HM Medical Clinic

The Analysis of Estrogens Using
Liquid Chromatography and Negative
Electrospray Ionization Mass
Katie estridge1,2, carol Babyak2, wendy lewis2
1environmental science Program, appalachian state University, Boone, nc 2a.R. smith department of chemistry, appalachian state University, Boone, nc [email protected] Abstract
Many estrogenic compounds enter the environment via wastewater treatment plants
(WWtps) that discharge treated water into aquatic systems. effluent from the Boone,
nC WWtp into the South Fork of the new river is responsible for 23-30% of its total
flow. to date, no field-testing has been conducted to determine if elevated concen-
trations of estrogenic compounds are being expelled from the wwtP when a larger
student female population is living in Boone during appalachian state University's
academic year. to investigate natural and synthetic estrogens sourced from the stu-
dent population, including estrone, estradiol, and ethinylestradiol, we developed an
analytical method for the identification and quantification of various estrogens using
high-performance liquid chromatography coupled to an electrospray ionization mass
spectrometer (hplC-eSi-MS) operating in negative ion mode. to determine the ap-
propriate cone voltage, probe temperature and mobile phase, a 100 part per million
(ppm) multi-component standard solution containing each of the target estrogens
was used. Optimal needle voltage was determined using three 100 ppm single com-
ponent estrogen solutions, which were manually injected into the esi-Ms. with re-
spect to these parameters, optimal results were obtained using a cone voltage of 120
v, a probe temperature of 450˚C, a mobile phase of degassed deionized water, and a
needle voltage of -2.0 kv.
approximately 1.0 part per billion (ppb), intersex the presence of pharmaceuticals in environ- gonads were evident; at concentrations of 5-10 mental and drinking waters has recently been ppb, total sex reversal was observed [1].
a growing concern internationally. specifically, waste water treatment plants exist to col- sex hormones may induce some of the most lect and treat or "clean" human sewage before readily observable and daunting effects when releasing it as a water effluent back into the they are present in elevated concentrations in environment. the major goal of such treatment nature [1]. while sex hormones occur naturally, facilities is to remove suspended solids, patho-an increased use of birth control drugs and the genic microbes, chlorine compounds, excessive application of hormones to livestock likely exac- phosphorus and nitrogen nutrients, heavy met- erbate this problem [2]. the effects of pharma- als, and pharmaceuticals prior to discharging the ceuticals in drinking water may have less of an water back into the ecosystem. a typical wwtP immediate impact on humans compared to the effluent report may include levels of dissolved immediacy of endocrine disruption on certain oxygen, suspended solids, ph, fecal coliform, aquatic species. in a controlled exposure experi- metals, nitrogen, etc., but these reports are not ment by Metcalfe et al. (2001) [3], male Japanese likely to include information on pharmaceuti- medake fish were exposed to various estrogen cal discharge content because measuring a vast compounds including estradiol, ethinylestradiol number of pharmaceutical compounds is obvi-(ee2), estrone, and estriol. At a concentration of ously impractical for most facilities. due to the Journal of Student research in environmental Science at Appalachian lack of attention given to pharmaceutical con- dardized mass chromatograms, esi is a relatively tent in wwtP effluent, some drugs may not be new technique without standardized conditions. completely removed during the treatment pro- therefore, esi operations have not been fully cess, and may exist in effluent at trace concentra- developed and are unique to a particular set of tions or higher. analytes for individual experiments. For example, in general, estrogens leave the body as con- electron impact (ei) ionization sources typically jugates (primarily glucuronides), which have the operate using 70 ev electron energy because it is potential to dissociate into their original free under these conditions that ion yield is generally estrogen form after being released into the en- maximized. Also, under conditions of 70 ev, stan- vironment or during treatment at a WWtp [2,4]. dardized ei mass spectra are readily available in Recent research suggests that these estrogens, databases such as the nist chemical webBook synthetic ee2, estradiol, and estrone (see Figure ( however, 1), were found to be the most potent estrogen- esi parameters within the ionization source, such ic compounds in WWtp effluent [4]. Synthetic as cone voltage, needle voltage, and probe tem- ee2, used in 90% of birth control pills, can pass perature, must be manipulated to achieve op- through the body either un-metabolized or can timum volatilization and fragmentation for the be metabolized into estradiol or estrone [5]. the target analytes.
