- Email: [email protected]
Can microRNA profiling in maternal blood identify women at risk for preterm birth?
Research
ajog.org
OBSTETRICS
Can microRNA profiling in maternal blood identify women at risk for preterm birth? Michal A. Elovitz, MD; Lauren Anton, PhD; Jamie Bastek, MD, MSCE; Amy G. Brown, PhD OBJECTIVE: MicroRNAs (miRNAs), which are highly conserved singlestranded noncoding RNAs that play a crucial role in gene regulation, have now been identified as important players in many diseases states. MiRNAs have also been demonstrated to be reliable and useful biomarkers to identify those women who are at risk for specific adverse outcomes. The objective of this study was to determine whether miRNA profiles in maternal blood are different in women who are destined to have a preterm, compared with a term, birth.
were analyzed with the significance of analysis of microarrays and principle components analyses. A false discovery rate of 20% was used to determine the most differentially expressed miRNAs.
STUDY DESIGN: A nested case-control study was performed with maternal serum that was collected as part of a larger prospective cohort. MiRNA expression profiles in maternal serum were compared between women who ultimately had a preterm birth (n ¼ 40) compared with term birth (n ¼ 40). MiRNA expression profiles were created with the use of the Affymetrix GeneChip miRNA Array. The data
CONCLUSION: MiRNA profiles in maternal blood were not significantly
RESULTS: Of the 5640 miRNAs that were analyzed on the array, 4
miRNAs were significantly different between cases and control subjects. Two of the 4 miRNAs were mature miRNAs. The fold difference in expression was <2 for all 4 miRNAs. different in women who were destined to have a preterm, compared with a term, birth. MiRNAs in maternal blood are unlikely to become clinically useful biomarkers for the prediction of preterm birth. Key words: biomarker, microRNA, preterm birth
Cite this article as: Elovitz MA, Anton L, Bastek J, et al. Can microRNA profiling in maternal blood identify women at risk for preterm birth? Am J Obstet Gynecol 2015;212:782.e1-5.
P
reterm birth remains the leading cause of childhood morbidity and death. Although there have been some clinical trials that have demonstrated a reduction in the preterm birth rate by targeting women at high risk, preventative or interventional strategies that significantly reduce the national and international preterm birth rate have not
From the Maternal and Child Health Research Program, Department of Obstetrics and Gynecology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA. Received Aug. 22, 2014; revised Dec. 15, 2014; accepted Jan. 19, 2015. Supported by the Penn Presbyterian George L. and Emily McMichael Harrison Fund for Research in Obstetrics and Gynecology and the March of Dimes Prematurity Initiative Research Award (both to M.A.E.). The authors report no conflict of interest. Corresponding author: Michal A. Elovitz, MD. [email protected] 0002-9378/$36.00 ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2015.01.023
yet been realized.1-4 Although women with a previous preterm birth are at high risk for recurrent spontaneous preterm birth, most preterm births occur in women without this history. As such, accurate identification of those women who are at risk for preterm birth, weeks if not months, before the clinical event, would allow for improved use of resources (use of steroids, transfer to appropriate hospital facilities) as well as the potential for targeting novel therapeutics to prevent preterm birth. Biomarkers have been used extensively in the medical field with notable success. The use of markers such as highsensitivity c-reactive protein or prostate specific antigen have led to risk stratification algorithms for treatment options in the cardiovascular and cancer fields, respectively. Biomarker use in the field of preterm birth has been less successful. Although some studies demonstrate the utility of some biomarkers, most studies have revealed poor test characteristics for most proposed biomarkers, thus limiting any clinical utility.5 Based on
782.e1 American Journal of Obstetrics & Gynecology JUNE 2015
available evidence that has demonstrated an association between inflammatory processes and preterm birth, much of the biomarker work in preterm birth has focused on inflammatory mediators. Reviews of these works report insufficient evidence for the use of any of these biomarkers in predicting preterm birth.6,7 MicroRNAs (miRNA) may provide a new opportunity for biomarker discovery in the field of preterm birth. In the last decade, miRNA biology has emerged as an important player in both physiologic and pathophysiologic responses.8-11 MiRNAs are now implicated in varied disease states that include cancer and cardiovascular disease and are considered to be important therapeutic targets.12-16 MiRNAs can be released into the circulation and have been targeted as potential biomarkers for diverse disease states. Our objective for this study was to determine whether there was a distinct miRNA profile in maternal blood collected from women who were
ajog.org
Obstetrics
TABLE 1
Demographics on the nested case-control study Demographic
Full term birth (n [ 40)
Preterm birth (n [ 40)
Age, ya
24.06 (21.05e28.85)
24.04 (21.35e31.00)
P value .89b c
Black race, n (%)
28 (70.00)
35 (87.50)
.056
Previous preterm birth, n (%)
2 (5.00)
14 (35.00)
.003c
Gestational age at delivery, wksa
39.9 (39.6e40)
27.3 (25.75e29.35)
Gestational age at blood draw, wksa
26.75 (25.25e28.45)
26.35 (24.4e28)
a
< .001b .23b
Data are presented as median (interquartile range); b Calculated by nonparametric comparison of medians; c Calculated by c2.
Elovitz. MiRNA profile in maternal blood and preterm birth. Am J Obstet Gynecol 2015.
destined to have a preterm birth compared with a term birth. For this study, we performed a nested casecontrol within a prospective cohort of women at high risk for preterm birth.
