Advantageous use of methanesulfonic acid in water-rich modifiers for peptide analysis

Gioacchino LucaLosaccoâ^,^b,JimmyOliviera DaSilva^c, Jinchu Liu^c,Erik L. Regalado^c, Jean-Luc Veutheyâ^,^b, Davy Guillarmeâ^,^b^,^∗

aSchool of Pharmaceutical Sciences, University of Geneva, CMU – Rue Michel-Servet 1, 1211 Geneva 4, Switzerland

bInstitute of Pharmaceutical Sciences of Western Switzerland, University of Geneva, CMU – Rue Michel-Servet 1, 1211 Geneva 4, Switzerland

cAnalytical Research and Development, MRL, Merck & Co, Inc., 126 E. Lincoln Ave, Rahway, NJ 07065, United States

a r t i cl e i n f o

Article history:

Received 25 January 2021 Revised 1 March 2021 Accepted 2 March 2021 Available online 9 March 2021 Keywords:

Ultra-high performance supercritical fluid chromatography

Theaimofthiswork wastoexpandtheapplicabilityrangeofUHPSFCtoseriesofsyntheticand com-mercializedpeptides.Initially,ascreeningofdifferentcolumnchemistriesavailableforUHPSFCanalysis wasperformed,in combinationwithadditivesof eitherbasicor acidicnature.Thecombinationofan acidicadditive(13mMTFA)withabasicstationaryphase(TorusDEAand2-PIC)wasfoundtobethe bestforaseriesofsixsyntheticpeptidespossessingeitheracidic,neutralorbasicisoelectricpoints. Sec-ondly,methanesulfonicacid(MSA)wasevaluatedasapotentialreplacementforTFA.Duetoitsstronger acidity,MSAgavebetterperformancethanTFAatthesameconcentrationlevel.Furthermore,theuseof reducedpercentagesofMSA, suchas8mM,yieldedsimilarresultstothoseobservedwith15 mMof MSA. TheoptimizedUHPSFCmethod was,then,usedtocomparetheperformanceof UHPSFCagainst RP-UHPLCforpeptideswith differentpIandwithincreasingpeptide chainlength.UHPSFCwasfound togiveaslightlybetterseparationofthepeptidesaccordingtotheirpIvalues,infewcasesorthogonal tothatobservedinUHPLC.Ontheotherhand,UHPSFCproducedamuchbetterseparationofpeptides withanincreasedaminoacidicchaincomparedtoUHPLC.Subsequently,UHPSFC-MSwassystematically comparedtoUHPLC-MSusingasetoflinearandcyclicpeptidescommerciallyavailable.Theoptimized UHPSFC methodwasabletogenerateat leastsimilar, andinsomecasesevenbetterperformanceto UHPLCwiththeadvantageofprovidingcomplementaryinformationtothatgivenbyUHPLCanalysis. Fi-nally,theanalyticalUHPSFCmethodwastransferredtoasemipreparativescaleusingaproprietarycyclic peptide,demonstratingexcellentpurityandhighyieldinlessthan15min.

ThisisanopenaccessarticleundertheCCBY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/)

1.Introduction

Peptides and peptide-like drugs are compounds which typi-callygeneratedalot ofinterestwithinthepharmaceutical indus-try.Theirpresenceinseveralkeybiologicalprocessesmakesthem an interesting class of molecules from which new drugs could bedeveloped[1,2]. Therehavebeenseveral developmentsin the use of peptides as therapeutic agents:originally,they were sim-plyusedinreplacementtherapies,whenpatientslackedaspecific peptidein theirorganism[3,4].Aclassicexampleof thisstrategy

∗Corresponding author at: School of Pharmaceutical Sciences, University of Geneva, CMU – Rue Michel-Servet 1, 1211 Geneva 4, Switzerland.

