
According to a KEGG pathway analysis, many HSP and NR genes (HSPA1A, HSPA1B, HSPA1L, HSPA2, HSPA6, and HSPA8), including three HSP90 genes (HSP90AA1, HSP90AB1, and HSP90B1), play a role in the estrogen signaling pathway. Estrogen signaling is one of the main pregnancy-associated pathways in which HSPs and NRs play a role. In the absence of estrogenic ligands, estrogen receptor (ER), like other nuclear hormone receptors, is assembled into an Hsp90-based chaperone protein complex, which keeps the ER in a ligand binding–competent but inactive state. A total of 21 HSPs and three HSP co-chaperones have been found to associate with the ER63. Estrogen signaling is necessary for a successful pregnancy, as estrogen is required for processes such as proliferation of the myometrium before term and the contractile response that leads to parturition at term64. Thus, changes in HSP genes or protein expression could affect estrogen signaling and cause pregnancy complications, promoting preterm birth.
SEC63 (of Hsp40 family) was identified to associate with SPTB in GWAS, WES and placental transcriptomic data. SEC63 act as a co-chaperone, is a component of the protein translocation machinery in endoplasmic reticulum and associates with decidualization in early pregnancy65,66. The damaging variants of SEC63 and lower expression levels of SEC63 in the placenta might affect implantation site and decidualization predisposing to SPTB. Another Hsp40 family member, SACS had damaging variants in both Finnish and Danish affected exomes and was upregulated in the placentas of SPTB. It has been suggested that SACS is a key player in cellular protein quality control system and in organizing proteins into bundles called intermediate filaments67,68. Being part of protein quality control, increased levels of SACS in the placenta could be due to a consequence of harmful conditions that are responsible for SPTB.
One of the important GWAS findings of the present study were in a region that has been shown to loop to DNAJB8 [DnaJ Heat Shock Protein Family (Hsp40) Member B8].The looping region has a genome-wide significant signal associated with SPTB23. DNAJB8 has a role in suppressing aggregation and toxicity of polyglutamine proteins [i.e., proteins containing polyglutamine (polyQ) regions that are encoded by repetitive CAG or CAA DNA sequences]. Proteins with expanded polyQ regions can cause pathogenic phenotypes (e.g., neurodegenerative phenotypes), and intermediate levels of polyQ expansion can influence host cell susceptibility to misfolded pathogenic protein; however, other studies have proposed that polyQ aggregates can be benign or even offer protection from toxicity associated with smaller, oligomeric conformers69.
HSPs have important roles starting at the beginning of pregnancy, and they are among the first proteins expressed by the zygote after fertilization. They are expressed during early pregnancy stages in both the embryo and maternal decidua. For example, they maintain the integrity of intracellular proteins14. A recent study revealed that HSPs from placental mitochondria may be associated with trophoblast differentiation70. As HSPs have many important functions throughout pregnancy, it is plausible that changes in either HSP genes or their expression could compromise maintenance of normal pregnancy, leading to SPTB. On the other hand, HSPs are involved in activation of the innate and adaptive proinflammatory immune response. It is well established that infection and inflammation represent a highly significant risk factor in preterm birth5. Thus, it is possible that the changes in HSP expression associated with preterm birth are due to activation of inflammation-related pathways.
There were some clear limitations in our study. First, we could not differentiate whether a specific gene associates with early preterm or late preterm birth. In our analysis, the preterm data consisted of specimens obtained both before 30 weeks and after 30 weeks of gestation Second, there are likely differences how placental samples were collected. There were three different data sets utilizing tissue samples from human placenta. Two of these data sets were already published33,34. It may be that not all the placental samples have been collected according to standard as proposed by the International Federation of Placenta Associations. These methodological limitations decrease likelihood to detect gestational age specific associations and complicates comparisons between the data sets originating from placental tissues.
In conclusion, GWAS, WES, and transcriptome datasets indicated that various HSPs and NRs are associated with SPTB susceptibility. Further studies are required to resolve the exact roles of different HSPs and NRs in SPTB. The length of pregnancy is a species-specific which makes it challenging to study human pregnancy. However, for example, gene knock-down in relevant cell line could allow in vivo studies of gene function and identify specific pathways in preterm birth. We propose that activation of HSP signaling disturbs maternal–fetal tolerance and promotes susceptibility to early labor. Mechanistic studies are needed to resolve how genetic variants of HSP and NR genes compromise maintenance of normal homeostasis to support normal pregnancy or promote initiation of events leading to SPTB.
