Gas chromatography/mass spectrometry of catechol estrogens Luigi A. Castagnetta, * Orazia M. Granata,? Felice P. Arcuri,* Lucia M. Polite,? Felice Rosati,$ and Gian P. Cartoni§ *Hormone Biochemistry Laboratories, School of Medicine, Policlinico, University of Palermo, Italy; tExperimenta1 Oncology and Molecular Endocrinology Units, Palermo Branch of National Cancer Institute of Genoa, clo “M. Ascoli” Cancer Hospital Center, Palermo, Italy; SAntidoping Laboratory, FMSI, Rome, Italy; and Khemistty Department,
University
“La Sapienza,
” Rome,
Italy
Catecholestrogens (CCEs), namely 2- or 4-hydroxyestradiol and hydroxyestrone, are highly polar, reactive, and extremely labile estrogen metabolites in many experimental conditions. For these reasons, indirect assay methods mainly have been used. Some experimental evidence suggests that CCEs are synthesized and biologically active mostly in target cells. At this level, unfortunately, the indirect assays cannot be used. We present a method of gas chromatographiclmass spectral (GCIMS) analysis for the identijication of individual CCEs; the major fragmentation ions of authentic estrogen standards as trimethylsilylether derivatives, and the MS patterns of the major CCEs, namely, 2-hydroxyestradiol and hydroxyestrone, are included. Few examples of CCEs detected in human breast cancer tissues and in breast cystfluids are reported. Sample extracts were submitted to reversed-phase, high-performance liquid chromatography (RP-HPLC) and were quantified by “on line” electrochemical (EC) detection; thereafter, either crude extracts or single eluted peaks were submitted to GCIMS, by which detection limits of less than 5 pmol were attained. As expected, the molecular ion was the most relevant molecule in all but one case. On the contrary, the other relative intensities of major fragmentation ions M - 15, M -30, M - 90, and M - 15 + ( - 90) were unevenly distributed, although represented in the majority of cases. In all cases, the GCIMS ofpeakfractions, purtsed by RP-HPLC and WV detection, conJirmed the results of liquid chromatographic analysis combined with EC detection. in contrast, GCIMS of crude extracts was not equally satisfactory. Comparison of a liquid chromatography system with EC detection and the GCIMS approach revealed some inconsistency in quantitation of individual CCEs. Despite the fact that RP-HPLC with EC detection is sensible and versatile enough to analyze a number of free estrogen metabolites, the results of present study strongly suggest the need for a preliminary sample purification for CCE identt$cation by GCIMS. This represents an essential requirement for intratissue concentration studies. (Steroids 57:437-443, 1992)
Keywords: hormones and cancer; steroids; catecholestrogens; phy; gas chromatography/mass spectrometry
Introduction Catecholestrogens (CCEs) are peculiar metabolites of the parent “classic” estrogens, namely, estradiol (E2) and estrone (E,), which are produced by hydroxylation at either the C-2 or C-4 position, and are considered a major group of active metabolites of estrogens.‘32
Address reprint requests to Prof. Luigi A. Castagnetta at the Experimental Oncology and Molecular Endocrinology Units, c/o “M. Ascoli” Cancer Hospital Center, P.O. Box 636, Via de1 Vespro, 90127 Palermo, Italy. Received December 2, 1991; accepted April 25, 1992.
0 1992 Butterworth-Heinemann
reversed-phase,
high-performance
liquid chromatogra-
The catechol function (ie, a double adjacent OHgroup) in the aromatic ring confers to CCE high reactivity. Because of these physicochemical properties, CCEs exhibit a very wide range of polarities and are difficult to handle; thus, indirect measuring approaches have been used almost exclusively.3-5 An important feature of CCEs is that they are widespread, being present in many different tissues and organs. In particular, very high levels have been found in both the central nervous system (CNS) and liver.6-8 In addition, CCEs represent the major product of estroeven though their blood concentragen excretion.‘-” tions are reported to be 10w.~
Steroids,
1992, vol. 57, September
437
Papers Table 1 Systematic and trivial names of classic and main metabolites of estrogens: TMS numbers, times of authentic standard derivatives in gas chromatography/mass spectrometry analysis Systematic
Trivial name
name
1,3,5(10)-Estratriene-3,17/3-diol 1,3,5(10)-Estratrien-3-ol-17-one 1,3,5(10)-Estratriene-3,16~1,17p-trio1 1,3,5( 1 O)-Estratriene-2,3,17@triol 1,3,5(10)-Estratriene-3,4,17/3-trio1 1,3,5(10)-Estratriene-2,3-diol-17-one 1,3,5(10)-Estratriene-3,4-diol-17-one 1,3,5(10)-Estratriene-2,3,16a,l7/3-tetrol 1,3,5(10)-Estratrien-3,16cu-diol-17-one 1,3,5(10)-Estratriene-2,3,16a,l7p-tetrol 2-methyl ether 1,3,5(10)-Estratriene-2,3,17/3-trio1 2-methyl ether 1,3,5(10)-Estratriene-3,4,17p-trio1 4-methyl ether 1,3,5(10)-Estratriene-2,3-diol-17-one Z-methyl ether 1,3,5(10)-Estratriene-3,4-diol-17-one 4-methyl ether Abbreviation:
TMS, trimethylsilyl;
MW, molecular
weight;
molecular
weights,
and retention
No. of TMS
MW
RT
2 2 3 3 3 3 3 4 3 3 2 2 2 2
416 414 504 504 504 502 502 592 502 534 446 446 444 444
8.06 7.66 13.85 11.45 12.82 10.95 12.15 17.00 13.82 16.30 11.30 10.15 10.60 9.60
Estradiol Estrone Estriol 2-Hydroxyestradiol 4-Hydroxyestradiol 2-Hydroxyestrone 4-Hydroxyestrone 2-Hydroxyestriol 16~~Hydroxyestrone 2-Methoxyestriol P-Methoxyestradiol 4-Methoxyestradiol 2-Methoxyestrone 4-Methoxyestrone RT, retention times.
These pieces of evidence strongly suggest an overall biologic relevance of CCEs, which are mainly synthesized and biologically active in target tissues and cells. We have knowledge of a number of biochemical functions exerted by CCEs, including their antiestrogenie action exerted in vitro,” and of their genotoxic potential.13 In addition, their levels may be strongly modified in certain pathologic conditions.R,9s’4 However, assay methods are unsatisfactory and the biologic role(s) of CCEs in endocrine pathophysiology is still inadequately understood. All the above concerns indicate that these highly polar phenolic steroids deserve to be quantified in target tissues, implying the need for the direct identification of single estrogen metabolites as previously carried out, for example, in urineJ5,‘” and in in vitro systems.” To this end, we have devised a method of gas chromatography/mass spectrometry (GC/MS) analysis for the identification of CCE derivatives after their measurement in benign (cyst fluids) and cancerous breast tissues by reversed-phase, high-performance liquid chromatography (RP-HPLC) and “on line” electrochemical (EC) detection. Examples of the assessment of the major CCEs by GC/MS, after single peak collection from the RPHPLC system and UV detection, are presented.
Experimental Specimens Breast cancer (BCA) tissue specimens, obtained through radical mastectomy, and breast cyst fluid (BCF) specimens, obtained by fine needle aspiration, were used.
[namely, water and acetonitrile]) were obtained from Carlo ErbaFarmitalia (Milan, Italy). Before setting the chromatographic analysis conditions, a citric acid solution of the eluting system was passed through a cellulose nitrate membrane filter (0.22 pm) purchased from Whatman (Kent, UK). Ultrasphere-ODS (Beckman Inc., Fullerton, CA, USA) or Spherisorb 5s ODS-2 (Phase SEP, Clwyd, UK) columns (5pm particle size, 250 x 4.6 mm internal diameter) were used. N-methyl-N-trimethylsilyltrifluoroacetamide (MSTFA) and N-trimethylsilylimidazole (TMSI) were from Pierce (Luton, UK) and dithioerythritol was obtained from Serva (Heidelberg, Germany).
