an spme–gc–ms method using an octadecyl silica fibre for the determination of the potential...

7
ORIGINAL PAPER An SPMEGCMS method using an octadecyl silica fibre for the determination of the potential angiogenesis modulators 17β-estradiol and 2-methoxyestradiol in culture media Federica Bianchi & Monica Mattarozzi & Maria Careri & Alessandro Mangia & Marilena Musci & Francesca Grasselli & Simona Bussolati & Giuseppina Basini Received: 4 December 2009 / Revised: 21 January 2010 / Accepted: 21 January 2010 / Published online: 20 February 2010 # Springer-Verlag 2010 Abstract A simple and easily automable method based on solid-phase microextraction followed by gas chromato- graphicmass spectrometric analysis was developed for the determination of two potential angiogenesis modulators 17β-estradiol (17-BE) and 2-methoxyestradiol (2-MEOE) in culture media. Trifluoroacetic anhydride was used as the derivatising agent. A homemade octadecyl silica coating, characterised by a coating thickness of 72±10 μm and a good thermal stability until 250°C, was prepared. Experi- mental design was used to optimise the extraction con- ditions in terms of derivatisation time, derivatisation temperature and time of extraction. As for method validation, lower limits of quantification of 0.17 and 0.015μg/l for 17β-estradiol and 2-methoxyestradiol, re- spectively, were obtained. Finally, the capabilities of the developed fibres were evaluated for the analysis of the investigated analytes developed by granulosa cells in culture media maintained under normoxic, hypoxic and anoxic conditions, in order to better elucidate their possible role in the angiogenic process. An increase of the production of both 17-BE and 2-MEOE in hypoxic and anoxic conditions seems to be related to the effect of oxygen deprivation. Keywords Angiogenesis . Solid-phase microextraction . 17β-Estradiol . 2-Methoxyestradiol . Derivatisation Introduction Angiogenesis is a matter of paramount importance: great attention is paid towards anti-angiogenic therapies to fight cancer and malignancies [1, 2], so a deeper understanding of the molecular control of angiogenesis is demanded not only to provide a novel approach to manipulate reproduc- tive function but also to control the factors responsible for the growth of solid tumours. The ovarian follicle represents a rare event of physiological neovascularisation and therefore is an outstanding model in which to study the molecular machinery responsible for new vessel growth. In particular, ovarian granulosa cells have been proven to be primarily involved in the angiogenesis regulation [3, 4], mainly by means of their steroid production [5, 6]. The presence of 17β-estradiol (17-BE) metabolites at low concentration levels in complex matrices like biological fluids requires the development of reliable and sensitive analytical methods in order to elucidate the role of these substances in health and diseases. Amongst estradiol metabolites, 2-methoxyestradiol (2-MEOE) has been recently investigated owing to its potential anti- angiogenic effect [7, 8]. Different methods based on liquid chromatography and gas chromatography have been developed to quantify this compound in biological samples [9-14]. F. Bianchi (*) : M. Mattarozzi : M. Careri : A. Mangia : M. Musci Dipartimento di Chimica Generale ed Inorganica, Chimica Analitica, Chimica Fisica, Università degli Studi di Parma, Viale Usberti 17/A, 43100 Parma, Italy e-mail: [email protected] F. Grasselli : S. Bussolati : G. Basini Dipartimento di Produzioni Animali, Biotecnologie Veterinarie, Qualità e Sicurezza degli Alimenti, Sezione di Fisiologia Veterinaria, Università degli Studi di Parma, Via del Taglio 8, 43100 Parma, Italy Anal Bioanal Chem (2010) 396:26392645 DOI 10.1007/s00216-010-3508-z

Upload: federica-bianchi

Post on 10-Jul-2016

215 views

Category:

Documents


3 download

TRANSCRIPT

ORIGINAL PAPER

An SPME–GC–MS method using an octadecyl silica fibrefor the determination of the potential angiogenesismodulators 17β-estradiol and 2-methoxyestradiolin culture media

Federica Bianchi & Monica Mattarozzi & Maria Careri & Alessandro Mangia &

Marilena Musci & Francesca Grasselli & Simona Bussolati & Giuseppina Basini

Received: 4 December 2009 /Revised: 21 January 2010 /Accepted: 21 January 2010 /Published online: 20 February 2010# Springer-Verlag 2010