Methods
GWAS data
We used multiple available sources of preterm birth GWAS data, including both maternal and fetal genomes. The first dataset we used comprised 43,568 mothers of European ancestry; these mothers were identified from among 23andMe’s research participants as described previously48. The second dataset included meta-analysis data for 4632 mothers and their 1960 infants from three independent Nordic (Finnish, Danish, and Norwegian) preterm birth case/control data sets of European ancestry31. In this dataset, preterm samples were enriched, and samples from borderline preterm and early term (gestation age 37–38 weeks), as well as post-term (gestation age > 42 weeks), births were excluded. The third dataset included 608 mothers with spontaneous preterm (gestation age < 36 weeks) or term (gestation age 38–41 weeks) deliveries and their preterm or term born children. This dataset originated exclusively from northern Finland, and a full data description was presented previously in detail32.
All preterm births included in the Nordic or Northern Finnish population sets were spontaneous. Obstetrical induction of labor, placental abnormalities, preeclampsia, congenital malformations, and multiple births were excluded. Pregnancies involving preexisting medical conditions known to be associated with preterm birth and pregnancies with complications were also excluded.
WES data
We had two population sets with WES data available. The first was a dataset of Northern Finnish mothers with preterm deliveries (n = 13)25 and their children (n = 23) who were born preterm (gestation age < 36 weeks). This population set comprised seven unrelated families with a strong family history of recurrent SPTBs; the set was selected retrospectively from the birth diaries of Oulu University Hospital from 1973 to 2003 and prospectively from 2003 to 2005. Selection criteria were described previously in detail43,71. Another population set of European ancestry with available WES data was from Denmark and included 192 women from 95 families: 93 affected sister pairs (both sisters had given birth preterm) and two sister triads with preterm deliveries occurring before 37 completed weeks of gestation. All women had experienced at least one PTB and, in the majority of the sister pairs (83%), both sisters had experienced a SPTB.
WES for both of the Finnish and Danish exomes was performed as previously described in detail25. In short, next-generation exome sequencing of the Finnish samples was performed at the Center for Pediatric Genomic Medicine, Children’s Mercy Hospital (CMH; Kansas City, MO, USA). Exome samples were prepared with the Illumina Nextera Rapid Capture Exome kit and sequenced with the Illumina HiSeq 2500 instrument. Sequence data were generated with Illumina RTA 1.18.64.0 and bcl2fastq-1.8.4 and aligned against the reference human genome (GRCh37.p5). Variant calls were made with the Genome Analysis Toolkit (GATK v4.2.0.0)72, and duplicate reads were identified and flagged with the Picard MarkDuplicates tool72. Exomes of the Danish sample set were sequenced with the Complete Genomics platform (BGI, Shenzhen, China) by using the manufacturer’s pipeline. Reads were aligned against the National Center for Biotechnology Information (NCBI) build 37 human reference genome.
For the Finnish exomes, we used variant annotation data from the Center for Pediatric Genomic Medicine’s CMH Variant Warehouse database (http://warehouse.cmh.edu), including frequency data for approximately 3900 individuals previously sequenced at the center73. Pathogenicity was categorized according to the ACMG35 as: 1, previously reported to be disease-causing; 2, expected to be pathogenic (loss of initiation, premature stop codon, disruption of stop codon, whole-gene deletion, frame shifting indel, and disruption of splicing); and 3, unknown significance but potentially disease-causing (nonsynonymous substitution, in-frame indel, disruption of polypyrimidine tract, and overlap with 5′ exonic, 5′ flank, or 3′ exonic splice contexts). Only variants that fit one of these criteria (1–3) were considered interesting. For the Danish exomes, we used Ingenuity Variant Analysis software (Qiagen) and included only rare (MAF < 1%) likely damaging variants that were shared by the affected sisters in each family.