Apparatus An HPLC system obtained from Beckman Instruments Inc. (Berkeley, CA, USA) equipped with a model 1I2 pump, a model 210 sample injection valve, a model 160 UV detector (18.5~1 sample cell, mercury lamp, and a 280-nm filter), a two-channel recorder Omniscribe (Houston Instruments, Gistel, Belgium), and a fraction collector model 202 (Gilson Medical Electronics S.A., Villiers le Bell, France) was used. To concentrate aqueous extracts, a Speed Vat System (Savant Instruments Inc., Farmingdale, NY, USA) was used as evaporator-concentrator-dryer. An EC detector Coulochem model 5100A (ESA, Bedford, MA, USA) equipped with a 5020 model guard cell set at +0.50 V and a 5010 model analytical cell (5-~1 cell size) set at +0.75 V and containing two porous graphite coulometric electrodes was used. The GC/MS detection was performed using a GC-MS model 5890-5970 from Hewlett Packard (Rockville, MD, USA), equipped with a 5% phenyl-methyl silicone capillary column (20 mm x 0.2 mm internal diameter, 0.33-km film thickness) HP 190915-105.
Extruction procedures Materials All standard estrogens (Table l), including deuterated E,, were purchased from Sigma Chemical Co. (St. Louis, MO, USA). Chemicals for estrogen extraction procedures (i.e., acid alumina, tris(hydroxymethyl)aminomethane, hydrochloric acid, glacial acetic acid, diethylether, citric acid, and HPLC-grade solvents
438
Steroids,
1992, vol. 57, September
Tissues were processed fresh or after storage at -20 C in a sucrose/glycerol buffer. Dissected BCA tissue (1 g) was homogenized in 3 ml of 0.01 N perchloric acid and 3 ml of 0.1 M Tris buffer (pH 8.6) at a final pH of 3.4 and centrifuged at 2,000 x g for 10 minutes. The supernatant was decanted into a test tube containing 40 mg of acid alumina (pH 4.0) in 5.0 ml of 0.1 M
GUMS of catechol estrogens: Table 2
Mass spectral analysis of authentic estrogen derivatives:
Estradiol Estrone Estriol P-Hydroxyestradiol 4-Hydroxyestradiol 2-Hydroxyestrone 4-Hydroxyestrone 2-Methoxyestriol P-Hydroxyestriol 16wHydroxyestrone 2-Methoxyestradiol 4-Methoxyestradiol 2-Methoxyestrone 4-Methoxyestrone
416 414 504 504 504 502 502 534 592 487 446 446 444 444
(100)” (100)” (1 OOP (100)” (1 OOY (lOOja (100)” (100)” (lOO)a (100)’ (100)’ (100)” (lOO)a (100)a
285 399 311 373 373 155 155 129 433 502 315 315 155 155
(61 I (67jc (62) (9) (19) (16) (22) (27) (5) (25ja (16) (40) (25) (50)
232 155 129 267 325 319 319 375 385 231 416 129 261 429
Castagnetta et at.
the most intense peaks and relative intensities
(51 (45) (55) (6) (9) (9) (15) (22) (5) (IO) (7P (25) (12) (25)’
326 309 386 489 489 306 169 259 147 147 431 261 429 261
(12jb (24P’ (50) (5P (5)’ (9) (12) (15) (5) (IO) (6)’ (12) (1OY (20)
129 231 297 325 267 169 487 327 399 397 209 431 339 339
(12) (15) (40) (3) (5) (8) (lOjc (15) (4) (3) (6) (IOY (8jd (12)d
401 (9)” 324 (3)“ 324 (28) 129 (3) 129 (3) 487 (6)’ 306 (10) 519 (5Y 577 (4P
414 (18)’
129 (6) 416 (10P 414 (5P 414 (3P
397 397
(6)“ (8)‘j
356 372
(8jb (5)
aM+. b(M “(M d(M e (M
- 90)+. - 15)‘. - 15 - 30)‘.
19)+.
Tris buffer (pH 8.6). The CCEs adsorbed by acid alumina were extracted with 200 ~1 of 0.2 M acetic acid and dried in a Speed Vat system, while all other estrogens were extracted from Tris buffer with diethyl ether and dried at 40 C. Both extracts were combined just before HPLC analysis. The detailed steps of the extraction procedure for BCF samples have been described elsewhere.‘*
Reversed-phase, high-performance liquid chromatography and gas chromatographylmass spectrometry analytical conditions Dried extracts were redissolved in 100 ~1 of acetonitrile. Aliquots (10 to 20 ~1) were injected for RP-HPLC chromatographic analysis using our computer-aided optimized isocratic conditions (acetonitrile/O.OOl M citric acid, 40 : 60 v/v; 1.O ml/min flow rate), as previously detailed.‘9*‘” The sensitivity and response linearity of the EC detector and the relative retention times (RRTs) of all standard estrogens of interest have been previously reported.” To verify by W/MS the identity of the metabolites quantified in RP-HPLC plus EC, the single peak fractions were collected from the RP-HPLC system after UV detection; dried extracts were stored for a maximum of 1 week if they were not immediately processed. Just before the GC/MS analysis, equilin (50 ng) was added as internal standard to dried extracts redissolved in 50 ~1 of MSTFA/TMSI (1,000 : 2 v/v) containing 0.2% dithioerythritol, according to the methods we used previously for the anabolic steroids and their hydroxylated metabolites.*’ Samples were then heated for 30 minutes at 60 C and finally injected into the W/MS apparatus. The temperature of the injector, ion source and transfer line were set at 280, 200, and 280 C, respectively. The oven temperature was initially kept at 250 C for 13 minutes, then increased to a final temperature of 280 C (at a rate of 10 C/min) and held constant for 7 minutes. Samples (1 to 2 ~1) were injected with a splitter in a 1: 10 ratio. Helium was used as carrier gas at a flow rate of 0.9 ml/min. The energy of ionizing electron beam was 70 eV and electron impact (EI) mass spectra were recorded in the full scan mode from 100 to 600 amu. Using equilin as internal standard (IS), GC/MS quantitation of single peaks was carried out on a single ion of either IS or the
specific external standard (20 ppm). Calculations were performed through area measurement integration using the MSD 5890 software package.
Results As expected, the retention times (RTs) of the CCETMS derivatives, reported in Table 1, increased with the number of TMS groups in the molecule. A good separation was obtained for all the compounds examined by GC. Table 2 shows the mass/energy (m/z) values with the relative intensities obtained from EI mass spectra (in decreasing order). Intensity values below 3% of the base peak were not included.
502
4
6
8
?o
12
14
16
min Figure 1 (A) Mass spectra and (B) total ion chromatogram the authentic 2-hydroxyestrone derivative.
Steroids,
1992, vol. 57, September
of
439
Papers groups
as well
as from
the cleavage
of the
A. B. or II
rings.
