Innovative Strategies for Stabilization of Therapeutic - TI Pharma

InnovatIve strategIes for stabIlIzatIon 

of therapeutIc peptIdes In aqueous 

formulatIons 

Christina Avanti

Paranimfen: Milica Stankovic 

Leviny Zachreina 

The research presented in this thesis was performed within the 

framework of project D6-202 of the Dutch Top Institute Pharma. 

Printing of this thesis was supported by generous 

contributions from: 

University of Groningen 

Faculty of Mathematics and Natural Sciences of the 

University of Groningen 

ISBN: 978-94-6182-122-5 

© Copyright 2012 C. Avanti 

All rights reserved. No part of this thesis may be reproduced, 

stored in a retrieval system or transmitted in any form or 

by any means, mechanically, by photocopying, recording or 

otherwise, without the written permission of the author. 

Cover design: Bao Tung Pham and Christina Avanti 

Layout & printing: Off Page, Amsterdam

RIJKSUNIVERSITEIT GRONINGEN 

InnovatIve strategIes for stabIlIzatIon 



Proefschrift 

ter verkrijging van het doctoraat in de 

Wiskunde en Natuurwetenschappen 

aan de Rijksuniversiteit Groningen 

op gezag van de 

Rector Magnificus, dr. E. Sterken, 

in het openbaar te verdedigen op 

maandag 2 juli 2012 

om 16.15 uur 

door 

Christina Avanti 

geboren op 3 april 1968 

Kota Baru, Indonesië

Promotor: Prof. dr. H.W. Frijlink 

Copromotor: Dr. W.L.J. Hinrichs 

Beoordelingscommissie: Prof. dr. G.M.M. Groothuis 

Prof. dr. A.J.M. Driessen 

Prof. dr. W. Jiskoot

contents 

PART ONE INTRODUCTION 

Chapter 1 General Introduction 9 

Chapter 2 Current Strategies for Stabilization of Therapeutic Peptides in 

Aqueous Formulations 13 

PART TWO THE USE OF DIVALENT METAL IONS AND CITRATE BUFFERS 

TO STABILIZE OXYTOCIN IN AQUEOUS SOLUTION 

Chapter 3 A New Strategy to Stabilize Oxytocin in Aqueous Solutions : 

I. The Effects of Divalent Metal Ions and Citrate Buffer 35 

Chapter 4 A New Strategy To Stabilize Oxytocin in Aqueous Solutions: 

II. Suppression of Cysteine-Mediated Intermolecular Reactions 

by a Combination of Divalent Metal Ions and Citrate 51 

PART THREE THE USE OF DIVALENT METAL IONS AND ASPARTATE 

BUFFER TO STABILIZE OXYTOCIN IN AQUEOUS SOLUTION 

Chapter 5 Insight into the Stability of the Zinc-Aspartate-Oxytocin 

Complex 79 

Chapter 6 Aspartate buffer and divalent metal ions affect the oxytocin 

conformation in aqueous solution and protect it from 

degradation 95 

PART FOUR THE USE OF EXTREMOLYTES TO STABILIZE PROTEIN IN 

AQUEOUS SOLUTION 

Chapter 7 Extremolytes: Are There Universal Stabilizers for Proteins in 

Aqueous Solution? 121 

Summary, Concluding Remarks, and Global Perspective 137 

Samenvatting, Conclusies, Aanbevelingen, en Mondiaal Perspectief 145 

Acknowledgments 153

for all the women in existence… 

for a better chance to live the life with the new born …

Every year 166 000 women die of bleeding after child birth, 

and more than 50% of these deaths occur in sub-Saharan Africa 

(Clyburn et. al., 2007)

general IntroductIon 

1

1 

10 

general IntroductIon 

Although successful developments in the field of peptide synthesis increased the availability 

of peptide drugs [1], a significant part of the world’s population is still facing serious 

problems associated with insufficient access to several essential drugs which are peptide in 

nature. The major cause for this problem is the lack of stable formulations that withstand 

transport, storage and distribution, particularly in rural and remote areas of developing 

(tropical) countries that lack a cold chain. 

An important active pharmaceutical peptide is oxytocin, a nonapeptide which is the 

drug of first choice to treat bleeding after childbirth or post-partum hemorrhage (PPH) [2]. 

Although the availability of this drug has greatly declined maternal mortality rates in 

the developed world, PPH remains a leading cause of maternal mortality elsewhere [3]. 

Current commercial formulations of oxytocin are insufficiently heat-stable to withstand 

tropical conditions [4,5]. Cleavage of the disulfide bridge was found to be the major 

degradation pathway [6,7]. 

The aim of the present thesis was to develop heat-stable formulations for polypeptide 

drugs, in particular oxytocin and investigate the mechanism of stabilization based on 

degradation products formed during thermal stress before and after formulation. 

This thesis is divided in four parts. In part 1 (Chapter 2) a general introduction is given 

on the stability of peptide drugs in aqueous solution and current stabilization technologies. 

Part 2 (Chapter 3 and 4) describes the use of a combination of divalent metal ions and citrate 

buffer to improve stability of oxytocin in aqueous solution, whereas part 3 (Chapter 5 and 6) 

is about the effect of zinc ions to improve stability of oxytocin in aspartate buffer. Finally, 

part 4 (Chapter 7) describes an attempt to find the universal stabilizer for polypeptides by 

investigating various extremolytes using lysozyme and insulin as model peptides. 

In Chapter 2, we review the most common degradation pathways of peptides, such 

as hydrolysis, deamidation, isomerization, racemization, oxidation, disulfide exchange, 

dimerization, and further aggregation that cause loss of potency of pharmaceutical peptides. 

In this chapter we also review different strategies to overcome these instabilities, such as pH 

optimization, the use of buffers, antioxidant and other additives. 

In Chapter 3, we describe an investigation on the effect of monovalent (Na + and K + ) and 

divalent (Ca 2+ , Mg 2+ , and Zn 2+ ) metal ions in combination with citrate and acetate buffers 

at pH of 4.5 on the stability of oxytocin in aqueous solution. The effect of combinations of 

buffers and metal ions on the stability of aqueous oxytocin solutions was determined by 

reversed-phase high performance liquid chromatography (RP-HPLC) and size exclusion 

chromatography (HP-SEC) after 4 weeks of storage at either 4°C or 55°C. We also measured 

the interaction between oxytocin and Ca 2+ , Mg 2+ , or Zn 2+ in citrate buffer in comparison 

with acetate buffer by using isothermal titration calorimetry (ITC). 

In Chapter 4, we identified various degradation products of oxytocin in citrate-buffered 

solution after thermal stress at a temperature of 70 °C for 5 days and the differences in degradation 

pattern in the presence and absence of divalent metal ions. Degradation products of oxytocin in 

the citrate buffer formulation with and without divalent metal ions were analyzed using liquid 

chromatography−mass spectrometry/mass spectrometry (LC−MS/MS). 

In Chapter 5, we investigated the effects of various metal ions (Ca 2+ , Mg 2+ and Zn 2+ ) on 

the stability of oxytocin in aspartate buffer pH 4.5 and determined their interaction with 

the peptide in aqueous solution. The effect of combinations of various metal ions on the 

stability of oxytocin in aspartate buffer solutions was determined by RP-HPLC and HP-SEC

after 4 weeks of storage at either 4°C or 55°C. We also investigated which degradation 

products of oxytocin were formed in the aspartate buffer formulation with and without 

divalent metal ions using LC−MS/MS and determined the interaction between oxytocin 

and Ca 2+ , Mg 2+ , or Zn 2+ in aspartate buffer by using ITC. 

In Chapter 6, we further explored the mechanism of stabilization of oxytocin by the 

combination of Zn 2+ and aspartate buffer. Therefore, we investigated the conformation of 

oxytocin in aspartate buffer in the presence of Zn 2+ in comparison with Mg 2+ using 2D 

NMR spectroscopy, i.e. NOESY, TOCSY, 1 H- 13 C HSQC and 1 H- 15 N HSQC with neither 13 C 

nor 15 N enrichment. 

In Chapter 7, we describe a study on the effects of extremolytes on the stabilization of two 

model proteins (the larger peptides) lysozyme and insulin in aqueous solutions. The effects of 

different extremolytes (betaine, hydroxyectoine, trehalose, ectoine, and firoin) on the stability 

of lysozyme were determined by Nile red Fluorescence Spectroscopy and a bioactivity assay. 

Insulin stability was determined by RP-HPLC and HP-SEC. The effects of extremolytes on the 

unfolding temperature of the proteins were analyzed using a thermal shift assay for lysozyme 

and liquid differential scanning microcalorimetry for insulin. The interaction between 

extremolytes and protein was studied by isothermal titration calorimetry (ITC). 

references 

1. B.A. Moss, Peptide Synthesis, in: S. Frokjaer, 

L. Hovgaard (Eds.), Pharmaceutical 

Formulation Development of Peptides 

and Proteins, CRC Press, United States of 

America, (2000) 1-12. 

2. R.A. Dyer, D. van Dyk, A. Dresner, The use of 

uterotonic drugs during caesarean section, 

Int. J. Obstet. Anesth. 19 (2010) 313-19. 

3. P. Clyburn, S. Morris, J. Hall, Anaesthesia 

and safe motherhood, Anaesthesia 62 (2007) 

21-5. 

4. H.V. Hogerzeil, G.J.A. Walker, M.J. De Goeje, 

Stability of injectable ocytocics in tropical 

climates, World Health Organization, 

Geneva WHO/DAP/93.6. (1993). 

5. A.N.J.A.D. Groot, T.B. Vree, H.V. Hogerzeil, 

G.J.A. Walker, WHO Action Programme on 

Essential Drugs: Stability of Oral Oxytocics 

in Tropical Climates : Results of Simulation 

Studies on Oral Ergometrine, Oral 

Methylergometrine, Buccal Oxytocin and 

Buccal Desamino-Oxytocin, World Health 

Organization, Geneva, (1994). 

6. A. Hawe, R. Poole, S. Romeijn, P. Kasper, 

R. van der Heijden, W. Jiskoot, Towards 

heat-stable oxytocin formulations: analysis 

of degradation kinetics and identification 

of degradation products, Pharm. Res. 26 

(2009) 1679-88. 

7. C. Avanti, H.P. Permentier, A.V. Dam, R. 

Poole, W. Jiskoot, H.W. Frijlink, W.L.J. 

Hinrichs, A new strategy to stabilize oxytocin 

in aqueous solutions: II. Suppression of 

cysteine-mediated intermolecular reactions 

by a combination of divalent metal ions and 

citrate, Mol. Pharmaceutics 9 (2012) 554-62. 

General IntroductIon 

1 

11

Christina Avanti 1 , Wim Jiskoot 2 , Wouter L. J. Hinrichs 1 , and Henderik W. Frijlink 1 

1 Department of Pharmaceutical Technology and Biopharmacy, 

University of Groningen, Groningen, The Netherlands 

2 Division of Drug Delivery Technology, 

Leiden/Amsterdam Center for Drug Research, Leiden University, Leiden, The Netherlands

current strategIes for stabIlIzatIon 



2

2 

abstract 

Parenteral administration is one of the most used routes to obtain systemic delivery of 

peptide drugs. Although most of therapeutic peptides are formulated in the dry powder 

for reconstitution, from economic point of view aqueous liquid formulations are preferred. 

However, therapeutic peptides are often unstable in the aqueous formulation of the 

injection. This review will focus on the formulation strategies that can be applied to stabilize 

therapeutic peptides in aqueous solution. We have organized this review as follows: first the 

main peptide stability problems in solution are described followed by a discussion on the 

known strategies to reduce peptide degradation, including recent research on improving 

peptide stability in liquid formulations.

1. IntroductIon 

There has been a rapid expansion in the use of peptides as potential drugs since the successful 

chemical synthesis of oxytocin by duVigneaud in 1953 [1]. The following decades witnessed 

the discovery of a large number of peptides as active pharmaceutical ingredients, and this 

process is likely to continue in the future. Currently over a hundred peptide drug candidates 

are in development for a wide variety of diseases such as several forms of cancer and metabolic 

diseases [2]. Examples of several peptides in the Dutch market are described in Table 1. 

Peptides differ from proteins in that they are smaller and typically lack a defined tertiary 

structure, but a clear distinction is difficult to make and several arbitrary definitions exist. 

For instance, Malavolta [3] defined peptides as molecules containing fewer than 40 amino 

acid residues, while proteins contain 50 residues or more, with in between a category 

called polypeptides (40-49 residues). In another source, amino acids joined together in 

chains of 50 amino acids or less are defined as peptides, 50-100 amino acids are defined 

as polypeptides, and over 100 amino acids are defined as proteins [4]. Although also other 

definitions of the term peptide have been proposed, in this review we will focus on peptides 

containing less than 50 amino acid residues. Within this definition peptides can range in 

size from dimers containing two (modified) amino acid residues, such as enalapril and 

lisinopril [5], through small peptides, such as oxytocin [6] and octreotide [7] containing 

fewer than 10 residues, to relatively large polypeptides, such as calcitonin (32 residues) [8]. 

A number of hormones, enzymes, antitumor agents, antibiotics and neurotransmitters 

are peptides. Peptides regulate many physiological processes, acting at some sites as 

endocrine or paracrine signals and at other sites as neurotransmitters or growth factors. 

Nowadays, peptides are used as therapeutic agents in diverse disease areas such as 

neurological, endocrinological and hematological disorders [9]. 

Therapeutic peptides pose a number of challenges for pharmaceutical scientists regarding 

their formulation and delivery. The sensitivity of many peptides to enzymatic breakdown 

(e.g. in the gastro-intestinal tract) [10] and their poor ability to pass absorbing membranes 

typically results in a poor bioavailability following non-parenteral administration [11]. 

Moreover, the lack of physical and chemical stability may lead to significant degradation 

during processing and storage of the (aqueous) formulations. Enalapril and lisinopril do 

not have the typical delivery issues and oral delivery is well possible. Therefore we do not 

discuss those modified dipeptides in this review. 

The lack of oral efficacy of peptides has prompted the examination of other noninvasive 

routes for peptide drug administration [4,12]. These routes include buccal [13-15], 

rectal [16,17], vaginal [18,19], percutaneous [20], ocular [21,22], transdermal [23,24], 

nasal [25-27] and the pulmonary route [28,29]. However, many of these routes of 

administration are still under investigation and they may be insufficiently efficient, 

especially when a rapid effect is desired. Therefore, for delivery of therapeutic peptides 

formulations the invasive parenteral routes are still often preferred [30]. Intravenous 

administration is the most efficient way to deliver peptide drugs directly into the systemic 

circulation, since no absorption process is involved. Intramuscular routes can also be used, 

however, the absorption is slower than after intravenous administration. Both intravenous 

and intramuscular routes do not easily allow self-administration and patient experiences 

pain after injection. The subcutaneous route can be employed when a more sustained action 

is acceptable or required and this route is more suitable for self-administration. 

StabIlIzatIon of therapeutIc peptIdeS 

2 

15

2 

16 

Table 1 Several therapeutic peptide-based products in the Dutch market in 2010-2011 

Generic 

names 

Trade 

names Principal 

Dosage 

form 

Self-life and 

storage cond. 

pH 

(adj. agent) 

Thyrotropin Releasing Hormones 

Protirelin 

Antibiotic peptides 

Relefact 

TRH® 

Sanovi-Aventis Liquid inj. 3 yrs at 15-25°C 6.5 (Phosphate) 

Daptomycin Cubicin® Novartis 

Powder 

for inj. 

3 yrs at 2-8°C 4-5 (NaOH) 

Teicoplanin Targocid® Sanofi-Aventis 

Powder 


4 yrs at 2-8°C 3.8 to 6.5 

Platelet aggregates inhibitors 

Eptifibatide Integrilin® GlaxoSmithKline 

Liquid for 

infusion 

3 yrs at 2-8°C 5.35 (citrate) 

Somatostatin analog 

Octreotide Sandostatin® Novartis Liquid inj. 3 yrs at 2-8°C 4.2 (lact/carb.) 

Vasopressins and analog 

Desmopressin DDAVP® Ferring Liquid inj. 4 yrs at 2-8°C 4-5 

Octostim® Ferring Liquid inj. 4 yrs at 2-8°C 4 (HCl) 

Minrin® Ferring Liquid inj. 4 yrs at 2-8°C 4 (HCl) 

Felypressin 

Oxytocins 

Citanest 3% 

Octapressin® 

Densply Liquid inj. 3 yrs at 15-25°C 3.2 

Oxytocin Syntocinon® Defiante Liquid inj. 4 yrs at 2-8°C 4 (acetate) 

carbetocin 

Oxytocin antagonist 

Pabal® Ferring Liquid inj. 2 yrs 2-8°C 3.8 (acetate) 

Atosiban Tractocile ®® Ferring Liquid inj. 4yrs at 2-8°C 4.5 (HCl) 

Gonadotropin Releasing Hormone (GNRH)/Luteinizing Releasing Hormone (LHRH) agonists 

Goserelin Zoladex® Div Liquid inj. 3 yrs < 25°C 

Gonadorelin 

Relefact 

LH-RH® 

Sanofi-Aventis Liquid inj. 15-25°C 

Powder 

Triptorelin 

Decapeptyl 

-CR® 

Ipsen 

and solvent 

for solution 


3 yrs at 2-8°C 

nafarelin Synarel® Pfizer 

Liquid 

nasal spray 

Powder 

2 yrs < 25*C 5-7 (acetate) 

Leuprolide Eligard ® Sanofi-Aventis 

and solvent 



2 yrs at 2-8°C 

Cetrorelix Cetrotide® Serono 

Powder 


3 yrs at 15-25°C 

Source: http://www.geneesmiddelenrepertorium.nl/repertorium 

continued next page

Table 1 continued Several therapeutic peptide-based products in the Dutch market in 2010-2011 

Generic 

names 

Trade 

names Principal 

Dosage 

form 

Self-life and 

storage cond. 

pH 

(adj. agent) 

Non-Steroidal Anti-Inflammatory Drugs 

Ziconotide 

Calcitonins 

Prialt® Elan Liquid inj. 3 yrs at 2-8°C 4-5 (HCl/NaOH) 

Salmon 

Calcitonin 

Calcitonin 

Sandoz® 

Novartis Liquid inj. 5 yrs at 2-8°C 3.3-3.7 

Human Parathyroid Hormone [hPTH (1–34) 

Teriparatide Forsteo® Eli Lily Liquid inj. 2 yrs at 2-8°C 4 (acetate) 

Fusion inhibitor of Human Immunodeficiency Virus (HIV) with Cluster Difference 4 (CD4) cells 

Powder 

Enfuvirtide Fuzeon® Roche 

and solvent 



4 yrs at 2-8°C 9-9.5 (carbonate) 

Adrenocorticotropin Hormone (ACTH) and derivatives 

Corticorelin 

CRH - 

Ferring® 

Ferring 

Powder 


3 yrs < 25°C 

Source: http://www.geneesmiddelenrepertorium.nl/repertorium 

Because of their instability, most peptide drugs have to be stored and transported at 

low temperatures. This has a big impact on the access to pharmaceutical active peptides, 

particularly in rural and tropical areas that lack a cold chain [31,32]. Therefore, an urgent 

need exists for new strategies to tackle the problems associated with peptide instability, 

especially for aqueous injectable formulations which are preferred over freeze-dried 

formulations. Although freeze-dried formulations seem an ideal strategy to maintain the 

peptide´s structural integrity [9], unfortunately, this strategy is economically not ideal. 

Particularly for developing countries, the lyophilization process may be too expensive. 

Production costs are high and the products that have to be stored and transported have a 

volume and mass up to twice as large as that of a liquid formulation, since the storage and 

transport includes both the vials containing the lyophilized powder and sterile water for 

reconstitution. Therefore, liquid formulations are preferred. 

Stability of peptide drugs in liquid formulations is the most important factor to be 

considered in the design and development of liquid parenteral peptide formulations. Table 1 

displays the shelf-life and storage conditions of the various marketed peptides, showing that 

the majority of the formulations is to be stored under refrigerated conditions. This is a clear 

indication for the poor stability of many products. Loss of potency of a peptide commonly 

finds its origin in physical degradation processes such as adsorption [33], aggregation and 

chemical degradation pathways (e.g. hydrolysis, oxidation, deamidation [34]. This process 

is not only specific for small peptide or protein but for peptides in general. The chemical 

instability of peptides poses a problem in their development as active pharmaceutical 

ingredients. Therefore, a better understanding of the underlying mechanisms of instability 

of a certain peptide is essential to design rational strategies that can be investigated during 

the pharmaceutical development process in order to optimize the stability of the peptide in 

the final formulation [9]. 


2 

17

2 

18 

2. peptIde InstabIlIty and the possIble 

causes of degradatIon 

During formulation, processing, storage and transportation of a peptide drug product, 

the peptide is exposed to conditions that can have significant effects on its chemical and 

physical integrity and many peptides have to be transported and stored under a “cold chain” 

regime [35]. There are several degradation pathways peptides may undergo when they are 

formulated as an aqueous solution. The major pathways of peptide degradation are broadly 

divided into: chemical and physical instability. 

Chemical instability can be defined as any process involving modification of the 

peptide by covalent bond formation or covalent bond cleavage, generating new chemical 

entities [36]. Chemical instability pathways include hydrolysis, oxidation, racemization, 

isomerization, deamidation, disulfide exchange and β-elimination [37]. Physical instability 

refers to any changes to the higher order structure (dimerization and further aggregation), 

i.e. changes in noncovalent bonds, not necessarily involving covalent bond modifications. 

This physical instability can result in denaturation which may lead to adsorption to surfaces, 

aggregation and precipitation [9]. In Table 2, reported degradation pathways of various 

peptides are presented. 

2.1 Hydrolytic pathways 

Hydrolysis is one of the potential degradation pathways of peptides. In aqueous 

solution peptide bonds can undergo chemical reactions, usually through an attack of 

an electronegative atom (nucleophile) on the carbonyl carbon, resulting in cleavage of 

peptide bonds. In an alkaline aqueous environment, hydroxyl ions are better nucleophiles 

than polar molecules such as water. Under acidic conditions, the carbonyl group becomes 

protonated, which leads to a much easier nucleophile attack [58]. 

2.1.1 Acid/base catalyzed hydrolysis 

The stability of peptides in aqueous solution strongly depends on pH. It was reported that 

at pH range of 1-3 gonadorelin and triptorelin mostly undergo acid-catalysed hydrolysis 

via deamidation of the C-terminal amide to form free acid deamidated products. In the 

pH range 5−6, the peptide backbones of gonadorelin and triptorelin are hydrolyzed at the 

N-terminal side of the serine residue probably involving nucleophilic addition of the serine 

hydroxyl group to the neighboring amide bond forming a cyclic intermediate resulting 

in fragmentation [47,59]. At pH values over 7, base- catalyzed epimerization is the main 

pathway of degradation. Serine is most likely involved in base-catalyzed epimerization 

through a carbanion intermediate. Its ability to form a relatively stable six membered 

intermediate with a hydrogen bridge explains the relative high rate of racemization of the 

L-serine residue compared to other amino acid residues. Parallel to the epimerization, 

base-catalysed hydrolysis of gonadorelin and triptorelin occurs. Epimerization of serine is 

considered the most important degradation reaction for hydroxyl-catalyzed degradation of 

gonadorelin and triptorelin [47,48,60]. 

Acid/base catalyzed hydrolysis has also been observed in the degradation of somatostatin 

analog octastatin in aqueous formulations. This hydrolysis is also affected by the buffer 

species. Jang et. al. [43] reported that the degradation rate of octastatin was higher in 

phosphate-buffered solution than in glutamate buffers. Increasing buffer concentrations 

resulted in a greater degradation of octastatin in phosphate-containing buffer, presumably

Table 2. Reported degradation pathways of different peptides 

Peptide 

Number of 

amino acid 

residues Degradation pathways Reference 

Thyrotropin Releasing Hormones 3 Hydrolysis [38] 

Ceftazidime 5 Hydrolysis [39,40] 

Eptifibatide 6 

Hydrolysis, isomerization, deamidation, 

oxidation, and dimerization 

[41] 

Octreotide 8 Hydrolysis [42,43] 

Oxytocin 9 

Hydrolysis, deamidation, oxidation, 

beta-elimination, and dimerization 

[31,44] 

Desmopressin 9 

Beta-elimination, deamidation, disulfide 

exchange, racemization, and oxidation 

[45] 

Leuprolide 10 

Hydrolysis, isomerization, oxidation, 

and aggregation 

[46] 

Goserelin 10 Acid-base catalyzed hydrolysis [47] 

Gonadorelin 10 Acid-base catalyzed hydrolysis [47,48] 

Triptorelin 10 Acid-base catalyzed hydrolysis [47,48] 

Somatostatin and analogues 14 Hydrolysis, [42,43] 

Calcitonin 32 

Deamidation, oxidation, and 

aggregation/fibrilation 

[8,49,50] 

Human Brain Natriuretic Peptide 

[hBNP (1–32)] 

32 

Aggregation, 

deamidation, and oxidation 

[51] 

Human Parathyroid Hormone 

[hPTH (1–34) 

34 Aggregation, deamidation, and oxidation [51] 

Enfuvirtide 36 Deamidation and aggregation [52,53] 

Adrenocorticotropin Hormone 

(ACTH) 

39 

Deamidation of Asn residues, and 

racemization 

[54] 

Corticotropin Heleasing factor 41 Oxidation [55] 

Amyloid-β (Aβ) Peptides 36-43 

Metal-catalyzed oxidation, 

dimerization and aggregation 

[56,57] 

because of catalytic effects of phosphate ions. In contrast, in the glutamate-buffered 

solution, increasing buffer concentrations caused greater stabilization, presumably due to 

a stabilizing effect of glutamic acid by both ionic and hydrophobic interactions between 

octastatin and glutamate ions. 

2.1.2 Deamidation of Asn and Gln residues 

Deamidation is regarded as the most common chemical degradation pathway for peptides and 

proteins. Peptides containing asparagine (Asn) and glutamine (Gln) residues are known to 

undergo spontaneous deamidation to form aspartic acid (Asp) and glutamic acid (Glu) residues 

under physiological conditions. Under acidic conditions (pH below 3), deamidation of Asn 

residues is reported to proceed through direct hydrolysis of the Asn side chain amide to form 

Asp. Similarly, Gln residues are converted to Glu by acid catalyzed direct hydrolysis [36]. 

At alkaline and neutral conditions, the deamidation of Asn residues predominantly 

occurs through a cyclic imide intermediate formed by an intermolecular reaction in which 

the carbonyl carbon in the side chain of Asn residue is attacked by the nitrogen in the amino 

acid residue following Asn. Therefore, the rate of deamidation via this pathway depends on 


2 

19

2 

20 

the nature of the amino acid residue on the carboxyl side of the Asn residue [54,61]. At 

the same condition, deamidation of Gln residues is less common than for Asn, because 

cyclization of Asn residues leads to a five-membered ring. With Gln, the corresponding 

intermediate is a six-membered ring, which formation is thermodynamically less favorable 

than the smaller, five-membered, ring [36]. 

The high rate of deamidation is caused by the high degree of peptide chain flexibility. 

The rate of deamidation also depends on the amino acid sequence in the peptides [62]. 

The amino acid residues following asparagine (Asn) such as glycine (Gly), alanine (Ala), 

serine (Ser), and aspartic acid (Asp) may seriously increase the reaction rate in the 

degradation of therapeutic peptides [63]. Catak et al. [64] reported that deamidation 

in aqueous solution can occur either through direct hydrolysis or through succinimidemediation. 

These two reactions are competitive even in the absence of acid or base 

catalysis (Figure 1). 

Adrenocorticotropic hormone (ACTH), a 39-amino acid polypeptide with a single Asn 

residue, was shown to be degraded via deamidation of Asn residues [54,65]. The deamidation 

of Asn or Gln residues was also observed in the degradation of salmon calcitonin [49]. 

The nonapeptide oxytocin is another example of a peptide that can undergo deamidation 

and involves the hydrolysis of Asn 5 and Gln 4 side chain amides. It was also reported that 

Gly-NH 2 of oxytocin undergoes deamidation at a pH of 2. [31]. 

2.1.3 Isomerization of Asp residues 

Deamidation of the Asn residue in neutral conditions proceeds through the formation of 

a cyclic succinimide intermediate by the intermolecular attack of the backbone nitrogen 

atom on the carbonyl group of the Asn side chain [66]. Hydrolysis of the succinimide 

intermediate generates the formation of isoaspartic (isoAsp) and aspartic acid (Asp) 

Figure 1. Deamidation pathways of Asn residue via A. direct hydrolysis and B. succinimide 

mediation [64].

esidues. Furthermore, direct dehydration occurs in a transformation of aspartic acid (Asp) 

into its isoform isoAsp through the same cyclic succinimide intermediate [67]. Additionally, 

L-succinimide may be racemized to D-succinimide to form the D-Asp and D-isoAsp 

enantiomers [61,68]. Thus, the mechanisms for the deamidation and the isomerization 

reactions are similar since they both proceed via an intra-molecular cyclic succinimide 

intermediate. The formation of succinimide intermediate is the rate-limiting step in both 

the deamidation and the isomerization reactions at physiological pH [69]. 

Under acidic conditions, cleavage at the Asp residue is the most important degradation 

pathway of recombinant human parathyroid hormone (rhPTH). While at pH values above 

5 the Asn deamidation is the major degradation route [70]. 

2.2 Oxidation pathways 

Oxidation is another primary chemical degradation pathway that can occur in a peptide. 

Oxidation of organic molecules is defined as an increase in oxygen or a decrease in hydrogen 

content [71]. Alternatively, oxidation can be defined as a reaction that increases the content 

of more electronegative atoms in a molecule, in which the electronegative heteroatoms are 

generally oxygen or halogens.[72]. Any peptide that contains cysteine (Cys), methionine (Met), 

histidine (His), tyrosine (Tyr) and tryptophan (Trp) residues can potentially be damaged due 

to their high reactivity with various reactive oxygen species (ROS). Cys and Met are at risk 

for oxidation because of their sulfur atoms, and His, Tyr and Trp because of their aromatic 

rings [73]. Figure 2 illustrates the oxidation reaction of Met and His. 

Figure 2. Oxidation reaction of A: methionine to methionine sulfoxide in acidic condition 

(HA) and B: histidine through an oxometallacycyclic intermediate to form various 

degradation products, such as 2-oxomidazoline, asparagine, and aspartate (adapted from 

Li, et. al, 1995) 


2 

21

2 

22 

Oxidation can be induced by contaminating oxidants and light exposure. Oxidation 

can be catalyzed by the presence of (traces of) transition metal ions during processing and 

storage. Oxidation of peptides may furthermore be influenced by pH, temperature, and 

buffer composition. [63]. 

2.2.1 Auto-oxidation 

A rapid oxidation reaction of a peptide may occur when the ROS is able to access the side 

chain, particularly for peptides with Met residues autooxidation may be a major degradation 

pathway [36]. Adrenocorticotropic hormone (ACTH) and human atrial natriuretic peptide 

(ANP) are examples of peptides that undergo spontaneous Met-oxidation [74]. 

Cysteine residues may also undergo spontaneous oxidation to form the molecular 

byproducts sulfinic acid and cysteic acid in the presence of metal ions or nearby thiol 

groups. Cysteine oxidation involves a nucleophilic attack of thiolate ions on disulfide bonds, 

generating new disulfide bonds and other thiolate ions. The newly formed thiolate ions can 

subsequently react with another disulfide bond to form cysteine [75]. 

