|The dietary consumption of foods rich in phytoestrogens is thought to reduce the risk of hormone-dependent cancers as a result of the estrogen-like behavior of phytoestrogens. Since 1999, the U.S. Centers for Disease Control and Prevention (CDC) has assessed phytoestrogen exposure in the U.S. population by measuring urinary levels of several isoflavones (daidzein, genistein, equol, O-desmethylangolensin) and lignans (enterolactone, enterodiol) as part of the National Health and Nutrition Examination Survey (NHANES) using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Most LC-MS/MS methods used in epidemiological studies to estimate phytoestrogen exposure are electrospray ionization (ESI) or atmospheric pressure chemical ionization (APCI) based procedures that require the use of solid phase extraction (SPE) or some other laborious cleanup and preconcentration procedure.|
In this study we sought to develop a simplified LC-MS/MS method for measuring urinary phytoestrogens that eliminated the need for SPE cleanup. We explored the possibility of measuring urinary phytoestrogens using atmospheric pressure photoionization (APPI). Preliminary experiments showed that the polyphenolic chemical structures of phytoestrogens were highly amenable to APPI, resulting in substantial (10-fold) sensitivity improvements over the ESI method that would permit the elimination of a preconcentration step. The APPI source also appeared to have a greater degree of ionization specificity, resulting in MS/MS chromatograms with far fewer concomitant peaks and presenting the possibility of injecting urine samples directly without SPE cleanup.
The method we propose uses a simplified sample preparation procedure that entails amending a urine sample with stable isotope labeled internal standards, β-glucuronidase/sulfatase (to convert conjugated analytes to their free forms) and umbelliferone glucuronide (a probe to monitor the deconjugation process), followed by incubation at 37°C for 4 hours. Chromatographic separation was achieved in 8 min using a conventional C18 column and a water-methanol gradient. The APPI source was operated in negative ion mode using toluene as a dopant. Limits of detection were in the range of 0.05–0.5 ng/mL urine. This is commensurate or better than our current ESI-based SPE procedure and suitable for biomonitoring purposes. Average spike recoveries ranged from 96%–105%. Between-run coefficients of variation (CVs) across six days were <10% for all analytes at 10 ng/mL and dropped with increasing urine concentration, in some cases to <2%. Within run CVs were on average <4%. Preliminary comparison between the new and the current method based on 80 urine samples showed excellent correlation (Pearson r = 0.992–0.998) and good agreement (Deming regression slopes = 0.86–0.95).