E-mail address: Davy.guillarme@unige.ch (D. Guillarme).

is theadministrationof insulin topatients suffering fromtype1 diabetes [3].Subsequently, syntheticanalogsof differentpeptides already present in the human body came along [5,6]. However, peptides present several issues as drugs,mainly related to their pharmacokineticproperties[7,8],becauseoftheirlow bioavailabil-ity due to their size, up to 5000 – 6000 Da for peptides with an amino acidic sequence of 40–50 amino acids, as well as an facile metabolism [9]. To improve their properties, modern syn-thetic peptides have started to differ, from a structural point of view, from their biological precursors, including new functional groupsintheirstructure(i.e.polymersandfattyacids)introduced todevelopabetterbioavailabilityviatheiroralformulation[10,11].

The analytical strategy to characterizethis class of molecules has revolved on the use of ultra-high performance liquid

https://doi.org/10.1016/j.chroma.2021.462048

G.L. Losacco, J.O. DaSilva, J. Liu et al. Journal of Chromatography A 1642 (2021) 462048

chromatography (UHPLC) as the preferred technique, mainly in reversed-phasemode(RPLC)[12–14].Itseaseofuse,high through-put capacity andability to couple with ultraviolet (UV) detector and,moreimportantly, massspectrometry (MS)madeita popu-lartechniqueforpeptide analysis[15–17].Despitethe advantages of UHPLC-UV-MS, a demand for greener, faster and complemen-taryanalyticaltechniquesisalwayspresent[18].Amongthem,one ofthemostpromisingstrategiesisultra-highperformance super-criticalfluidchromatography(UHPSFC).Thankstothedevelopment ofdedicatedsub-2µmstationaryphases,aswellasthereleaseof chromatographicsystemsabletowithstandthebackpressures gen-eratedbythesecolumns,UHPSFChas showna greatpotentialas acomplementary alternativeto UHPLC.This waspossible thanks to the use of a mobile phase consisting in a mixture of super-criticalcarbondioxidewithpolarorganicmodifier[19].Moreover, itpresents an easy hyphenationto variousdetectors such as UV and MS[20] and can provide fast analyses as the mobile phase presentslowviscosity,enablinghigherflow-rateswithout experi-encinghighbackpressures.Finally,ahighdegreeof orthogonality existsbetweenUHPSFCandUHPLC,especiallywiththeRPLCmode [21].

The analysis of peptides in UHPSFCis describedin the litera-ture,andthere havealready beenstudiesdemonstrating the use ofUHPSFCfortheiranalysis[22–25].However,asystematic com-parisonbetweenUHPSFCandUHPLChasnotbeenmadeuntilnow.

ThisisprobablybecauseUHPSFC isdifficulttouseforthe analy-sisofhighlypolarcompoundshavinghighmolecularweight (par-tialelutionfromthecolumn,solubilityissues,distortedpeaks…).

Nonetheless,inthelast2–3yearsanewtrendappearsinUHPSFC, consistingin the useof gradient profilesreachingpercentagesof organic modifierup to 90–100%[26–28]. Furthermore, the addi-tion of water, upto 5–7% in the organic co-solvent has enabled UHPSFCtogiveimprovedperformanceintheanalysisofpolarand ionizedmetabolites,asitincreasestheelutionstrengthofthe mo-bilephase[28,29].Thesenewtrends inUHPSFC could,therefore, reinvigorateitsapplicabilityfortheanalysisofpeptides.

Theaimof thisstudy wastoevaluatetheperformanceof UH-PSFC,coupledtodifferentdetectors(UVandMS),fortheanalysis of a seriesof synthetic and therapeuticpeptides. Different chro-matographicaspects,suchasretention,selectivityandpeakshape, aswellascompatibilitywithMSdetectionand,finally,scale-upto thepreparativescale,havebeeninvestigated. The impactof pep-tideisoelectricpoint,hydrophobicityandaminoacidschainlength,

ontheUHPSFCseparationwasassessed.Asystematiccomparison toUHPLCin theRPLCmodewasalsoperformedwiththegoalof highlighting possible advantages and disadvantages of the newly developedmethod.

