Influence of Solvents upon Diketopiperazine Formation of FPG8KZhi-chao Zhang Shannon A Raab David A Hales and David E Clemmer

Cite This httpsdxdoiorg101021acsjpcb1c00269 Read Online

ACCESS Metrics amp More Article Recommendations sı Supporting Information

ABSTRACT Ion mobility spectrometry (IMS) and massspectrometry (MS) techniques were used to monitor diketopiper-azine (DKP) formation from the peptide FPG8K at multipledefined temperatures in methanol ethanol propanol and waterwith the motivation to study the effect of solvent polarity onspontaneous solution dissociation The reaction rate increases withdecreasing solvent polarity The observed rates of trans rarr cisisomerization of Phe1minusPro2 and the cis-Pro2 isomer dissociationresult in the cis isomer growing in abundance relative to the transisomer throughout the reaction in all solvents Analysis of rateconstants derived from the data using a sequential unimolecularkinetics model that includes hidden intermediate states yields transition state thermodynamic values for both trans rarr cisisomerization of Phe1minusPro2 and dissociation The measured thermochemistry appears to be closely correlated with these solventsrsquodielectric constants a lower solvent dielectric constant accelerates the reaction by reducing the enthalpic barrier albeit with slightentropic restriction

INTRODUCTION

Proteins and peptides can decompose during synthesis andstorage through several spontaneous chemical reactions12 Oneof these reactions is diketopiperazine (25-dioxopiperazineDKP) formation which involves an N-terminal aminenucleophilic attack on the carbonyl carbon between thesecond and third amino acid residues breaking the peptidechain and forming a cyclic peptide composed of the first tworesidues from the N-terminus and a truncated sequence of thepeptide3minus6 DKP formation happens especially when there is aproline residue at the second position from the N-terminus(ie penultimate proline) There are a few reports on thisspontaneous degradation including the dissociation ofrecombinant DNA-derived human growth hormone (rhGH)resulting in DKP and a truncated variant of rhGH7

degradation of the eleven-residue neuropeptide substance Pin solution and solid forms8 and a recent report about self-cleavage of bradykinin at elevated temperature which involvesbreaking the bond that is the most difficult to cleaveenzymatically9

Understanding the DKP formation reaction is of importancefor other reasons as well For example it is a way of generatingbiologically active species throughout the body such ascyclo(HisminusPro)10 Also preventing intramolecular aminolysisis important for the storage of pharmaceutically importantproteins and peptides since there are reports of decompositionof such compounds1112 In addition a recent finding by theHunt group shows that a variety of MHC class I-associatedphosphorylated peptides cause immune response13 Thesepeptides have the potential to be used as novel vaccines for

immunotherapy However many of these peptides contain apenultimate proline which can lead to DKP formation andcompromise the function of these peptidesSeveral reports show that acids and bases as well as buffer

species can influence the DKP formation rate in aqueoussolution61214 However few reports analyzed the solventeffects on the DKP formation rate1516 Also the exactrelationship between the rate of trans rarr cis isomerization ofXaa1minusPro2 and DKP formation has not been established Thisis largely because almost all the previous research utilizedtechniques that were not able to separate the cis and transisomers The ion mobility spectrometry (IMS)minusmass spec-trometry (MS) technique has proven to be a useful tool toseparate and analyze different conformations based ondifferences in moleculesrsquo overall mobility due to differentcollision cross sections17 18 Recent work used IMSminusMS toanalyze the dissociation kinetics of bradykinin and substanceP919 in which different isomers due to proline trans rarr cisisomerization of the Xaa1minusPro2 bond are monitored during thedegradation process Analyzing the isomers from each speciesduring dissociation provided evidence that at least three andfour intermediates are involved in spontaneous solutiondissociation of bradykinin and substance P respectively In

Received January 11 2021Revised February 26 2021

ArticlepubsacsorgJPCB

copy XXXX American Chemical SocietyA

httpsdxdoiorg101021acsjpcb1c00269J Phys Chem B XXXX XXX XXXminusXXX

Dow

nloa

ded

via

IND

IAN

A U

NIV

BL

OO

MIN

GT

ON

on

Mar

ch 1

9 2

021

at 1

238

18

(UT

C)

See

http

sp

ubs

acs

org

shar

ingg

uide

lines

for

opt

ions

on

how

to le

gitim

atel

y sh

are

publ

ishe

d ar

ticle

s

each case a configurationally coupled protonation process isobserved as the penultimate proline that undergoes trans rarr cisisomerization similarly to a previous Pro7 study20 Here thepeptide FPG8K is used as a model to analyze DKP formationkinetics in four solvents at multiple defined temperaturesFPG8K was selected because the conformations resulting frompenultimate proline trans rarr cis isomerization have sufficientlydifferent collision cross sections to be separable by IMS thusenabling easy monitoring of changes in the cistrans ratio ofproline during degradation Transition state thermochemistryvalues are also calculated for isomerization in all four solventsand for dissociation in three of the four solvents Thisthermochemistry appears to be closely correlated with thesesolventsrsquo dielectric constants The implications of this arediscussed below