2.2.2 Metal ion-catalyzed oxidation 

Metal-catalyzed oxidation occurs when a redox active metal ion binds to the peptide. The 

amino acid residues that are most susceptible to metal ion catalyzed oxidations are His, Arg, 

Lys, Pro, Met, and Cys. Of these amino acids, His and Cys are sensitive to oxidative damage, 

as the ROS generated at the metal center does not have to diffuse very far before reacting 

with the peptide [36,76]. Several metals, such as iron and copper have a strong catalytic 

power to generate highly reactive hydroxyl radicals if reacted with hydrogen peroxide 

(Fenton reaction). The hydroxyl radicals may react with His residue to form 2-oxo-His[77]. 

Metal ion-catalysis has been observed in the oxidation of amyloid-β (Aβ) peptides which 

mediates the neurotoxicity of Alzheimer’s disease. The amino acid residue involved in this 

pathway is His [56]. 

2.2.3 Light-induced oxidation 

In 2007, Kerwin and Remmel [78] summarized light-induced degradation in 

biopharmaceuticals. These reactions may occur at several points from production to 

delivery of the products. Light-induced oxidation is initiated when a compound absorbs 

a certain wavelength of light, which provides energy to raise the molecule to an excited 

state. The excited molecule can then transfer that energy to molecular oxygen, converting 

it to reactive singlet oxygen atoms. This is how tryptophan, histidine, and tyrosine can 

be modified under light in the presence of oxygen [37,79]. Tyrosine photo-oxidation can 

produce mono-, di-, tri-, and tetrahydroxyl tyrosine as byproducts [80]. Aggregation is 

observed in some peptides due to cross-linking between oxidized tyrosine residues [81]. 

2.3 β-elimination reactions 

Disulfide-bridge peptides might undergo destruction to form free thiol groups through 

β-elimination. This reaction is commonly observed when peptides are incubated at 

higher pH values or elevated temperature [82,83]. At neutral and alkaline pH levels 

peptides can undergo a disulfide exchange reaction which is catalyzed by thiol groups 

[84]. When cystine-containing proteins are heated at 100°C, even at a neutral pH, they 

undergo destruction to form free thiols [85]. Salmon calcitonin (sCT) degrades via 

β-elimination at a disulfide bridge between the cysteine residues at positions 1 and 7. The

insertion of an extra sulfur to form a trisulfide bridge has also been reported as a result 

of a β-elimination reaction [8]. 

2.4 Disulfide exchange reactions 

Disulfide exchange reactions contribute to the formation of dimers and larger aggregates. 

These reactions may occur at the cysteine–cysteine link. In a study on the degradation of 

sCT, dimeric products were found to be formed via disulfide exchange reactions, however, 

the disulfide-linked dimers may undergo further disulfide exchange reactions that will 

eventually regenerate sCT monomers [8]. Disulfide interchange at aqueous acid conditions 

proceeds through sulfenium ions, which arise from the acid hydrolysis of the disulfide bond 

[86]. Several studies on the disulfide exchange reaction and the importance of disulfidebridges 

for the stability of peptides have been reported [8,82,86,87]. 

2.5 Dimerization, aggregation, and precipitation 

The Cys residues in the oxytocin molecule are able to form disulfide bonds due to thiol 

exchange between two oxytocin molecules [31,44]. Thiol exchange occurs at neutral and 

alkaline conditions via a nucleophilic attack of free thiolate on one of the sulfur atoms within 

the disulfide bridge. At lower pH values the disulfide exchange progresses via a sulfenium 

cation, formed following protonation of the disulfide bridge [36,86]. In either case, disulfide 

exchange can result in dimerization and progressive aggregation [31]. As mentioned above, 

peptide molecules are also able to form dimers due to light or metal induced oxidation 

of tyrosine residues to form dityrosine-linked dimers [31,88,89]. Aggregation may be 

induced by several stress condition, such as heating, freezing or agitation. Aggregates can 

form either via covalent bonds, such as disulfide bonds, ester, or amide linkage or noncovalent 

bonds occurring via hydrophobic interaction or charge-charge complexation. The 

relative weakness of non-covalent protein bonds can lead to aggregate disruption during 

the analytical process [90]. 

There is no single pathway by which peptides can form aggregates [36]. Both mechanisms 

can occur simultaneously leading to the formation of either soluble or insoluble aggregates 

[91]. Furthermore higher concentrations of the peptide in solution will bring aggregates more 

easily to a later stage of physical instability i.e. precipitation [12]. Calcitonin, β-amyloid peptide, 

leuprolide, and deterelix have been reported to form gel-like aggregates. The gel formation is a 

function of peptide concentration, temperature, time, pH, and agitation. However the chemical 

stability and dimerization of leuprolide was not affected by gelation [92]. 

3. strategIes to Improve peptIde stabIlIty 

In aqueous formulatIons 

Many efforts have been made to improve the stability of peptides in aqueous solution. 

Examples are: the use of buffers, metal ions or organic solvents, and oxygen removal. However, 

an optimal use of these strategies requires first of all knowledge of the peptide´s structure and 

thorough understanding of the possible degradation pathways of the specific peptide. 

Examining the amino acid sequence of a peptide can provide an insight into how the 

molecule may degrade. This will be of real use for peptides of which the amino acid sequence 

is generally exposed on the aqueous environment and therefore able to undergo chemical 


2 

23

2 

24 

degradation. The degradation pathways and involved amino acid sequences as well as the 

formulation strategies to inhibit the specific degradation pathway are summarized in Table 3. 

3.1 Optimizing hydrolytic stability 

3.1.1 pH optimization and buffer species 

Stability of peptides depends on the pH, therefore the common strategy to avoid instability 

of peptides in aqueous solution is adjusting the pH using buffers. In table 1 the pH values 

and the pH adjusting agent or buffer of several marketed products are given. Considering 

modern formulation development strategies it may be assumed that these values represent 

a value which is (close to) the optimum pH stability of the different peptides. The acceptable 

pH range for (slow) intravenous administration is 3-10.5 and 4-9 for other parenteral routes 

to minimize discomfort at the injection site [93,94]. Therefore, it is important to study the 

pH stability of a peptide in the range of pH 3-10 with different buffers in the early stages 

of formulation development [95,96]. Hydrolysis of the side chain amide on glutamine and 

asparagine residues could be reduced by formulating at a pH below neutral. However, 

the pH should not be lower than 3 to prevent direct hydrolysis of the Asn and Gln side 

chains and to minimize hydrolytic fragmentation [36]. Oxytocin was described to have the 

highest stability at pH 4.5 [31]. Calcitonin undergoes hydrolysis at basic pH, but no such 

degradation is observed at a pH of 7 even at room temperature [97]. Other mechanisms 

for stabilization by buffers have also been reported: In some cases they can act as radical 

scavengers [98]. Even more important is the fact that some buffers are able to bind directly 

to peptides, thereby increasing their conformational stability [36]. Citric acid has been 

reported to bind with oxytocin forming N-cytril oxytocin which increases the stability of 

oxytocin in the presence of divalent metal ions [44]. 

The pH of the formulation can also substantially affect deamidation. Deamidation 

is known to be sensitive to both the pH and the composition and concentration of the 

buffer. In general, formulations with a pH in the range from 3 to 5 minimize peptide 

deamidation [65,98,105]. Buffer species can also affect deamidation. It has been reported 

that the rate of deamidation is faster in phosphate and bicarbonate buffers than in acetate 

and pyruvate buffers [96]. Isomerization is greatest at low pH values for asparagine and 

glutamine residues [106]. 

3.1.2 Ionic strength 

The total concentration of dissolved electrolytes might affect the rate of hydrolysis. The 

increase in ionic strength could have a stabilizing or destabilizing effect on a peptide, 

depending on the nature of the charge–charge interactions within the peptide [107-109]. 

However, in another study, no significant effect of the ionic strength on the rate of 

deamidation or hydrolysis in small peptides was found [110]. 

3.1.3 Cosolvent 

Cosolvents such as low molecular weight polyethylene glycol (PEG) were shown to reduce 

the aggregation of several peptides [9,12,111]. Peptide deamidation can be modestly 

inhibited in an aqueous solution by the addition of glycerol [100], or ethanol [112]. Addition 

of organic solvents decreases the dielectric constant of an aqueous solution. Reduction 

of the solvent´s dielectric strength leads to significantly lower rates of isomerization and 

deamidation [69]. The use of polyols includes polyhydric alcohols and carbohydrates

Table 3. Degradation pathways and possible stabilization strategies for peptides including amino acid 

residues involved in the degradation. 

Degradation pathway Stabilization strategy 

Chemical Instability 

Acid/base catalyzed hydrolysis pH, buffer species, co-solvents 

Amino acid 

residue(s) involved Ref. 

Ser Trp 

Asn-Pro 

Asn-Tyr 

[47,60] 

[42] 

Deamidation pH 3-5, increased solvent viscosity Asn, Gln 

[62,69] 

[99,100] 

Isomerization pH 6

2 

26 

adjusting the primary and secondary packaging to protect from light, and the use of antioxidants 

in the formulation. Watermann and co-workers have made a guideline on the use of excipients to 

optimize oxidative stability including their recommended concentrations [101]. 

3.2.1 Buffers 

The side chains of tryptophan, methionine and cysteine and, to a lesser extent, tyrosine 

and histidine are potential oxidation sites. The modification of indole group of Trp, thiol 

group of Cys, imidazole group of His, and phenolic side chain of Tyr by the reactive oxygen 

species is most significant at neutral and alkaline condition. Therefore, a lower pH reduces 

oxidation of peptides containing these amino acids [114]. In contrast, the thioether group 

of methionine can be readily oxidized by certain reagents under acidic conditions [58]. 

3.2.2 Air exclusion 

Removal of oxygen by bubbling another gas, for example nitrogen, through the solution 

may be effective to exclude oxygen. For some oxidation-sensitive peptide drugs, processing 

and filling steps should be carried out in the presence of an inert gas such as nitrogen, argon 

or helium. To further minimize air oxidation violent stirring should be avoided [115]. 

3.2.3 Antioxidants 

Antioxidants are commonly used to protect peptides from oxidation during processing 

and formulation. Some antioxidants, however, can be problematic in peptide 

formulations [37,114]. For example, bisulfite is not a suitable agent because it is a strong 

nucleophile, which may interact with peptides. Furthermore, addition of antioxidants to 

trace metal ions contaminated peptide solutions will not protect the peptide from oxidative 

modification. In contrast, it can even accelerate the oxidation process, as demonstrated by 

the use of ascorbic acid which promotes rather than inhibits oxidation of the Met residue of 

small model peptides in the presence of metal ions [116,117]. Methionine, a sulfur containing 

amino acid, is readily oxidized to methionine sulfoxide by many reactive oxygen species and 

oxidizes more easily than the peptide, and therefore can act as a sacrificial antioxidant [118]. 

3.2.4 Chelating agents 

Another method to reduce free radical oxidation is to include chelating agents. Chelating 

agents are used to inhibit oxidation by complexation of trace metal ions. The most commonly 

used chelating agents in pharmaceutical formulations are ethylenediaminetetraacetic acid 

(EDTA), desferal, diethylenetriaminepentaacetic acid (DTPA), inositol hexaphosphate, 

ethylenediamine bis(o-hydroxyphenylacetic acid), tris(hydroxymethyl)aminomethane 

(TRIS), citric acid, and tartaric acid. EDTA has been recommended as a chelating agent for 

copper ions, whereas desferal, DTPA, inositol hexaphosphate and ethylenediamine bis(ohydroxy-phenylacetic 

acid) were recommended for iron ions [113]. It is well known that 

chelating agents are generally effective in stabilizing peptides against oxidation. However, 

it cannot be assumed that addition of a certain chelating agents will be able to interact 

with all trace metal ions and completely eliminate oxidation. Under certain circumstances, 

chelating agents may even accelerate the oxidation process [37,119]. 

3.2.5 Polyols 

Polyols such as mannitol, trehalose, sucrose, maltose, and raffinose can prevent oxidation 

of therapeutic peptides. For example, mannitol was shown to inhibit the iron-catalyzed 

oxidation of Met-containing peptides [102]. Sucrose has been shown to decrease the

ate of oxidation of parathyroid hormone hPTH (1–34) and brain natriuretic hormones 

hBNP (1–32) in liquid formulations [51]. 

3.3 Protection against disulfide exchange reaction 

Formulating octreotide in glycine with pharmaceutically acceptable salts and HCl was 

found to be effective in protecting its disulfide-bridge from cleavage, and was also reported 

to be better tolerated than the formulation in acetate, lactate or bicarbonate [120]. Divalent 

metal ions in combination with specific buffers may protect peptide drugs against disulfide 

exchange reaction. Recently, it was shown that Zn 2+ , Ca 2+ and Mg 2+ ions in combination 

with dicarboxylic and tricarboxylic acids improved the stability of oxytocin [44,104]. 

The addition of at least 2 mM CaCl 2 , MgCl 2 , or ZnCl 2 to 5 and 10 mM citrate-buffered 

solutions at pH 4.5 increased oxytocin stability in aqueous formulations and the stability is 

further increased with increasing concentration of the divalent metals ions up to 50 mM. 

The improved stability is due to the reactivity of carboxyl ions in citrate buffer which can 

attack the N-terminal group of Cys to form an adduct and together with divalent metal ions 

suppress the disulfide exchange reaction [44,104]. 

3.4 Inhibition of dimerization, aggregation, and precipitation 

Dimerization and aggregation may involve the interchange of covalent bonds, such as 

disulfide bridges or non-covalent forces such as hydrophobic interactions. Both soluble 

and insoluble aggregates may occur. A peptide in aqueous solution can be stabilized against 

dimerization and aggregation by optimizing the pH and ionic strength of the solution. 

Furthermore, dimerization can be prevented by preferential exclusion using sugars, amino 

acids, and/or polyols, and by using surfactants [36]. The use of a combination of divalent 

metal ions and citrate buffer has been found to inhibit the cysteine-mediated dimerization of 

oxytocin [44]. PEG has been shown to reduce the aggregation of several peptides [9,12,111]. 

The stabilizing effects increased with increasing concentration and increasing molecular 

weight [99,100]. Dicarboxylic amino acids such as aspartic and glutamic acid have been 

used to reduce aggregation [121] and glycine, arginine and lysine have also been reported to 

prevent aggregation [121-123]. Polysorbate 20 and 80 [124,125] can reduce agitation-induced 

aggregation of peptides. However, surfactants are reported to be less effective in reducing 

thermally-induced aggregation. Therefore, the optimum concentration required to protect a 

specific peptide should be evaluated for each different type of stress that may be encountered. 

3.5 Preferential exclusion 

Another strategy to minimize peptide degradation is the use of extremolytes. 

Extremolytes are small organic molecules produced by extremophilic microorganisms, 

to protect their biological macromolecules from damage by external stresses [126]. 

Examples of extremolytes are the polyol derivatives ectoin and hydroxyectoin [127], 

betain [128], trehalose [129], amino acids (e.g. proline), and the mannose derivative 

mannosylglycerate [130]. Mannosylglycerate was reported to have the ability to inhibit 

β-amyloid peptide aggregation [131]. Extremolytes are known to stabilize peptides by 

forming solute hydrate clusters that are preferentially excluded from the hydrate shell of 

the peptide as a result of the repulsive interactions between extremolytes and the backbone 

of the peptides. Accumulation of water near peptide domains assembles the peptide into 

a more compact structure with a reduced surface area [132-134]. However, there are no 


2 

27

2 

28 

extremolytes that can be used as a universal stabilizer for all peptides in aqueous solution. 

In addition, it should be realized that extremolytes may also act as destabilizer for certain 

peptides under specific conditions [135]. 

Surfactants have also been described as stabilizers for peptides. Pluronic® 

F68 has 

been used to enhance stability of peptide drugs ceftazidime in parenteral solutions. In this 

approach, the peptides are kept in a microenvironment that shields them from exterior 

conditions, limiting the effect of moisture and pH [40]. 

5. conclusIon 

In aqueous solutions peptides are often unstable. Peptides have unique structures that differ 

from proteins in that they are smaller and rarely have a tendency to physical degradation e.g. 

unfolding. Peptides often do not possess a higher order structure that can sequester reactive 

groups, the side chain of nearly all of the amino acid residues are fully solvent exposed, 

allowing maximal contact with solvents and the degradation rates appear to correlate 

with the degree of solvent exposure. Based on knowledge of the peptide’s structure and an 

understanding of the predominant degradation pathways, the strategies may be developed 

to achieve adequate stability of the formulation. The degradation pathways of peptides are 

mainly dependent on the amino acid sequence. The most prominent degradation pathways 

for peptides are hydrolysis, oxidation, and dimerization. Formulating peptides in a specific 

pH with a specific buffer, avoiding oxygen reactive species, and minimizing solvent exposure 

eliminate chemical degradation. Increasing solution viscosity by using sugars or polymers 

reduces peptide mobility and further decelerates physical degradation. 

references 

1. V. du Vigneaud, C. Ressler, S. Trippett, The 

sequence of amino acids in oxytocin, with a 

proposal for the structure of oxytocin, J. Biol. 

Chem. 205 (1953) 949-57. 

2. J. Sterling, GEN’s 30th Anniversary: Peptide 

Therapeutics, Editor’s Note, Genetic 

Engineering and Biotechnology News 31 

(2011). 

3. L. Malavolta, F.R. Cabral, Peptides: 

Important tools for the treatment of central 

nervous system disorders, Neuropeptides 45 

(2011) 309-16. 

4. V.H.L. Lee (Ed.), Peptide and Protein Drug 

Delivery, Marcel Dekker Inc., New York, 

1991. 

5. S.Y. Lin, S.L. Wang, Advances in simultaneous 

DSC-FTIR microspectroscopy for rapid 

solid-state chemical stability studies: Some 

dipeptide drugs as examples, Adv. Drug 

Deliv. Rev. 64 (2012) 461-78. 

6. T. Wieland, M. Bodanszky (Eds.), The Worlds 

of Peptides: A Brief History of Peptide 

Chemistry, Springer-Verlag, Berlin, 1991. 

7. J. Pless, The history of somatostatin analogs, 

J. Endocrinol. Invest. 28 (2005) 1-4. 

8. V. Windisch, F. DeLuccia, L. Duhau, 

F. Herman, J.J. Mencel, S.Y. Tang, M. 

Vuilhorgne, Degradation pathways of 

salmon calcitonin in aqueous solution, J. 

Pharm. Sci. 86 (1997) 359-64. 

9. S. Frokjaer, L. Hovgaard (Eds.), 

Pharmaceutical Formulation Development 

of Peptides and Proteins, CRC Press, United 

States America, 2000. 

10. P.A. Zhou XH, Peptide and protein drugs. 

Therapeutic applications, absorption and 

parenteral administration, Int J Pharm 75 

(1991) 97-115. 

11. A.M. Papini, Peptide Chemistry Revolution, 

Chemistry Today 29 (2011). 

12. M.F. Powell, L.M. Sanders, A. Rogerson, V. Si, 

Parenteral peptide formulations: chemical 

and physical properties of native luteinizing 

hormone-releasing hormone (LHRH) and 

hydrophobic analogues in aqueous solution, 

Pharm. Res. 8 (1991) 1258-63. 

13. F. Veuillez, Y.N. Kalia, Y. Jacques, J. 

Deshusses, P. Buri, Factors and strategies for 

improving buccal absorption of peptides, 

Eur. J. Pharm. Biopharm. 51 (2001) 93-109.

14. D.H. Oh, K.H. Chun, S.O. Jeon, J.W. Kang, 

S. Lee, Enhanced transbuccal salmon 

calcitonin (sCT) delivery: Effect of chemical 

enhancers and electrical assistance on in 

vitro sCT buccal permeation, Eur. J. Pharm. 

Biopharm. 79 (2011) 357-63. 

15. K.V. Patel, N.D. Patel, H.D. Dodiya, P.K. 

Shelat, Buccal Bioadhesive Drug Delivery 

System: An Overview, Int. J. Pharm. Biol. 

Arch. 2 (2011) 600-9. 

16. A. Nasr, A.Y. Shahin, A.M. Elsamman, 

M.S. Zakherah, O.M. Shaaban, Rectal 

misoprostol versus intravenous oxytocin for 

prevention of postpartum hemorrhage, Int. 

J. Gynaecol. Obstet. 105 (2009) 244-7. 

17. K. Morimoto, H. Akatsuchi, R. Aikawa, M. 

Morishita, K. Morisaka, Enhanced rectal 

absorption of [Asu’.‘]-eel calcitonin in rats 

using polyacrylic acid aqueous gel base, J. 

Pharm. Sci. 73 (1984) 1366-8. 

18. A.P. Sayani, Y.W. Chien, Systemic delivery 

of peptides and proteins across absorptive 

mucosae, Rev. Ther. Drug Carrier Syst. 13 

(1996) 85-184. 

19. J.L. Richardson, L. Illum, (D) Routes of 

delivery: Case studies: (8) The vaginal route 

of peptide and protein drug delivery, Adv. 

Drug Deliv. Rev. 8 (1992) 341-66. 

20. N. Seo, M. Takigawa, The current status 

and future direction of percutaneous 

peptide immunization against melanoma, J. 

Dermatol. Sci. 48 (2007) 77-85. 

21. S.P. Read, S.M. Cashman, R. Kumar-Singh, 

A poly(ethylene) glycolylated peptide 

for ocular delivery compacts DNA into 

nanoparticles for gene delivery to postmitotic 

tissues in vivo, J. Gene Med. 12 

(2010) 86-96. 

22. D. Harris, J.H. Liaw, J.R. Robinson, (D) 

Routes of Delivery: Case Studies (7) Ocular 

delivery of peptide and protein drugs, Adv. 

Drug Deliv. Rev. 8 (1992) 331-9. 

23. H.E. Bodde, J.C. Verhoef, M. Ponec, 

Transdermal peptide delivery, Biochem. 

Soc. Trans. 17 (1989) 943-5. 

24. M.R. Prausnitz, A peptide chaperone 

for transdermal drug delivery, Nature 

Biotechnology 24 (2006) 416-7. 

25. A.E. Pontiroli, Peptide hormones: Review of 

current and emerging uses by nasal delivery, 

Adv Drug Deliv Rev 29 (1998) 81-7. 

26. S. Chhajed, S. Sangale, S.D. Barhate, 

Advantageous Nasal Drug Delivery System: 

A Review, IJPSR 2 (2011) 1322-36. 

27. S. Turker, E. Onur, Y. Ozer, Nasal route and 

drug delivery systems, Pharm. World Sci. 26 

(2004) 137-42. 

28. S.A. Shoyele, A. Slowey, Prospects of 

formulating proteins/peptides as aerosols 

for pulmonary drug delivery, Int. J. Pharm. 

314 (2006) 1-8. 

29. R.I. Henkin, Inhaled insulinintrapulmonary, 

intranasal, and other routes 

of administration: mechanisms of action, 

Nutrition 26 (2010) 33-9. 

30. X.H. Zhou, A. Li Wan Po, Peptide and 

protein drugs: II. Non-parenteral routes of 

delivery, Int. J. Pharm. 75 (1991) 117-30. 






(2009) 1679-88. 

32. International Conference on Harmonisation; 

Stability Data Package for Registration 

Applications in Climatic Zones III and IV; 

Stability Testing of New Drug Substances 

and Products; availability. Notice, Fed. 

Regist. 68 (2003) 65717-8. 

33. S.T. Anik, J.Y. Hwang, Adsorption of D- 

Nal(2)6LHRH, a decapeptide, onto glass 

and other surface., Int. J. Pharm. 16 (1983) 

181-190. 

34. V.G. Badelin, O.V. Kulikov, V.S. Vatagin, 

Physico-chemical properties of peptides and 

their solutions, Thermochimica Acta 169 

(1990) 81-93. 

35. M. Zahn, P.W. Kallberg, G.M. Slappendel, 

H.M. Smeenge, A risk-based approach to 

establish stability testing conditions for 

tropical countries, J. Pharm. Sci. 95 (2006) 

946-65. 

36. M.C. Manning, D.K. Chou, B.M. Murphy, 

R.W. Payne, D.S. Katayama, Stability of 

protein pharmaceuticals: an update, Pharm. 

Res. 27 (2010) 544-75. 

37. S. Li, C. Schoneich, R.T. Borchardt, Chemical 

instability of protein pharmaceuticals: 

Mechanisms of oxidation and strategies for 

stabilization, Biotechnol. Bioeng. 48 (1995) 

490-500. 

38. T. Hashimoto, K. Ohki, N. Sakura, Hydrolytic 

cleavage of pyroglutamyl-peptide bond. I. 

The susceptibility of pyroglutamyl-peptide 

bond to dilute hydrochloric acid, Chem. 

Pharm. Bull. (Tokyo) 43 (1995) 2068-74. 

39. J.M. Hwang, T.E. Piccinini, C.J. Lammel, 

W.K. Hadley, G.F. Brooks, Effect of storage 

temperature and pH on the stability of 

antimicrobial agents in MIC trays, J. Clin. 

Microbiol. 23 (1986) 959-61. 

40. C.A.D. Santos, G.B. Ribeiro, M.C. Knirsch, 

A.P. Junior, T.C.V. Penna, Influence of 

Pluronic® F68 on ceftazidime biological 

activity in parenteral solutions, J. Pharm. 

Sci. 100 (2011) 715-20. 


2 

29

2 

30 

41. L. Zhao, S.H. Yalkowsky, Stabilization of 

eptifibatide by cosolvents, Int. J. Pharm. 218 

(2001) 43-56. 

42. R. Krishnamoorthy, A.K. Mitra, Kinetics 

and mechanism of degradation of a cyclic 

hexapeptide (somatostatin analogue) in 

aqueous solution, Pharm. Res. 10 (1992) 

1314-20. 

43. S.W. Jang, B.H. Woo, J.T. Lee, S.C. Moon, 

K.C. Lee, P.P. DeLuca, Stability of Octastatin, 

a somatostatin analog cyclic octapeptide, in 

aqueous solution, Pharm. Dev. Technol. 2 

(1997) 409-14. 



Hinrichs, A new strategy to stabilize oxytocin 

in aqueous solutions: II. Suppression of 

cysteine-mediated intermolecular reactions 

by a combination of divalent metal ions and 

citrate, Mol. Pharmaceutics 9 (2012) 554-62. 

45. S.L. Law, K.J. Huang, V.H. Chou, Stability 

of desmopressin loaded in liposomes, J. 

Liposome Res. 13 (2003) 269-77. 

46. S.C. Hall, M.M. Tan, J.J. Leonard, 

C.L. Stevenson, Characterization and 

comparison of leuprolide degradation 

profiles in water and dimethyl sulfoxide, J. 

Pept. Res. 53 (1999) 432-41. 

47. M.A. Hoitink, J.H. Beijnen, M.U.S. Boschma, 

A. Bult, E. Hop, J. Nijholt, C. Versluis, 

Wiese,G.,Underberg,W.J.M., Identification 

of the degradation products of gonadorelin 

and three analogues in aqueous solution, 

Anal. Chem. 69 (1997) 4972-8. 

48. V.J. Helm, B.W. Muller, Stability of 

gonadorelin and triptorelin in aqueous 

solution, Pharm. Res. 7 (1990) 1253-6. 

49. K.C. Lee, Y.J. Lee, H.M. Song, C.J. Chun, P.P. 

DeLuca, Degradation of synthetic salmon 

calcitonin in aqueous solution, Pharm. Res. 

9 (1992) 1521-3. 

50. S. Seyferth, G. Lee, Structural studies of 

EDTA-induced fibrillation of salmon 

calcitonin, Pharm. Res. 20 (2003) 73-80. 

51. M. Kamberi, Y.J. Kim, B. Jun, C.M. Riley, 

The effects of sucrose on stability of human 

brain natriuretic peptide [hBNP (1-32)] and 

human parathyroid hormone [hPTH (1- 

34)], J. Pept. Res. 66 (2005) 348-56. 

52. H. Stocker, C. Kloft, N. Plock, A. Breske, G. 

Kruse, C. Herzmann, H. Schulbin, P. Kreckel, 

C. Weber, F. Goebel, J. Roeling, S. Staszewski, 

A. Plettenberg, C. Moecklinghoff, K. 

Arasteh, M. Kurowski, Pharmacokinetics 

of enfuvirtide in patients treated in typical 

routine clinical settings, Antimicrob. Agents 

Chemother. 50 (2006) 667-73. 

53. R.W. Payne, R. Nayar, R. Tarantino, S. Del 

Terzo, J. Moschera, J. Di, D. Heilman, B. Bray, 

M.C. Manning, C.S. Henry, Second virial 

coefficient determination of a therapeutic 

peptide by self-interaction chromatography, 

Biopolymers 84 (2006) 527-33. 

54. N.P. Bhatt, K. Patel, R.T. Borchardt, 

Chemical pathways of peptide degradation. 

I. Deamidation of adrenocorticotropic 

hormone, Pharm. Res. 7 (1990) 593-9. 

55. I.Q. Assil, M.E. Shomali, A.B. Abou-Samra, 

An oxidation resistant radioligand for 

corticotropin-releasing factor receptors, 

Peptides 22 (2001) 1055-61. 

56. K. Inoue, A. Nakagawa, T. Hino, H. Oka, 

Screening assay for metal-catalyzed 

oxidation inhibitors using liquid 

chromatography-mass spectrometry with 

an N-terminal beta-amyloid peptide, Anal. 

Chem. 81 (2009) 1819-25. 

57. D.P. Smith, G.D. Ciccotosto, D.J. Tew, M.T. 

Fodero-Tavoletti, T. Johanssen, C.L. Masters, 

K.J. Barnham, R. Cappai, Concentration 

dependent Cu2+ induced aggregation and 

dityrosine formation of the Alzheimer’s 

disease amyloid-beta peptide, Biochemistry 

46 (2007) 2881-91. 

58. M.C. Manning, K. Patel, R.T. Borchardt, 

Stability of protein pharmaceuticals, Pharm 

Res 6 (1989) 903-18. 

59. R.G. Strickley, M. Brandl, K.W. Chan, K. 

Straub, L. Gu, High-performance liquid 

chromatographic (HPLC) and HPLCmass 

spectrometric (MS) analysis of the 

degradation of the luteinizing hormonereleasing 

hormone (LH-RH) antagonist RS- 

26306 in aqueous solution., Pharm. Res. 7 

(1990) 530-6. 

60. M.A. Hoitink, J.H. Beijnen, A. Bult, 

O.A.G.J. Van der Houwen, J. Nijholt, 

W.J.M. Underberg, Degradation kinetics of 

gonadorelin in aqueous solution, J. Pharm. 

Sci. 85 (1996) 1053-59. 

61. T. Geiger, S. Clarke, Deamidation, 

isomerization, and racemization at 

asparaginyl and aspartyl residues in 

peptides: Succinimide-linked reactions that 

contribute to protein degradation, J. Biol. 

Chem. 262 (1987) 785-94. 

62. K. Patel, R.T. Borchardt, Chemical pathways 

of peptide degradation. III. Effect of primary 

sequence on the pathways of deamidation 

of asparaginyl residues in hexapeptides, 

Pharm. Res. 7 (1990) 787-93. 

63. R.W. Payne, M.C. Manning, Peptide 

formulation: challenges and strategies, 

Innov. Pharm. Technol. 28 (2009) 64-8. 

64. S. Catak, G. Monard, V. Aviyente, M.F. Ruiz- 

Lopez, Deamidation of asparagine residues: 

direct hydrolysis versus succinimide-

mediated deamidation mechanisms, J Phys 

Chem A 113 (2009) 1111-20. 