2. Materialsandmethods

2.1. Chemicals,reagentsandsamplepreparationprocedures

For all experiments performed at the University of Geneva, methanol(MeOH) andacetonitrile(ACN) ofOPTIMALC-MS grade and water (H2O) of UHPLC grade were purchased from Fischer Scientific (Loughborough,UK). Carbondioxide (CO2) of 4.5 grade (99.995%puritylevel)waspurchasedfromPanGas(Dagmerstellen, Switzerland). Metanilyellow and methyl orange,lysine, arginine, asparticacid,glutamicacid,ammoniasolutionat25%ofMSgrade, trifluoroacetic acid (TFA) of MS grade and methanesulfonic acid (MSA) at a puritylevel of 99.5% or higher were purchased from Sigma-Aldrich(Buchs,Switzerland).Syntheticpeptides1N,2N,1B, 2B, 1A,2A,6mer,9mer,12mer,15mer,18merand21merat a pu-rity level of ≥95%have been purchasedfrom GenScriptBiotech (Leiden, Netherlands). More information regarding their amino acidsequences,molecularweightsaswellaspredictedisoelectric points (pI)and GRAVY numbers areprovided in Table 1. GRAVY number is a measure of the grade of hydrophilicity of a pro-tein/peptide basedonits hydropathyindex, avalue whichvaries between −2 to 2 for most proteins; the higher the hydropathy index, the higher the hydrophobicity.GRAVY numberand pI val-ues wereobtainedusingthe ProtParamtoolavailableonthe pro-teomic serverExPASy[30,31].Commercial pharmaceutical formu-lations of liraglutide, leuprorelin, glucagon, cyclosporin A, eptifi-batideandlinaclotide(Table1)havebeenpurchasedfromthe hos-pital pharmacyat theGenevaUniversity Hospitals (HUG,Geneva, Switzerland).

For all purification experiments, methanol (HPLC Grade) and water (HPLC grade) were purchased from Fisher Scientific (Fair Lawn, NJ,USA). Methanesulfonicacid(MSA), 99%extra pure was purchasedfromARCOS¯ Organics(MorrisPlains,NJ,USA).Thecyclic peptide wasobtainedin-house(Merck&Co.,Inc., Kenilworth,NJ, USA).Bonedry-gradeCO2wasobtainedfromAirGas(New Hamp-shire,USA).

Detailsregardingthesamplepreparationandstressprocedures usedinthisstudycanbefoundinthesupplementarymaterial.

Table 1

List of synthetic and commercial peptides used in this study.

Name Sequence

Peptide 1N Trp-Asn-Ser-Val-Lys-Tyr-Asp-Ile-Ser-Tyr-His-Thr 1512 12 6.74 −0.93

Peptide 2N Ala-Tyr-His-Asp-Gln-Trp-Lys-Tyr-His-Phe-Cys 1497 11 6.95 −1.24

Peptide 1B Trp-Gln-Ser-Thr-Tyr-His-Asp-Lys-Phe-Ala-Trp-Arg-Tyr 1788 13 8.50 −1.53

Peptide 2B Phe-Lys-Asn-Ser-Tyr-His-Gln-Ile-Arg-Trp-Val-Tyr-Asn-Phe 1902 14 9.70 −0.86

Peptide 1A Phe-Asn-Glu-Cys-Tyr-Arg-Ser-Asp-Ala-Tyr-Ser-Asn-Thr-Phe 1717 14 4.37 −0.96

Peptide 2A Tyr-Asn-Ser-Phe-Asp-Glu-Trp-Lys-Cys-Thr-Phe-Ser-Trp 1713 13 4.37 −0.90

Peptide 6mer Leu-Trp-His-Gly-Ser-Asn 713 6 6.74 −0.83

Peptide 9mer Leu-Trp-His-Gly-Ser-Asn-Lys-Trp-Asp 1142 9 6.74 −1.48

Peptide 12mer Leu-Trp-His-Gly-Ser-Asn-Lys-Trp-Asp-Asn-Gly-Gln 1441 12 6.74 −1.73

Peptide 15mer Leu-Trp-His-Gly-Ser-Asn-Lys-Trp-Asp-Asn-Gly-Gln-Trp-Ser-Asn 1829 15 6.74 −1.73

Peptide 18mer Leu-Trp-His-Gly-Ser-Asn-Lys-Trp-Asp-Asn-Gly-Gln-Trp-Ser-Asn-Gly-Thr-Gln 2115 18 6.74 −1.69 Peptide 21mer Leu-Trp-His-Gly-Ser-Asn-Lys-Trp-Asp-Asn-Gly-Gln-Trp-Ser-Asn-Gly-Thr-Gln-Ala-Asn-Ser 2387 21 6.74 −1.57 Liraglutide His-Ala-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Lys( γ

