- Email: [email protected]
Noninvasive metabolomic profiling of embryo culture media using proton nuclear magnetic resonance correlates with reproductive potential of embryos in women undergoing in vitro fertilization
Noninvasive metabolomic profiling of embryo culture media using proton nuclear magnetic resonance correlates with reproductive potential of embryos in women undergoing in vitro fertilization Emre Seli, M.D.,a Lucy Botros, M.Sc.,b,c Denny Sakkas, Ph.D.,a,d and David H. Burns, Ph.D.b a Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, New Haven, Connecticut; and b Department of Chemistry, McGill University, Montreal, Quebec, Canada; c Molecular Biometrics LLC, Quebec, Canada; and d Molecular Biometrics LLC, New Haven, Connecticut
Objective: To identify biomarkers associated with reproductive outcome using proton nuclear magnetic resonance (1H NMR) metabolomic profiling of embryo culture media. Design: Retrospective study. Setting: An academic assisted reproductive technology (ART) program; a university research center. Patient(s): Women undergoing ART treatment. Intervention(s): Spent media samples from embryos that resulted in pregnancy and delivery (n ¼ 17) and samples (n ¼ 17) from embryos that failed to implant were individually collected on day 3, and evaluated using 1H NMR spectroscopy. The spectra obtained were quantified by integrating six biomarker signals in the aliphatic region after baseline subtraction. Using a multivariate analysis, a model that calculates a viability index for each spectrum was developed. Sensitivity and specificity of predicting pregnancy (described as implantation and delivery) were calculated. Main Outcome Measure(s): The 1H NMR metabolomic profile of embryo culture media and embryo viability. Result(s): Glutamate concentrations determined by 1H NMR were significantly higher in spent culture media of embryos that resulted in pregnancy and delivery compared to those that failed to implant. Similarly, viability indices calculated by 1H NMR using the weighted coefficients of glutamate and alanine/lactate ratio quantities were higher for embryos that implanted and resulted in a delivery. Proton NMR spectroscopy predicted viability of individual embryos with a sensitivity of 88.2% and a specificity of 88.2%. Conclusion(s): Metabolomic profile of spent embryo culture media using 1H NMR correlates with the reproductive potential of embryos. (Fertil Steril 2008;90:2183–9. 2008 by American Society for Reproductive Medicine.) Key Words: Metabolomics, proton NMR, spectroscopy, embryo viability, reproductive potential, culture media, IVF
Infertility, defined as the inability to conceive after 1 year of unprotected intercourse, is estimated to affect 15% of the reproductive age population (1). Among the treatment modalities offered to infertile couples, those using assisted reproductive technologies (ART) are associated with the highest success rates. Consequently, ART use has been increasing steadily within the past decade. In the United States, more than 125,000 ART cycles started in 2004 (2), accounting for 1% of births (2) and 18% of multiple births (3). Despite the widening use of ART, approximately 2 of 3 ART cycles fail to result in pregnancy, and 8 of 10 transferred embryos fail to implant (2, 4, 5). Failed ART cycles seem to Received April 2, 2008; revised July 5, 2008; accepted July 18, 2008. E.S. is a member of the scientific advisory board for Molecular Biometrics. L.B. has nothing to disclose. D.S. has nothing to disclose. D.H.B. has nothing to disclose. Conducted in collaboration with Molecular Biometrics, LLC, Chester, New Jersey. Reprint requests: Emre Seli, M.D., Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, 300 George Street, Suite 770J, New Haven, CT 06511 (FAX: 203-785-7134; E-mail: [email protected]).
0015-0282/08/$34.00 doi:10.1016/j.fertnstert.2008.07.1739
be at least in part due to our inability to determine the embryos with the highest reproductive potential as ART cycles using thawed embryos following failed fresh cycles result in 7%–11 % implantation rate, and 13%-17% ongoing pregnancy rate (PR) per transfer (6, 7). To prevent failures, centers have historically chosen to perform simultaneous transfer of multiple embryos, accepting the related risk of multiple pregnancies. In the United States, a mean number of 2.45 embryos were transferred in ART cycles using fresh nondonor oocytes in 2006, leading to a 34.3% live birth rate per transfer, of which 32% were multiple infant live births (2). Similarly, a mean number of 2.3 embryos were transferred in ART cycles using fresh donor oocytes, achieving a 52.3% live birth rate per transfer, 40.8% of which were multiple infant live births (2). In total, more than 30% of ART pregnancies are twins or higher-order multiple gestations, and 51% of all ART neonates are the products of multiple gestations (3), a frequency 15- to 20-fold greater than with spontaneous conceptions (8). The high multiple PRs associated with ART has significant public health consequences. The increased rate of preterm
Fertility and Sterility Vol. 90, No. 6, December 2008 Copyright ª2008 American Society for Reproductive Medicine, Published by Elsevier Inc.
2183
delivery in neonates from multiple infant pregnancies compromises their survival chances and increases their risk of lifelong disability. Although multiple births constitute only 3% of all births in the United States, they account for 13% of preterm births (<37 weeks), and 25% of very low birthweight infants (<2,000 g) (9). The incidence of cerebral palsy is increased 8-fold in twins and 47-fold in triplets (10), whereas infant deaths (birth to 1 year) are increased 6-fold in twins, and 17-fold in triplets and higher-order gestations (11). Multiple pregnancies also constitute a health risk for the mother including a 2- to 4-fold increase in pregnancy-induced hypertension and postpartum hemorrhage (12). To limit the rate of multiple pregnancies and associated complications, a number of countries have restricted the number of embryos transferred in ART cycles (5). However, the majority of countries have not yet adopted strict regulations with respect to ART practice; most probably due to high cost of treatment and patients’ desire to increase their chances to achieve pregnancy in a given cycle. Consequently, decreasing multiple gestations while maintaining or increasing overall PRs is an important goal of contemporary infertility treatment, and an improvement over the currently used embryo assessment methods would be beneficial for this purpose. The currently used embryo grading systems, largely based on embryo cleavage rate and morphology (13–18) were developed soon after the report of the first successful pregnancy after IVF (19), and the development of controlled ovarian hyperstimulation (COH) (20). Although these grading systems led to significant improvements in implantation rates and PRs (21), their accuracy remained insufficient to compel most patients and clinicians to reduce the number of embryos transferred to a point where twins are uncommon and high-order multiple gestations are rare or eliminated. The limitations of embryo assessment based on morphology and cleavage rate have led many investigators to pursue adjunctive technologies to determine an individual embryo’s reproductive potential. Several metabolic parameters of developing embryos and of the spent embryo culture media have been studied using a variety of noninvasive techniques (22). In one such study, Conaghan et al. (23) found an inverse relationship between pyruvate uptake and human embryo viability, whereas Gardner and colleagues (24) reported that glucose uptake was greatest in human blastocysts of highest grade. Brison et al. (25) determined the levels of 18 amino acids in embryo culture media, using high performance liquid chromatography (HPLC), and found that elevated asparagine and decreased glycine and leucine levels in embryo culture media correlate with pregnancy. We have recently reported that noninvasive metabolomic profiling of embryo culture media using Raman and nearinfrared (NIR) spectroscopy correlates with pregnancy outcome in women undergoing IVF (26). Our initial findings 2184
Seli et al.
1
H NMR analysis of embryo culture media
were confirmed in a prospective blinded study (27). In the current study, we applied proton nuclear magnetic resonance (1H NMR) spectroscopy to analyze the metabolomic profile of embryo culture media and to identify components of the media that correlate with reproductive potential.