methods developed, as described herein, focus the goal of this work is to describe the meth- on the detection and quantification of the syn- od development process for the detection of thetic estrogen ee2, as well as two natural estro- estrogenic compounds using hPlc-esi-Ms. with gens: estradiol and estrone. the addition of a mass spectrometry component, the dilute estrogen concentrations in treat- analyte detection may be improved with better ed effluent are further reduced when they reach detection limits. with the ability to achieve de- natural sources such as rivers or streams; as a tection in two ways (via pDA detector and MS), result, it is necessary to employ the use of in- one technique can be used to aid in validating strumentation with appropriate sensitivity [4]. the other. Because the high-performance liquid Previously, an isocratic high-performance liquid chromatograph - ultraviolet-visible spectroscopy chromatograph (hplC) with a photo diode ar- (hplC-Uv-viS) method has already been devel- ray (pDA) detector was used to achieve separa- oped successfully [6], any mass spectral incon- tion and detection of analytes [6]; however, due sistencies obtained during esi method develop- to a lack of sensitivity, this method was revised. ment were compared to lc results, which were in order to gain a second element of analyte generally more consistent. detection, a gradient elution method described by Croley et al (2000) [1] was more desirable be- 2.0 Experimental Methods
cause it is compatible with hPlc coupled to elec-trospray ionization mass spectrometry (eSi-MS). 2.1 Reagents & Laboratory Equipment
Unlike more traditional ionization sources estrogen stock solutions were prepared using used in mass spectrometry, which have pre-de- powdered estrogens from sigma-aldrich veloped standardized conditions as well as stan- (17-α-ethinylestradiol- minimum 98 % hplC, Figure 1. structures of estradiol, estrone, and ee2
Molecular Weight (amu)
volume 2, 1st edition • Spring 2012 β-estradiol- minimum 98%, estrone- minimum parafilm and aluminum foil to minimize evapo- 99%). Mobile phases were prepared using ration and light exposure, and placed in a freezer. deionized (Di) water (nanopure Diamond, Prior to their analysis, approximately 1 ml of the Barnstaed hollow fiber filter, gamma irradiated, standard solutions were placed in amber hPlc 0.2 µm pore size rating), acetonitrile (omniSolv, vials and allowed to equilibrate to ambient tem- lC-MS grade), methanol (eMD, hplC grade), acetic acid (glacial, eMD), and ammonium acetate (ominipur, eMD). solid mass measurements were achieved us- separations were achieved using a dionex hPlc ing an analytical balance (Adventurer, ohaus, ser- equipped with the following Ultimate 3000 com- viced certified: 6-15-09) and liquid aliquots were ponents: pump (with Smart Flow), autosampler, measured using glass micro syringes (hamilton column compartment, and photodiode array micro syringes). Mobile phases were degassed detector (with deuterium and tungsten lamps). (sparged) with helium gas (Airgas, he com- a Phenomenex column was used in separa- pressed, Uhp) and filtered using Whatman filters tions (luna, 5µ, C18, 100A, 250mm x 4.6 mm), (cellulose acetate membrane: 0.45 µm pore size, equipped with a guard column (phenomenex, and nylon membrane: 0.2 µm pore size). Security guard hplC, C18, 4 x 3.0 mm). Column all glassware was washed with isopropyl oven temperature was set at 30ºC and a normal alcohol three times, rinsed with di water three injection mode (20 µl injection volume) was em-times, and then solvent rinsed. ployed. At a flow rate of 0.667 ml/min, a gradient elution was used to optimize estrogen separa- 2.2 Mobile Phase & Standard Preparation
tion: at time= 0 min (100 % mobile phase A, zero all mobile phases were filtered using vacuum hold time), at time = 3 min (70% mobile phase filtration and then degassed. degassing was B, hold 7 min), at time = 10 min (100% mobile initially accomplished using nitrogen gas; how- phase A, hold 10 min) for a total separation time ever, mobile phases are now degassed using he- of 20 min. Mobile phase a is considered the pri- lium gas, which is thought to be more effective mary mobile phase (Di water was determined because of its low mass. Mobile phase shelf life best) and mobile phase B is an organic modi-integrity was questionable because degradation fier solution of 1:3 methanol: acetonitrile. the and re-gassing became apparent with age; as a lc separation chromatograms were obtained at result, new mobile phases were prepared at the wavelengths of 217 and 230 nm.