-Glu-palmitoyl)-Glu-Phe-Ile-Ala-Trp-Leu-Val-Arg-Gly-Arg-Gly

3751 32 4.96 −0.36

Leuprorelin pGlu-His-Trp-Ser-Tyr- d -Leu-Leu-Arg-Pro-NHEt 1209 9 8.75 −0.52

Glucagon His-Ser-Gln-Gly-Thr-Phe-Thr-Ser-Asp-Tyr-Ser-Lys-Tyr-Leu-Asp-Ser-Arg-Arg-Ala-Gln-Asp-Phe- Val-Gln-Trp-Leu-Met-Asn-Thr

3483 29 6.75 −0.99

Cyclosporin A Abu-Sar-Leu-Val-Leu-Ala-dAla-Leu-Leu-Val-Bmt 1203 11 NA NA

Eptifibatide Cys-hArg-Gly-Asp-Trp-Pro-Cys 832 7 3.80 −0.20

Linaclotide Cys-Cys-Glu-Tyr-Cys-Cys-Asn-Pro-Ala-Cys-Thr-Gly-Cys-Tyr 1527 14 4.00 0.32

G.L. Losacco, J.O. DaSilva, J. Liu et al. Journal of Chromatography A 1642 (2021) 462048 2.2. ChromatographicandMSinstrumentationsandconditions

At University of Geneva, for UHPSFC analyses, five different columnshavebeen initiallyemployed,namely Torus2-PIC, Torus DEA,TorusDIOL,BEHsilica(Waters,Milford,MA,USA),allpacked with 1.7 µm fully porous silica particles, and Nucleoshell HILIC (Macherey-Nagel, Düren, Germany), packedwith 2.7 µm superfi-ciallyporoussilicaparticles.All columnspossessthesame geom-etry of 100×3.0 mm I.D. Agenericgradient wasdeveloped for theanalysisofsyntheticpeptides,from10to100%organic modi-fierinthemobilephaseover7min,followedbyanisocratichold at100%ofco-solventfor1min,thenareturntoinitialconditions in0.1min,andafinalisocraticstepwith10%oforganicmodifier for 2.9min, giving a total run time of 11 min (section 3.1.). Or-ganicmodifieremployedatthisstagewasamixtureofMeOH/H₂O 95:5v/vcontainingeither13mM(0.1%)TFA,15mM(0.1%)MSAor 52mM (0.2%)NH₄OH.Flow-rate wasfixedat 0.7 mL.min⁻¹. Fol-lowing thispreliminary step,an optimized method for the anal-ysisof syntheticpeptides wasdevelopedandusedin the second partofthestudy(section3.2),basedontheTorus2-PICstationary phasewith mixture of MeOH/H2O 95:5v/v + 8mM MSA as the co-solvent.Theoptimizedmethodfollowsadifferentgradient pro-file,startingfrom30to80% organicmodifierover5min,thenan isocratic stepat 80% of co-solventfor 0.5min,followed bya re-turntoinitialconditionsin0.1minandasecondisocraticstepof 1.9minforatotalanalysistimeof7.5min.Flow-ratewasfixed,in thiscaseat0.9mL.min⁻¹.Forthecommerciallyavailablepeptides (i.e.liraglutide, leuprorelin,glucagon, linaclotideandeptifibatide), amodifiedversionoftheoptimizedgradientwasemployed:start at35%co-solvent,reaching90%in5min,thenanisocraticstepat 90%ofco-solventfor0.5min,followedbyareturntoinitial condi-tionsin0.1minandasecondisocraticstepundertheseconditions of1.9min foratotal analysistime of7.5min.For cyclosporinA, athird gradient was chosen,starting from2to 40% over 5min, withan isocratic step at 40% of co-solvent for0.5 min, then re-turntoinitialconditionsin0.1minandasecondisocraticstepfor 1.9min,givingatotalruntimeof7.5min.