METHODS

Peptide Synthesis Peptides are synthesized according toa traditional Fmoc solid-phase peptide synthesis procedure asdetailed elsewhere21 Briefly 001 mmol Fmoc-Lys(Boc)-Wangresin is deprotected using piperidineDMF (20 by volume)Then the second amino acid from the C-terminus is activatedusing 3-(diethoxyphosphoryloxy)-123-benzotriazin-4(3H)-one (DEPBT) and NN-diisopropylethylamine (DIEA) beforemixing with the resin in 10 mL of DMF for sim15 h Aftermixing the resulting product is washed with DMF DCM andMeOH three times each This procedure is repeated until thepeptide reaches the desired length Cleavage from the resin isperformed with TFAH2Otriisopropylsilane (952525 VVV) The peptide is precipitated with diethyl etherand purified with HPLCSample Preparation and Kinetics Experiment A

detailed description can be found in a previous paper9 Brieflythe peptide is dissolved into pure propanol ethanol methanolor water to make 1 mM stock solutions which are stored atminus20 degC Stock solutions are diluted to 20 μM with 1 aceticacid (by volume) for use The resulting alcohol solutions werethen incubated at 65 70 and 75 degC and tested periodically byIMSminusMS In water we were able to monitor isomerization at75 and 90 degC but the dissociation process was too slow to beobservedIMS Instrumentation A home-built instrument in which a

2 m drift tube is coupled with a time-of-flight (TOF) massspectrometer (see Supporting Information Figure S14) is usedto monitor the dissociation Detailed descriptions of theinstrument and theory appear elsewhere1822minus29 Briefly ionsare produced by nanoelectrospray in a NanoMate autosampler(Advion Ithica NY) and then transferred into the sourceregion Ions that are stored in the source funnel are periodicallypulsed into the drift tube for separation based on their overallstructural differences The drift tube is filled with sim30 torr(sim40 mbar) He buffer gas and operated under a weak electricfield of roughly 10 Vmiddotcmminus1 The ions are analyzed using a TOFmass spectrometer after migrating through the 2 m drift tubeIt is useful to convert the ion drift time tD into the collisionalcross section (CCS) (Ω in Aring2) according to the followingequation25

πΩ = +Auml

Ccedil

AringAringAringAringAringAringAringAringAringAring

Eacute

Ouml

NtildeNtildeNtildeNtildeNtildeNtildeNtildeNtildeNtildeNtildek T m mt E

L PT

N(18 )

16ze

( )1 1 760

2732112

b12

I B

12D

(1)

where the relevant terms are ion charge (ze) Boltzmannrsquosconstant (kb) mass of the ion (mI) mass of the buffer gas(mB) temperature of buffer gas (T) electric field value (E)drift tube length (L) pressure of buffer gas (P) and buffer gasneutral number density (N)

RESULTS AND DISCUSSIONMass Spectral Data for [FPG8K + 2H]2+ Dissociation in

Different Solvents Figure 1 shows the mass spectra acquired

after incubating FPG8K in n-propanol ethanol and methanol(with 1 acetic acid) at 75 degC In Figure 1a peptide FPG8K isincubated in n-propanol At 0 min there is one peak at mz =4257 in the mass spectrum corresponding to [FPG8K +2H]2+ At 21 min a new peak appears at mz = 3019representing [G8K + 2H]2+ As time progresses the [FPG8K +2H]2+ peak decreases in abundance while the [G8K + 2H]2+

peak grows in abundance The degradation process is completeat around 125 min when only [G8K + 2H]2+ remains Thedissociation half-life is around 48 min Figure 1bc showspeptide FPG8K dissociation mass spectra in ethanol andmethanol respectively Similar dissociation patterns areobserved as incubation time increases the initial [FPG8K +2H]2+ ion disappears yielding the [G8K + 2H]2+ ion The

Figure 1 Peptide FPG8K dissociation mass spectra in n-propanol (a)ethanol (b) and methanol (c) with 1 acetic acid (by volume) at 75degC Self-cleavage of the Pro2minusGly3 bond happens during theincubation The dissociation half-lives are 48 60 and 291 min in n-propanol ethanol and methanol respectively The dissociation isassumed to follow the DKP formation mechanism as illustrated inScheme 1

Scheme 1 Cleavage Mechanism of Pro2minusGly3 by DKPFormation with Phe1minusPro2 in the Cis configurationa

aThe cis and trans isomers mentioned in this paper are b and arespectively when R = Phe

The Journal of Physical Chemistry B pubsacsorgJPCB Article

httpsdxdoiorg101021acsjpcb1c00269J Phys Chem B XXXX XXX XXXminusXXX

B

dissociation half-lives in ethanol and methanol are 60 and 291min respectively Peptide FPG8K is also incubated in water Ittakes 50 h for the dissociation to begin which is much longerthan in the other three solvents (see Supporting InformationFigure S1) It turns out that the FPG8K dissociation rateincreases with decreasing solvent polarityCollision Cross Section Distributions for [FPG8K +

2H]2+ The structural transition of FPG8K can be directlymonitored by examining IMS cross section distributions usingthe IMSminusMS instrument Pierson et al30 assigned prolineisomers in bradykinin based on the fact that an alanine-substituted residue can only exist in the trans configurationWe took advantage of this approach here Figure 2 shows the

CCS distribution of the peptide FPG8K and its alanine-substituted analogue FAG8K The