beginning of each week. instrumental interface was achieved be- single component estrogen stock solu- tween the hPlc and an esi-Ms also manufac- tions were prepared by dissolving appropriate tured by Dionex (Surveyor MSQ Mass Spectrom-weights of each estrogen in a 100.00 ml volu- eter System, MSQ20826), and equipped with a metric flask with methanol to achieve a final con- single quadrupole mass analyzer and nitrogen centration of 2000 ppm. eventually the estrone generator (nitrogen, peak Scientific, n418lA). stock solution was prepared at a concentration Mass spectrometry was carried out in negative of 1000 ppm because estrone would not dissolve ion mode and the following parameters for op- at higher concentrations. also, estrone would oc- timum ionization were determined throughout casionally precipitate out of solution after being the course of the experiment: probe tempera- stored in the freezer. Multi-component estrogen ture: 450ºC, needle voltage: -2.0 kv, selected ion standards were prepared by adding appropri- monitoring (SiM) cone voltage: -120 v , and total ate aliquots of stock solutions to five 10.00-ml ion count (full scan ion monitoring) cone volt-volumetric flasks, and diluted with methanol. age: -110 v. Because the ionization was carried concentrations of multi-component calibration out in negative ion mode, molecular ion peaks of standards included the following: 5, 10, 25, 50 [M-h]- were prominent. and 100 ppm. stock solutions were prepared no Prior to interfacing the instruments, the Pda less than every six months and calibration stan- was allowed to equilibrate using wavelengths of dards were prepared according to how often 230 and 217 nm, and a basic function test (BFt) calibrations were needed for experimental pur- was performed on the Ms. during a BFt, mass poses. stock solutions and standards were stored spectra are viewed in real time using the follow- in volumetric flasks, capped and wrapped with ing conditions: 0.25 ml/min flow rate, positive Journal of Student research in environmental Science at Appalachian ion mode, probe temperature: 350ºC, cone volt- -110 v was used in the analysis of estrogens. age: 70 v, needle voltage: +3.0 kv, and a mobile phase of 50/50 acetonitrile: water. these analyses 3.2 Probe Temperature Parameterization
were performed at the beginning of each day to next, a series of experiments were conducted ensure the Ms was functioning properly, which to examine the probe temperature effects on was confirmed by the presence of two major compound/estrogen fragmentation and ioniza-mass-to-charge ratio (m/z) peaks (at 42 and 83). tion. like cone voltage, probe temperature also affects fragmentation during the ionization pro- 3.0 Results and Discussion
cess. increasing in increments of 50ºC, trial tem- For many of the experiments described, the ob- peratures ranged from 250 to 500ºC. in order to jective was to achieve optimal and reproducible assess and ensure reproducibility in obtaining mass spectral data after altering parameters target analytes as molecular ion base peaks, a within the esi-Ms. Because analytes were not be- total of three trials were conducted for each tem- ing quantified, rigorous calibrations were rarely perature examined. conducted. a 100 ppm multi-component estro- Upon analysis of the total ion chromato- gen standard containing estradiol, estrone, and grams (tiC), probe temperatures of 400 ºC and ee2 was used throughout each experiment. sub- 450ºC produced the greatest ion yield for each sequent experiments are listed in the chronolog- estrogen (see Figure 2); however, mass spectra ical order in which they were performed. show the most desirable fragmentation when probe temperature was set to 450ºC. it should 3.1 Cone Voltage Parameterization
be noted that at 500ºC, mass spectra from each electrospray ionization is used to fragment large source were too noisy to effect molecular ion molecules into smaller ions, so one of the param- peaks. also, estrone was consistently abundant eters of esi that influences the fragmentation of in mass spectra for all temperatures, while the ions is the cone voltage; different cone voltages appearance and abundance of estradiol and ee2 cause compounds to ionize and fragment in dif- fluctuated greatly between different tempera- ferent ways. in order to determine the optimal tures. Based on these results, a probe tempera-cone voltage to use for the analysis of the target ture of 450ºC was used for the remaining sets of estrogenic compounds, varying cone voltages experiments described in the following sections.