As reported both in BCA
12
10
8
6
16
14
18
min Figure 2 (A) Mass spectra and (B) total ion chromatogram the authentic Z-hydroxyestradiol derivative.
of
7 t ;i a E m3 z ii 4 1 050lii~
, 8
10
..,...,, 12
z 5811
7
can be seen from Table 3 and from Figure 3. when crude extracts were submitted to GCIMS, no CCE derivatives were detectable. This was to be expected, since an adequate sensitivity of detection” may be achieved in the presence of low concentrations of single metabolites, only after preliminary purification. The GCYMS analysis confirmed identity of all peak fractions were collected from RP-HPLC after UV detection, with the exception of two BCA and one BCF cases, which had concentrations less than 0.25 nmolig or 0.25 nmol/ml, respectively. Comparison of quantitation of individual CCEs by EC detection “on line” with RP-HPLC and GCiMS did not yield equivalent results. This was especially true for crude extracts, even in the presence of an appreciable amount of CCEs. Both acid alumina extraction’* (Table 4) and RP-HPLC purification”’ greatly improved the recoveries and signal to background ratio for MS analysis of free CCEs. The evidence of 20H-E, in the breast cancer tissues examined is confirmed by both the presence of two characteristic ions (SO2 and 487 amu), obtained by the Selected Ion Monitoring (SIM) search (Figure 4). and the good coincidence of the RTs in the traces recorded. The same is shown in Figure 5 for the 20H-E, content of BCF. Again, the presence of two characteristic ions (504 and 489 amu) emerges, and RTs are coincidental.
_ 14--iF;l;,liJ
Table 3 Concentrations of catecholestrogens as calculated by (A) reversed-phase, high-performance liquid chromatography and electrochemical detection and (B) gas chromatography/ mass spectrometry
Ion 412 amu
3.104
in Table 3, CCEs have been detected tissues and in gross cyst fluids. As also
8
10
9 Ion 446amu
,r. ws.4~. .. . .-9.--w---.#.,, .h*.,..,_au_ , ..--7 8 9 10
Code no.
CCEs
A
B
5 oI.‘-. 4: 738
Ion 504 amu ~..*C~~~.~wC_~YU*.n~.~~ I”“#’ ‘.r-...r 13 12 11
+I---
T---y 15
min
Figure 3 Total ion chromatogram of individual ion trace of investigated CCEs from crude extract without any prior HPLC purification. Ions monitored at m/z 412 (M)‘, 446 (MI+, and 504 (M)+ for the equilin, 4-methoxyestradiol, and 4-hydroxyestrone derivatives, respectively.
The mass spectra of both 2-hydroxy-E, (20H-E,) and 2-hydroxy-E, (20H-EJ derivatives of authentic estrogen standards (recorded on the elution peak) and their total ion chromatograms (TICS), are illustrated in Figures 1 and 2, respectively. These data clearly indicate that in almost all the mass spectra, the molecular ion is the most intense peak. Formation of other ions may result from the loss of OH-Si-(CH3)3 (M-90) and/or CH3 (M-15, M-30) 440
Steroids,
1992,
vol.
57, September
Crude extracts 81251 81291 B1314 81314
(nmol/g) 20H-E, 20H-E, 20H-E, 40H-E,
2.76 0.76 1.79 0.89
ND ND ND ND
Purified peak fractions from BCA tissues (nmol/g) 81223 40H-E, 0.22 81265 20H-E, 2.46 81314 20H-E, 1.78 81314 40H-E, 0.89 81343 20H-E2 0.24 81548 20H-E2 2.38 81582 20H-E, 0.89
ND 2.35 1.67 0.91 ND 2.15 1.10a
Purified peak fractions from BCF (nmol/ml) BCFI 06 20H-E, 5.70 BCF115 40H-E, 0.11 BCFI 18 20H-E, 1 .20c
5.23’ 0.09 ND
Abbreviation: ND, not detected a See also Figure 4. ‘See also Figure 5. c Reported as pmoliml.
GUMS
of catechol
4
11
d
amu 12
mu
13
s . :! I
14 min
12
14 min
A typical EC detection of a RP-HPLC chromatographic profile of free estrogens from a sample of BCA tissue is presented in Figure 6.
Discussion The CCEs exert a number of biochemical actions, including (1) a wide spectrum of binding to estrogen receptors23*24;(2) the inhibition, for instance, of catecholamines or tyrosine hydroxylase2s; (3) the ability to compete for dopamine receptors in the pituitary26 and thus to modulate synthesis or secretion of gonadotropin, prolactin, and prostaglandins27*28; and, most important, (4) the ability to covalently bind macromolecules and even to produce DNA adducts.13 Other interesting pieces of evidence are that 20HE, may reduce proliferative activity of BCA cells in
Table 4 Recovery values (%) of some authentic dards using acid extraction method (pg)
1 0.5 0.05 0.01 Mean
t
SD
. . . . . . 11
11
Figure 4 Selected ion gas chromatogram of the purified peak fraction from a breast cancer tissue extract (sample 81582 in Table 3). Ions monitored at m/z 502 (M)’ and 487 (M - 15)+ for the 2-hydroxyestrone derivative.
Amount
. .
estrogen
stan-
20H-E,
El
2MeO-E,
82 90 90 a4
82 84 92 87
93 90 80 86
86.5 -c 4.1
86.5 2 4.0
87.3 2 5.6
14
Ion 504 amu .._........... 12 13 Ion
13
Castagnetta
et al.
16 min
l8
12
4...s.-U--” 11
12
10
8 Ion 502
estrogens:
12
489
amu 13
I4 min
_
.I
14 min
Figure 5 Selected ion gas chromatogram of the purified peak fraction from a BCF extract (sample BCF106 in Table 3). Ions monitored at m/z 504 (M)+ and 489 (M - 15)’ for the 2-hydroxyestradiol derivative.
vitro; thus, CCEs may behave as true natural antiestrogens.‘2,29 Furthermore, the early formation of CCEs by hormone-responsive human mammary cancer cell lines has been observed previously.30 As a consequence, the balance between 2/4 hydroxylation versus 16a hydroxylation, as previously shown3’ could be understood as a kind of regulation of proliferative activity of cancer cells, at least in the “endocrine growth control.” The inconsistency observed between plasma and urine levels reinforces the assumption that CCEs are mainly synthesized and exert a major biologic activity in target tissues. Consequently, it seems of more importance to analyze, identify, and quantify these metabolites in target tissues than to evaluate their blood levels. Identification and quantification of several endogenous estrogens are hardly practicable without a precise, undeniable confirmation of their identity. The data reported here indicate a good concordance between RP-HPLC with “on line” EC detection and GUMS in the identification of specific CCE metabolites. Looking at quantification, a certain inconsistency between the two different detection systems emerges. On one hand, this discrepancy could be attributed, in part, to nonoptimal recovery and losses during extraction and purification procedures; on the other hand, high background levels may strongly reduce the sensitivity limits of MS analysis. In spite of sufficient reproducibility, the recoveries for hydroxyestrogens appear to be limited to 50%32
Steroids.
1992. vol. 57. SeDtember
441
Papers
tion procedure
allows improved recovery. These reour combined RP-HPLC and GUMS approach for the simple, time-saving identification and quantitation of intratissular hydroxyestrogens.
5 i
sults validate
Acknowledgments The authors are indebted to the late Professor V. Pozzi for his enthusiastic encouragement and the provision of BCF samples. The authors also thank the Italian Association for Cancer Research (AIRC) for financial support. F. Arcuri is the recipient of a fellowship from the AIRC. References I. 2.
6
Fishman J, Bradlow HJL, Gallagher TF (1960). Oxidative metabolism of estradiol. J Biol Chem 235:3104-3107. Ball P, Knuppen R (1980). Catechol estrogens (2. and 4-hydroxyestrogens). Chemistry, biogenesis, metabolism, occurrence and physiological significance. Acta Endocrinol (Copenh) 93:1-127.
3.
Kono S, Merriam GR, Brandon DD, Loriaux DL, Lipsett MB (1982). Radioimmunoassay and metabolism of the catechol estrogen 2-hydroxyestradiol. J Ckn Endocrinol Metab 54:150-154.