MATERIALS AND METHODS Patient Selection, Treatment, and Sample Collection All patients participating in the study were treated at the Yale Fertility Center, New Haven, CT. Institutional Review Board approval was obtained before the initiation of the study. All patients undergoing IVF were considered for participation in the study. Patients with abnormal endometrial development (<6 mm in peak thickness or failure to develop a trilaminar pattern before hCG administration) were excluded. Controlled ovarian hyperstimulation was performed using a variety of protocols as previously published (26–28). Patients were monitored per established protocol and were judged to have sufficient follicular maturation when they had 2 or more follicles 18 mm or greater in maximal diameter. Oocytes were collected by transvaginal ultrasoundguided needle aspiration of the follicles under deep conscious sedation. Retrieved oocytes were rinsed, graded, and placed in N-2-hydroxyethylpiperazine-N0 -2-ethanesulfonic acid (HEPES)-buffered human tubal fluid (HTF) (Irvine Scientific, Santa Ana, CA) at 37 C under 6% CO2 in air. Conventional insemination or intracytoplasmic sperm injection (ICSI) was used as indicated. On the day after oocyte retrieval and insemination (day 1), each oocyte was examined for evidence of fertilization. Those that were found to have two pronuclei were placed into individual droplets for culture to the cleavage stage. For culture from days 1–3, 30 mL of Scandinavian G1 media (VitroLife, Englewood, CO) supplemented with 5% human serum albumin (HSA; Irvine Scientific) was used. Embryos were cultured individually. An embryo scoring system based on cleavage rate and morphology was used for the evaluation of embryo quality as previously described (15, 26, 27). The patients who did not have sufficient high quality embryos to justify extended culture (5 or more 4-cell grade 1 or 2 embryos on day 2) had their embryos selected for transfer on day 3. After removal of the embryos in preparation for transfer, the spent media were placed individually into labeled cryovials, snap frozen in liquid nitrogen, and then stored at 80 C. Upon completion of the treatment cycle, the patients were characterized relative to their implantation rates. The transferred embryos from patients who had no implantation were labeled as having a zero percent implantation rate. The transferred embryos from patients who had all transferred embryos implant and subsequently delivered were labeled as having a 100% sustained implantation rate. Vol. 90, No. 6, December 2008
Specific patients were identified for inclusion in the study population by first identifying those patients with 100% sustained implantation rates. The specimens from these patients and the next consecutive patient with a zero percent implantation rate were selected for inclusion in the study. Samples included in the study derived from women who had two or three embryos transferred. Collected samples were immediately frozen and stored at 80 C. The specimens were shipped on dry ice to the metabolomics laboratory at McGill University for analysis. NMR Spectral Acquisition Samples were thawed at room temperature for 30 minutes before analysis, and 18 mL of media was diluted in 600 mL of solvent prepared by diluting 3.7 mg of TSP (sodium 3-(trimethyl-2,2,3,3)-1-propionic acid-d4) in 250 mL of 10% D2O/90% H2O. The TSP also acted as a chemical shift reference molecule for 1H NMR analysis. The total diluted sample was transferred to a 5-mm diameter NMR tube. Free inductive decays of the culture media solutions were acquired on a 500-MHz Varian INOVA NMR spectrometer (Varian Medical Systems, Palo Alto, CA) with a HCN triple resonance cold probe having Z-axis gradients at 20.0 0.1 C. A 90 radio frequency pulse (7.5 ms) and a 20-second relaxation delay were used. To enhance low concentration signals, the intense water peak at approximately 4.8 ppm was suppressed using a pulse sequence outlined by Hwang (29), termed excitation sculpting. NMR Quantification The 360 radiofrequency pulse lengths were determined as described by Burton et al. (30) for all IVF culture media samples, where p360 ¼ 29.2 0.2 ms. As the 360 radiofrequency pulse lengths were similar for IVF samples no further processing of the signal was used. The NMR signal integral was calculated using equation 1. Ix ¼ Cx Ks
(1)
where Ix is the NMR signal integration for component x, Cx is the concentration of component x within the sample, and KS is a proportionality constant resulting from parameters of the spectrometer (31). Changes in biomarker quantities between culture media samples were computed by first subtracting out neighboring baseline contributions to their respective signals. Six biomarker integrals were then ratioed to the TSP integral as an internal standard for each sample spectrum. Multivariate Analysis Quantification of the sample properties from 1H NMR spectra consisted of determining a parsimonious combination of independent biomarker quantities that estimate pregnancy outcome by inverse least-squares regression. Fertility and Sterility
Multivariate models investigated were described by the formula: Y ¼ b 0 þ b1 X 1 þ b 2 X 2 þ . þ b K X K ;
(2)
where Y is the pregnancy outcome of each sample (1 ¼ 100% sustained implantation rate, 0 ¼ zero percent implantation rate), X1, X2, ., XK are ratios or multiplication of two biomarker integral values, or single biomarker quantities. Weighting coefficients of these independent variables are denoted by b0, b1, ., bK. The multivariate analysis was carried out using the following steps: 1. Compute all possible ratio and multiplication of two independent biomarker quantities. Using an ‘‘N choose K’’ method, 15 possible ratios and 15 possible multiplicative variables were computed between two of six biomarker integral values. 2. Compute all possible multivariate regression models. All possible regression variables (ratio, multiplicative and single quantities) were then combined in another ‘‘N choose K’’ method, where N is the number of variables available and K is the number of variables included in the model, as given by equation 2. For each possible model, the weighting coefficients (b0, b1, ., bn) were calculated by inverse least-squares regression, using known pregnancy outcomes per sample. 3. Select optimal multivariate regression model. Viability index values were determined using equation 2 for all possible multivariate regression models. A model that discriminated the most between zero percent and 100% implantation rate samples was selected by optimizing for accuracy, Accuracy ¼
TP þ TN TP þ TN þ FP þ FN
(3)
where TP is the number of true positives, TN, the number of true negatives, FP, the number of false positive, and FN, the number of false negatives. Values of TP, TN, FP, and FN were calculated by selecting the optimal threshold as determined by a receiver operator characteristic curve. 4. Repeat steps 2 and 3 by increasing variables included in model. 5. Select the number of variables to be included in the model.
Statistical Analysis Subsequent to NMR quantification and viability index estimation, notched box plots and Student’s t-test results were used to separately determine trends within the pregnancy outcome groups. Alpha error of less than 0.05 was considered significant for all comparisons. All statistical analysis was done using the Matlab (The MathWorks Inc., Natick, MA) programming package. 2185
RESULTS A total of 34 day 3 spent embryo culture media samples from 18 patients with known pregnancy outcomes (zero percent or 100% sustained implantation rates) were evaluated using 1H NMR spectroscopy. Of the 34 embryos transferred, 17 embryos (from 10 patients) implanted and led to delivery (100% sustained implantation), whereas 17 embryos (from 8 patients) did not implant (zero percent implantation).
regression. Using the weighted coefficients of glutamate and alanine-to-lactate ratio quantities (equation 2), a viability index was calculated for each sample. Mean viability index of embryos with proven reproductive potential was significantly higher (0.6201 0.1619) compared with those that failed to implant (0.3799 0.2660) (P<.002) (Fig. 2). Proton NMR spectroscopy identified implantation/pregnancy with a sensitivity of 88.2% and a specificity of 88.2%.
Proton NMR spectral measurements were made. Phase and baseline were corrected manually and key signals were assigned using previously published data (32, 33). Proton NMR spectra were quantified by integrating six biomarker signals in the aliphatic region after baseline subtraction (31). Subtracted spectral baselines and biomarker integration regions are summarized in Table 1. Integration areas were ratioed to their associated TSP signal integral. For plotting purposes, all TSP ratioed biomarker quantities were normalized to the mean ratio of the zero percent implantation group. The relative ratio changes between zero percent and 100% pregnancy outcomes are given in Figure 1.
The two groups did not differ significantly in the mean embryo score calculated as previously described (34). DISCUSSION In this study we found that noninvasive metabolomic profiling of human embryo culture media using 1H NMR spectroscopy correlates with pregnancy outcome. Our findings, together with our recent reports using other forms of spectral analysis (26, 27), suggest that in vitro cultured embryos that have a high reproductive potential alter their environment differently compared with embryos that do not result in a pregnancy.
A significant increase (P¼.002) in glutamate concentrations was detected in spent culture media of embryos that resulted in pregnancy compared with those that did not. In addition, signal integrations for alanine, pyruvate, and glucose were decreased in culture media of embryos that resulted in a pregnancy compared with those that did not; however, these only reached a confidence interval (CI) of 90%.
Metabolome is the complete array of small molecule metabolites that are found within a biological system and reflects the functional phenotype (35). Metabolomics, studies this dynamic inventory of metabolites as small molecular biomarkers using various forms of analytical approaches, in order to determine and quantify metabolites associated with physiologic and pathologic states (36).
