Advertisement

Human embryos secrete microRNAs into culture media—a potential biomarker for implantation

      Objective

      To determine whether human blastocysts secrete microRNA (miRNAs) into culture media and whether these reflect embryonic ploidy status and can predict in vitro fertilization (IVF) outcomes.

      Design

      Experimental study of human embryos and IVF culture media.

      Setting

      Academic IVF program.

      Patient(s)

      91 donated, cryopreserved embryos that developed into 28 tested blastocysts, from 13 couples who had previously completed IVF cycles.

      Intervention(s)

      None.

      Main Outcome Measure(s)

      Relative miRNA expression in IVF culture media.

      Result(s)

      Blastocysts were assessed by chromosomal comparative genomic hybridization analysis, and the culture media from 55 single-embryo transfer cycles was tested for miRNA expression using an array-based quantitative real-time polymerase chain reaction analysis. The expression of the identified miRNA was correlated with pregnancy outcomes. Ten miRNA were identified in the culture media; two were specific to spent media (miR-191 and miR-372), and one was only present in media before the embryos had been cultured (miR-645). MicroRNA-191 was more highly concentrated in media from aneuploid embryos, and miR-191, miR-372, and miR-645 were more highly concentrated in media from failed IVF/non-intracytoplasmic sperm injection cycles. Additionally, miRNA were found to be more highly concentrated in ICSI and day-5 media samples when compared with regularly inseminated and day-4 samples, respectively.

      Conclusion(s)

      MicroRNA can be detected in IVF culture media. Some of these miRNA are differentially expressed according to the fertilization method, chromosomal status, and pregnancy outcome, which makes them potential biomarkers for predicting IVF success.

      Key Words

      Discuss: You can discuss this article with its authors and with other ASRM members at http://fertstertforum.com/rosenbluthem-human-embryos-micrornas-biomarker-implantation/
      MicroRNAs (miRNAs) are small (approximately 22 nucleotides) noncoding RNAs that regulate gene expression and have been implicated in a wide array of biologic processes including early embryo development and stem cell differentiation (
      • Laurent L.C.
      MicroRNAs in embryonic stem cells and early embryonic development.
      ). Recently, miRNAs have been found to be packaged into small vesicles called exosomes and subsequently secreted into the extracellular space (
      • Valadi H.
      • Ekström K.
      • Bossios A.
      • Sjöstrand M.
      • Lee J.J.
      • Lötvall J.O.
      Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells.
      ). Encapsulated miRNAs are protected from degradation and, consequently, can be detected after extended periods of time (
      • Jung M.
      • Schaefer A.
      • Steiner I.
      • Kempkensteffen C.
      • Stephan C.
      • Erbersdobler A.
      • et al.
      Robust MicroRNA stability in degraded RNA preparations from human tissue and cell samples.
      ). Although the role of exosomal miRNAs is still being elucidated, growing evidence suggests that packaged miRNAs can reach distant cells and affect gene expression (
      • Hergenreider E.
      • Heydt S.
      • Treguer K.
      • Boettger T.
      • Horrevoets A.J.G.
      • Zeiher A.M.
      • et al.
      Atheroprotective communication between endothelial cells and smooth muscle cells through miRNAs.
      ).
      Regardless of their physiologic role, distinct patterns of secreted miRNAs have been found to correlate with a variety of diseases including cancer (
      • Chen X.
      • Ba Y.
      • Ma L.
      • Cai X.
      • Yin Y.
      • Wang K.
      • et al.
      Characterization of microRNAs in serum: a novel class of biomarkers for diagnosis of cancer and other diseases.
      ), diabetes (
      • Zampetaki A.
      • Kiechl S.
      • Drozdov I.
      • Willeit P.
      • Mayr U.
      • Prokopi M.
      • et al.
      Plasma microRNA profiling reveals loss of endothelial miR-126 and other microRNAs in type 2 diabetes.
      ), and tissue injury (
      • Wang K.
      • Zhang S.
      • Marzolf B.
      • Troisch P.
      • Brightman A.
      • Hu Z.
      • et al.
      Circulating microRNAs, potential biomarkers for drug-induced liver injury.
      ,
      • Redell J.B.
      • Moore A.N.
      • Ward 3rd, N.H.
      • Hergenroeder G.W.
      • Dash P.K.
      Human traumatic brain injury alters plasma microRNA levels.
      ). They have been detected in virtually all bodily fluids including breast milk, amniotic fluid, tears, cerebrospinal fluid, peritoneal fluid, blood, pleural fluid, saliva, semen, and urine (
      • Weber J.A.
      • Baxter D.H.
      • Zhang S.
      • Huang D.Y.
      • Huang K.H.
      • Lee M.J.
      • et al.
      The microRNA spectrum in 12 body fluids.
      ). Consequently, there is great interest in identifying miRNAs within these fluids to be used as biomarkers for the early detection of diseases.
      MicroRNAs are highly expressed in rapidly growing and undifferentiated cells such as cancer cells and embryonic stem cells. This led us to discover that miRNAs are highly expressed in human embryos and that intracellular miRNA expression patterns differ between euploid and aneuploid embryos (
      • Rosenbluth E.M.
      • Shelton D.N.
      • Sparks A.E.T.
      • Devor E.
      • Christenson L.
      • Van Voorhis B.J.
      Human blastocyst miRNA expression.
      ). Because miRNAs are known to be secreted from cells grown in culture into the surrounding media (
      • Hergenreider E.
      • Heydt S.
      • Treguer K.
      • Boettger T.
      • Horrevoets A.J.G.
      • Zeiher A.M.
      • et al.
      Atheroprotective communication between endothelial cells and smooth muscle cells through miRNAs.
      ), we first determined whether human embryos secrete miRNAs into in vitro fertilization (IVF) culture media, and if so, whether they were differentially secreted according to embryo chromosomal status. We further hypothesized that secreted miRNAs could be used as biomarkers to determine embryonic health before embryo transfer, with the ultimate goal of improving live-birth rates. Our final goal was to see whether specific patterns of media miRNA correlated with clinical IVF pregnancy outcomes.