4.
.s.
6. Figure 6 Reversed-phase, high-performance liquid chromatography estrogen profile by EC detection of a BCA tissue extract. (For analysis conditions, see text.) Peaks and quantities (nmol/ g tissue): 1, estriol (0.08); 2, 2-methoxyestriol (0.11); 3, 16~ hydroxyestrone (0.22); 4, P-hydroxyestradiol (1.04); 5, 4-hydroxyestradiol(4.16); 6,2-hydroxyestrone’ (1 ,011; 7,4-methoxyestradiol (0.06); 8, 2-methoxyestradiol (0.05). *Peak no. 6 was also identified by GUMS (see sample 81582 in Table 3 and Figure 4).
Jellinck PH, Hahn EF, Norton BI, Fishman J (1984). Catechol estrogen formation and metabolism in brain tissue: comparison of tritium release from different positions in ring A of the steroid. Endocrinology 115:1850-1856. Hershcopf RJ, Bradlow HL, Fishman J (1986). Differential hydroxylation of estrone and estradiol in man. J Clin Endocrino/ Mrrtib 62:170-173. Fishman J, Norton B (1975). Catechol estrogen formation in the central nervous system of the rat. Endocrinology 96:1054-1059.
7.
8. 9.
Ball P, Haupt M, Knuppen R (1978). Comparative studies on the metabolism of oestradiol in the brain, the pituitary and liver of the rat. AC.IN Endocrinol (Copenh) 87: l-l 1. MacLusky NJ, Naftolin F, Krey LC, Franks S (1981). The catechol estrogens. J Steroid Biochem 15:l I I-124. Zumoff B, Fishman J, Cassouto J, Hellman L, Gallagher TF (1966). Estradiol transformation in men with breast cancer. J C/in Endocrinol
IO.
when using conventional techniques (i.e., gel filtration, enzymic hydrolysis, extensive column extraction and purification, ion exchange, and partition chromatography “j. To improve recovery and limit losses, we used acid alumina extraction”,‘* and optimized RP-HPLC’9,2” procedures for investigating free estrogens only. We have previously devised a computer-aided simulation method to optimize the mobile phase; on the basis of only seven chromatograms, it dictates the best chromatographic conditions for optimal resolution of several estrogen metabolites. ‘9-20*33 Therefore, RP-HPLC was used at the same time as analytic chromatography, performing the estrogen profile with EC detection, and as preparative chromatography, using single purified fractions for MS identification. We report the detection and clear identification of less than 5 pmol of any single CCE in both BCA tissue and BCF extracts; in addition, the acid alumina extrac442
Steroids,
1992, vol. 57, September
II.
12.
13. 14.
Metab
26:960-966.
Chattoraj SC, Fanous AS, Cecchini D, Lowe EW (1978). A radioimmunoassay method for urinary catechol estrogens. Steroids 31:375-391. Shimada K, Tanaka T, Nambara T (1981). Studies on steroids. CLXV. Determination of isomeric catechol estrogens in pregnancy urine by high-performance liquid chromatography with electrochemical detection. J Chromatogr 223:33-39. Schneider J, Huh MM, Bradlow HL, Fishman J (1984). Antiestrogen action of 2-hydroxyestrone on MCF-7 human breast cancer cells. J Bio/ Chem 259:4840-4845. Liehr JG (1990). Genotoxic effects of estrogens. Mutut Res 238~269-276. Reddy VVR, Hanjani P, Rajan R (1981). Synthesis of catechol estrogens by human uterus and leiomyoma. Steroids 37: 195-203.
IS.
16.
17.
Fotsis T, Adlercreutz H (1987). The multicomponent analysis of estrogens in urine by ion exchange chromatography and GC-MS-I. Quantitation of estrogens after initial hydrolysis of conjugates. J. Steroid Biochem 28:203-213. Adlercreutz H, Fotsis T, Hiickerstedt K, Hamalainen E, Bannwart C, Bloigu S, Valtonen A, Ollus A (1989). Diet and urinary estrogen profile in premenopausal omnivorous and vegetarian women and in premenopausal women with breast cancer. J Steroid Biochem 34:527-530. Spink DC, Lincoln DW II, Dickerman HW, Gierthy JF (1990). 2,3,7,8-Tetrachlorodibenzo-p-dioxin causes an extensive al-
GUMS
18.
19. 20.
21. 22. 23.
24. 25.
teration of 17/3-estradiol metabolism in MCF-7 breast tumor cells. Proc Nat1 Acad Sci USA 87~6917-6921. Castagnetta L, Granata OM, Brignone G, Blasi L, Arcuri F, Me&i M, ~Aquino A, Preitano W (1990). Steroid patterns of benign breast disease. Ann N Y Acad Sci 586:12f-136. D’Agostino G, Castagnetta L, Mitchell F, O’Hare M (1985). Computer aided mobile phase optimization and chromatogram simulation in HPLC: a review. J Chromarogr 338tl-23. Castagnetta L, Granata OM, Lo Casto M. Mitchell F. D’Agostino G, O’Hare M (1986). Steroid profiles and optimization of HPLC chromatog~ph~c analytic procedure. Ann N Y Acad Sci 464~316-330. Cartoni GP, Ciardi M, Giarrusso A, Rosati F (1983). Capillary aas chromatographic-mass spectrometric detection of anabolic steroids. J Chrokalogr 2791515-522. Juniewicz PE. Pallante Morel1 S. Moser A. Ewing LL (1988). Identification of phytoestrogens’in the urine of male dogs. J Sreroid Biochem 31:987-994. Davies IJ, Naftolin F, Ryan KJ, Fishman J, Siu 3 (1975). The alktity of catechol estrogens for estrogen receptors in the pituitary and anterior hypothalamus of the rat. Endocrinology 97:554-557. Vandewalle B, Peyrat JP, Bonneterre J, Lefebvre J (1985). Catechol estrogen binding sites in breast cancer. J Steroid Biochem 23~603-610. Lloyd T, Weisz J (1978). Direct inhibition of tyrosine hydroxylase activity by catechol estrogens. J Bioi Cham 25% 4841-4843.
26.
27.
28. 29. 30.
31.
32.
33.
of catechol estrogens:
Castagnetta
et al.
SchaefferJM, Hsueh AJW (1979). 2-Hydroxyestradiol interaction with dopamine receptor binding in rat anterior pituitary. J Biol Gem 2545606-5608. Naftolin F, Morishita H, Davies IJ, Todd R, Ryan KJ (1975). 2-Hydroxyestrone induced rise in serum luteinizing hormone in the immature male rat. Biochem Biophys Res Common 64:90.5-910. Merriam GR, Pfeiffer DG, Loriaux DL, Lipsett MB (1983). Catechol estrogen and the control of gonadotropin and prolactin secretion in man. J Steroid Biochem 19:619-625. Vandewalle B, Lefebvre J (f989). Opposite effects of estrogen and catecholestrogen on hormone-sensitive breast cancer cell growth and differentiation. Mot Cell Endocrinol61:239-246. Castagnetta L, Lo Casto M, Granata OM, Carruba G, Miserendino V, Calb M (1986). Metabolism of estradiol by estrogen responsive ZR75-1 cells. In: Baulieu EE, Iacobelli S, McGuire WL (eds), Endocrinology and Malignancy. Parthenon Publishing, Camforth, pp. 191-200. Martucci C, Fishman J (1977). Direction of estradioi metabolism as a control of its hormonal action-uterotrophic activity of estradiol metabolites. Endocrinolooav 101:1709-1715. Fotsis T, Jarvenpiil P, Adlercreutz H-(1980). Purification of urine for quantification of the complete estrogen profile. J Steroid Biochem 12503-508. D’Agostino G, Mitchel F, Castagnetta L, O’Hare MJ (1984). Solvent optimization for reverse-phase high-~~ormance liquid chromatography for polar adrenal steroids using computerpredicted retentions. J Chromarogr 30513-26.