DTCCSHe values for the twoconformations are Ω = 197 plusmn 3 and 207 plusmn 2 Aring2 for the peptideFPG8K Only the trans isomer is present in FAG8K with Ω =206 plusmn 2 Aring2 after correction for differences in the intrinsic sizeparameters between proline and alanine as describedpreviously3132 Therefore we assign the peak at Ω = 207 plusmn2 Aring2 as the isomer with a trans penultimate proline while theisomer at Ω= 197 plusmn 3 Aring2 has a cis penultimate prolineFigure 3 shows the cross section distributions for FPG8K in

water methanol ethanol and propanol Both cis and transisomers are observed and the cis isomer is the majorcomponent The percentages of the cis isomer are 88 80 75and 60 in propanol ethanol methanol and waterrespectively Although all these solvents favor the cis isomeroverall the amount of trans increases with solvent polarityAn interesting question arises is there a relationship

between dissociation kinetics and the isomerization ratioFigure 4 (top) shows the normalized intensity plots of FPG8Kisomerization and dissociation in n-PrOH EtOH MeOH andH2O at 75 degC As dissociation progresses both the cis and transisomers of [FPG8K + 2H]2+ decrease in abundance while theintensity of the fragment ion (ie [G8K + 2H]2+) increases Aprevious study shows that peptides with a penultimate prolinein the trans configuration cannot undergo the DKP formation

process33 therefore the trans isomer of [FPG8K + 2H]2+ mustundergo trans rarr cis isomerization of the Phe1minusPro2 bondbefore dissociation The plot here shows that both theisomerization and fragmentation rates increase with decreasingpolarity of the solventFigure 4 (bottom) shows the percent of the cis isomer in the

total distribution of [FPG8K + 2H]2+ during the degradationprocess in different solvents at 75 degC The trans rarr cisisomerization of the Phe1minusPro2 bond continues even while thepopulation of cis molecules is dissociating The relative rates ofthe two steps result in the cis isomer growing in abundancerelative to the trans isomer throughout the process Unlikesome protein folding processes in which the trans rarr cisisomerization of proline is the rate-limiting step34minus36 isomer-ization and cis isomer dissociation of FPG8K have comparablerates with isomerization being somewhat faster thandissociation The cis isomer abundance of the plots in Figure4 increases with decreasing solvent polarity implying that theeffect of the solvent on the rates is stronger for isomerizationthan for fragmentation

Characterizing the Dynamics and Pathways of FPG8KDissociation In order to further understand the dynamicsbehind the dissociation we modeled the experimental kineticsfor several potential pathways and compared the sums ofsquares of the fitting residuals This is conducted in a mannersimilar to that used in previous studies of dissociation ofbradykinin and substance P919 Details of this type of fittinghave been discussed previously37 We begin by considering thesimplest pathway FPG8K with a trans penultimate prolineundergoes trans rarr cis isomerization while FPG8K with a cispenultimate proline (both from the initial cis conformerpopulation and the cis conformer converted from trans)dissociates to form cFP and G8K as shown by reactions 2 and3

[ ‐ + ] rarr [ ‐ + ]+ +trans FPG K 2H cis FPG K 2H82

82

(2)

[ ‐ + ] rarr [ + ] ++ +cis FPG K 2H G K 2H cFP82

82

(3)

In Figure 4 (top) the green lines show the best fits of thesetwo models with the experimental data for methanol These

Figure 2 CCS distribution for Pro2 rarr Ala of the [FPG8K + 2H]2+

ion Distributions are obtained in ethanol with 1 acetic acid (byvolume) from FPG8K and FAG8K Alanine-substituted distribution isshown on top of [FPG8K + 2H]2+ as a comparison Collision crosssection distribution for the alanine-substituted peptide is corrected forthe size difference between Ala and Pro

Figure 3 CCS distribution of the doubly charged peptide [FPG8K +2H]2+ ions in water propanol ethanol and methanol Distributionswere collected from IMSminusMS instruments by integrating the driftbins for the mz range of [FPG8K + 2H]2+ ions

The Journal of Physical Chemistry B pubsacsorgJPCB Article

httpsdxdoiorg101021acsjpcb1c00269J Phys Chem B XXXX XXX XXXminusXXX

C

models fail to represent the general trend of the data Wecompared 10 different model mechanisms for each process ineach solvent The sum of squares of the fitting residuals

(ΣRSS) are listed with each candidate mechanism in Tables S1(isomerization) and S2 (dissociation) for methanol Thesingle-step processes in 2 and 3 appear as model 1 in theirrespective tables The sequential unimolecular models withthree and four intermediates have the lowest ΣRSS and thusbest reproduce the abundance profiles of trans rarr cisisomerization of the Phe1minusPro2 bond and cis isomerdissociation respectively as shown by the blue curves inFigure 4 The data for these two processes in ethanol andpropanol are also best reproduced (ΣRSS at a minimum) bysequential unimolecular models although with differentnumbers of unseen intermediates (see Supporting InformationTables S3minusS6) These intermediates are referred as unseenintermediates because they are not detected by IMSminusMS butare derived from the best-fitting kinetics model937 FPG8Ktrans rarr cis isomerization of the Phe1minusPro2 bond and cis-isomer dissociation require two and three intermediates inEtOH while one and two intermediates yield the best fits tothe 1-PrOH data These fits for 75 degC are also shown in Figure4 with data and fits for 70 and 65 degC found in SupportingInformation Figures S3minusS7 The same type of figure and tablefor FPG8K isomerization in water is shown in Table S7 andFigures S8 and S9 respectivelyThe numbers of intermediates involved in dissociation in

ethanol are similar to our previous study of bradykinin andsubstance P dissociation in the same solvent in which in totalat least three and four intermediates are involved in BK(2H+)rarr BK(3minus9)(2H