      Materials and methods

      All experiments were performed and monitored via approved protocols under the University of Iowa Institutional Review Board. Specifically, patients donated embryos for research on biomarkers of embryo quality, and the patients were approached prospectively to allow the collection and testing of embryo culture media to find potential markers of implantation. All University of Iowa studies using human embryos are initially approved by the University's IVF ethics committee before institutional review board review.

       Embryo Culture

      The overall study design (Fig. 1) was to screen culture media from a cohort of donated, IVF embryos for relative and differential miRNA expression using an array-based quantitative real-time PCR (qRT-PCR) method. To confirm the miRNA array findings, the media from an expanded set of embryos were tested for miRNA expression with single miRNA qRT-PCR assays. Finally, miRNAs that were identified as being expressed in the initial experiments were then measured in embryonic conditioned media (hereafter referred to as spent media) samples from single-embryo transfer (SET) cycles to see whether the concentrations correlated with pregnancy outcomes.
      Figure thumbnail gr1
      Figure 1Flow diagram of experimental design.
      Patients who donated excess cryopreserved embryos for this study had used IVF for a variety of infertility diagnoses. However, these were undisclosed because of institutional review board protocol constraints. Only embryos fertilized by intracytoplasmic sperm injection (ICSI) were used in the initial studies to prevent possible sample contamination by accessory sperm. Pronuclear-stage embryos had been cryopreserved by controlled rate freezing 18 to 22 hours after ICSI in 1.5 M 1,2 propanediol (PROH; Sigma) as previously described elsewhere (
      • Testart J.
      • Lassalle B.
      • Belaisch-Allart J.
      • Hazout A.
      • Forman R.
      • Rainhorn J.D.
      • et al.
      High pregnancy rate after early human embryo freezing.
      ). Embryos were thawed by air warming for 40 seconds followed by 10-second exposure to 30°C sterile water. Cryoprotectants were removed in a stepwise dilutional fashion. Surviving embryos were cultured in groups of three to four in 50-μL microdrops of IVC-One (In VitroCare) supplemented with 20% SPS (Serum Protein Substitute; CooperSurgical/Sage) under oil (Cook Medical) in 5.5% to 6.0% CO2 in air at 37°C. Embryos were moved to fresh drops of IVC-One supplemented with 20% SPS on day 3.
      On the morning of day 4, embryos were moved to individual culture in 8 μL of IVC-Three (In VitroCare) supplemented with 20% SPS. Blank media control drops (hereafter referred to as controls) were incubated in the same dishes as those with media drops containing embryos. Before moving the embryos to fresh drops, we rinsed them through a series of five wash drops. All embryos were cultured to the blastocyst stage and were graded according a standardized classification system (

      Jansen R, Mortimer D, eds. Towards reproductive certainty: fertility and genetics beyond 1999: the plenary proceedings of the 11th World Congress on In Vitro Fertilization & Human Reproductive Genetics. Pearl River, NY: Parthenon, 1999.

      ). On the morning of the fifth day of culture, the embryos that had reached at least the early blastocyst stage and had an inner cell mass grade of B or better were chosen for assisted hatching. These embryos were moved to fresh culture media drops, and 6 μL of the spent media were collected and stored at −80°C for miRNA analysis. On the afternoon of day 5 or on the morning of day 6 of culture, hatching blastocysts were selected for biopsy.

       Embryo Biopsy and Determination of Chromosomal Makeup of Donated Pronuclear Embryos

      To determine the chromosomal makeup of donated embryos, a 10-μm channel was opened in the zona pellucida with a series of three to five laser pulses of 5-ms duration (Octax Microscience, GmbH). Approximately five herniating trophectoderm cells per embryo were aspirated into a biopsy pipette and detached by firing laser pulses at the area of constriction. After several passes through a wash solution, the biopsied cells were placed into a PCR tube with lysis buffer supplied by the Genesis Genetics Institute. The biopsy samples then were shipped on dry ice to their facility for array comparative genomic hybridization (aCGH) by their standard proprietary diagnostic technique using the BlueGnome 24sure V3 microarray platform. Mosaicism was determined by using standard clinical protocols, which includes a cutoff of 25% or greater deviation of fluorescent ratios between the hybridized chromosomes (
      • Mamas T.M.
      • Gordon A.
      • Brown A.
      • Harper J.
      • SenGupta S.
      Detection of aneuploidy by array comparative genomic hybridization using cell lines to mimic a mosaic trophectoderm biopsy.
      ).