M ATERIALS
AND
M ETHODS
For this discovery work, we performed a nested case-control study with patients who were enrolled as part of prospective cohort study (The Biomarker Study, clinical trials NCT01148654) at the Hospital of the University of Pennsylvania.17,18 The study was approved (protocol # 807678) by the institutional review board. The parent study was a prospective cohort study that consisted of women with singleton pregnancies from 22-336/7 weeks’ gestational age who presented to the labor and delivery triage unit with complaints concerning preterm labor. Patients were excluded for multiple-gestation, major fetal anomaly, intrauterine fetal death, severe preeclampsia before enrollment, chronic steroid or immunosuppressive drug use, active immunologic disease, acute systemic febrile illness, and/or pregestational diabetes mellitus. Patients who were either not delivered at our institution or whose infants were transferred to a different hospital for care were also excluded from these analyses. Over the study period, 1067 women were enrolled. Of those women, 39% had a preterm birth at <37 weeks’ gestation.
Patients were enrolled in the study by trained clinical research coordinators who obtained informed consent at the time of enrollment. Patients were enrolled at any point during their assessment or admission for preterm labor. Therefore, some patients were enrolled at the time of admission to the hospital; others might have been enrolled after being admitted for treatment of preterm labor. For this study, 57% of women received betamethasone for fetal benefit. Considering enrollment into the study could occur at any point from presentation through admission, the timing of the maternal blood draw to betamethasone administration varied about the cases and control subjects. Once a patient was enrolled in the study, all management decisions were made by the treating physician according to the standard of care at our institution. Women were enrolled from April 2009 through March 2012. After informed consent was given, maternal blood was obtained by routine venipuncture. Specimens were processed to obtain serum. Serum samples were aliquoted and then immediately frozen. Samples were stored at e80 C until use. For the study described herein, we identified all women who had a preterm birth at <37 weeks’ gestation. From that group of women, we further identified women who delivered at <30 weeks’ gestation to provide a more homogenous phenotype of preterm birth. Through detailed chart review, we confirmed that
Research
these preterm births were spontaneous and that gestational age at delivery was <30 weeks. Forty control subjects were then selected at random.
Data analyses Pearson c2 analyses or Fisher exact, as appropriate, were used to determine associations between categoric demographic variables and preterm birth. Nonparametric comparisons that included Wilcoxon-Rank Sum tests were performed to assess associations between continuous variables and preterm birth. A probability value < .05 was considered statistically significant. MiRNA profiling Total RNA isolation Total RNA, which included miRNA and other small RNAs, was isolated from human serum by phenol/chloroform extraction followed by column-based purification according to the manufacturer’s instructions (miRNeasy Serum/ plasma kit; Qiagen, Valencia, CA). The RNA was eluted in 14 mL of RNase-free water. The RNA concentration was then determined with a spectrophotometer (Nanadrop 2000 Spectrophotometer; Nanodrop, Rockland, DE) to understand potential limitations of quantifying RNA in this manner from biofluids. RNA integrity was not measured because these samples are composed of extracellular RNA. The isolated RNA was then used for RNA profiling with the use of a miRNA array. MiRNA array methods and analysis MicroRNA profiling was performed on term (n ¼ 40) and spontaneous preterm birth samples (n ¼ 40). Total RNA (400 ng) was 30 labeled with the use of the FlashTag Biotin HSR RNA Labeling Kit, per the manufacturer’s protocol (Affymetrix, Santa Clara, CA). Samples were then hybridized onto Genechip miRNA 3.0 arrays for 18 hours (Affymetrix). Hybridization images were scanned and digitized with the Genechip Scanner 3000 (Affymetrix). The normalized signal intensity was log2 transformed, and data analysis was performed with Partek Genomic Suite software (version 6.6; Copywrite, Partek Inc, St. Louis MO).
JUNE 2015 American Journal of Obstetrics & Gynecology
782.e2
Research
ajog.org
Obstetrics
Global sample variation was assessed by principle components analysis (PCA). Statistical analysis of microarrays (http:// www-stat.stanford.edu/wtibs/SAM/) was used to analyze the data and generate false discovery rates (FDRs) for individual microRNAs. After using SAM, an FDR <20% was selected. MicroRNAs with a >2-fold expression change and FDR <20% were considered significantly altered.
FIGURE
Principle components analyses plot for the miRNA results
R ESULTS Clinical demographics Maternal age and race were not significantly different between the cases and control subjects (Table 1). As expected with our study design, gestational age at delivery was significantly different in cases and control subjects. As might be anticipated, more cases had a history of spontaneous preterm birth than the control subjects. The gestational age of blood draw was not significantly different between cases and control subjects. MiRNA array The results of the PCA plot reveal that there is no clustering of women based on the array results (Figure). Of the 5640 miRNAs that were analyzed on the array, 4 microRNAs (stem-loop or mature) were altered between the preterm and term birth patients (Table 2). MiR-200a* and miR-4695-5P were mature miRNA species, whereas miR665 and miR-887 were detected in their stem-loop structure, which suggests that these last 2 species were not in their active form. All 4 targets had a <2-fold change in expression between preterm and term delivering women. All other targets had an FDR of 65% that suggests that there is no significant difference in expression between cases and control subjects.