UnderUHPLCconditions,a 50×2.1mm I.D.BEH C18 station-aryphasepackedwith1.7µmfullyporousparticles(Waters)was used.MobilephaseAwasH₂O+13mMTFA,whilemobilephase Bwas ACN+13 mM TFA. Anoptimizedgradient was employed forallsyntheticandtherapeuticpeptides(withtheonlyexception ofcyclosporinA),consistingina 5min gradientfrom5to65%B, aholdupfortwominutes at65%B,then areturntoinitial con-ditionsin 0.1 min andan isocratic holdfor 1.9 min at 5% for a totalruntime of 9min.ForcyclosporinAthe gradienttime and totalruntimewere thesame,howeverthehighestpercentage of Breachedduringthegradientwas95%.Inalltheseconditions,the flow-ratewasfixedat0.4mL.min⁻¹.

The column screening consisted of eight different stationary phases, namely Chiralpak IC and Chiralcel OZ (both of geome-try of 100 × 4.6 mm I.D. – 3.0µm fullyporous particles); Chi-ralcel OJ, Chiralpak IG and DCpak P4VP (all with geometry of 150×4.6mmI.D.– fullyporousparticlesizesof3.0µmfor Chi-ralcelOJandof 5.0µm ChiralpakIGandDCpak P4VP)from Chi-ralTechnologies(West Chester,PA,USA);CELERIS 4EPfromRegis Technologies(MortonGrove,IL,USA)andTorusDEAandTorus 2-PICfromWaters Corp.(Milford, MA,USA),all withthe geometry of250×4.6mmI.D.andpackedwith5.0µmfullyporous parti-cles.SFCscreeningswerecarriedoutonthediversesetofcolumns describedintheabovesectionbygradientelutionataflowrateof 2mL.min⁻¹ withthe backpressureregulator(BPR)set at103 bar (1500psi).TheSFCeluentsconsistedofCO₂andorganicmodifier, whichconsistedofMeOH/H₂O95/5v/v+8mMMSA.Themobile phaseswereprogrammedasfollows:35%Bat0min,linear gradi-entfrom35%to90%Bin5min,aholdat90%Bfor0.5min,then

return to35%Bin 0.1min andfinally holdat35% Bfor1.9min.

The PDAscansfrom190to400 nmandthechromatogramis ex-tractedat 210nm. The MSscansthe massrangeof 100to1200 withasamplingfrequencyof 2Hz, conevoltagesof10and50V in ESI (+) and acone voltage of 10Vin ESI (-).Preparative SFC purification was performed on a Waters Torus 2-PIC 30.0 mm x 250 mm, 5 µm column with a mobile phase of 35% MeOH/H₂O 95/5 v/v+ 8mM MSA / CO₂. The flow ratewas 140 mL.min⁻¹, mobile phaseandcolumn oven temperatureat 35 °C,back pres-sureregulatorsetto103 bar(1500psi),UVdetectionat 210nm.

Thesamplewaspreparedat20mg/mLinmethanolwithaloadof 1mL.

SFCanalysisofthecyclic peptidewascarriedoutonaWaters Torus2-PIC4.6mmI.D.x250mm5µmcolumnataflowrateof 2mL.min⁻¹withthebackpressureregulator(BPR)setat100bar;

TheSFCeluentsolventwas40%MeOH/H₂O95/5v/v+8mMMSA /CO₂.The PDAscansfrom190to400 nmandthe chromatogram wasextractedat210nm.

Allinformation regardingthe chromatographicandMS instru-ments conditions, as wellas onthe software employed for data treatmentcanbefoundinthesupplementarymaterial.

3. Resultsanddiscussion

3.1. DevelopmentoftheUHPSFCchromatographicmethod

3.1.1. Impactoftheadditivenatureonthestationaryphase performance

To ensure the elution of peptides using a CO₂-based mobile phase, variousparameters have tobe considered.Firstly, the ad-dition of water in the co-solvent seems necessary to ensure ac-ceptable peak shapes as well as elution within reasonable time [25,32–34]. Secondly, additives areneeded to furtherreduce the tailingfactorandpeakwidths[23,24,35].To choosethe most ap-propriate stationaryphase,a screeningof the severalchemistries availablewasoftenneeded.Overall,analyticalconditionsfor pep-tideanalysisunderSFCcanbesummarized asfollows:amixture of methanoland waterasthe organicco-solvent, in combination withanadditive(inmostcasesTFA). However,the applicationof such conditionsismostlylimitedtotheanalysisof peptideswith relatively shortaminoacidicsequences (often10–12or less)[22–