Placental transcriptomics data
We used available transcriptomics data to explore HSP and NR levels in human placenta. Placental tissues were collected at Oulu University Hospital in 2012–2014, and all samples were from uncomplicated preterm or term pregnancies as described previously74. Each placenta was inspected in terms of morphology; weight, size, cord position, infarcts and calcification were recorded. In short, samples were collected from the basal plate immediately underneath the placental surface (the maternal side of placenta). The transcriptomic set consisted of placenta samples resulting from spontaneous vaginal deliveries that occurred either preterm (SPTB; gestation age < 36 weeks, n = 6) or term (STB; gestation age > 38 weeks, n = 6), and from elective caesarean deliveries without signs or symptoms of labor at term (ETB; gestation age > 38 weeks, n = 6).
Transcriptomics data were generated from RNA isolated from the basal plate of placentas with the Qiagen Rneasy Micro kit. RNA quality was assessed with the Agilent RNA 6000 Nano kit in the Agilent 2100 Bioanalyzer instrument. The samples were sequenced with the HiSeq2500 instrument using paired-end sequencing chemistry with 100 bp read length. Number of lanes used in sequencing was 1. The reads obtained from the instrument were base called using the instrument manufacturer’s Bcl2fastq version 1.8.4 base calling software. Read quality droped at the ends of the reads and thus quality trimming was needed. Trimming of reads was done with Trimgalore version 0.3.3 (https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) and Cutadapt (v1.1)75. The reads were aligned against the human reference genome (hg19 assembly, downloaded from UCSC) using TopHat version 2.0.10.1 (https://ccb.jhu.edu/software/tophat/index.shtml). Only uniquely aligned reads were used for the further analysis.
Next the reads were associated with known genes based on RefSeq annotations derived from UCSC database and the number of reads associated with each gene was counted using HTSeq tool version 0.6.1 (https://htseq.readthedocs.io/en/master/). Here the counts were normalised using the TMM normalisation method of the edgeR R/Bioconductor package76. For statistical testing the data were further transformed using the voom approach in the limma package77.
Data were normalized to remove variations among the samples. Finally, there was a very high correlation among the samples. The thresholds used in filtering the differentially expressed genes were p values of < 0.05. The placental transcriptomics data were deposited in NCBI’s Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo) and are accessible through GEO Series accession number GSE120480.
RNA expression data from placental villous and decidual cells; discovery dataset
We used publicly available RNA sequencing data (GEO dataset ID: GSE73714)33 from paired villous trophoblast and decidua basalis specimens collected from spontaneous idiopathic preterm birth (SPTB; gestation age 30–33 weeks, n = 5) or term birth (by caesarean section, absence of labor; i.e., ETB; gestation age 38–39 weeks, n = 5). Details of the RNA sequencing and data generation were described previously33. In short, total RNA was extracted from flash-frozen specimens with TRIzol. RNA-seq libraries were constructed with the TruSeq Stranded Total RNA Sample Prep Kit with Ribo-Zero Gold (Illumina) and sequenced with the Illumina HiSeq 2500 platform. Illumina Analysis pipeline in HiSeq Control Software v2.2.38 was used for image analysis, base calling, and error estimation. Reads were mapped to UCSC hg38, and data were normalized by using Bioconductor statistical packages (https://www.bioconductor.org/).
RNA expression data from villous trophoblasts; replication/metadataset
This dataset included placental villous samples collected from pregnancies that resulted in SPTB (gestation age 29–36 weeks, n = 8) or term (gestation age 38–42 weeks, n = 9) deliveries. Full data description, generation, and analysis were explained in detail elsewhere34. In short, total RNA was prepared from snap-frozen biobank placental samples and submitted to the University of Cincinnati Genomics, Epigenomics and Sequencing Core for RNA sequencing with the RiboZero kit (Illumina) and Illumina High Seq 2100 system for library preparation and sequencing, respectively. To increase statistical power, RNA sequence data obtained from these samples were combined with previously published placental villous transcriptomes (GEO GSE73714 term, n = 5 and SPTB, n = 5)33. We compared RNA expression levels (counts per million, CPM) between SPTB and term births and generated heatmaps with GraphPad Prism version 8.0 (https://www.graphpad.com/scientific-software/prism/).