       miRNA Isolation and Detection

      To maximize the total amount of RNA available from each spent media sample collected, the direct Cells-to-Ct method was used for reverse transcription (TaqMan Micro RNA Cells-to-CT Kit; Applied Biosystems). The 6 μL of day-5 spent media collected from each sample was placed into an equal amount of Cells-to-Ct lysis buffer with dilute deoxyribonuclease I. After 8 minutes, stop solution was added, and the samples were stored at −80°C. To allow the simultaneous reverse transcription of 754 human miRNAs, three endogenous miRNA controls, and one nonhuman negative control for each sample, two master mixes consisting of the A and B Megaplex RT primer pools, respectively (Human Pools Set v3.0; Applied Biosystems), were made per the manufacturer's Megaplex Pools protocol. We mixed 3 μL of lysate with 4.5 μL of master-mix for each reaction for a total volume of 7.5 μL. Thermal-cycling conditions were as follows: 40 cycles at 16°C for 2 minutes, 42°C for 1 minute, and 50°C for 1 second, then 85°C for 5 minutes and held at 4°C. Samples were stored at −80°C.
      We preamplified 2.5 μL of the reverse transcription product per A and B primer pool set using TaqMan PreAmp Master Mix (2×) and Megaplex PreAmp Primers (10×) (Applied Biosystems). The total reaction volume was 25 μL under these thermal-cycling conditions: an initial step of 95°C for 10 minutes, 55°C for 2 minutes, and 72°C for 2 minutes, followed by 12 cycles of 95°C for 15 seconds and 60°C for 4 minutes. The reaction was terminated at 99°C for 10 minutes and held at 4°C. The final preamplified product was not diluted.
      The preamplified reverse transcription products from each of the day-5 spent media were profiled using two 384-well TaqMan Low Density Array (TLDA) microfluidic cards with a final dilution of 1:16 (Human miRNA A+B Cards Set v3.0; Applied Biosystems). Additionally, two control media samples were also analyzed for a total of 17 TLDA arrays (34 cards). The arrays were loaded with a total of 50 μL of preamplification product per TLDA card. The PCR was performed with TaqMan Universal PCR Master Mix, No AmpErase UNG on a 7900HT Fast Real-Time PCR System (Applied Biosystems) under the following thermal-cycling conditions: 95°C for 10 minutes followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute.

       Array statistical analysis

      The TLDA real-time data were analyzed by using SDS RQ manager v2.4 and DataAssist v3.0 software (Applied Biosystems). Array miRNA expression data were analyzed relative to the expression of the small nuclear RNA (snRNA) U6 control probe that had been previously validated (
      • Rosenbluth E.M.
      • Shelton D.N.
      • Sparks A.E.T.
      • Devor E.
      • Christenson L.
      • Van Voorhis B.J.
      Human blastocyst miRNA expression.
      ). Comparisons between experimental groups were performed using the ΔCt method where fold change was expressed as 2−ΔΔCt. The statistical significance of fold changes was determined by performing a two-sample, two-tailed Student's t-test of the ΔCt values. P values were adjusted for false discovery using the Benjamini-Hochberg procedure built into the software package. The differentially expressed miRNAs in each group were determined by the relative expression to snRNA U6. MicroRNAs with statistically significant differences (at least fourfold changes with an adjusted P<.05) were confirmed in the following experiments.

       Confirmation of Differentially Expressed MicroRNA

      An additional group of donated cryopreserved embryos were thaw, cultured, and biopsied for chromosome analysis as previously described (see Fig. 1). Both culture medium and protein supplement were derived from the identical lot as the initially screened embryos. We collected 6 μL of day-5 media from these blastocysts for miRNA analysis and placed it in an equal amount of Cells-to-Ct lysis buffer. The lysates of these additional samples were added to the original group of media sample lysates to form an extended panel consisting of 28 day-5 media sample lysates. These were then processed for miRNA isolation, reverse transcribed, and preamplified as previously described.
      To confirm the differential expression of miRNAs in euploid versus aneuploid embryo media samples found by TLDA array, the miRNAs having the greatest differential expression were analyzed by qRT-PCR. All miRNAs were expressed relative to snRNA U6 in the expanded panel of blastocyst media. Blank media control drops were again incubated in the same dishes as those with media drops containing embryos, and the controls were analyzed in an identical manner as the test media.
      We performed qRT-PCR on 384-well plates using a dilution of 1:30 preamplification products in 10-μL reactions with no-template controls, TaqMan Universal PCR Master Mix, and No AmpErase UNG on the 7900HT Fast Real-Time PCR System (Applied Biosystems) with the following thermal cycling conditions: 95°C for 10 minutes followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. Each reaction was performed in triplicate. Standard efficiency curves were calculated for each miRNA probe set using serial dilutions of cDNA in a final 10-μL reaction volume.