Abstract A simple and easily automable method based onsolid-phase microextraction followed by gas chromato-graphic–mass spectrometric analysis was developed forthe determination of two potential angiogenesis modulators17β-estradiol (17-BE) and 2-methoxyestradiol (2-MEOE)in culture media. Trifluoroacetic anhydride was used as thederivatising agent. A homemade octadecyl silica coating,characterised by a coating thickness of 72±10 μm and agood thermal stability until 250°C, was prepared. Experi-mental design was used to optimise the extraction con-ditions in terms of derivatisation time, derivatisationtemperature and time of extraction. As for methodvalidation, lower limits of quantification of 0.17 and0.015µg/l for 17β-estradiol and 2-methoxyestradiol, re-spectively, were obtained. Finally, the capabilities of thedeveloped fibres were evaluated for the analysis of theinvestigated analytes developed by granulosa cells inculture media maintained under normoxic, hypoxic andanoxic conditions, in order to better elucidate their possiblerole in the angiogenic process. An increase of theproduction of both 17-BE and 2-MEOE in hypoxic and

anoxic conditions seems to be related to the effect ofoxygen deprivation.

Keywords Angiogenesis . Solid-phasemicroextraction . 17β-Estradiol . 2-Methoxyestradiol .

Derivatisation

Introduction

Angiogenesis is a matter of paramount importance: greatattention is paid towards anti-angiogenic therapies to fightcancer and malignancies [1, 2], so a deeper understandingof the molecular control of angiogenesis is demanded notonly to provide a novel approach to manipulate reproduc-tive function but also to control the factors responsible forthe growth of solid tumours.

The ovarian follicle represents a rare event of physiologicalneovascularisation and therefore is an outstanding model inwhich to study the molecular machinery responsible for newvessel growth. In particular, ovarian granulosa cells have beenproven to be primarily involved in the angiogenesis regulation[3, 4], mainly by means of their steroid production [5, 6].

The presence of 17β-estradiol (17-BE) metabolites atlow concentration levels in complex matrices likebiological fluids requires the development of reliableand sensitive analytical methods in order to elucidate therole of these substances in health and diseases. Amongstestradiol metabolites, 2-methoxyestradiol (2-MEOE) hasbeen recently investigated owing to its potential anti-angiogenic effect [7, 8]. Different methods based onliquid chromatography and gas chromatography have beendeveloped to quantify this compound in biologicalsamples [9-14].

F. Bianchi (*) :M. Mattarozzi :M. Careri :A. Mangia :M. MusciDipartimento di Chimica Generale ed Inorganica, ChimicaAnalitica, Chimica Fisica, Università degli Studi di Parma,Viale Usberti 17/A,43100 Parma, Italye-mail: [email protected]

F. Grasselli : S. Bussolati :G. BasiniDipartimento di Produzioni Animali, Biotecnologie Veterinarie,Qualità e Sicurezza degli Alimenti, Sezione di FisiologiaVeterinaria, Università degli Studi di Parma,Via del Taglio 8,43100 Parma, Italy

Anal Bioanal Chem (2010) 396:2639–2645DOI 10.1007/s00216-010-3508-z

As for the extraction procedure, the possibility ofoperating with small sample volumes is another aspectparticularly appealing in the study of physiological phe-nomena like angiogenesis. Sample handling, selectivity andpossibility of automation of the utilised methods canrepresent the bottleneck in the development of a reliableanalytical procedure. Recently, a solid-phase extraction(SPE) gas chromatographic–mass spectrometric (GC–MS)method has been developed and validated by our researchgroup to quantify 17-BE and its metabolites in porcinefollicular fluids [13]. Solid-phase microextraction (SPME)[15] can represent a valid alternative for the extraction ofthe investigated compounds, being able to perform anextraction/preconcentration step using reduced amount ofsample and being easily automable. Coating selectivity isanother parameter of great importance: many coatings havebeen recently developed to enhance the extraction perfor-mance of SPME towards target analytes using differentapproaches, from physical deposition to sol-gel technology[16-22].