Using the multivariate analysis approach as described in the Materials and Methods section, an optimal multivariate regression model was selected as most discriminatory between the two study groups. This model determined a linear relationship between implantation rate and quantities of glutamate and alanine-to-lactate ratio. The optimal weightings of these regions were calculated by inverse least squares
We have recently reported that metabolomic profile of spent culture media by Raman or NIR spectroscopy correlates with reproductive potential of individual embryos (26). The model developed was then tested by blinded analysis of spent embryo culture media samples collected at a different clinical center using different type and volume of embryo culture media, and correlated with pregnancy
TABLE 1 The IVF culture media biomarker chemical shifts in 1H NMR spectrum, with their respective signal multiplicity, functional group, and integrated regions. Chemical shift (multiplicity)a
Compound
Function
TSP
0.00 (s)
Si(CH3) 3
Isoleucine, leucine Lactate
0.89 (t)/0.92 (t) 1.33 (d), 4.12 (q)
CH3 CH2, CHOH
Alanine Glutamate Pyruvate a,b-Glucose
1.47 (d), 3.67 (q) 2.34 (t) 2.38 (s) 3.30–3.94 (various)
CH3 CH2 CH3 CHOH
a
Baseline subtracted (ppm) 0.23–0.28 0.23–0.28 Average of: 3.93–4.03 4.20–4.30 3.79–3.88 2.42–2.51 2.42–2.51 3.79–3.88
Integrated region (ppm) 0.01–0.08 0.40–0.96 4.00–4.25
3.62–3.71 2.31–2.37 2.37–2.40 3.71–3.81
Chemical shifts are expressed in ppm. The multiplicities are: singlet (s), doublet (d), triplet (t), quartet (q), multiplet (m).
Seli. 1H NMR analysis of embryo culture media. Fertil Steril 2008.
2186
Seli et al.
1
H NMR analysis of embryo culture media
Vol. 90, No. 6, December 2008
FIGURE 1 Box plots of the relative quantities of six individual compounds. Shaded boxes are embryos that resulted in zero percent implantation and open boxes are those that resulted in 100% implantation/delivery. P-values are for a Student’s t-test comparing the two groups for each compound. TSP ¼ sodium 3-(trimethyl-2,2,3,3)-1propionic acid-d4.
Seli. 1H NMR analysis of embryo culture media. Fertil Steril 2008.
outcome (26, 27). These studies used Raman and NIR spectroscopy for rapid noninvasive analysis of culture media with the ultimate aim of developing a technology that is suitable for clinical practice. In the current study we investigated the molecular constituents of spent embryo culture media using quantitative 1H NMR to determine which molecules are altered in association with increasing or decreasing reproductive potential of individual embryos. The NMR spectroscopy has proven to be one of the most powerful technologies for biological fluid metabolomic analysis. It is capable of studying intact tissues and fluids, producing a comprehensive profile of metabolites. An important aspect of NMR spectroscopy is that the fundamental physicochemical mechanism is completely different from other common analytical techniques and provides a different scientific perspective (37). In particular, NMR spectroscopy is distinctly different from chromatographic procedures (such as HPLC), which, in combination with mass spectrometry, are currently the most commonly used methods for identification and quantification of molecular species in a sample. Because chromatography requires thorough knowledge of the sample’s chemical properties, it is limited to the chemical dynamic range and separation methods (37), whereas NMR is limited only by magnetic field strength and does not require sample separation. Fertility and Sterility
Vibrational spectroscopic methods, such as Raman and NIR, have also been used to analyze molecular species in biological samples. Most recently we have successfully used Raman and NIR spectroscopies to assess reproductive potential of in vitro cultured embryos (26, 27). However, absolute quantification using Raman or infrared spectroscopy requires external calibration (38), whereas NMR can be used to determine relative and absolute quantities of resolved pure components without the need for external calibrations. Although vibrational spectroscopy has the advantage of being easily incorporated into a clinical setting as a noninvasive diagnostic tool, NMR has the advantage of identifying and quantifying specific metabolites simultaneously (39–41). Additional studies will be needed to determine whether 1H NMR spectral regions of embryo culture media identified as correlating with reproductive potential of embryos can be validated by Raman or NIR, and to determine whether a combination of these modalities may provide increased diagnostic ability. In the current study, using 1H NMR spectral integration, we found that glutamate levels in spent culture media samples of embryos that result in pregnancy are higher compared with those that fail to implant (Fig. 1). Glutamate may be generated from a-ketoglutarate and ammonium in a transamination reaction catalyzed by glutamate dehydrogenase in the presence 2187
FIGURE 2 Box plots of viability indices calculated from 1H NMR of spent IVF culture media. Shaded boxes are embryos that resulted in zero percent implantation and open boxes are those that resulted in 100% implantation/delivery (P< .01).
Seli. 1H NMR analysis of embryo culture media. Fertil Steril 2008.
of reduced nicotinamide adenine dinucleotide (42). Therefore, the elevated levels of glutamate in culture media of embryos with higher reproductive potential may be a reflection of their ability to lower the levels of potentially detrimental ammonium in culture media better than nonviable embryos (42). We also observed a trend toward decreased alanine levels in the culture media of embryos that resulted in pregnancy (Fig. 1). Our findings are consistent with those of Houghton et al. (43) who showed that alanine production from days 2–3 of development is significantly lower in culture media of embryos that developed into blastocysts compared with those that arrested. Also, consistent with our findings, they observed a reverse trend toward increased glutamate production by cleavage stage embryos that developed into blastocysts (43). It is noteworthy that these investigators measured in vitro embryo development assessed by morphology as their main outcome. More recently, Brison et al. (25) used HPLC to evaluate amino acid concentrations in spent embryo culture media and found elevated asparagine, and decreased glycine and leucine levels to correlate with pregnancy. The difference between our findings and theirs can be explained by the fact that they used a different culture media, and a significantly smaller culture volume (4 mL). In addition, embryos were cultured for only 24 hours before embryo transfer on day 2 after retrieval (25), whereas in our study, embryos were cultured for 2 days before transfer on day 3 (25). In addition, the differences among our findings suggest that observed associations between amino acid levels in culture media and pregnancy outcome may be dependent on the type and volume of 2188
Seli et al.
1
H NMR analysis of embryo culture media
culture media and the day of transfer. Nevertheless, within our culture system, glutamate emerges as the small molecule that most significantly correlates with reproductive outcome. During early cleavage, pyruvate, rather than glucose, is the predominant energy substrate for embryos (44). A switch to glucose as the preferred energy substrate occurs at the 8cell/morula stage, and glucose becomes the primary energy source at the blastocyst stage (45). Although not statistically significant, we observed a trend toward lower glucose in culture media of embryos that resulted in pregnancy (Fig. 1). This finding could be reflective of the earlier expression of mediators involved in glycolysis and glucose transport in embryos with higher reproductive potential. We also observed a trend toward lower pyruvate levels in day 3 culture media of embryos that resulted in pregnancy compared with those that failed to implant (Fig. 1). Our findings are consistent with previous reports by Hardy et al. (46) and Gott et al. (47) who showed that pyruvate and, to a lesser extent, glucose uptake is higher on day 3 embryos that develop into blastocysts compared with those that arrest at cleavage stage. However, our findings are in conflict with another study from the same group that reported that pyruvate uptake was higher in day 2 and day 3 embryos that failed to implant (23). We agree with their conclusions that the observed differences among studies may be due to culture media differences and methodologies applied, and that the range of individual values in study groups would prevent selection of embryos for transfer on the basis of pyruvate uptake alone (23). This conclusion strengthens the view that a model (Fig. 2) examining the input of various components is better able to provide discrimination between the individual reproductive potential of embryos. Our current findings, together with our recent reports using Raman and NIR spectroscopy (26, 27), suggest that metabolomic assessment of spent embryo culture media may provide additional information that may help better assess the reproductive potential of individual embryos in ART. A major benefit of an improved embryo selection methodology would be an increase in implantation rates and PRs. Such a benefit would lead to a decrease in the cost, as well as in the physical and psychological morbidity, associated with prolonged infertility treatment. We could also speculate that improved embryo selection would promote the transfer of a lower number of embryos and ultimately contribute to a decrease in multiple infant births that result from ART. To determine whether such effects will ensue, randomized prospective trials comparing the current embryo selection methodology based on cleavage rate and morphology, to embryo selection based on spectroscopy alone or in combination with morphologic evaluation will have to be conducted.
REFERENCES 1. Mosher WD, Pratt WF. Fecundity and infertility in the United Sates: incidence and trends. Fertil Steril 1991;56:192–3. 2. Society for Assisted Reproductive Technology (SART). Assisted reproductive technology success rates. National summary and fertility clinic reports. Centers for Disease Control, Atlanta, GA, 2006.