C OMMENT Although the importance of miRNA biology to health and disease is being realized across disciplines, there is a paucity of research investigating the potential role of miRNAs in regulating the molecular mechanisms that is contributing to the pathophysiologic condition
The blue dots represent women who experienced a term birth; the red dots represent women who had a preterm birth. PCA, principle components analyses. Elovitz. MiRNA profile in maternal blood and preterm birth. Am J Obstet Gynecol 2015.
of preterm birth; no data, to date, has explored the ability of miRNAs to serve reliable biomarkers for predicting preterm birth. Novel to this work, we have demonstrated that miRNA profiles in maternal serum are not significantly different in women who are destined to have a preterm delivery compared with a term birth. For future clinical implications, we propose that the PCA results demonstrate a lack of clinical utility in pursuing miRNA profiling as a serum biomarker for preterm birth prediction. It is possible that future miRNAs might be discovered that are not included on the array that was used for this study. However, based on current miRNA knowledge, the array that was used in this study is very broad and encompassing. Likewise, there may
782.e3 American Journal of Obstetrics & Gynecology JUNE 2015
be cell-specific miRNAs (as opposed to circulating miRNAs) that may be useful in the prediction of preterm birth. As exosome technology advances, it may be possible to pursue the investigation of exosome encapsulated miRNAs and their role in contributing to spontaneous preterm birth. However, considering our current technology and understanding of miRNA biology, our data suggest that there are not systemic changes in the miRNA profile in the maternal circulation that could serve as reliable biomarkers for the prediction of preterm birth. Of the 4 miRNAs found to be expressed differentially, 2 are mature miRNAs (miR-4695 and miR-200a). Very little is known about the downstream targets and hence function of miR-4695. Conversely, miR-200a
ajog.org
Obstetrics
TABLE 2
MicroRNA array results MicroRNA
Sequence type
Fold-changea
P value
mIR-200a
miRNA
1.16
.0001
mIR-4695-5P
miRNA
1.61
.002
mIR-665
Stem-loop
1.14
.001
mIR-887
Stem-loop
1.11
.0004
a
Fold change in miRNA expression between patients that did and did not have a preterm birth.
Elovitz. MiRNA profile in maternal blood and preterm birth. Am J Obstet Gynecol 2015.
is actually a fairly well-researched epithelial-specific miRNA.19,20 MiR200a has been found to be elevated in the serum of patients with either prostate or ovarian cancers.21,22 This research suggests that an increased tissue-specific production of miR-200 has the potential to be released into the maternal circulation. The question for our study was whether the observed small, but significant fold, decrease in miR-200a is a biologically relevant finding. Previous work in mice has demonstrated that miR-200a is involved in myometrial activity. Specifically, miR-200a alters myometrial activity by being a regulator of progesterone metabolism.23 As is consistent with miRNA biology, an increase in miRNA expression usually results in the repression of a network of specific target genes. As such, it might be expected that an increased production of miR-200a in the uterus before preterm birth would lead to an increased presence of in maternal serum. Notably, we found the opposite result with a significant decrease in serum miR-200a in women who were destined to have a preterm birth. However, because we assessed systemic miR-200a, not uterine miR-200a, and the miR200 family has pleotropic effects, it is possible that the biologic relevance of this finding is not related to the effects of miR-200a in uterine tissue. We also believe that it is important to note that the fold change in miR-200a expression as observed in this study is very small. Consequently, we do not propose that miR-200a would be able to serve as a clinically relevant biomarker of preterm birth. Unlike the small changes found in miR-200a in maternal serum, we
recently have demonstrated that there are large changes in specific miRNAs that were isolated from cervical cells that were collected from women who were destined to have a preterm birth. The findings of the current study are in contrast to our recent cervical cell work in which the PCA plot revealed distinct miRNA profiles for women who were destined to deliver preterm.24 Notably, in that study, 99 miRNAs were significantly different weeks before a preterm birth.24 That previous work suggested that diverse changes in the miRNA profile in the cervix may be involved in premature cervical remodeling. Because miRNAs can be secreted and/or released into the circulation, it might be argued that some of the altered miRNAs that were observed in that study might have been differentially expressed in the maternal serum. However, we did not observe this finding. Because we found differences in the miRNA profile in cervical cells and not from a core biopsy of the cervix, it is possible, if not more likely, that the miRNA differences would not be detectable systemically. Taken together, these 2 studies suggest that preterm birth may be a “local” disease with molecular and cellular changes at the level of the reproductive tissues. Furthermore, our finding of miRNA differences in the cervix, but not in maternal serum, would suggest that future biomarker work should focus locally on the vagina, cervix, and uterus if we hope to find clinically relevant biomarkers. This study has notable strengths. We used a cohort of well-phenotyped,
Research