+) + cRP(H+) and subP(3minus11)(2H+)rarr cKP(H+)

+ subP(5minus11) respectively In addition one and zerointermediates are derived from the dissociation curves forpenultimate proline trans rarr cis isomerization in bradykininand substance P which are close to the two intermediatesfound in this study for the same process in ethanol A previousPro13 study shows that the all-cis-configured right-handedhelical PPI conformer can undergo transrarr cis isomerization ofeach peptide bond into an all-trans-configured PPII structureduring which six distinct long-lived intermediates areobserved38 The proline chain in HisPro13 undergoes thesame process via a cooperative two-state transition with sim15unseen intermediates37 These results suggest that intermediateconformations are likely inherent in proline trans rarr cisisomerization

Transition State Thermodynamics After fitting the dataat various temperatures Arrhenius plots are generated forFPG8K trans rarr cis isomerization of the Phe1minusPro2 bond andcis-FPG8K dissociation in each solvent by plotting the naturallog of the rate constant against inverse temperature accordingto eq 4

Figure 4 Top Normalized abundances of the doubly charged peptide[FPG8K + 2H]2+ as incubation progresses in propanol (black)ethanol (red) methanol (blue) and water (cyan) at 75 degC Thehollow triangles and inverted triangles represent cis and trans [FPG8K+ 2H]2+ ions respectively The abundance of [FPG8K + 2H]2+ ionsrsquocis and trans conformations is extracted from the CCS distributionobtained by IMSminusMS Kinetic fits from several models are shownincluding model 1 (for methanol green curves) and the best-fittingmodel for each process (see Tables S1minusS6 for details) BottomPercentage of cis relative to trans [FPG8K + 2H]2+ ions throughoutthe degradation process in propanol (rectangle) ethanol (circle)methanol (triangle) and water (diamond) at 75 degC Cis graduallyincreases as dissociation progresses to near 100 at the end of thedissociation The rate of the cis growth is proportional to thedegradation rate The entire dataset for the water system can be foundin Supporting Information Figure S16

Table 1 Summary of Transition State Thermodynamic Values at 298 K for FPG8K Trans rarr Cis Isomerization and cis-FPG8KDissociation in Different Solvents

solvents processa ΔG⧧ (kJmiddotmolminus1) ΔH⧧ (kJmiddotmolminus1) ΔS⧧ (Jmiddotmolminus1middotKminus1) Ea (kJmiddotmolminus1)

n-PrOH trans rarr cis (1) 1001 plusmn 21 1043 plusmn 16 141 plusmn 47 1018 plusmn 16n-PrOH dissociation (2) 994 plusmn 24 1056 plusmn 22 209 plusmn 35 1031 plusmn 22EtOH trans rarr cis (2) 1005 plusmn 11 1060 plusmn 05 184 plusmn 34 1035 plusmn 05EtOH dissociation (3) 999 plusmn 37 1095 plusmn 34 319 plusmn 55 1070 plusmn 34MeOH trans rarr cis (3) 1038 plusmn 17 1100 plusmn 13 210 plusmn 37 1076 plusmn 13MeOH dissociation (4) 1041 plusmn 04 1164 plusmn 02 411 plusmn 11 1139 plusmn 02H2O trans rarr cis (7) 1100 plusmn 32 1204 plusmn 25 350 plusmn 67 1179 plusmn 25

aThe number in the parenthesis represents the number of unseen intermediates derived from the best-fitting model

The Journal of Physical Chemistry B pubsacsorgJPCB Article

httpsdxdoiorg101021acsjpcb1c00269J Phys Chem B XXXX XXX XXXminusXXX

D

=minus

middot +kE

R TAln( )

1ln( )a

(4)

(Supporting Information Figures S10minusS13) activationenergy (Ea) and the pre-exponential factor (A) are determinedfrom the slope and y-intercept respectively of the best-fittingline Transition state theory then allows us to parse out thecontributions of ΔH⧧ and entropy (ΔS⧧) to the transitionbarriers in each solvent based on eqs 5 and 6

Δ = +DaggerH E RTa (5)

= Δ DaggerA

ek Th

e S RB (6)

Δ = Δ minus ΔDagger DaggerG H T S (7)

where h is Planckrsquos constant kB is Boltzmannrsquos constant R isthe gas constant T is temperature of the reaction and e isEulerrsquos number We calculated the Gibbs free energy accordingto eq 7 for each transition state (ΔG⧧) from ΔH⧧ and ΔS⧧determined here The derived values are shown in Table 1The consecutive-step analysis assumes identical rate constantsfor all steps through a group of unseen intermediates that leadto an observed changeeither isomerization or dissociationso these are single-step barriersThe derived enthalpy of activation for dissociation is higher

in each solvent than that for the isomerization step and thedifference grows with solvent polarity In propanol the ΔH⧧

values are essentially the same for the two steps in that theuncertainty ranges of the two values overlap These differenceseasily exceed the uncertainty ranges for ethanol and methanolso the solvent polarity effect is clearly real Not only is eachbarrier higher for the dissociation step than for trans rarr cisisomerization of Phe1minusPro2 but also dissociation involves moreconsecutive barriers (due to the unseen intermediates) in allthree solventsThe entropy of activation for both steps depends similarly

on solvent polarity We find that the value of ΔS⧧ for both theisomerization and dissociation steps increases with solventpolarity In addition the difference between ΔS⧧ forisomerization and dissociation in a given solvent increaseswith solvent polarity Unlike enthalpy although a higherentropy value eases the process The higher the value of ΔS⧧the greater the loss of molecular order as the combinedpeptideminussolvent system approaches the transition state Thedissociation transition state is looser than that of isomerizationthe channel on the potential energy surface through which a cisconformer dissociates is less entropically restrictive than thatthrough which trans to cis isomerization proceeds In otherwords in a potential energy diagram a broader range ofconformations is compatible with crossing the transition for cisconformer dissociation than for a trans isomer isomerizationIn each solvent the uncertainty range of ΔG⧧ for trans rarr cisisomerization overlaps with that for dissociation so the twovalues are statistically indistinguishable The barriers for transrarr cis isomerization at penultimate proline determined here aresimilar to previously reported values for other peptides939