Steroids,
1992, vol. 57, September
443
University
“La Sapienza,
” Rome,
Italy
Catecholestrogens (CCEs), namely 2- or 4-hydroxyestradiol and hydroxyestrone, are highly polar, reactive, and extremely labile estrogen metabolites in many experimental conditions. For these reasons, indirect assay methods mainly have been used. Some experimental evidence suggests that CCEs are synthesized and biologically active mostly in target cells. At this level, unfortunately, the indirect assays cannot be used. We present a method of gas chromatographiclmass spectral (GCIMS) analysis for the identijication of individual CCEs; the major fragmentation ions of authentic estrogen standards as trimethylsilylether derivatives, and the MS patterns of the major CCEs, namely, 2-hydroxyestradiol and hydroxyestrone, are included. Few examples of CCEs detected in human breast cancer tissues and in breast cystfluids are reported. Sample extracts were submitted to reversed-phase, high-performance liquid chromatography (RP-HPLC) and were quantified by “on line” electrochemical (EC) detection; thereafter, either crude extracts or single eluted peaks were submitted to GCIMS, by which detection limits of less than 5 pmol were attained. As expected, the molecular ion was the most relevant molecule in all but one case. On the contrary, the other relative intensities of major fragmentation ions M - 15, M -30, M - 90, and M - 15 + ( - 90) were unevenly distributed, although represented in the majority of cases. In all cases, the GCIMS ofpeakfractions, purtsed by RP-HPLC and WV detection, conJirmed the results of liquid chromatographic analysis combined with EC detection. in contrast, GCIMS of crude extracts was not equally satisfactory. Comparison of a liquid chromatography system with EC detection and the GCIMS approach revealed some inconsistency in quantitation of individual CCEs. Despite the fact that RP-HPLC with EC detection is sensible and versatile enough to analyze a number of free estrogen metabolites, the results of present study strongly suggest the need for a preliminary sample purification for CCE identt$cation by GCIMS. This represents an essential requirement for intratissue concentration studies. (Steroids 57:437-443, 1992)
Keywords: hormones and cancer; steroids; catecholestrogens; phy; gas chromatography/mass spectrometry
Introduction Catecholestrogens (CCEs) are peculiar metabolites of the parent “classic” estrogens, namely, estradiol (E2) and estrone (E,), which are produced by hydroxylation at either the C-2 or C-4 position, and are considered a major group of active metabolites of estrogens.‘32
Address reprint requests to Prof. Luigi A. Castagnetta at the Experimental Oncology and Molecular Endocrinology Units, c/o “M. Ascoli” Cancer Hospital Center, P.O. Box 636, Via de1 Vespro, 90127 Palermo, Italy. Received December 2, 1991; accepted April 25, 1992.
0 1992 Butterworth-Heinemann
reversed-phase,
high-performance
liquid chromatogra-
The catechol function (ie, a double adjacent OHgroup) in the aromatic ring confers to CCE high reactivity. Because of these physicochemical properties, CCEs exhibit a very wide range of polarities and are difficult to handle; thus, indirect measuring approaches have been used almost exclusively.3-5 An important feature of CCEs is that they are widespread, being present in many different tissues and organs. In particular, very high levels have been found in both the central nervous system (CNS) and liver.6-8 In addition, CCEs represent the major product of estroeven though their blood concentragen excretion.‘-” tions are reported to be 10w.~
Steroids,
1992, vol. 57, September
437
Papers Table 1 Systematic and trivial names of classic and main metabolites of estrogens: TMS numbers, times of authentic standard derivatives in gas chromatography/mass spectrometry analysis Systematic
Trivial name
name
1,3,5(10)-Estratriene-3,17/3-diol 1,3,5(10)-Estratrien-3-ol-17-one 1,3,5(10)-Estratriene-3,16~1,17p-trio1 1,3,5( 1 O)-Estratriene-2,3,17@triol 1,3,5(10)-Estratriene-3,4,17/3-trio1 1,3,5(10)-Estratriene-2,3-diol-17-one 1,3,5(10)-Estratriene-3,4-diol-17-one 1,3,5(10)-Estratriene-2,3,16a,l7/3-tetrol 1,3,5(10)-Estratrien-3,16cu-diol-17-one 1,3,5(10)-Estratriene-2,3,16a,l7p-tetrol 2-methyl ether 1,3,5(10)-Estratriene-2,3,17/3-trio1 2-methyl ether 1,3,5(10)-Estratriene-3,4,17p-trio1 4-methyl ether 1,3,5(10)-Estratriene-2,3-diol-17-one Z-methyl ether 1,3,5(10)-Estratriene-3,4-diol-17-one 4-methyl ether Abbreviation:
TMS, trimethylsilyl;
MW, molecular
weight;
molecular
weights,
and retention
No. of TMS
MW
RT
2 2 3 3 3 3 3 4 3 3 2 2 2 2
416 414 504 504 504 502 502 592 502 534 446 446 444 444
8.06 7.66 13.85 11.45 12.82 10.95 12.15 17.00 13.82 16.30 11.30 10.15 10.60 9.60
Estradiol Estrone Estriol 2-Hydroxyestradiol 4-Hydroxyestradiol 2-Hydroxyestrone 4-Hydroxyestrone 2-Hydroxyestriol 16~~Hydroxyestrone 2-Methoxyestriol P-Methoxyestradiol 4-Methoxyestradiol 2-Methoxyestrone 4-Methoxyestrone RT, retention times.
These pieces of evidence strongly suggest an overall biologic relevance of CCEs, which are mainly synthesized and biologically active in target tissues and cells. We have knowledge of a number of biochemical functions exerted by CCEs, including their antiestrogenie action exerted in vitro,” and of their genotoxic potential.13 In addition, their levels may be strongly modified in certain pathologic conditions.R,9s’4 However, assay methods are unsatisfactory and the biologic role(s) of CCEs in endocrine pathophysiology is still inadequately understood. All the above concerns indicate that these highly polar phenolic steroids deserve to be quantified in target tissues, implying the need for the direct identification of single estrogen metabolites as previously carried out, for example, in urineJ5,‘” and in in vitro systems.” To this end, we have devised a method of gas chromatography/mass spectrometry (GC/MS) analysis for the identification of CCE derivatives after their measurement in benign (cyst fluids) and cancerous breast tissues by reversed-phase, high-performance liquid chromatography (RP-HPLC) and “on line” electrochemical (EC) detection. Examples of the assessment of the major CCEs by GC/MS, after single peak collection from the RPHPLC system and UV detection, are presented.
Experimental Specimens Breast cancer (BCA) tissue specimens, obtained through radical mastectomy, and breast cyst fluid (BCF) specimens, obtained by fine needle aspiration, were used.
[namely, water and acetonitrile]) were obtained from Carlo ErbaFarmitalia (Milan, Italy). Before setting the chromatographic analysis conditions, a citric acid solution of the eluting system was passed through a cellulose nitrate membrane filter (0.22 pm) purchased from Whatman (Kent, UK). Ultrasphere-ODS (Beckman Inc., Fullerton, CA, USA) or Spherisorb 5s ODS-2 (Phase SEP, Clwyd, UK) columns (5pm particle size, 250 x 4.6 mm internal diameter) were used. N-methyl-N-trimethylsilyltrifluoroacetamide (MSTFA) and N-trimethylsilylimidazole (TMSI) were from Pierce (Luton, UK) and dithioerythritol was obtained from Serva (Heidelberg, Germany).
