B American Society for Mass Spectrometry, 2017 J. Am. Soc. Mass Spectrom. (2017) 28:1657Y1665DOI: 10.1007/s13361-017-1670-7

RESEARCH ARTICLE

DMSO Assisted Electrospray Ionization for the Detectionof Small Peptide Hormones in Urineby Dilute-and-Shoot-Liquid-Chromatography-HighResolution Mass Spectrometry

Péter Judák,1 Janelle Grainger,2 Catrin Goebel,2 Peter Van Eenoo,1 Koen Deventer1

1Department of Clinical Chemistry, Microbiology and Immunology, Doping Control Laboratory, Ghent University,Technologiepark 30 B, B-9052, Zwijnaarde, Belgium2Australian Sports Drug Testing Laboratory, National Measurement Institute, 105 Delhi Road, North Ryde, New South Wales2113, Australia

Abstract. The mobile phase additive (DMSO) has been described as a useful tool toenhance electrospray ionization (ESI) of peptides and proteins. So far, this techniquehas mainly been used in proteomic/peptide research, and its applicability in a routineclinical laboratory setting (i.e., doping control analysis) has not been described yet. Thiswork provides a simple, easy to implement screeningmethod for the detection of dopingrelevant small peptides (GHRPs, GnRHs, GHS, and vasopressin-analogues) withmolecular weight less than 2 kDa applying DMSO in the mobile phase. The gain insensitivity was sufficient to inject the urine samples after a 2-fold dilution step omitting atime consuming sample preparation. The employed analytical procedure was validatedfor the qualitative determination of 36 compounds, including 13 metabolites. The detec-

tion limits (LODs) ranged between 50 and 1000 pg/mL andwere compliant with the 2 ng/mLminimumdetection levelrequired by the World Anti-Doping Agency (WADA) for all the target peptides. To demonstrate the feasibility of thework, urine samples obtained from patients who have been treated with desmopressin or leuprolide and urinesamples that have been declared as adverse analytical findings were analyzed.Keywords: Peptides, High-resolution mass spectrometry, Doping control, HRMS, Urine, Dilute-and-shoot, DMSO

Received: 17 January 2017/Revised: 16 March 2017/Accepted: 19 March 2017/Published Online: 19 April 2017

Introduction

Over the past few years, peptides have gained a markedlyrelevant role in pharmaceutical research and development

[1]. Even though these new drugs are designed for specifictherapeutic indications, some of them can be used to improveperformance in sports. Consequently, several of them are in-cluded in the prohibited list published by WADA [2]. Withinthis class of peptide drugs, there is a group of peptides calledBsmall peptides.^ These peptides have a molecular weight ofless than 2000 Da. Their structure consists of short chains ofamino acids (AAs) linked together by peptide bonds as baseunits. In contrast to some bigger peptides, the naturally occur-ring amino acids are not the only building blocks, an entirerange of strategies based on chemical modifications (includingsynthetic AAs) are used to prolong their half-life [3].

These small peptides cover growth hormone releasing pep-tides (GHRPs), gonadotropin releasing hormones (GnRHs),vasopressin analogues, and agonists of growth hormone secre-tagogue receptors (GHSRs). It has been reported that GHRPscan boost testosterone levels by stimulating the release ofgrowth hormone [4], GnHR agonists can alter the productionof gonadotropin-releasing hormone [5], and vasopressin and itssynthetic analogue desmopressin can regulate urine productionand thus can be used as a masking agent [6]. The thymosinbeta-4 analogue TB-500 might help in faster recovery of mus-cle fibers and cells [7, 8]. Taking into account the potentialbenefits, these compounds are being used by athletes, andmethods to detect them should be developed.

Without doubt, sophisticated chromatographic techniquescoupled with mass spectrometry currently offer the most sen-sitive analytical platform to detect peptides at low concentra-tions [9]. Also in the field of anti-doping analysis, this tech-nique has become the gold standard for the analysis of peptideCorrespondence to: Péter Judák; e-mail: peter.judak@ugent.be

drugs [10]. Hence, several comprehensive LC-MS detectionmethods have been described for the detection of small pep-tides in human urine [11–16] and in equine urine [7, 17, 18].

As a sample clean-up, all of them use a solid phase extrac-tion (SPE) technology. Indeed, the polar nature of the peptidesdoes not allow them to be extracted by liquid–liquid extractionand SPE is a popular technique for peptide drugs [19].