prospectively collected patients, thus limiting many biases. Our cases were all documented to have spontaneous preterm birth, thus preventing misclassification bias that might occur with such case-control studies. Our discovery approach was strong with the 40 women who were included in each group. This study was limited in that it explored miRNA profiling in women who were symptomatic with preterm labor. Although many of the women ultimately have a term delivery, perhaps there is something about preterm labor itself that alters the miRNA profile and, hence, dilutes out a significant finding. Noting this possibility, if there are large differences in the miRNA profile of women who are destined to have a preterm birth, compared with a term birth, it might be expected to be evident in this study. Another potential bias for this study would be whether the administration of steroid (betamethasone) was able to limit miRNA expression in the women who ultimately had a preterm birth; hence, the study would be biased to the null. Although this is a possibility, several other studies have failed to find any significant effect of the use of betamethasone (in the doses used clinically) on maternal immune or hormonal responses. Therefore, we believe that the administration of steroids to some of the women did not significantly alter our findings. In summary, these findings suggest that further investigation of miRNA profiling in maternal serum is unlikely to advance our understanding of preterm birth and/or serve as clinically relevant biomarkers of this disease. In contrast, previous work by us and others argues that exploring the role of local miRNAs in reproductive tissues might serve to significantly advance our understanding of preterm birth. -
REFERENCES 1. Meis PJ, Klebanoff M, Thom M, et al. Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate. N Engl J Med 2003;348:2379-85. 2. Fonseca EB, Celik E, Parra M, Singh M, Nicolaides KH. Progesterone and the risk of
JUNE 2015 American Journal of Obstetrics & Gynecology
782.e4
Research
ajog.org
Obstetrics
preterm birth among women with a short cervix. N Engl J Med 2007;357:462-9. 3. Fonseca EB, Bittar RE, Carvalho MH, Zugaib M. Prophylactic administration of progesterone by vaginal suppository to reduce the incidence of spontaneous preterm birth in women at increased risk: a randomized placebo-controlled double-blind study. Am J Obstet Gynecol 2003;188: 419-24. 4. Owen J, Hankins G, Iams JD, et al. Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length. Am J Obstet Gynecol 2009;201:375.e1-8. 5. Bastek JA, Elovitz MA. The role and challenges of biomarkers in spontaneous preterm birth and preeclampsia. Fertil Steril 2013;99: 1117-23. 6. Wei SQ, Fraser W, Luo ZC. Inflammatory cytokines and spontaneous preterm birth in asymptomatic women: a systematic review. Obstet Gynecol 2010;116:393-401. 7. Conde-Agudelo A, Papageorghiou AT, Kennedy SH, Villar J. Novel biomarkers for the prediction of the spontaneous preterm birth phenotype: a systematic review and metaanalysis. BJOG 2011;118:1042-54. 8. Eacker SM, Dawson TM, Dawson VL. The interplay of microRNA and neuronal activity in health and disease. Front Cell Neurosci 2013;7: 136.
9. Farazi TA, Hoell JI, Morozov P, Tuschl T. MicroRNAs in human cancer. Adv Exp Med Biol 2013;774:1-20. 10. Hagiwara S, Kantharidis P, Cooper ME. MicroRNA as biomarkers and regulator of cardiovascular development and disease. Curr Pharm Des 2014:2347-70. 11. Hulsmans M, Holvoet P. MicroRNAs as early biomarkers in obesity and related metabolic and cardiovascular diseases. Curr Pharm Des 2013;19:5704-17. 12. Angulo M, Lecuona E, Sznajder JI. Role of microRNAs in lung disease. Arch Bronconeumol 2012;48:325-30. 13. Abdellatif M. Differential expression of microRNAs in different disease states. Circ Res 2012;110:638-50. 14. Bratkovic T, Glavan G, Strukelj B, Zivin M, Rogelj B. Exploiting microRNAs for cell engineering and therapy. Biotechnol Adv 2012;30:753-65. 15. Fiedler J, Gupta SK, Thum T. MicroRNAbased therapeutic approaches in the cardiovascular system. Cardiovasc Ther 2012;30:e9. 16. Gambari R, Fabbri E, Borgatti M, et al. Targeting microRNAs involved in human diseases: a novel approach for modification of gene expression and drug development. Biochem Pharmacol 2011;82:1416-29. 17. Bastek JA, Weber AL, McShea MA, Ryan ME, Elovitz MA. Prenatal inflammation is associated with adverse neonatal outcomes. Am J Obstet Gynecol 2014;210:450.e1-10.
782.e5 American Journal of Obstetrics & Gynecology JUNE 2015
18. Bastek JA, Sammel MD, Srinivas SK, et al. Clinical prediction rules for preterm birth in patients presenting with preterm labor. Obstet Gynecol 2012;119:1119-28. du E, Katchy A, Tsouko E, et al. 19. Aydog MicroRNA-regulated gene networks during mammary cell differentiation are associated with breast cancer. Carcinogenesis 2012;33: 1502-11. 20. Cao Q, Lu K, Dai S, Hu Y, Fan W. Clinicopathological and prognostic implications of the miR-200 family in patients with epithelial ovarian cancer. Int J Clin Exp Pathol 2014;7:2392-401. 21. Kan CW, Hahn MA, Gard GB, et al. Elevated levels of circulating microRNA-200 family members correlate with serous epithelial ovarian cancer. BMC Cancer 2012;12:627. 22. Cheng HH, Mitchell PS, Kroh Em, et al. Circulating microRNA profiling identifies a subset of metastatic prostate cancer patients with evidence of cancer-associated hypoxia. PloS One 2013;8:e69239. 23. Williams KC, Renthal NE, Condon JC, Gerard RD, Mendelson CR. MicroRNA-200a serves a key role in the decline of progesterone receptor function leading to term and preterm labor. Proc Natl Acad Sciences U S A 2012;109: 7529-34. 24. Elovitz MA, Brown AG, Anton L, et al. Distinct cervical microRNA profiles are present in women destined to have a preterm birth. Am J Obstet Gynecol 2014;210:221.e1-11.