       Single assay statistical analysis

      The single miRNA assay real-time data were analyzed by using SDS RQ manager v2.4 and DataAssist v3.0 software (Applied Biosystems). The miRNA expression data were analyzed relative to the expression of snRNA U6 in each of the individual blastocyst media samples. We chose U6 for normalization because of the consistent expression in all media samples and its presence in both the A and B Megaplex pool sets.
      Comparisons between experimental groups were performed using the ΔCt method, where fold change was expressed as 2−ΔΔCt. The statistical significance of fold changes was determined by performing a two-sample, two-tailed Student's t-test of the ΔCt values. P<.05 was considered statistically significant differential expression.

       Fresh SET IVF Cycles

      Expression of identified miRNAs from the previous experiments were examined in clinical IVF media samples and then correlated with IVF pregnancy outcomes. Spent media from a total of 55 consented patients undergoing fresh SET IVF cycles were collected. Patients underwent standard controlled ovarian stimulation followed by ultrasound-guided oocyte retrieval. Oocytes were fertilized either by ICSI or regular insemination, as determined by standard clinic protocol. Embryos were cultured in microdrops under light mineral oil in an environment of 5% to 6% CO2 in air at 37°C as follows. Day-1 embryos were cultured in groups of four in 50-μL drops of IVC-One medium (InVitroCare) supplemented with 20% serum protein substitute (Cooper Surgical Inc.) for 48 hours. Day-3 embryos were transferred to individual 15-μL drops of IVC-One and 10% serum protein substitute for 24 hours. Day-4 embryos were moved to individual 15-μL drops of IVC-Three with 10% serum protein substitute for 24 hours. Empty drops of IVC-Three with SPS served as controls and were incubated in the same dishes as those with media drops containing embryos. Before moving embryos to fresh drops, they were rinsed through five wash drops. We collected 12 μL of spent media from individual culture drops on days 4 and 5, which was stored at −80°C. For the miRNA analysis, 6 μL of spent media were thawed and processed using the same Cells-to-Ct isolation, reverse transcription, and preamplification as previously described. MicroRNA expression was determined by qRT-PCR and single miRNA assays, as described in the confirmation step.

      Results

       Donated Embryos

      Thirteen couples donated 91 cryopreserved pronuclear-stage embryos for this study (experiments 1 and 2, Fig. 1). Of these, 35 developed in culture to the blastocyst stage with trophectoderm biopsies performed upon them for the aCGH determination of chromosomal makeup. Seven embryos were excluded from further analysis: three for lack of an aCGH signal and four for mosaicism. Mosaic embryos were excluded to unambiguously discriminate between euploid and aneuploid embryo miRNA expression.

       MicroRNA expression in IVF culture media of cultured embryos

      The media from a total of 15 embryos (five female, five male, and five aneuploid) were initially screened for miRNA expression by individual TLDA arrays. Ten miRNAs were consistently detected in the spent IVF culture media using a cycle threshold (Ct) value cutoff of 38 (Table 1) where the Ct value is the number of PCR cycles needed for the fluorescence of the sample to be detected above the fluorescence threshold. The threshold value is set above the background fluorescence and in the exponential phase of the PCR reaction. The Ct value is inversely proportional to the original relative expression level of the miRNA of interest. Although high Ct values (38–40) generally represent extremely small levels of target transcripts and can be difficult to discern from the background, we used an initial high Ct value during our screening process to ensure that we captured all potential meaningful signals from such small amounts of starting material.
      Table 1MicroRNA found in spent media (embryo exposed) and control media.
      AssayAverage cycle threshold value
      Spent mediaControl media
      hsa-miR-106b-437315528.628.4
      hsa-miR-191–439541036.340
      hsa-miR-30c-437306024.424.6
      hsa-miR-372–437302925.540
      hsa-miR-376a-437302637.535.4
      hsa-miR-548a-3p-438094829.029.3
      hsa-miR-548c-3p-438099324.624.0
      hsa-miR-548d-3p-438100832.230.9
      hsa-miR-576–3p-439546231.330.5
      hsa-miR-603–00156627.730.6
      hsa-miR-64539.531.0
      Of the 10 miRNAs identified, only two miRNAs (miR-372 and miR-191) were confirmed by later single assay qRT-PCR analysis to be solely in spent media samples. The rest were present in the control media before embryo exposure. One miRNA, miR-645, was detected in unexposed media samples with an average Ct value of 31.0 but was undetected in all spent media samples. To determine the source of miRNA in media unexposed to embryos, we assayed both protein-free media and media with added serum protein substitute and only detected miRNA in the latter.

       MicroRNA expression in media based on chromosomal content of the embryo

      We next tested whether the miRNA expression in the media varied based on whether the embryo was euploid. For this experiment (experiment 2, Fig. 1), we tested the media from 28 embryos in which the chromosomal content was known by aCGH (Table 2). The only difference we found was that miR-191 was more highly expressed in media from aneuploid embryos than media from euploid embryos (4.7 fold, P=.031; Table 2). No differences in miRNA expression were detected when comparing media from euploid male and female embryos.
      Table 2MicroRNA cycle threshold values in spent culture media from euploid and aneuploid embryos.
      Chromosome complementnmiR-191miR-372miR-645
      46 XY834.9 (29–40)29.0 (22.7–40)39.2 (36.3–40)
      46 XX1136.8 (29.2–40)31.1 (24.1–40)38.7 (34.6–40)
      46 XX (19q 12.3–qter)127.524.140
      45 X del (X1)128.824.437.4
      47 XX; +1 del (2q 21.3–qter)129.225.940
      45 XY −17127.824.040
      46 XX; del (4q 13.2–qter)

      del (9q 21.11–qter)
      14029.240
      45 XX −141404039
      47 XXY135.930.737.2
      48 XXY +191404038.8
      45 XY −1614027.839
      Note: Cycle threshold (Ct) values or average Ct values with ranges are presented where appropriate. The Ct value is defined as the number of polymerase chain reaction (PCR) cycles needed for the fluorescence of the sample to be detected above the fluorescence threshold. The threshold value is set above the background fluorescence and in the exponential phase of the PCR reaction. The Ct is inversely proportional to the original relative expression level of the miRNA of interest.