Incorporating a sample preparation is often based uponcompromises to be able to extract as many compounds, eachwith a sufficient recovery and in line with required sensitivity.For example, Mazzarino et al. describes a laborious SPE opti-mization procedure, coming to the conclusion that two types ofelution mixtures are necessary for the simultaneous extractionof analytes with different properties [16]. Such off-line samplepreparations are time-consuming, expensive, in cartridges, andprone to increase the likelihood of human error as well. Theseextraction issues were avoided by Thomas et al. in 2016 using atwo-dimensional liquid chromatographic system coupled withan ion mobility mass spectrometer [14]. The disadvantage ofsuch a system is the need for a double pump, which increasesinstrument and maintenance costs. Another promising ap-proach is using a simple dilution as sample preparation (di-lute-and-shoot) and subsequent injection in a conventional LC-MS system. Such an approach has been reviewed before [20].However, due to the absence of a preconcentration step, thesensitivity of a dilute-and-shoot LC-MS method relies for amajor part on the ionization of the targeted compounds [20].Hence, a higher ionization efficiency has the potential to over-come sensitivity issues and consequently provide desired de-tection limits. Recently, an approach was presented to improveionization efficiency of small peptides by Hahne et al. [21].They proposed that the addition of low percentages of DMSOto the mobile phase enables the detection of peptides with poorionization properties. Therefore, the aim of this work was toinvestigate the sensitivity of a dilute-and-shoot LC-MSmethodfor small peptides using DMSO as a mobile phase additive.

ExperimentalChemicals and Materials

[deamino Cys1, Val4, D-Arg8] Vasopressin was purchasedfrom Sigma Aldrich (Bornem, Belgium). TB 500 and its me-tabolites were synthesized in-house [8]. All the other peptideswere obtained from the Australian Sports Drug Testing Labo-ratory (Sydney, Australia) and were synthesized by Auspep(Tullamarine, Australia). Bovine insulin was purchased fromSigma Aldrich (Bornem, Belgium). LC-MS grade acetonitrile(ACN) and water were obtained from J.T. Baker (Deventer,The Netherlands), formic acid (HCOOH) from Fisher Chemi-cal (Madrid, Spain), acetic acid (HOAc) and dimethyl sulfox-ide (DMSO) from Sigma Aldrich (Bornem, Belgium).

Stock and working solutions were prepared and stored inlow binding microcentrifugal vials, type Eppendorf(Eppendorf, Hamburg, Germany). For performing the dilutionsteps, low retention pipette tips (Eppendorf) were used.

A mixture consisting of 90% water, 10% acetonitrile, andbovine insulin at the concentration of 1 μg/mL was preparedand used for the dilution of the samples. This solution isreferred to as Bdilution mixture^ in this study.

Instrumentation

The HPLC system consisted of an Accela LC (Thermo Scien-tific, Bremen, Germany) equipped with a degasser, Accela1250 UPLC pump, an autosampler thermostated at 15 °C anda thermostated (35 °C) column compartment. An AgilentZorbax RX C8 (2.1 × 150 mm) 5 μm particle size columnwas used for LC separation. Mobiles phases were A: H2O(containing 0.2% formic acid and 1% DMSO), and B: aceto-nitrile (containing 0.2% formic acid/ 1% DMSO). Gradientelution was as follows: 1% B, held for 4 min and increased to35% in 20min, then further increased to 90% in 2 min and heldthere for 3 min, decreased to 1% in 0.1 min and equilibrated for3.9 min, giving a total runtime of 29 min. The flow rate was 0.3mL/min.

The LC-system was coupled with a Q-Exactive benchtopquadrupole Orbitrap tandem mass spectrometer (Thermo Sci-entific, Bremen, Germany). The mass spectrometer wasequipped with an electrospray ionization source. The parame-ters of the ion source were the following: spray voltage 4 kV;capillary temperature 250 °C; heater 30 °C; sheath gas 50;auxiliary gas 30; S-lens 100. The instrument was set to operateboth in positive full scan and in positive multiplexed targetedion monitoring. In full scan mode, the settings were the fol-lowing: resolution 70,000; scan range 300–1075; agc target1e6; maximum ion injection time 200 ms. In target sim mode,the settings were the following: resolution 70,000; scan range300–1075; agc target 2e5; maximum ion injection time 200ms;isolation window 2.5 m/z; msx 10. The mass extraction wasdone at 5 ppm. For the experiments to study charge statecoalescence, the above described full scanmonitoring was usedwith the change of scan range to 200–1500 m/z.