Apparatus An HPLC system obtained from Beckman Instruments Inc. (Berkeley, CA, USA) equipped with a model 1I2 pump, a model 210 sample injection valve, a model 160 UV detector (18.5~1 sample cell, mercury lamp, and a 280-nm filter), a two-channel recorder Omniscribe (Houston Instruments, Gistel, Belgium), and a fraction collector model 202 (Gilson Medical Electronics S.A., Villiers le Bell, France) was used. To concentrate aqueous extracts, a Speed Vat System (Savant Instruments Inc., Farmingdale, NY, USA) was used as evaporator-concentrator-dryer. An EC detector Coulochem model 5100A (ESA, Bedford, MA, USA) equipped with a 5020 model guard cell set at +0.50 V and a 5010 model analytical cell (5-~1 cell size) set at +0.75 V and containing two porous graphite coulometric electrodes was used. The GC/MS detection was performed using a GC-MS model 5890-5970 from Hewlett Packard (Rockville, MD, USA), equipped with a 5% phenyl-methyl silicone capillary column (20 mm x 0.2 mm internal diameter, 0.33-km film thickness) HP 190915-105.
Extruction procedures Materials All standard estrogens (Table l), including deuterated E,, were purchased from Sigma Chemical Co. (St. Louis, MO, USA). Chemicals for estrogen extraction procedures (i.e., acid alumina, tris(hydroxymethyl)aminomethane, hydrochloric acid, glacial acetic acid, diethylether, citric acid, and HPLC-grade solvents
438
Steroids,
1992, vol. 57, September
Tissues were processed fresh or after storage at -20 C in a sucrose/glycerol buffer. Dissected BCA tissue (1 g) was homogenized in 3 ml of 0.01 N perchloric acid and 3 ml of 0.1 M Tris buffer (pH 8.6) at a final pH of 3.4 and centrifuged at 2,000 x g for 10 minutes. The supernatant was decanted into a test tube containing 40 mg of acid alumina (pH 4.0) in 5.0 ml of 0.1 M
GUMS of catechol estrogens: Table 2
Mass spectral analysis of authentic estrogen derivatives:
Estradiol Estrone Estriol P-Hydroxyestradiol 4-Hydroxyestradiol 2-Hydroxyestrone 4-Hydroxyestrone 2-Methoxyestriol P-Hydroxyestriol 16wHydroxyestrone 2-Methoxyestradiol 4-Methoxyestradiol 2-Methoxyestrone 4-Methoxyestrone
416 414 504 504 504 502 502 534 592 487 446 446 444 444
(100)” (100)” (1 OOP (100)” (1 OOY (lOOja (100)” (100)” (lOO)a (100)’ (100)’ (100)” (lOO)a (100)a
285 399 311 373 373 155 155 129 433 502 315 315 155 155
(61 I (67jc (62) (9) (19) (16) (22) (27) (5) (25ja (16) (40) (25) (50)
232 155 129 267 325 319 319 375 385 231 416 129 261 429
Castagnetta et at.
the most intense peaks and relative intensities
(51 (45) (55) (6) (9) (9) (15) (22) (5) (IO) (7P (25) (12) (25)’
326 309 386 489 489 306 169 259 147 147 431 261 429 261
(12jb (24P’ (50) (5P (5)’ (9) (12) (15) (5) (IO) (6)’ (12) (1OY (20)
129 231 297 325 267 169 487 327 399 397 209 431 339 339
(12) (15) (40) (3) (5) (8) (lOjc (15) (4) (3) (6) (IOY (8jd (12)d
401 (9)” 324 (3)“ 324 (28) 129 (3) 129 (3) 487 (6)’ 306 (10) 519 (5Y 577 (4P
414 (18)’
129 (6) 416 (10P 414 (5P 414 (3P
397 397
(6)“ (8)‘j
356 372
(8jb (5)
aM+. b(M “(M d(M e (M
- 90)+. - 15)‘. - 15 - 30)‘.
19)+.
Tris buffer (pH 8.6). The CCEs adsorbed by acid alumina were extracted with 200 ~1 of 0.2 M acetic acid and dried in a Speed Vat system, while all other estrogens were extracted from Tris buffer with diethyl ether and dried at 40 C. Both extracts were combined just before HPLC analysis. The detailed steps of the extraction procedure for BCF samples have been described elsewhere.‘*
Reversed-phase, high-performance liquid chromatography and gas chromatographylmass spectrometry analytical conditions Dried extracts were redissolved in 100 ~1 of acetonitrile. Aliquots (10 to 20 ~1) were injected for RP-HPLC chromatographic analysis using our computer-aided optimized isocratic conditions (acetonitrile/O.OOl M citric acid, 40 : 60 v/v; 1.O ml/min flow rate), as previously detailed.‘9*‘” The sensitivity and response linearity of the EC detector and the relative retention times (RRTs) of all standard estrogens of interest have been previously reported.” To verify by W/MS the identity of the metabolites quantified in RP-HPLC plus EC, the single peak fractions were collected from the RP-HPLC system after UV detection; dried extracts were stored for a maximum of 1 week if they were not immediately processed. Just before the GC/MS analysis, equilin (50 ng) was added as internal standard to dried extracts redissolved in 50 ~1 of MSTFA/TMSI (1,000 : 2 v/v) containing 0.2% dithioerythritol, according to the methods we used previously for the anabolic steroids and their hydroxylated metabolites.*’ Samples were then heated for 30 minutes at 60 C and finally injected into the W/MS apparatus. The temperature of the injector, ion source and transfer line were set at 280, 200, and 280 C, respectively. The oven temperature was initially kept at 250 C for 13 minutes, then increased to a final temperature of 280 C (at a rate of 10 C/min) and held constant for 7 minutes. Samples (1 to 2 ~1) were injected with a splitter in a 1: 10 ratio. Helium was used as carrier gas at a flow rate of 0.9 ml/min. The energy of ionizing electron beam was 70 eV and electron impact (EI) mass spectra were recorded in the full scan mode from 100 to 600 amu. Using equilin as internal standard (IS), GC/MS quantitation of single peaks was carried out on a single ion of either IS or the
specific external standard (20 ppm). Calculations were performed through area measurement integration using the MSD 5890 software package.
Results As expected, the retention times (RTs) of the CCETMS derivatives, reported in Table 1, increased with the number of TMS groups in the molecule. A good separation was obtained for all the compounds examined by GC. Table 2 shows the mass/energy (m/z) values with the relative intensities obtained from EI mass spectra (in decreasing order). Intensity values below 3% of the base peak were not included.
502
4
6
8
?o
12
14
16
min Figure 1 (A) Mass spectra and (B) total ion chromatogram the authentic 2-hydroxyestrone derivative.
Steroids,
1992, vol. 57, September
of
439
Papers groups
as well
as from
the cleavage
of the
A. B. or II
rings.
As reported both in BCA
12
10
8
6
16
14
18
min Figure 2 (A) Mass spectra and (B) total ion chromatogram the authentic Z-hydroxyestradiol derivative.
of
7 t ;i a E m3 z ii 4 1 050lii~
, 8
10
..,...,, 12
z 5811
7
can be seen from Table 3 and from Figure 3. when crude extracts were submitted to GCIMS, no CCE derivatives were detectable. This was to be expected, since an adequate sensitivity of detection” may be achieved in the presence of low concentrations of single metabolites, only after preliminary purification. The GCYMS analysis confirmed identity of all peak fractions were collected from RP-HPLC after UV detection, with the exception of two BCA and one BCF cases, which had concentrations less than 0.25 nmolig or 0.25 nmol/ml, respectively. Comparison of quantitation of individual CCEs by EC detection “on line” with RP-HPLC and GCiMS did not yield equivalent results. This was especially true for crude extracts, even in the presence of an appreciable amount of CCEs. Both acid alumina extraction’* (Table 4) and RP-HPLC purification”’ greatly improved the recoveries and signal to background ratio for MS analysis of free CCEs. The evidence of 20H-E, in the breast cancer tissues examined is confirmed by both the presence of two characteristic ions (SO2 and 487 amu), obtained by the Selected Ion Monitoring (SIM) search (Figure 4). and the good coincidence of the RTs in the traces recorded. The same is shown in Figure 5 for the 20H-E, content of BCF. Again, the presence of two characteristic ions (504 and 489 amu) emerges, and RTs are coincidental.