65. K. Patel, R.T. Borchardt, Chemical pathways 

of peptide degradation. II. Kinetics of 

deamidation of an asparaginyl residue in a 

model hexapeptide, Pharm. Res. 7 (1990) 

703-11. 

66. H. Yang, R.A. Zubarev, Mass spectrometric 

analysis of asparagine deamidation and 

aspartate isomerization in polypeptides, 

Electrophoresis 31 (2010) 1764-72. 

67. N.E. Robinson, Protein deamidation, Procl. 

Natl. Acad. Sci. 99 (2002) 5283-8. 

68. N.P. Sargaeva, A.A. Goloborodko, 

P.B. O’Connor, E. Moskovets, M.V. 

Gorshkov, Sequence-specific predictive 

chromatography to assist mass spectrometric 


aspartate isomerization in peptides, 


69. A.A. Wakankar, R.T. Borchardt, Formulation 

considerations for proteins susceptible 

to asparagine deamidation and aspartate 

isomerization, J. Pharm. Sci. 95 (2006) 2321- 

36. 

70. Y. Nabuchi, E. Fujiwara, H. Kuboniwa, Y. 

Asoh, H. Ushio, The stability and degradation 

pathway of recombinant human parathyroid 

hormone: deamidation of asparaginyl 

residue and peptide bond cleavage at 

aspartyl and asparaginyl residues, Pharm. 

Res. 14 (1997) 1685-90. 

71. T.W.G. Solomons (Ed.), Organic Chemistry, 

5th ed., John Wiley and Sons, New York, 

1992. 

72. [72] J. March (Ed.), Advanced Organic 

Chemistry, 4th ed., John Wiley and Sons, 

New York, 1992. 


pathways of peptide degradation. VIII. 

Oxidation of methionine in small model 

peptides by prooxidant/transition metal ion 

systems: influence of selective scavengers for 

reactive oxygen intermediates, Pharm. Res. 

3 (1995) 348-55. 

74. Z. Guan, N.A. Yates, R. Bakhtiar, Detection 

and characterization of methionine 

oxidation in peptides by collisioninduced 

dissociation and electron capture 

dissociation, J. Am. Soc. Mass Spectrom. 14 

(2003) 605-13. 

75. J. Patel, R. Kothari, R. Tunga, N.M. Ritter, 

B.S. Tunga, Stability considerations for 

biopharmaceuticals. Part 1 Overview of 

protein and peptide degradation pathways 

, Bioprocess International (2011) 20-31. 

76. M. Khossravi, R.T. Borchardt, Chemical 

pathways of peptide degradation. X: Effect 

of metal-catalyzed oxidation on the solution 

structure of a histidine-containing peptide 

fragment of human relaxin, Pharm. Res. 17 

(2000) 851-8. 

77. M. Khossravi, R.T. Borchardt, Chemical 

pathways of peptide degradation: IX. Metalcatalyzed 

oxidation of histidine in model 

peptides, Pharm. Res. 15 (1998) 1096-1102. 

78. B.A. Kerwin, R.L. Remmele Jr, Protect 

from light: photodegradation and protein 

biologics, J. Pharm. Sci. 96 (2007) 1468- 

1479. 

79. A.J. Grosvenor, J.D.D. Morton J.M., 

Profiling of residue-level photooxidative 

damage in peptides 

, Amino Acids 39 (2010) 285-96. 

80. J. Zhang, D.S. Kalonia, The effect of 

neighboring amino acid residues and 

solution environment on the oxidative 

stability of tyrosine in small peptides, AAPS 

PharmSciTech 8 (2007) 176-83. 

81. N. Rathore, R.S. Rajan, Current perspectives 

on stability of protein drug products during 

formulation, fill and finish operations, 

Biotechnol. Prog. 24 (2008) 504-14. 

82. A.K. Galande, J.O. Trent, A.F. Spatola, 

Understanding base-assisted desulfurization 

using a variety of disulfide-bridged peptides, 

Peptide Science 71 (2003) 534-51. 

83. S.L. Cohen, C. Price, J. Vlasak, Betaelimination 

and peptide bond hydrolysis: 

two distinct mechanisms of human IgG1 

hinge fragmentation upon storage, J. Am. 

Chem. Soc. 129 (2007) 6976-7. 

84. A.P. Ryle, F. Sanger, Disulphide interchange 

reactions, Biochem. J. 60 (1955) 535. 

85. D.B. Volkin, A.M. Klibanov, Thermal 

Destruction Processes in Proteins Involving 

Cystine Residues, J. Biol. Chem. 262 (1987) 

2945-50. 

86. R.E. Benesch, R. Benesch, The mechanism of 

disulfide interchange in acid solution; role of 

sulfenium ions, J. Am. Chem. Soc. 80 (1958) 

1666-9. 

87. K. Wakabayashi, H. Nakagawa, A. Tamura, 

S. Koshiba, K. Hoshijima, M. Komada, T. 

Ishikawa, Intramolecular disulfide bond is a 

critical check point determining degradative 

fates of ATP-binding cassette (ABC) 

transporter ABCG2 protein, J. Biol. Chem. 

282 (2007) 27841-6. 

88. D.A. Malencik, S.R. Anderson, Dityrosine 

formation in calmodulin: conditions for 

intermolecular cross-linking, Biochemistry 

33 (1994) 13363-72. 

89. I.C. Smith, R. Deslauriers, H. Saito, R. 

Walter, C. Garrigou-Lagrange, H. McGregor, 

D. Sarantakis, Carbon-13 NMR studies of 


2 

31

2 

32 

peptide hormones and their components, 

Ann. N Y Acad. Sci. 222 (1973) 597-627. 

90. M. Kamberi, P. Chung, R. DeVas, L. Li, Z. 

Li, X.S. Ma, S. Fields, C.M. Riley, Analysis of 

non-covalent aggregation of synthetic hPTH 

(1-34) by size-exclusion chromatography 

and the importance of suppression of 

non-specific interactions for a precise 

quantitation, J. Chromatogr. B. Analyt 

Technol. Biomed. Life. Sci. 810 (2004) 151-5. 

91. W. Wang, Protein aggregation and its 

inhibition in biopharmaceutics, Int. J. 

Pharm. 289 (2005) 1-30. 

92. M.M. Tan, C.A. Corley, C.L. Stevenson, 

Effect of gelation on the chemical stability 

and conformation of leuprolide, Pharm. Res. 

15 (1998) 1442-8. 

93. G.A. Brazeau, B. Cooper, K.A. Svetic, C.L. 

Smith, P. Gupta, Current perspectives on 

pain upon injection of drugs, J. Pharm. Sci. 

87 (1998) 667-77. 

94. Extemp.ie, Section 2: Sterile Pharmaceuticals 

- Parenterals, accessed April, 01 2012. 

95. L. Jorgensen, S. Hostrup, E.H. Moeller, 

H. Grohganz, Recent trends in stabilising 

peptides and proteins in pharmaceutical 

formulation - considerations in the choice 

of excipients, Expert Opin. Drug Deliv. 6 

(2009) 1219-30. 

96. J.L. Cleland, M.F. Powell, S.J. Shire, The 

development of stable protein formulations: 

a close look at protein aggregation, 

deamidation, and oxidation, Crit. Rev. Ther. 

Drug Carrier Syst. 10 (1993) 307-77. 

97. M. Cholewinski, B. Lückel, H. Horn, 

Degradation pathways, analytical 

characterization, and formulation strategies 

of a peptide and a protein: calcitonin and 

human growth hormone in comparison, 

Pharm. Acta Helv. 71 (1996) 405-19. 

98. N.E. Good, G.D. Winget, W. Winter, T.N. 

Connolly, S. Izawa, R.M. Singh, Hydrogen 

ion buffers for biological research, 

Biochemistry 5 (1966) 467-77. 

99. R. Li, M.J. Hageman, E.M. Topp, Effect of 

viscosity on the deamidation rate of a model 

Asn-hexapeptide, J. Pept. Res. 59 (2002) 

211-20. 

100. R. Li, A.J. D’Souza, B.B. Laird, R.L. Schowen, 

R.T. Borchardt, E.M. Topp, Effects of 

solution polarity and viscosity on peptide 

deamidation. 2000 Nov;56(5):326-334., J 

Pept Res 56 (2000) 326. 

101. K.C. Waterman, R.C. Adami, K.M. Alsante, 

J. Hong, M.S. Landis, F. Lombardo, C.J. 

Roberts, Stabilization of pharmaceuticals to 

oxidative degradation, Pharm. Dev. Technol. 

7 (2002) 1-32. 

102. S. Li, T.W. Patapoff, T.H. Nguyen, R.T. 

Borchardt, Inhibitory effect of sugars and 

polyols on the metal-catalyzed oxidation of 

human relaxin, J. Pharm. Sci. 85 (1996) 868- 

72. 

103. L.P. Stratton, R.M. Kelly, J. Rowe, J.E. Shively, 

D.D. Smith, J.F. Carpenter, M.C. Manning, 

Controlling deamidation rates in a model 

peptide: Effects of temperature, peptide 

concentration, and additives, J. Pharm. Sci. 

90 (2001) 2141-8. 

104. C. Avanti, J.P. Amorij, D. Setyaningsih, A. 

Hawe, W. Jiskoot, J. Visser, A. Kedrov, A.J. 

Driessen, W.L. Hinrichs, H.W. Frijlink, A 

new strategy to stabilize oxytocin in aqueous 

solutions: I. The effects of divalent metal ions 

and citrate buffer, AAPS J. 13 (2011) 284-90. 

105. W. Wang, S. Martin-Moe, C. Pan, L. Musza, 

Y.J. Wang, Stabilization of a polypeptide in 

non-aqueous solvents, Int. J. Pharm. 351 

(2008) 1-7. 

106. Y.C.J. Wang, M.A. Hanson, Parenteral 

formulations of proteins and peptides: 

stability and stabilizers, J. Parenter. Sci. 

Technol. 42 (1987) S2-S26. 

107. K. Zheng, C.R. Middaugh, T.J. Siahaan, 

Evaluation of the physical stability of the 

EC5 domain of E-cadherin: effects of pH, 

temperature, ionic strength, and disulfide 

bonds, J. Pharm. Sci. 98 (2009) 63-73. 

108. S.A. Bursakov, C. Carneiro, M.J. Almendra, 

R.O. Duarte, J. Caldeira, I. Moura, J.J. Moura, 

Enzymatic properties and effect of ionic 

strength on periplasmic nitrate reductase 

(NAP) from Desulfovibrio desulfuricans 

ATCC 27774, Biochem. Biophys. Res. 

Commun. 239 (1997) 816-22. 

109. P. O’Neill, S. Davies, E.M. Fielden, L. 

Calabrese, C. Capo, F. Marmocchi, G. Natoli, 

G. Rotilio, The effects of pH and various salts 

upon the activity of a series of superoxide 

dismutases, Biochem. J. 251 (1988) 41-6. 

110. R. Tyler-Cross, V. Schirch, Effects of amino 

acid sequence, buffers, and ionic strength on 

the rate and mechanism of deamidation of 

asparagine residues in small peptides, J. Biol. 

Chem. 266 (1991) 22549-56. 

111. K. Neelon, H.J. Schreier, H. Meekins, P.M. 

Robinson, M.F. Roberts, Compatible solute 

effects on thermostability of glutamine 

synthetase and aspartate transcarbamoylase 

from Methanococcus jannaschii, Biochim. 

Biophys. Acta 1753 (2005) 164-73. 

112. T.V. Brennan, S. Clarke, Spontaneous 

degradation of polypeptides at aspartyl and 

asparaginyl residues: effects of the solvent 

dielectric, Protein Sci. 2 (1993) 331-8.

113. D.A. Parkins, U.T. Lashmar, The Formulation 

of Biopharmaceutical Products, Pharm. Sci. 

Technol. Today 3 (2000) 129-37. 


instability of protein pharmaceuticals: 

Mechanisms of oxidation and strategies for 

stabilization, Biotechnol. Bioeng. 48 (1995) 

490-500. 

115. S. Landi, H.R. Held, Effect of oxidation on 

the stability of tuberculin purified protein 

derivative (PPD), Dev. Biol. Stand. 58 (1986) 

545-52. 

116. C. Schoneich, F. Zhao, G.S. Wilson, R.T. 

Borchardt, Iron-thiolate-induced oxidation 

of methionine to methionine sulfoxide in 

small model peptides. Catalysis by histidine., 

Biochim. Biophys. Acta 1158 (1993) 307-22. 

117. S. Li, C. Schoneich, G.S. Wilson, R.T. 

Borchardt, Chemical pathways of peptide 

degradation. V. Ascorbic acid promotes 

rather than inhibits the oxidation of 

methionine to methionine sulfoxide in 

small model peptides, Pharm. Res. 10 (1993) 

1572-9. 

118. R.L. Levine, J. Moskovitz, E.R. Stadtman, 

Oxidation of methionine in proteins: roles in 

antioxidant defense and cellular regulation, 

IUBMB Life 50 (2000) 301-7. 

119. P.K. Tsai, D.B. Volkin, J.M. Dabora, K.C. 

Thompson, M.W. Bruner, J.O. Gress, B. 

Matuszewska, M. Keogan, J.V. Bondi, C.R. 

Middaugh, Formulation design of acidic 

fibroblast growth factor, Pharm. Res. 10 

(1993) 649-59. 

120. B.P. Obiols, G.J. Farres, F.J.C. Rodriguez, 

S.P. Fernandez, J.B. Cabado, Stable 

pharmaceutical formulation for intravenous 

or intramuscular administration of active 

peptide compound (2003) March, 20, 2012. 

121. T. Arakawa, S.J. Prestrelski, W.C. Kenney, J.F. 

Carpenter, Factors affecting short-term and 

long-term stabilities of proteins, Adv. Drug 

Deliv. Rev. 10 (1993) 1-28. 

122. T. Matsuoka, S. Tomita, H. Hamada, 

K. Shiraki, Amidated amino acids are 

prominent additives for preventing heatinduced 

aggregation of lysozyme, J. Biosci. 

Bioeng. 103 (2007) 440-3. 

123. R. Quinn, J.D. Andrade, Minimizing the 

aggregation of neutral insulin solutions, J. 

Pharm. Sci. 72 (1983) 1472-3. 

124. D.K. Chou, R. Krishnamurthy, T.W. 

Randolph, J.F. Carpenter, M.C. Manning, 

Effects of Tween 20 and Tween 80 on the 

stability of Albutropin during agitation, J. 

Pharm. Sci. 94 (2005) 1368-81. 

125. A. Lahlou, B. Blanchet, M. Carvalho, P. P M., 

A. Astier, Mechanically-induced aggregation 

of the monoclonal antibody cetuximab, Ann. 

Pharm. Fr. 67 (2009) 340-52. 

126. G. Lentzen, T. Schwarz, Extremolytes: 

Natural compounds from extremophiles 

for versatile applications, Appl. Microbiol. 

Biotechnol. 72 (2006) 623-34. 

127. I. Yu, Y. Jindo, M. Nagaoka, Microscopic 

understanding of preferential exclusion of 

compatible solute ectoine: direct interaction 

and hydration alteration, J. Phys. Chem. B 

111 (2007) 10231-8. 

128. S. Knapp, R. Ladenstein, E.A. Galinski, 

Extrinsic protein stabilization by the 

naturally occurring osmolytes betahydroxyectoine 

and betaine, Extremophiles 

3 (1999) 191-8. 

129. A. Hedoux, J.F. Willart, L. Paccou, Y. Guinet, 

F. Affouard, A. Lerbret, M. Descamps, 

Thermostabilization mechanism of bovine 

serum albumin by trehalose, J. Phys. Chem. 

B 113 (2009) 6119-26. 

130. T.Q. Faria, S. Knapp, R. Ladenstein, 

A.L. Macanita, H. Santos, Protein 

stabilisation by compatible solutes: effect 

of mannosylglycerate on unfolding 

thermodynamics and activity of ribonuclease 

A, ChemBioChem 4 (2003) 734-41. 

131. J. Ryu, M. Kanapathipillai, G. Lentzen, 

C.B. Park, Inhibition of β-amyloid peptide 

aggregation and neurotoxicity by α-dmannosylglycerate, 

a natural extremolyte, 

Peptides 29 (2008) 578-84. 

132. E.A. Galinski, M. Stein, B. Amendt, M. 

Kinder, The kosmotropic (structureforming) 

effect of compensatory solutes, 

Comp. Biochem. Physiol. 117A (1997) 357- 

65. 

133. T. Arakawa, S.N. Timasheff, The stabilization 

of proteins by osmolytes, Biophys. J. 47 

(1985) 411-4. 

134. D.W. Bolen, G.D. Rose, Structure and 

Energetics of the Hydrogen-Bonded 

Backbone in Protein Folding, Annu. Rev. 

Biochem. 77 (2008) 339-62. 

135. N. Borges, A. Ramos, N.D. Raven, R.J. 

Sharp, H. Santos, Comparative study 

of the thermostabilizing properties of 

mannosylglycerate and other compatible 

solutes on model enzymes, Extremophiles 6 

(2002) 209-16. 


2 

33

Christina Avanti 1 , Jean-Pierre Amorij 1 , Dewi Setyaningsih 1 , Andrea Hawe 4 , 

Wim Jiskoot 4 , Jan Visser 2 , Alexej Kedrov 3 , Arnold J. M. Driessen 3 , 

Wouter L. J. Hinrichs 1 , and Henderik W. Frijlink 1 


2 Department of Pharmacokinetics, Toxicology, and Targeting, 

3 Department of Molecular Microbiology, 

University of Groningen, Groningen, The Netherlands. 

4 Division of Drug Delivery Technology, Leiden/Amsterdam Center for Drug Research, 

Leiden University, Leiden, The Netherlands.

a new strategy to stabIlIze oxytocIn 

In aqueous solutIons: I. the effects of 

dIvalent metal Ions and cItrate buffer 

AAP S Journal 2011; 13(2): 284–290 

3

3 

abstract 

In the current study, the effect of metal ions in combination with buffers (citrate, acetate, 

pH 4.5) on the stability of aqueous solutions of oxytocin was investigated. Both monovalent 

metal ions (Na + and K + ) and divalent metal ions (Ca 2+ , Mg 2+ , and Zn 2+ ) were tested all as 

chloride salts. The effect of combinations of buffers and metal ions on the stability of aqueous 

oxytocin solutions was determined by RP-HPLC and HP-SEC after 4 weeks of storage at 

either 4°C or 55°C. Addition of sodium or potassium ions to acetate- or citrate-buffered 

solutions did not increase stability, nor did the addition of divalent metal ions to acetate 

buffer. However, the stability of aqueous oxytocin in aqueous formulations was improved in 

the presence of 5 and 10 mM citrate buffer in combination with at least 2 mM CaCl 2 , MgCl 2 , 

or ZnCl 2 and depended on the divalent metal ion concentration. Isothermal titration 

calorimetric measurements were predictive for the stabilization effects observed during the 

stability study. Formulations in citrate buffer that had an improved stability displayed a 

strong interaction between oxytocin and Ca 2+ , Mg 2+ , or Zn 2+ , while formulations in acetate 

buffer did not. In conclusion, our study shows that divalent metal ions in combination with 

citrate buffer strongly improved the stability of oxytocin in aqueous solutions.


According to the World Health Organization, half a million of women in Africa, Asia, and 

Latin America die each year due to problems during pregnancy and childbirth. At least 25% 

of those deaths can be attributed to bleeding after child birth (post-partum hemorrhage), 

mainly caused by failure of the uterus to contract adequately after child birth (atonicity) [1]. 

The preferred drug to prevent post-partum hemorrhage is oxytocin. Oxytocin is a cyclic 

nonapeptide hormone [sequence: cyclo (Cys 1 -Tyr 2 -Ile 3 -Gln 4 -Asn 5 -Cys 6 ), -Pro 7 -Leu 8 -Gly 9 -NH 2 ], 

which is naturally produced in the hypothalamus. It is involved primarily in uterine contraction 

and stimulation of milk release from the mammary tissue [2]. Oxytocin, which is currently 

available in synthetic form [3], has been widely used for indications such as induction of 

labor, augmentation of labor, post-partum hemorrhage, or uterine atony, and also for other 

indications such as diabetes insipidus and vasodilatory shock. Reported additional functions 

for oxytocin include an antiduretic effect and blood vessel contraction [2,4,5]. 

Unfortunately, oxytocin preparations are highly unstable at elevated temperatures, 

which is an issue particularly in tropical countries [6]. Stability studies conducted by 

Groot et al. [6] have shown that injectable oxytocin formulations are rapidly degraded as 

the storage temperature rises to 30°C or higher. Oxytocic tablets for oral administration 

(ergometrine, ethylergometrine, oxytocin, and desamino-oxytocin) are also not stable 

under simulated tropical conditions. Because of its poor stability at elevated temperatures, 

the use of oxytocin in many developing countries is limited. Thus, there is a clear need for a 

heat-stable oxytocin formulation, preferably an aqueous injectable solution, with improved 

thermal stability. 

One way to effectively improve stability of several peptides in aqueous solution is using 

metal salts in combination with a suitable buffer [7]. To investigate the effect of metal ions 

in buffered solutions on the stability of oxytocin, we screened various combinations of 

unbuffered and buffered solutions with monovalent or divalent metal ions. Hawe et al. [8] 

observed that the degradation of oxytocin strongly depends on the pH of the formulation, 

with the highest stability at pH 4.5. Therefore, all formulations will be set to pH 4.5. The 

purpose of this study is to investigate whether specific combinations of buffer and metal 

ions can stabilize oxytocin. 

2. materIals and method 

2.1 Materials 

The following materials were used in this study: oxytocin monoacetate powder (Diosynth, 

Oss, The Netherlands), citric acid, calcium chloride (Riedel-de Haen, Seelze, Germany), 

acetic acid, magnesium chloride, zinc chloride (Fluka, Steinheim, Germany), sodium 

hydroxide, sodium chloride, potassium chloride, sodium dihydrogen phosphate dihydrate, 

acetonitrile, formic acid (Merck, Darmstadt, Germany) and Baxter Viavlo Ringer’s lactate 

solution for intravenous infusion (Baxter, Utrecht, The Netherlands). 

2.2 Formulation and Stability Study 

Oxytocin was formulated at a concentration of 0.1 mg/ml in citrate (5 or 10 mM) or 

acetate (10 mM) buffer at pH 4.5 (pH adjusted with sodium hydroxide) with different 

additions of metal ions. pH samples were controlled and remained within ±0.1 pH units 

StabIlIty of the dIvalent Metal-cItrate-oxytocIn coMplex 

3 

37

3 

38 

during the stability study. The initial concentration of oxytocin was determined using UV 

spectrophotometry [9] at 280 nm with an extinction coefficient of 1.52 ml mg −1 cm −1 . All 

metal ion solutions were prepared using their chloride salts at concentrations of 2, 5, 10, 

and 50 mM. Control solutions were formulated in water (pH 6.9±0.2) and Ringer’s lactate 

(6.4±0.2). Ringer’s lactate solution consists of 131 mM sodium, 5 mM potassium, 2 mM 

calcium, 111 mM chloride, and 29 mM bicarbonate (as lactate). In this report, the following 

codes were used: First character(s) refer to the type of buffer or water; CB (citrate), 

AC (acetate), RL (Ringer’s lactate), and W (water). Following digit(s) refer to buffer 

concentration in mM, following character(s) to the type of metal ion, and last digit(s) to 

metal ion concentration in mM. Thus, e.g., CB10Mg10 means 10 mM citrate buffer (pH 4.5) 

and 10 mM MgCl 2 . After preparation, the solutions were stored in 6R glass type 1 vials for 

4 weeks at either 4°C or 55°C, and protected from light. 

Based on the results of the screening study, oxytocin formulations in 10 mM citrate 

buffer pH 4.5 with 10 or 50 mM divalent metal salts were selected for a longer period 

of stability study for 6 months at 40°C according to ICH guidelines for long-term and 

accelerated stability study for climatic zone III and IV [10]. 

2.3 Reversed-Phase High-Performance Liquid Chromatography 

The recovery of oxytocin (remaining oxytocin as percentage of initial amount) was 

determined by RP-HPLC. RPHPLC was performed according to the procedure described 

by Hawe et al. [8] An Alltima C-18 RP column with 5 μm particle size, inner diameter of 

4.6 mm, and length of 150 mm (Alltech, Ridderkerk, Netherlands), a Waters (Millipore) 

680 Automated Gradient Controller, two Waters 510 HPLC pumps, a Waters 717 Plus 

Autosampler, and a Waters 486 Tunable Absorbance UV Detector were used. Samples of 

20 μl were injected and the separation was carried out at a flow rate of 1.0 ml/min and UV 

detection at 220 nm. Samples were eluted using 15% (v/v) acetonitrile in 65 mM phosphate 

buffer pH 5.0 as solvent A and 60% (v/v) acetonitrile in 65 mM phosphate buffer pH 5.0 as 

solvent B. The acetonitrile concentration was linearly increased from 15% at the beginning, 

to 20% at 10 min, to 30% at 20 min, and finally to 60% at 25 min. 

2.4 Size Exclusion HPLC 

The fraction of monomeric oxytocin (percentage of total remaining oxytocin) was assessed 

by Size Exclusion HPLC (HP-SEC). HP-SEC was carried out using a Superdex peptide 

10/300 GL column (GE Healthcare Inc., Brussels, Belgium) on an isocratic HPLC system, 

according to the method previously reported by Hawe et al. [8]. A Waters 510 pump, a 

Waters 717 plus auto sampler, a Waters 474 Scanning Fluorescence Detector and Waters 484 

Tunable Absorbance Detector (Waters, Milford Massachusetts, USA) were used. Samples 

of 50 μl were injected, and separation was performed at a flow rate of 1 ml/min. Peaks 

were detected by UV absorption at 274 nm, as well as fluorescence detection at excitation 

wavelength of 274 nm and emission wavelength of 310 nm. The mobile phase consisted of 

30% acetonitrile and 70% 0.04 M formic acid. 

2.5 Isothermal Titration Calorimetry 

Isothermal titration calorimetry (ITC) was used to investigate the interaction between 

oxytocin and divalent metal ions in the presence of citrate buffer and acetate buffer. 

Microcalorimetric titrations of divalent metal ions to oxytocin were conducted by using a

MicroCal ITC 200 Microcalorimeter (Northampton, MA 01060 USA). A solution of 300 μL 

of 5 mM oxytocin in 10 mM of either citrate or acetate pH 4.5 was placed in the sample cell, 

while 30 μL of 125 mM divalent metal chloride either calcium, magnesium, or zinc in 10 mM 

citrate or acetate buffer pH 4.5 was placed in the syringe. The reference cell contained 300 μL 

of the corresponding buffer. Experiments were performed at 55°C. Automated titrations 

were conducted up to a divalent metal ion/oxytocin molar ratio of 5:1. The effective heat of 

the peptide-metal ion interaction upon each titration step was corrected for dilution and 

mixing effects, as measured by titrating the divalent metal ion solution into buffer and by 

titrating buffer into oxytocin solution. To investigate the possibility of oxytocin or metal ion 

binding to the buffer components, control experiments were performed in water. The heats 

of bimolecular interactions were obtained by integrating the peak following each injection. 

All measurements were performed in triplicate. 

ITC data were analyzed by using the ITC non-linear curve fitting functions for one 

or two binding sites from MicroCal Origin 7.0 software (MicroCal Software, Inc.). The 

calculated curve was determined by the best-fit parameter, which was used to determine 

the molar enthalpy change for binding and the corresponding association constant, K a . The 

molar free energy of binding ΔG° and the molar entropy change ΔS° were derived from the 

fundamental equations of thermodynamics ΔG° = -RTlnK a and ΔG° = ΔH°- TΔS°. 

3. results 

3.1 Influence of Divalent Metal Ions on Oxytocin Stability in 

Unbuffered Solutions 

First, the effect of divalent metal ions on the stability of oxytocin in water without any buffer 

salt was investigated. RP-HPLC results (Fig. 1a) showed that after 4 weeks of storage at 4°C, 

oxytocin recovery was almost 100% in the presence of 2–50 mM zinc or 50 mM calcium 

ions. No stabilizing effect was observed from the presence of magnesium (2–50 mM) 

and calcium (2–10 mM), where the recovery was reduced to about 65%, similar to levels 

found for oxytocin solutions in water. HPSEC results (Fig. 1b) showed a similar trend in 

the recovery of monomeric oxytocin. However, when the solutions were stored at 55°C, 

both RP-HPLC and HP-SEC measurements showed substantial degradation of oxytocin 

after 4 weeks. These results demonstrate that divalent metal ions in non-buffered aqueous 

oxytocin formulations have only a limited stabilizing effect at elevated temperature. 

3.2 Oxytocin Stability in Buffered Solutions 

To determine the effect of buffer on stability, oxytocin was formulated in 5 and 10 mM 

citrate buffer and 10 mM acetate buffer.As a reference, the stability of oxytocin in pure water 

and in Ringer’s lactate buffer was investigated. Figure 2a shows the oxytocin recovery in 

RP-HPLC after 4 weeks of storage either at 4°C or 55°C in the buffered solutions. Compared 

to pure water, the stability of oxytocin was substantially increased at 4°C in the presence of 

the buffer salts. After storage at 55°C, the recovery of oxytocin after 4 weeks in citrate and 

acetate buffer was much higher as compared to water or Ringer’s lactate solution. However, 

the recovery of oxytocin was still poor. In addition, only about 20% of oxytocin remained 

in its monomeric form after storage (Fig. 2b). 


3 

39

3 

40 

Figure 1. Recovery of oxytocin in the presence of divalent metal ions in non-buffered pure 

water, stored for 4 weeks at pH 4.5 and a temperature of 4°C (light gray bars) or 55°C (dark 

gray bars). The divalent metal ions (Ca 2+ , Mg 2+ , and Zn 2+) were used in concentrations of 

2, 5, 10, and 50 mM. a recovery determined by RP-HPLC. b oxytocin monomer recovery 

determined by HP-SEC. The results are depicted as averages of three independent 

measurements±SD 

3.3 Oxytocin Stability in Buffered Solutions Containing Monovalent 

Metal Ions 

To investigate the stability of oxytocin in the presence of a combination of buffer and 

monovalent metal ions, citrate and acetate buffer were used at a concentration of 10 mM, 

in combination with the monovalent metal ions, sodium and potassium, added at a 

concentration of 10 and 20 mM (excluding sodium from the buffer component). The

Figure 2. Recovery of oxytocin in pure water, with or without a buffer. Citrate buffer at a 

concentration of 5, 10, or 50 mM, acetate buffer at a concentration of 10 mM, or Ringer’s 

lactate solution were used. The formulations contained no metal ions and were stored for 

4 weeks at pH 4.5 or 6.4 for Ringer’s lactate solution at a temperature of 4°C (light gray 

bars) or 55°C (dark gray bars). a recovery determined by RP-HPLC. b oxytocin monomer 

recovery determined by HP-SEC. The results are depicted as averages of three independent 

measurements±SD 

presence of monovalent metal ions had only a minor effect on the stability of oxytocin 

(stored at 55°C). Although the oxytocin recovery was slightly improved compared to 

oxytocin in the presence of buffer alone, the maximum recovery of oxytocin was only 35% 

in 10 mM acetate buffer with 10 mM sodium chloride. In addition, only about 30% of 

oxytocin in this formulation remained monomeric (data not shown). These results clearly 

indicate that the presence of a combination of buffer and monovalent metal ions is not 

sufficient to substantially stabilize oxytocin in aqueous solution. 