Preparation of Stock and Working Solutions

All peptide stock standard solutions were prepared at 0.1 mg/mL using aqueous 2% acetic acid and were stored at –20 °C.Working solutions (50, 25, 10 ng/mL) were prepared using thedilution mixture.

Excretion Study and Positive Urine Samples

As a proof of concept, five urine samples from patients whoreceived desmopressin were obtained from the Department ofInternal Medicine (Endocrinology) and Hematology of theGhent University Hospital, with the approval of theEthical Committee of the Ghent University (reference:B67020108809). Three patients received desmopressin intra-nasally (~10 mg/dose), two orally (200 mg).

Five urine samples from healthy male volunteers, 18 y andolder, who were treated with leuprolide (1 mg subcutaneouslyinjected) were analyzed. Ethics approval was granted by the

1658 P. Judák et al.: Detection of Small Peptides by Ds-Lc-Hrms

Sydney South West Area Health Service Human ResearchEthics Review Committee (Concord) consistent with NationalHealth and Medical Research Council guidelines forethical research involving humans and registered at the Aus-tral ian and New Zealand Clinical Trial Registry(ACTRN12609000629235). Two samples that were declaredpositive for GHRP 6 and one for GHRP 2 were also analyzed.Samples were stored at –80 °C awaiting analysis.

Sample Preparation

To 100 μL urine, 100 μL dilution mixture was added. Allsamples were mixed and centrifuged at 12,500 g-force for5 min preceding the LC-HRMS analysis. Then, themicrocentrifuge tubes were placed in the autosampler and 30μL of the supernatant was injected. [deamino Cys1, Val 4, D-Arg8] AVP and 13C, 15N GHRP-2 (1–3) were used as internalstandards and were added to the samples prior to the centrifu-gation step (2 ng/mL).

Preparation of Samples for Peptide CarrierExperiment

To investigate the effect of bovine insulin as a carrier peptide,two dilution series were prepared in the following manner: a 1μg/mL mix of all peptides was prepared in duplicate, bothcontaining ephedrine at the concentration of 100 μg/mL, butonly to one of them bovine insulin was added at the concen-tration of 1 μg/mL. These solutions were further diluted to 250;50; 10; 5; 2,5 ng/mL, making sure that the concentration of thecarrier peptide was kept constantly at 1 μg/mL in one of thedilution series. At each dilution step, bovine insulin was addedfirst to the vial. Finally to 100 μL of each solution 100 μL of anaqueous solution containing betamethasone was added andvortexed prior to the LC-MS analysis. Including ephedrineand betamethasone in the experiment guaranteed that the in-creased peak areas can only be attributed to the use of bovineinsulin, not to low precision of injection volumes of theautosampler, or to human error committed during the dilutionof the solutions.

Validation

Validation was carried out according to Eurachem Guidelines[22]. Therefore, 10 different blank urine samples were collect-ed in a representative way covering a pH range of 5.08–6.24and density of 1.012–1.031 g/cm3. Each urine sample wasspiked at 2, 1, 0.5, 0.250, 0.125, 0.050 ng/mL to determinethe LOD of the method, which was considered the lowestconcentration at which the targeted analyte could be detectedin all 10 urine samples with a signal-to-noise ratio greater than3 and a retention time difference of less than 0.3 min to thereference.

To obtain concentrations of 2, 1, 0.5 ng/ml, 480, 490, 495μL of blank urine was spiked from the 50 ng/mL workingsolution by the addition of 20, 10, 5 μL, respectively, to obtainpeptide concentrations of 0.250, 0.125 ng/mL 495, 497, 5 of

μL of blank urine was spiked from the 25 ng/mL workingsolution by adding 5, 2,5 μL, respectively, to obtain concen-trations of 0.050 ng/mL 497,5 μL of blank urine was spikedfrom the 10 ng/mL working solution by the addition of 2.5 μL.

Retention time stability was assessed by comparing theretention time of the target analytes in matrix and in neatsolutions without matrix.