ajog.org
OBSTETRICS
Can microRNA profiling in maternal blood identify women at risk for preterm birth? Michal A. Elovitz, MD; Lauren Anton, PhD; Jamie Bastek, MD, MSCE; Amy G. Brown, PhD OBJECTIVE: MicroRNAs (miRNAs), which are highly conserved singlestranded noncoding RNAs that play a crucial role in gene regulation, have now been identified as important players in many diseases states. MiRNAs have also been demonstrated to be reliable and useful biomarkers to identify those women who are at risk for specific adverse outcomes. The objective of this study was to determine whether miRNA profiles in maternal blood are different in women who are destined to have a preterm, compared with a term, birth.
were analyzed with the significance of analysis of microarrays and principle components analyses. A false discovery rate of 20% was used to determine the most differentially expressed miRNAs.
STUDY DESIGN: A nested case-control study was performed with maternal serum that was collected as part of a larger prospective cohort. MiRNA expression profiles in maternal serum were compared between women who ultimately had a preterm birth (n ¼ 40) compared with term birth (n ¼ 40). MiRNA expression profiles were created with the use of the Affymetrix GeneChip miRNA Array. The data
CONCLUSION: MiRNA profiles in maternal blood were not significantly
RESULTS: Of the 5640 miRNAs that were analyzed on the array, 4
miRNAs were significantly different between cases and control subjects. Two of the 4 miRNAs were mature miRNAs. The fold difference in expression was <2 for all 4 miRNAs. different in women who were destined to have a preterm, compared with a term, birth. MiRNAs in maternal blood are unlikely to become clinically useful biomarkers for the prediction of preterm birth. Key words: biomarker, microRNA, preterm birth
Cite this article as: Elovitz MA, Anton L, Bastek J, et al. Can microRNA profiling in maternal blood identify women at risk for preterm birth? Am J Obstet Gynecol 2015;212:782.e1-5.
P
reterm birth remains the leading cause of childhood morbidity and death. Although there have been some clinical trials that have demonstrated a reduction in the preterm birth rate by targeting women at high risk, preventative or interventional strategies that significantly reduce the national and international preterm birth rate have not
From the Maternal and Child Health Research Program, Department of Obstetrics and Gynecology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA. Received Aug. 22, 2014; revised Dec. 15, 2014; accepted Jan. 19, 2015. Supported by the Penn Presbyterian George L. and Emily McMichael Harrison Fund for Research in Obstetrics and Gynecology and the March of Dimes Prematurity Initiative Research Award (both to M.A.E.). The authors report no conflict of interest. Corresponding author: Michal A. Elovitz, MD. [email protected] 0002-9378/$36.00 ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2015.01.023
yet been realized.1-4 Although women with a previous preterm birth are at high risk for recurrent spontaneous preterm birth, most preterm births occur in women without this history. As such, accurate identification of those women who are at risk for preterm birth, weeks if not months, before the clinical event, would allow for improved use of resources (use of steroids, transfer to appropriate hospital facilities) as well as the potential for targeting novel therapeutics to prevent preterm birth. Biomarkers have been used extensively in the medical field with notable success. The use of markers such as highsensitivity c-reactive protein or prostate specific antigen have led to risk stratification algorithms for treatment options in the cardiovascular and cancer fields, respectively. Biomarker use in the field of preterm birth has been less successful. Although some studies demonstrate the utility of some biomarkers, most studies have revealed poor test characteristics for most proposed biomarkers, thus limiting any clinical utility.5 Based on
782.e1 American Journal of Obstetrics & Gynecology JUNE 2015
available evidence that has demonstrated an association between inflammatory processes and preterm birth, much of the biomarker work in preterm birth has focused on inflammatory mediators. Reviews of these works report insufficient evidence for the use of any of these biomarkers in predicting preterm birth.6,7 MicroRNAs (miRNA) may provide a new opportunity for biomarker discovery in the field of preterm birth. In the last decade, miRNA biology has emerged as an important player in both physiologic and pathophysiologic responses.8-11 MiRNAs are now implicated in varied disease states that include cancer and cardiovascular disease and are considered to be important therapeutic targets.12-16 MiRNAs can be released into the circulation and have been targeted as potential biomarkers for diverse disease states. Our objective for this study was to determine whether there was a distinct miRNA profile in maternal blood collected from women who were
ajog.org
Obstetrics
TABLE 1
Demographics on the nested case-control study Demographic
Full term birth (n [ 40)
Preterm birth (n [ 40)
Age, ya
24.06 (21.05e28.85)
24.04 (21.35e31.00)
P value .89b c
Black race, n (%)
28 (70.00)
35 (87.50)
.056
Previous preterm birth, n (%)
2 (5.00)
14 (35.00)
.003c
Gestational age at delivery, wksa
39.9 (39.6e40)
27.3 (25.75e29.35)
Gestational age at blood draw, wksa
26.75 (25.25e28.45)
26.35 (24.4e28)
a
< .001b .23b
Data are presented as median (interquartile range); b Calculated by nonparametric comparison of medians; c Calculated by c2.
Elovitz. MiRNA profile in maternal blood and preterm birth. Am J Obstet Gynecol 2015.
destined to have a preterm birth compared with a term birth. For this study, we performed a nested casecontrol within a prospective cohort of women at high risk for preterm birth.