_ 14--iF;l;,liJ
Table 3 Concentrations of catecholestrogens as calculated by (A) reversed-phase, high-performance liquid chromatography and electrochemical detection and (B) gas chromatography/ mass spectrometry
Ion 412 amu
3.104
in Table 3, CCEs have been detected tissues and in gross cyst fluids. As also
8
10
9 Ion 446amu
,r. ws.4~. .. . .-9.--w---.#.,, .h*.,..,_au_ , ..--7 8 9 10
Code no.
CCEs
A
B
5 oI.‘-. 4: 738
Ion 504 amu ~..*C~~~.~wC_~YU*.n~.~~ I”“#’ ‘.r-...r 13 12 11
+I---
T---y 15
min
Figure 3 Total ion chromatogram of individual ion trace of investigated CCEs from crude extract without any prior HPLC purification. Ions monitored at m/z 412 (M)‘, 446 (MI+, and 504 (M)+ for the equilin, 4-methoxyestradiol, and 4-hydroxyestrone derivatives, respectively.
The mass spectra of both 2-hydroxy-E, (20H-E,) and 2-hydroxy-E, (20H-EJ derivatives of authentic estrogen standards (recorded on the elution peak) and their total ion chromatograms (TICS), are illustrated in Figures 1 and 2, respectively. These data clearly indicate that in almost all the mass spectra, the molecular ion is the most intense peak. Formation of other ions may result from the loss of OH-Si-(CH3)3 (M-90) and/or CH3 (M-15, M-30) 440
Steroids,
1992,
vol.
57, September
Crude extracts 81251 81291 B1314 81314
(nmol/g) 20H-E, 20H-E, 20H-E, 40H-E,
2.76 0.76 1.79 0.89
ND ND ND ND
Purified peak fractions from BCA tissues (nmol/g) 81223 40H-E, 0.22 81265 20H-E, 2.46 81314 20H-E, 1.78 81314 40H-E, 0.89 81343 20H-E2 0.24 81548 20H-E2 2.38 81582 20H-E, 0.89
ND 2.35 1.67 0.91 ND 2.15 1.10a
Purified peak fractions from BCF (nmol/ml) BCFI 06 20H-E, 5.70 BCF115 40H-E, 0.11 BCFI 18 20H-E, 1 .20c
5.23’ 0.09 ND
Abbreviation: ND, not detected a See also Figure 4. ‘See also Figure 5. c Reported as pmoliml.
GUMS
of catechol
4
11
d
amu 12
mu
13
s . :! I
14 min
12
14 min
A typical EC detection of a RP-HPLC chromatographic profile of free estrogens from a sample of BCA tissue is presented in Figure 6.
Discussion The CCEs exert a number of biochemical actions, including (1) a wide spectrum of binding to estrogen receptors23*24;(2) the inhibition, for instance, of catecholamines or tyrosine hydroxylase2s; (3) the ability to compete for dopamine receptors in the pituitary26 and thus to modulate synthesis or secretion of gonadotropin, prolactin, and prostaglandins27*28; and, most important, (4) the ability to covalently bind macromolecules and even to produce DNA adducts.13 Other interesting pieces of evidence are that 20HE, may reduce proliferative activity of BCA cells in
Table 4 Recovery values (%) of some authentic dards using acid extraction method (pg)
1 0.5 0.05 0.01 Mean
t
SD
. . . . . . 11
11
Figure 4 Selected ion gas chromatogram of the purified peak fraction from a breast cancer tissue extract (sample 81582 in Table 3). Ions monitored at m/z 502 (M)’ and 487 (M - 15)+ for the 2-hydroxyestrone derivative.
Amount
. .
estrogen
stan-
20H-E,
El
2MeO-E,
82 90 90 a4
82 84 92 87
93 90 80 86
86.5 -c 4.1
86.5 2 4.0
87.3 2 5.6
14
Ion 504 amu .._........... 12 13 Ion
13
Castagnetta
et al.
16 min
l8
12
4...s.-U--” 11
12
10
8 Ion 502
estrogens:
12
489
amu 13
I4 min
_
.I
14 min
Figure 5 Selected ion gas chromatogram of the purified peak fraction from a BCF extract (sample BCF106 in Table 3). Ions monitored at m/z 504 (M)+ and 489 (M - 15)’ for the 2-hydroxyestradiol derivative.
vitro; thus, CCEs may behave as true natural antiestrogens.‘2,29 Furthermore, the early formation of CCEs by hormone-responsive human mammary cancer cell lines has been observed previously.30 As a consequence, the balance between 2/4 hydroxylation versus 16a hydroxylation, as previously shown3’ could be understood as a kind of regulation of proliferative activity of cancer cells, at least in the “endocrine growth control.” The inconsistency observed between plasma and urine levels reinforces the assumption that CCEs are mainly synthesized and exert a major biologic activity in target tissues. Consequently, it seems of more importance to analyze, identify, and quantify these metabolites in target tissues than to evaluate their blood levels. Identification and quantification of several endogenous estrogens are hardly practicable without a precise, undeniable confirmation of their identity. The data reported here indicate a good concordance between RP-HPLC with “on line” EC detection and GUMS in the identification of specific CCE metabolites. Looking at quantification, a certain inconsistency between the two different detection systems emerges. On one hand, this discrepancy could be attributed, in part, to nonoptimal recovery and losses during extraction and purification procedures; on the other hand, high background levels may strongly reduce the sensitivity limits of MS analysis. In spite of sufficient reproducibility, the recoveries for hydroxyestrogens appear to be limited to 50%32
Steroids.
1992. vol. 57. SeDtember
441
Papers
tion procedure
allows improved recovery. These reour combined RP-HPLC and GUMS approach for the simple, time-saving identification and quantitation of intratissular hydroxyestrogens.
5 i
sults validate
Acknowledgments The authors are indebted to the late Professor V. Pozzi for his enthusiastic encouragement and the provision of BCF samples. The authors also thank the Italian Association for Cancer Research (AIRC) for financial support. F. Arcuri is the recipient of a fellowship from the AIRC. References I. 2.
6
Fishman J, Bradlow HJL, Gallagher TF (1960). Oxidative metabolism of estradiol. J Biol Chem 235:3104-3107. Ball P, Knuppen R (1980). Catechol estrogens (2. and 4-hydroxyestrogens). Chemistry, biogenesis, metabolism, occurrence and physiological significance. Acta Endocrinol (Copenh) 93:1-127.
3.
Kono S, Merriam GR, Brandon DD, Loriaux DL, Lipsett MB (1982). Radioimmunoassay and metabolism of the catechol estrogen 2-hydroxyestradiol. J Ckn Endocrinol Metab 54:150-154.
4.
.s.