Selectivity was tested during the validation procedure toprobe for interfering peaks in the selected ion chromatogramsat the expected retention times of the different peptides. For thisreason, during the validation, 13 different mixtures of sub-stances (in total 327 substances) from other doping classeswere checked. Additionally, the 10 urine samples not spikedwith the peptides were analyzed for interfering peaks. Matrixeffects were expressed as the ratio of the average peak area ofan analyte in 10 spiked blank urine samples (2 ng/mL) to thepeak area of the same analyte in neat standard solution, multi-plied by 100. A value greater than 100% indicates ionizationenhancement and a value less than 100% indicates ionizationsuppression. Carryover was assessed by injecting blank sam-ples (solvent used for dilution) after each urine sample duringthe validation process. The robustness of the method wasdemonstrated by analyzing several batches of routine urinesamples that have been analyzed in our laboratory. Identicalanalytical columns but different in state of use were tested tocheck for increase of pressure.

Results and DiscussionMethod Development

Since no preconcentration step is used in a dilute-and-shootstrategy, sensitivity relies on the ionization behavior of thetarget compounds, sensitivity of the mass spectrometer, andthe stability of the compounds in solution. Therefore, these keyelements were investigated during method development phase.

Effect of DMSO on the Electrospray Ionization

It has been revealed before that the nature of the chromato-graphic solvent can influence the process of the formation ofelectrospray droplets [23–25]. DMSO is one of the mobile-phase additives that has been used in proteomic experiments byseveral research groups to enhance electrospray response [21,26, 27]. One of the beneficial effects of including such solventsin the eluent is a remarkable gain in sensitivity that has beenreported to be mainly caused by improved ionization efficien-cy, simplymeaning that more ions enter the mass spectrometer.[21]

To investigate this, a solution containing several of thepeptide analytes at 2 ng/mL was injected twice using the samemethod but with the change of the addition of DMSO to boththe aqueous and organic solvents. In Figure 1, a comparison ofthe signal intensities of the most abundant charge states arepresented.

P. Judák et al.: Detection of Small Peptides by Ds-Lc-Hrms 1659

On average signal intensities were 2–5 times higher with theDMSO-containing mobile phase. The improved sensitivity wassignificant but was less than the 10-fold increase observedelsewhere [21]. The highest increase was observed in the inten-sity of Lys-vasopressin (which could not be detected with con-ventional mobile phase compositions). This peptide has not beenincluded in previously published methods, probably because oflimited ionization capacities in absence of DMSO [14, 16].

Besides the increased signal intensities, we observed that theuse of DMSO also resulted in the coalescence of ion currentinto fewer charge states, more precisely in charge state reduc-tion leading to more abundant lower charge states. Meyer andKomives explained the observed ESI MS signal improvementwith this phenomenon, while Hahne et al. stated that it onlyaccounts for merely 10%–20% of the gain in sensitivity and iscaused by the high gas-phase basicity of DMSO (charge-stripping) [21, 26].

To study this phenomenon further, several peptides at ahigher concentration (1μg/mL) were injected in the respective

eluents and data was collected for each observable charge state.Examples of charge state reduction are demonstrated in Fig-ure 2. For alexamorelin (Figure 2a) without DMSO, the [M +H]3+ ion was the most abundant; with DMSO, the [M + H]2+

ion became the most abundant. For desmopressin (Figure 2b), areduction from +2 to +1 was noticed. A less prominent chargecompression was observed for GHRP 6 (Figure 2c). Chargestate reduction could not be observed (or it was too low to bedetected) for perforelin (Figure 2d) when the comparison of thecharge states was carried out. The highest detectable +3 chargestate was the most abundant, irrespective of the mobile phase.This less outstanding effect could possibly be explained by thefact that perforelin has a structure with the charge (protons) tobe more localized; therefore, it is less prone to be stripped ingas phase [21].

However; it has to be noted that as the focus of this work ison molecules with relatively short AA chain (5–10 AAs), ourinvestigation has its limitations. The peptides of our interest arealmost all present in +1 or +2 charge state. The infusion of

Figure 1. Peptide ionization boosting effect of DMSO: Comparison of MS signal intensities of different GHRPs. MS signalintensities indicated with red correspond to the experiments that were carried out with DMSO in the chromatographic mobile phase

1660 P. Judák et al.: Detection of Small Peptides by Ds-Lc-Hrms

analytes with longer AA chains would probably lead to a betterand more comprehensive understanding of the observed com-pression of charge states but that was not the goal of thepresented work.