Vol. 90, No. 6, December 2008
3. Wright V, Chang J, Jeng G, Macaluso M. Assisted reproductive technology surveillance—United States, 2003. MMWR Surveil Summ 2006;55:1–22. 4. Kovalevsky G, Patrizio P. High rates of embryo wastage with the use of assisted reproductive technology: a look at the trends between 1995 and 2001 in the United States. Fertil Steril 2005;84:325–30. 5. Bromer JG, Seli E. Assessment of embryo viability in assisted reproductive technologies: shortcomings of current approaches and the emerging role of metabolomics. Curr Opin Obstet Gynecol 2008;20:234–41. 6. Toner JP, Veeck L, Acosta AA, Muasher SJ. Predictive value of pregnancy during original in vitro fertilization cycle on implantation and pregnancy in subsequent cryothaw cycles. Fertil Steril 1991;56:505–8. 7. El-Thouky T, Khalaf Y, Al-Darazi K, O’Mahony F, Wharf E, Taylor A, et al. Cryo-thawed embryos obtained from conception cycles have double the implantation and pregnancy potential of those from successful cycles. Hum Reprod 2003;6:1313–8. 8. Reddy UM, Wapner RJ, Rebar RW, Tasca RJ. Infertility, assisted reproductive technology, and adverse pregnancy outcomes. Obstet Gynecol 2007;109:967–77. 9. Luke B, Martin J. The rise in multiple births in the United States: who, what, when, where, and why. Clin Obstet Gynecol 2004;47:118–33. 10. Keith LG, Oleszczuk JJ, Keith DM. Multiple gestation: reflections on epidemiology, causes, and consequences. Int J Fertil Womens Med 2000;45:206–14. 11. US Department of Health and Human Services PHS. US vital statistics, 1998 and from US Public Health Service. Healthy People 2000: National Health Promotion and Disease Prevention Objectives, DHHS Pub No. (PHS) 90–50212. 12. Luke B, Brown MB. Contemporary risks of maternal morbidity and adverse outcomes with increasing maternal age and plurality. Fertil Steril 2007;88:283–93. 13. Scott LA, Smith S. The successful use of pronuclear embryo transfer the day following oocyte retrieval. Hum Reprod 1998;13:1003–13. 14. Tesarik J, Greco E. The probability of abnormal preimplantation development can be predicted by a single static observation on pronuclear state morphology. Hum Reprod 1999;14:1318–23. 15. Veeck L. An atlas of human gametes and conceptuses: an illustrated reference for assisted reproductive technology. New York: Parthenon Publishing, 1999. 16. Gerris J, De Neubourg D, Mangelschots K, Van Royen E, Van de Meerssche M, Valkenburg M. Prevention of twin pregnancy after in-vitro fertilization or intracytoplasmic sperm injection based on strict embryo criteria: a prospective randomized clinical trial. Hum Reprod 1999;14: 2581–7. 17. VanRoyan E, Mangelschots K, De Neubourg D, Valkenburg M, Van de Meerssche M, Ryckaert G, et al. Characterization of a top quality embryo, a step towards single-embryo transfer. Hum Reprod 1999;14:2345–9. 18. Gardner DK, Schoolcraft WB. In vitro culture of human blastocysts. In: Jansen R, ed. Towards reproductive certainty: fertility and genetics beyond. Carnforth: Parthenon Publishing, 1999:378–88. 19. Steptoe PC, Edwards RG. Birth after the reimplantation of a human embryo. Lancet 1978;2:366. 20. Trounson A, Leeton J, Wood C, Webb J, Wood J. Pregnancies in humans by fertilization in vitro and embryo transfer in the controlled ovulatory cycle. Science 1981;216:681–2. 21. Toner JP. Progress we can be proud of: U.S. trends in assisted reproduction over the first 20 years. Fertil Steril 2002;78:943–50. 22. Sakkas D, Gardner DK. Noninvasive methods to assess embryo quality. Curr Opin Obstet Gynecol 2005;17:283–8. 23. Conaghan J, Hardy K, Handyside A, Winston RML, Leese HJ. Selection criteria for human embryo transfer: a comparison of pyruvate uptake and morphology. J Assist Reprod Genet 1993;10:21–30. 24. Gardner DK, Lane M, Stevens J, Schoolcraft WB. Noninvasive assessment of human embryo nutrient consumption as a measure of developmental potential. Fertil Steril 2001;76:1175–80. 25. Brison DR, Houghton FD, Falconer D, Roberts SA, Hawkhead J, Humpherson PG, et al. Identification of viable embryos in IVF by noninvasive measurement of amino acid turnover. Hum Reprod 2004;19: 2319–24.
Fertility and Sterility
26. Seli E, Sakkas D, Scott R, Kwok JS, Rosendahl S, Burns DH. Non-Invasive metabolomic profiling of human embryo culture media using Raman and near infrared spectroscopy correlates with reproductive potential of embryos in women undergoing in vitro fertilization. Fertil Steril 2007;88:1350–7. 27. Scott RT, Seli E, Miller K, Sakkas D, Scott K, Burns DH. Non-Invasive metabolomic profiling of human embryo culture media using Raman spectroscopy predicts embryonic reproductive potential: a prospective blinded pilot study. Fertil Steril 2008;90:77–83. 28. Muasher SJ, Abdallah RT, Hubayter ZR. Optimal stimulation protocols for in vitro fertilization. Fertil Steril 2006;86:267–73. 29. Hwang AJS. Water suppression that works. excitation sculpting using arbitrary waveforms and pulsed field gradients. J Magnetic Res 1995;A 112:275–9. 30. Burton IW, Quilliam MA, Walter JA. Quantitative 1H NMR with external standards: use in preparation of calibration solutions for algal toxins and other natural products. Anal Chem 2005;776:3123–31. 31. Maltz F. Validation of quantitative NMR. J Pharm Biomed Anal 2005;38: 813–23. 32. Groenen PM, Engelke UF, Wevers RA, Hendriks JC, Eskes TK, Merkus HM, et al. High-resolution 1H NMR spectroscopy of amniotic fluids from spina bifida fetuses and controls. Eur J Obstet Gynecol Reprod Biol 2004;112:16–23. 33. Le Moyec L, Muller F, Eugene M, Spraul M. Proton magnetic resonance spectroscopy of human amniotic fluids sampled at 17-18 weeks of pregnancy in cases of decreased digestive enzyme activities and detected cystic fibrosis. Clin Biochem 1994;27:475–83. 34. Cummins JM, Breen TM, Harrison KL, Shaw JM, Wilson LM, Hennessey JF. A formula for scoring human embryo growth rates in in vitro fertilization: its value in predicting pregnancy and in comparison with visual estimates of embryo quality. J In Vitro Fert Embryo Transf 1986;3:284–95. 35. Oliver SG, Winson MK, Kell DB, Baganz F. Systematic functional analysis of the yeast genome. Trends Biotechnol 1998;16:373–8. 36. Ellis DI, Goodacre R. Metabolic fingerprinting in disease diagnosis: biomedical applications of infrared and Raman spectroscopy. Analyst 2006;131:875–85. 37. Pauli GF, Jaki BU, Lankin DC. Quantitative 1H NMR: development and potential of a method for natural products analysis. J Nat Prod 2005;68: 133–49. 38. Kan-Zhi Liu M, Mantsch HH. Simultaneous quantitation from infrared spectra of glucose concentrations lactate concentrations, and lecithin/ sphingomyelin ratios in amniotic fluid. Am J Obstet Gynecol 1999;180: 696–702. 39. Bock JL. Metabolic profiling of amniotic fluid by proton nuclear magnetic resonance spectroscopy: correlation with fetal maturation and other clinical variables. Clinical Chemistry 1994;40:56–61. 40. Szantay CJ. High-field NMR spectroscopy as an analytical tool for quantitative determinations: pitfalls, limitations, and possible solutions. Trends Anal Chem 1992;11:332–44. 41. Gruetter R, Weisdorf SA, Rajanayagan V, Terpstra M, Merkle H, Truwit CL, et al. Resolution improvements in in vivo 1H NMR spectra with increased magnetic field strength. J Magn Reson 1998;135:260–4. 42. Gardner DK, Lane M. Amino acids and ammonium regulate mouse embryo development in culture. Biol Reprod 1993;48:377–85. 43. Houghton FD, Hawkhead JA, Humpherson PG, Hogg JE, Balen AH, Rutherford AJ, et al. Non-invasive amino acid turnover predicts human embryo developmental capacity. Hum Reprod 2002;17:999–1005. 44. Brinster RL. Studies on the development of the mouse embryos in vitro. II. The effect of energy source. J Exp Zool 1965;158:59–68. 45. Gardner DK, Leese HJ. Non-invasive measurement of nutrient uptake by single cultured preimplantation mouse embryos. Hum Reprod 1986;1:25–7. 46. Hardy K, Hooper MAK, Handyside AH, Rutherford AJ, Winston RML, Leese HJ. Non-invasive measurement of glucose and pyruvate uptake by individual human oocytes and preimplantation embryos. Hum Reprod 1989;4:188–91. 47. Gott AL, Hardy K, Winston RML, Leese HJ. Non-invasive measurement of pyruvate and glucose uptake and lactate production by single human preimplantation embryos. Hum Reprod 1990;5:104–8.