To our knowledge, the information is lacking about thedevelopment of adequate analytical tools able to quantifythese estrogens in culture media: this knowledge appears ofoutstanding value to unravel the molecular events responsiblefor new vessel growth and regression.

The aim of this study was the use of a homemadeoctadecyl silica coating for SPME for the determination of17-BE and 2-MEOE potentially developed by granulosacells in culture media under normoxic, partial and totalhypoxic conditions to better elucidate their possible role inthe angiogenic process.

The capabilities of the obtained material were checked interms of film thickness, porosity, pH and thermal stability.Finally, the results obtained demonstrate the suitability ofthe octadecyl silica coating with respect to commerciallyavailable fibres for sampling 17-BE and 2-MEOE at tracelevels in biological samples.

Experimental

Chemicals

2-Fluoroestradiol (2-FE, internal standard 99.9% purity),17-BE and 2-MEOE, all 98% purity, were purchased fromSteraloids (London, UK). Trifluoroacetic anhydride (TFA,>99% purity), pyridine anhydrous (99.8% purity), sodiumchloride (99.5% purity), octane (>99% purity), decane(>98% purity) and Supelclean ENVI-18 SPE bulk packingwere from Sigma-Aldrich (Milan, Italy). Dulbecco’s mod-ified Eagle’s medium (DMEM)/Ham’s F12 was fromSigma-Aldrich and stored at −20°C. Heparin, ammoniumchloride, sodium bicarbonate, bovine serum albumin,

penicillin, streptomycin, amphotericin B, selenium andtransferrin were from Sigma-Aldrich.

17-BE and 2-FE were maintained at room temperature,whereas 2-MEOE and TFAwere stored at −20°C and undernitrogen at 4°C, respectively.

Stock solutions were prepared in acetone at the concen-tration of 1,000 mg/l and stored at −20°C for up to 5 weeks,whereas standard and working solutions were prepareddaily by dilution from the stock solutions.

SPME bare fused silica fibres with assembly, polydime-thylsiloxane (PDMS) 7 and 100 μm, PDMS/divinylbenzene(DVB) 65 μm and polyacrylate (PA) 85 μm fibres werepurchased from Supelco (Bellefonte, PA, USA)

Octadecyl silica fibre preparation

The octadecyl (C18) silica coating was immobilised on thesilica support of SPME fibres by using an epoxy resin glue.The coating was obtained by vertically dipping the silicasupports into the glue and subsequently in the SupelcleanENVI-18 silica for five times. Prior to use, the fibres weremaintained at room temperature for 12 h. Four fibres wereprepared.

Fibre characterisation

Thermogravimetric analysis (TGA) was performed using aTGA 7 instrument (Perkin-Elmer, Waltham, MA, USA)over the temperature range 30–400°C (heating rate 5°C/min) under inert (N2) atmosphere. Coating thickness andsurface morphology were investigated by using scanningelectron microscopy (SEM) with a Leica 430i instrument(Leica, Solms, Germany). Prior to use, all the fibres wereconditioned in the GC injection port at 250°C for 1 h undera helium flow.

pH resistance was evaluated by sampling 20 ng/l ofoctane and decane in water at pH=2, pH=7 and pH=11.Five replicated measurements for each pH value wereperformed.

Sample preparation and collection

Granulosa cells were aseptically harvested by aspiration ofswine ovarian follicles (diameter > 5 mm) with a 26-gaugeneedle, released in medium containing heparin (50 IU/ml),centrifuged for pelleting and then treated with 0.9%prewarmed ammonium chloride at 37°C for 1 min toremove red blood cells. Cell number and viability wereestimated using a haemocytometer under a phase-contrastmicroscope after vital staining with Trypan blue (0.4%) ofan aliquot of the cell suspension. Cells were seeded inDMEM/Ham’s F12 supplemented with sodium bicarbonate(2.2 mg/ml), bovine serum albumin (0.1%), penicillin

2640 F. Bianchi et al.

(100 IU/ml), streptomycin (100 μg/ml), amphotericin B(2.5 μg/ml), selenium (5 ng/ml) and transferrin (5 μg/ml).Seeded in 1 ml 24-well plates incubated at 37°C underhumidified atmosphere (5% CO2) for 24 h and thensubjected to normoxic (19% O2), hypoxic (5% O2) oranoxic (1% O2) conditions were 106 cells. Total hypoxiawas achieved by employing an Anaerocult® A mini,whilst partial hypoxia was obtained by means of anAnaerocult® C mini (Merck KgaA, Darmstadt, Germany);these experimental conditions were maintained for 18 h asrecommended by the manufacturer. In both cases, thesystem consisted of plastic pouches and a paper gas-generating sachet.

