Abstract
Age has a detrimental effect on reproduction and as an increasing number of women postpone motherhood, it is imperative to assess biological age in terms of fertility prognosis and optimizing fertility treatment individually. Horvath’s epigenetic clock is a mathematical algorithm that calculates the biological age of human cells, tissues or organs based on DNA methylation levels. The clock, however, was previously shown to be highly inaccurate for the human endometrium, most likely because of the hormonal responsive nature of this tissue. The aim of this study was to determine if epigenetically based biological age of the human endometrium correlated with chronological age, when strictly timed to the same time point in the menstrual cycle. Endometrial biopsies from nine women were obtained in two consecutive cycles, both strictly timed to the LH surge (LH + 7) and additionally, peripheral whole blood samples were analyzed. Using the Illumina HumanMethylation 450 K array and Horvath’s epigenetic clock, we found a significant correlation between the biological age of the endometrium and the chronological age of the participants, although the endometrial biological age was accelerated by comparison with blood and chronological age. Moreover, similar biological ages were found in pairs of consecutive biopsies, indicating that an endometrial biopsy does not alter the biological age in the following cycle. In conclusion, as long as endometrial samples are timed to the same time point in the menstrual cycle, Horvath’s epigenetic clock could be a powerful new biomarker of reproductive aging in the human endometrium.
Introduction
With embryo culture conditions continuing to improve, the rate-limiting step in assisted reproduction technology (ART) remains the implantation of the embryo in the endometrium (Macklon et al. 2006, Dekel et al. 2010). The human endometrium is a unique dynamic tissue that cyclically sheds, regenerates and undergoes complex molecular, cellular and functional changes in response to sex hormones, preparing for embryo implantation (Psychoyos 1986, Salamonsen 2008). As an increasing number of women postpone their motherhood (Baird et al. 2005, Schmidt et al. 2012), it is becoming increasingly important to have a tool that can assess biological age in terms of fertility prognosis (Alviggi et al. 2009).
Recently, an accurate biomarker of biological age was developed and termed Horvath’s epigenetic clock (Horvath 2013). It uses the methylation status of 353 CpG sites in the human genome to calculate an epigenetically based biological age of an individual (also called methylation age or DNAm age), which has shown to be remarkably closely correlated with chronological age irrespective of the DNA source within the organism. The epigenetic clock allows one to compare the biological age of different tissues or cells of the human body and by contrasting the biological age of different tissues with an individual’s chronological age, any differences in tissue biological age can be estimated. A discrepancy between biological and chronological age could potentially be attributed to pathological processes, underlying disease or lifestyle choices (Horvath 2013). The relevance of Horvath’s epigenetic clock for human health and disease has already been demonstrated. An increased age acceleration in blood correlates with increased all-cause mortality (Marioni et al. 2015, Chen et al. 2016, Christiansen et al. 2016, Perna et al. 2016). Further, obesity was shown to be associated with increased age acceleration in the liver (Horvath et al. 2014). The first study addressing the importance of the biological age in reproduction comprised methylation data from four large studies including the Women’s Health Initiative, where they found that postmenopausal women with early-onset menopause had increased age acceleration in their blood compared to women with a later onset of menopause (Levine et al. 2016). Surgical menopause, i.e. bilateral oophorectomy, was also associated with increased age acceleration in both blood and saliva, overall suggesting that loss of female sex hormones accelerates the aging rate in blood and possibly in other tissues (Levine et al. 2016).
While Horvath’s epigenetic clock has been widely used for many different tissues today as a proxy for biological age, it has not been used to investigate the endometrium and its implications for reproductive aging. This is most likely because of the poor correlation with chronological age reported specifically for this tissue in the initial publication of Horvath (Horvath 2013).
The aim of this study was to investigate if the biological age of the human endometrium can be estimated using Horvath’s epigenetic clock. We collected blood and paired endometrial biopsies from nine women in two consecutive cycles, with both biopsies strictly timed to LH + 7, corresponding to the window of implantation. This approach allowed us to assess the correlation between biological and chronological age in the normal human endometrium and to investigate if biological age was different between the two biopsies.