were evaluated by incrementally increasing 10 trial voltages ranging from 60 to 130 v. 3.3 Mobile Phase Modification Experiment
in general, peaks from Uv-vis chromato- in an effort to optimize the esi response in nega- grams were very distinguishable with consis- tive ion mode, mobile phase modification was tent retention times for beta-estradiol near 11.0 examined. Research/work conducted by wu et minutes, ee2 around 11.5 minutes, and estrone al. (2004) [7] describes a phenomenon referred around 12.4 minutes. Molecular ion peaks for be- to as the "wrong way round" concept where the ta-estradiol were observed at m/z = 271 and be- addition of weak acids to mobile phases actually gan to appear at -90 v and are present at highest aid in the deprotonation of analytes in negative abundances from -110v to -120 v. For ee2, strong ion esi-Ms. Up until this point, the mobile phase molecular ion peaks with m/z = 295 were appar- used in our experimentation was a solution of ent from -110 v to -130 v, with a cone voltage of 10 mM ammonium acetate. challenging wu's -110 v showing the best fragmentation. Molecu- results, varying concentrations of acetic acid lar ion peaks for estrone were seen at m/z=269 were used as mobile phase modifiers (in place and were present with more consistency, be- of the original ammonium acetate solution). the coming observable starting at -80 v, but optimiz- modified mobile phase concentrations included ing at cone voltages of -120 and -130 v. Based on 1uM, 10uM, 100uM, 1mM, 10mM, and 100mM these data alone, the optimal cone voltage to be acetic acid solutions, as well as a sample of pure used for these estrogenic compounds was deter- di water as a control. acetic acid solutions were mined to be around -110 to -120 v. For the rest prepared in de-ionized (Di) water, and degassed of experimentation, a siM cone voltage of 120 v before being run through the lc-Ms.
and a full scan monitoring cone voltage of -110 v analysis of mass spectra indicated that the were used. these values are similar to work done best results were obtained when the di water by Croley et al. (2000) [1] where a cone voltage of control was used as mobile phase. as acetic acid volume 2, 1st edition • Spring 2012 Figure 2. tic Peak areas vs. Probe temperature for estrone, ß-estradiol, and ee2
concentration gradually increased, so did the above was used to confirm optimal cone voltage.