SPME analysis and on-fibre derivatisation

All the SPME experiments were performed by using amanual injection device. The C18 fibre was firstlyimmersed in a 2-ml amber vial containing the analytesand the internal standard in 1.5 ml of DMEM, for 45 min at40°C. A constant magnetic stirring was applied. The fibrewas then exposed in the headspace of a 50-ml vialcontaining 5 µl of the derivatising agent (TFA) and 5 µlof pyridine for 1 h at the temperature of 70°C. Desorptionwas carried out at the temperature of 250°C for 2 min. Afibre blank was run between each sample to avoid carryovereffects. The same procedure was applied using the PDMS7- and 100-μm fibres, the PA 85-μm fibre and the PDMS/DVB 65-μm fibre.

GC–MS analysis

An HP 6890 Series Plus gas chromatograph (AgilentTechnologies, Milan, Italy) equipped with an MSD 5973mass spectrometer (Agilent Technologies) was used for GC–MS analysis. Helium was used as the carrier gas at a constantflow rate of 1 ml/min; the gas chromatograph was operated insplitless mode with a Programmable Temperature Vapor-isation (PTV) injector (Agilent Technologies) maintained atthe temperature of 250°C and equipped with a PTV multi-baffled liner (ID 1.5 mm, Agilent Technologies). Chromato-graphic separation was performed on a 30 m×0.25 mm, df0.25 μm Factor Four-5MS capillary column (Varian, Turin,Italy). The following GC oven temperature programme wasapplied: 70°C for 0.5 min, 30°C/min to 210°C, 210°C for2 min, 50°C/min to 270°C, 30°C/min to 310°C for 1 min.Transfer line and source were maintained at the temperatureof 280 and 250°C, respectively. Preliminarily, full-scanelectron impact data were acquired to determine appropriatemasses for selected ion monitoring (SIM) mode under thefollowing conditions: ionisation energy 70 eV, mass range35–350 amu, scan time three scans per second. The massspectrometer was finally operated in time-scheduled SIM

mode by applying a delay time of 6 min and by recordingthe current of the following acetylated ions: from 6.00 to10.20 minm/z 482, 369 and 256 for 2-FE and m/z 464, 351and 309 for 17-BE and from 10.20 to 15 min m/z 494, 381and 339 for 2-MEOE.

For all the investigated analytes, the corresponding ionratios were used for confirmation purposes. A dwell time of100 ms was used for all the ions. All the analyses wereperformed by setting the electron multiplier voltage at1,900 V.

Signal acquisition and elaboration were performed usingthe HP Chemstation (Agilent Technologies).

Experimental design and optimisation procedure

The experiments were carried out on blank DMEM samplesspiked with 10 μg/l of both 17-BE and 2-MEOE.

A 23 two-level full-factorial design (FFD) was per-formed [23] to investigate the effects of time of extraction(te), temperature of derivatisation (Td) and time of deriva-tisation (td). Low and high levels were: te=15–45 min, Td=40–70°C and td=15–60 min. The order of experiments wasrandomised in order to avoid possible carryover effects ofthe analytical apparatus. An F test comparing the experi-mental and calculated responses at the centre of theexperimental domain was performed to evaluate theexistence of relevant quadratic effects, whereas a stardesign [24] was added to the factorial design experimentssince relevant quadratic effects were observed.

The final regression models were then calculated usingthe central composite design experiments obtained bothfrom the FFD and the star design. The best regressionmodels were obtained by a forward search stepwisevariable selection algorithm, and the optimal conditionswere evaluated by the global desirability D [25]: themaximum of D was determined by a grid search algorithm,estimating the responses by means of the regressionmodels.