Ethics statement
All experiments were performed in accordance with relevant guidelines and regulations. Written informed consent was obtained from all of the individuals or their guardians who participated in this study. Study methods for the Northern Finnish WES, GWAS, and placental transcriptomics data were approved by the ethics committee of Oulu University Hospital (79/2003, 14/2010, and 73/2013), and for the Danish WES by the University of Southern Denmark (NVK#1302824) and the University of Iowa (IRB#200608748). Individuals in the large GWAS (European American women) were research participants of 23andMe, Inc., a personal genetics company. All of the 23andMe participants provided informed consent and participated in the research online, under a protocol approved by the external AAHRPP-accredited institutional review board Ethical & Independent Review Services (E&I Review). For the Nordic GWAS data, the study was approved by the ethics committees of Oulu University Hospital and Helsinki University Central Hospital (Finnish subset), as well as The Regional Committee for Medical Research Ethics in Norway, 7021/2010/2683 REK South-Eastern (The Norwegian Mother, Father and Child Cohort Study subset). Written informed consent was given by all participants. The study involving villous tissue specimens was approved by the Cincinnati Children’s Hospital Medical Center institutional review board (#IRB 2013-2243, 2015-8030, 2016-2033).
References
Hartl, F. U. Molecular chaperones in cellular protein folding. Nature 381, 571–579 (1996).
Schlesinger, M. J. Heat shock proteins. J. Biol. Chem. 265, 12111–12114 (1990).
Chaiworapongsa, T. et al. Amniotic fluid heat shock protein 70 concentration in histologic chorioamnionitis, term and preterm parturition. J. Matern. Fetal. Neonatal Med. 21, 449–461 (2008).
Hartl, F. U., Bracher, A. & Hayer-Hartl, M. Molecular chaperones in protein folding and proteostasis. Nature 475, 324–332 (2011).
Cappelletti, M., Della Bella, S., Ferrazzi, E., Mavilio, D. & Divanovic, S. Inflammation and preterm birth. J. Leukoc. Biol. 99, 67–78 (2016).
Nollen, E. A. & Morimoto, R. I. Chaperoning signaling pathways: Molecular chaperones as stress-sensing “heat shock” proteins. J. Cell. Sci. 115, 2809–2816 (2002).
Schoneveld, O. J., Gaemers, I. C. & Lamers, W. H. Mechanisms of glucocorticoid signalling. Biochim. Biophys. Acta 1680, 114–128 (2004).
Morken, N. H. et al. Reference population for international comparisons and time trend surveillance of preterm delivery proportions in three countries. BMC Womens Health 8, 16–16 (2008).
Chawanpaiboon, S. et al. Global, regional, and national estimates of levels of preterm birth in 2014: A systematic review and modelling analysis. Lancet Glob. Health. 7, e37–e46 (2019).
Ravanos, K. et al. Factors implicated in the initiation of human parturition in term and preterm labor: A review. Gynecol. Endocrinol. 31, 679–683 (2015).
Romero, R., Dey, S. K. & Fisher, S. J. Preterm labor: One syndrome, many causes. Science 345, 760–765 (2014).
Monangi, N. K., Brockway, H. M., House, M., Zhang, G. & Muglia, L. J. The genetics of preterm birth: Progress and promise. Semin. Perinatol. 39, 574–583 (2015).
Christians, E. S., Zhou, Q., Renard, J. & Benjamin, I. J. Heat shock proteins in mammalian development. Semin. Cell Dev. Biol. 14, 283–290 (2003).
Makri, A. et al. Early changes of the heat-shock protein 60 to 70 ratio as prediction of miscarriage in pregnancy. Am. J. Reprod. Immunol. 81, e13087 (2019).
Luft, J. C. & Dix, D. J. Hsp70 expression and function during embryogenesis. Cell Stress Chaperones 4, 162–170 (1999).
Shah, M., Stanek, J. & Handwerger, S. Differential localization of heat shock proteins 90, 70, 60 and 27 in human decidua and placenta during pregnancy. Histochem. J. 30, 509–518 (1998).
Wataba, K. et al. Changed expression of heat shock proteins in various pathological findings in placentas with intrauterine fetal growth restriction. Med. Electron. Microsc. 37, 170–176 (2004).