Materials and methods
Participants
Nine female volunteers with a regular menstrual cycle of 28–32 days were included in the study during the period of January 1st 2015 to November 1st 2015 at the Fertility Clinic, Horsens Regional Hospital, Horsens, Denmark. The participants were recruited through fertility patient’s social network and through adverts in the municipality.
Inclusion criteria were no prior use of intrauterine devices or contraception pills, age of 18–40 years and a BMI of 18–32 kg/m2. Women with congenital uterine abnormalities, presence of fibroids or polyps, as well as women with suspected hydrosalpinges, adenomyosis, undergoing fertility treatment or expressing a wish to be pregnant were excluded from the study.
Ethical approval
The study was approved by the National Research Ethics Committee and the Danish Data Protection Agency. The study was registered at www.clinicaltrials.gov (NCT02219425). Informed consent was obtained from all participants.
Endometrial biopsies
All participants underwent a therapeutic endometrial scratching with a Pipelle de Cornier (Laboratoires Prodimed, Neuilly-en-Thelle, France) in two consecutive cycles. For each participant, the length of their menstrual cycle was calculated and afterwards they were asked to start testing their urine with a commercial LH surge urine test (Clinic Exclusive, Unigroup, Denmark) in the evening 3 days before their estimated LH surge and repeated until positive, thereby defining cycle day LH + 0. The first endometrial biopsy was then taken at cycle day LH + 7 (Biopsy 1) followed by a second biopsy (Biopsy 2) in the next cycle at day LH + 7. The scratching was carried out with the patient lying in a lithotomy position and was performed sterile once in each quadrant of the endometrium. Biopsies were snap frozen at −80°C. Each frozen endometrial biopsy was sectioned into two pieces, one for endometrial dating and one for DNA extraction. Venous peripheral whole blood samples (4 mL) were taken in a K2 EDTA Vacutainer from each participant and stored at 4°C until DNA extraction, which was performed the same day (Fig. 1).
Study overview. Nine female volunteers were recruited and paired endometrial biopsies obtained in two consecutive cycles. Blood samples were also collected at the time of the first visit. Biopsies were timed to LH + 7 the assumed window of implantation.
Citation: Reproduction 155, 2; 10.1530/REP-17-0601
Endometrial dating
The endometrial biopsies for dating were thawed, fixed in 5% buffered formalin for at least 24 h and consecutively embedded in paraffin. The embedded samples were cut in 4 μm thick sections, mounted on SuperFrost/Plus slides (Menzel, Germany) and subsequently deparaffinized and rehydrated. Hematoxylin eosin counterstaining was followed by mounting the cover slide with Histofluid. The staining was performed in triplicate and images were taken at the same exposure times by a Leica CTR5000 (Wetzlar, Germany) upright and a MetaSystems (Altußheim, Germany) microscope. The evaluation was performed according to Noyes criteria (Noyes et al. 1975) by two independent investigators who were blinded to the experimental paradigm (Fig. 2).
Paraffin-embedded and hematoxylin & eosin counterstained paired endometrial samples. Biopsy 1 was taken at LH + 7 while biopsy 2 taken at LH + 7 in the following menstrual cycle. The scale bar indicates 100 µm.
Citation: Reproduction 155, 2; 10.1530/REP-17-0601
DNA extraction
The endometrial biopsies for DNA extraction were homogenized into powder with a mortar and pestle in liquid nitrogen to avoid thawing. Genomic DNA was extracted from approximately 10 mg of homogenized tissue and from 100 μL of fresh peripheral blood using Qiagen Blood and Tissue kit (Qiagen) according to the manufacturer’s protocol. An agarose gel electrophoresis was performed on all DNA samples, demonstrating the presence of large amounts of RNA in the DNA samples from endometrial biopsies most likely because of the highly proliferative nature of the tissue. The genomic DNA from the endometrial samples was therefore each treated with 4 μL RNAase (Qiagen) and a second agarose gel electrophoresis was performed to confirm that RNA was removed from the samples, as well as that DNA was intact and of good quality. All DNA samples were further quantified using a NanoDrop (ND-1000, Nanodrop Technologies, Wilmington, DE, USA) and stored at −20°C until further use.