2189
Objective: To identify biomarkers associated with reproductive outcome using proton nuclear magnetic resonance (1H NMR) metabolomic profiling of embryo culture media. Design: Retrospective study. Setting: An academic assisted reproductive technology (ART) program; a university research center. Patient(s): Women undergoing ART treatment. Intervention(s): Spent media samples from embryos that resulted in pregnancy and delivery (n ¼ 17) and samples (n ¼ 17) from embryos that failed to implant were individually collected on day 3, and evaluated using 1H NMR spectroscopy. The spectra obtained were quantified by integrating six biomarker signals in the aliphatic region after baseline subtraction. Using a multivariate analysis, a model that calculates a viability index for each spectrum was developed. Sensitivity and specificity of predicting pregnancy (described as implantation and delivery) were calculated. Main Outcome Measure(s): The 1H NMR metabolomic profile of embryo culture media and embryo viability. Result(s): Glutamate concentrations determined by 1H NMR were significantly higher in spent culture media of embryos that resulted in pregnancy and delivery compared to those that failed to implant. Similarly, viability indices calculated by 1H NMR using the weighted coefficients of glutamate and alanine/lactate ratio quantities were higher for embryos that implanted and resulted in a delivery. Proton NMR spectroscopy predicted viability of individual embryos with a sensitivity of 88.2% and a specificity of 88.2%. Conclusion(s): Metabolomic profile of spent embryo culture media using 1H NMR correlates with the reproductive potential of embryos. (Fertil Steril 2008;90:2183–9. 2008 by American Society for Reproductive Medicine.) Key Words: Metabolomics, proton NMR, spectroscopy, embryo viability, reproductive potential, culture media, IVF
Infertility, defined as the inability to conceive after 1 year of unprotected intercourse, is estimated to affect 15% of the reproductive age population (1). Among the treatment modalities offered to infertile couples, those using assisted reproductive technologies (ART) are associated with the highest success rates. Consequently, ART use has been increasing steadily within the past decade. In the United States, more than 125,000 ART cycles started in 2004 (2), accounting for 1% of births (2) and 18% of multiple births (3). Despite the widening use of ART, approximately 2 of 3 ART cycles fail to result in pregnancy, and 8 of 10 transferred embryos fail to implant (2, 4, 5). Failed ART cycles seem to Received April 2, 2008; revised July 5, 2008; accepted July 18, 2008. E.S. is a member of the scientific advisory board for Molecular Biometrics. L.B. has nothing to disclose. D.S. has nothing to disclose. D.H.B. has nothing to disclose. Conducted in collaboration with Molecular Biometrics, LLC, Chester, New Jersey. Reprint requests: Emre Seli, M.D., Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, 300 George Street, Suite 770J, New Haven, CT 06511 (FAX: 203-785-7134; E-mail: [email protected]).
0015-0282/08/$34.00 doi:10.1016/j.fertnstert.2008.07.1739
be at least in part due to our inability to determine the embryos with the highest reproductive potential as ART cycles using thawed embryos following failed fresh cycles result in 7%–11 % implantation rate, and 13%-17% ongoing pregnancy rate (PR) per transfer (6, 7). To prevent failures, centers have historically chosen to perform simultaneous transfer of multiple embryos, accepting the related risk of multiple pregnancies. In the United States, a mean number of 2.45 embryos were transferred in ART cycles using fresh nondonor oocytes in 2006, leading to a 34.3% live birth rate per transfer, of which 32% were multiple infant live births (2). Similarly, a mean number of 2.3 embryos were transferred in ART cycles using fresh donor oocytes, achieving a 52.3% live birth rate per transfer, 40.8% of which were multiple infant live births (2). In total, more than 30% of ART pregnancies are twins or higher-order multiple gestations, and 51% of all ART neonates are the products of multiple gestations (3), a frequency 15- to 20-fold greater than with spontaneous conceptions (8). The high multiple PRs associated with ART has significant public health consequences. The increased rate of preterm
Fertility and Sterility Vol. 90, No. 6, December 2008 Copyright ª2008 American Society for Reproductive Medicine, Published by Elsevier Inc.
2183
delivery in neonates from multiple infant pregnancies compromises their survival chances and increases their risk of lifelong disability. Although multiple births constitute only 3% of all births in the United States, they account for 13% of preterm births (<37 weeks), and 25% of very low birthweight infants (<2,000 g) (9). The incidence of cerebral palsy is increased 8-fold in twins and 47-fold in triplets (10), whereas infant deaths (birth to 1 year) are increased 6-fold in twins, and 17-fold in triplets and higher-order gestations (11). Multiple pregnancies also constitute a health risk for the mother including a 2- to 4-fold increase in pregnancy-induced hypertension and postpartum hemorrhage (12). To limit the rate of multiple pregnancies and associated complications, a number of countries have restricted the number of embryos transferred in ART cycles (5). However, the majority of countries have not yet adopted strict regulations with respect to ART practice; most probably due to high cost of treatment and patients’ desire to increase their chances to achieve pregnancy in a given cycle. Consequently, decreasing multiple gestations while maintaining or increasing overall PRs is an important goal of contemporary infertility treatment, and an improvement over the currently used embryo assessment methods would be beneficial for this purpose. The currently used embryo grading systems, largely based on embryo cleavage rate and morphology (13–18) were developed soon after the report of the first successful pregnancy after IVF (19), and the development of controlled ovarian hyperstimulation (COH) (20). Although these grading systems led to significant improvements in implantation rates and PRs (21), their accuracy remained insufficient to compel most patients and clinicians to reduce the number of embryos transferred to a point where twins are uncommon and high-order multiple gestations are rare or eliminated. The limitations of embryo assessment based on morphology and cleavage rate have led many investigators to pursue adjunctive technologies to determine an individual embryo’s reproductive potential. Several metabolic parameters of developing embryos and of the spent embryo culture media have been studied using a variety of noninvasive techniques (22). In one such study, Conaghan et al. (23) found an inverse relationship between pyruvate uptake and human embryo viability, whereas Gardner and colleagues (24) reported that glucose uptake was greatest in human blastocysts of highest grade. Brison et al. (25) determined the levels of 18 amino acids in embryo culture media, using high performance liquid chromatography (HPLC), and found that elevated asparagine and decreased glycine and leucine levels in embryo culture media correlate with pregnancy. We have recently reported that noninvasive metabolomic profiling of embryo culture media using Raman and nearinfrared (NIR) spectroscopy correlates with pregnancy outcome in women undergoing IVF (26). Our initial findings 2184
Seli et al.
1
H NMR analysis of embryo culture media
were confirmed in a prospective blinded study (27). In the current study, we applied proton nuclear magnetic resonance (1H NMR) spectroscopy to analyze the metabolomic profile of embryo culture media and to identify components of the media that correlate with reproductive potential.