All statistical analyses were carried out by using thestatistical package SPSS 10.0 for Windows (SPSS, Bologna,Italy).

Validation

Method validation was carried out to meet the acceptancecriteria for bioanalytical method validation [26]. DMEMwas used as blank matrix.

The lower limit of quantification (LLOQ) was calculatedas signal to noise ratio, S/N=5, using eight independentsamples and tested for accuracy and precision to meet thepreviously cited international criteria. The calibration curvewas evaluated by analysing blank DMEM samples spikedwith the investigated analytes (six concentration levels:

A SPME–GC–MS method using a C18 fibre for angiogenesis studies 2641

LLOQ, 0.5, 1.5, 2.75, 3.5 and 5 μg/l for 17-BE and LLOQ,0.04, 0.07, 0.11, 0.2 and 0.3 μg/l for 2-MEOE—threereplicated measurements for each level). Homoscedasticitywas verified by applying the Bartlett test. Lack-of-fit andMandel’s fitting tests were also performed to check thegoodness of fit and linearity [27]. The significance of theintercept (significance level 5%) was established byrunning a Student t test. For both the analytes, intra-batchand inter-batch precision were calculated, for each analytein terms of relative standard deviation percentage (RSD%)on three concentration levels (at the LLOQ, at 0.3 and10 μg/l), performing seven replicates at each level.Accuracy was calculated in terms of recovery rate (RR%)as follows:

RR% ¼ c1c2

� 100

where c1 is the measured concentration and c2 is theconcentration calculated from the quantity spiked into thesample. Three different concentration levels (low, mediumand high) with five replicated measurements were analysed.The extraction yield in terms of percent recovery wascalculated by comparing the results obtained from theSPME analysis of standard solutions (n=3) with thoserelated to the analysis of DMEM samples containing thesame amount of analytes (n=3).

Results and discussion

Characterisation of the octadecyl silica fibre

On the basis of the results achieved in an our previousstudy dealing with the determination of 17-BE and 2-MEOE in porcine follicular fluids in which a octadecylcartridge proved to be the best sorbent for solid-phaseextraction [13], a C18 SPME coating was prepared in ourlaboratory, with the aim to detect the investigated analytesusing a simple and easily automable procedure, thusimproving the more complex SPE approach. Taking intoaccount that C18 coatings are not commercially available,the fibres were obtained following a similar procedurereported by Lee et al. [21], in order to obtain in a verysimple and quick way the desired coating.

The thermal stability of the developed coating wasstudied by means of TGA, observing a good stability fromroom temperature to 250°C with a negligible weight loss.The thermal capability of the coating was also evaluated byconditioning the fibres in the GC injector port underdifferent temperatures, i.e., 100 and 250°C. Since underthese conditions no significant bleeding was observed, fourfibres were prepared and tested for the determination of 17-BE and 2-MEOE in culture media.

The morphology of the coating was also investigated bySEM under different magnifications (Fig. 1). The averagethickness (n=4) was found to be 72±10 μm with ahomogenous distribution of the silica C18 particles on theentire fibre surface. The coating resulted in a large surfacearea, able to provide a high extractive capacity.

pH resistance was also checked by using the developedfibres for the sampling of octane and decane in aqueoussolutions at different pH (pH=2, pH=7 and pH=11) for15 min at 30°C; each pH value was obtained using properbuffer solutions (NaHSO4/Na2SO4 for pH 2, NaH2PO4/Na2HPO4 for pH 7 and NaHCO3/Na2CO3 for pH 11) andmaintaining constant the ionic strength. ANOVA did notshow significant differences (p>0.05) amongst the obtainedmean responses (n=5 for each pH value), thus assessing thecapabilities of the developed coating for the sampling ofsolutions also at extreme pH conditions.

The performance of the developed fibres was alsoevaluated in terms of batch-to-batch repeatability: byoperating under these conditions, the four fibres were usedfor the analysis of 17-BE and 2-MEOE in a blank DMEMmatrix. RSD% lower than 10% were obtained also whensampling was performed along different days. The obtainedresults were very satisfying, taking into account thepresence of a derivatisation step, thus allowing us to assessthe feasibility of the used coating procedure in thedevelopment of chemically stable fibre.