While the ΔH⧧ values account for the bulk of each free energybarrier the entropic easing effect is the largest where theenthalpic barriers are the largest This results in both thesimilarity of ΔG⧧ for isomerization and dissociation in a givensolvent and the fact that all seven ΔG⧧ values across the threesolvents are in the same neighborhood There appears to be an

enthalpy and entropy compensation that results in similar freeenergy barriers for each step as observed in previousresearch40minus44

Of these three solvents FPG8K degradation kinetics is thefastest in propanol Propanol also yields the smallest number ofintermediates involved in both FPG8K trans rarr cis isomer-ization and cis-FPG8K dissociation processes We can makesense of this by noting that the strength of hydrogen bondinginteractions between the peptide and solvent should increasewith solvent polarity Therefore a more polar solvent requiresmore energy for the peptide to undergo structural changewhether trans rarr cis isomerization or cis conformationdissociation This manifests as a larger ΔH⧧ in a more polarsolvent The tighter solventminuspeptide interactions in a morepolar solvent also may play into the higher number of hiddenintermediate states as the peptide and surrounding solventmolecules both rearrange on the way to completing each of thetwo observable steps The peptideminussolvent and solventminussolvent interactions that must be overcome as the peptiderearranges have a greater initial ordering effect in a solvent thatmakes stronger hydrogen bonds The observed trend of higherΔS⧧ in a more polar solvent thus seems more likely to growout of greater order in the potential well before the transitionbarrier than from additional disorder in the transition stateitself

Transition State Thermochemistry and the DielectricConstant of the Solvent Because of the changes inthermochemistry observed with variations in solvent polarityit is interesting to consider the relationship of the measuredthermochemistry and the dielectric constants of the solventsPlots of transition state thermochemistry versus dielectricconstant for both isomerization and dissociation of the peptideFPG8K are shown in Figures 5 and 6 respectively Thecorresponding regression values and uncertainties are listed inTable 2 Transition state thermochemistry appears to becorrelated with the dielectric constant of the solvent BothΔH⧧ and TΔS⧧ increase with increases in the dielectricconstant

Figure 5 ΔH⧧ and TΔS⧧ vs dielectric constant (values are used at 25degC) diagram for the isomerization step of each solvent systeminvestigated The correlation equations for ΔH⧧ vs dielectric constantand TΔS⧧ vs dielectric constant are y = 026x + 9996 and y = 099x +279 respectively Each value is the average of triplicate measure-ments and error bars represent the standard deviation about themean T is 29815 K

The Journal of Physical Chemistry B pubsacsorgJPCB Article

httpsdxdoiorg101021acsjpcb1c00269J Phys Chem B XXXX XXX XXXminusXXX

E

In Figures 5 and 6 trend lines are derived based ontransition state enthalpy and entropy values Extrapolation to ε= 1 yields transition state thermochemistry corresponding tovacuum for the isomerization step ΔH⧧ ΔS⧧ and ΔG⧧ are1002 kJmiddotmolminus1 97 Jmiddotmolminus1middotKminus1 and 973 kJmiddotmolminus1 respec-tively and for the dissociation step ΔH⧧ ΔS⧧ and ΔG⧧ are887 kJmiddotmolminus1 minus97 Jmiddotmolminus1middotKminus1 and 916 kJmiddotmolminus1respectively These results represent the contributions fromthe peptide itself to the total transition state thermochemistryThe values here suggest that both steps are enthalpicallydisfavored However these extrapolations to vacuum predictΔH⧧ for isomerization to be higher than for dissociationwhich is the opposite of our solution-phase results In additionwhile both steps are entropically favored in solution invacuum the dissociation step is predicted to be restrictedentropically This comparison is also reversed from thesolution-phase results A negative ΔS⧧ makes sense fordissociation in that the reaction requires a precise attack ofan amine nitrogen on a carbonyl carbon These trend linesmight also allow predictions about the reaction rate andtransition state thermochemistry in biological environmentswith varying local dielectric characteristics It is worthmentioning that the correlation curves are different with orwithout the derived transition state thermochemistry values ofwater for the isomerization step The derived values for thepeptide isomerization without including the data with water asa solvent are as follows ΔH⧧ ΔS⧧ and ΔG⧧ values are 944 kJmiddotmolminus1 410 Jmiddotmolminus1middotKminus1 and 933 kJmiddotmolminus1 respectivelyAlthough these values vary once we include the data

associated with water as the solvent the general thermochem-ical trends remain consistent and this does not change ouroverall conclusionsBoth ΔH⧧ and TΔS⧧ grow with increasing dielectric