Methylation arrays
Genomic DNA (500 ng/sample) from endometrial and blood samples was bisulfite converted with EZ Methylation Gold Kit (Zymo Research, Irvine, CA, USA) and analyzed using the Infinium HumanMethylation450 BeadChips (Illumina, San Diego, CA, USA) array according to the manufacturer’s protocol.
Epigenetically based biological age
The biological age was calculated using Horvath’s epigenetic clock (Horvath 2013). The intensity files from the Illumina arrays were imported into Partek Genomic Suite 6.6 (St. Louis, MO, USA) for quality control. No technical outliers were found in the intensity plot. The un-normalized beta values were uploaded to the online DNA methylation Age Calculator available at https://dnamage.genetics.ucla.edu/.
Statistical analysis
To test for differences in means between groups (biological age of first biopsy and second biopsy) a paired t-test was used. The assumption of normal distribution was confirmed by a QQ-plot and a Shapiro–Wilk test.
The correlation between the biological age and chronological age was evaluated with Pearson’s correlation test. For the endometrium, the mean of the two biopsies were used in the correlation test. In two-sided P values, all P < 0.05 were considered significant. Mean values are given with 95% confidence intervals (CI).
Results
Nine Caucasian women provided a blood sample and two endometrial biopsies. The chronological age of the study participants ranged from 19.1 to 38.6 years. Six of the women had a proven pregnancy (PP) with ≥1 childbirth, while the remaining three were presumably fertile but had no prior pregnancy (NP). The clinical characteristics of the nine participants are listed in Table 1. The endometrial biopsies were taken in two consecutive cycles, and both strictly timed to LH + 7 (Fig. 1), The histological staging confirmed that the biopsies were in the mid-secretory phase (Fig. 2).
The biological age of the endometrial samples, calculated using Horvath’s epigenetic clock ranged from 20.5 to 41.6 years (Table 1). The mean biological age was 35.3 years (s.d. = 4.7; CI = 31.7–39.0) for the first biopsies (Biopsy 1) and 35.5 years, (s.d. = 3.6; CI = 32.8–38.3) for the second biopsies (Biopsy 2). No significant difference in biological age was found between the first and the second biopsy (paired t-test delta = −0.2 years; P = 0.85; CI = −2.5-2.1). We therefore decided to take the average of biopsy 1 and biopsy 2 for each participant to increased accuracy in the following analysis. The overall mean biological age was 35.4 years (s.d. = 3.9; CI = 32.4–38.4) in the endometrium, 33.9 years (s.d. = 5.9; CI = 29.3–38.4) in blood, while mean chronological age was 31.0 years (s.d. = 6.6; CI = 26.0–36.1). The biological age of the endometrium was on average 1.5 years higher than whole blood (P = 0.23) and 4.4 years higher than the corresponding chronological age (P = 0.012) (Fig. 3), suggesting a moderate age acceleration in the endometrium at day LH + 7. This observation is in line with data obtained for other hormone responsive tissues as noted by Horvath (Horvath 2013).
Chronological age and biological age of blood and endometrium for the nine participating woman. Samples from the same individual are connected with a line. A paired t-test was used to test for differences between groups. s.d.: standard deviation (s.d.)
Citation: Reproduction 155, 2; 10.1530/REP-17-0601
Baseline characteristics for volunteers.