Another advantage of the utilised coating relies on theabsence of swelling on exposure to the derivatising agent aswell as upon exposure to different solvent like water,methanol, acetone and their mixtures, thus making possiblethe use of the fibre for about 80 analyses. By contrast, thisphenomenon was particularly evident using other commer-cial devices like the PDMS/DVB or the PA ones (thePDMS 7- and 100-µm fibres were discharged as aconsequence of their low affinity towards the analytes),

Fig. 1 SEM image of the octadecyl silica coating

2642 F. Bianchi et al.

for which the swelling resulted in stripping of the coatingfrom the SPME assembly when it was retracted inside theneedle.

SPME optimisation and validation

A 23 two-level factorial design was preliminarily used toevaluate the significance of the main and interaction effectsof the parameters investigated. The experimental domainwas defined taking into account operative limits, namely:derivatisation temperature values higher than 70°C favourthe loss of the derivatised analytes; extraction times andderivatisation times greater than 45 and 60 min, respective-ly, would determine long analysis times.

In order to evaluate the repeatability of the measure-ments over the time, five replicates at the centre of theexperimental domain were added before and after

performing the factorial design experiments. For eachcompound, the main and interaction effects were calculated.

The presence of a significant curvature, being Fcalc

values higher than the Ftab value (Ftab(α=0.05; 1; 4)=7.7),indicated that a quadratic model had to be used, thusrequiring us to perform additional measurementscorresponding to a star design. The regression models(Table 1), calculated by stepwise regression analysis, werethen used to depict the response surfaces (Fig. 2) and tosearch for the highest global SPME–GC–MS responsewithin the explored domain. A global desirability function(D=0.86) was calculated, and the optimal extractionconditions were found in correspondence to an extractiontime of 45 min, a derivatisation temperature of 70°C and aderivatisation time of 60 min.

The method was then validated by using the experimen-tal setting providing the optimised conditions. LLOQvalues of 0.17 and 0.015µg/l were obtained for 17-BEand 2-MEOE, respectively, thus proving the potentiality ofthe method for the determination of the investigatedanalytes at trace levels.

Good linearity was proved in the LLOQ 5µg/l andLLOQ 0.3µg/l range for 17-BE and 2-MEOE, respectively,by applying Mandel’s fitting test. As for method precision,good results were achieved both in terms of intra-batch andinter-batch repeatability with RSD always lower than 10%with the exception of the LLOQ levels where RSD of 20%were obtained. However, taking into account that therequirements of the guidelines for the validation ofbioanalytical methods [26] state that, at the LLOQ levels,the analyte response has to be reproducible with a precisionof 20% and that these results were obtained by using four

Table 1 Regression coefficients (standard error in parentheses) of thepolynomial functions—expressed as GC peak areas (counts)—obtained during the optimisation process

17-BE

y=4,400 (±1,300) + 3,600 (±1,100) × te + 2,000 (±1,000) × Td + 3,000(±1,200) × te × td + 2,400 (±1,200) × Td × td + 2,700 (±1,200) × te ×Td × td + 3,100 (±1,700) × Td × Tdr2=0.77

2-MEOE

y=7,700 (±1,800) + 6,200 (±1,500) × te +8,500 (±1,500) × Td + 5,600(±2,300) × Td × Tdr2=0.81

te time of extraction, Td temperature of derivatisation, td time ofderivatisation

Fig. 2 Response surfaces of 17-BE and 2-MEOE depicted from the regression models calculated by stepwise regression analysis using DMEM asblank matrix

A SPME–GC–MS method using a C18 fibre for angiogenesis studies 2643

different fibres, the goodness of the developed coating/method can be assessed.

Finally, extraction yields of 101% (±10%) and 88%(±9%) were obtained for 17-BE and 2-MEOE, respectively,whereas recoveries in the 75% (±12%)–58% (±2%) (n=5)range proved the accuracy of the developed method.

Method application

Reliability of the developed method to the analysis of foursamples of granulosa cell culture media maintained undernormoxic, hypoxic and anoxic conditions (Table 2, Fig. 3)was assessed.