M ATERIALS
AND
M ETHODS
For this discovery work, we performed a nested case-control study with patients who were enrolled as part of prospective cohort study (The Biomarker Study, clinical trials NCT01148654) at the Hospital of the University of Pennsylvania.17,18 The study was approved (protocol # 807678) by the institutional review board. The parent study was a prospective cohort study that consisted of women with singleton pregnancies from 22-336/7 weeks’ gestational age who presented to the labor and delivery triage unit with complaints concerning preterm labor. Patients were excluded for multiple-gestation, major fetal anomaly, intrauterine fetal death, severe preeclampsia before enrollment, chronic steroid or immunosuppressive drug use, active immunologic disease, acute systemic febrile illness, and/or pregestational diabetes mellitus. Patients who were either not delivered at our institution or whose infants were transferred to a different hospital for care were also excluded from these analyses. Over the study period, 1067 women were enrolled. Of those women, 39% had a preterm birth at <37 weeks’ gestation.
Patients were enrolled in the study by trained clinical research coordinators who obtained informed consent at the time of enrollment. Patients were enrolled at any point during their assessment or admission for preterm labor. Therefore, some patients were enrolled at the time of admission to the hospital; others might have been enrolled after being admitted for treatment of preterm labor. For this study, 57% of women received betamethasone for fetal benefit. Considering enrollment into the study could occur at any point from presentation through admission, the timing of the maternal blood draw to betamethasone administration varied about the cases and control subjects. Once a patient was enrolled in the study, all management decisions were made by the treating physician according to the standard of care at our institution. Women were enrolled from April 2009 through March 2012. After informed consent was given, maternal blood was obtained by routine venipuncture. Specimens were processed to obtain serum. Serum samples were aliquoted and then immediately frozen. Samples were stored at e80 C until use. For the study described herein, we identified all women who had a preterm birth at <37 weeks’ gestation. From that group of women, we further identified women who delivered at <30 weeks’ gestation to provide a more homogenous phenotype of preterm birth. Through detailed chart review, we confirmed that
Research
these preterm births were spontaneous and that gestational age at delivery was <30 weeks. Forty control subjects were then selected at random.
Data analyses Pearson c2 analyses or Fisher exact, as appropriate, were used to determine associations between categoric demographic variables and preterm birth. Nonparametric comparisons that included Wilcoxon-Rank Sum tests were performed to assess associations between continuous variables and preterm birth. A probability value < .05 was considered statistically significant. MiRNA profiling Total RNA isolation Total RNA, which included miRNA and other small RNAs, was isolated from human serum by phenol/chloroform extraction followed by column-based purification according to the manufacturer’s instructions (miRNeasy Serum/ plasma kit; Qiagen, Valencia, CA). The RNA was eluted in 14 mL of RNase-free water. The RNA concentration was then determined with a spectrophotometer (Nanadrop 2000 Spectrophotometer; Nanodrop, Rockland, DE) to understand potential limitations of quantifying RNA in this manner from biofluids. RNA integrity was not measured because these samples are composed of extracellular RNA. The isolated RNA was then used for RNA profiling with the use of a miRNA array. MiRNA array methods and analysis MicroRNA profiling was performed on term (n ¼ 40) and spontaneous preterm birth samples (n ¼ 40). Total RNA (400 ng) was 30 labeled with the use of the FlashTag Biotin HSR RNA Labeling Kit, per the manufacturer’s protocol (Affymetrix, Santa Clara, CA). Samples were then hybridized onto Genechip miRNA 3.0 arrays for 18 hours (Affymetrix). Hybridization images were scanned and digitized with the Genechip Scanner 3000 (Affymetrix). The normalized signal intensity was log2 transformed, and data analysis was performed with Partek Genomic Suite software (version 6.6; Copywrite, Partek Inc, St. Louis MO).
JUNE 2015 American Journal of Obstetrics & Gynecology
782.e2
Research
ajog.org
Obstetrics
Global sample variation was assessed by principle components analysis (PCA). Statistical analysis of microarrays (http:// www-stat.stanford.edu/wtibs/SAM/) was used to analyze the data and generate false discovery rates (FDRs) for individual microRNAs. After using SAM, an FDR <20% was selected. MicroRNAs with a >2-fold expression change and FDR <20% were considered significantly altered.
FIGURE
Principle components analyses plot for the miRNA results
R ESULTS Clinical demographics Maternal age and race were not significantly different between the cases and control subjects (Table 1). As expected with our study design, gestational age at delivery was significantly different in cases and control subjects. As might be anticipated, more cases had a history of spontaneous preterm birth than the control subjects. The gestational age of blood draw was not significantly different between cases and control subjects. MiRNA array The results of the PCA plot reveal that there is no clustering of women based on the array results (Figure). Of the 5640 miRNAs that were analyzed on the array, 4 microRNAs (stem-loop or mature) were altered between the preterm and term birth patients (Table 2). MiR-200a* and miR-4695-5P were mature miRNA species, whereas miR665 and miR-887 were detected in their stem-loop structure, which suggests that these last 2 species were not in their active form. All 4 targets had a <2-fold change in expression between preterm and term delivering women. All other targets had an FDR of 65% that suggests that there is no significant difference in expression between cases and control subjects.