Molvarec, A. et al. Association of elevated serum heat-shock protein 70 concentration with transient hypertension of pregnancy, preeclampsia and superimposed preeclampsia: A case-control study. J. Hum. Hypertens. 20, 780–786 (2006).
Fekete, A., Ver, A., Bogi, K., Treszl, A. & Rigo, J. Is preeclampsia associated with higher frequency of HSP70 gene polymorphisms?. Eur. J. Obstet. Gynecol. Reprod. Biol. 126, 197–200 (2006).
Dvorakova, L., Ivankova, K., Krofta, L. & Hromadnikova, I. Expression profile of heat shock proteins in placental tissues of patients with preterm prelabor rupture of membranes and spontaneous preterm labor with intact membranes. Am. J. Reprod. Immunol. 78, 10.1111/aji.12698. Epub 2017 May 12 (2017).
Bezold, K. Y., Karjalainen, M. K., Hallman, M., Teramo, K. & Muglia, L. J. The genomics of preterm birth: From animal models to human studies. Genome Med. 5, 34 (2013).
Strauss, J. F. et al. Spontaneous preterm birth: Advances toward the discovery of genetic predisposition. Am. J. Obstet. Gynecol. 218, 294-314.e2 (2018).
Zhang, G. et al. Genetic associations with gestational duration and spontaneous preterm birth. N. Engl. J. Med. 377, 1156–1167 (2017).
Liu, X. et al. Variants in the fetal genome near pro-inflammatory cytokine genes on 2q13 associate with gestational duration. Nat. Commun. 10, 3927–3928 (2019).
Huusko, J. M. et al. Whole exome sequencing reveals HSPA1L as a genetic risk factor for spontaneous preterm birth. PLoS Genet. 14, e1007394 (2018).
Anacker, C., Zunszain, P. A., Carvalho, L. A. & Pariante, C. M. The glucocorticoid receptor: Pivot of depression and of antidepressant treatment?. Psychoneuroendocrinology 36, 415–425 (2011).
Byrns, M. C. Regulation of progesterone signaling during pregnancy: Implications for the use of progestins for the prevention of preterm birth. J. Steroid Biochem. Mol. Biol. 139, 173–181 (2014).
Molvarec, A. et al. Circulating heat shock protein 70 (HSPA1A) in normal and pathological pregnancies. Cell Stress Chaperones 15, 237–247 (2010).
UniProt Consortium. UniProt: A worldwide hub of protein knowledge. Nucleic Acids Res. 47, D506–D515 (2019).
Challis, J. R. G., Matthews, S. G., Gibb, W. & Lye, S. J. Endocrine and paracrine regulation of birth at term and preterm. Endocr. Rev. 21, 514–550 (2000).
Zhang, G. et al. Assessing the causal relationship of maternal height on birth size and gestational age at birth: A Mendelian randomization analysis. PLoS Med. 12, e1001865 (2015).
Tiensuu, H. et al. Risk of spontaneous preterm birth and fetal growth associates with fetal SLIT2. PLoS Genet. 15, e1008107 (2019).
Ackerman, W. E. et al. Comprehensive RNA profiling of villous trophoblast and decidua basalis in pregnancies complicated by preterm birth following intra-amniotic infection. Placenta 44, 23–33 (2016).
Brockway, H. M. et al. Unique transcriptomic landscapes identified in idiopathic spontaneous and infection related preterm births compared to normal term births. PLoS One 14, e0225062 (2019).
Richards, S. et al. Standards and guidelines for the interpretation of sequence variants: A joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet. Med. 17, 405–424 (2015).
Takahashi, S. et al. De novo and rare mutations in the HSPA1L heat shock gene associated with inflammatory bowel disease. Genome Med. 9, 8–9 (2017).
Grieco, G. S. et al. Novel SACS mutations in autosomal recessive spastic ataxia of Charlevoix-Saguenay type. Neurology 62, 103–106 (2004).
Criscuolo, C. et al. A novel mutation in SACS gene in a family from southern Italy. Neurology 62, 100–102 (2004).