It has long been recognised that oxygen deprivation isthe prominent driving force of neovascularisation mediatedby angiogenic factors that recruit new blood vessels to theischaemic area.

In a previous work [28], we showed that during folliclegrowth swine granulosa cells are physiologically exposed tooxygen shortage as the increasing thickness of the avasculargranulosa layer results in a decrease of pO2 within thefollicle. Moreover, we evidenced [29] that vascular endothe-lial growth factor (VEGF) output by granulosa cellsincreases as the follicle grows: this production can be furtherstimulated by culturing in low-oxygen environment, thussuggesting that physiological hypoxia could be responsiblefor the highest levels of VEGF produced by cells from large

follicles. No information exists on the effect of oxygendeprivation on the synthesis of the previously demonstratedangiogenesis inhibitor 2-MEOE [7]. The data obtained byanalysing the cellular culture media samples (Table 2) seemto indicate an increase of the production of both 17-BE and2-MEOE in hypoxic and anoxic conditions, thus suggestinga well-known angiogenesis-promoting stimulus as oxygendeprivation could be able to modulate the synthesis ofhormones potentially involved in angiogenesis balancing. Itis of outmost interest that, despite considerable progressmade in the understanding of the pathways, which areactivated during cellular hypoxia, no consensus has beenreached on the mechanism by which O2 sensing is achieved[30, 31]. Since the mitochondrion is the major oxygen-consuming organelle, it might be expected to play a centralrole in oxygen-sensitive processes by varying the productionof reactive oxygen species (ROS) during hypoxia [32].These molecules, mainly in the form of O2

−, have long beenregarded as toxic products. However, nowadays, both invitro and in vivo studies indicate that angiogenic response istriggered by ROS signalling in a highly coordinated manner[32] pointing out a role for ROS as signal transducers [33].In particular, in a previous study [28], we evidenced that astimulation of O2

− triggers angiogenetic response in gran-ulosa cells and that 2-MEOE stimulates superoxide dismu-tase activity, thus inhibiting O2

− generation. Therefore,present data would help to investigate the importance ofredox-regulated signalling in angiogenesis.

However, due to the low number of samples and to thehigh intrinsic variability of the biological matrix, furtheranalyses as well as the correlation with other physiologicalparameters will be carried out in order to achieve a betterunderstanding of the investigated phenomena.

The development of this analytical method would leadus to improve the knowledge on the involvement of theexamined molecules in the angiogenesis balance.

Conclusions

A simple and easily automable headspace solid-phasemicroextraction method based on the use of an efficientoctadecyl silica fibre and on-fibre derivatisation with TFA

Fig. 3 SPME–GC–MS (SIM) chromatogram of a granulosa cellsample maintained under normoxic conditions

Table 2 Analysis of cell culture media samples maintained under normoxic (19% O2), hypoxic (5% O2) and anoxic (1% O2) conditions

Normoxia 17-BE (µg/l) 2-MEOE (µg/l) Hypoxia 17-BE (µg/l) 2-MEOE (µg/l) Anoxia 17-BE (µg/l) 2-MEOE (µg/l)

A 0.54±0.12 0.017±0.002 1.04±0.23 0.04±0.01 1.15±0.35 0.10±0.07

B 0.23±0.12 nd 0.21±0.03 0.02±0.01 0.27±0.03 0.045±0.007

C nd nd 0.37±0.01 nd 0.40±0.08 nd

D 0.34±0.03 0.018±0.003 0.39±0.02 nd 0.37±0.09 0.04±0.01

nd not detected

2644 F. Bianchi et al.

was optimised and validated for the GC–MS determinationof 17-BE and 2-MEOE at trace levels in culture medium.The preliminary results obtained by the analysis of culturemedia samples maintained under normoxic, hypoxic andanoxic conditions are expected to be useful in pathologicalevents where angiogenesis control is lacking, such asduring neoplastic processes. In particular, the developedmethod should help to improve knowledge about thepotential role of estradiol metabolites in the genesis ofmammary and other hormone-dependent cancers.