| ID | Pregnancy status | Chronological age | Biological age blood, years | Biological age endometrium, years | |||||
|---|---|---|---|---|---|---|---|---|---|
| Blood | Difference (Blood chronological age) | Endometrium (Biopsy 1) | Endometrium (Biopsy 2) | Difference (Biopsy 2–1) | Endometrium mean (Biopsy 1 and 2) | Difference (Endometrium mean – chronological age) | |||
| 001 | PP | 34.2 | 35.2 | 1.0 | 38.3 | 37.4 | −0.9 | 37.8 | 3.6 |
| 002 | PP | 35.4 | 39.8 | 4.4 | 41.6 | 40.1 | −1.5 | 40.9 | 5.5 |
| 003 | PP | 35.4 | 38.6 | 3.2 | 36.6 | 37.4 | 0.8 | 37.0 | 1.6 |
| 004 | NP | 21.5 | 29.8 | 8.3 | 32.2 | 34.3 | 2.1 | 33.2 | 11.7 |
| 005 | PP | 30.2 | 38.2 | 8.0 | 38.8 | 34.6 | −4.2 | 36.7 | 6.5 |
| 006 | PP | 38.6 | 34.1 | −4.5 | 39.4 | 39.3 | −0.1 | 39.4 | 0.8 |
| 007 | PP | 32.8 | 35.9 | 3.1 | 29.9 | 36.2 | 6.3 | 33.0 | 0.2 |
| 008 | NP | 31.9 | 32.8 | 0.9 | 33.7 | 31.3 | −2.4 | 32.5 | 0.6 |
| 009 | NP | 19.1 | 20.5 | 1.4 | 27.7 | 29.2 | 1.6 | 28.4 | 9.3 |
PP, Proven pregnancy;NP, no pregnancy.
Finally, the correlation between chronological and biological age in both tissues was investigated. To first confirm the previously well-established correlation between the biological age in blood and chronological age, we analyzed the nine blood samples and found a significant correlation (r = 0.81, P = 0.0086) (Fig. 4A), which is in line with other studies of blood (Horvath 2013). Encouraged by this, we analyzed the endometrial samples. Here, a significant correlation between biological age and chronological age was also found (r = 0.80, P = 0.0093) (Fig. 4B), suggesting that the correlation is of similar magnitude in the endometrium as in blood.
Correlation between chronological age and biological age in (A) blood (r = 0.81, P = 0.0086) (B) endometrium at LH + 7 (r = 0.80, P = 0.0093).
Citation: Reproduction 155, 2; 10.1530/REP-17-0601
Discussion
By timing all biopsies to the same time point in the menstrual cycle, we show for the first time that the human endometrium appears to have a biological clock despite its strong sex hormone responsive nature.
We believe the poor correlation between the chronological age and endometrial biological age reported earlier (Horvath 2013) is a result of samples being from different time points in the menstrual cycle or disease states, as was also suggested by the author (Horvath 2013). If this is indeed the case, the implications are interesting. It could suggest that there is a systematic shift up and down in the biological age across the menstrual cycle in the endometrium. This could be a consequence of the large changes in the endometrial DNA methylation profile across the menstrual cycle (Munro et al. 2010, Houshdaran et al. 2014), which could affect methylation levels at CpG sites used by Horvath’s epigenetic clock. In addition, DNA methylation changes associated with gynecological diseases such as endometriosis (Guo 2009, Houshdaran et al. 2016, Saare et al. 2016) could also have an influence on the biological age in the endometrium. Further studies are needed to address this.
For the two youngest women in the study, the difference between biological age in the endometrium and chronological age was relatively higher than that for the remaining participants (i.e. difference of 11.7 and 9.3 years) (Fig. 4). Interestingly, this result echoes a recent finding in breast tissue, where the strongest discrepancy between biological age of breast tissue and the biological age in blood could be observed in younger women (Sehl et al. 2017). Although our sample size is too small to claim any statistical significance of this observation, it is consistent with previous findings, where the rate of the ‘ticking biological clock’ was found to be faster from birth to early adulthood, slowing down to a constant rate from around the age of 20 years (Horvath 2013).
Importantly, we found no significant change in the biological age between first and second biopsies in our study. Studies have suggested that an endometrial biopsy, or scratching, has a positive effect on pregnancy rates (Barash et al. 2003, Karimzadeh et al. 2009, Narvekar et al. 2010, Nastri et al. 2013). The molecular mechanisms, however, are unknown. Our study shows that biological age is not changed in response to a biopsy and subsequently, any fertility effect of the scratching procedure is in our opinion unlikely to be a result of biological rejuvenation of the endometrium, at least when using Horvath’s epigenetic clock.