3 

41

3 

42 

3.4 Oxytocin Stability in Citrate-Buffered Solutions Containing 

Divalent Metal Ions 

To study the stability of oxytocin in the presence of citrate buffers and divalent metal ions, 

formulations containing 5 and 10 mM citrate buffer in combination with divalent metal ions 

(calcium, magnesium, and zinc) added at concentrations of 2, 5, 10, or 50 mM were used. 

The results of RP-HPLC and HP-SEC of formulations in 10 mM citrate buffer are presented 

in Fig. 3a and b. The stability of oxytocin solutions was clearly improved when formulating 

them with citrate buffer in combination with calcium ions. The oxytocin stability increased 

Figure 3. Effect of Ca 2+ (squares), Mg 2+ (circles), and Zn 2+ (triangles) concentration on the 

recovery of oxytocin in citrate buffer at the concentration of 10 mM after 4 weeks of storage 

at either 4°C or 55°C and pH 4.5. Solid symbols denoted 4°C storage, while open symbols 

correspond to 55°C. a recovery determined by RP-HPLC. b oxytocin monomer recovery 

determined by HP-SEC. The results are depicted as averages of three independent 

measurements±SD

with increasing calcium ion concentrations. The recovery of oxytocin and the remaining 

percentage of oxytocin monomers after 4 weeks of storage at 55°C were increased up to 

almost 80% in the presence of 50 mM calcium. 

Similar results were obtained for the combination of citrate with magnesium. The 

degradation of oxytocin in citrate buffer at 5 and 10 mM decreased with an increasing 

concentration of magnesium ions. Formulations with zinc ions in citrate buffer also 

preserved oxytocin during storage. These combinations exert even a stronger effect on 

oxytocin stability than combinations of citrate buffer and calcium or magnesium ions. 

Stability was strongly improved at zinc concentrations as low as 2 or 5 mM. Both oxytocin 

recovery (RP-HPLC) and the monomeric oxytocin fraction (HP-SEC) were substantially 

higher (up to 90%) in the presence of 10 mM zinc ions (CB5Zn10) after storage for 4 weeks 

at 55°C. When citrate buffer was used at a concentration of 5 mM, similar results were 

found (data not shown). 

Beside citrate, we also carried out several further experiments to investigate whether 

divalent metal ions affect the stability in the presence of acetate buffer. Acetate buffer at a 

concentration of 10 mM was used with and without calcium, magnesium, and zinc ions at 

various concentrations (2, 5, 10, 50 mM). The combination of these ions with acetate buffer 

was found to be less efficient in stabilizing oxytocin (data not shown). 

3.5 Long-term Stability of Oxytocin in Selected Formulations 

Containing Citrate and Divalent Metal Ions 

For the combination of citrate with Ca 2+ , Zn 2+ , and Mg 2+ , a long-term stability study 

for 6 months at 40°C was conducted. A temperature of 40°C was chosen to simulate 

tropical conditions [10,11]. The long-term stability study at 40°C clearly demonstrates 

the synergistic stabilizing effect of citrate buffer and divalent metal ions. Even though the 

oxytocin recovery decreased gradually with time, the recovery of oxytocin (Fig. 4a) and the 

remaining percentage of oxytocin monomers (Fig. 4b) after 6 months storage at 40°C were 

increased up to 80%in the presence of 50 mM calcium, and even higher (up to 90%) in the 

presence of 50 mM magnesium. 

Formulations with 10 mM zinc ions in citrate buffer exerted the same effect on oxytocin 

stability as combinations of citrate buffer and 50 mM magnesium ions. This shows that zinc 

ions at lower concentrations have already a higher impact on increasing oxytocin stability 

compared with calcium or magnesium ions. 

This result also confirms the short-term stability study that showed up to 90%remaining 

oxytocin in the presence of 10mM zinc ions after storage for 4 weeks at 55°C. 

3.6 ITC to Study the Interaction between Oxytocin and Divalent 

Metal Ions 

To examine the interaction between oxytocin and the metal ions, ITC experiments were 

carried out that are summarized in Table 1. The titration of calcium ions into an oxytocin 

solution in citrate buffer resulted in an exothermic reaction (Fig. 5a) with a K a value of 

about 400 M −1 and an apparent ion/oxytocin stoichiometry close to 4:1 (Table 1). 

Remarkably, when titrating magnesium into oxytocin we observed heat absorption (Fig. 5a) 

and this endothermic reaction occurred with an identical apparent stochiometry (4:1) and a similar 

K a value of about 200 M −1 as with calcium ions. However, with magnesium ions the ITC trace was 

complex and showed an exothermic phase at magnesium concentrations below 5 mM. Due to 


3 

43

3 

44 

Figure 4. Oxytocin recovery over time storage at 40°C and pH 4.5 in the presence of 10 

mM citrate buffer, without (star) and with divalent metal ions. Ca 2+ (square), Mg 2+ (triangle), 

and Zn 2+ (circle) were used in concentrations of 10 mM (open symbols), and 50 mM (solid 

symbols). a recovery determined by RP-HPLC. b oxytocin monomer recovery determined 

by HP-SEC. The results are depicted as averages of three independent measurements±SD 

the rapid saturation of this early phase—typically, within three injections—and the low enthalpy 

of the ion/oxytocin interaction, it was not possible to quantitatively analyze the exothermic stage. 

To analyze the following endothermic phase we omitted the first 4-titration steps and fitted the 

remaining data using a single site model. A similar dual-phase response was observed for the 

zinc: oxytocin interaction (Fig. 5a, open triangles), but the exothermic and endothermic stages 

were well-resolved and suitable for the analysis using a two sites model (Table 1). In contrast, in 

the presence of acetate buffer there was no measurable interaction between oxytocin and calcium, 

magnesium, or zinc ion (Fig. 5b).

Table 1. Thermodynamics of Divalent Metal binding to Oxytocin as Determined by Isothermal Titration 

Calorimetry in 10 mM Citrate Buffer 

Metal Phase N (sites) Ka (M−1 ) ΔH°(cal mol−1 ) ΔS° (cal/mol/deg) 

Ca2+ 1 0.26 400 −2,100 5.6 

Mg2+ 1 n.d. n.d.

3 

46 

a b 

Figure 5. Least squares fit of the data from calorimetric titration profiles of aliquots of 125 

mM divalent metal ions: Ca 2+ (solid square), Mg 2+ (open square), and Zn 2+ (open triangle) 

into 5 mM oxytocin in 10 mM a citrate buffer and b acetate buffer pH 4.5. The heat absorbed 

per mol of titrant is plotted versus the ratio of the total concentration of divalent metal ions 

to the total concentration of oxytocin 

from water molecules. These conformational changes can increase the stability of oxytocin in 

aqueous medium as they will prevent dimerization and further aggregation by hydrophobic 

interactions among oxytocin molecules. Metal salts are often used to stabilize peptides or 

proteins by chelation or ionic interactions [22]. Wang et al. examined the peptide (P66) 

stability in the presence of ZnCl 2 , MgCl 2 , and CaCl 2 in non aqueous solution, and found that 

in the presence of 1 mM ZnCl 2 , P66 was significantly stabilized. However in the aqueous 

solution (pure water), these ions did not show any stabilizing effect [22]. In our experiments, 

the addition of calcium, magnesium, and zinc ions in combination with citrate buffer had 

a large impact on oxytocin stability in contrast to similar experiments in pure water or 

acetate buffer. This study suggests that there is a synergistic effect between citrate buffer and 

the divalent metal ions, possibly due to the protection of the disulfide bridge by complex 

formation of divalent metal ion and citrate with oxytocin which suppressed intermolecular 

reaction leading to tri/tetrasulfide formation as well as dimerization (unpublished data). 

ITC is a sensitive method for studying the thermodynamics of binding events and 

quantifying binding reactions. When divalent metal ion are added to oxytocin, the ITC 

data indicate an interaction between oxytocin and Ca 2+ , Mg 2+ , or Zn 2+ ions in the presence 

of citrate buffer. Both Mg 2+ and Zn 2+ ions demonstrated complex, dual-phase interaction 

profile, while a single phase was observed for Ca 2+ . Each interaction was entropy driven, 

while both exothermic and endothermic reactions were observed. It may be speculated that 

the solvation effect, i.e., release of structured water molecules plays a key role in binding, 

while the specific ion–oxytocin interaction further contributes to the complex stability. 

The latter is also predicted by molecular dynamic simulations [20,21]. Remarkably, no

interaction between oxytocin and either of the tested ions was detected in the acetate buffer 

or deionized water. These observations underscore the role of a particular environment 

in the ion–oxytocin interaction and agree well with our findings on the peptide stability. 

Isothermal titration calorimetric measurements were predictive for the effects observed 

during the stability study. 

In conclusion, this study shows that with a combination of divalent metal salts and citrate 

buffer, the stability of oxytocin in aqueous solution can be strongly improved. The increased 

stability of oxytocin aqueous formulations was achieved in the presence of citrate acid buffer and 

2 mM or more of the salts CaCl 2 , MgCl 2 , or ZnCl 2 . The oxytocin stability is further increased with 

increasing concentration of the divalent metals ions up to 50 mM. In combination with citrate 

buffer, Zn 2+ has a superior stabilizing effect as compared with Ca 2+ or Mg 2+ . 

acknowledgments 

The authors want to thank MSD Oss for providing oxytocin for the study. This study was 

performed within the framework of the Dutch Top Institute Pharma project: number 

D6–202. Open Access This article is distributed under the terms of the Creative Commons 

Attribution Noncommercial License which permits any noncommercial use, distribution, 

and reproduction in any medium, provided the original author(s) and source are credited. 

references 

1. K.S. Khan, D. Wojdyla, L. Say, A.M. 

Gulmezoglu, P.F. Van Look, WHO analysis 

of causes of maternal death: a systematic 

review, Lancet 367 (2006) 1066-74. 

2. V.S. Ananthanarayanan, K.S. Brimble, 

Interaction of oxytocin with Ca2+: I. CD and 

fluorescence spectral characterization and 

comparison with vasopressin, Biopolymers 

40 (1996) 433-43. 

3. E.H. Bishop, Synthetic oxytocin; a clinical 

evaluation, Obstet Gynecol 11 (1958) 290-4. 

4. K.P. Conrad, M. Gellai, W.G. North, H. 

Valtin, Influence of oxytocin on renal 

hemodynamics and electrolyte and water 

excretion, Am. J. Physiol. 251 (1986) F290-6. 

5. A.V. Somlyo, C.Y. Woo, A.P. Somlyo, 

Responses of Nerve-Free Vessels to 

Vasoactive Amines and Polypeptides, Am J 

Physiol 208 (1965) 748-53. 

6. A.N.J.A.D. Groot, T.B. Vree, H.V. Hogerzeil, 

G.J.A. Walker, WHO Action Programme on 

Essential Drugs: Stability of Oral Oxytocics 

in Tropical Climates : Results of Simulation 

Studies on Oral Ergometrine, Oral 

Methylergometrine, Buccal Oxytocin and 

Buccal Desamino-Oxytocin, World Health 

Organization, Geneva, 1994. 




Res. 27 (2010) 544-75. 






(2009) 1679-88. 

9. S.C. Gill, P.H. von Hippel, Calculation of 

protein extinction coefficients from amino 

acid sequence data, Anal. Biochem. 182 

(1989) 319-26. 

10. International Conference on Harmonisation; 

Stability Data Package for Registration 

Applications in Climatic Zones III and IV; 

Stability Testing of New Drug Substances 

and Products; availability. Notice, Fed. 

Regist. 68 (2003) 65717-8. 

11. W. Grimm, Extension of the International 

Conference on Harmonization Tripartite 

Guideline for Stability Testing of New Drug 

Substances and Products to countries of 

climatic zones III and IV, Drug Dev. Ind. 

Pharm. 24 (1998) 313-25. 





13. L.A. Trissel, Y. Zhang, K. Douglass, E. 

Kastango, Extended Stability of Oxytocin 

in common infusion solution, International 

Journal of Pharmaceutical Compounding 10 

(2006) 156-8. 


3 

47

3 

48 

14. A.B. Joshi, M. Sawai, W.R. Kearney, L.E. 

Kirsch, Studies on the mechanism of aspartic 

acid cleavage and glutamine deamidation in 

the acidic degradation of glucagon, J. Pharm. 

Sci. 94 (2005) 1912-27. 

15. H. Yang, R.A. Zubarev, Mass spectrometric 


aspartate isomerization in polypeptides, 


16. N.E. Robinson, Protein deamidation, Procl. 

Natl. Acad. Sci. 99 (2002) 5283-8. 

17. C. Leeuwenburgh, J.E. Rasmussen, F.F. Hsu, 

D.M. Mueller, S. Pennathur, J.W. Heinecke, 

Mass spectrometric quantification of 

markers for protein oxidation by tyrosyl 

radical, copper, and hydroxyl radical in low 

density lipoprotein isolated from human 

atherosclerotic plaques, J. Biol. Chem. 272 

(1997) 3520-6. 

18. A. Fiser, I. Simon, Predicting the oxidation 

state of cysteines by multiple sequence 

alignment, Bioinformatics 16 (2000) 251-6. 

19. M. Klingenberg, M. Appel, The uncoupling 

protein dimer can form a disulfide cross-link 

between the mobile C-terminal SH groups, 

Eur J Biochem 180 (1989) 123-31. 

20. V.S. Ananthanarayanan, M.P. Belciug, 

B.S. Zhorov, Interaction of oxytocin with 

Ca2+: II. Proton magnetic resonance 

and molecular modeling studies of 

conformations of the hormone and its Ca2+ 

complex, Biopolymers 40 (1996) 445-64. 

21. D. Liu, A.B. Seuthe, O.T. Ehrler, X. Zhang, 

T. Wyttenbach, J.F. Hsu, M.T. Bowers, 

Oxytocin-receptor binding: why divalent 

metals are essential, J. Am. Chem. Soc. 127 

(2005) 2024-5. 

22. W. Wang, S. Martin-Moe, C. Pan, L. Musza, 

Y.J. Wang, Stabilization of a polypeptide in 

non-aqueous solvents, Int. J. Pharm. 351 

(2008) 1-7.

Christina Avanti 1 , Hjalmar P. Permentier 2 , Annie van Dam 2 , Robert Poole 3 , 

Wim Jiskoot 3 , Henderik W. Frijlink 1 , and Wouter L. J. Hinrichs 1 

1 Department of Pharmaceutical Technology & Biopharmacy and 

2 Mass Spectrometry Core Facility, 


3 Division of Drug Delivery Technology, Leiden/Amsterdam Center for Drug Research, 

Leiden University, Leiden, The Netherlands

a new strategy to stabIlIze oxytocIn 

In aqueous solutIons: II. suppressIon 

of cysteIne-medIated Intermolecular 

reactIons by a combInatIon of 

dIvalent metal Ions and cItrate 

Mol. Pharmaceutics. 2012 ; 9(3): 554-62 

4

4 

abstract 

A series of studies have been conducted to develop a heat-stable liquid oxytocin formulation. 

Oxytocin degradation products have been identified including citrate adducts formed in a 

formulation with citrate buffer. In a more recent study we have found that divalent metal 

salts in combination with citrate buffer strongly stabilize oxytocin in aqueous solutions 

(Avanti, C.; et al. AAPS J. 2011, 13, 284−290). The aim of the present investigation was 

to identify various degradation products of oxytocin in citrate-buffered solution after 

thermal stress at a temperature of 70°C for 5 days and the changes in degradation pattern 

in the presence of divalent metal ions. Degradation products of oxytocin in the citrate 

buffer formulation with and without divalent metal ions were analyzed using liquid 

chromatography−mass spectrometry/ mass spectrometry (LC−MS/MS). In the presence 

of divalent metal ions, almost all degradation products, in particular citrate adduct, tri- 

and tetrasulfides, and dimers, were greatly reduced in intensity. No significant difference 

in the stabilizing effect was found among the divalent metal ions Ca 2+ , Mg 2+ , and Zn 2+ . The 

suppressed degradation products all involve the cysteine residues. We therefore postulate 

that cysteine-mediated intermolecular reactions are suppressed by complex formation of 

the divalent metal ion and citrate with oxytocin, thereby inhibiting the formation of citrate 

adducts and reactions of the cysteine thiol group in oxytocin.


Oxytocin is a neurohypophyseal hormone, which was first discovered by H. H. Dale in 

1909 [1,2]. Oxytocin is produced by neurons of the posterior lobe of the hypophysis and 

pulsatively released into the periphery. In clinical practice, oxytocin has been prescribed 

primarily for labor induction and augmentation, control of postpartum hemorrhage and 

uterine hypotonicity in the third stage of labor [3,4]. Oxytocin is commonly administered 

by intravenous infusion [5]. 

The oxytocin structure was elucidated in 1951 [6-8] and the characterization and 

biosynthesis of oxytocin were reported in 1953 by du Vigneaud [9]. Oxytocin consists 

of nine amino acids: cyclo-(Cys 1 -Tyr 2 -Ile 3 -Gln 4 -Asn 5 -Cys 6 )-Pro 7 -Leu 8 -Gly 9 -NH 2 with a 

disulfide bridge between Cys residues 1 and [6,10,11]. The primary structure of oxytocin is 

shown in Figure 1. A major problem of the compound is its intrinsic instability in aqueous 

formulations [12]. Recently significant attention was focused on efforts to overcome the 

instability of oxytocin [13,14]. We have conducted several studies with the aim to develop a 

heat-stable oxytocin formulation. 

We identified the main degradation products of oxytocin stressed at a temperature of 

70°C in various buffers, pH values and storage time [13] as well as citryl oxytocin in citratebuffered 

formulations [14]. The degradation reactions and target residues of oxytocin are 

indicated in Figure 1. 

In a recent publication we described that formulations containing divalent metal salts 

in combination with citrate buffer strongly stabilize oxytocin in aqueous solutions [15]. 

However, the role by which divalent metal ions stabilize oxytocin in citrate buffer and 

their influence on (inhibition of) formation of degradation products were not elucidated. 

Therefore, the aim of the present study was to identify by LC−MS/MS the degradation 

products of oxytocin solutions after thermal stress and to investigate which degradation 

pathways are suppressed by the formulations containing combinations of divalent metal 

ions and citrate buffer. 

Figure 1. Molecular structure of oxytocin with its ring and tail fragment ions formed upon 

MS/MS fragmentation. 

SuppreSSIon of cySteIne-MedIated InterMolecular reactIonS 

4 

53

4 

54 

2. materIals and methods 

2.1 Materials 

The following materials were used in this study: oxytocin monoacetate powder (Diosynth, 

Oss, The Netherlands), citric acid, calcium chloride (Riedel-de Haen, Seelze, Germany), 

magnesium chloride, zinc chloride (Fluka, Steinheim, Germany), sodium hydroxide, 

acetonitrile, ammonium acetate (Merck, Darmstadt, Germany), DL-dithiothreitol and 

ammonium bicarbonate (Sigma-Aldrich Chemie Gmbh, Steinheim, Germany). 

2.2 Formulation and Stability Study 

Three independent batches of each of the four formulations given in Table 1 were prepared. 

The concentration of the citrate buffer was 10 mM, and the pH was adjusted to 4.5. All metal 

ion (M 2+ ) solutions were prepared using their respective chloride salts at concentrations of 

10 mM (MCl 2 ). The solutions were stored for 5 days protected from light at either 4 or 70 

°C. Prior to analysis, the samples were diluted 10-fold with water. 

2.3 Reversed-Phase High Performance Liquid Chromatography 

HPLC was performed using a Shimadzu LC system, consisting of LC-20AD gradient pumps 

and a SIL-20AC autosampler. Chromatographic separation was achieved on an Alltima C18 

column (2.1 × 150 mm 5 μm, Grace Davison Discovery Sciences). The injection volume was 

50 μL. Elution was performed by a linear gradient from 5% to 60% B in 30 min, followed 

by an increase to 90% B in 1 min, where it was kept 4 min, after which it returned to the 

starting conditions. Eluent A was 95% water/5% acetonitrile (AcN), and eluent B was 5% 

water/95% AcN, both containing 0.05% v/v acetic acid and 10 mM ammonium acetate. The 

flow rate was 0.2 mL/min. The UV signal was recorded at 220 nm. 

2.4 Liquid Chromatography−Mass Spectrometry 

The HPLC system was coupled to an API 3000 triple-quadrupole mass spectrometer 

(Applied Biosystems/MDS Sciex) via a turbo ion spray source. The ionization was performed 

by electrospray in the positive mode. Full scan spectra were recorded at a scan rate of 2 s 

from m/z 500 to 1300 and a step size of 0.2 amu with a declustering potential (DP) of 40 V 

and a focusing potential (FP) of 250 V. Product ion scans were acquired in specified time 

windows with a DP of 60 V, a FP of 300 V, a collision energy of 40 V and a collision cell exit 

Table 1. The Composition of Oxytocin Liquid Formulation 

formulation a oxytocin metal ion buffer 

OCB 0.1 mM citrate 10 mM 

OCBCa 0.1 mM Ca2+ 10 mM citrate 10 mM 

OCBMg 0.1 mM Mg2+ 10 mM citrate 10 mM 

OCBZn 0.1 mM Zn2+ 10 mM citrate 10 mM 

a OCB = oxytocin in the absence of divalent metal ions in 10 mM citrate-buffered solution at pH 4.5. 

OCBCa = oxytocin in the presence of Ca 2+ in 10 mM citrate-buffered solution at pH 4.5. 

OCBMg = oxytocin in the presence of Mg 2+ in 10 mM citrate-buffered solution at pH 4.5. 

OCBZn = oxytocin in the presence of Zn2+ in 10 mM citratebuffered solution at pH 4.5.

potential of 20 V. Data acquisition and processing was performed using Analyst version 

1.4.2 and 1.5 software (Applied Biosystems/MDS Sciex). 

LC−MS/MS were used to identify the oxytocin degradation products observed by 

RP-HPLC. The analyses were carried out on purified fractions corresponding to each of the 

major degradation peaks for oxytocin in citrate buffer formulation without divalent metal 

salts as well as by LC−MS for the formulation with and without divalent metal salts. 

Reduction of disulfide bonds was performed by adding 10 mM DL-dithiothreitol (DTT) 

in 200 mM ammonium carbonate. The samples were analyzed after an incubation time of at 

least 10 min at room temperature. 

3. results 

3.1 Degradation Products of Oxytocin in Citrate Buffer with and 

without Divalent Metal Ions 

Assignment of degradation products was done based on the m/z value of each compound 

from the LC−MS data, using known assignments and retention time order from previous 

studies [13,14]. MS/MS data were used for the more intense peaks to confirm the identification 

and to determine in which part of the molecule the modification had occurred. 

The highest degradation product intensities were found in the formulation without 

divalent metal salts stressed at 70°C for 5 days. The HPLC profiles based on ultraviolet (UV) 

absorbance at 220 nm and the total ion current (TIC) of mass spectra with m/z between 

900 and 1200 are shown in Figure 2a and Figure 2b, respectively. The UV profile shows 

11 main peaks and the TIC 12 main peaks (see labels in Figure 2b). The LC−MS contour 

plot (Figure 2c) reveals that some peaks contain several molecular species, which cannot be 

distinguished in the TIC, notably peaks 2 and 3 reveal two compounds at 12.0 and 12.1 min, 

peaks 10 and 11 reveal two compounds at 16.7 and 16.8 min, and peaks 14 and 15 reveal two 

compounds at 18.7 and 18.9 min. All 15 assigned molecular species are presented in Table 2, 

and MS/MS product ion spectra are shown in Figures S1−S10 in the Supporting Information. 

The protonated molecular ion of unmodified oxytocin was found at a retention time of 

12.6 min. Figure 2d shows its extracted ion chromatogram (XIC) at m/z 1007.6. MS/MS 

fragmentation of unmodified oxytocin results in fragments of m/z 723.4 and 285.2, from 

the disulfide-linked ring of residues 1−6 and the C-terminal residues 7−9, respectively 

(Figure 1 and Figure S4 in the Supporting Information). 

The degradation peaks eluting before the oxytocin peak (labeled 1 and 2 in Figure 2b) 

were identified as amide- and imide-linked N-citryl oxytocin with m/z of 1181.8 and 

1163.8, respectively.14 The MS/MS spectrum of m/z 1181.6 showed that the ring fragment 

was shifted from m/z 723.4 to 897.4, but the tail fragment was still observed at m/z 285.2. 

It shows that the citrate modification is in the ring fragment and the most likely location is 

on the N-terminal amine (see Figure 1 and Figure S2 in the Supporting Information) [14]. 

At a retention time of 12.0 and 12.1 min (Figure 3b and 3c), there is overlap of two 

different compounds (peaks 2 and 3), namely a monodeamidated species of oxytocin with 

a protonated molecular ion at m/z 1008.6 in a very low intensity and dehydrated (imidelinked) 

N-citryl oxytocin with the protonated molecular ion at m/z 1163.6 [14]. The MS/ 

MS spectrum of m/z 1008.6 showed a shift of the ring fragment from m/z 723.4 to 724.4, but 

no shift of the tail fragment mass (Figure S3 in the Supporting Information), which means 

that the deamidation occurred at Asn 5 or Gln 4 [13]. The 13 C denoted as peak in Figure 3c is 


4 

55

4 

56 

Figure 2. LC−UV trace (a), total ion chromatogram (m/z 900−1200) (b), and LC−MS contour 

plot (c) of stressed oxytocin in citrate buffered solution with the extracted ion current of 

oxytocin (d) and acetylated oxytocin (e).

Table 2. Summary of the Observed Degradation Products of Oxytocin after Stressing at 70°C for 5 Days 

in 10 mM Citrate Buffer 

no. t (min) R m/z charge Assignment 

1 11.8 1181.6 1+ oxytocin N-citryl amide 

2 12.0 1008.6 1+ monodeamidation at Asn5 or Gln4 3 12.1 1163.6 1+ oxytocin N-citryl imide 

4 12.6 1007.6 1+ oxytocina 5 13.5 1039.6 1+ oxytocin trisulfide 

6 13.9 1195.6 1+ oxytocin trisulfide N-citryl imide 

7 14.4 1049.6 1+ N-acetyl oxytocina 8 14.8 1071.6 1+ oxytocin tetrasulfide 

9 15.5 975.6 2+ dimer 1 

10 16.7 975.6 2+ dimer 2 

11 16.8 1008.6 2+ dimer 3 

12 17.4 1024.6b 2+ dimer 4 

13 17.8 1008.6 2+ dimer 5 

14 18.7 1024.6b 2+ dimer 6 

15 18.9 1008.6 2+ dimer 7 

a No degradation products; compounds were also found in the original oxytocin preparation. b Ammoniated. 

the second isotope peak of oxytocin (m/z 1007.6), which is also found at m/z 1008.6. The 

three other peaks in Figure 3b, at retention times of 16.8, 17.8, and 18.9 min, are the dimer 

degradation products described in Figure 4c. 

A compound at m/z 1049.6 was found at a retention time of 14.4 min (Figure 2e) in every 

sample including the original oxytocin preparation. The difference of +42 Da with respect 

to oxytocin suggests acetylation, presumably at the N-terminal amine (see Figure S6 in the 

Supporting Information). This acetylation reaction may have occurred during synthesis 

since oxytocin is supplied as the monoacetate salt. 

At a retention time of 13.5 min a trisulfide degradation product was identified from 

the m/z 1039.6 peak. Upon MS/ MS fragmentation, the ring fragment shifted to m/z 755.4, 

while the tail fragment mass did not change (see Figure S5 in the Supporting Information). 

The mass difference of +32 Da can be assigned to a trisulfide modification in the oxytocin 

ring [13]. At a retention time of 13.9 min, a product of m/z 1195.6 was found, which is 

tentatively identified as the N-citryl oxytocin with the trisulfide modification. The formation 

of a tetrasulfide product (m/z 1071.6) was found at the retention time of 14.8 min. Molecular 

structure N-citryl oxytocin, trisulfide and tetrasulfide modification produced from the 

degradation of stressed oxytocin in citrate buffered solution are shown in Figure 5. 

Seven peaks of dimers were evident as doubly charged ions in MS eluting between 15.5 

and 18.9 min (Table 2 and Figure 4). Dimer 1 and 2 had identical masses (m/z 975.6) and 

were also observed in a previous study of stressed oxytocin solutions [13] where they were 

assigned to the doubly protonated form of a sulfur-linked dimeric oxytocin species that has 

been doubly deamidated and which has lost two sulfur atoms through β-elimination [13]. 

Identification of dimers is difficult due to the overlapping LC peaks and multiple ion forms 

(protonated and ammoniated, and doubly charged, Figure 2c). In addition, their MS/MS 


4 

57

4 

58 

Figure 3. LC−MS extracted ion currents for oxytocin N-citryl amide (a) and N-citryl imide 

(b), monodeamidation (c), trisulfide (d), trisulfide N-citryl imide (e), and tetrasulfide products 

from the degradation of stressed oxytocin in citrate buffered solution.

Figure 4. LC−MS extracted ion currents for dimer degradation products. 

fragmentation spectra only show the tail fragment of m/z 285.2 (Figures S7−10 in the 

Supporting Information) and consequently they are not informative. 

In order to distinguish disulfide linked dimers from other types of dimers, we have 

done the LC−MS analysis on the most degraded solution after disulfide bond reduction 

with DTT. All monomeric products observed were reduced, as summarized in Table 3, 

producing oxytocin, oxytocin acetate, and N-citryl amide/imide in the reduced thiol 

form (Figures S11−S17 in the Supporting Information). Tri- and tetrasulfide forms were 

also absent after reduction and are expected to have been converted to reduced oxytocin. 

The product ion spectrum after DTT reduction of m/z 993.6 at a retention time of 16.6 

min (Figure S18 in the Supporting Information) showed that it is a monomer which has 

undergone β-elimination at Cys 1 and deamidation at Asn 5 or Gln 4 . This is as evidenced by 

the presence of an unmodified y 4 ion and a b 6 ion (Figure S1 in the Supporting Information) 

mass of −33 Da with respect to oxytocin. β-Elimination leads to a loss of 34 Da (-H 2 S), and 


4 

59

4 

60 

Figure 5. Molecular structures of oxytocin N-citryl amide (a), N-citryl imide (b), trisulfide (c) and 

tetrasulfide (d) produced from the degradation of stressed oxytocin in citrate buffered solution. 

deamidation increases the mass by 1 Da. The presence of this reduced monomer at high 

intensity confirms the assignment of dimers 1 and 2, which proves that these dimers are 

disulfide linked at the unmodified Cys 6 residues. 

In the LC−MS profile (Figure 6) it was also found that the LC−MS peaks of all dimers 

had disappeared after reduction. However there were 2 new peaks present at retention 

times of 17.9 and 18.1 min with m/z 994.0 and 993.6 respectively, which are doubly charged 

ion (Figures S19−S20 in the Supporting Information). The presence of dimers after DTT 

reduction, combined with the observation that all original 7 dimer peaks were absent after 

reduction, shows that all dimers had at least one disulfide bond and some had an additional 

thio-ether bond (which may form after β-elimination and reaction with another Cys 

residue). Tyrosine-linked dimers were unlikely to be present, since no monomeric tyrosine 

oxidation products were observed. 

3.2 Effects of Divalent Metal Ions on Oxytocin’s Degradation Profile 

Addition of divalent metal salts in the formulation did not result in any additional degradation 

products compared to samples without metal salts. As expected from our previous study [14], 

it resulted in a significant reduction of most degradation peaks. Figure 7 shows that, of all 

formulations tested, oxytocin formulated in 10 mM citrate buffer (pH 4.5) without divalent 

metal ions (OCB) was most degraded with only approximately 35% oxytocin recovered after 

incubation at 70°C for 5 days. However, when divalent metal ions were added, the degradation 

was decelerated and a recovery of about 70% oxytocin was found, irrespective of the nature of 

the divalent metal ion, which is in line with our previous study [15].