References

1. Folkman J, Klagsbrun M (1987) Science 235:442–4472. Folkman J (1996) Sci Am 275:150–1543. Santini SE, Basini G, Bussolati S, Grasselli F (2009) J Biomed

Biotechnol 2009:4198914. Bianco F, Basini G, Grasselli F (2005) Domest Anim Endocrinol

28:308–3195. Basini G, Santini SE, Grasselli F (2006) Ann N Y Acad Sci

1091:34–406. Basini G, Bussolati S, Santini SE, Bianchi F, Careri M, Mangia A,

Musci M, Grasselli F (2008) J Endocrinol 199:127–1357. Basini G, Bussolati S, Santini SE, Bianchi F, Careri M, Mangia A,

Musci M, Grasselli F (2007) Steroids 72:660–6658. Ricker JL, Chen Z, Yang XP, Pribluda VS, Swartz GM, Van Waes

C (2004) Clin Cancer Res 10:8665–86739. Zacharia LC, Dubey RK, Jackson EK (2004) Steroids 69:255–261

10. Castagnetta LA, Granata OM, Arcuri FP, Polito LM, Rosati F,Cartoni GP (1992) Steroids 57:437–443

11. Lakhani NJ, Sparreboom A, Dahut WL, Venitz J, Figg WD (2004)J Chromatogr B 806:289–293

12. Lakhani NJ, Lepper ER, Sparreboom A, Dahut WL, Venitz J, FiggWD (2005) RCM 19:1176–1182

13. Bianchi F, Careri M, Mangia A, Musci M, Santini SE, Basini G(2007) J Pharm Biomed Anal 44:711–717

14. Pinnella KD, Cranmer BK, Tessari JD, Cosma GN, VeeramachaneniDNR (2001) J Chromatogr B 758:145–152

15. Pawliszyn J (1997) Solid phase microextraction. Wiley-VCH,New York

16. Bianchi F, Bisceglie F, Careri M, Di Berardino S, Mangia A,Musci M (2008) J Chromatogr A 1196–1197:15–22

17. Bianchi F, Mattarozzi M, Betti P, Bisceglie F, Careri M, MangiaA, Sidisky L, Ongarato S, Dalcanale E (2008) Anal Chem80:6423–6430

18. Zeng J, Yu B, Chen W, Lin Z, Zhang L, Lin Z, Chen X, Wang X(2008) J Chromatogr A 1188:26–33

19. Djozan DD, Baheri T (2007) J Chromatogr A 1166:16–2320. Zhang Z, Duan H, Zhang L, Chen X, Liu W, Chen G (2009)

Talanta 78:1083–108921. Liu Y, Shen Y, Lee ML (1997) Anal Chem 69:190–19522. Azenha M, Ornelas M, Silva AF (2009) J Chromatogr A

1216:2302–230623. Box GEP, Hunter WG, Hunter JS (1978) Statistic for experi-

menters: an introduction to design. Data analysis and modelbuilding. Wiley, Hoboken

24. Khuri AI, Cornell JA (1987) Response surfaces: design andanalyses. Marcel Dekker, New York

25. Carlson R (1992) Design and optimization in organic synthesis.Elsevier, Amsterdam

26. US Department of Health and Human Services, Food and DrugAdministration (2001) Guidance for industry, bioanalytical meth-od validation. http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM070107.pdf

27. Funk W, Dammann V, Donnevert G (1995) Quality assurance inanalytical chemistry. VHC, New York

28. Basini G, Bianco F, Grasselli F, Tirelli M, Bussolati S, Tamanini C(2004) Regul Pept 120:69–75

29. Grasselli F, Basini G, Tirelli M, Cavalli V, Bussolati S, TamaniniC (2003) J Physiol Pharmacol 54:361–370

30. Chandel NS, Schumacker PT (2000) J Appl Physiol 88:1880–1889

31. Maxwell PH, Ratcliff PJ (2002) Semin Cell Dev Biol 13:29–3732. Maulik N, Das DK (2002) Free Radic Biol Med 33:1047–106033. Schroedl C, McClintock DS, Scott Budinger GR (2002) A J

Physiol Lung Cell Mol Physiol 283:922–931

A SPME–GC–MS method using a C18 fibre for angiogenesis studies 2645