Peptide-Stability, Storage Conditions

Urinary stability of small peptides has been investigated byseveral research groups. Mazzarino et al. report that fortifiedurine samples (5 ng/mL) are stable at least for 2 weeks at 4 °C[16]. Timms et al. observed that after 7 d of storing the spikedhuman urine samples (50 ng/mL) at room temperature, GHRPlevels between 38% and 100% were detected [17]. Thomaset al. found urine samples (20 ng/mL) stable for 3 wk at –20 °C[14]. Ekatrine et al. recommend storage of urine samples at –80°C instead of –20 °C since a drastic decrease was reported after6 mo for GHRP 1 metabolites [12]. Besides this urinary stabil-ity, it has also been published that peptides can be lost from asolution due to absorption to recipient surfaces [28]. To avoidthese losses and degradation effects, several precautions weretaken. Because of the more basic properties of the targetanalytes, an aqueous solution containing 2% of acetic acidwas used to dissolve the compounds. Addition of smallamounts of organic solvents is also advised if hydrophobicresidues approach 50% or more. Hence also 10% ACN wasadded to all solutions [29].

Another phenomenon is that with increasing the dilutionfactor, unspecific losses can increase. To avoid this, a carrierpeptide can be used [30]. Bovine insulin was arbitrarily chosenas carrier peptide since it does not interfere with the analysis. InFigure 3, the effect of the carrier peptide is shown through theexample of GHRP 5. It is clearly seen that decreased signalintensities might not be noticed when working with high con-centrations; the issue becomes significant at low solute

concentrations. While at 2,5 ng/mL the peak area is increasedup to five times in the sample with bovine carrier, at 250 ng/mLthe improvement is less significant (1, 20 times). Maybe theaforementioned stability issues observed by the other researchgroups in urine can be due to this phenomenon.

It has to be noted that while the majority of the investigatedanalytes followed the same trend (2- to 5-fold peaks are in-creased at the lowest concentration and less or no improvementat higher concentrations), not all the of them could be recoveredfrom the solution as good as the displayed GHRP 5 peptide. Noimproved peak area was observed in the case of ipamorelin (1–4) free acid, GHRP 2 (1–3) free acid, leuprolide (5–9), GHRP 1(2–4) free acid, ipamorelin, leuprolide (1–3), terlipressin, Arg 8vasopressine, hexarelin (1–3), Lys-vasopressin, TB-500 or itsmetabolites.

Method Validation

To guarantee that all doping laboratories can report the pres-ence of prohibited substances, their metabolites, or markers in auniform way, a minimum routine detection capability for test-ing methods has been established by WADA called the mini-mum required performance limit (MRPL) [2]. A developedmethod should be compliant to this MRPL.

The currentMRPL level for small peptides is set at 2 ng/mL.The validated method presented in this paper shows goodsensitivity and is capable of detecting 36 peptides including12 metabolites at least at this MPRL. Table 1 summarizes theresults of the validation. For 23 out of 36 compounds, theLODs were at least 10 times lower than the MRPL. TheseLODs were in the same order of magnitude as previousstudies, but were obtained without solid phase extractionor with the aid of a 2D system [12, 14, 16]. It wasobserved that for some compounds, LODs of parent drug

Figure 2. Comparison of peak areas of different charge states of alexamorelin (a), desmopressin (b), GHRP 6 (c), and perforelin (d).Peak areas that were generated using DMSO are indicated with red, peak areas that were generated without DMSO are indicated byblue

P. Judák et al.: Detection of Small Peptides by Ds-Lc-Hrms 1661

and metabolites show a significant difference. While theLOD value for GHRP 6 parent ion is only 1000 pg/mL,

that of the metabolites was found to be five or even 10times better.

Figure 3. The effect of using a carrier peptide. Columns in red correspond to samples with bovine carrier, columns in bluecorrespond to the samples without using bovine carrier

Table 1. The Investigated Peptides, Showing Elemental Composition, Most Abundant Charge State in the Presence of DMSO, LOD, and Matrix Effect Values.Nonvalidated Peptides are Labeled as NV