Table 3. Summary of the Observed Degradation Products of Oxytocin after Stressing at 70 °C for 5 Days in 

10 mM Citrate Buffer Followed by Reduction by DTT a 

no. t (min) R m/z charge Assignment 

a 12.8 1007.4 1+ oxytocin (disulfide) 

b 13.1 1200.6b 1+ oxytocin N-citryl amide 

c 13.5 1009.6 1+ oxytocin 

d 14.6 1182.6b 1+ oxytocin N-citryl imide 

e 15.5 1068.8b 1+ N-acetyl oxytocin 

f 16.6 993.6b 1+ β-elimination at Cys1 , monodeamidation at Asn5 or Gln4 g 17.9 994.0 2+ dimer 

h 18.1 993.6 2+ dimer 

a Assigned products are in reduced form. b Ammoniated. 

From each stressed formulation of oxytocin in citrate buffer in the presence of calcium, 

magnesium, or zinc ions, we recorded the intensity of each degradation product from their 

respective extracted MS ion currents. The relative intensities in percentage of the peak area 

of each degradation product in the presence of divalent metal ions with respect to that of 

the same product in the absence of divalent metal ions are listed in Table 4. 

Addition of divalent metal ions reduced the peak intensity of almost all degradation 

products. The formation of the N-citryl oxytocin and the tri- and tetrasulfide was reduced 

by 20−70%. The formation of the dehydrated N-citryl oxytocin and dimmers 1, 2, 4, 5 was 

even more strongly reduced, by 90%. No clear reduction of the intensity of the deamidated 

species (m/z 1008.6, retention time 12.1 min) and dimer 3 (m/z 1024.8, retention time 

17.4 min) was observed, possibly because of their low signal intensity which made peak 

integration difficult (Figures 3b and 4b). 

The acetylated oxytocin species (peak number 7) appeared to be very stable: no 

significant difference in its relative intensity was observed either in the presence or in the 

absence of divalent metal ions. In addition, there is no significant difference of the intensity 

of this product in unstressed or stressed condition. We conclude that this molecule is not a 

degradation product formed under stress conditions but, as mentioned above, an impurity 

of oxytocin formed during synthesis. 

4. dIscussIon 

Several biological and physicochemical methods have been described to monitor 

the stability of oxytocin. HPLC with UV/vis detection is the most frequently used 

physicochemical method to monitor oxytocin stability. Especially when combined with 

mass spectrometry (MS) detection, most degradation products can be identified and 

quantified [16,17]. Hyphenation of the LC unit to MS via an electrospray ionization (ESI) 

interface allows sensitive and selective confirmation of degradation products by extracting 

corresponding ion chromatograms from the recorded total ion current (TIC)[14] and 

by using MS/MS for the identification of degradation products. Furthermore, previous 

studies have demonstrated that LC−MS can be used to monitor the stability of oxytocin in 

pharmaceutical dosage forms [13,18]. 


4 

61

4 

62 

Figure 6. Total ion chromatogram (m/z 900−1200) of stressed oxytocin in citrate buffered 

solution followed by reduction using DTT. 

Figure 7. Recovery of oxytocin in the absence (OCB) and presence of 10 mM Ca 2+ (OCBCa), 

Mg 2+ (OCBMg) and Zn 2+ (OCBZn) in 10 mM citrate-buffered solution at pH 4.5 under 

stressed condition at a temperature of 70 °C for 5 days. Oxytocin recovery determined by 

LC−MS. The results are depicted as averages of three independent measurements±SD. 

The degradation of oxytocin in various formulations was analyzed after 5 days 

of incubation at 70°C. A temperature of 70°C was chosen to ensure the formation of 

qualitatively similar degradation products as found by Poole et al., who incubated oxytocin 

in citrate buffered solutions at 70°C for 2 days [14]. The aim of the present study was to 

investigate the effect of stabilizing divalent metal ions on the degradation profile. 

In order to observe sufficient amounts for characterization of all degradation products in 

the stabilized formulations, it was necessary to prolong the thermal stress from 2 to 5 days. 

The strongly decreased intensity of the dimeric products suggests that formulation with 

divalent metal ions protects against thiol exchange, hence avoiding dimerization. At low 

pH, dimerization occurs via thiol exchange in the disulfide bridge, which progresses via

Table 4. The Effect of Divalent Metal Ions on the Intensity of Degradation Products of Oxytocin Found 

after Stressing at 70°C for 5 Days in 10 mM Citrate Buffer a 

Assignment 

relative peak intensity (%) 

OCBCa OCBMg OCBZn 

oxytocin N-citryl amide 71 ± 19 67 ± 20 52 ± 12 

monodeamidation at Asn5 or Gln 4 534 ± 86 b 436 ± 65 b 906 ± 50 b 

oxytocin N-citryl imide 58 ± 19 73 ±9 50± 8 

oxytocin trisulfide 40 ± 39 37 ±5 30± 28 

oxytocin trisulfide N-citryl imide 3 ±4 4±5 1± 1 

N-acetyl oxytocin c 107 ±7 89± 13 114 ± 13 

oxytocin tetrasulfide 53 ± 64 31 ±5 33± 31 

dimer 1 5 ±1 5±4 2± 1 

dimer 2 7 ±1 10± 7 3± 2 

dimer 3 15 ± 11 14 ±7 6± 4 

dimer 4 172 ± 42 b 150 ± 54 b 90 ± 7 b 

dimer 5 15 ± 11 15 ±5 6± 5 

dimer 6 27 ± 21 17 ±9 17± 16 

dimer 7 43 ± 42 b 42 ± 11 b 31 ± 30 b 

a The level of degradation products is expressed relative to the levels found in the solution without metal ions. 

b Weak MS signal. c No degradation product; compound was also found in the original oxytocin preparation. 

a sulfonium cation, which is formed following protonation of the disulfide bridge [19,20]. 

This could also explain the ability of these formulations to inhibit the formation of tri/ 

tetrasulfide oxytocin species. 

The combination of citrate buffer and divalent metal ions greatly reduces the formation 

of most dimers after thermal stress. The absence of a significant decrease in the intensity 

of dimer 4 hardly contributes to the total amount of dimers because its intensity in the 

formulation without divalent metal ions after thermal stress was already low. 

In this study, the buffer used was citrate, which is considered to be safe and occurs in 

many foods and is also a normal metabolite in the body. Its calcium, potassium and sodium 

salts do not constitute a significant hazard to humans [21]. After thermal stress, covalent 

citrate-oxytocin adducts were formed through a mechanism involving the intermediate 

production of citrate anhydride, which reacts with the N-terminal amino group from the 

cysteine residue [14]. Inhibition of the formation of N-citryl oxytocin by divalent metal 

ions as found in the present study might be due to a differential interaction of oxytocin and 

divalent metal ions for citrate. Wyttenbach et al [22] found that under acidic conditions 

(pH 3.0) divalent metal ions bind to the carbonyl groups in the ring structure of oxytocin. 

The presence of doubly charged cations, such as Zn 2+ , Mg 2+ , Ni 2+ , Mn 2+ and Co 2+ , has been 

found to be essential in increasing the potency of the specific binding of oxytocin to its 

receptor. Therefore the complex formed might increase the biological activity [23]. Further, 

isothermal titration calorimetry data showed that citrate interacts with divalent metal ions, 

and this interaction is stronger than that of citrate with oxytocin (Tables SI and SII and 

Figures S21−S22 in the Supporting Information). Free citrate ions are probably more reactive 

toward the N-terminal amino group from the cysteine residue than divalent metal−citrate 


4 

63

4 

64 

adducts. In the presence of divalent metal ions, the formation of metal−citrate adducts 

reduces the concentration of free citrate, which reduces the driving force for citrate adduct 

formation with oxytocin [15]. A second possibility is that binding of a divalent metal ion to 

oxytocin may change the position of the N-terminal amino group from the cysteine residue 

rendering it less accessible for binding with citrate. 

It can be concluded that citrate has two opposite effects on oxytocin stability. First, as 

also shown by Poole et al.,[14] it is reactive itself and can attack the N-terminal amino group 

from the cysteine residue to form an adduct. Second, as shown in our previous study, [15] 

it protects oxytocin from degradation in the presence of divalent metal ions. In this study, 

we clearly show that the stabilization is due to the suppression of N-citryl oxytocin, tri/ 

tetrasulfide and dimer formation. Furthermore, all reactions that were suppressed occurred 

on Cys 1 , and possibly Cys 6 . Cysteine is susceptible to oxidation and β-elimination, and 

degradation of oxytocin involving cysteine leads to dimerization, formation of tri/ 

tetrasulfide and β-elimination followed by thio-ether formation. 

Divalent metal ions in combination with citrate buffer suppress intermolecular reactions 

in the ring structure of oxytocin presumably by forming a complex in the region where the 

degradation reaction occurs. No significant difference was observed among the three tested 

divalent metal ions, Ca 2+ , Mg 2+ , and Zn 2+ , suggesting that divalency is the most important 

property of the metal contributing to stabilization of the oxytocin-metal-citrate cluster. 

assocIated content 

*Supporting Information 

Figures depicting the MS/MS series of reduced oxytocin, numerous product ion spectra, 

and calorimetric titration profiles and tables of thermodynamics data. 


The authors want to thank MSD Oss for providing oxytocin for the study. This study was 

performed within the framework of the Dutch Top Institute Pharma project: number D6−202. 

references 

1. H.H. Dale, The Action of Extracts of the 

Pituitary Body, Biochem. J. 4 (1909) 427-47. 

2. C.M. Karbiwnyk, K.C. Faul, S.B. Turnipseed, 

W.C. Andersen, K.E. Miller, Determination of 

oxytocin in a dilute IV solution by LC-MS(n), 

J harm Biomed Anal 48 (2008) 672-7. 

3. J. Owen, J.C. Hauth, Oxytocin for the 

induction or augmentation of labor, Clin 

Obstet Gynecol 35 (1992) 464-75. 

4. F.A. Chaibva, R.B. Walker, Development and 

validation of a stability-indicating analytical 

method for the quantitation of oxytocin 

in pharmaceutical dosage forms, J Pharm 

Biomed Anal 43 (2007) 179-85. 

5. J.W. Gard, J.M. Alexander, R.E. Bawdon, 

J.T. Albrecht, Oxytocin preparation stability 

in several common obstetric intravenous 

solutions, Am. J. Obstet. Gynecol. 186 

(2002) 496-8. 

6. R.A. Turner, J.G. Pierce, V. du Vigneaud, The 

purification and the amino acid content of 

vasopressin preparations, J. Biol. Chem. 191 

(1951) 21-8. 

7. H. Davoll, R.A. Turner, J.G. Pierce, V. du 

Vigneaud, An investigation of the free 

amino groups on oxytocin and desulfurized 

oxytocin preparations, J. Biol. Chem. 193 

(1951) 363-70. 

8. M. Winkler, W. Rath, A risk-benefit 

assessment of oxytocics in obstetric practice, 

Drug. Saf. 20 (1999) 323-45.

9. V. du Vigneaud, C. Ressler, S. Trippett, The 



Chem. 205 (1953) 949-57. 





40 (1996) 433-43. 

11. G. Gimpl, F. Fahrenholz, The oxytocin 

receptor system: structure, function, and 

regulation, Physiol. Rev. 81 (2001) 629-83. 










(2009) 1679-88. 

14. R.A. Poole, P.T. Kasper, W. Jiskoot, Formation 

of amide- and imide-linked degradation 

products between the peptide drug oxytocin 

and citrate in citrate-buffered formulations, 

J. Pharm. Sci. 100 (2011) 3018-22. 







16. The United States Pharmacopeial 

Convention,Rockville, USP-29 NF-24, MD, 

2005. (2006). 

17. G.S. Shaw, Synthetic calcium-binding 

peptides, Methods Mol Biol 173 (2002) 175- 

82. 

18. C.W. Huck, V. Pezzei, T. Schmitz, G.K. 

Bonn, A. Bernkop-Schnurch, Oral peptide 

delivery: are there remarkable effects on 

drugs through sulfhydryl conjugation?, J 

Drug Target 14 (2006) 117-25. 




Res. 27 (2010) 544-75. 

20. G. Bulaj, Formation of disulfide bonds 

in proteins and peptides, Biotechnology 

Advances 23 (2005) 87-92. 

21. Joint FAO/WHO Expert Committee on 

Food Additives, Citric acid and its calcium, 

potassium and sodium salts 539 (1974). 

22. T. Wyttenbach, D. Liu, M.T. Bowers, 

Interactions of the hormone oxytocin with 

divalent metal ions, J. Am. Chem. Soc. 130 

(2008) 5993-6000. 

23. A.F. Pearlmutter, M.S. Soloff, Characterization 

of the metal ion requirement for oxytocinreceptor 

interaction in rat mammary gland 

membranes, J Biol Chem 254 (1979) 3899-906. 


4 

65

4 

66 

supportIng InformatIon 

Figure S1. Oxytocin MS/MS series 

ggg 

ggg 

Figure S2. Product ion spectrum of m/z 1198.8 at a retention time of 11.8 min of oxytocin 

N-citryl amide

ggg 

Figure S3. Product ion spectrum of m/z 1007.6 at a retention time of 12.0 min of 

monodeamidated species of oxytocin 

Figure S4. Product ion spectrum of m/z 1007.6 at a retention time of 12.6 min of unmodified 

oxytocin 


4 

67

4 

68 

ggg 


trisulfide 


acetate (ammoniated)

ggg 


dimer 1(ammoniated) 


dimer 3 


4 

69

4 

70 

ggg 


dimer 5 


dimer 6

ggg 

TIC of +Q1: from Sample 14 (july 3 DTT) of Q1.wiff (Turbo Spray) Max. 1.9e7 cps. 

Figure S11. Overlay of UV-RP-HPLC traces of oxytocin in citrate buffer after storage at 70°C 

for 5 days detected at a wavelength of 220 nm before DTT reduction (light gray) and after 

DTT reduction (dark gray). 

2.0e7 

1.8e7 

1.6e7 

1.4e7 

1.2e7 

1.0e7 

8.0e6 

6.0e6 

4.0e6 

2.0e6 

0.0 

no DTT 

with DTT 

6 8 10 12 14 16 18 20 22 

Time, min 

Figure S12. Overlay of LC-MS TIC traces of oxytocin in citrate buffer after storage at 70°C 

for 5 days before and after DTT reduction 


4 

71

4 

72 

ggg 

Figure S13. Product ion spectrum after DTT reduction of m/z 1007.4 at a retention time of 

12.9 min of oxytocin disulfide 


13.1 min of N-citryl oxytocin reduced

ggg 


13.5 min of oxytocin reduced 


14.7 min of N-citryl oxytocin dehydrated (ammoniated) 


4 

73

4 

74 

ggg 


15.6 min of Oxytocin acetate ammoniated 


16.6 min of β-elimination and deamidation at Asn 5 or Gln 4

ggg 


18.0 min of dimer 


18.1 min of dimer 


4 

75

4 

76 

(a) Ca-OT-water (b) Mg-OT-water (c) Zn-OT-water 

Figure S21. Calorimetric titration profiles of aliquots of 125 mM divalent metal ions: 

(a) Ca 2+ , (b) Mg 2+ , and (c) Zn 2+ into 5 mM oxytocin in water pH 4.5. The heat absorbed per 

mol of titrant is plotted versus the ratio of the total concentration of divalent metal ions to 

the total concentration of oxytocin 

(a) Ca-citrate-water (b) Mg- citrate-water (c) Zn-citrate-water 

Figure S22. Calorimetric titration profiles of aliquots of 125 mM divalent metal ions: (a) 

Ca 2+ , (b) Mg 2+ , and (c) Zn 2+ into 10 mM citrate buffer pH 4.5. The heat absorbed per mol 

of titrant is plotted versus the ratio of the total concentration of divalent metal ions to the 

total concentration of oxytocin

Table SI. Thermodynamics of divalent metal binding to oxytocin as determined by isothermal titration 

calorimetry in the absence of buffer. 

Metal Phase N (sites) K a (M -1 ) ΔH° (cal mol -1 ) ΔS° (cal/mol/deg) 

Ca 2+ 

1 

2 

1 

0.27 

30 

300 

Mg2+ No binding observed 

Zn2+ No binding observed 

-4000 

-3516 

The results are depicted as averages of three independent measurements with relative standard deviations 

below 10% 

Table SII. Thermodynamics of divalent metal binding to citrate as determined by isothermal titration 

calorimetry in the absence of oxytocin. 

Metal Phase N (sites) Ka (M -1 ) ΔH° (cal mol -1 ) ΔS° (cal/mol/deg) 

Ca 2+ 

1 

2 

0.3 

1.1 

800 

38 

5700 

-1800 

Mg2+ 1 0.2 300 5800 29 

Zn2+ 1 0.42 540 1600 18 

The results are depicted as averages of three independent measurements with relative standard deviations 

below 10% 


-8 

1 

15 

1.5 

4 

77

Christina Avanti 1 , Wouter L.J. Hinrichs 1 , Angela Casini 2 , Anko C. Eissens 1 , 

Annie van Dam 3 , Alexej Kedrov 4 Arnold J. M. Driessen 4 , Henderik W. Frijlink 1 , 

and Hjalmar P. Permentier 3 

1 Department of Pharmaceutical Technology & Biopharmacy, 

2 Pharmacokinetics, Toxicology and Targeting, Research Institute of Pharmacy, 

3 Mass Spectrometry Core Facility and 

4 Department of Molecular Microbiology, 

University of Groningen, The Netherlands

InsIght Into the stabIlIty of the 

zInc-aspartate-oxytocIn formulatIon 

5

5 

abstract 

The stability of the peptide hormone oxytocin in pharmaceutical formulations during 

prolonged storage at high temperatures is strongly dependent on the solution composition. 

The aim of this study was to investigate the effect of divalent metal ions (Ca 2+ , Mg 2+ 

and Zn 2+ ) on the stability of oxytocin in aspartate buffer (pH 4.5) and determine their 

interaction with the peptide in aqueous solution. Reversed phase and size exclusion-high 

performance liquid chromatography (RP-HPLC and HP-SEC) measurements indicated 

that after 4 weeks of storage at 55°C all tested divalent metal ions improved the stability of 

oxytocin in aspartate buffered solutions (pH 4.5). However, the stabilizing effects of Zn 2+ 

were by far superior compared to Ca 2+ and Mg 2+ . Liquid chromatography-tandem mass 

spectrometry (LC-MS/MS) showed that the combination of aspartate and Zn 2+ in particular 

suppressed the formation of peptide dimers. As shown by isothermal titration calorimetry, 

Zn 2+ interacted with oxytocin in the presence of aspartate buffer while Ca 2+ or Mg 2+ did 

not. In conclusion, the stability of oxytocin in the aspartate buffered-solution is strongly 

improved in the presence of zinc ions, and the stabilization effect is correlated with the 

ability of the divalent metal ions in aspartate buffer to interact with oxytocin. The reported 

results are discussed in relation to the possible mode of interactions between the peptide, 

zinc and buffer components leading to the observed stabilization effects.


Pregnant women may face life-threatening blood loss at the time of delivery. As stated in 

the ICM-FIGO joint statement, the drug of choice to prevent bleeding after child-birth 

(post-partum hemorrhage) is oxytocin [1]. Oxytocin (Figure 1) is a nonapeptide hormone 

that is composed of a cyclic sequence of Cys 1 -Tyr 2 -Ile 3 -Gln 4 - Asn 5 -Cys 6 with an N-terminal 

amino group, and of a linear Pro 7 -Leu 8 -Gly 9 with a C-terminal amide [2]. 

Unfortunately, a major problem in practice is that injectable oxytocin formulations are 

highly unstable as the storage temperature rises to 30°C or higher [3]. Therefore, oxytocin 

should be stored and transported refrigerated, the so-called cold chain, which is not 

always guaranteed especially in rural and tropical areas [4]. At a molecular level oxytocin 

can undergo degradation via deamidation, oxidation or thiol exchange. In particular, the 

stability of the Cys 1 -Cys 6 disulfide bridge with respect to thiol exchange, or to oxidation due 

to the presence of oxygen, light and/or metal ions, is crucial to avoid oxytocin degradation 

and progressive aggregation [5]. 

Several studies have been conducted to improve the stability of oxytocin 

formulations [6-8]. Within this frame, we have previously demonstrated that the use of 

combinations of divalent metal ions with citrate buffer greatly improve the stability of 

oxytocin in aqueous solution, while divalent metal ions added to non-buffered aqueous 

oxytocin formulation have only limited stabilizing effects at 40 or 55°C [7]. In a consecutive 

study, we have shown that formation of a complex of divalent metal ions and citrate with 

oxytocin leads to the suppression of cysteine-mediated inter- or intramolecular reactions, 

thus suppressing tri/tetrasulfide and dimer formation [8]. 

We have also observed that different type of buffers may lead to a different interaction between 

divalent metal ions and oxytocin, therefore having a different impact on oxytocin stability. For 

example, we demonstrated that addition of divalent metal ions to oxytocin solutions in either 

acetic or citric acid buffers results in different stabilization effects of the peptide in aqueous 

solution [7]. While Ca 2+ , Mg 2+ and Zn 2+ ions in combination with citrate buffer were successful in 

Figure 1. Oxytocin structure. 

StabIlIty of the dIvalent Metal -aSpartate-oxytocIn coMplex 

5 

81

5 

82 

stabilizing oxytocin, no stabilizing effect was observed for the combination of the same divalent 

metal ions and acetate buffer [7]. Acetic acid and citric acid are carboxylic acids with one and 

three carboxylate groups, respectively. Thus, the above mentioned results suggested that one 

carboxylate group is not sufficient for stabilizing the oxytocin-metal-buffer salt cluster. 

The aim of the present study is to investigate the effects of divalent metal ions (Ca 2+ , 

Mg 2+ and Zn 2+ ) on the stability of oxytocin in aspartate buffer. Aspartate is a commonly 

used buffer in parenteral products approved by the FDA for formulation purposes [9]. It 

has one amine and two carboxylate groups, which may establish different interactions and 

affect differently the peptide stability with respect to citric acid. 

It should be noted that both the pH and concentration of metal ions in the formulation 

were found to be crucial in the effectiveness of oxytocin stabilization in citrate buffer [7]. 

Thus, since the optimum stability of oxytocin was reported to be at pH 4.5 [6], the peptide 

formulations were maintained at that pH, while various divalent metal ion concentrations 

were tested. Afterwards, we have also investigated the effect of divalent metal ions on the 

degradation profile of oxytocin in aspartate buffer by liquid chromatography-tandem mass 

spectrometry (LC-MS/MS), as well as the thermodynamics of the oxytocin-metal-buffer 

system by isothermal titration calorimetry (ITC). 


2.1. Materials 

Oxytocin monoacetate powder (Diosynth. Oss, The Netherlands) was kindly provided by MSD, 

Oss, The Netherlands. L-aspartic acid, magnesium chloride, and zinc chloride were purchased 

from Fluka, Steinheim, Germany. Calcium chloride was from Riedel-de Haen, Seelze, Germany, 

and sodium hydroxide, sodium dihydrogen phosphate dihydrate, acetonitrile (supergradient 

grade), as well as formic acid were purchased from Merck, Darmstadt, Germany. 

2.2. Methods 

2.2.1. Formulation of oxytocin solution and stability study 

Oxytocin solution was formulated at a concentration of 0.1 mg/mL (0.094 mM) in 10 mM 

aspartate buffer at pH 4.5 (pH adjusted with sodium hydroxide) with different concentrations 

of divalent metal ions Ca2+ , Mg2+ and Zn2+ . All divalent metal ion solutions were prepared 

using their chloride salts at concentrations of 2, 5, 10 and 50 mM. The concentration of 

oxytocin was determined using a UV spectrometer as described previously [7,10]. After 

preparation, the solutions were stored in 6R glass type 1 vials for 4 weeks at either 4 or 

55°C, and protected from light. Some selected formulations were also stored for 5 days at 

70°C and the samples were diluted 10-fold with water for LC-MS/MS analysis. During the 

stability study controlled pH levels in the samples were within 0.1 pH units. 

It must be noted that oxytocin is commonly formulated at very low concentration, which 

is about 1/10 of the concentration within this study (RP-HPLC and HP-SEC). The higher 

concentration was chosen to have better intensity of signal for the degradation products 

produced by the heat stress. 

2.2.2. Reversed-Phase High-Performance Liquid Chromatography (RP-HPLC) 

RP-HPLC was carried out according to the procedure described earlier [6,7]. An Alltima 

C-18 RP column with 5 μm particle size, inner diameter of 4.6 mm, and length of 150 mm

(Alltech, Ridderkerk, Netherlands), a Waters (Millipore) 680 Automated Gradient 

Controller, two Waters 510 HPLC pumps, a Waters 717 Plus Autosampler, and a Waters 

486 Tunable Absorbance UV Detector were used. Samples of 20 μL were injected and 

the separation was carried out at a flow rate of 1.0 mL/min and UV detection at 220 nm. 

Samples were eluted using 15% (v/v) acetonitrile in 65 mM phosphate buffer, pH 5.0 as 

solvent A and 60% (v/v) acetonitrile in 65 mM phosphate buffer, pH 5.0 as solvent B. The 

acetonitrile concentration was linearly increased from 15% at the beginning, to 20% at 10 

min, to 30% at 20 min and finally to 60% at 25 min. The recovery of oxytocin is expressed 

as the percentage of initial amount. 

2.2.3. Size Exclusion HPLC (HP-SEC) 

HP-SEC was performed according to the method previously reported [6,7]. A Superdex 

Peptide 10/300 GL column (GE Healthcare Inc., Brussels, Belgium) was used on an 

isocratic HPLC system with a Waters 510 pump, a Waters 717 plus auto sampler, a Waters 

474 Scanning Fluorescence Detector and Waters 484 Tunable Absorbance Detector 

(Waters, Milford Massachusetts, USA). Samples of 50 μL were injected, and separation was 

performed at a flow rate of 1 mL/min. Chromatograms were obtained using fluorescence 

detection at excitation wavelength of 274 nm and emission wavelength of 310 nm. The 

mobile phase consisted of 30% acetonitrile and 70% 0.04 M formic acid. The recovery of 

monomeric oxytocin is expressed as the percentage of initial amount. 

2.2.4. Liquid Chromatography-Mass Spectrometry/Mass Spectrometry 

(LC-MS/ MS) 

The LC-MS/MS system was set up according to the method described earlier [6,8]. A 

Shimadzu LC system equipped with LC-20AD gradient pumps and a SIL-20AC autosampler 

was used. The chromatographic separation was carried out on an Alltima C18 column 

(internal diameter 2.1 mm, length 150 mm, particle size 5 µm, Grace Davison Discovery 

Sciences). The gradient mobile phase composition was a mixture of solvent A consisting of 

95% water / 5% acetonitrile and solvent B consisting of 5% water / 95% acetonitrile, both 

containing 0.05% (v/v) acetic acid and 10 mM ammonium acetate. Elution was performed 

by a linear gradient from 5 to 60% of the solvent B in 30 min, followed by an increase to 

90% solvent B in 1 min, where it was kept 4 min, after which it returned to the starting 

conditions. The flow rate was 0.2 mL/min. The UV signal was recorded at 220 nm. The 

injection volume was 50 μL. 

The HPLC system was coupled to an API 3000 triple-quadrupole mass spectrometer 

(Applied Biosystems/MDS Sciex) via a Turbo Ion Spray source. The ionization was 

performed by electrospray in the positive mode. Full scan spectra were recorded at a scan 

rate of 2 s from m/z 500 to 1300 and a step size of 0.2 amu with a Declustering Potential 

(DP) of 40 V and a Focusing Potential (FP) of 250 V. Product ion scans were acquired in 

specified time windows with a DP of 60 V, a FP of 300 V, a Collision Energy of 40 V and a 

Collision Cell Exit Potential of 20 V. Data acquisition and processing was performed using 

Analyst version 1.5 software (Applied Biosystems/MDS Sciex). 

LC-MS/MS was used to identify the oxytocin degradation products observed by 

RP-HPLC with UV detection. The analyses were carried out on each of the major degradation 

peaks for oxytocin in aspartate buffer formulation without divalent metal salts as well as by 

LC-MS for the formulation with divalent metal salts. 


5 

83

5 

84 

2.2.5. Isothermal Titration Calorimetry (ITC) 

Microcalorimetric titrations of divalent metal ions to oxytocin in aspartate buffer were 

performed by using a MicroCal ITC 200 microcalorimeter (Northampton, MA 01060 USA) 

as described previously [7]. A solution of 30 µL of 125 mM divalent metal chloride (calcium, 

magnesium, or zinc chloride) in 10 mM aspartate buffer, pH 4.5 was placed in the syringe, while 

300 µL of 5 mM oxytocin in 10 mM of aspartate buffer, pH 4.5 was placed in the sample cell. 

The reference cell contained 300 µL of aspartate buffer. Experiments were performed at 55°C. 

The effective heat of the peptide-metal ion interaction upon each titration step was corrected 

for dilution and mixing effects, as measured by titrating the divalent metal ion solution into 

the buffer and by titrating the buffer into the oxytocin solution (reference measurement). To 

investigate the possibility of oxytocin or metal ion binding to the buffer components, control 

experiments were performed in water. The heats of bimolecular interactions were obtained by 

integrating the peak following each injection. All measurements were performed in triplicate. 

ITC data were analyzed by using the ITC non-linear curve fitting functions for one or two 

binding sites from Origin 7.0 software (MicroCal Software, Inc.). 

3. results 

3.1. The effect of divalent metal ions on oxytocin stability in 

aspartate buffer solution 

To investigate the stability of oxytocin in the aspartate buffered solutions in the presences of 

divalent metal ions, oxytocin formulation in 10 mM aspartate buffer (pH 4.5) with various 

concentrations of divalent metal ions were prepared. Oxytocin is commonly formulated at 

a concentration of about 0.01 mg/mL, which is 10 times lower than the concentration used 

in this study. The higher concentration was chosen to have better intensity of degradation 

product produced by the heat stress in RP-HPLC and HP-SEC analysis. Oxytocin recovery 

and the presence of oxytocin monomer were determined for all the samples after 4 weeks at 

different storage temperatures by RP-HPLC and HP-SEC, respectively, as described in the 

experimental section. The obtained results are shown in Figure 2. 

After 4 weeks of storage at 4°C, oxytocin remained stable in all formulations. Instead, 

after 4 weeks of storage at 55°C oxytocin stability increased with increasing divalent metal 

ions concentration. Ca 2+ and Mg 2+ had similar effects on improving the oxytocin stability: 

the recoveries of oxytocin, as well as the remaining percentage of oxytocin monomer were 

increased up to 45% in the presence of 50 mM Mg 2+ , and up to 35% in the presence of 50 

mM Ca 2+ . Notably, Zn 2+ was much more effective in stabilizing oxytocin: at a concentration 

of 2 mM Zn 2+ increased oxytocin recovery up to 35% (i.e. to a comparable extent as 50 

mM Ca 2+ or Mg 2+ ), and almost ca. 75% of oxytocin monomer was recovered. Overall, both 

RP-HPLC and HP-SEC showed that Zn 2+ has superior stabilizing effects in aspartate buffer 

compared to Ca 2+ and Mg 2+ 

3.2. The effect of divalent metal ions on the degradation profile of 

oxytocin in aspartate buffer 

To assign the degradation products of oxytocin in the formulations, oxytocin in10 mM 

aspartate buffer (pH 4.5) in the absence and presence of 10 mM divalent metal ions was 

analyzed by LC-MS/MS after incubation of the samples at 70°C for 5 days to ensure the

. 