24,32].Therefore,the goal of the presentstudy wastofind con-ditionssuitableforawiderrangeofpeptides,throughthe screen-ingofdifferentstationaryphasechemistries,incombinationwith the use of acidic and basic additives. Such a screening strategy was firstly appliedto aseries of synthetic peptides describedin Table1(peptides1N,2N,1B,2B,1Aand2A).Thesepeptidesall possessasequence ofalengthbetween11– 14aminoacidsand witha molecularweightranging from1500to1900 Da. Further-more, these differentpeptides possess either an acidic(pI < 7), neutral (pI≈7),or basicnature(pI>7)andtheyallpossessan important polarcharacter (GRAVY number between−1and −2).

Indeed, compoundswiththesepropertieshavealwaysbeen chal-lenging toanalyzeunderUHPSFCconditions, astheyarestrongly retainedonthe(polar)stationaryphase,andpoorlysolublein mo-bilephaseswithapredominantpresenceofsupercriticalCO2.Each stationary phase (i.e. Torus 2-PIC, Torus DEA, Nucleoshell HILIC, Torus DIOL, BEHsilica) was tested with the sameorganic modi-fier composition(MeOH/H₂O 95:5)in whicheither13mM (0.1%) of TFA or 52 mM (0.2%) of NH₄OH was added. InFig. 1a, a ta-ble summarizing the data ispresented.Stationary phases with a

“basic” nature (having oneor more positively charged functional groups)arethoseprovidingthebestresults,yieldingcomplete elu-tion ofallsyntheticpeptideswithgoodpeak shape,asillustrated inFig.1bforpeptides1Band2AontheTorus2-PIC.Betweenthe Torus 2-PIC and Torus DEA, nomajordifferences were observed,

G.L. Losacco, J.O. DaSilva, J. Liu et al. Journal of Chromatography A 1642 (2021) 462048

Fig. 1. a) A classification of the combination between the nature of the additive and the properties of the stationary phase chemistries evaluated in this study, on the quality of the chromatographic separation and elution for a series of synthetic peptides. b) Chromatograms, for peptides 1B and 2A, obtained on a “neutral”, “zwitterionic” and

“basic” stationary phase using the best combination between stationary phase and nature of the additive in the mobile phase.

but the Torus 2-PIC gave a slightly faster elution. As expected, thesecolumnsgavegoodresultsonlywhenanacidicadditivewas used. The addition of 52 mM NH₄OH in the mobile phase pro-videdasevereloss of performanceonthetwo“basic” stationary phases(i.e.Torus2-PICandTorusDEA)(Fig.S1).Thecombination of abare silica (BEH silica) stationaryphasewith acidicadditive such as TFA, or even basic additives (52 mM NH₄OH) provided inferiorperformance to those witnessedon the Torus 2-PIC/DEA columns(Fig.1a-b).Withacidicpeptides(peptide2A),theBEH sil-icagavecomparablepeakshapestothatobservedontheTorus 2-PIC(Fig.1b),butdidnotforpeptideswithhigherisoelectricpoints (peptide 1B – Fig. 1b). Finally, the two remaining columns em-ployedinthis study,namely the TorusDIOL(neutral)and Nucle-oshellHILIC (zwitterionic), were thoseoffering the worst perfor-manceoverall.Morespecifically,theuseofazwitterionic station-aryphaseperformedratherpoorlywith13mMTFA,whilethe ad-dition52mMNH4OHensuredtheproperelutionofpeptides,but withextremelypoorpeakshapesasshownforpeptides1Band2A (Fig. 1b).Inconclusion,the combination of a columnhaving ba-sicproperties (Torus2-PIC) withanacidic additive(13mM TFA)

Dans le document Evolution of sub/supercritical fluid chromatography – mass spectrometry for the analysis of highly polar compounds and biological matrices (Page 117-130)