       SET clinical samples

      To determine whether secreted miRNA correlated with clinical IVF outcomes, media samples were collected on day 4 and day 5 of culture from patients undergoing IVF with fresh blastocyst embryo transfers (experiment 3, Fig. 1). To easily pair media samples with pregnancy outcomes, we cultured the embryos in individual media droplets, and patients receiving only a single embryo were included. There were a total of 55 patients recruited, with their cycle outcomes listed in Figure 2. MicroRNAs previously identified to be uniquely secreted into the culture media (miR-372 and miR-191) were analyzed by single assay qRT-PCR. MicroRNA-645 was also investigated owing to the unique finding of its being present in control media samples but undetectable in conditioned media from morphologically good-quality embryos.
      Figure thumbnail gr2
      Figure 2MicroRNAs in spent media from single-embryo transfer cycles separated by insemination method and clinical outcomes. Data are presented as ΔCt values (and ranges) to control for differences in starting concentrations of RNA. (Note: The ΔCt values are derived by comparing the Ct value of the miRNA of interest with the Ct value of control RNA (snRNA U6) in the sample. In these particular experiments, smaller numbers indicate a higher concentration of miRNA.)

       SET samples by fertilization method

      To determine whether the method of fertilization affected the secretion of miRNA into culture media, comparisons were made between regularly inseminated embryos and embryos fertilized by ICSI. In day-5 media, miR-191 and miR-372 were found to be 4.4 fold (P=.014) and 7.1 fold (P=.045) more highly concentrated in media from embryos inseminated by ICSI (n = 28) when compared with embryos fertilized with regular insemination (n = 27). There were no differences between these two groups when day-4 media was analyzed.

       SET samples by day of analysis

      To determine whether there were differences in miRNA secretion on different days of embryo culture, media samples from day 4 (n = 55) and from day 5 (n = 55) of embryo culture were analyzed. Both media sample groups had exposure to individual embryos for 24 hours. We found that miR-372 was 5.4 fold (P<.01) more highly concentrated in day-5 media than in day-4 media. When comparing only spent media from ICSI-inseminated embryos, we found miR-191 was 1.9-fold (P=.024) and miR-372 was 12.0-fold (P<.01) more highly concentrated in day-5 media.

       SET samples by pregnancy outcomes

      Media samples from single-embryo blastocyst transfers were compared between successful (live-birth) samples and failed (biochemical pregnancies, spontaneous abortions, or implantation failure) samples. There were no statistically significant differences in miRNA concentrations between these groups when all the samples were analyzed. However, when analyzing media samples from embryos that were regularly inseminated only (ICSI embryos excluded), miR-191, miR-645, and miR-372 were 5.1-fold (P=.018), 6.0-fold (P=.024), and 7.1-fold (P=.046), respectively, more highly detected in day-5 media from failed IVF cycle embryos (n = 9) when compared with media from embryos that led to live birth (n = 18).