noise apparent in mass spectra (see Figures 3a, 3b, 3c). results obtained using a Di water mobile 3.5 Adduct Formation
phase are slightly better than results obtained Occasionally the molecular ion peak [M-h]- for previously when the ammonium acetate solu- each estrogen was less reproducible as the base tion was used as the mobile phase. peak (if present at all) in mass spectra. Mass to charge ratios of the most abundant adducts 3.4 Needle Voltage Experiments
observed include 61.9, 207.1, 297.1, 311.2 and like cone voltage and probe temperature, nee- 325.2. the source of these adducts is currently dle voltage also influences the molecular frag- unknown; however, it is suspected that they are mentation during the ionization process. when related to a structure (see Figure 5), which each operating in negative ion mode, needle voltages of the three estrogens have in common. were measured in negative kilovolts (-kv). prior adducts only appear when samples are ex- to the investigation of needle voltage, a needle amined (apparent in spectra for each estrogen) voltage of -3.0 kv had been used, in accordance and not when mobile phases are examined alone with the croley method [1]. single component -- with the exception of m/z= 62. A BFt modified estrogen standards (100 ppm) were injected to negative ion mode was performed on each manually into the esi-Ms in triplicate for the fol- mobile phase (freshly prepared) and in each test lowing needle voltages: -2.0, -2.5, -3.0, -3.5, -4.0 real time chromatograms displayed base peaks and -4.5 kv. Manually injecting single compo- at m/z= 62. Since the adduct present at m/z=62 nent estrogen solutions was a faster alternative is present during sample analysis and mobile to carrying out the entire lc-Ms separation and phase analysis, we believe that this structure detection. also, using this method, mass spectra may be related to nitrate (no3-, 62.0049 g/mol) could be viewed in real time as estrogens were formation during the ionization process. despite introduced to the detector. Mass spectral analy- good BFts, replacing the esi capillary and follow- sis indicated that increasing needle voltage re- ing the manufacturers maintenance protocol for sulted in an undesirable amount of molecular cleaning the entrance cone, cone wash, capillary fragmentation (see Figure 4). thus, for this work, sleeve and RF lens, mass spectral data continued a needle voltage of -2.0 kv (the lowest our instru- to contain major peaks (often base peaks) at m/ ment is capable of) provided optimal results. the same manual injection technique described the presence of m/z=62 often interfered Journal of student Research in environmental science at appalachian Figure 3a. Mass spectra for each estrogen using the di water mobile phase.
volume 2, 1st edition • Spring 2012 Figure 3b. Mass spectra for each estrogen using the 1uM acetic acid mobile phase.
Journal of student Research in environmental science at appalachian Figure3c. Mass spectra for each estrogen using the 100 mM acetic acid mobile phase.
volume 2, 1st edition • Spring 2012 with our ability to obtain mass spectra with tar- amdar, and diana s. aga. "simultaneous get analytes as the most abundant molecular ion analysis of Free and conjugated estrogens, peaks (see Figure 6). to confirm the suspected sulfonamides, and tetracyclines in Runoff identity of molecular ion peaks at 62, the MS car- water and soils Using solid-Phase extrac- rier gas (nitrogen) was modified with the addi- tion and liquid chromatography−tandem tion of helium gas. Because our instrument did Mass spectrometry." Journal of agricultural not function using 100% helium gas, both gases and Food Chemistry (2011): 2213-222. print.
were introduced to the Ms at varying flow rates. [3] Metcalf, Chris D., tracy l. Metcalfe, yiannis to begin, nitrogen flow was highest with a slight Kiparissis, Brenda g. Koenig, colin Khan, helium flow, which was slowly adjusted such that Richard J. hughes, timothy R. croley, Ray- nitrogen flow was minimized and helium flow mond e. March, and thomas Potter. "estro- was maximized. during carrier gas adjustments, genic Potency of chemicals detected in mass spectra were observed in real time. as he- sewage treatment Plant effluents as de- lium flow was increased and nitrogen flow was termined by in vivo assays with Japanese decreased, the abundance of m/z=62 decreased Medaka (oryzias latipes)."environmental to the point where it was barely detectable. toxicology 20.2 (2001): 297-308. print.