Figure 2 Effect of Ca 2+ (squares), Mg 2+ (circles), and Zn 2+ Figure 2. Effect of Ca 

(triangles) concentration on the 

recovery of oxytocin in 10 mM aspartate buffer, pH 4.5, after 4 weeks of storage at either 4°C 

(solid symbols) or 55°C (open symbols). A: Oxytocin recovery as determined by RP-HPLC. 

B: Oxytocin monomer remaining as determined by HP-SEC. The results are depicted as 

averages of three independent measurements ± SD 

2+ (squares), Mg2+ (circles), and Zn2+ (triangles) concentration on 

the recovery of oxytocin in 10 mM aspartate buffer, pH 4.5, after 4 weeks of storage at 

either 4°C (solid symbols) or 55°C (open symbols). A: Oxytocin recovery as determined 

by RP-HPLC. B: Oxytocin monomer remaining as determined by HP-SEC. The results are 

depicted as averages of three independent measurements ± SD 

formation of sufficient degradation products intensity. MS/MS analysis was applied to 

confirm the identification of the most intense peaks, and degradation products were assigned 

following our previous reports [6,8,11]. Figure 3A shows the total ion-current (TIC) of mass 

spectra in the m/z range between 900 and 1200 for formulation of oxytocin without divalent 

metal ions (solid line), and with 10 mM Zn2+ After 4 weeks of storage at 4°C, oxytocin remained stable in all formulations. Instead, 

after 4 weeks of storage at 55°C oxytocin stability increased with increasing divalent 

metal ions concentration. Ca (dashed line). Spectra of formulations with 

calcium and magnesium ions are reported in the supplementary material available. 

Nine peaks are observed in the MS profile, which revealed 10 molecular species in the 

LC-MS contour plot (Figure 3B). The protonated molecular ion of unmodified oxytocin was 

found at a retention time of 12.8 min with m/z of 1007.4. A peak at a retention time of 14.5 

min with m/z 1066.6 corresponded to oxytocin acetate (ammoniated), but also appeared in 

the LC-MS of the unstressed oxytocin preparation. Most likely, oxytocin acetate was formed 

during synthesis as the oxytocin is supplied as a monoacetate salt. Complete assignment of 

the degradation products is listed in Table 1. At retention time of 13.7 and 15.0 min tri- and 80 

tetrasulfide form degradation products were identified from the m/z 1039.6 and 1071.6 

peaks, respectively (see supplementary material for the structures of tri- and tetrasulfide 

oxytocin species). Other peaks (Dimers 1 to 6 in Table 1) are interpreted as various forms of 

disulfide or thioether-linked oxytocin dimers as described previously [8]. 

2+ and Mg 2+ had similar effects on improving the 

oxytocin stability: the recoveries of oxytocin, as well as the remaining percentage of 

oxytocin monomer were increased up to 45% in the presence of 50 mM Mg 2+ , and up 

to 35% in the presence of 50 mM Ca 2+ . Notably, Zn 2+ was much more effective in 

stabilizing oxytocin: at a concentration of 2 mM Zn 2+ increased oxytocin recovery 

up to 35% (i.e. to a comparable extent as 50 mM Ca 2+ or Mg 2+ ), and almost ca. 75% 

of oxytocin monomer was recovered. Overall, both RP-HPLC and HP-SEC showed 


5 

85

5 

86 

Figure 3. A: Total ion chromatogram (m/z 900-1200) and B: Contour plot of oxytocin and 

its degradation products in 10 mM aspartate buffer after 5 days of storage at 70°C and pH 

4.5 (solid-line) and in the presence of 10 mM Zn 2+ (dashed-line). 

Addition of divalent metal salts in the formulation with the aspartate buffer did not 

result in any additional degradation products compared to samples without metal salts (see 

dashed line in Figure 3A for zinc salt addition). Conversely, as expected from the RP-HPLC 

and HP-SEC data, it resulted in a significant reduction of the major degradation peaks. 

The intensity of each degradation product from each stressed formulation of oxytocin in 

aspartate buffer in combination with Ca 2+ , Mg 2+ , and Zn 2+ was also recorded. Table 1 reports 

the relative intensities in percentage of the peak area of each degradation product in the 

presence of divalent metal ions with respect to the same formulations without divalent 

metal ions.Addition of Ca 2+ and Mg 2+ reduced the formation of dimer 1 and 2 of about 30%, 

an effect that was even more marked in the presence of Zn 2+ : i.e., 53 and 60% reduction for 

dimer 1 and 2, respectively. 

Zinc is the most efficient element in suppressing the total amount of the degradation 

products of oxytocin in aspartate buffer. In fact, at variance with Ca 2+ and Mg 2+ , Zn 2+ also 

reduced the formation of other dimers: dimer 3 was reduced by 14% and dimer 5 was 

reduced by 30%. However, an increase in the formation of tri and tetrasulfide species was

Table 1. Effect of divalent metal ions on the intensity of oxytocin in aspartate buffered solution. Relative 

peak intensity is expressed with respect to that of the same product in the absence of divalent metal ions. 

Assignment 

retention 

time (min) m/z 

Relative peak intensity of oxytocin a (%) 

Ca 2+ Mg 2+ Zn 2+ 

oxytocin trisulfide 13.7 1039.4 111 ± 4 101 ± 8 146 ± 5 

Oxytocin tetrasulfide 15.0 1071.6 135 ± 13 114 ± 10 187 ± 12 

dimer 1 15.7 993.2b 68 ± 3 68 ± 2 47 ± 0 

dimer 2 16.7 993.2b 69 ± 3 76 ± 2 39 ± 2 

dimer 3 16.9 1008.6 115 ± 4 111 ± 2 86 ± 2 

dimer 4 17.6 1024.6b 147 ± 19 135 ± 3 127 ± 5 

dimer 5 18.0 1008.6 111 ± 2 111 ± 1 71 ± 3 

dimer 6 18.9 1024.6b 152 ± 8 144 ± 3 115 ± 18 

a Aspartate buffer solution with 10 mM of the indicated divalent metal ion 

b ammoniated. 

Figure 4. Recovery of oxytocin in the absence (OAP) and presence of 10 mM Ca 2+ (OAPCa), 

Mg 2+ (OAPMg) and Zn 2+ (OAPZn) in 10 mM aspartate-buffered solution at pH 4.5 under 

stressed condition at a temperature of 70 °C for 5 days. Oxytocin recovery was determined 

by LC-MS. The results are depicted as averages of three independent measurements ± SD 

also observed. The signal intensity of the tetrasulfide form (m/z 1071.6, retention time 

15.0 min) is very low compared to that of the dimers, therefore peak integration is less 

accurate. We have no explanation for the increased intensity of the trisulfide form (m/z 

1039.4 retention time 13.7 min) in the presence of Zn 2+ . 

Oxytocin formulated in aspartate buffer after 5 days at 70°C without divalent metal ions 

was the most degraded of all formulations tested, and only approximately 20% oxytocin 

could be recovered (Figure 4). Interestingly, addition of 10 mM Ca 2+ and Mg 2+ did not 

markedly improve the peptide recovery. However, when Zn 2+ was added, the degradation 

was reduced and a recovery of about 45% oxytocin was found. 


5 

87

3.3. Interaction of oxytocin and divalent metal ions in aspartate 

buffer 

The interaction of oxytocin with divalent metal ions was investigated by ITC at a temperature 

of 55°C. Figure 5 shows the exothermic (DH obs ) events when Ca 2+ or Mg 2+ was titrated into 

a solution of oxytocin in 10 mM aspartate buffer. However, the effects were very small (not 

more than 0.05 kcal/mole of injectant) and only slightly different from the corresponding 

reference measurements. When Zn2+ was titrated into oxytocin in aspartate buffer, a strong 

endothermic binding reaction was observed and the heat effects upon the titration reached 

more than 0.5 kcal/mole for the first injection. The shape of the titration curve for the 

Zn2+ -oxytocin in aspartate is indicative for a binding reaction. Using the analysis model 

5 

with two distinct types of binding sites, the binding constants K at 55°C for Zn a 2+ -oxytocin 

interaction are 6.2 ± 1.0 × 103 M-1 and 72 ± 17 M-1 Mg 

, respectively. Thus, the only system that 

shows thermodynamics interactions has the most pronounced effect on oxytocin stability 

in formulations. 

2+ was titrated into a solution of oxytocin in 10 mM aspartate buffer. However, the 

effects were very small (not more than 0.05 kcal/mole of injectant) and only slightly 

different from the corresponding reference measurements. When Zn 2+ was titrated 

into oxytocin in aspartate buffer, a strong endothermic binding reaction was observed 

and the heat effects upon the titration reached more than 0.5 kcal/mole for the first 

injection. The shape of the titration curve for the Zn 2+ -oxytocin in aspartate is 

indicative for a binding reaction. Using the analysis model with two distinct types of 

binding sites, the binding constants Ka at 55°C for Zn 2+ -oxytocin interaction are 6.2 ± 

4. dIscussIon 

1.0 × 10 

In a previous study, we have found that oxytocin can be stabilized by a combination of divalent 

metal ions and citrate buffer at pH 4.5. Divalent metal ions in combination with citrate 

buffer suppressed intermolecular reactions in the ring structure of oxytocin presumably by 

3 M -1 and 72 ± 17 M -1 , respectively. Thus, the only system that shows 

thermodynamics interactions has the most pronounced effect on oxytocin stability in 

formulations. 

Figure 5 Least squares fit of the data from calorimetric titration profiles of aliquots of 125 

mM divalent metal ions: Ca 2+ (solid square), Mg 2+ (open square), and Zn 2+ Figure 5. Least squares fit of the data from calorimetric titration profiles of aliquots of 125 

mM divalent metal ions: Ca 

(open triangle) 

into 5 mM oxytocin in 10 mM aspartate buffer pH 4.5. The heat absorbed per mole of titrant 

is plotted versus the ratio of the total concentration of divalent metal ions to the total 

concentration of oxytocin. 

2+ (solid square), Mg2+ (open square), and Zn2+ (open triangle) 

into 5 mM oxytocin in 10 mM aspartate buffer pH 4.5. The heat absorbed per mole of 

titrant is plotted versus the ratio of the total concentration of divalent metal ions to the total 

concentration of oxytocin. 

88 

4 Discussion

forming a complex in the disulfide bridge region where the degradation reaction occurred. 

There were no significant differences observed among the three tested divalent metal ions, 

Ca 2+ , Mg 2+ , and Zn 2+ [7]. Our present study indicates that formulations with a combination 

of divalent metal ions and aspartate buffer behave differently in improving oxytocin 

stability in aqueous solution. In fact, we have shown that in aspartate buffer, only addition 

of Zn 2+ results in a comparable stabilizing effect on oxytocin as citrate with Ca 2+ , Mg 2+ and 

Zn 2+ [7]. The LC-MS/MS results show that formation of oxytocin dimers is hampered by the 

presence of Zn 2+ in aspartate, while it is not so affected by Ca 2+ and Mg 2+ . A previous study 

by us [8] showed that all the dimers are produced from the thiol exchange in the disulfide 

bridge, and there are no additional dimers produced from this formulation. Thus, it appears 

that the combination of Zn 2+ and aspartate was able to protect the disulfide Cys 1,6 bridge on 

oxytocin. In line with these results, ITC data demonstrate that only zinc, among the tested 

divalent metal ions, is able to strongly interact with oxytocin in the formulation conditions. 

The observed stabilization effects might be ascribed to the formation of divalent metal 

ion adducts to oxytocin. In fact, several studies reported on the ability of Ca 2+ , Mg 2+ and 

Zn 2+ to form complexes with oxygen atoms from the carbonyl backbone of the peptide, but 

the strength of the interactions is different [12-14]. The fact that the three metals have a 

smaller ionic radius than the oxytocin ring [15,16] suggested that Ca 2+ , Mg 2+ and Zn 2+ are 

located within the ring of oxytocin, and adduct formation, arising from coordination of the 

metal ions via the backbone carbonyl oxygens of oxytocin, induces a compact structure of 

an oxytocin-metal octahedral complex [12,17,18]. 

Most importantly, the binding affinity of oxytocin for Mg 2+ has been reported to be lower than 

in the case of Ca 2+ and far below Zn 2+ [12,17,18]. Indeed, Zn 2+ ions very effectively coordinate 

with oxytocin and strongly affect its conformation in physiological environment [18]. 

The oxytocin stabilizing effect of Zn 2+ with respect to Ca 2+ and Mg 2+ might be due to the 

higher stability of the oxytocin-Zn 2+ adduct in the reported formulation conditions. 

In addition, our results indicate that also the type of buffer used in the formulations 

plays an important role in the peptide stabilization effects. At a pH of 4.5, citrate has one 

carboxylate ion (–COO - ) in α position and other two carboxylates that can contribute to 

the coordination of a metal or to binding to oxytocin itself. At the same pH of 4.5, aspartate 

has only two carboxylate ions that could act as electron donors towards a metal ion or 

oxytocin. This difference of available electron donor groups between the two buffers might 

influence the oxytocin-metal adduct formation, as well as its stabilization. The lack of an 

additional electron donor oxygen moiety in aspartate is likely to discriminate the binding of 

Ca 2+ or Mg 2+ with respect to Zn 2+ , as suggested by the ITC results, which did not show any 

significant interactions between oxytocin and Ca 2+ or Mg 2+ in aspartate buffer. 

5. conclusIon 

Our study clearly shows that the stability of oxytocin in the aspartate buffered-solution is 

strongly improved in the presence of zinc ions, and the stabilization effect is correlated with 

the strength of interaction between oxytocin and divalent metal ions. Further studies using 

NMR and molecular modeling are initiated to characterize the mechanisms of oxytocin 

stabilization at a molecular level. We can conclude that Zn 2+ binding to peptide in aspartate 

solution suppress intermolecular degradation reactions near the Cys 1,6 disulfide bridge that 

is responsible for oxytocin degradation. 


5 

89

acknowledgement 

The authors want to thank MSD, Oss, The Netherlands for providing oxytocin for the study. 

This research was performed within the framework of the Dutch Top Institute Pharma 

(project number D6–202) and Rosalind Franklin fellowship funding for Angela Casini. 

references 

1. International Confederation of Midwives, 

International Federation of Gynaecology 

and Obstetrics. Joint statement management 

of the third stage of labour to prevent 

post-partum haemorrhage, Http://www. 

Internationalmidwives. org/Whatwedo/ 

P ro g r a m m e s / P O P P H I / Po s t Pa r t u m - 

Haemorrhage/tabid/339/Default. Aspx 2012 

(2012). 

2. A. Ohno, N. Kawasaki, K. Fukuhara, H. 

Okuda, T. Yamaguchi, Complete NMR 

analysis of oxytocin in phosphate buffer, 

Magn. Reson. Chem. 48 (2010) 168-72. 





(2002) 496-8. 

4. H.V. Hogerzeil, G.J. Walker, Instability of 

(methyl)ergometrine in tropical climates: an 

overview, Eur. J. Obstet. Gynecol. Reprod. 

Biol. 69 (1996) 25-9. 




Res. 27 (2010) 544-75. 






(2009) 1679-88. 









Hinrichs, A New Strategy To Stabilize 

Oxytocin in Aqueous Solutions: II. 

Suppression of Cysteine-Mediated 

Intermolecular Reactions by a Combination 

of Divalent Metal Ions and Citrate, Mol. 

Pharmaceutics 9 (3) (2012) 554-62. 

9. P. Jurgens, C. Panteliadis, G. Fondalinski, 

Total parenteral nutrition of premature 

infants: metabolic effects of an exogenous 

supply of L-aspartic acid and L-glutamic 

acid, Z. Ernahrungswiss. 21 (1982) 225-45. 

10. [10] S.C. Gill, P.H. von Hippel, Calculation 

of protein extinction coefficients from 

amino acid sequence data, Anal. Biochem. 

182 (1989) 319-26. 

11. R.A. Poole, P.T. Kasper, W. Jiskoot, Formation 

of amide- and imide-linked degradation 

products between the peptide drug oxytocin 

and citrate in citrate-buffered formulations, 

J. Pharm. Sci. 100 (2011) 3018-22. 





40 (1996) 433-43. 

13. J.P. Glusker, A.K. Katz, C.W. Bock, Metal ions 

in biological systems, The Rigaku Journal 16 

(1999) 8-19. 

14. H. Einspahr, C.E. Bugg, The Geometry 

of Calcium-Carboxylate Interactions in 

Crystalline Complexes, Acta Cryst. B37 

(1981) 1044-52. 

15. A.K. Katz, J.P. Glusker, S.A. Beebe, C.W. Bock, 

Calcium Ion Coordination: A Comparison 

with That of Beryllium, Magnesium, and 

Zinc, J. Am. Chem. Soc. 118 (1996) 5752-63. 

16. R.H. Holm, P. Kennepohl, E.I. Solomon, 

Structural and Functional Aspects of Metal 

Sites in Biology, Chem. Rev. 96 (1996) 2239- 

314. 

17. V.S. Ananthanarayanan, M.P. Belciug, 

B.S. Zhorov, Interaction of oxytocin with 

Ca2+: II. Proton magnetic resonance 

and molecular modeling studies of 

conformations of the hormone and its Ca2+ 

complex, Biopolymers 40 (1996) 445-64. 

18. D. Liu, A.B. Seuthe, O.T. Ehrler, X. Zhang, 

T. Wyttenbach, J.F. Hsu, M.T. Bowers, 

Oxytocin-receptor binding: why divalent 

metals are essential, J. Am. Chem. Soc. 127 

(2005) 2024-5.

5 

92 


Figure S1. Total ion chromatogram (m/z 900-1200) of oxytocin and its degradation 

products in 10 mM aspartate buffer after 5 days of storage at 70°C and pH 4. in the 

presence of 10 mM Ca 2+ . 

Figure S2. Total ion chromatogram (m/z 900-1200) of oxytocin and its degradation 

products in 10 mM aspartate buffer after 5 days of storage at 70°C and pH 4. in the 

presence of 10 mM Ca 2+ .

Figure S3. Structure of oxytocin trisulfide 

Figure S4. Structure of oxytocin tetrasulfide 


5 

93

Christina Avanti 1,* , Nur Alia Oktaviani 2,* , Wouter L.J. Hinrichs 1 , 

Henderik W. Frijlink 1 , and Frans A.A. Mulder 2,3 

# Equally contributed first author 


2 Groningen Biomolecular Sciences and Biotechnology Institute, 


3 Department of Chemistry and Interdisciplinary Nanoscience Center iNANO, 

University of Aarhus, Aarhus C, Denmark

aspartate buffer and dIvalent 

metal Ions affect the oxytocIn 

conformatIon In aqueous solutIon 

and protect It from degradatIon 

6

6 

abstract 

Oxytocin is a peptide drug used to induce labor and prevent bleeding after child birth. 

Due to its instability, transport and storage of oxytocin formulations under tropical 

condition is problematic. In a previous study, we described that the stability of oxytocin 

in aspartate buffered formulation is improved by the addition of divalent metal ions. The 

stabilizing effect of Zn 2+ was by far superior compared to that of Mg 2+ . In addition, it was 

found that stabilization correlated well with the ability of the divalent metal ions to interact 

with oxytocin in aspartate buffer. Furthermore, LC-MS (MS) measurements indicated 

that the combination of aspartate buffer and Zn 2+ in particular suppressed intermolecular 

degradation reactions near the Cys 1,6 disulfide bridge. These results lead to the hypothesis 

that in aspartate buffer, Zn 2+ changes the conformation of oxytocin in such a way that the 

Cys 1,6 disulfide bridge is shielded from its environment thereby suppressing intermolecular 

reactions involving this region of the molecule. 

To verify this hypothesis, in this study the conformation of oxytocin in aspartate buffer 

in the presence of Mg 2+ or Zn 2+ , was investigated by 2D NMR spectroscopy, i.e. NOESY, 

TOCSY, 1 H- 13 C HSQC and 1 H- 15 N HSQC. Almost all 1 H, 13 C and 15 N resonances could 

be assigned using HSQC spectroscopy of oxytocin without 13 C or 15 N enrichment. 1 H- 13 C 

and 1 H- 15 N HSQC spectra showed that aspartate buffer alone induces a minor change in 

oxytocin in D 2 O with the largest chemical shift changes are observed in Cys 1 . Zn 2+ causes 

more extensive changes in oxytocin in aqueous solution than Mg 2+ . Our findings suggest 

that the carboxylate group of aspartate neutralizes the positive charge of the N terminus of 

Cys 1 , allowing the interactions with Zn 2+ to become more favorable. These interactions may 

explain the protection of the disulfide bridge against intermolecular reactions that lead to 

dimerization and inactivation.


Oxytocin is a nonapeptide hormone secreted by the posterior lobe of the pituitary gland 

which is involved in the control of labor and bleeding cessation after child birth [1]. The 

peptide consists of nine amino acids (Cys-Tyr-Ile-Gln-Asn-Cys-Pro-Leu- Gly) and an 

amidated C-terminus [2]. Oxytocin is the preferred drug to prevent postpartum hemorrhage 

and commonly formulated in aqueous solution for parenteral administration [3]. Instability 

of oxytocin in aqueous solution under severe conditions, particularly in tropical conditions, 

presents a significant challenge for pharmaceutical scientists [4]. Oxytocin instability in 

aqueous solution has been reported in several studies [5,6] and the degradation strongly 

depends on the pH of the formulation, with the highest stability reported at pH 4.5 [4]. 

Several studies have been aimed at the improvement of the stability of oxytocin in aqueous 

solution [4,7]. The most recent finding is the use of divalent metal ions in combination with 

certain buffers that strongly increases the stability of oxytocin in aqueous solution [8,9]. 

In a previous study [9] we found that Zn 2+ in combination with aspartate buffer strongly 

stabilizes oxytocin in aqueous solutions, while Ca 2+ and Mg 2+ only have minor effects. The 

stabilization occurred as a result of complex formation of Zn 2+ ions and aspartate with 

oxytocin, which, by protecting the Cys 1,6 disulfide bridge, suppressed dimerization. In line 

with those results, isothermal titration calorimetry data demonstrate that only Zn 2+ ions, 

among the tested divalent metal ions (Zn 2+ , Ca 2+ , and Mg 2+ ) is able to strongly interact with 

oxytocin in the formulation conditions [9]. Those results lead to the hypothesis that Zn 2+ 

induces a conformational change, thereby stabilizing oxytocin in aspartate buffer. Aspartic 

acid is one of the non-essential amino acids which normally synthesized in the body. It 

consists of two carboxylate groups with pKa 1 and pKa 2 of 2.1 and 3.9, and one amine group 

(pKa 3 of 9.8). Aspartate is a commonly used buffer in parenteral products approved by 

the FDA for formulation purposes [10]. To investigate the conformation of oxytocin in 

aspartate buffer in the presence of divalent metal ions (Zn 2+ and Mg 2+ ), two-dimensional 

Nuclear Magnetic Resonance (NMR) spectroscopy was used. 

Nuclear Magnetic Resonance spectroscopy is a suitable technique to study the 

conformational details of proteins or peptides in solutions. Several one-dimensional NMR 

studies of oxytocin and vasopressin analogs have previously been performed in various 

solvents such as dimethyl sulfoxide [13], deuterated dimethylsulfoxide [14,15], deuterated 

trifuoroethanol [16,17], and aqueous solutions [14,18,19]. Since resonance overlap is much 

reduced in two-dimensional NMR spectra in comparison with one-dimensional NMR, we 

used 2D NMR spectroscopy. 

Our most recent study [9] showed that Zn 2+ in combination with aspartate buffer 

strongly stabilizes oxytocin in aqueous solutions, while Ca 2+ and Mg 2+ only have minor 

effects. The stabilization occurred as a result of complex formation of Zn 2+ ions and 

aspartate with oxytocin, which, by protecting the Cys 1,6 disulfide bridge, suppressed 

dimerization. In line with those results, ITC data demonstrate that only Zn 2+ ions, among 

the tested divalent metal ions (Zn 2+ , Ca 2+ , and Mg 2+ ) is able to strongly interact with 

oxytocin in the formulation conditions [9]. Those results lead to the hypothesis that Zn 2+ 

ions induce a different conformation, thereby stabilizing oxytocin in aspartate buffer. To 

investigate the conformation of oxytocin in aspartate buffer in the presence of divalent 

metal ions (Zn 2+ and Mg 2+ ), two-dimensional Nuclear Magnetic Resonance (NMR) 

spectroscopy has been used. 

conforMatIon of dIvalent Metal-aSpartate-oxytocIn coMplex 

6 

97

6 

98 

Nuclear Magnetic Resonance spectroscopy is a suitable technique to study 

conformational detail of proteins or peptides in solution. Several one-dimensional NMR 

studies of oxytocin and vasopressin analogs have been previously performed in various 

solvents such as dimethyl sulfoxide [13], deuterated dimethylsulfoxide [14,15], deuterated 

trifuoroethanol [16,17], and water [14,18,19]. Since resonance overlap is much reduced 

in two-dimensional NMR spectra in comparison with one-dimensional NMR, we used 

2D NMR spectroscopy. 

Oxytocin solutions are commonly formulated in the concentration of 5 IE/mL 

or approximately 0.01 mM, while the stability studies previously have been done at a 

concentration of 0.1 mM. The concentration used in this study (10 mM) was higher, since it 

enables NMR measurements without 13 C or 15 N enrichment, relying only on the low natural 

abundance of these isotopes. 

A complete NMR analysis of oxytocin in phosphate buffer has been reported [20]. 

However, no reports are available on the 1 H, 13 C, and 15 N resonance assignments of oxytocin 

in aspartate buffer in the absence and presence of divalent metal ions. In this study, we used 

2D NMR spectroscopy to investigate the conformation of oxytocin in the presence of Zn 2+ 

or Mg 2+ in aspartate buffer at pH 4.5. 


2.1 Materials 

Oxytocin monoacetate powder (Diosynth. Oss, The Netherlands) was kindly provided 

by MSD, Oss, The Netherlands. Deuterium oxide (D 2 O, isotopic purity 99.9 atom % D) 

containing 0.75% TSP(3-(trimethylsilyl) propionic-2,2,3,3,-d4 acid, sodium salt) was 

purchased from Aldrich, Steinheim, Germany and deuterated L-aspartic acid-2,3,3-d3 was 

purchased from Medical Isotope, Inc, NH. TSP was used as an internal standard having a 

chemical shift (d) of 0.0 ppm. Zinc chloride was purchased from Fluka, Steinheim, Germany. 

All reagents used for the NMR experiments were of analytical grade (purity > 99%), and 

were used without further purification. 

2.2 Sample preparation 

Two different types of NMR samples were prepared with the following compositions: 

1. 2D 13 C- 1 H HSQC NMR samples 

10 mM oxytocin (natural abundance) in 10 mM deuterated aspartate buffer (pH 4.5 = pD 5.1) 

in D 2 O containing 0.75% TSP in the absence (OT-AP) and presence of 100 mM ZnCl 2 

(OT-AP-Zn) or 100 mM MgCl 2 (OT-AP-Mg). Reference solution was oxytocin in D 2 O. 

2. 2D 15 N- 1 H HSQC, 1 H- 1 H TOCSY, and 1 H- 1 H NOESY samples. 

10 mM oxytocin (natural abundance) in 10 mM aspartate buffer (pH 4.5) in H 2 O 

containing 0.75% TSP in the absence and presence of 100 mM ZnCl 2 or 100 mM MgCl 2 . 

Reference solution was oxytocin in water. 

2.3 Sample preparation 

Spectra of these samples were recorded using a Varian Unity INOVA 600 MHz NMR 

spectrometer equipped with pulsed field-gradient probes. The spectra were recorded at 

278 K, processed using NMR Pipe [21] and analyzed using Sparky [22]. All information 

about the NMR measurements is summarized in Table 1.

Table 1. Information about the NMR experiments 

Experiment Correlation 

15 1 N and H 

separated by one 

bond 

15 1 N- H HSQC (NH-HN Nε-Hε for 

Gln, Nd-Hd for Asn, 

N-H for amide of 

Gly9 ) 

13 1 C and H 

separated by one 

13 1 C- H HSQC bond 

(Ha-Ca, Hβ-Cβ, 

Hγ-Cγ, etc.) 

Correlates all protons 

1 1 H- H TOCSY in a J-coupled spin 

system 

Correlates all protons 

1 1 H- H NOESY which are close in 

space (

6 

100 

Figure 1. 2D NMR spectra of oxytocin in aspartate buffer: 15 N- 1 H HSQC spectrum which 

shows the correlation between amide protons and amide nitrogens in the backbone of 

the peptide. The correlation between amide protons and amide nitrogens in the side 

chain of Gln 4 , Asn 5 and the C-terminal Gly 9 (which has a carboxy-amide group instead of a 

regular carboxylate, indicated by N1-H11 and N1-H12)are also indicated (A); 13 C- 1 H HSQC 

spectrum shows the correlation between all side chain carbon signals with signals due to 

the attached protons. The Tyr 2 Cd and Cε signals are aliased in this spectrum. The real 

Tyr 2 Cd and Cε chemical shifts are 132.69 and 117. 95 ppm, respectively (B) 

Figure 1 2D NMR spectra of 

aspartate buffer: 15 N- 1 H HSQC spe 

shows the correlation between a 

and amide nitrogens in the back 

peptide. The correlation between a 

and amide nitrogens in the side c 

Asn 5 and the C-terminal Gly 9 ( 

carboxy-amide group instead o 

carboxylate, indicated by N 

N1-H12)are also indicated (A); 13 

spectrum shows the correlation bet 

chain carbon signals with signal 

attached protons. The Tyr 2 Cδ an 

are aliased in this spectrum. The 

and Cε chemical shifts are 132.69 

ppm, respectively (B)

Figure 2. 2D 1 H- 1 H NOESY spectrum of oxytocin in aspartate buffer. 


6 

101

6 

102 

Figure 3. Short distances observed in the 2D 1 H- 1 H NOESY spectrum of oxytocin in 

aspartate buffer. 

3.3 Chemical shift difference (Dd) 

The chemical shift differences induced by the divalent metal ion upon complexation with 

oxytocin in aspartate buffer were analyzed to get information about which residues are 

involved in binding of the divalent metal ions to oxytocin and to learn about the extent of 

the perturbation caused by these ions. 

3.3.1 Influence of aspartate buffer on the conformation of oxytocin in water. 

Aspartate buffer is one of the buffers known to stabilize oxytocin if divalent metal ions are 

present in the liquid formulation. To investigate the influence of aspartate buffer on the 

conformation of oxytocin, the chemical shift (d) of Ca and Ha backbone resonances of 

oxytocin in deuterated aspartate buffer was compared with the chemical shift of Ca and Ha 

resonances of oxytocin in D 2 O. The difference between those chemical shifts was expressed 

as chemical shift difference (Dd) in ppm. 