      Discussion

      Selection of the best embryo for transfer during an IVF cycle is imprecise. The most commonly used selection method is to choose an embryo based on morphologic criteria. At the blastocyst stage, embryos with better morphology are more likely to have a normal chromosomal content and are more likely to implant and produce a pregnancy. However, it is known that many well-developed, morphologically normal blastocysts will still be chromosomally abnormal or will not implant (
      • Alfarawati S.
      • Goodall N.N.
      • Fragouli E.
      • Daphnis D.D.
      • Gordon A.
      • Griffiths T.
      • et al.
      Cytogenetic analysis of human blastocysts with the use of FISH, CGH and aCGH: scientific data and technical evaluation.
      ). Thus, there is interest in finding a biomarker that will allow even better selection of the best embryo beyond morphologic criteria. Such a biomarker could lead to higher pregnancy rates per transfer and reduce the rate of multiple gestations with IVF by allowing more successful implementation via SET.
      Several potential embryonic biomarkers have been recently investigated. Secreted proteins have been identified in embryo culture media with some correlating with embryo morphology (
      • Mains L.M.
      • Christenson L.
      • Yang B.
      • Sparks A.E.
      • Mathur S.
      • Van Voorhis B.J.
      Identification of apolipoprotein A1 in the human embryonic secretome.
      ). However, there are currently no protein biomarkers that reliably correlate with embryo implantation or pregnancy. These studies are complicated because the detection of small changes in secreted proteins is especially difficult in the background of high concentrations of protein required to be supplemented into the culture media for optimal in vitro embryo development. Metabolomic profiles have been identified in media surrounding embryos that are associated with a “healthy” embryo, but to date this technology has not led to an improved ability to select embryos that will implant (
      • Hardarson T.
      • Ahlstro A.
      • Rogberg L.
      • Botros L.
      • Hillensjo T.
      • Westlander G.
      • et al.
      Non-invasive metabolomics profiling of day 2 and 5 embryo culture medium: a prospective randomized trial.
      ,
      • Vergouw C.G.
      • Kieslinger D.C.
      • Kostelijk E.H.
      • Botros L.L.
      • Schats R.
      • Hompes P.G.
      • et al.
      Day 3 embryo selection by metabolomics profiling of culture medium with near-infrared spectroscopy as an adjunct to morphology: a randomized controlled trial.
      ). Preimplantation genetic screening (PGS) of embryos has promise, but it is invasive, expensive, and requires embryo biopsy. Long-term effects from the procedure are still unknown. When using older methods for chromosomal analysis, PGS has not improved embryo selection, as evidenced by the same or even worse pregnancy rates in prospective randomized trials (
      • Zamora S.
      • Clavero A.
      • Gonzalvo M.C.
      • Luna del Castillo J.D.D.
      • Roldan-Nofuentes J.A.
      • Mozas J.
      • et al.
      PGS-FISH in reproductive medicine and perspective directions for improvement: a systematic review.
      ).
      The ideal biomarker would allow noninvasive analysis of the embryo by analyzing the media surrounding the embryo. The marker would be stable over time, specific to embryos, and easily measured to allow rapid assessment before embryo transfer. We hypothesized that miRNAs might be good candidates because they are stable, resistant to degradation, consistently expressed, easily detected, and already been shown to correlate with a variety of pathologic conditions (
      • Chen X.
      • Ba Y.
      • Ma L.
      • Cai X.
      • Yin Y.
      • Wang K.
      • et al.
      Characterization of microRNAs in serum: a novel class of biomarkers for diagnosis of cancer and other diseases.
      ,
      • Zampetaki A.
      • Kiechl S.
      • Drozdov I.
      • Willeit P.
      • Mayr U.
      • Prokopi M.
      • et al.
      Plasma microRNA profiling reveals loss of endothelial miR-126 and other microRNAs in type 2 diabetes.
      ,
      • Wang K.
      • Zhang S.
      • Marzolf B.
      • Troisch P.
      • Brightman A.
      • Hu Z.
      • et al.
      Circulating microRNAs, potential biomarkers for drug-induced liver injury.
      ,
      • Redell J.B.
      • Moore A.N.
      • Ward 3rd, N.H.
      • Hergenroeder G.W.
      • Dash P.K.
      Human traumatic brain injury alters plasma microRNA levels.
      ,
      • Alevizos I.
      • Gabor G.
      MicroRNAs as biomarkers in rheumatic diseases.
      ). Our objective was to characterize the miRNA content of media around human blastocysts and search for differential expression of miRNAs based on the genetic makeup of the embryo. We further sought to determine if miRNA concentration correlated with pregnancy outcomes.
      We found miRNAs to be readily detectable in IVF culture media. However, the majority of miRNAs detected were also present in the culture media before embryo culture. Further analysis showed that the miRNAs were derived from the protein supplement used in our culture media. Because miRNA containing exosomes are 30–90 nm in diameter, it is feasible that they readily pass through the 200 nm filtration process the manufacturer uses for sterilization. This is particularly intriguing considering the recent findings that miRNAs packaged into exosomes can target and affect gene transcription in remote cells (
      • Hergenreider E.
      • Heydt S.
      • Treguer K.
      • Boettger T.
      • Horrevoets A.J.G.
      • Zeiher A.M.
      • et al.
      Atheroprotective communication between endothelial cells and smooth muscle cells through miRNAs.
      ,
      • Yang M.
      • Chen J.
      • Su F.
      • Yu B.
      • Su F.
      • Lin L.
      • et al.
      Microvesicles secreted by macrophages shuttle invasion-potentiating microRNAs into breast cancer cells.