Based on these observations, and the fact that [4] pedrouzo, Marta, Francesc Borrull, eva po-nitrate was often the most abundant molecular curull, and Rosa Maria Marcé. "estrogens ion within our scan range, it was determined that and their conjugates: determination in nitrate (m/z=62) was responsible for consistent water samples by solid-phase extraction interference with analyte detection. Otherwise, and liquid chromatography–tandem Mass the presence of nitrate is not interfering with our Spectrometry." talanta 78.4-5 (2009): 1327- target analytes . For future work, the scan range will be adjusted such that values at or below m/ [5] Benijts, tom, riet Dams, Wolfgang gunther, z=62 are excluded.
willy lambert, and andre de leenheer. "analysis of estrogenic contaminants in 4.0 Conclusions & Future Work
River water Using liquid chromatography the purpose of this research was to develop an coupled to ion trap Based Mass spectrom- instrumental method suitable for the detection etry." Rapid communications in Mass spec- of estrogens in treated wastewater and in natural trometry 16.14 (2002): 1358-364. print.
water systems receiving effluent. Presently, it has [6] Carter, Bethany. "effects of estrogenic Com- been determined that a primary mobile phase pounds in waste water treatment Plant ef- of Di water, a probe temperature of 450ºC, a SiM fluent on Fish vitellogenesis." appalachian cone voltage of -120 v, a full scan monitoring State University. August 2, 2007 cone voltage of -110 v, and a needle voltage of [7] Wu, Zengru, Wenqing gao, Mitch A. phelps, -2.0 kv are the optimal instrumental conditions di wu, and duane d. Miller. "Favorable ef- for the ionization of estrogens of interest. a solid fects of weak acids on negative-ion elec- phase extraction method similar to the one used trospray ionization Mass spectrometry." by Croley, et al. (2000) is currently being inves- Analytical Chemistry 76.3 (2004): 839-47. tigated and a sample collection method must also be established. a deuterium labeled internal standard has been purchased and is intended for use during actual sample analysis.
[1] croley, timothy R., Richard J. hughes,
Brenda g. Koenig, chris d. Metcalfe, and Raymond e. March."Mass spectrometry ap-plied to the analysis of estrogens in the en-vironment." Rapid communications in Mass Spectrometry 14.13 (2000): 1087-093. print. 2011. [2] tso, Jerry, sudarshan dutta, shreeram in- Journal of student Research in environmental science at appalachian Figure 4. Mass spectra for estrone at needle voltages of 3.0 kv (top), 3.5 kv (middle), 4.0 kv (bottom).
the trend of increasing fragmentation with increasing needle voltage is also evident in estradiol and
Figure 5. structure shared by estrogens
identified in the text.
volume 2, 1st edition • Spring 2012 Figure 6. the mass spectra for estradiol (top), ee2 (middle) and estrone (bottom) with the
interference of m/z=62. the presence of m/z=62 often interfered with our ability to obtain mass
spectra with target analytes as the most abundant molecular ion peaks.
Journal of student Research in environmental science at appalachian volume 2, 1st edition • Spring 2012


IT'S MORE Table of Contents Introduction . . . . . . . . . . . . . . . . 37 Aging and Nutritional Well-Being . . . . . . . 38 Good Nutrition for Seniors . . . . . . . . . 39 Determining Nutritional Risk . . . . . . . . . 39 Nutritional Risk Checklist . . . . . . . . . . 40 Promoting Fluid Intake . . . . . . . . . . . 42

Model based drug development - what is it good for?

LIFE SCIENCE TECHNICAL BULLETIN MODEL BASED DRUG DEVELOPMENT – WHAT IS IT GOOD FOR?AUTHOR: JANET R WADE, PHD, SENIOR CONSULTANT, SGS EXPRIMO The FDA ‘‘critical path'' document characterizes model-based drug development (MBDD) as the development and application of pharmaco-statistical models of drug efficacy and safety from preclinical and clinical data to improve both drug development knowledge management and decision-making (1), (2). A less formal description of MBDD would be the development and use of mathematical models to aid answer existing and future project team questions that arise during drug development. As an analogy, many of us probably played with Lego bricks during our childhood. We could make many different models out of those bricks, but we made what we wanted to make at that point in time. In MBDD, the Lego bricks represent pieces of information and the models they are used to create (castle, house, farm, etc.), depends on what is needed (a place to defend yourself from an attack, somewhere warm to sleep, somewhere to raise cows and chickens, respectively).