A major strength in our study is that we succeeded in timing the biopsies to a narrow timespan of LH + 7, the assumed window of implantation in each woman’s menstrual cycle. This allows us to compare the endometrial biopsies, from an otherwise very dynamic tissue (Munro et al. 2010, Houshdaran et al. 2014). An additional strength is that each woman contributed with paired endometrial samples. A major limitation is the small sample size, and thus, replication of our results in a larger independent cohort is warranted. In conclusion, we have found that the normal human endometrium has a biological age that can be measured using Horvath’s epigenetic clock. Whether the biological age changes across the normal cycle and whether accelerated biological age plays a role in infertile women remain to be determined.
Declaration of interests
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was supported by The Engell Friis Foundation; The Jørgen W Schou and wife Else Marie Schou Foundation; The Aase and Ejnar Danielsen Foundation and The Health Research Fund of Central Denmark. The foundations had no influence on the study design, collection of samples, analysis or interpretation of data, writing the article or in the decision to submit it for publication.
Acknowledgements
The authors would like to thank the volunteers for their participation in this study.
References
Alviggi C, Humaidan P, Howles CM, Tredway D & Hillier SG 2009 Biological versus chronological ovarian age: implications for assisted reproductive technology. Reproductive Biology and Endocrinology 7 101. (https://doi.org/10.1186/1477-7827-7-101)
Baird DT, Collins J, Egozcue J, Evers LH, Gianaroli L, Leridon H, Sunde A, Templeton A, Van Steirteghem A & Cohen J et al. 2005 Fertility and ageing. Human Reproduction Update 11 261–276. (https://doi.org/10.1093/humupd/dmbib6)
Barash A, Dekel N, Fieldust S, Segal I, Schechtman E & Granot I 2003 Local injury to the endometrium doubles the incidence of successful pregnancies in patients undergoing in vitro fertilization. Fertility and Sterility 79 1317–1322. (https://doi.org/10.1016/S0015-0282(03)00345-5)
Chen BH, Marioni RE, Colicino E, Peters MJ, Ward-Caviness CK, Tsai PC, Roetker NS, Just AC, Demerath EW & Guan W et al. 2016 DNA methylation-based measures of biological age: meta-analysis predicting time to death. Aging 8 1844–1865. (https://doi.org/10.18632/aging.101020)
Christiansen L, Lenart A, Tan Q, Vaupel JW, Aviv A, McGue M & Christensen K 2016 DNA methylation age is associated with mortality in a longitudinal Danish twin study. Aging Cell 15 149–154. (https://doi.org/10.1111/acel.12421)
Dekel N, Gnainsky Y, Granot I & Mor G 2010 Inflammation and implantation. American Journal of Reproductive Immunology 63 17–21. (https://doi.org/10.1111/j.1600-0897.2009.00792.x)
Guo SW 2009 Epigenetics of endometriosis. Molecular human reproduction 15 587–607. (https://doi.org/10.1093/molehr/gap064)
Horvath S 2013 DNA methylation age of human tissues and cell types. Genome Biology 14 R115. (https://doi.org/10.1186/gb-2013-14-10-r115)
Horvath S, Erhart W, Brosch M, Ammerpohl O, von Schonfels W, Ahrens M, Heits N, Bell JT, Tsai PC & Spector TD et al. 2014 Obesity accelerates epigenetic aging of human liver. PNAS 111 15538–15543. (https://doi.org/10.1073/pnas.1412759111)
Houshdaran S, Zelenko Z, Irwin JC & Giudice LC 2014 Human endometrial DNA methylome is cycle-dependent and is associated with gene expression regulation. Molecular Endocrinology 28 1118–1135. (https://doi.org/10.1210/me.2013-1340)
Houshdaran S, Nezhat CR, Vo KC, Zelenko Z, Irwin JC & Giudice LC 2016 Aberrant Endometrial DNA Methylome and Associated Gene Expression in Endometriosis. Biology of Reproduction 95 1–16. (https://doi.org/10.1095/biolreprod.116.140434)
Karimzadeh MA, Ayazi RM & Tabibnejad N 2009 Endometrial local injury improves the pregnancy rate among recurrent implantation failure patients undergoing in vitro fertilisation/intra cytoplasmic sperm injection: a randomised clinical trial. Australian and New Zealand Journal of Obstetrics and Gynaecology 49 677–680.