constant This means the entropic barrier is eased as theenthalpic barrier grows Conversely the dissociation in thesolvent with a lower dielectric constant is favored largely bylowering the enthalpic barrier while slightly increasing theentropic barrier This opposed trend of entropic and enthalpicbarriers with a dielectric constant is reminiscent of enzymebehavior The dielectric environment of the active pocket of anenzyme depends on which residues are present but will ingeneral differ significantly from the dielectric environment ofthe solvent which could contribute to high catalyticactivity45minus47 This difference of permittivity leads to a decreasein activation energy at the expense of entropic restrictionwhich parallels the behavior we see here with changes insolvent dielectric constantThe fact that the fragmentation rate increases with

decreasing dielectric constant leads us to offer a finalsuggestion In the nonpolar environment of a biologicalmembrane the DKP formation rate should be faster than thesame reaction under aqueous conditions This implies that aspeptides are screened for possible drug or antigen use in vivosusceptibility to DKP formation should be considered and ifpossible such studies should consider a variety of environ-ments

Proton Concentration from Acetic Acid Dissociationin Different Solvent Considering the fact that DKPformation reaction is influenced by the protonation of theN-terminus (a free amino group is necessary for thenucleophilic attack) the dissociation of acetic acid in differentsolvents which contributes to the donation of protons to thesolvent could affect dissociation The pKa values of acetic acidin PrOH EtOH MeOH and H2O are 1045 1032 963 and475 respectively48minus50 Since 1 acetic acid (175 times 104 μM)is added to each solution the calculated [H+] in H2O MeOHEtOH and PrOH are 55 times 10minus4 20 times 10minus6 92 times 10minus7 and79 times 10minus7 M respectively The correlation between log[H+]and the solvent dielectric constant appears to be linear (shownin Figure S15) The concentration of the peptide FPG8K is 20μM sim004 10 22 and 25 times the concentration of [H+] inH2O MeOH EtOH and PrOH respectively Overall we seethat this trend is opposed to experimental finding suggestingthat excess protons have little if any influence on this system

CONCLUSIONSIMSminusMS techniques were used to investigate the dissociationkinetics of FPG8K in different solvents at multiple definedtemperatures The CCS distribution shows that two con-formations in solution result from the peptidersquos trans rarr cisisomerization at proline Using IMSminusMS we are able tomonitor changes in abundance of the cis and trans isomersalongside the dissociation process We found that there are twoprocesses involved in the dissociation trans rarr cis isomer-ization and cis isomer degradation We found that DKPformation is more rapid in less polar solvents Decreasedsolvent polarity also enhances the preference for the cisconformation over trans at the penultimate proline and speedsup the trans rarr cis isomerization process By monitoring thedissociation process at multiple temperatures and modeling thedissociation pathway we found that a number of unseenintermediates are involved in both isomerization and

Figure 6 ΔH⧧ and TΔS⧧ vs dielectric constant (values are used at 25degC) diagram for the dissociation step of each solvent systeminvestigated The correlation equation for ΔH⧧ vs dielectric constantand TΔS⧧ vs dielectric constant are y = 094x + 8780 and y = 050x minus340 respectively Each value is the average of triplicate measure-ments and error bars represent the standard deviation about themean T is 29815 K

Table 2 Regression Values and Uncertainties of Correlationof ΔH⧧ and TΔS⧧ Versus Dielectric Constant forIsomerization and Dissociation Processes

process equation R-square

ΔH⧧ vs ε isomerization y = 026x + 9996 0975ΔH⧧ vs ε dissociation y = 094x + 8780 1000TΔS⧧ vs ε isomerization y = 099x + 279 0974TΔS⧧ vs ε dissociation y = 050x minus 340 0949

The Journal of Physical Chemistry B pubsacsorgJPCB Article

httpsdxdoiorg101021acsjpcb1c00269J Phys Chem B XXXX XXX XXXminusXXX

F

dissociation This modeling also allowed us to determinetransition state thermodynamic values There appears to be astrong correlation between the dielectric constant andtransition state enthalpy and entropy The solvent with alower dielectric constant speeds up the reaction by loweringthe enthalpic barrier although slightly increasing the entropicbarrier The Gibbs free energy of activation in all the solventsare primarily controlled by transition state enthalpy

ASSOCIATED CONTENTsı Supporting InformationThe Supporting Information is available free of charge athttpspubsacsorgdoi101021acsjpcb1c00269

Dissociation mass spectra of FPG8K in water kineticsdata of [FPG8K + 2H]2+ in different solutions undermultiple elevated temperatures Arrhenius plots for thepeptide FPG8K in different solutions schematic diagramof a 2 m instrument log[H+] versus dielectric constantdiagram percentage of cis relative to the trans isomerthroughout the degradation process in water andresidue sums of square values for FPG8K dissociationand trans rarr cis isomerization processes in differentsolutions (PDF)

AUTHOR INFORMATIONCorresponding AuthorsDavid A Hales minus Department of Chemistry IndianaUniversity Bloomington Indiana 47401 United StatesDepartment of Chemistry Hendrix College ConwayArkansas 72032 United States Email haleshendrixedu

David E Clemmer minus Department of Chemistry IndianaUniversity Bloomington Indiana 47401 United Statesorcidorg0000-0003-4039-1360 Email clemmer

indianaedu

AuthorsZhi-chao Zhang minus Department of Chemistry IndianaUniversity Bloomington Indiana 47401 United States

Shannon A Raab minus Department of Chemistry IndianaUniversity Bloomington Indiana 47401 United States