Name Elemental composition Charge state LODng/mL

Matrix effect%

Alexamorelin C50H63O7N13 2 0.125 49.3Anamorelin C31H42N6O3 1 0.125 41.1AOD9604 C78H123N23O23S2 3 0.25 28.6Arg 8 Vasopressin C46H65N15O12S2 2 0.5 25.0Buserelin C60H86N16O13 2 0.125 40.1Deslorelin C64H83N17O12 2 0.125 35.2Desmopressin C46H64N14O12S2 1 0.5 59.0Goserelin C59H84N18O14 2 0.25 37.1GHRP-1 C51H62N12O7 2 0.25 34.4GHRP-1 (2-4) free acid C22H25N5O4 1 0.5 55.4GHRP-1 (3-7) NH2 C42H50N8O5 2 1 33.0GHRP-2 C45H55N9O6 2 0.05 29.2GHRP-2 (1-3) free acid C19H23N3O4 1 0.5 41.0GHRP-4 C34H37N7O4 1 0.05 51.8GHRP-5 C43H46N8O6 1 1 22.8GHRP-6 C46H56N12O6 2 1 38.6GHRP-6 free acid C46H55N11O7 2 0.125 35.4GHRP-6 (2-5) free acid C34H36N6O5 1 0.5 40.3GHRP-6 (2-6) free acid C40H50N9O5 2 0.5 30.7Hexarelin C47H58N12O6 2 0.25 43.7Hexarelin (1-3) free acid C21H26N6O4 1 0.5 35.5Hexarelin (2-5) free acid C35H38N6O5 1 0.25 32.3Histrelin C66H86N18O12 3 0.125 40.8Ipamorelin C38H49N9O5 2 0.05 47.6Ipamorelin (1-4) free acid C32H36N6O5 1 0.25 48.4Leuprolide C59H84N16O12 2 0.5 36.6Leuprolide (5-9) C34H57N9O6 2 1 42.9Leuprolide (1-3) C22H24N6O5 1 0.05 50.7LHRH C55H75N17O13 2 0.25 36.3Lys-Vasopressin C46H65N13O12S2 2 0.5 4.9Nafarelin C66H83N17O13 2 0.25 34.3Nafarelin (5-10) C41H56N10O7 2 0.25 33.0Peforelin C59H74N18O14 3 0.05 39.6TB 500 (NV) C38H68N10O14 2 0.5 52.2TB 500 (1-3) C20H39N5O5 1 0.25 61.8TB 500 (1-5) (NV) C29H53N7O10 2 0.5 75.2Terlipressin C52H74N16O15S2 2 0.25 27.3Triptorelin C64H82N18O13 2 0.25 27.8

1662 P. Judák et al.: Detection of Small Peptides by Ds-Lc-Hrms

The test for the assessment of the selectivity of the methodshowed no interfering peaks at the expected retention times ofthe analytes. Nevertheless, for confirmation purposes, a meth-od with a higher degree of selectivity and using tandem massspectrometry is recommended [31].

The results of matrix effect test are listed in Table 1. Greatvariation was observed across the 10 analyzed urine samples.For example, for AOD9604 the calculated values varied be-tween 5% and 55%, giving an average of 28.6%.

This was also the observation of a previous research usingdirect injection [14] and was discussed by Panuwet et al.,stating that urine concentration can vary drastically acrossindividuals, depending on a variety of factors, including dietand hydration status [32].

No carryover was observed in the blank samples injectedafter the urine samples. Despite the relatively large (30 μL)injection volume and only ½ dilution factor, no increase inpressure was noticed while running several batches of valida-tion and routine urine samples. Possibly, non-dissolved parti-cles are efficiently removed by the centrifugation step.

Evaluating the retention time criteria showed that 36 out of38 compounds met the set requirement. For the early elutingTB-500 and one of its metabolite TB-500 (1–5), a retentiontime shift of more than 1 min in combination with deformedpeak shape was observed. Timms et al. reported also peakdeformation for small peptides due to the presence of the urinematrix [17]. The two substances were considered as not-validated (NV). The detection of TB-500 was covered by

Figure 4. Extracted ion chromatograms of a blank urine sample (a), an adverse analytical finding (b), and a blank urine samplefortified at 2 ng/mL with the parent and metabolite peptides (c). The first row corresponds to GHRP-6, the second to GHRP-6 freeacid, and the third to GHRP-6 (2–5)

P. Judák et al.: Detection of Small Peptides by Ds-Lc-Hrms 1663

monitoring the TB-500 (1–3), which did not show such aproblem. Besides, the latter substance showed also a betterLOD.

Excretion Study Samples-Routine Samples

The suitability of the developed method for target analysis wasproven by analyzing samples previously declared positive forthe presence of small peptides and authentic samples fromcontrolled administration studies. GHRP 2 and GHRP 6 parentions could be detected in all corresponding excretion urinesamples. Additionally GHRP-6 free acid and GHRP-6 (2–5)free acid metabolite for GHRP-6 were also present in thesample (Figure 4). The identification and characterization ofthese metabolites have been studied before [13]. The presenceof desmopressin and leuprolide was also confirmed in allcorresponding excretion samples.