6. Figure 6 Reversed-phase, high-performance liquid chromatography estrogen profile by EC detection of a BCA tissue extract. (For analysis conditions, see text.) Peaks and quantities (nmol/ g tissue): 1, estriol (0.08); 2, 2-methoxyestriol (0.11); 3, 16~ hydroxyestrone (0.22); 4, P-hydroxyestradiol (1.04); 5, 4-hydroxyestradiol(4.16); 6,2-hydroxyestrone’ (1 ,011; 7,4-methoxyestradiol (0.06); 8, 2-methoxyestradiol (0.05). *Peak no. 6 was also identified by GUMS (see sample 81582 in Table 3 and Figure 4).
Jellinck PH, Hahn EF, Norton BI, Fishman J (1984). Catechol estrogen formation and metabolism in brain tissue: comparison of tritium release from different positions in ring A of the steroid. Endocrinology 115:1850-1856. Hershcopf RJ, Bradlow HL, Fishman J (1986). Differential hydroxylation of estrone and estradiol in man. J Clin Endocrino/ Mrrtib 62:170-173. Fishman J, Norton B (1975). Catechol estrogen formation in the central nervous system of the rat. Endocrinology 96:1054-1059.
7.
8. 9.
Ball P, Haupt M, Knuppen R (1978). Comparative studies on the metabolism of oestradiol in the brain, the pituitary and liver of the rat. AC.IN Endocrinol (Copenh) 87: l-l 1. MacLusky NJ, Naftolin F, Krey LC, Franks S (1981). The catechol estrogens. J Steroid Biochem 15:l I I-124. Zumoff B, Fishman J, Cassouto J, Hellman L, Gallagher TF (1966). Estradiol transformation in men with breast cancer. J C/in Endocrinol
IO.
when using conventional techniques (i.e., gel filtration, enzymic hydrolysis, extensive column extraction and purification, ion exchange, and partition chromatography “j. To improve recovery and limit losses, we used acid alumina extraction”,‘* and optimized RP-HPLC’9,2” procedures for investigating free estrogens only. We have previously devised a computer-aided simulation method to optimize the mobile phase; on the basis of only seven chromatograms, it dictates the best chromatographic conditions for optimal resolution of several estrogen metabolites. ‘9-20*33 Therefore, RP-HPLC was used at the same time as analytic chromatography, performing the estrogen profile with EC detection, and as preparative chromatography, using single purified fractions for MS identification. We report the detection and clear identification of less than 5 pmol of any single CCE in both BCA tissue and BCF extracts; in addition, the acid alumina extrac442
Steroids,
1992, vol. 57, September
II.
12.
13. 14.
Metab
26:960-966.
Chattoraj SC, Fanous AS, Cecchini D, Lowe EW (1978). A radioimmunoassay method for urinary catechol estrogens. Steroids 31:375-391. Shimada K, Tanaka T, Nambara T (1981). Studies on steroids. CLXV. Determination of isomeric catechol estrogens in pregnancy urine by high-performance liquid chromatography with electrochemical detection. J Chromatogr 223:33-39. Schneider J, Huh MM, Bradlow HL, Fishman J (1984). Antiestrogen action of 2-hydroxyestrone on MCF-7 human breast cancer cells. J Bio/ Chem 259:4840-4845. Liehr JG (1990). Genotoxic effects of estrogens. Mutut Res 238~269-276. Reddy VVR, Hanjani P, Rajan R (1981). Synthesis of catechol estrogens by human uterus and leiomyoma. Steroids 37: 195-203.
IS.
16.
17.
Fotsis T, Adlercreutz H (1987). The multicomponent analysis of estrogens in urine by ion exchange chromatography and GC-MS-I. Quantitation of estrogens after initial hydrolysis of conjugates. J. Steroid Biochem 28:203-213. Adlercreutz H, Fotsis T, Hiickerstedt K, Hamalainen E, Bannwart C, Bloigu S, Valtonen A, Ollus A (1989). Diet and urinary estrogen profile in premenopausal omnivorous and vegetarian women and in premenopausal women with breast cancer. J Steroid Biochem 34:527-530. Spink DC, Lincoln DW II, Dickerman HW, Gierthy JF (1990). 2,3,7,8-Tetrachlorodibenzo-p-dioxin causes an extensive al-
GUMS
18.
19. 20.
21. 22. 23.
24. 25.
teration of 17/3-estradiol metabolism in MCF-7 breast tumor cells. Proc Nat1 Acad Sci USA 87~6917-6921. Castagnetta L, Granata OM, Brignone G, Blasi L, Arcuri F, Me&i M, ~Aquino A, Preitano W (1990). Steroid patterns of benign breast disease. Ann N Y Acad Sci 586:12f-136. D’Agostino G, Castagnetta L, Mitchell F, O’Hare M (1985). Computer aided mobile phase optimization and chromatogram simulation in HPLC: a review. J Chromarogr 338tl-23. Castagnetta L, Granata OM, Lo Casto M. Mitchell F. D’Agostino G, O’Hare M (1986). Steroid profiles and optimization of HPLC chromatog~ph~c analytic procedure. Ann N Y Acad Sci 464~316-330. Cartoni GP, Ciardi M, Giarrusso A, Rosati F (1983). Capillary aas chromatographic-mass spectrometric detection of anabolic steroids. J Chrokalogr 2791515-522. Juniewicz PE. Pallante Morel1 S. Moser A. Ewing LL (1988). Identification of phytoestrogens’in the urine of male dogs. J Sreroid Biochem 31:987-994. Davies IJ, Naftolin F, Ryan KJ, Fishman J, Siu 3 (1975). The alktity of catechol estrogens for estrogen receptors in the pituitary and anterior hypothalamus of the rat. Endocrinology 97:554-557. Vandewalle B, Peyrat JP, Bonneterre J, Lefebvre J (1985). Catechol estrogen binding sites in breast cancer. J Steroid Biochem 23~603-610. Lloyd T, Weisz J (1978). Direct inhibition of tyrosine hydroxylase activity by catechol estrogens. J Bioi Cham 25% 4841-4843.
26.
27.
28. 29. 30.
31.
32.
33.
of catechol estrogens:
Castagnetta
et al.
SchaefferJM, Hsueh AJW (1979). 2-Hydroxyestradiol interaction with dopamine receptor binding in rat anterior pituitary. J Biol Gem 2545606-5608. Naftolin F, Morishita H, Davies IJ, Todd R, Ryan KJ (1975). 2-Hydroxyestrone induced rise in serum luteinizing hormone in the immature male rat. Biochem Biophys Res Common 64:90.5-910. Merriam GR, Pfeiffer DG, Loriaux DL, Lipsett MB (1983). Catechol estrogen and the control of gonadotropin and prolactin secretion in man. J Steroid Biochem 19:619-625. Vandewalle B, Lefebvre J (f989). Opposite effects of estrogen and catecholestrogen on hormone-sensitive breast cancer cell growth and differentiation. Mot Cell Endocrinol61:239-246. Castagnetta L, Lo Casto M, Granata OM, Carruba G, Miserendino V, Calb M (1986). Metabolism of estradiol by estrogen responsive ZR75-1 cells. In: Baulieu EE, Iacobelli S, McGuire WL (eds), Endocrinology and Malignancy. Parthenon Publishing, Camforth, pp. 191-200. Martucci C, Fishman J (1977). Direction of estradioi metabolism as a control of its hormonal action-uterotrophic activity of estradiol metabolites. Endocrinolooav 101:1709-1715. Fotsis T, Jarvenpiil P, Adlercreutz H-(1980). Purification of urine for quantification of the complete estrogen profile. J Steroid Biochem 12503-508. D’Agostino G, Mitchel F, Castagnetta L, O’Hare MJ (1984). Solvent optimization for reverse-phase high-~~ormance liquid chromatography for polar adrenal steroids using computerpredicted retentions. J Chromarogr 30513-26.
Steroids,
1992, vol. 57, September
443