C OMMENT Although the importance of miRNA biology to health and disease is being realized across disciplines, there is a paucity of research investigating the potential role of miRNAs in regulating the molecular mechanisms that is contributing to the pathophysiologic condition
The blue dots represent women who experienced a term birth; the red dots represent women who had a preterm birth. PCA, principle components analyses. Elovitz. MiRNA profile in maternal blood and preterm birth. Am J Obstet Gynecol 2015.
of preterm birth; no data, to date, has explored the ability of miRNAs to serve reliable biomarkers for predicting preterm birth. Novel to this work, we have demonstrated that miRNA profiles in maternal serum are not significantly different in women who are destined to have a preterm delivery compared with a term birth. For future clinical implications, we propose that the PCA results demonstrate a lack of clinical utility in pursuing miRNA profiling as a serum biomarker for preterm birth prediction. It is possible that future miRNAs might be discovered that are not included on the array that was used for this study. However, based on current miRNA knowledge, the array that was used in this study is very broad and encompassing. Likewise, there may
782.e3 American Journal of Obstetrics & Gynecology JUNE 2015
be cell-specific miRNAs (as opposed to circulating miRNAs) that may be useful in the prediction of preterm birth. As exosome technology advances, it may be possible to pursue the investigation of exosome encapsulated miRNAs and their role in contributing to spontaneous preterm birth. However, considering our current technology and understanding of miRNA biology, our data suggest that there are not systemic changes in the miRNA profile in the maternal circulation that could serve as reliable biomarkers for the prediction of preterm birth. Of the 4 miRNAs found to be expressed differentially, 2 are mature miRNAs (miR-4695 and miR-200a). Very little is known about the downstream targets and hence function of miR-4695. Conversely, miR-200a
ajog.org
Obstetrics
TABLE 2
MicroRNA array results MicroRNA
Sequence type
Fold-changea
P value
mIR-200a
miRNA
1.16
.0001
mIR-4695-5P
miRNA
1.61
.002
mIR-665
Stem-loop
1.14
.001
mIR-887
Stem-loop
1.11
.0004
a
Fold change in miRNA expression between patients that did and did not have a preterm birth.
Elovitz. MiRNA profile in maternal blood and preterm birth. Am J Obstet Gynecol 2015.
is actually a fairly well-researched epithelial-specific miRNA.19,20 MiR200a has been found to be elevated in the serum of patients with either prostate or ovarian cancers.21,22 This research suggests that an increased tissue-specific production of miR-200 has the potential to be released into the maternal circulation. The question for our study was whether the observed small, but significant fold, decrease in miR-200a is a biologically relevant finding. Previous work in mice has demonstrated that miR-200a is involved in myometrial activity. Specifically, miR-200a alters myometrial activity by being a regulator of progesterone metabolism.23 As is consistent with miRNA biology, an increase in miRNA expression usually results in the repression of a network of specific target genes. As such, it might be expected that an increased production of miR-200a in the uterus before preterm birth would lead to an increased presence of in maternal serum. Notably, we found the opposite result with a significant decrease in serum miR-200a in women who were destined to have a preterm birth. However, because we assessed systemic miR-200a, not uterine miR-200a, and the miR200 family has pleotropic effects, it is possible that the biologic relevance of this finding is not related to the effects of miR-200a in uterine tissue. We also believe that it is important to note that the fold change in miR-200a expression as observed in this study is very small. Consequently, we do not propose that miR-200a would be able to serve as a clinically relevant biomarker of preterm birth. Unlike the small changes found in miR-200a in maternal serum, we
recently have demonstrated that there are large changes in specific miRNAs that were isolated from cervical cells that were collected from women who were destined to have a preterm birth. The findings of the current study are in contrast to our recent cervical cell work in which the PCA plot revealed distinct miRNA profiles for women who were destined to deliver preterm.24 Notably, in that study, 99 miRNAs were significantly different weeks before a preterm birth.24 That previous work suggested that diverse changes in the miRNA profile in the cervix may be involved in premature cervical remodeling. Because miRNAs can be secreted and/or released into the circulation, it might be argued that some of the altered miRNAs that were observed in that study might have been differentially expressed in the maternal serum. However, we did not observe this finding. Because we found differences in the miRNA profile in cervical cells and not from a core biopsy of the cervix, it is possible, if not more likely, that the miRNA differences would not be detectable systemically. Taken together, these 2 studies suggest that preterm birth may be a “local” disease with molecular and cellular changes at the level of the reproductive tissues. Furthermore, our finding of miRNA differences in the cervix, but not in maternal serum, would suggest that future biomarker work should focus locally on the vagina, cervix, and uterus if we hope to find clinically relevant biomarkers. This study has notable strengths. We used a cohort of well-phenotyped,
Research