MATERIALS AND METHODS Patient Selection, Treatment, and Sample Collection All patients participating in the study were treated at the Yale Fertility Center, New Haven, CT. Institutional Review Board approval was obtained before the initiation of the study. All patients undergoing IVF were considered for participation in the study. Patients with abnormal endometrial development (<6 mm in peak thickness or failure to develop a trilaminar pattern before hCG administration) were excluded. Controlled ovarian hyperstimulation was performed using a variety of protocols as previously published (26–28). Patients were monitored per established protocol and were judged to have sufficient follicular maturation when they had 2 or more follicles 18 mm or greater in maximal diameter. Oocytes were collected by transvaginal ultrasoundguided needle aspiration of the follicles under deep conscious sedation. Retrieved oocytes were rinsed, graded, and placed in N-2-hydroxyethylpiperazine-N0 -2-ethanesulfonic acid (HEPES)-buffered human tubal fluid (HTF) (Irvine Scientific, Santa Ana, CA) at 37 C under 6% CO2 in air. Conventional insemination or intracytoplasmic sperm injection (ICSI) was used as indicated. On the day after oocyte retrieval and insemination (day 1), each oocyte was examined for evidence of fertilization. Those that were found to have two pronuclei were placed into individual droplets for culture to the cleavage stage. For culture from days 1–3, 30 mL of Scandinavian G1 media (VitroLife, Englewood, CO) supplemented with 5% human serum albumin (HSA; Irvine Scientific) was used. Embryos were cultured individually. An embryo scoring system based on cleavage rate and morphology was used for the evaluation of embryo quality as previously described (15, 26, 27). The patients who did not have sufficient high quality embryos to justify extended culture (5 or more 4-cell grade 1 or 2 embryos on day 2) had their embryos selected for transfer on day 3. After removal of the embryos in preparation for transfer, the spent media were placed individually into labeled cryovials, snap frozen in liquid nitrogen, and then stored at 80 C. Upon completion of the treatment cycle, the patients were characterized relative to their implantation rates. The transferred embryos from patients who had no implantation were labeled as having a zero percent implantation rate. The transferred embryos from patients who had all transferred embryos implant and subsequently delivered were labeled as having a 100% sustained implantation rate. Vol. 90, No. 6, December 2008
Specific patients were identified for inclusion in the study population by first identifying those patients with 100% sustained implantation rates. The specimens from these patients and the next consecutive patient with a zero percent implantation rate were selected for inclusion in the study. Samples included in the study derived from women who had two or three embryos transferred. Collected samples were immediately frozen and stored at 80 C. The specimens were shipped on dry ice to the metabolomics laboratory at McGill University for analysis. NMR Spectral Acquisition Samples were thawed at room temperature for 30 minutes before analysis, and 18 mL of media was diluted in 600 mL of solvent prepared by diluting 3.7 mg of TSP (sodium 3-(trimethyl-2,2,3,3)-1-propionic acid-d4) in 250 mL of 10% D2O/90% H2O. The TSP also acted as a chemical shift reference molecule for 1H NMR analysis. The total diluted sample was transferred to a 5-mm diameter NMR tube. Free inductive decays of the culture media solutions were acquired on a 500-MHz Varian INOVA NMR spectrometer (Varian Medical Systems, Palo Alto, CA) with a HCN triple resonance cold probe having Z-axis gradients at 20.0 0.1 C. A 90 radio frequency pulse (7.5 ms) and a 20-second relaxation delay were used. To enhance low concentration signals, the intense water peak at approximately 4.8 ppm was suppressed using a pulse sequence outlined by Hwang (29), termed excitation sculpting. NMR Quantification The 360 radiofrequency pulse lengths were determined as described by Burton et al. (30) for all IVF culture media samples, where p360 ¼ 29.2 0.2 ms. As the 360 radiofrequency pulse lengths were similar for IVF samples no further processing of the signal was used. The NMR signal integral was calculated using equation 1. Ix ¼ Cx Ks
(1)
where Ix is the NMR signal integration for component x, Cx is the concentration of component x within the sample, and KS is a proportionality constant resulting from parameters of the spectrometer (31). Changes in biomarker quantities between culture media samples were computed by first subtracting out neighboring baseline contributions to their respective signals. Six biomarker integrals were then ratioed to the TSP integral as an internal standard for each sample spectrum. Multivariate Analysis Quantification of the sample properties from 1H NMR spectra consisted of determining a parsimonious combination of independent biomarker quantities that estimate pregnancy outcome by inverse least-squares regression. Fertility and Sterility
Multivariate models investigated were described by the formula: Y ¼ b 0 þ b1 X 1 þ b 2 X 2 þ . þ b K X K ;
(2)
where Y is the pregnancy outcome of each sample (1 ¼ 100% sustained implantation rate, 0 ¼ zero percent implantation rate), X1, X2, ., XK are ratios or multiplication of two biomarker integral values, or single biomarker quantities. Weighting coefficients of these independent variables are denoted by b0, b1, ., bK. The multivariate analysis was carried out using the following steps: 1. Compute all possible ratio and multiplication of two independent biomarker quantities. Using an ‘‘N choose K’’ method, 15 possible ratios and 15 possible multiplicative variables were computed between two of six biomarker integral values. 2. Compute all possible multivariate regression models. All possible regression variables (ratio, multiplicative and single quantities) were then combined in another ‘‘N choose K’’ method, where N is the number of variables available and K is the number of variables included in the model, as given by equation 2. For each possible model, the weighting coefficients (b0, b1, ., bn) were calculated by inverse least-squares regression, using known pregnancy outcomes per sample. 3. Select optimal multivariate regression model. Viability index values were determined using equation 2 for all possible multivariate regression models. A model that discriminated the most between zero percent and 100% implantation rate samples was selected by optimizing for accuracy, Accuracy ¼
TP þ TN TP þ TN þ FP þ FN
(3)
where TP is the number of true positives, TN, the number of true negatives, FP, the number of false positive, and FN, the number of false negatives. Values of TP, TN, FP, and FN were calculated by selecting the optimal threshold as determined by a receiver operator characteristic curve. 4. Repeat steps 2 and 3 by increasing variables included in model. 5. Select the number of variables to be included in the model.
Statistical Analysis Subsequent to NMR quantification and viability index estimation, notched box plots and Student’s t-test results were used to separately determine trends within the pregnancy outcome groups. Alpha error of less than 0.05 was considered significant for all comparisons. All statistical analysis was done using the Matlab (The MathWorks Inc., Natick, MA) programming package. 2185
RESULTS A total of 34 day 3 spent embryo culture media samples from 18 patients with known pregnancy outcomes (zero percent or 100% sustained implantation rates) were evaluated using 1H NMR spectroscopy. Of the 34 embryos transferred, 17 embryos (from 10 patients) implanted and led to delivery (100% sustained implantation), whereas 17 embryos (from 8 patients) did not implant (zero percent implantation).
regression. Using the weighted coefficients of glutamate and alanine-to-lactate ratio quantities (equation 2), a viability index was calculated for each sample. Mean viability index of embryos with proven reproductive potential was significantly higher (0.6201 0.1619) compared with those that failed to implant (0.3799 0.2660) (P<.002) (Fig. 2). Proton NMR spectroscopy identified implantation/pregnancy with a sensitivity of 88.2% and a specificity of 88.2%.
Proton NMR spectral measurements were made. Phase and baseline were corrected manually and key signals were assigned using previously published data (32, 33). Proton NMR spectra were quantified by integrating six biomarker signals in the aliphatic region after baseline subtraction (31). Subtracted spectral baselines and biomarker integration regions are summarized in Table 1. Integration areas were ratioed to their associated TSP signal integral. For plotting purposes, all TSP ratioed biomarker quantities were normalized to the mean ratio of the zero percent implantation group. The relative ratio changes between zero percent and 100% pregnancy outcomes are given in Figure 1.
The two groups did not differ significantly in the mean embryo score calculated as previously described (34). DISCUSSION In this study we found that noninvasive metabolomic profiling of human embryo culture media using 1H NMR spectroscopy correlates with pregnancy outcome. Our findings, together with our recent reports using other forms of spectral analysis (26, 27), suggest that in vitro cultured embryos that have a high reproductive potential alter their environment differently compared with embryos that do not result in a pregnancy.
A significant increase (P¼.002) in glutamate concentrations was detected in spent culture media of embryos that resulted in pregnancy compared with those that did not. In addition, signal integrations for alanine, pyruvate, and glucose were decreased in culture media of embryos that resulted in a pregnancy compared with those that did not; however, these only reached a confidence interval (CI) of 90%.
Metabolome is the complete array of small molecule metabolites that are found within a biological system and reflects the functional phenotype (35). Metabolomics, studies this dynamic inventory of metabolites as small molecular biomarkers using various forms of analytical approaches, in order to determine and quantify metabolites associated with physiologic and pathologic states (36).
Using the multivariate analysis approach as described in the Materials and Methods section, an optimal multivariate regression model was selected as most discriminatory between the two study groups. This model determined a linear relationship between implantation rate and quantities of glutamate and alanine-to-lactate ratio. The optimal weightings of these regions were calculated by inverse least squares
We have recently reported that metabolomic profile of spent culture media by Raman or NIR spectroscopy correlates with reproductive potential of individual embryos (26). The model developed was then tested by blinded analysis of spent embryo culture media samples collected at a different clinical center using different type and volume of embryo culture media, and correlated with pregnancy
TABLE 1 The IVF culture media biomarker chemical shifts in 1H NMR spectrum, with their respective signal multiplicity, functional group, and integrated regions. Chemical shift (multiplicity)a
Compound
Function
TSP
0.00 (s)
Si(CH3) 3
Isoleucine, leucine Lactate
0.89 (t)/0.92 (t) 1.33 (d), 4.12 (q)
CH3 CH2, CHOH
Alanine Glutamate Pyruvate a,b-Glucose
1.47 (d), 3.67 (q) 2.34 (t) 2.38 (s) 3.30–3.94 (various)
CH3 CH2 CH3 CHOH
a
Baseline subtracted (ppm) 0.23–0.28 0.23–0.28 Average of: 3.93–4.03 4.20–4.30 3.79–3.88 2.42–2.51 2.42–2.51 3.79–3.88
Integrated region (ppm) 0.01–0.08 0.40–0.96 4.00–4.25
3.62–3.71 2.31–2.37 2.37–2.40 3.71–3.81
Chemical shifts are expressed in ppm. The multiplicities are: singlet (s), doublet (d), triplet (t), quartet (q), multiplet (m).