Figure 4A shows that aspartate buffer induced a minor change in oxytocin in D 2 O. The 

largest chemical shift changes were observed in the Ca and Hα resonances of Cys 1 . Ca’s 

of Cys 1 and Cys 6 were more shielded and Ca’s of Tyr 2 and Ile 3 were more deshielded in 

the presence of aspartate buffer. Ca’s of Gln 4 , Asn 5 , Pro 7 and Leu 8 were not affected by the 

presence of aspartate buffer. As shown by black bars in Figure 4A, the Ha’s of Cys 1 , Ile 3

Figure 4. Chemical shift difference (Dd) of Ca (light grey bars) and Ha (black bars) of A. 

oxytocin in deuterated aspartate buffer (pD 5.1) and B. oxytocin in D 2O in the presence 

of Zn 2+ relative to the chemical shifts of the same spins measured in D 2O. Oxytocin in 

deuterated aspartate buffer (pD 5.1) in the presence of C. Zn 2+ and D. Mg 2+ relative to the 

chemical shifts of the same spins measured in deuterated aspartate buffer pD 5, analyzed 

by 2D 13 C- 1 H HSQC spectroscopy. 

and Cys 6 are also affected. The opposite effects of aspartate on the Ca and Ha chemical 

shifts of Cys 1 may be due to an electrostatic interaction between the positive charge of the 

N terminus of Cys 1 and the negative charge of the carboxylate group of the aspartate at 

pD 5.1. In order to test this hypothesis, we recorded 2D 13 C- 1 H HSQC of oxytocin in the 

presence and absence of aspartate buffer at pD 2.0. The same chemical shifts were found 

in the two spectra (see Supporting information). This result shows that at a very low pH, 

where the carboxylate groups of aspartate are totally protonated, there is no effect at Cys 1 , 

apparently because the electrostatic interaction between the aspartate and the N terminus 

of Cys 1 has disappeared. 

3.3.2 Influence of zinc ions on the conformation of oxytocin in water. 

Figure 4B shows that Ca and Ha of the amino acid residues of oxytocin other than Leu 8 

are shifted in the presence of Zn 2+ . The largest chemical shift changes are observed in Ca of 

Tyr 2 . In Tyr 2 , Gln 4 , Cys 6 , Pro 7 , and Gly 9 the Ca spins are more deshielded in the presence of 

Zn 2+ . In contrast, in Ile 3 and Asn 5 the Ca spins are more shielded in the presence of Zn 2+ . 

The presence of Zn 2+ does not cause changes in the Ca and Ha chemical shifts of Leu 8 . 

The largest Dd observed for Ha is in Cys 1 . Interestingly, similar to the effect of aspartate 


6 

103

6 

104 

buffer on oxytocin, in many residues the chemical shift change of Ca induced by Zn 2+ is of 

opposite sign compared to that of Ha. 

3.3.3 Influence of divalent metal ions on the conformation of oxytocin in 

aspartate buffer 

As depicted in Figure 4C, the effects of Zn 2+ on chemical shift of oxytocin in aspartate 

are very similar to the effects observed in D 2 O (Figure 4B), suggesting that a similar 

conformational change is induced in both circumstances. A small chemical shift change 

was observed only in the Ca of Cys 1 . 

The effects of Mg 2+ on the chemical shifts of Ca and Ha resonances of oxytocin in 

aspartate buffer are much smaller than those of Zn 2+ ions. Strikingly, also in the case of 

Mg 2+ , the effects on Ca resonances are opposite to the effects on the Ha resonance of the 

same residue, except for Cys 6 and Leu 8 . The effects of Zn 2+ and Mg 2+ ions on the Ca and Ha 

chemical shifts of oxytocin in the presence of aspartate are displayed in Figure 6. 

Figure 5. Overlay of Ca and Ha signals from 13C-1H HSQC spectra of oxytocin in deuterated 

aspartate buffer without (red) and with Mg2+ (green) (A) or Zn2+ (blue) (B). The Pro7 Cd-Hd 

correlation is also shown in these spectra. 

Figure 5 Overlay of Cα and Hα signals from 13 C- 1 H HSQC spectra of oxytocin in 

deuterated aspartate buffer without (red) and with Mg 2+ (green) (A) or Zn 2+ (blue) (B). 

The Pro 7 Cδ-Hδ correlation is also shown in these spectra. 

4 Discussion 

The results of this study indicate that Zn 2+ and aspartate buffer changes the

4 dIscussIon 

The results of this study indicate that Zn 2+ and aspartate buffer changes the conformation 

of oxytocin. These conformational changes most likely contribute to the stabilization of 

oxytocin in aqueous solution. Our study presents nearly complete NMR assignment 

of oxytocin in aspartate buffer in the presence and absence of divalent metal ions, Zn 2+ 

and Mg 2+ . 2D 1 H- 1 H NOESY spectra of oxytocin in aspartate buffer clearly demonstrate 

that the residue pairs Ile 3 -Asn 5 and Ile 3 -Cys 6 are in near proximity. These NOEs are also 

present in a 2D 1 H- 1 H NOESY spectrum of oxytocin in water. Molecular dynamic studies 

by Wyttenbach et al. [11] suggested an interaction between Ha of Tyr 2 and the Gly 9 amide 

group of oxytocin in water, but we could not find evidence for this interaction. 

Aspartate buffer induces a minor change in the NMR spectrum oxytocin in D 2 O with the 

largest chemical shift changes are observed in Cys 1 . We suggest that there is an electrostatic 

interaction between the positively charged N terminus of Cys 1 and the negatively charged 

carboxylate groups of aspartate. This hypothesis is supported by the fact that in pD 2, when 

aspartate is no longer negatively charged, no changes in chemical shifts are induced by aspartate. 

Zn 2+ has similar effects on oxytocin in aspartate buffer and in D 2 O, suggesting that a 

similar conformational change is induced in both circumstances. In our previous study [7], 

Zn 2+ was shown to improve oxytocin stability only slightly in water, but much more in the 

presence of aspartate. The small effect of Zn 2+ on the chemical shift of the Ca of Cys 1 , which 

was observed only in the presence of aspartate, may be relevant in this respect. We suggest 

that the carboxylate group of aspartate neutralizes the positive charge of the N terminus 

of Cys 1 , allowing the interactions with Zn 2+ to become more favorable. The arrangement 

of carbonyl/carboxyl groups around the Zn 2+ might also play a role and may explain 

the observed chemical-shift changes and the protection of the disulfide bridge against 

intermolecular reactions that lead to dimerization and inactivation. 

The observation of chemical-shift changes in Ca and Ha resonances suggest that small 

changes in the backbone dihedral angles occur to accommodate Zn 2+ . However drastic 

changes in the overall structure of the molecule are not likely, since no drastic changes 

were observed in the pattern of NOEs upon addition of Zn 2+ (see supporting information). 

Increased propensities for the formation of helical or extended secondary-structure 

elements are not likely, because these would be accompanied by correlated changes in Ca 

and Ha chemical shifts [29], while in our study we found these changes to be mostly anticorrelated. 

On the other hand, the presence of a positively charged Zn 2+ ion per se will cause 

polarization of nearby chemical bonds and thus may explain the observed anti-correlation 

in the Ca and Ha chemical-shift changes. 

The effects of Mg 2+ on the chemical shifts of Ca and Ha resonances of oxytocin in 

aspartate buffer are much smaller than those of Zn 2+ . This result is in agreement with our 

finding from isothermal titration calorimetry that Mg 2+ show only very weak heat effects 

when added to solutions of oxytocin [9]. The small chemical-shift changes in the backbone 

of residues Cys 1 , Tyr 2 and Ile 3 in the presence of Mg 2+ in aspartate buffer may be due to 

a cation-pi interaction between Mg 2+ and the aromatic side chain of Tyr 2 . However, the 

interactions between Mg 2+ and oxytocin in aspartate buffer are too weak to cause a significant 

shift in the conformational equilibrium that is essential for protection against inactivation. 


6 

105

6 

106 

5 conclusIon 

In conclusion, our NMR studies revealed that Zn 2+ cause a shift in the conformational 

equilibrium of oxytocin in aqueous solution. Zn 2+ causes changes in the chemical shifts of 

almost all residues of oxytocin while Mg 2+ only induces very little chemical-shift changes 

in some residues. In contrast an analysis of the NOESY spectra in the presence and absence 

of Zn 2+ showed that the same NOEs are present under both circumstances, although with 

slightly different intensities. The exact nature of the changes induced by Zn 2+ is not yet 

clear. Small changes in the backbone dihedral angles to accommodate the Zn 2+ may explain 

the chemical shift changes, but drastic changes in the overall structure of the molecule 

are not likely, since no drastic changes in the pattern of NOEs could be observed upon 

addition of Zn 2+ . Alternatively, the presence of a positively charged zinc ion per se will 

cause polarization of nearby chemical bonds and thus may explain the observed chemical 

shift changes. A definitive explanation of these observations must await a more detailed, 

dynamic description of the peptide in the absence and presence of Zn 2+ , possibly from MD 

simulations steered by distance- and angle-restraints derived from the NMR spectra. 


This study was performed within the framework of the Dutch Top Institute Pharma project: 

number D6–202. 

The authors want to thank MSD Oss for providing oxytocin for the study and Ruud M. 

Scheek at the Groningen Biomolecular Sciences and Biotechnology Institute for helpful 

discussions. 

references 

1. K.L. Maughan, S.W. Heim, S.S. Galazka, 

Preventing postpartum hemorrhage: 

managing the third stage of labor, Am. Fam. 

Physician 73 (2006) 1025-8. 

2. du Vigneaud V, Ressler C,Trippett S., The 



Chem 205 (1953) 949-57. 





(2002) 496-8. 

4. A. Hawe, R. Poole, S. Romeijn, P. Kasper, R. 

van der Heijden, W. Jiskoot, Towards heatstable 

oxytocin formulations: analysis of 

degradation kinetics and identification of 

degradation products, Pharm Res 26 (2009) 

1679-88. 




Geneva WHO/DAP/93.6 (1993). 

6. Trissel LA, Zhang Y, Douglass K,Kastango 

E., Extended Stability of Oxytocin in 

common infusion solution, International 

Journal of Pharmaceutical Compounding 10 

(2006) 156-8. 









Hinrichs, A New Strategy To Stabilize 

Oxytocin in Aqueous Solutions: II. 

Suppression of Cysteine-Mediated 

Intermolecular Reactions by a Combination 

of Divalent Metal Ions and Citrate, Mol. 

Pharmaceutics 9 (3) (2012) 554-62. 

9. C. Avanti, W.L.J. Hinrichs, A. Casini, A.C. 

Eissens, A.V. Dam, A. Kedrov, A.J.M. 

Driessen, H.W. Frijlink, H.P. Permentier, 

Insight into the stability of the zincaspartate-oxytocin 

formulation, manuscript 

in preparation. 

10. P. Jurgens, C. Panteliadis, G. Fondalinski, 

Total parenteral nutrition of premature

infants: metabolic effects of an exogenous 

supply of L-aspartic acid and L-glutamic 

acid, Z. Ernahrungswiss. 21 (1982) 225-45. 

11. T. Wyttenbach, D. Liu, M.T. Bowers, 

Interactions of the hormone oxytocin with 

divalent metal ions, J. Am. Chem. Soc. 130 

(2008) 5993-6000. 

12. A.F. Pearlmutter, M.S. Soloff, Characterization 

of the metal ion requirement for oxytocinreceptor 

interaction in rat mammary gland 

membranes, J. Biol. Chem. 254 (1979) 3899- 

906. 

13. E. Sikorska, M.J. Slusarz, B. Lammek, 

Conformational studies of vasopressin 

analogues modified with Nmethylphenylalanine 

enantiomers in 

dimethyl sulfoxide solution, Biopolymers 82 

(2006) 603-14. 

14. J.D. Glickson, R. Rowan, T.P. Pitner, J. Dadok, 

A.A. Bothner-By, R. Walter, 1H nuclear 

magnetic resonance double resonance 

study of oxytocin in aqueous solution, 

Biochemistry 15 (1976) 1111-9. 

15. L.F. Johnson, I.L. Schwartz, R. Walter, 

Oxytocin and neurohypophyseal peptides 

spectral assignment and conformational 

analysis by 220 M1Hz nuclear magnetic 

resonance, Proc. Nat. Acad. Sci 64 (1969) 

1269-75. 

16. M. Rholam, P. Cohen, N. Brakch, L. Paolillo, 

A. Scatturin, C. Di Bello, Evidence for 

beta-turn structure in model peptides 

reproducing pro-ocytocin/neurophysin 

proteolytic processing site, Biochem, 

Biophys, Res, Commun, 168 (1990) 1066-73. 

17. V.S. Ananthanarayanan, M.P. Belciug, B.S. 

Zhorov, Interaction of oxytocin with Ca 2+ : II. 

Proton magnetic resonance and molecular 

modeling studies of conformations of the 

hormone and its Ca 2+ complex, Biopolymers 

40 (1996) 445-64. 

18. E. Sikorska, S. Rodziewicz-Motowidlo, 

Conformational studies of vasopressin 

and mesotocin using NMR spectroscopy 

and molecular modelling methods. Part I: 

Studies in water, J. Pept. Sci. 14 (2008) 76-84. 

19. I.C. Smith, R. Deslauriers, H. Saito, R. 

Walter, C. Garrigou-Lagrange, H. McGregor, 

D. Sarantakis, Carbon-13 NMR studies of 

peptide hormones and their components, 

Ann. N. Y. Acad. Sci. 222 (1973) 597-627. 

20. A. Ohno, N. Kawasaki, K. Fukuhara, H. 

Okuda, T. Yamaguchi, Complete NMR 

analysis of oxytocin in phosphate buffer, 

Magn. Reson. Chem. 48 (2010) 168-172. 

21. F. Delaglio, S. Grzesiek, G.W. Vuister, 

G. Zhu, J. Pfeifer, A. Bax, NMR pipe - a 

multidimensional spectral processing 

system based on unix pipes., J Biomol NMR 

6 (1995) 277-293. 

22. T.D. Goddard, D.G. Kneller, SPARKY 3, 

University of California, San Francisco, 

2003. 

23. L.E. Kay, P. Keifer, T. Saarinen, Pure Absorption 

Gradient Enhanced Heteronuclear Single 

Quantum Spectroscopy with Improved 

Sensitivity, J. Am. Chem. Soc. 114 (1992) 

10633-5. 

24. F.A. Mulder, R. Otten, R.M. Scheek, Origin 

and removal of mixed-phase artifacts in 

gradient sensitivity enhanced heteronuclear 

single quantum correlation spectra, J. 

Biomol. NMR 51 (2011) 199-207. 

25. K.B. John, D. Plant, R.E. Hurd, Improved 

Proton-Detected Heteronuclear Correlation 

Using Gradient-Enhanced z and zz filters, J. 

Magn. Reson. A101 113-7. 

26. D.G. Davis, A. Bax, Assignment of Complex 

1H NMR spectra via Two-Dimensional 

Homonuclear Hartman-Hahn Spectroscopy, 

J. Am. Chem. Soc. 107 (1985) 2820-1. 

27. J. Boyd, G.R. Moore, G. Williams, 

Correlation of proton chemical shifts in 

proteins using two-dimensional exchange 

correlated spectroscopy, J. Magn. Reson. 58 

(1984) 511-6. 

28. C. Cavanagh, W.J. Fairbrother, A.G. Palmer 

III, N.J. Skelton (Eds.), Protein NMR 

Spectroscopy, Second Edition: Principles 

and Practice, 2nd ed., Elsevier Inc., United 

State of America, 2006. 

29. D.S. Wishart, Interpreting protein chemical 

shift data, Prog Nucl Magn Reson Spectrosc 

58 (2011) 62-87. 


6 

107

6 

108 


Table S1. Resonance list of oxytocin in aspartate buffer 

Group Atom Nuc Shift SDev Assignments 

C1 Ca 13 C 54.963 0 1 

C1 Cβ 13 C 42.593 0.009 2 

C1 Ha 1 H 4.225 0.013 2 

C1 Hβ2 1 H 3.279 0 1 

C1 Hβ3 1 H 3.477 0 1 

Y2 Ca 13 C 58.13 0 1 

Y2 Cβ 13 C 38.843 0.009 2 

Y2 Cd 13 C 132.686 0 1 

Y2 Cε 13 C 117.951 0 1 

Y2 H 1 H 9.118 0.009 10 

Y2 Ha 1 H 4.781 0 1 

Y2 Hβ2 1 H 3.002 0.005 4 

Y2 Hβ3 1 H 3.2 0.007 4 

Y2 Hd 1 H 7.247 0.017 5 

Y2 Hε 1 H 6.875 0.006 6 

Y2 N 15 N 123.84 0 1 

I3 Ca 13 C 62.55 0 1 

I3 Cβ 13 C 38.694 0 1 

I3 Cd 13 C 13.49 0 1 

I3 Cγ 13 C 27.28 0.021 2 

I3 Cγ1 13 C 17.67 0 1 

I3 H 1 H 8.103 0.052 16 

I3 Ha 1 H 4.073 0.011 3 

I3 Hβ 1 H 1.94 0.002 3 

I3 Hd 1 H 0.888 0.009 4 

I3 Hγ1 1 H 0.901 0.004 2 

I3 Hγ2 1 H 1.011 0.004 2 

I3 Hγ3 1 H 1.237 0.003 4 

I3 N 15 N 120.224 0 1 

Q4 Ca 13 C 57.793 0 1 

Q4 Cβ 13 C 28.472 0 1 

Q4 Cγ 13 C 33.688 0 2 

Q4 H 1 H 8.368 0.016 12 

Q4 Ha 1 H 4.116 0.005 3 

Q4 Hβ 1 H 2.079 0.004 3 

Q4 Hε21 1 H 6.989 0.013 6 

Q4 Hε22 1 H 7.705 0.012 8 

continued next page


Q4 HG2 1 H 2.41 0.003 2 

Q4 HG3 1 H 2.426 0.003 2 

Q4 N 15 N 120.399 0 1 

Q4 Nε 15 N 112.68 0.001 2 

N5 Ca 13 C 53.044 0 1 

N5 Cβ 13 C 38.413 0 1 

N5 H 1 H 8.455 0.006 13 

N5 Ha 1 H 4.767 0.022 5 

N5 Hβ 1 H 2.878 0.003 2 

N5 Hd21 1 H 7.068 0.01 8 

N5 Hd22 1 H 7.759 0.009 7 

N5 N 15 N 116.465 0 1 

N5 Nd 15 N 112.933 0.004 2 

C6 Ca 13 C 54.019 0 1 

C6 Cβ 13 C 40.798 0.004 2 

C6 H 1 H 8.332 0.006 11 

C6 Ha 1 H 4.914 0.022 2 

C6 Hβ2 1 H 3.005 0.064 3 

C6 Hb3 1 H 3.245 0.003 2 

C6 N 15 N 119.907 0 1 

P7 Ca 13 C 63.203 0 1 

P7 Cβ 13 C 32.035 0.002 2 

P7 Cd 13 C 50.569 0.001 2 

P7 Cγ 13 C 27.467 0 1 

P7 Ha 1 H 4.46 0.004 2 

P7 Hβ2 1 H 1.941 0 1 

P7 Hβ3 1 H 2.314 0 2 

P7 Hd2 1 H 3.744 0.011 2 

P7 Hd3 1 H 3.774 0 1 

P7 Hγ 1 H 2.049 0 1 

L8 Ca 13 C 55.346 0 1 

L8 Cβ 13 C 41.873 0.005 2 

L8 Cd1 13 C 24.903 0 1 

L8 Cd2 13 C 23.345 0 1 

L8 Cγ 13 C 27.046 0 1 

L8 H 1 H 8.705 0.012 11 

L8 Ha 1 H 4.313 0.004 3 

L8 Hβ2 1 H 1.619 0.005 2 

L8 Hβ3 1 H 1.713 0 1 

L8 Hd1 1 H 0.950 0.004 3 

L8 Hd2 1 H 0.911 0.014 2 

L8 Hγ 1 H 1.699 0.004 3 

L8 N 15 N 122.909 0 1 

continued next page 


6 

109

6 

110 


G9 Ca 13 C 44.757 0.006 2 

G9 H 1 H 8.589 0.007 7 

G9 H11 1 H 7.51 0.005 6 

G9 H12 1 H 7.215 0.004 10 

G9 Ha1 1 H 3.875 0.004 2 

G9 Ha2 1 H 3.957 0.01 3 

G9 N 15 N 111.181 0 1 

G9 N1 15 N 107.471 0.002 2 

Table S2. Resonance list of oxytocin in the presence of Zn 2+ in aspartate buffer 


C1 Ca 13 C 55.064 0 1 

C1 Cβ 13 C 43.085 0.011 2 

C1 Ha 1 H 4.317 0.044 2 

C1 Hβ2 1 H 3.295 0.008 2 

C1 Hβ3 1 H 3.554 0.057 2 

Y2 Ca 13 C 58.649 0 1 

Y2 Cβ 13 C 38.689 0.005 2 

Y2 Cd 13 C 132.905 0 1 

Y2 Cε 13 C 118.028 0 1 

Y2 Ha 1 H 4.748 0.029 2 

Y2 Hβ2 1 H 3.022 0.021 4 

Y2 Hβ3 1 H 3.189 0.003 4 

Y2 Hd 1 H 7.235 0.011 7 

Y2 Hε 1 H 6.872 0.006 6 

Y2 N 15 N 123.932 0 1 

I3 Ca 13 C 61.979 0 1 

I3 Cβ 13 C 38.842 0 1 

I3 Cd 13 C 13.663 0 1 

I3 Cγ 13 C 27.233 0.001 2 

I3 Cγ1 13 C 17.826 0 1 

I3 H 1 H 8.131 0.009 15 

I3 Ha 1 H 4.098 0.025 3 

I3 Hβ 1 H 1.945 0.011 3 

I3 Hd 1 H 0.884 0.011 4 

I3 Hγ1 1 H 0.894 0.002 3 

I3 Hγ2 1 H 1.042 0.021 2 

I3 Hγ3 1 H 1.234 0.006 4 

I3 N 15 N 120.225 0 1 

Q4 Ca 13 C 57.986 0 1 

continued next page


Q4 Cβ 13 C 28.54 0 1 

Q4 Cγ 13 C 33.781 0 2 

Q4 H 1 H 8.374 0.027 13 

Q4 Ha 1 H 4.096 0.017 3 

Q4 Hβ 1 H 2.07 0.005 3 

Q4 Hε21 1 H 6.995 0.017 6 

Q4 Hε22 1 H 7.704 0.014 7 

Q4 Hγ2 1 H 2.399 0.009 2 

Q4 Hγ3 1 H 2.423 0.005 2 

Q4 N 15 N 120.391 0 1 

Q4 Nε 15 N 112.749 0.006 2 

N5 Ca 13 C 52.976 0 1 

N5 Cβ 13 C 38.542 0 1 

N5 H 1 H 8.456 0.006 13 

N5 Ha 1 H 4.758 0.012 4 

N5 Hβ 1 H 2.88 0.003 3 

N5 Hd21 1 H 7.068 0.009 6 

N5 N 15 N 116.474 0 1 

N5 Nd 15 N 113.046 0 2 

C6 Ca 13 C 54.282 0 1 

C6 Cβ 13 C 40.909 0.001 2 

C6 H 1 H 8.33 0.005 11 

C6 Ha 1 H 4.878 0.022 2 

C6 Hβ2 1 H 2.97 0.01 2 

C6 Hβ3 1 H 3.264 0.004 2 

C6 N 15 N 119.856 0 1 

P7 Ca 13 C 63.595 0 1 

P7 Cβ 13 C 32.115 0 2 

P7 Cd 13 C 50.779 0.003 2 

P7 Cγ 13 C 27.628 0 1 

P7 Ha 1 H 4.449 0.004 2 

P7 Hβ2 1 H 1.936 0 1 

P7 Hβ3 1 H 2.308 0 2 

P7 Hd2 1 H 3.738 0.017 2 

P7 Hd3 1 H 3.77 0 1 

P7 Hγ 1 H 2.036 0 1 

L8 Ca 13 C 55.351 0 1 

L8 Cβ 13 C 41.604 0.003 2 

L8 Cd1 13 C 25.116 0 1 

L8 Cd2 13 C 23.383 0 1 

L8 Cγ 13 C 27.151 0 1 

L8 H 1 H 8.699 0.005 10 

L8 Ha 1 H 4.308 0.007 3 



6 

111

6 

112 


L8 Hβ2 1 H 1.611 0.01 3 

L8 Hβ3 1 H 1.768 0 1 

L8 Hd1 1 H 0.941 0.032 3 

L8 Hd2 1 H 0.879 0.01 2 

L8 Hγ 1 H 1.688 0.022 3 

L8 N 15 N 122.855 0 1 

G9 Ca 13 C 44.726 0.002 2 

G9 H 1 H 8.572 0.006 7 

G9 H11 1 H 7.51 0.004 5 

G9 H12 1 H 7.217 0.003 9 

G9 Ha1 1 H 3.877 0.011 2 

G9 Ha2 1 H 3.94 0 1 

G9 N 15 N 111.157 0 1 

G9 N1 15 N 107.5 0.016 2 

Table S3. Resonance list of oxytocin in the presence of Mg 2+ in aspartate buffer 


C1 Ca 13 C 54.873 0 1 

C1 Cβ 13 C 42.513 0 2 

C1 Ha 1 H 4.287 0.015 2 

C1 Hβ2 1 H 3.295 0.011 2 

C1 Hβ3 1 H 3.51 0.003 2 

Y2 Ca 13 C 58.233 0 1 

Y2 Cβ 13 C 38.786 0.002 2 

Y2 Cd 13 C 132.706 0 1 

Y2 Cε 13 C 117.98 0 1 

Y2 H 1 H 9.133 0.005 12 

Y2 Ha 1 H 4.771 0.003 2 

Y2 Hβ2 1 H 3.012 0.008 3 

Y2 Hβ3 1 H 3.184 0.004 4 

Y2 Hd 1 H 7.233 0.009 4 

Y2 Hε 1 H 6.872 0.007 7 

Y2 N 15 N 123.963 0 1 

I3 Ca 13 C 62.434 0 1 

I3 Cβ 13 C 38.714 0 1 

I3 Cd 13 C 13.585 0.084 2 

I3 Cγ 13 C 27.312 0 2 

I3 Cγ1 13 C 17.739 0.076 2 

I3 H 1 H 8.1 0.063 13 

continued next page


I3 Ha 1 H 4.074 0.008 3 

I3 Hβ 1 H 1.938 0.003 3 

I3 Hd 1 H 0.877 0.007 7 

I3 Hγ1 1 H 0.894 0.005 3 

I3 Hγ2 1 H 1.018 0.008 2 

I3 Hγ3 1 H 1.23 0.006 4 

I3 N 15 N 120.247 0 1 

Q4 Ca 13 C 57.874 0 1 

Q4 Cβ 13 C 28.457 0 1 

Q4 Cγ 13 C 33.668 0 2 

Q4 H 1 H 8.359 0.013 10 

Q4 Ha 1 H 4.108 0.002 2 

Q4 Hβ 1 H 2.069 0.006 3 

Q4 Hε21 1 H 6.974 0 2 

Q4 Hε22 1 H 7.696 0.01 6 

Q4 Hγ2 1 H 2.404 0.003 2 

Q4 Hγ3 1 H 2.435 0 1 

Q4 N 15 N 120.407 0 1 

N5 Ca 13 C 53.012 0 1 

N5 Cβ 13 C 38.448 0 1 

N5 H 1 H 8.452 0.007 10 

N5 Ha 1 H 4.756 0.01 4 

N5 Hβ 1 H 2.881 0.005 3 

N5 Hd21 1 H 7.065 0.009 4 

N5 Hd22 1 H 7.763 0 5 

N5 N 15 N 116.458 0 1 

N5 Nd 15 N 113.06 0.008 2 

C6 Ca 13 C 54.063 0 1 

C6 Cβ 13 C 40.784 0.007 2 

C6 H 1 H 8.325 0.004 7 

C6 Ha 1 H 4.897 0.004 2 

C6 Hβ2 1 H 2.96 0.006 2 

C6 Hβ3 1 H 3.262 0.008 2 

C6 N 15 N 119.877 0 1 

P7 Ca 13 C 63.284 0 1 

P7 Cβ 13 C 32.052 0.002 2 

P7 Cd 13 C 50.61 0 2 

P7 Cγ 13 C 27.497 0 1 

P7 Ha 1 H 4.453 0.006 2 

P7 Hβ2 1 H 1.938 0 1 

P7 Hβ3 1 H 2.312 0 1 

P7 Hd2 1 H 3.745 0.004 2 

P7 Hd3 1 H 3.749 0 1 



6 

113

6 

114 


P7 Hγ 1 H 2.044 0 1 

L8 Ca 13 C 55.332 0 1 

L8 Cβ 13 C 41.813 0.003 2 

L8 Cd1 13 C 24.926 0 1 

L8 Cd2 13 C 23.365 0 1 

L8 Cγ 13 C 27.07 0 1 

L8 H 1 H 8.693 0.003 9 

L8 Ha 1 H 4.309 0.006 3 

L8 Hβ2 1 H 1.622 0.007 2 

L8 Hβ3 1 H 1.725 0 1 

L8 Hd1 1 H 0.932 0.029 3 

L8 Hd2 1 H 0.904 0.002 2 

L8 Hγ 1 H 1.695 0.008 2 

L8 N 15 N 122.836 0 1 

G9 Ca 13 C 44.806 0.001 2 

G9 H 1 H 8.564 0.001 5 

G9 H12 1 H 7.214 0.003 9 

G9 Ha1 1 H 3.87 0.006 2 

G9 Ha2 1 H 3.942 0 1 

G9 N 15 N 111.19 0 1 

G9 N1 15 N 107.554 0.009 2 

Table S4. Resonance List of oxytocin in water 


C1 Ca 13 C 55.228 0 1 

C1 Cβ 13 C 42.996 0.004 2 

C1 Ha 1 H 4.199 0.051 3 

C1 Hβ2 1 H 3.22 0 1 

C1 Hβ3 1 H 3.392 0 1 

Y2 Ca 13 C 58.032 0 1 

Y2 Cβ 13 C 38.834 0 2 

Y2 Cd 13 C 132.686 0 1 

Y2 Cε 13 C 117.951 0 1 

Y2 H 1 H 9.125 0.004 11 

Y2 Ha 1 H 4.78 0 1 

Y2 Hβ2 1 H 3.002 0.006 4 

Y2 Hβ3 1 H 3.21 0.039 5 

Y2 Hd 1 H 7.244 0.015 5 

Y2 Hε 1 H 6.876 0.005 6 

Y2 N 15 N 123.788 0 1 

continued next page


I3 Ca 13 C 62.415 0 1 

I3 Cβ 13 C 38.757 0 1 

I3 Cd 13 C 13.49 0 2 

I3 Cγ 13 C 27.259 0.002 2 

I3 Cγ1 13 C 17.67 0 2 

I3 H 1 H 8.109 0.055 16 

I3 Ha 1 H 4.072 0.012 3 

I3 Hβ 1 H 1.941 0.006 3 

I3 Hd 1 H 0.885 0.008 5 

I3 Hγ1 1 H 0.902 0.005 3 

I3 Hγ2 1 H 1.01 0.015 2 

I3 Hγ3 1 H 1.228 0.007 4 

I3 N 15 N 120.158 0 1 

Q4 Ca 13 C 57.793 0 1 

Q4 Cγ 13 C 33.683 0.002 2 

Q4 H 1 H 8.371 0.023 12 

Q4 Ha 1 H 4.114 0.006 3 

Q4 Hβ 1 H 2.078 0.005 3 

Q4 Hε21 1 H 6.989 0.014 6 

Q4 Hε22 1 H 7.704 0.009 8 

Q4 HG2 1 H 2.413 0.002 2 

Q4 HG3 1 H 2.423 0.009 2 

Q4 N 15 N 120.41 0 1 

Q4 Nε 15 N 112.664 0.005 2 

N5 Ca 13 C 53.044 0 1 

N5 Cβ 13 C 38.356 0 1 

N5 H 1 H 8.458 0.006 13 

N5 Ha 1 H 4.763 0.016 5 

N5 Hβ 1 H 2.877 0.003 2 

N5 Hd22 1 H 7.756 0.01 7 

N5 N 15 N 116.463 0 1 

N5 Nd 15 N 112.86 0.001 2 

C6 Ca 13 C 54.093 0 1 

C6 Cβ 13 C 40.903 0.007 2 

C6 H 1 H 8.337 0.004 11 

C6 Ha 1 H 4.895 0.015 2 

C6 Hβ2 1 H 3.02 0.088 3 

C6 Hb3 1 H 3.244 0.007 2 

C6 N 15 N 119.936 0 1 

P7 Ca 13 C 63.203 0 1 

P7 Cβ 13 C 32.034 0.002 2 

P7 Cd 13 C 50.57 0.002 2 

P7 Cγ 13 C 27.467 0 1 



6 

115

6 

116 


P7 Ha 1 H 4.46 0.003 2 

P7 Hβ2 1 H 1.94 0 1 

P7 Hβ3 1 H 2.312 0.002 2 

P7 Hd2 1 H 3.744 0.011 2 

P7 Hd3 1 H 3.774 0 1 

P7 Hγ 1 H 2.049 0 1 

L8 Ca 13 C 55.346 0 1 

L8 Cβ 13 C 41.873 0.005 2 

L8 Cd1 13 C 24.879 0 1 

L8 Cd2 13 C 23.383 0 1 

L8 Cγ 13 C 27.046 0 1 

L8 Ha 1 H 4.311 0.005 3 

L8 Hβ2 1 H 1.657 0.042 2 

L8 Hβ3 1 H 1.713 0 1 

L8 Hd1 1 H 0.958 0.005 3 

L8 Hd2 1 H 0.911 0 1 

L8 Hγ 1 H 1.696 0.003 3 

L8 N 15 N 122.962 0 1 

G9 Ca 13 C 44.763 0 2 

G9 H 1 H 8.592 0.013 8 

G9 H11 1 H 7.515 0.003 6 

G9 H12 1 H 7.213 0.005 10 

G9 Ha1 1 H 3.872 0.005 2 

G9 Ha2 1 H 3.955 0.009 3 

G9 N 15 N 111.18 0 1 

G9 N1 15 N 107.426 0.005 2

Figure S1. Overlay of Ca and Ha 

signals from 13 C- 1 H HSQC spectra 

of oxytocin in D 2 O pD 2 (red) and 

oxytocin in deuterated aspartate 

buffer pD 2 (green). 