Complete contact information is available athttpspubsacsorg101021acsjpcb1c00269

NotesThe authors declare no competing financial interest

ACKNOWLEDGMENTSThis work is supported by funds from the National Institutesof Health grant 5R01GM121751-03 (DEC) the Robert andMarjorie Mann Graduate Research Fellowship (ZZ andSAR) from Indiana University and a Faculty Project Grantfrom Hendrix College (DAH) The authors acknowledgeNavneet Sahota Dr Daniel Fuller and Dr Chris Conant forhelp and NS CRC and DRF were supported by fellowshipsfrom the Robert and Marjorie Mann Chair

REFERENCES(1) Manning M C Chou D K Murphy B M Payne R WKatayama D S Stability of protein pharmaceuticals an updatePharm Res 2010 27 544minus575(2) Manning M C Patel K Borchardt R T Stability of proteinpharmaceuticals Pharm Res 1989 06 903minus918

(3) Capasso S Vergara A Mazzarella L Mechanism of 25-Dioxopiperazine Formation J Am Chem Soc 1998 120 1990minus1995(4) Marsden B J Nguyen T M Schiller P W Spontaneousdegradation via diketopiperazine formation of peptides containing atetrahydroisoquinoline-3-carboxylic acid residue in the 2-position ofthe peptide sequence Int J Pept Protein Res 1993 41 313minus316(5) Steinberg S M Bada J L Peptide decomposition in the neutralpH region via the formation of diketopiperazines J Org Chem 198348 2295minus2298(6) Beyermann M Bienert M Niedrich H Carpino L A Sadat-Aalaee D Rapid continuous peptide synthesis via FMOC amino acidchloride coupling and 4-(aminomethyl)piperidine deblocking J OrgChem 1990 55 721minus728(7) Battersby J E Hancock W S Canova-Davis E Oeswein JOrsquoConnor B Diketopiperazine formation and N-terminal degrada-tion in recombinant human growth hormone Int J Pept Protein Res1994 44 215minus222(8) Kertscher U Bienert M Krause E Sepetov N F Mehlis BSpontaneous chemical degradation of substance P in the solid phaseand in solution Int J Pept Protein Res 1993 41 207minus111(9) Fuller D R Conant C R El-Baba T J Brown C JWoodall D W Russell D H Clemmer D E Conformationallyregulated peptide bond cleavage in bradykinin J Am Chem Soc2018 140 9357minus9360(10) Moslashss J Bundgaard H Kinetics and mechanism of the facilecyclization of histidyl-prolineamide to cyclo (His-Pro) in aqueoussolution and the competitive influence of human plasma J PharmPharmacol 1990 42 7minus12(11) Oyler A R Naldi R E Lloyd J R Graden D A Shaw CJ Cotter M L Characterization of the solution degradation productsof histrelin a gonadotropin releasing hormone (LHRH) agonist JPharm Sci 1991 80 271minus275(12) Straub J A Akiyama A Parmar P Musso G F Chemicalpathways of degradation of the bradykinin analog RMP-7 Pharm Res1995 12 305minus308(13) Cobbold M De La Pena H Norris A Polefrone J MQian J English A M Cummings K L Penny S Turner J ECottine J et al MHC Class I-associated hosphopeptides are thetargets of memory-like immunity in leukemia Sci Transl Med 20135 203ra125(14) Sepetov N F Krymsky M A Ovchinnikov M VBespalova Z D Isakova O L Soucek M Lebl M Rearrangementracemization and decomposition of peptides in aqueous solutionPept Res 1991 4 308minus313(15) Capasso S Mazzarella L Activation of diketopiperazineformation by alkylammonium carboxylate salts and aprotic dipolarprotophobic solvents Peptides 1998 19 389minus391(16) Capasso S Mazzarella L Solvent effects on diketopiperazineformation from N-terminal peptide residues J Chem Soc PerkinTrans 2 1999 329minus332(17) Clemmer D E Jarrold M F Ion mobility measurements andtheir applications to clusters and biomolecules J Mass Spectrom1997 32 577minus592(18) Bohrer B C Merenbloom S I Koeniger S L HilderbrandA E Clemmer D E Biomolecule analysis by ion mobilityspectrometry Annu Rev Anal Chem 2008 1 293minus327(19) Conant C R Fuller D R El-Baba T J Zhang Z RussellD H Clemmer D E Substance P in solution trans-to-cisconfigurational changes of penultimate prolines initiate non-enzymatic peptide bond cleavages J Am Soc Mass Spectrom 201930 919minus931(20) Shi L Holliday A E Khanal N Russell D H Clemmer DE Configurationally-coupled protonation of polyproline-7 J AmChem Soc 2015 137 8680minus8683(21) Coin I Beyermann M Bienert M Solid-phase peptidesynthesis from standard procedures to the synthesis of difficultsequences Nat Protoc 2007 2 3247minus3256