      ).
      One of the miRNAs detected in this study, miR-645, was present and expressed in all control samples. However, miR-645 was found to be undetectable in the media from several healthy embryos. Furthermore, higher levels of miR-645 correlated to poor pregnancy outcomes in our clinical group of non-ICSI inseminated embryos. Together, this provides indirect evidence that miRNAs might be incorporated and used by developing embryos. Future studies could confirm this finding and determine whether IVF culture media enriched or deprived of specific miRNAs could improve embryonic development.
      The two other miRNAs that correlated with IVF pregnancy outcome, miR-191 and miR-372, were not present in IVF culture media before exposure to the embryos. Higher levels of miR-191 correlated with both aneuploid media samples and failed IVF cycles, suggesting that miR-191 may be a biomarker of embryo aneuploidy and subsequent pregnancy failure. High levels of miR-372 also correlated with IVF failure. However, miR-372 did not correlate with embryonic ploidy status. MicroRNA-372 is known to be highly expressed in embryonic stem cells and has been recently found to be the most highly expressed miRNA in human embryos (
      • Rosenbluth E.M.
      • Shelton D.N.
      • Sparks A.E.T.
      • Devor E.
      • Christenson L.
      • Van Voorhis B.J.
      Human blastocyst miRNA expression.
      ). The exact role these miRNAs play within embryo development has yet to be determined. However, target prediction software reveals miR-191 and miR-372 both may regulate mitogen-activated protein kinase kinase kinase 1 (MAP3K1) and cyclin-dependent kinase 6 (CDK6), genes critical in cell cycle, signaling, and apoptotic pathways (
      • Maragkakis M.
      • Reczko M.
      • Simossis V.A.
      • Alexiou P.
      • Papadopoulos G.L.
      • Dalamagas T.
      • et al.
      DIANA-microT web server: elucidating microRNA functions through target prediction.
      ). MicroRNA-372 has recently been shown to help catalyze the induction of human fibroblasts to induced pluripotent stem cells and points to the importance of miRNAs in regulating cell differentiation (
      • Subramanyam D.
      • Lamouille S.
      • Judson R.L.
      • Liu J.Y.
      • Bucay N.
      • Derynck R.
      • et al.
      Multiple targets of miR-302 and miR-372 promote reprogramming of human fibroblasts to induced pluripotent stem cells.
      ).
      MicroRNAs 372 and 191 were found to be higher in the media of embryos fertilized by ICSI when compared with regularly inseminated embryos. Possible explanations could be that physical damage to the zona pellucida and oolemma after ICSI permits the leakage of miRNAs into the culture media. MicroRNAs have been found to be more highly expressed under conditions of cell stress (
      • Weber M.
      • Baker M.B.
      • Moore J.P.
      • Searles C.D.
      MiR-21 is induced in endothelial cells by shear stress and modulates apoptosis and eNOS activity.
      ). Embryos fertilized by ICSI have already endured physical insult that could possibly mediate higher levels of miRNA expression. Regardless of the mechanism, when designing future studies, the method of fertilization will have to be taken into consideration. Although we found miRNAs that correlated with pregnancy outcomes, we could only find this correlation from embryos that were regularly inseminated, suggesting that the process of ICSI alters miRNA secretion patterns, which may confound the ability to use miRNA as a biomarker when this method of fertilization is used.
      Due to the extended period of time between collecting media from clinical SET cycles and culturing donated experimental embryos, different lots of media were required. Although no differences were noted in the blank media controls from the different lots using our target single miRNA assays, it is feasible that there were differences in levels of other untested miRNAs. In addition to batch-to-batch variation, there will inevitably be differences between media from varying manufacturers which must be accounted for in future studies. Additionally, due to protocol changes, different protein concentrations were used in our clinical and experimental culture media (10% vs. 20%). No direct comparisons were made between these experimental groups. However, it is feasible that different protein concentrations could affect miRNA secretion into the culture media, and we may have missed potential miRNAs markers in either group.
      Although no difference in miRNA media content could be detected between male and female embryos, the number of samples analyzed by TLDA was low. Differences between embryos are likely to be extremely small, making detection in a limited number of samples exceedingly difficult. Future studies using a greater number of samples and more sensitive detection methods might ferret out subtle differences missed in these experiments.
      This is the first study that we are aware of to investigate miRNA secretion of human embryos. We found miRNAs that are secreted from IVF embryos into culture media correlate with embryonic aneuploidy and pregnancy outcomes. We were hopeful to replicate our findings in day-4 media samples to be able to test embryos before a day-5 blastocyst transfer. However, miRNA expression was statistically significantly lower on day 4, making early detection too difficult. Regardless, qRT-PCR can be performed inexpensively in-house in a matter of hours, making fresh blastocyst transfers still feasible even with same-day media analysis. We were encouraged by the initial findings of this study, and we believe that this work provides early proof-of-principle evidence that miRNA can be measured in day-5 culture medium. Larger clinical trials are needed to determine whether measurement of miRNA can be used as a novel, noninvasive approach to assessing embryonic health.