ConclusionA simple comprehensive screening method for the detection ofsmall peptides has been successfully developed and validated.The addition of DMSO reagent to the mobile phase enhancespeptide ionization and allows applying a dilute-and-shoot strat-egy. The main benefits of the dilute-and-shoot strategy lie intime- and cost-saving by the omission of the sample prepara-tion step. Taking into account the progressive increase of thesensitivity of recent mass spectrometers, even lower limits ofdetection are expected to be achievable in the near future usinga dilute-and-shoot LC-MS strategy.

AcknowledgmentsThe authors are grateful to Richard Caldwell for providingthem with excretion samples, Ken De Fauw, a recent graduateof Gent University, for preliminary research carried out in theframework of his Master thesis.

References

1. Fosgerau, K., Hoffmann, T.: Peptide therapeutics: current status andfuture directions. Drug Discov. Today 20, 122–8 (2015)

2. WADALaboratory Expert Group: minimum required performance levelsfor detection and identification of non-threshold substances. Available at:https://www.wada-ama.org/en/resources/science-medicine/td2015-mrpl.

3. Pollaro, L., Heinis, C.: Strategies to prolong the plasma residence time ofpeptide drugs. Med. Chem. Commun. 1, 319 (2010)

4. Argente, J., García-Segura, L.M., Pozo, J., Chowen, J.A.: Growthhormone-releasing peptides: clinical and basic aspects. Horm. Res. 46,155–9 (1996)

5. Kauffold, J., Rohrmann, H., Boehm, J., Wehrend, A.: Effects of long-term treatment with the GnrH agonist deslorelin (Suprelorin) on sexualfunction in boars. Theriogenology 74, 733–40 (2010)

6. Sanchis-Gomar, F., Martinez-Bello, V.E., Nascimento, A.L., Perez-Quilis, C., Garcia-Gimenez, J.L., Viña, J., Gomez-Cabrera, M.C.:Desmopresssin and hemodilution: implications in doping. Int. J. SportsMed. 31, 5–9 (2010)

7. Ho, E.N.M., Kwok,W.H., Lau,M.Y.,Wong, A.S.Y.,Wan, T.S.M., Lam,K.K.H., Schiff, P.J., Stewart, B.D.: Doping control analysis of TB-500, asynthetic version of an active region of thymosin β4, in equine urine and

plasma by liquid chromatography-mass spectrometry. J. Chromatogr. A1265, 57–69 (2012)

8. Esposito, S., Deventer, K., Goeman, J., Van der Eycken, J., Van Eenoo,P.: Synthesis and characterization of the N-terminal acetylated 17-23fragment of thymosin beta 4 identified in TB-500, a product suspectedto possess doping potential. Drug Test. Anal. 4, 733–738 (2012)

9. Tuli, L., Ressom, W.H.: LC-MS based detection of differential proteinexpression. J. Proteom. Bioinform. 2, 416–438 (2009)

10. Thevis, M., Schänzer, W.: Current role of LC-MS(/MS) in doping con-trol. Anal. Bioanal. Chem. 388, 1351–1358 (2007)

11. Sigdel, T.K., Nicora, C.D., Hsieh, S.-C., Dai, H., Qian, W.-J., Camp,D.G., Sarwal, M.M.: Optimization for peptide sample preparation forurine peptidomics. Clin. Proteom. 11, 7 (2014)

12. Semenistaya, E., Zvereva, I., Thomas, A., Thevis, M., Krotov, G.,Rodchenkov, G.: Determination of growth hormone releasing peptidesmetabolites in human urine after nasal administration of GHRP-1, GHRP-2, GHRP-6, Hexarelin, and Ipamorelin. Drug Test. Anal. 7, 919–925(2015)

13. Thomas, A., Delahaut, P., Krug, O., Schänzer, W., Thevis, M.: Metabo-lism of growth hormone releasing peptides. Anal. Chem. 84, 10252–10259 (2012)

14. Thomas, A., Görgens, C., Guddat, S., Thieme, D., Dellanna, F., Schänzer,W., Thevis, M.: Simplifying and expanding the screening for peptides2 kDa by direct urine injection, liquid chromatography, and ion mobilitymass spectrometry. J. Sep. Sci. 39, 333–341 (2016)