Levine ME, Lu AT, Chen BH, Hernandez DG, Singleton AB, Ferrucci L, Bandinelli S, Salfati E, Manson JE & Quach A et al. 2016 Menopause accelerates biological aging. PNAS 113 9327–9332. (https://doi.org/10.1073/pnas.1604558113)
Macklon NS, Stouffer RL, Giudice LC & Fauser BC 2006 The science behind 25 years of ovarian stimulation for in vitro fertilization. Endocrine Reviews 27 170–207. (https://doi.org/10.1210/er.2005-0015)
Marioni RE, Shah S, McRae AF, Chen BH, Colicino E, Harris SE, Gibson J, Henders AK, Redmond P & Cox SR et al. 2015 DNA methylation age of blood predicts all-cause mortality in later life. Genome Biology 16 25. (https://doi.org/10.1186/s13059-015-0584-6)
Munro SK, Farquhar CM, Mitchell MD & Ponnampalam AP 2010 Epigenetic regulation of endometrium during the menstrual cycle. Molecular Human Reproduction 16 297–310. (https://doi.org/10.1093/molehr/gaq010)
Narvekar SA, Gupta N, Shetty N, Kottur A, Srinivas M & Rao KA 2010 Does local endometrial injury in the nontransfer cycle improve the IVF-ET outcome in the subsequent cycle in patients with previous unsuccessful IVF? A Randomized Controlled Pilot Study 3 15–19. (https://doi.org/10.4103/0974-1208.63116)
Nastri CO, Ferriani RA, Raine-Fenning N & Martins WP 2013 Endometrial scratching performed in the non-transfer cycle and outcome of assisted reproduction: a randomized controlled trial. Ultrasound in Obstetrics and Gynecology 42 375–382.
Noyes RW, Hertig AT & Rock J 1975 Dating the endometrial biopsy. American Journal of Obstetrics and Gynecology 122 262–263. (https://doi.org/10.1016/S0002-9378(16)33500-1)
Perna L, Zhang Y, Mons U, Holleczek B, Saum KU & Brenner H 2016 Epigenetic age acceleration predicts cancer, cardiovascular, and all-cause mortality in a German case cohort. Clinical Epigenetics 8 64. (https://doi.org/10.1186/s13148-016-0228-z)
Psychoyos A 1986 Uterine receptivity for nidation. Annals of the New York Academy of Sciences 476 36–42. (https://doi.org/10.1111/j.1749-6632.1986.tb20920.x)
Salamonsen LA 2008 Menstrual and estrous cycles. In The Endometrium Molecular, Cellular and Clinical Perspectives, pp 25–45. Eds Aplin JD, Fazleabas AT, Glasser SR, Giudice LC. UK: Informa.
Saare M, Modhukur V, Suhorutshenko M, Rajashekar B, Rekker K, Soritsa D, Karro H, Soplepmann P, Soritsa A, Lindgren CM, Rahmioglu N, Drong A, Becker CM, Zondervan KT, Salumets A & Peters M 2016 The influence of menstrual cycle and endometriosis on endometrial methylome. Clinical Epigenetics 8 2-015–0168-z. eCollection 2016. (https://doi.org/10.1186/s13148-015-0168-z)
Schmidt L, Sobotka T, Bentzen JG, Nyboe Andersen A & ESHRE Reproduction and Society Task Force 2012 Demographic and medical consequences of the postponement of parenthood. Human Reproduction Update 18 29–43. (https://doi.org/10.1093/humupd/dmr040)
Sehl ME, Henry JE, Storniolo AM, Ganz PA & Horvath S 2017 DNA methylation age is elevated in breast tissue of healthy women. Breast Cancer Research and Treatment 164 209–219. (https://doi.org/10.1007/s10549-017-4218-4)