      References

        • Laurent L.C.
        MicroRNAs in embryonic stem cells and early embryonic development.
        J Cell Mol Med. 2008; 12: 2181-2188
        • Valadi H.
        • Ekström K.
        • Bossios A.
        • Sjöstrand M.
        • Lee J.J.
        • Lötvall J.O.
        Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells.
        Nat Cell Biol. 2007; 9: 654-659
        • Jung M.
        • Schaefer A.
        • Steiner I.
        • Kempkensteffen C.
        • Stephan C.
        • Erbersdobler A.
        • et al.
        Robust MicroRNA stability in degraded RNA preparations from human tissue and cell samples.
        Clin Chem. 2010; 56: 998-1006
        • Hergenreider E.
        • Heydt S.
        • Treguer K.
        • Boettger T.
        • Horrevoets A.J.G.
        • Zeiher A.M.
        • et al.
        Atheroprotective communication between endothelial cells and smooth muscle cells through miRNAs.
        Nat Cell Biol. 2012; 3: 249-256
        • Chen X.
        • Ba Y.
        • Ma L.
        • Cai X.
        • Yin Y.
        • Wang K.
        • et al.
        Characterization of microRNAs in serum: a novel class of biomarkers for diagnosis of cancer and other diseases.
        Cell Res. 2008; 18: 997-1006
        • Zampetaki A.
        • Kiechl S.
        • Drozdov I.
        • Willeit P.
        • Mayr U.
        • Prokopi M.
        • et al.
        Plasma microRNA profiling reveals loss of endothelial miR-126 and other microRNAs in type 2 diabetes.
        Circ Res. 2010; 107: 810-817
        • Wang K.
        • Zhang S.
        • Marzolf B.
        • Troisch P.
        • Brightman A.
        • Hu Z.
        • et al.
        Circulating microRNAs, potential biomarkers for drug-induced liver injury.
        Proc Natl Acad Sci USA. 2009; 106: 4402-4407
        • Redell J.B.
        • Moore A.N.
        • Ward 3rd, N.H.
        • Hergenroeder G.W.
        • Dash P.K.
        Human traumatic brain injury alters plasma microRNA levels.
        J Neurotrauma. 2010; 27: 2147-2156
        • Weber J.A.
        • Baxter D.H.
        • Zhang S.
        • Huang D.Y.
        • Huang K.H.
        • Lee M.J.
        • et al.
        The microRNA spectrum in 12 body fluids.
        Clin Chem. 2010; 56: 1733-1741
        • Rosenbluth E.M.
        • Shelton D.N.
        • Sparks A.E.T.
        • Devor E.
        • Christenson L.
        • Van Voorhis B.J.
        Human blastocyst miRNA expression.
        Fertil Steril. 2013; 99: 855-861.e3
        • Testart J.
        • Lassalle B.
        • Belaisch-Allart J.
        • Hazout A.
        • Forman R.
        • Rainhorn J.D.
        • et al.
        High pregnancy rate after early human embryo freezing.
        Fertil Steril. 1986; 46: 268-272
      1. Jansen R, Mortimer D, eds. Towards reproductive certainty: fertility and genetics beyond 1999: the plenary proceedings of the 11th World Congress on In Vitro Fertilization & Human Reproductive Genetics. Pearl River, NY: Parthenon, 1999.

        • Mamas T.M.
        • Gordon A.
        • Brown A.
        • Harper J.
        • SenGupta S.
        Detection of aneuploidy by array comparative genomic hybridization using cell lines to mimic a mosaic trophectoderm biopsy.
        Fertil Steril. 2012; 97: 943-947
        • Alfarawati S.
        • Goodall N.N.
        • Fragouli E.
        • Daphnis D.D.
        • Gordon A.
        • Griffiths T.
        • et al.
        Cytogenetic analysis of human blastocysts with the use of FISH, CGH and aCGH: scientific data and technical evaluation.
        Hum Reprod. 2011; 26: 480-490
        • Mains L.M.
        • Christenson L.
        • Yang B.
        • Sparks A.E.
        • Mathur S.
        • Van Voorhis B.J.
        Identification of apolipoprotein A1 in the human embryonic secretome.
        Fertil Steril. 2011; 96: 422-427.e2
        • Hardarson T.
        • Ahlstro A.
        • Rogberg L.
        • Botros L.
        • Hillensjo T.
        • Westlander G.
        • et al.
        Non-invasive metabolomics profiling of day 2 and 5 embryo culture medium: a prospective randomized trial.
        Hum Reprod. 2012; 27: 89-96
        • Vergouw C.G.
        • Kieslinger D.C.
        • Kostelijk E.H.
        • Botros L.L.
        • Schats R.
        • Hompes P.G.
        • et al.
        Day 3 embryo selection by metabolomics profiling of culture medium with near-infrared spectroscopy as an adjunct to morphology: a randomized controlled trial.
        Hum Reprod. 2012; 27: 2304-2311
        • Zamora S.
        • Clavero A.
        • Gonzalvo M.C.
        • Luna del Castillo J.D.D.
        • Roldan-Nofuentes J.A.
        • Mozas J.
        • et al.
        PGS-FISH in reproductive medicine and perspective directions for improvement: a systematic review.
        J Assist Reprod Genet. 2011; 28: 747-757
        • Alevizos I.
        • Gabor G.
        MicroRNAs as biomarkers in rheumatic diseases.
        Nat Rev Rheum. 2010; 6: 391-398
        • Yang M.
        • Chen J.
        • Su F.
        • Yu B.
        • Su F.
        • Lin L.
        • et al.
        Microvesicles secreted by macrophages shuttle invasion-potentiating microRNAs into breast cancer cells.
        Mol Can. 2011; 10: 117
        • Maragkakis M.
        • Reczko M.
        • Simossis V.A.
        • Alexiou P.
        • Papadopoulos G.L.
        • Dalamagas T.
        • et al.
        DIANA-microT web server: elucidating microRNA functions through target prediction.
        Nucleic Acids Research. 2009 Jul 1; 37: W273-W276
        • Subramanyam D.
        • Lamouille S.
        • Judson R.L.
        • Liu J.Y.
        • Bucay N.
        • Derynck R.
        • et al.
        Multiple targets of miR-302 and miR-372 promote reprogramming of human fibroblasts to induced pluripotent stem cells.
        Nat Biotechnol. 2011; 29: 443-448
        • Weber M.
        • Baker M.B.
        • Moore J.P.
        • Searles C.D.
        MiR-21 is induced in endothelial cells by shear stress and modulates apoptosis and eNOS activity.
        Biochem Biophys Res Commun. 2010; 393: 643-648