15. Esposito, S., Deventer, K., T’Sjoen, G., Vantilborgh, A., Van Eenoo, P.:Doping control analysis of desmopressin in human urine by LC-ESI-MS/MS after urine delipidation. Biomed. Chromatogr. 27, 240–245 (2013)

16. Mazzarino, M., Calvaresi, V., de la Torre, X., Parrotta, G., Sebastianelli,C., Botrè, F.: Development and validation of a liquid chromatography–mass spectrometry procedure after solid-phase extraction for detection of19 doping peptides in human urine. Forensic. Toxicol. 33, 321–337(2015)

17. Timms, M., Hall, N., Levina, V., Vine, J., Steel, R.: A high-throughputLC-MS/MS screen for GHRP in equine and human urine, featuringpeptide derivatization for improved chromatography. Drug Test. Anal.6, 985–995 (2014)

18. Steel, R., Timms, M., Levina, V., Vine, J.: A high throughput screen for17 Dermorphin peptides in equine and human urine and equine plasma.Drug Test. Anal. 6, 909–921 (2014)

19. Hørning, O.B., Theodorsen, S., Vorm, O., Jensen, O.N.: Solid phaseextraction-liquid chromatography (SPE-LC) interface for automated pep-tide separation and identification by tandem mass spectrometry. Int. J.Mass Spectrom. 268, 147–157 (2007)

20. Deventer, K., Pozo, O.J., Verstraete, A.G., Van Eenoo, P.: Dilute-and-shoot-liquid chromatography-mass spectrometry for urine analysis indoping control and analytical toxicology. TrAC Trends Anal. Chem.55, 1–13 (2014)

21. Hahne, H., Pachl, F., Ruprecht, B., Maier, S.K., Klaeger, S., Helm, D.,Médard, G., Wilm, M., Lemeer, S., Kuster, B.: DMSO enhanceselectrospray response, boosting sensitivity of proteomic experiments.Nat. Methods 10, 989–991 (2013)

22. Magnusson, B., Örnemark, U.: The fitness for purpose of analyticalmethods: a laboratory guide to method validation and related topics.Available at: https://www.eurachem.org/index.php/publications/guides/mv. Accessed 02 Nov 2016

23. Banerjee, S., Mazumdar, S.: Electrospray ionization mass spectrometry: atechnique to access the information beyond the molecular weight of theanalyte. Int. J. Anal. Chem. 2012, 1–40 (2012)

24. Tjernberg, A.: DMSO-related effects in protein characterization. J.Biomol. Screen. 11, 131–137 (2005)

25. Kruve, A.: Influence of mobile phase, source parameters, and source typeon electrospray ionization efficiency in negative ion mode. J. MassSpectrom. 51, 596–601 (2016)

26. Meyer, J.G., Komives, E.A.: Charge state coalescence during electrosprayionization improves peptide identification by tandem mass spectrometry.J. Am. Soc. Mass Spectrom. 23, 1390–1399 (2012)

27. Strzelecka, D., Holman, S.W., Eyers, C.E.: Evaluation of dimethyl sulf-oxide (DMSO) as a mobile phase additive during top 3 label-free quan-titative proteomics. Int. J. Mass Spectrom. 391, 157–160 (2015)

28. Goebel-Stengel, M., Stengel, A., Taché, Y., Reeve, J.R.: The importanceof using the optimal plasticware and glassware in studies involvingpeptides. Anal. Biochem. 414, 38–46 (2011)

1664 P. Judák et al.: Detection of Small Peptides by Ds-Lc-Hrms

29. National Institute for Biological Standards and Control: Peptide handling,dissolution and storage. Available at: http://www.nibsc.org/science_and_research/virology/cjd_resource_centre/available_samples/peptide_library/peptide_storage.aspx. Accessed 23 Nov 2016

30. Stejskal, K., Potěšil, D., Zdráhal, Z.: Suppression of peptide sample lossesin autosampler vials. J. Proteome Res. 12, 3057–3062 (2013)

31. World Anti-Doping Agency: International standard for laboratories.Available at: https://www.wada-ama.org/sites/default/files/resources/files/isl_june_2016.pdf. Accessed 02 Nov 2016

32. Panuwet, P., Hunter, R.E. Jr., Souza, P.E.D., Chen, X., Radford, A.,Cohen, J.R., Marder, M.E., Kartavenka, K., Barry, P., Barr, D.B.: HHSPublic Access. 46, 93–105 (2016)

P. Judák et al.: Detection of Small Peptides by Ds-Lc-Hrms 1665