Seli. 1H NMR analysis of embryo culture media. Fertil Steril 2008.
2186
Seli et al.
1
H NMR analysis of embryo culture media
Vol. 90, No. 6, December 2008
FIGURE 1 Box plots of the relative quantities of six individual compounds. Shaded boxes are embryos that resulted in zero percent implantation and open boxes are those that resulted in 100% implantation/delivery. P-values are for a Student’s t-test comparing the two groups for each compound. TSP ¼ sodium 3-(trimethyl-2,2,3,3)-1propionic acid-d4.
Seli. 1H NMR analysis of embryo culture media. Fertil Steril 2008.
outcome (26, 27). These studies used Raman and NIR spectroscopy for rapid noninvasive analysis of culture media with the ultimate aim of developing a technology that is suitable for clinical practice. In the current study we investigated the molecular constituents of spent embryo culture media using quantitative 1H NMR to determine which molecules are altered in association with increasing or decreasing reproductive potential of individual embryos. The NMR spectroscopy has proven to be one of the most powerful technologies for biological fluid metabolomic analysis. It is capable of studying intact tissues and fluids, producing a comprehensive profile of metabolites. An important aspect of NMR spectroscopy is that the fundamental physicochemical mechanism is completely different from other common analytical techniques and provides a different scientific perspective (37). In particular, NMR spectroscopy is distinctly different from chromatographic procedures (such as HPLC), which, in combination with mass spectrometry, are currently the most commonly used methods for identification and quantification of molecular species in a sample. Because chromatography requires thorough knowledge of the sample’s chemical properties, it is limited to the chemical dynamic range and separation methods (37), whereas NMR is limited only by magnetic field strength and does not require sample separation. Fertility and Sterility
Vibrational spectroscopic methods, such as Raman and NIR, have also been used to analyze molecular species in biological samples. Most recently we have successfully used Raman and NIR spectroscopies to assess reproductive potential of in vitro cultured embryos (26, 27). However, absolute quantification using Raman or infrared spectroscopy requires external calibration (38), whereas NMR can be used to determine relative and absolute quantities of resolved pure components without the need for external calibrations. Although vibrational spectroscopy has the advantage of being easily incorporated into a clinical setting as a noninvasive diagnostic tool, NMR has the advantage of identifying and quantifying specific metabolites simultaneously (39–41). Additional studies will be needed to determine whether 1H NMR spectral regions of embryo culture media identified as correlating with reproductive potential of embryos can be validated by Raman or NIR, and to determine whether a combination of these modalities may provide increased diagnostic ability. In the current study, using 1H NMR spectral integration, we found that glutamate levels in spent culture media samples of embryos that result in pregnancy are higher compared with those that fail to implant (Fig. 1). Glutamate may be generated from a-ketoglutarate and ammonium in a transamination reaction catalyzed by glutamate dehydrogenase in the presence 2187
FIGURE 2 Box plots of viability indices calculated from 1H NMR of spent IVF culture media. Shaded boxes are embryos that resulted in zero percent implantation and open boxes are those that resulted in 100% implantation/delivery (P< .01).
Seli. 1H NMR analysis of embryo culture media. Fertil Steril 2008.
of reduced nicotinamide adenine dinucleotide (42). Therefore, the elevated levels of glutamate in culture media of embryos with higher reproductive potential may be a reflection of their ability to lower the levels of potentially detrimental ammonium in culture media better than nonviable embryos (42). We also observed a trend toward decreased alanine levels in the culture media of embryos that resulted in pregnancy (Fig. 1). Our findings are consistent with those of Houghton et al. (43) who showed that alanine production from days 2–3 of development is significantly lower in culture media of embryos that developed into blastocysts compared with those that arrested. Also, consistent with our findings, they observed a reverse trend toward increased glutamate production by cleavage stage embryos that developed into blastocysts (43). It is noteworthy that these investigators measured in vitro embryo development assessed by morphology as their main outcome. More recently, Brison et al. (25) used HPLC to evaluate amino acid concentrations in spent embryo culture media and found elevated asparagine, and decreased glycine and leucine levels to correlate with pregnancy. The difference between our findings and theirs can be explained by the fact that they used a different culture media, and a significantly smaller culture volume (4 mL). In addition, embryos were cultured for only 24 hours before embryo transfer on day 2 after retrieval (25), whereas in our study, embryos were cultured for 2 days before transfer on day 3 (25). In addition, the differences among our findings suggest that observed associations between amino acid levels in culture media and pregnancy outcome may be dependent on the type and volume of 2188
Seli et al.
1
H NMR analysis of embryo culture media
culture media and the day of transfer. Nevertheless, within our culture system, glutamate emerges as the small molecule that most significantly correlates with reproductive outcome. During early cleavage, pyruvate, rather than glucose, is the predominant energy substrate for embryos (44). A switch to glucose as the preferred energy substrate occurs at the 8cell/morula stage, and glucose becomes the primary energy source at the blastocyst stage (45). Although not statistically significant, we observed a trend toward lower glucose in culture media of embryos that resulted in pregnancy (Fig. 1). This finding could be reflective of the earlier expression of mediators involved in glycolysis and glucose transport in embryos with higher reproductive potential. We also observed a trend toward lower pyruvate levels in day 3 culture media of embryos that resulted in pregnancy compared with those that failed to implant (Fig. 1). Our findings are consistent with previous reports by Hardy et al. (46) and Gott et al. (47) who showed that pyruvate and, to a lesser extent, glucose uptake is higher on day 3 embryos that develop into blastocysts compared with those that arrest at cleavage stage. However, our findings are in conflict with another study from the same group that reported that pyruvate uptake was higher in day 2 and day 3 embryos that failed to implant (23). We agree with their conclusions that the observed differences among studies may be due to culture media differences and methodologies applied, and that the range of individual values in study groups would prevent selection of embryos for transfer on the basis of pyruvate uptake alone (23). This conclusion strengthens the view that a model (Fig. 2) examining the input of various components is better able to provide discrimination between the individual reproductive potential of embryos. Our current findings, together with our recent reports using Raman and NIR spectroscopy (26, 27), suggest that metabolomic assessment of spent embryo culture media may provide additional information that may help better assess the reproductive potential of individual embryos in ART. A major benefit of an improved embryo selection methodology would be an increase in implantation rates and PRs. Such a benefit would lead to a decrease in the cost, as well as in the physical and psychological morbidity, associated with prolonged infertility treatment. We could also speculate that improved embryo selection would promote the transfer of a lower number of embryos and ultimately contribute to a decrease in multiple infant births that result from ART. To determine whether such effects will ensue, randomized prospective trials comparing the current embryo selection methodology based on cleavage rate and morphology, to embryo selection based on spectroscopy alone or in combination with morphologic evaluation will have to be conducted.
REFERENCES 1. Mosher WD, Pratt WF. Fecundity and infertility in the United Sates: incidence and trends. Fertil Steril 1991;56:192–3. 2. Society for Assisted Reproductive Technology (SART). Assisted reproductive technology success rates. National summary and fertility clinic reports. Centers for Disease Control, Atlanta, GA, 2006.