prospectively collected patients, thus limiting many biases. Our cases were all documented to have spontaneous preterm birth, thus preventing misclassification bias that might occur with such case-control studies. Our discovery approach was strong with the 40 women who were included in each group. This study was limited in that it explored miRNA profiling in women who were symptomatic with preterm labor. Although many of the women ultimately have a term delivery, perhaps there is something about preterm labor itself that alters the miRNA profile and, hence, dilutes out a significant finding. Noting this possibility, if there are large differences in the miRNA profile of women who are destined to have a preterm birth, compared with a term birth, it might be expected to be evident in this study. Another potential bias for this study would be whether the administration of steroid (betamethasone) was able to limit miRNA expression in the women who ultimately had a preterm birth; hence, the study would be biased to the null. Although this is a possibility, several other studies have failed to find any significant effect of the use of betamethasone (in the doses used clinically) on maternal immune or hormonal responses. Therefore, we believe that the administration of steroids to some of the women did not significantly alter our findings. In summary, these findings suggest that further investigation of miRNA profiling in maternal serum is unlikely to advance our understanding of preterm birth and/or serve as clinically relevant biomarkers of this disease. In contrast, previous work by us and others argues that exploring the role of local miRNAs in reproductive tissues might serve to significantly advance our understanding of preterm birth. -
REFERENCES 1. Meis PJ, Klebanoff M, Thom M, et al. Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate. N Engl J Med 2003;348:2379-85. 2. Fonseca EB, Celik E, Parra M, Singh M, Nicolaides KH. Progesterone and the risk of
JUNE 2015 American Journal of Obstetrics & Gynecology
782.e4
Research
ajog.org
Obstetrics
preterm birth among women with a short cervix. N Engl J Med 2007;357:462-9. 3. Fonseca EB, Bittar RE, Carvalho MH, Zugaib M. Prophylactic administration of progesterone by vaginal suppository to reduce the incidence of spontaneous preterm birth in women at increased risk: a randomized placebo-controlled double-blind study. Am J Obstet Gynecol 2003;188: 419-24. 4. Owen J, Hankins G, Iams JD, et al. Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length. Am J Obstet Gynecol 2009;201:375.e1-8. 5. Bastek JA, Elovitz MA. The role and challenges of biomarkers in spontaneous preterm birth and preeclampsia. Fertil Steril 2013;99: 1117-23. 6. Wei SQ, Fraser W, Luo ZC. Inflammatory cytokines and spontaneous preterm birth in asymptomatic women: a systematic review. Obstet Gynecol 2010;116:393-401. 7. Conde-Agudelo A, Papageorghiou AT, Kennedy SH, Villar J. Novel biomarkers for the prediction of the spontaneous preterm birth phenotype: a systematic review and metaanalysis. BJOG 2011;118:1042-54. 8. Eacker SM, Dawson TM, Dawson VL. The interplay of microRNA and neuronal activity in health and disease. Front Cell Neurosci 2013;7: 136.
9. Farazi TA, Hoell JI, Morozov P, Tuschl T. MicroRNAs in human cancer. Adv Exp Med Biol 2013;774:1-20. 10. Hagiwara S, Kantharidis P, Cooper ME. MicroRNA as biomarkers and regulator of cardiovascular development and disease. Curr Pharm Des 2014:2347-70. 11. Hulsmans M, Holvoet P. MicroRNAs as early biomarkers in obesity and related metabolic and cardiovascular diseases. Curr Pharm Des 2013;19:5704-17. 12. Angulo M, Lecuona E, Sznajder JI. Role of microRNAs in lung disease. Arch Bronconeumol 2012;48:325-30. 13. Abdellatif M. Differential expression of microRNAs in different disease states. Circ Res 2012;110:638-50. 14. Bratkovic T, Glavan G, Strukelj B, Zivin M, Rogelj B. Exploiting microRNAs for cell engineering and therapy. Biotechnol Adv 2012;30:753-65. 15. Fiedler J, Gupta SK, Thum T. MicroRNAbased therapeutic approaches in the cardiovascular system. Cardiovasc Ther 2012;30:e9. 16. Gambari R, Fabbri E, Borgatti M, et al. Targeting microRNAs involved in human diseases: a novel approach for modification of gene expression and drug development. Biochem Pharmacol 2011;82:1416-29. 17. Bastek JA, Weber AL, McShea MA, Ryan ME, Elovitz MA. Prenatal inflammation is associated with adverse neonatal outcomes. Am J Obstet Gynecol 2014;210:450.e1-10.
782.e5 American Journal of Obstetrics & Gynecology JUNE 2015
18. Bastek JA, Sammel MD, Srinivas SK, et al. Clinical prediction rules for preterm birth in patients presenting with preterm labor. Obstet Gynecol 2012;119:1119-28. du E, Katchy A, Tsouko E, et al. 19. Aydog MicroRNA-regulated gene networks during mammary cell differentiation are associated with breast cancer. Carcinogenesis 2012;33: 1502-11. 20. Cao Q, Lu K, Dai S, Hu Y, Fan W. Clinicopathological and prognostic implications of the miR-200 family in patients with epithelial ovarian cancer. Int J Clin Exp Pathol 2014;7:2392-401. 21. Kan CW, Hahn MA, Gard GB, et al. Elevated levels of circulating microRNA-200 family members correlate with serous epithelial ovarian cancer. BMC Cancer 2012;12:627. 22. Cheng HH, Mitchell PS, Kroh Em, et al. Circulating microRNA profiling identifies a subset of metastatic prostate cancer patients with evidence of cancer-associated hypoxia. PloS One 2013;8:e69239. 23. Williams KC, Renthal NE, Condon JC, Gerard RD, Mendelson CR. MicroRNA-200a serves a key role in the decline of progesterone receptor function leading to term and preterm labor. Proc Natl Acad Sciences U S A 2012;109: 7529-34. 24. Elovitz MA, Brown AG, Anton L, et al. Distinct cervical microRNA profiles are present in women destined to have a preterm birth. Am J Obstet Gynecol 2014;210:221.e1-11.