Kircher, M. et al. A general framework for estimating the relative pathogenicity of human genetic variants. Nat. Genet. 46, 310–315 (2014).
Sim, N. L. et al. SIFT web server: Predicting effects of amino acid substitutions on proteins. Nucleic Acids Res. 40, 452 (2012).
Adzhubei, I. A. et al. A method and server for predicting damaging missense mutations. Nat. Methods 7, 248–249 (2010).
Schwarz, J. M., Cooper, D. N., Schuelke, M. & Seelow, D. MutationTaster2: Mutation prediction for the deep-sequencing age. Nat. Methods 11, 361–362 (2014).
Karjalainen, M. K. et al. A potential novel spontaneous preterm birth gene, AR, identified by linkage and association analysis of X chromosomal markers. PLoS One 7, e51378 (2012).
Peavey, M. C. et al. Progesterone receptor isoform B regulates the Oxtr-Plcl2-Trpc3 pathway to suppress uterine contractility. Proc. Natl. Acad. Sci. U. S. A. https://doi.org/10.1073/pnas.2011643118 (2021).
Ehn, N. L. et al. Evaluation of fetal and maternal genetic variation in the progesterone receptor gene for contributions to preterm birth. Pediatr. Res. 62, 630–635 (2007).
Mesiano, S. et al. Progesterone withdrawal and estrogen activation in human parturition are coordinated by progesterone receptor A expression in the myometrium. J. Clin. Endocrinol. Metab. 87, 2924–2930 (2002).
Blencowe, H. et al. National, regional, and worldwide estimates of preterm birth rates in the year 2010 with time trends since 1990 for selected countries: A systematic analysis and implications. Lancet 379, 2162–2172 (2012).
Hallman, M. et al. Spontaneous premature birth as a target of genomic research. Pediatr. Res. 85, 422–431 (2019).
Chang, A. et al. Alteration of heat shock protein 70 expression levels in term and preterm delivery. J. Matern. Fetal. Neonatal Med. 26, 1581–1585 (2013).
Fukushima, A. et al. Changes in serum levels of heat shock protein 70 in preterm delivery and pre-eclampsia. J. Obstet. Gynaecol. Res. 31, 72–77 (2005).
Bloshchinskaya, I. A. & Davidovich, I. M. Nitric oxide and HSP70 proteins during normal pregnancy, gestosis, and preclinical gestosis. Bull. Exp. Biol. Med. 135, 241–243 (2003).
Pockley, A. G., Bulmer, J., Hanks, B. M. & Wright, B. H. Identification of human heat shock protein 60 (Hsp60) and anti-Hsp60 antibodies in the peripheral circulation of normal individuals. Cell Stress Chaperones 4, 29–35 (1999).
Lewthwaite, J., Owen, N., Coates, A., Henderson, B. & Steptoe, A. Circulating human heat shock protein 60 in the plasma of British civil servants: Relationship to physiological and psychosocial stress. Circulation 106, 196–201 (2002).
Molvarec, A. et al. Serum heat shock protein 70 levels are decreased in normal human pregnancy. J. Reprod. Immunol. 74, 163–169 (2007).
Monreal-Flores, J., Espinosa-Garcia, M. T., Garcia-Regalado, A., Arechavaleta-Velasco, F. & Martinez, F. The heat shock protein 60 promotes progesterone synthesis in mitochondria of JEG-3 cells. Reprod. Biol. 17, 154–161 (2017).
Mendelson, C. R., Gao, L. & Montalbano, A. P. Multifactorial Regulation of Myometrial Contractility During Pregnancy and Parturition. Front. Endocrinol. (Lausanne) 10, 714 (2019).
Makieva, S., Saunders, P. T. & Norman, J. E. Androgens in pregnancy: Roles in parturition. Hum. Reprod. Update 20, 542–559 (2014).
Aruna, M. et al. Role of androgen receptor CAG repeat polymorphism and X-inactivation in the manifestation of recurrent spontaneous abortions in Indian women. PLoS One 6, e17718 (2011).
Hsu, T. Y. et al. Expression of androgen receptor in human placentas from normal and preeclamptic pregnancies. Taiwan J. Obstet. Gynecol. 48, 262–267 (2009).