Vol. 90, No. 6, December 2008
3. Wright V, Chang J, Jeng G, Macaluso M. Assisted reproductive technology surveillance—United States, 2003. MMWR Surveil Summ 2006;55:1–22. 4. Kovalevsky G, Patrizio P. High rates of embryo wastage with the use of assisted reproductive technology: a look at the trends between 1995 and 2001 in the United States. Fertil Steril 2005;84:325–30. 5. Bromer JG, Seli E. Assessment of embryo viability in assisted reproductive technologies: shortcomings of current approaches and the emerging role of metabolomics. Curr Opin Obstet Gynecol 2008;20:234–41. 6. Toner JP, Veeck L, Acosta AA, Muasher SJ. Predictive value of pregnancy during original in vitro fertilization cycle on implantation and pregnancy in subsequent cryothaw cycles. Fertil Steril 1991;56:505–8. 7. El-Thouky T, Khalaf Y, Al-Darazi K, O’Mahony F, Wharf E, Taylor A, et al. Cryo-thawed embryos obtained from conception cycles have double the implantation and pregnancy potential of those from successful cycles. Hum Reprod 2003;6:1313–8. 8. Reddy UM, Wapner RJ, Rebar RW, Tasca RJ. Infertility, assisted reproductive technology, and adverse pregnancy outcomes. Obstet Gynecol 2007;109:967–77. 9. Luke B, Martin J. The rise in multiple births in the United States: who, what, when, where, and why. Clin Obstet Gynecol 2004;47:118–33. 10. Keith LG, Oleszczuk JJ, Keith DM. Multiple gestation: reflections on epidemiology, causes, and consequences. Int J Fertil Womens Med 2000;45:206–14. 11. US Department of Health and Human Services PHS. US vital statistics, 1998 and from US Public Health Service. Healthy People 2000: National Health Promotion and Disease Prevention Objectives, DHHS Pub No. (PHS) 90–50212. 12. Luke B, Brown MB. Contemporary risks of maternal morbidity and adverse outcomes with increasing maternal age and plurality. Fertil Steril 2007;88:283–93. 13. Scott LA, Smith S. The successful use of pronuclear embryo transfer the day following oocyte retrieval. Hum Reprod 1998;13:1003–13. 14. Tesarik J, Greco E. The probability of abnormal preimplantation development can be predicted by a single static observation on pronuclear state morphology. Hum Reprod 1999;14:1318–23. 15. Veeck L. An atlas of human gametes and conceptuses: an illustrated reference for assisted reproductive technology. New York: Parthenon Publishing, 1999. 16. Gerris J, De Neubourg D, Mangelschots K, Van Royen E, Van de Meerssche M, Valkenburg M. Prevention of twin pregnancy after in-vitro fertilization or intracytoplasmic sperm injection based on strict embryo criteria: a prospective randomized clinical trial. Hum Reprod 1999;14: 2581–7. 17. VanRoyan E, Mangelschots K, De Neubourg D, Valkenburg M, Van de Meerssche M, Ryckaert G, et al. Characterization of a top quality embryo, a step towards single-embryo transfer. Hum Reprod 1999;14:2345–9. 18. Gardner DK, Schoolcraft WB. In vitro culture of human blastocysts. In: Jansen R, ed. Towards reproductive certainty: fertility and genetics beyond. Carnforth: Parthenon Publishing, 1999:378–88. 19. Steptoe PC, Edwards RG. Birth after the reimplantation of a human embryo. Lancet 1978;2:366. 20. Trounson A, Leeton J, Wood C, Webb J, Wood J. Pregnancies in humans by fertilization in vitro and embryo transfer in the controlled ovulatory cycle. Science 1981;216:681–2. 21. Toner JP. Progress we can be proud of: U.S. trends in assisted reproduction over the first 20 years. Fertil Steril 2002;78:943–50. 22. Sakkas D, Gardner DK. Noninvasive methods to assess embryo quality. Curr Opin Obstet Gynecol 2005;17:283–8. 23. Conaghan J, Hardy K, Handyside A, Winston RML, Leese HJ. Selection criteria for human embryo transfer: a comparison of pyruvate uptake and morphology. J Assist Reprod Genet 1993;10:21–30. 24. Gardner DK, Lane M, Stevens J, Schoolcraft WB. Noninvasive assessment of human embryo nutrient consumption as a measure of developmental potential. Fertil Steril 2001;76:1175–80. 25. Brison DR, Houghton FD, Falconer D, Roberts SA, Hawkhead J, Humpherson PG, et al. Identification of viable embryos in IVF by noninvasive measurement of amino acid turnover. Hum Reprod 2004;19: 2319–24.
Fertility and Sterility
26. Seli E, Sakkas D, Scott R, Kwok JS, Rosendahl S, Burns DH. Non-Invasive metabolomic profiling of human embryo culture media using Raman and near infrared spectroscopy correlates with reproductive potential of embryos in women undergoing in vitro fertilization. Fertil Steril 2007;88:1350–7. 27. Scott RT, Seli E, Miller K, Sakkas D, Scott K, Burns DH. Non-Invasive metabolomic profiling of human embryo culture media using Raman spectroscopy predicts embryonic reproductive potential: a prospective blinded pilot study. Fertil Steril 2008;90:77–83. 28. Muasher SJ, Abdallah RT, Hubayter ZR. Optimal stimulation protocols for in vitro fertilization. Fertil Steril 2006;86:267–73. 29. Hwang AJS. Water suppression that works. excitation sculpting using arbitrary waveforms and pulsed field gradients. J Magnetic Res 1995;A 112:275–9. 30. Burton IW, Quilliam MA, Walter JA. Quantitative 1H NMR with external standards: use in preparation of calibration solutions for algal toxins and other natural products. Anal Chem 2005;776:3123–31. 31. Maltz F. Validation of quantitative NMR. J Pharm Biomed Anal 2005;38: 813–23. 32. Groenen PM, Engelke UF, Wevers RA, Hendriks JC, Eskes TK, Merkus HM, et al. High-resolution 1H NMR spectroscopy of amniotic fluids from spina bifida fetuses and controls. Eur J Obstet Gynecol Reprod Biol 2004;112:16–23. 33. Le Moyec L, Muller F, Eugene M, Spraul M. Proton magnetic resonance spectroscopy of human amniotic fluids sampled at 17-18 weeks of pregnancy in cases of decreased digestive enzyme activities and detected cystic fibrosis. Clin Biochem 1994;27:475–83. 34. Cummins JM, Breen TM, Harrison KL, Shaw JM, Wilson LM, Hennessey JF. A formula for scoring human embryo growth rates in in vitro fertilization: its value in predicting pregnancy and in comparison with visual estimates of embryo quality. J In Vitro Fert Embryo Transf 1986;3:284–95. 35. Oliver SG, Winson MK, Kell DB, Baganz F. Systematic functional analysis of the yeast genome. Trends Biotechnol 1998;16:373–8. 36. Ellis DI, Goodacre R. Metabolic fingerprinting in disease diagnosis: biomedical applications of infrared and Raman spectroscopy. Analyst 2006;131:875–85. 37. Pauli GF, Jaki BU, Lankin DC. Quantitative 1H NMR: development and potential of a method for natural products analysis. J Nat Prod 2005;68: 133–49. 38. Kan-Zhi Liu M, Mantsch HH. Simultaneous quantitation from infrared spectra of glucose concentrations lactate concentrations, and lecithin/ sphingomyelin ratios in amniotic fluid. Am J Obstet Gynecol 1999;180: 696–702. 39. Bock JL. Metabolic profiling of amniotic fluid by proton nuclear magnetic resonance spectroscopy: correlation with fetal maturation and other clinical variables. Clinical Chemistry 1994;40:56–61. 40. Szantay CJ. High-field NMR spectroscopy as an analytical tool for quantitative determinations: pitfalls, limitations, and possible solutions. Trends Anal Chem 1992;11:332–44. 41. Gruetter R, Weisdorf SA, Rajanayagan V, Terpstra M, Merkle H, Truwit CL, et al. Resolution improvements in in vivo 1H NMR spectra with increased magnetic field strength. J Magn Reson 1998;135:260–4. 42. Gardner DK, Lane M. Amino acids and ammonium regulate mouse embryo development in culture. Biol Reprod 1993;48:377–85. 43. Houghton FD, Hawkhead JA, Humpherson PG, Hogg JE, Balen AH, Rutherford AJ, et al. Non-invasive amino acid turnover predicts human embryo developmental capacity. Hum Reprod 2002;17:999–1005. 44. Brinster RL. Studies on the development of the mouse embryos in vitro. II. The effect of energy source. J Exp Zool 1965;158:59–68. 45. Gardner DK, Leese HJ. Non-invasive measurement of nutrient uptake by single cultured preimplantation mouse embryos. Hum Reprod 1986;1:25–7. 46. Hardy K, Hooper MAK, Handyside AH, Rutherford AJ, Winston RML, Leese HJ. Non-invasive measurement of glucose and pyruvate uptake by individual human oocytes and preimplantation embryos. Hum Reprod 1989;4:188–91. 47. Gott AL, Hardy K, Winston RML, Leese HJ. Non-invasive measurement of pyruvate and glucose uptake and lactate production by single human preimplantation embryos. Hum Reprod 1990;5:104–8.
2189