The Journal of Physical Chemistry B pubsacsorgJPCB Article

httpsdxdoiorg101021acsjpcb1c00269J Phys Chem B XXXX XXX XXXminusXXX

G

(22) Koeniger S L Merenbloom S I Sevugarajan S ClemmerD E Transfer of structural elements from compact to extended statesin unsolvated ubiquitin J Am Chem Soc 2006 128 11713minus11719(23) Merenbloom S I Koeniger S L Valentine S J PlasenciaM D Clemmer D E IMSminusIMS and IMSminusIMSminusIMSMS forseparating peptide and protein fragment ions Anal Chem 2006 782802minus2809(24) Mesleh M F Hunter J M Shvartsburg A A Schatz G CJarrold M F Structural information from ion mobility measurementseffects of the long-range potential J Phys Chem 1996 100 16082minus16086(25) Mason E A McDaniel E W Transport Properties of Ions inGases John Wiley amp Sons Hoboken NJ 1988(26) Revercomb H E Mason E A Theory of plasmachromatographygaseous electrophoresis Review Anal Chem 197547 970minus983(27) Wyttenbach T von Helden G Batka J J Carlat DBowers M T Effect of the long-range potential on ion mobilitymeasurements J Am Soc Mass Spectrom 1997 8 275minus282(28) McLean J A Ruotolo B T Gillig K J Russell D H Ionmobility-mass spectrometry a new paradigm for proteomics Int JMass Spectrom 2005 240 301minus315(29) Kanu A B Dwivedi P Tam M Matz L Hill H H Jr Ionmobility-mass spectrometry J Mass Spectrom 2008 43 1minus22(30) Pierson N A Chen L Russell D H Clemmer D ECisminusTrans Isomerizations of proline residues are key to bradykininconformations J Am Chem Soc 2013 135 3186minus3192(31) Valentine S J Counterman A E Hoaglund-Hyzer C SClemmer D E Intrinsic amino acid size parameters from a series of113 lysine-terminated tryptic digest peptide ions J Phys Chem B1999 103 1203minus1207(32) Srebalus Barnes C A Clemmer D E Assessing intrinsic sidechain interactions between i and i + 4 residues in solvent-freepeptides a combinatorial gas-phase approachdagger J Phys Chem A 2003107 10566minus10579(33) Fuller D R Conant C R El-Baba T J Brown C JWoodall D W Russell D H Clemmer D E Conformationallyregulated peptide bond cleavage in bradykinin J Am Chem Soc2018 140 9357minus9360(34) Brandts J F Halvorson H R Brennan M Consideration ofthe possibility that the slow step in protein denaturation reactions isdue to cis-trans isomerism of proline residues Biochemistry 1975 144953minus4963(35) Brandts J F Brennan M Lung-Nan Lin L-N Unfolding andrefolding occur much faster for a proline-free proteins than for mostproline-containing proteins Proc Natl Acad Sci USA 1977 744178minus4181(36) Lin L-N Brandts J F Further evidence suggesting that theslow phase in protein unfolding and refolding is due to prolineisomerization a kinetic study of carp parvalbumins Biochemistry1978 17 4102minus4110(37) El-Baba T J Kim D Rogers D B Khan F A Hales D ARussell D H Clemmer D E Long-lived intermediates in acooperative two-state folding transition J Phys Chem B 2016 12012040minus12046(38) Shi L Holliday A E Shi H Zhu F Ewing M A RussellD H Clemmer D E Characterizing intermediates along thetransition from polyproline I to polyproline II using ion mobilityspectrometry-mass spectrometry J Am Chem Soc 2014 13612702minus12711(39) Shi L Holliday A E Glover M S Ewing M A Russell DH Clemmer D E Ion mobility-mass spectrometry reveals theenergetics of intermediates that guide polyproline folding J Am SocMass Spectrom 2016 27 22minus30(40) Liu L Guo Q-X Isokinetic relationship isoequilibriumrelationship and enthalpyminusentropy compensation Chem Rev 2001101 673minus696

(41) Pan A Biswas T Rakshit A K Moulik S PEnthalpyminusentropy compensation (EEC) effect a revisit J PhysChem B 2015 119 15876minus15884(42) Sharp K Entropyndashenthalpy compensation Fact or artifactProtein Sci 2001 10 661minus667(43) Chodera J D Mobley D L Entropy-enthalpy compensationrole and ramifications in biomolecular ligand recognition and designAnnu Rev Biophys 2013 42 121minus142(44) El-Baba T J Fuller D R Hales D A Russell D HClemmer D E Solvent mediation of peptide conformationspolyproline structures in water methanol ethanol and 1-propanolas determined by ion mobility spectrometry-mass spectrometry J AmSoc Mass Spectrom 2019 30 77minus84(45) Krishtalik L The specific features of enzyme as a polar mediumand their role in the mechanism of the enzymatic process Mol Biol1974 8 91minus99(46) Krishtalik L I Globule size and the activation energy of anenzymatic process Mol Biol 1979 13 577minus581(47) Krishtalik L I Catalytic acceleration of reactions by enzymesEffect of screening of a polar medium by a protein globule J TheorBiol 1980 86 757minus771(48) Goffredi M Shedlovsky T Electrolytic conductance inalcohol-water mixtures V Ionization constant of acetic acid in 1-propanol-water mixtures at 15 25 and 35degree J Phys Chem B1967 71 4436minus4442(49) Shedlovsky T Kay R L The ionization constant of acetic acidin waterminusmethanol mixtures at 25deg from conductance measurementsJ Phys Chem B 1956 60 151minus155(50) Grunwald E Berkowitz B J The measurement andcorrelation of acid dissociation constants for carboxylic acids in thesystem ethanolminuswater Activity Coefficients and Empirical ActivityFunctions1a J Am Chem Soc 1951 73 4939minus4944

The Journal of Physical Chemistry B pubsacsorgJPCB Article

httpsdxdoiorg101021acsjpcb1c00269J Phys Chem B XXXX XXX XXXminusXXX

H