Lo, B. C. et al. The orphan nuclear receptor RORalpha and group 3 innate lymphoid cells drive fibrosis in a mouse model of Crohn’s disease. Sci. Immunol. 1, eaaf8864 (2016).
Qiu, C. et al. Circadian clock-related genetic risk scores and risk of placental abruption. Placenta 36, 1480–1486 (2015).
Qiu, C. et al. Placental genetic variations in circadian clock-related genes increase the risk of placental abruption. Int. J. Mol. Epidemiol. Genet. 7, 32–40 (2016).
Dhamad, A. E., Zhou, Z., Zhou, J. & Du, Y. Systematic Proteomic Identification of the Heat Shock Proteins (Hsp) that Interact with Estrogen Receptor Alpha (ERalpha) and Biochemical Characterization of the ERalpha-Hsp70 Interaction. PLoS One 11, e0160312 (2016).
Anamthathmakula, P. et al. Estrogen receptor alpha isoform ERdelta7 in myometrium modulates uterine quiescence during pregnancy. EBioMedicine 39, 520–530 (2019).
Lang, S. et al. Different effects of Sec61alpha, Sec62 and Sec63 depletion on transport of polypeptides into the endoplasmic reticulum of mammalian cells. J. Cell. Sci. 125, 1958–1969 (2012).
Su, R. W. et al. The uterine expression of SEC63 gene is up-regulated at implantation sites in association with the decidualization during the early pregnancy in mice. Reprod. Biol. Endocrinol. 7, 12–12 (2009).
Morani, F. et al. Functional Transcriptome Analysis in ARSACS KO Cell Model Reveals a Role of Sacsin in Autophagy. Sci. Rep. 9, 11878–x (2019).
Gentil, B. J. et al. Sacsin, mutated in the ataxia ARSACS, regulates intermediate filament assembly and dynamics. FASEB J. 33, 2982–2994 (2019).
Zurawel, A. A. et al. CAG Expansions Are Genetically Stable and Form Nontoxic Aggregates in Cells Lacking Endogenous Polyglutamine Proteins. mBio. https://doi.org/10.1128/mBio.01367-16 (2016).
Fisher, J. J. et al. Proteomic analysis of placental mitochondria following trophoblast differentiation. Front. Physiol. 10, 1536 (2019).
Haataja, R. et al. Mapping a new spontaneous preterm birth susceptibility gene, IGF1R, using linkage, haplotype sharing, and association analysis. PLoS Genet. 7, e1001293 (2011).
McKenna, A. et al. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 20, 1297–1303 (2010).
Saunders, C. J. et al. Rapid whole-genome sequencing for genetic disease diagnosis in neonatal intensive care units. Sci. Transl. Med. 4, 154ra135 (2012).
Karjalainen, M. K. et al. CXCR3 polymorphism and expression associate with spontaneous preterm birth. J. Immunol. 195, 2187–2198 (2015).
Didion, J. P., Martin, M. & Collins, F. S. Atropos: Specific, sensitive, and speedy trimming of sequencing reads. PeerJ 5, e3720 (2017).
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: A Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140 (2010).
Ritchie, M. E. et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).
Acknowledgements
We express our sincere thanks to all of the families who participated in this study. We would like to thank Maarit Haarala of the University of Oulu for sample preparation and laboratory work. Sonja Eeli, Miia Lehto, and Riitta Vikeväinen of Oulu University Hospital are acknowledged for sample and data collection. We would like to offer our sincere thanks to Emily G. Farrow and Neil A. Miller from the Center for Pediatric Genomic Medicine, Children's Mercy Hospital, for their invaluable work with the Finnish exomes. Frans L. Bødker from the Institute of Public Health, Denmark, as well as Bruce Bedell, Patrick Breheny, and Noah W. Brown from the University of Iowa are thanked for their contributions to sample collection and data curation of the Danish exomes. We also thank the participants in the Finnish birth cohort (FIN), The Norwegian Mother, Father and Child Cohort Study (MoBa), and Danish National Birth Cohort (DNBC). The Norwegian Mother, Father and Child Cohort Study is supported by the Norwegian Ministry of Health and Care Services and the Ministry of Education and Research. We are grateful to all the participating families in Norway who take part in this on-going cohort study. The GWAS, WES, and transcriptomics studies that used specimens donated by families from the Northern Finland were supported by the grants from the Jane and Aatos Erkko Foundation (MH, MR), Foundation for Pediatric Resarch (MR), the Sigrid Jusélius Foundation (MH) and Competitive State Research Financing of the Expert Responsibility Area of Oulu University Hospital (MR). GZ is supported by grants from NIH (1R01HD101669-01A1), March of Dimes (22-FY17-889), and the Burroughs Wellcome Fund (10172896). GWAS data for the preterm birth project within the Norwegian Mother and Child Cohort Study were supported by the Norwegian Ministry of Health and Care Services and the Ministry of Education and Research, NIH/NIEHS (contract no. N01-ES-75558), and NIH/NINDS (grant no. 1 UO1 NS 047537-01 and grant no. 2 UO1 NS 047537-06A1). GWAS data for the preterm birth project DNBC samples were generated within the GENEVA consortium with funding provided through the NIH Genes, Environment, and Health Initiative (GEI, preterm birth: U01HG004423, dbGaP accession number phs000103.v1.p1). The GENEVA Coordinating Center (U01HG004446) provided assistance with genotype cleaning and general study coordination. We thank the research participants and employees of 23andMe for making this work possible, including the following members of the 23andMe Research Team: Michelle Agee, Babak Alipanahi, Adam Auton, Robert K. Bell, Katarzyna Bryc, Sarah L. Elson, Pierre Fontanillas, Nicholas A. Furlotte, David A. Hinds, Bethann S. Hromatka, Youna Hu, Karen E. Huber, Pan-Pan Jiang, Aaron Kleinman, Nadia K. Litterman, Matthew H. McIntyre, Joanna L. Mountain, Carrie A.M. Northover, Steven J. Pitts, Laura Russell, J. Fah Sathirapongsasuti, Olga V. Sazonova, Janie F. Shelton, Suyash Shringarpure, Chao Tian, Joyce Y. Tung, Vladimir Vacic, and Catherine H. Wilson. We thank the Finnish Functional Genomics Centre supported by the University of Turku, Åbo Akademi University, and Biocenter Finland, and the Medical Bioinformatics Centre of Turku Bioscience Centre for the sequencing data analysis. The Medical Bioinformatics Centre is supported by the University of Turku, Åbo Akademi University, Biocenter Finland, and Elixir-Finland.
Author information
Authors and Affiliations
Contributions
Conceptualization: J.M.H., A.M.H., S.F.K., M.H., L.J.M., M.R. Formal Analysis: J.M.H., H.T., A.M.H. Funding Acquisition: J.M.H., M.H., L.J.M., M.R. Investigation: J.M.H., H.T., A.M.H., A.P., P.T., M.K.K., G.Z. Methodology: J.M.H., H.T., A.M.H. Project Administration: J.M.H., A.M.H. Resources: G.Z., K.C., B.J., K.K.R., J.C.M., S.F.K., M.H., L.J.M., M.R. Supervision: B.J., J.C.M., S.F.K., M.H., L.J.M., M.R. Validation: J.M.H., H.T., A.M.H., A.P., M.K.K. Visualization: H.T. Writing – Original Draft: J.M.H., H.T., A.M.H. Writing – Review & Editing: J.M.H., H.T., A.M.H., A.P., P.T., M.K.K., G.Z., K.C., K.K.R., B.J., J.C.M., S.F.K., M.H., L.J.M., M.R.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Huusko, J.M., Tiensuu, H., Haapalainen, A.M. et al. Integrative genetic, genomic and transcriptomic analysis of heat shock protein and nuclear hormone receptor gene associations with spontaneous preterm birth. Sci Rep 11, 17115 (2021). https://doi.org/10.1038/s41598-021-96374-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-021-96374-9
This article is cited by
-
Elevated human placental heat shock protein 5 is associated with spontaneous preterm birth
Pediatric Research (2023)
-
The Role of Genetics in Preterm Birth
Reproductive Sciences (2023)
-
Human placental proteomics and exon variant studies link AAT/SERPINA1 with spontaneous preterm birth
BMC Medicine (2022)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