13 C (ppm) 

5.0 

4.5 

4.5 

4.0 

4.0 

1 H (ppm) 

3.5 

45 45 

50 P7Cδ-Hδ 50 

N5Cα-Hα 

C6Cα-Hα 

C1Cα-Hα 

55 55 

Y2Cα-Hα 

L8Cα-Hα 

60 60 

5.0 

G9Cα-Hα 

Q4Cα-Hα 

I3Cα-Hα 

P7Cα-Hα 

3.5 


6 

117

6 

118 

1 H (ppm) 

0 

2 

4 

6 

8 

10 

9.5 

Y2Hβ3-H 

Y2Hβ2-H 

C1Hβ2-Y2H 

C1Hβ3-Y2H 

L8Hγ-H 

I3Hβ-Q4H 

L8Hδ-N5H 

I3Hβ-H 

L8Hδ-H 

Q4Hβ-H 

Q4Hβ-N5H 

Q4Hγ2-H 

P7Hβ3-L8H 

L8Hα-H 

C1Hα-Y2H 

P7Hα-L8H 

Y2Hα-H 

I3Hδ-N5H 

I3Hγ1-Q4H 

I3Hδ-H 

I3Hγ2-H 

I3Hγ3-N5H I3Hγ3-H 

I3Hγ3-Q4H 

L8Hβ2-H 

L8Hβ2-G9H 

N5Hβ3-C6H 

N5Hβ-H 

N5Hβ2-Hδ22 

C6Hβ2-H 

Y2Hβ2-Hδ 

C6Hβ3-H Y2Hβ2-I3H 

P7Hδ2-C6H 

G9Hα1-H 

Q4Hα-N5H 

I3Hα-H 

Q4Hα-C6H 

I3Hα-Q4H 

L8Hα-G9H 

N5Hα-H N5Hα-Q4H 

C6Hα-H 

Y2Hβ3-I3H 

Y2Hβ3-Hδ 

N5Hδ21-Hδ22 

Y2Hε-Hδ 

Q4Hε21-Hε22 

G9H12-H11 G9H12-H12 

Y2Hε-Hε 

N5Hδ21-Q4Hε22 

Y2Hδ-Hδ 

Q4Hε21-Hε21 

Q4Hε22-Hε22 

N5Hδ21-Hδ21 

N5Hδ22-Hδ22 

Y2Hδ-Y2Hε 

I3H-Y2H I3H-C6H 

I3H-Q4H I3H-H 

Q4H-H 

Q4H-N5H 

G9H-H 

L8H-H 

C6H-H 

G9H11-H12 

G9H11-H11 

N5Hδ22-Hδ21 

Q4Hε21-Hε22 

Q4H-I3H 

Q4Hε22-Hε21 

Y2H-H 

N5H-H N5H-C6H 

Y2H-I3H 

Y2H-Hδ 

9.0 

8.5 

8.0 

7.5 

1 H (ppm) 

I3Hδ-Y2Hδ I3Hγ1-Y2Hε 

Y2Hα-Hδ 

N5Hβ-Hδ21 

7.0 

6.5 

Figure S2. 2D 1 H- 1 H NOESY spectrum of oxytocin in the presence of Zn 2+ in aspartate 

buffer.

Christina Avanti 1,*, Vinay Saluja 1,2,* , Erwin L.P. van Streun 1 , Imma Boyten 1 , 

Henderik W. Frijlink 1 , Wouter L.J. Hinrichs 1 

*Equally contributed first author 

1 Department of Pharmaceutical Technology &Biopharmacy, 


2 Vaccinology, National Institute for Public Health and the Environment, 

Bilthoven, The Netherlands

extremolytes: are there unIversal 

stabIlIzers for proteIns In aqueous 

solutIon? 

7

7 

abstract 

The purpose of this study was to investigate the ability of extremolytes to stabilize the model 

proteins lysozyme and insulin in aqueous solutions. The effects of the extremolytes, betaine, 

hydroxyectoine, trehalose, ectoine, and firoin on the stability of lysozyme were determined by 

Nile red Fluorescence Spectroscopy and a bioactivity assay. Insulin stability was determined 

by RP-HPLC and HP-SEC. The effects of extremolytes on the unfolding temperature of 

the proteins were analyzed using a thermal shift assay for lysozyme and liquid differential 

scanning microcalorimetry for insulin. The interaction between extremolytes and protein 

was studied by isothermal titration calorimetry (ITC). During storage at 70ºC for 10 min, 

firoin protected lysozyme against inactivation better than the other extremolytes. During 

storage at 55ºC for 4 weeks, firoin also acted as a stabilizer, however, betaine, hydroxyectoine, 

trehalose and ectoine, destabilized lysozyme. These findings surprisingly indicate that 

some extremolytes can stabilize proteins under certain stress conditions but destabilize 

the same proteins under other stress conditions. The increased stability caused by firoin 

is explained by the observed increased unfolding temperature of lysozyme in the presence 

of firoin. After storage at 40ºC for 4 weeks, trehalose and ectoine protected insulin against 

degradation, while betaine, hydroxyectoine and in particular firoin seemed to destabilize 

insulin. Furthermore, it was shown that firoin sharply decreased the unfolding temperature 

of insulin. The interaction of firoin with lysozyme and ectoine or trehalose with insulin 

was negligible as determined by ITC. Thus, firoin appears to be an excellent stabilizer for 

lysozyme but strongly destabilizes insulin, whereas ectoine showed the opposite behaviour. 

This study clearly shows that there is not one extremolyte that can acts as a universal 

stabilizer for proteins in aqueous solution.


Protein instability is one of the main issues in the administration of therapeutic protein 

based medicines, in particular in aqueous formulations. Protein stability can be achieved if 

there is a balance between intramolecular interaction of protein functional groups and their 

interaction with solvent environment [1]. A number of experimental studies have been 

done to overcome protein instability. One of the most promising results is the discovery of 

extremolytes, small organic osmolytes found in extremophiles. 

Extremophiles are microorganisms which are capable of surviving under extreme 

conditions, such as high or low temperatures, extreme pressure and high salt concentrations. 

Extremolytes are accumulated in response to these extreme conditions and minimize the 

denaturation of the biological macromolecules and proteins [2]. The polyols derivatives 

ectoine and hydroxyectoine are the first extremolytes that are currently produced in a large 

scale and are already being used as protein stabilizers. Furthermore, also other molecules 

such as betaine[3] various amino acids, carbohydrates such as trehalose[4]and the mannose 

derivative firoin[5,6] have been found in extremophyles and have been identified as stabilizers 

in these species. Figure 1 shows the molecular structures of some of these extremolytes. 

Extremolytes stabilize proteins by forming solute hydrate clusters [7] that are 

preferentially excluded from the hydrate shell of the protein [8]. Exclusion occurs as a result 

of the repulsive interactions between extremolytes and the backbone of the proteins [9] and 

generates accumulation of water near protein domains. Thus, proteins turn into a more 

compact tertiary structure with a reduced surface area. 

According to this mechanism, in the presence of extremolytes the native state of proteins 

is in the lower free energy state than in the unfolded state. Proteins lose their conformational 

Figure 1. Molecular structure of the extremolytes (a) mannosylglycerate (firoin), (b) betaine, 

(c) ectoine, (d) hydroxyectoine, and (e) and trehalose 

extreMolyteS aS proteInS StabIlIzerS 

7 

123

7 

124 

Figure 2. Mode of action extremolytes, adapted from Lentzen G., 2006 (2) 

entropy with the greater entropic loss in the unfolded state (S u ), leading to an overall shift 

in equilibrium towards the native state [10]. The entropy difference between the two states 

increases, thus the folded protein (S f ) is less likely to unfold as ilustrated in Figure 3 [11]. 

The aim of this study was to investigate whether known extremolytes such as betaine, 

hydroxyectoine, trehalose, ectoine, and firoin are able to stabilize the model proteins 

lysozyme and insulin at elevated temperatures and whether they can be used as universal 

stabilizers for proteins in aqueous solutions. 


2.1 Materials 

The following materials were used in this study: Hen egg-white lysozyme (Sigma- 

Aldrich, Steinheim, Germany), Recombinant human insulin, USP (provided by MSD, 

Oss, The Netherlands), ultrapure trehalose(Cargill, Krefeld, Germany), ultrapure betaine 

(Fluka Biochemika, Buchs, Switzerland), firoin(Biotop, Berlin-Brandenburg, Germany), 

ultrapure ectoine and hydroxyectoine(Biomol, Hamburg, Germany), citric acid (Riedel-de 

Haen, Seelze, Germany), sodium hydroxide, acetonitrile (Merck, Darmstadt, Germany), 

dimetylsulfoxide(DMSO), Nile red and M. Lysodeickticus(Sigma-Aldrich, Steinheim, 

Germany), PBS buffer (66 mM phosphate, pH 6.2), SYPRO orange protein gel stain, 

(invitrogenEugene, Oregon, USA), sodium sulphate, concentrated phosphoric acid, 

ethanolamine, hydrochloric acid (Merck, Darmstad, Germany), phosphate buffered saline 

(Fluka Analytical, Steinheim, Germany). 

2.2 The effect of extremolytes on the stability of lysozyme 

2.2.1 Heat shock stability study 

Lysozyme solutions at a concentration of 100 μg/ml were prepared in 10 mM citrate buffer 

(pH 5.0) with and without extremolytes at a concentration of 0.5 M [2,5]. Lysozyme solutions 

were incubated at 70ºC for 10 minutes. The effect of extremolytes on the stability of lysozyme 

was determined by Nile red fluorescence spectroscopy and by measuring its bioactivity.

Figure 3. Proposed mechanism of 

stabilization of extremolytes, adapted 

from Arakawa, 2006 (12) 

2.2.2 Accelerated stability studies 

Lysozyme solutions at a concentration of 100 μg/ml were prepared in 10 mM citrate buffer 

(pH 5.0) with and without 0.5 M stabilizers. Lysozyme solutions were incubated at 55ºC. 

The effect of extremolytes on the stability of lysozyme was determined by measuring its 

bioactivity. Bioassays were performed every week for 4 weeks. 

2.2.3 Nile red Fluorescence Spectroscopy 

Fluorescence studies were performed on a SLM-Aminco AB2 Spectrofluorometer at 

the temperature of 25°C using 5 mm cubical quartz cuvette (Hellma GmbH). Prior 

to measurement, 5 μl of 20 μg/ml Nile red solution in DMSO was added to a 1 ml of 

100 μg/ml lysozyme solution, to obtain a Nile red : lysozyme weight ratio of 1:1000. An 

excitation wavelength of 550 nm was used, with a band pass of 4.0 nm for the excitation 

monochromator and an emission wavelength of 610 nm with a band pass of 4.0 nm for the 

emission monochromator. Data were recorded at 1 nm intervals over the range 560–700 nm 

with a scanning speed of 100 nm/min. Spectra were corrected for background signal caused 

by buffer and stabilizers [12]. 

2.2.4 Bioassay 

The bioactivity of lysozyme was determined by measuring the rate of lysis of Micrococcus 

lysodeikticus by using a method as described by Shugar[13] with some modifications. 

Briefly, 1.3 ml of a 200 µg/ml of Micrococcus lysodeikticus suspension in sterile PBS buffer 

(66 mM phosphate, pH 6.2) was mixed with 10 µl of the lysozyme solutions in plastic disposable 

cuvettes. Immediately after mixing, the cuvette was placed in a UV/VIS spectrophotometer 

and the absorbance was measured at a wavelength of 450 nm. Subsequently, the change of the 

absorbance was recorded over time. Remaining activity of the stressed samples was obtained 

by comparing them with the unstressed samples. Statistical analyses were performed using 

Student’s t test with p < 0.05 as the minimal level of significance. The results are presented as 

mean ± standard error mean unless indicated otherwise. 

2.3 The effect of extremolytes on the stability of insulin 

Insulin was formulated at a concentration of 20 IE/ml in 2 mM PBS buffer at pH 7.4 with 

and without extremolytes at a concentration of 0.5 M. After preparation, the solutions were 

stored at either 4°C or 40°C for 4 weeks, and protected from light. The pH of the samples 

was measured regularly and was found to be remained at 7.4 ± 0.1 during the stability study. 


7 

125

7 

126 

2.3.1 Reversed-Phase High-Performance Liquid Chromatography (RP-HPLC) 

The recovery of insulin was determined by RP-HPLC as described in the USP. A Chrompack 

C18 column with 5 µm particle size, inner diameter of 4.6 mm and length of 250 mm, a 

Waters 510 HPLC pump, a Waters 717 Autosampler and a Waters 486 tunable absorbance 

UV detector were used. Samples of 10 µl were injected and the separation was carried 

out at a flow rate of 1.0ml/min and UV detection at 214 nm. Samples were eluted using 

27% acetonitrile in buffer pH 2.3. The buffer for the mobile phase was prepared by dissolving 

71.0 g of sodium sulphate in 2100 ml of Milli Q water. After adding 6.75 ml of phosphoric 

acid, the pH was adjusted to 2.3 with ethanolamine or 0.5 N phosphoric acid. The total 

volume was completed to 2500 mL and filtered through a 0.45µm filter. Integration of the 

chromatograms was done with Chromeleon software. The runtime was 30.5 minutes [14]. 

2.3.2 High-Performance Liquid-Size Exclusion Chromatography (HP-SEC) 

The monomeric fraction of insulin was assessed by HP-SEC[15]. HP-SEC was carried 

out using a Waters Insulin HMWP column, inner diameter of 7.8 mm, length 200 mm, 

Waters 510 HPLC pump, Waters 717 plus autosampler, Waters 486 tunable absorbance 

UV detector (Waters, Milford Massachusetts, USA)was used. Samples of 20 µl were 

injected, and separation was performed at a flow rate of 0.5 ml/min. Peaks were detected 

by UV absorption at 276 nm. Samples were eluted using mobile phase consisted of 40% of 

L-arginine 1 mg/ml solution, 50% of acetonitrile and 10% glacial acetic acid. The runtime 

was 30.5 minutes. 

2.4 The effect of extremolytes on protein unfolding temperature (T m ) 

2.4.1 Thermal shift assay 

The unfolding temperature of lysozyme in solutions was analyzed by a thermal shift assay 

using a real-time PCR machine [16]. Solutions of 17.5 μl of 1.0 mg/ml lysozyme with 

or without stabilizers (1 M) and 7.5 μl of 300 fold diluted SYPRO Orange solution as a 

molecular probes were added to the wells of a 96-well thin-wall PCR plate (Bio-Rad). 

The plates were sealed with optical-quality sealing tape (Bio-Rad Laboratories BV, 

Veenendaal, The Netherlands), inserted into a real-time PCR machine (iCycler, Bio-Rad 

Laboratories BV, Veenendaal, The Netherlands) and heated from 20 to 90°C with a 0.2°C 

increaseper 20 s. The fluorescence changes of the SYPRO Orange probe in the wells of 

the plate were monitored simultaneously with a fluorescence detector (MyIQ single-colour 

RT-PCR detection system, Bio-Rad) at excitation and emission wavelengths of 490 and 

575 nm, respectively. The midpoint of the transition was taken as the T m . 

2.4.2 Liquid Differential Scanning calorimetry 

The T m of insulin in solutions in a concentration of 1.5 mg/mL with or without 

extremolytes (0.1 M) was analyzed by differential scanning microcalorimetry. Data collection 

was performed using a VP-DSC differential scanning microcalorimeter (MicroCal, LLC, 

Northampton, MA) [17,18]. All insulin scans were performed with 2 mM PBS buffer pH 7.4 

in the reference cell from 25 to 110°C at a scan rate of 90°C/hour and an excess pressure of 

25-26 Psi. All samples and references were degassed immediately before use. A buffer-buffer 

reference scan was subtracted from each sample scan prior to concentration normalization. 

Data analysis was carried out using Origin 7.0 (OriginLab, Northampton, MA).

2.5 Interaction of extremolytes with protein determined by 

Isothermal Titration Calorimetry (ITC) 

Microcalorimetric titrations of extremolytes to lysozyme in citrate buffer pH 5.0 and 

insulin in PBS buffer were performed by using a MicroCal ITC 200 microcalorimeter 

(Northampton, MA 01060 USA). A solution of 20 mM of some selected extremolytes in 

10 mM citrate buffer, pH 5.0 was placed in the syringe, while 300 µl of 1 mM lysozyme in 

10 mM of citrate buffer, pH 5.0 was placed in the sample cell. The reference cell contained 

300 µl of citrate buffer. Experiments were performed at 10, 25, and 55°C. The initial delay 

time was 60 s. The reference power and the filter were set to 5 µcal/s and 2 s respectively. 

A typical titration experiment consisted of 20 injections of 2 µl extremolytes solutions 

with duration of 4 s and the time interval between two consecutive injections was set to 

180 s. During the experiments, the sample solution was continuously stirred at 1000 rpm. 

The effective heat of the protein-extremolytes interaction upon each titration step was 

corrected for dilution and mixing effects, as measured by titrating the extremolytes 

solution into the buffer and by titrating the buffer into the protein solution. The heats of 

bimolecular interactions were obtained by integrating the peak following each injection. 

All measurements were performed in triplicate. ITC data were analyzed by using the ITC 

non-linear curve fitting functions for one or two binding sites from MicroCal Origin 7.0 

software (MicroCal Software, Inc.) [19]. 

3. results and dIscussIon 

3.1 The effect of extremolytes on the stability of proteins 

In order to get a clear and unambiguous picture of the effect of the different extremolytes 

on the stability of the two proteins it was decided to test the stability effects on each of 

the proteins with two different methods. For the lysozyme the Nile Red Fluorescence 

Spectroscopy and the activity assay were used, whereas for the insulin the RP-HPLC and 

HP-SEC were combined. In all studies the extremolytes concentration of 0,5 M was used. 

3.1.1 Stability of lysozyme as measured by Nile Red Fluorescence Spectroscopy 

Nile red fluorescence can be employed to probe changes in protein conformations that are 

related to the formation of hydrophobic surfaces, such as during aggregation or protein 

unfolding because of its sensitivity to the polarity of its environment [20,21]. 

The effect of a heat shock on lysozyme without extremolytes in citrate buffer solution 

pH 5.0 is shown in Figure 4A. A huge increase in the fluorescence intensity of Nile red was 

observed when lysozyme without extremolytes as stressed at 70°C for 10 minutes indicating 

substantial denaturation of lysozyme. Apparently, heating lysozyme without extremolytes 

for 10 minutes at 70 ºC caused collapse of its secondary and/or tertiary structure. Stressed 

lysozyme solutions in the presence of betaine(Figure 4B) or hydroxyectoine (Figure 4C) 

showed similar Nile red fluorescence spectra. However, when trehalosewas added, we 

observed a slight difference in the Nile red fluorescence spectra of the lysozyme sample 

before and after heat shock. Figure 4D shows that there was a minor shift in the maximum 

peak after stress, however, the huge increase in the fluorescence intensities of Nile red 

as found for lysozyme formulations without extremolytes or in the presence betaine or 

hydroxyectoine was not observed. This indicates that trehalose was able to inhibit substantial 

denaturation of lysozyme. The stabilization effect of trehalose might be due to the fact that 


7 

127

7 

128 

Figure 4. The effect of extremolytes on the tryptophanyl fluorescence of the lysozymenile 

red complex after stressing lysozyme at 70°C for 10 minutes in citrate buffer pH 5.0. 

Fluorescence spectra recorded of lysozyme unstressed (solid line) and stressed (dotted line) 

of lysozyme A: without extremolytes, and with B: betaine, C: hydroxyectoine, D: trehalose, 

E: ectoine, and F: firoin

trehalose has a propensity to depress the formation of aggregates and chemical reactions of 

lysozyme by inducing α-helical structures and some tertiary structures [22]. 

Also the Nile red fluorescence spectrum of the lysozyme solution formulated with ectoine 

showed a minor shift in the maximum peak after stress (Figure 4E). Furthermore, also 

significantly higher fluorescence intensity was observed in the stressed lysozyme samples 

containing ectoine than those containing trehalose. However, this increased intensity was 

much smaller than that found for formulations without extremolytes or in the presence 

betaine or hydroxyectoine. Therefore, these measurements indicate that ectoine protected 

lysozyme against denaturation, however, to a somewhat lesser extent than trehalose. 

In conclusion, trehalose and ectoineare able to partially prevent the denaturation of lysozyme,. 

In contrast, when firoin was added to lysozyme solution, as the Nile red fluorescence spectra of 

unstressed and stressed lysozyme solution were fully identical (Figure 4 F) indicating that firoin 

completely inhibited the denaturation of lysozyme during heat shock. 

3.1.2 Stability of lysozyme determined by using Bioassay 

In order to further evaluate the stabilizing effects of extremolytes during a heat shock at a 

temperature of 70ºC for 10 minutes and during storage at a temperature of 55ºC for 4 weeks, 

the biological activity of lysozyme was determined. After heat shock the ability of lysozyme 

to inactivate the bacterium M. Lysodeickticus decreased dramatically. 

Figure 5 shows that lysozyme only maintained about 20-40% of its original activity. 

When betaine, trehalose, and ectoine were added, however, lysozyme maintained about 

70% of its original activity. There was no significant difference observed between the 

stabilizing effects of hydroxyectoine, trehalose, and ectoine but it is difficult to draw a 

conclusion from these data since the level of significance was low. It seems, however, that 

ectoine stabilized lysozyme better than trehalose and hydroxyectoine, also hydroxyectoine 

was not significantly different from control. 

Figure 5. The effect of lysozyme on M. Lysodeickticus (bioactivity) after stressed at 70°C 

for 10 minutes and the effect of extremolytes on the bioactivity of lysozyme. * is the level 

of significance 


7 

129

7 

130 

The bioactivity assay was also used to monitor the effects of extremolytes on the stability 

of lysozyme solution during incubation for 4 weeks at 55°C (accelerated stability study). 

Figure 6 shows that the addition of the extremolytes betaine, hydroxyectoine, trehalose 

and ectoine destabilized lysozyme. These results are in contrast to the heat shock experiments 

where these extremolytes lead to minor or substantial stabilizing effects. On the other hand, 

the bioactivity assay indicated that firoin stabilized lysozyme during storage at a 55°C for 4 

weeks, which is completely in line with the results of the heat shock experiments. 

3.1.3 Stability of insulin as measured by RP-HPLC and HP-SEC 

Figure 7A shows the results of insulin recovery analyzed by RP-HPLC method after 

incubation at 4°C and 40°C for 4 weeks. In PBS buffer alone approximately 95% insulin 

was recovered after 4 weeks at 4°C. The same result was obtained when hydroxyectoine, 

trehalose or ectoine was added to the solution. When incubated at 40°C for 4 weeks 

approximately 30% insulin was recovered from the PBS solution.. Addition of firoin, betaine 

and hydroxyectoine showed a destabilizing effect as these samples showed no insulin 

recovery at all after 4 weeks at 40°C. However, ectoine and trehalose showed significant 

stabilizing effects, resulting in an insulin recovery of 62% and 72%, respectively 

Figure 7B shows the effects of extremolytes on the amount of insulin monomer remaining 

after incubation at 4°C and 40°C for 4 weeks as measured by HP-SEC. At 4°C, about 90% 

insulin remained as monomer in the absence or presence of extremolytes, although firoin 

resulted in a lower extent of monomer remaining. The results after incubation at 40°C for 

4 weeks are in line with the RP-HPLC measurements. In the absence of extremolytes, insulin 

showed substantial degradation resulting in a remaining amount of insulin monomer of 

about 35%. The HP-SEC results confirm that betaine, hydroxyectoine and firoin destabilized 

insulin. However, ectoine and trehalose strongly stabilized insulin, resulting in a recovery 

of 80% and 88% insulin monomer, respectively. 

Figure 6. The effect of extremolytes on the bioactivity of lysozyme during 4 weeks storage 

at 55 °C

Figure 7. The effect of stabilizers on the concentration of insulin in a liquid formulation 

after storage at 4°C (light grey bars) or 40°C (dark grey bars) for 4 weeks determined by A. 

RP-HPLC, B HP-SEC 

In conclusion, the RP-HPLC and HP-SEC measurements indicate that trehalose and 

ectoine are able to protect insulin against chemical and physical degradation in liquid 

formulations, whereas betaine or hydroxyectoine and in particular firoin destabilized insulin. 

3.2 The effect of extremolytes on protein unfolding temperature (T m ) 

The unfolding temperature (T m ) of lysozyme in solution was analyzed by a thermal shift 

assay using a real-time PCR machine. The T m for lysozyme was found to be 82°C. While, 

the addition of firoin, resulted in a much higher increase in T m i.e. 9°C. 

These results may explain the improved stability of lysozyme in the presence of firoin. 

The firoin effect may be due to the increase of the melting temperature of lysozyme by 9ºC. 

It may be that a rise in T m of 4ºC is not enough to protect lysozyme for a longer period 

of time at 55ºC, but that a rise of 9-ºC is enough to stabilize lysozyme for at least a week 

at 55ºC. Santoro et al showed that up to 8.2 molar of extremolytes, including betaine, 

were able to increase the T m of lysozyme to about 23ºC [23]. It is possible that higher 

extremolytes concentrations are required to improve the stability of lysozyme during 

storage at a high temperature. 

Figure 9 shows the T m of insulin in the absence and presence of extremolytes as 

determined by liquid differential calorimetry. A slight but insignificant increase of the 

T m of insulin was found when ectoine was added to the solution. Addition of trehalose 

and betaine had no effect on the T m of insulin, whereas hydroxyectoine and in particular 

firoin even decrease the T m of insulin. The decreased T m found for firoin may explain the 

destabilizing effect of this extremolyte on insulin. 


7 

131

7 

132 

Figure 8. The effect of stabilizers on the unfolding temperature (T m ) of lysozyme. Unfolding 

temperatures was measured as transition midpoint analyzed by thermal shift assay 

using RT-PCR machine of lysozyme on the concentration of 1.0 mg/ml with A: betaine, 

B: hydroxyectoine, C: trehalose, D: ectoine, and E: firoin 

Figure 9. The effect of stabilizers on the unfolding temperature (T m ) of insulin. Unfolding 

temperatures was measured as transition midpoint analysed by liquid differential 

calorimetry of insulin on the concentration of 1.5 mg/ml with A: betaine, B: hydroxyectoine, 

C: trehalose, D: ectoine, and E: firoin 

3.3 Interaction of proteins with extremolytes 

The interaction of extremolytes with proteins was analyzed using Isothermal Titration 

Calorimetry at different temperatures for the most stable formulations. 

Figure 10 shows the similarity in the magnitude of the reaction and dilution enthalpy for 

the titration of lysozyme by firoin and the titration of insulin by ectoine or trehalose at the 

temperatures of 10, 25, or 55°C. The results indicate that in aqueous solution extremolytes 

do not show any significant interaction with the proteins. The stabilizing effects are rather 

due to modification of the water properties. This is in line with the preferential hydration 

theory that the repulsion between the amide backbone of the protein and the extremolytes 

is due to the influence of the extremolytes on the water structure and therefore do not 

interact directly with the proteins [2].

Figure 10. The effect of firoin (A) on the heat capacity of lysozyme, and the effect of ectoine 

(B) and trehalose (C) to the heat capacity of insulin determined by ITC 

4 conclusIon 

This study clearly shows that there are no extremolytes that can be used as a universal stabilizer 

for all proteins in aqueous solution. We even found that certain extremolytes, (firoin) can act 

as a stabilizer for a particular protein (lysozyme), but destabilizes another protein (insulin). 

Even worse, certain extremolytes (betaine) can stabilize a protein (lysozyme) under certain 

conditions (heat shock) but destabilize the same protein under other stress conditions 

(accelerated stability conditions). This implies not only that for each protein the stabilizing 

effects of different extremolytes should be screened but also that the envisaged storage 

conditions should be taken into account. Furthermore, for screening extremolytes for the 

stabilization of proteins, measuring the T m of the protein can be useful to predict stabilizing 

effects but only if the extremolyte induces a substantial change in the T m . 

acknowledgment 

This study was performed within the framework of the Dutch Top Institute Pharma 

(projectnumber D6−202). 

The authors want to thank G.K. Schuurman-Wolters, A. Kedrov and Prof. A.J.M. 

Driessen at the Groningen Institute Biomolecular Sciences & Biotechnology Groningen for 

helpful discussions on L-DSC and ITC. 


7 

133

7 

134 

references 

1. G. Nemethy, Orientation of amino acid side 

chains: intraprotein and solvent interactions, 

Methods. Enzymol. 127 (1986) 183-96. 

2. G. Lentzen, T. Schwarz, Extremolytes: 

Natural compounds from extremophiles 

for versatile applications, Appl. Microbiol. 

Biotechnol. 72 (2006) 623-34. 

3. S. Knapp, R. Ladenstein, E.A. Galinski, 

Extrinsic protein stabilization by the 

naturally occurring osmolytes betahydroxyectoine 

and betaine, Extremophiles 

3 (1999) 191-8. 

4. A. Hedoux, J.F. Willart, L. Paccou, Y. Guinet, 

F. Affouard, A. Lerbret, M. Descamps, 

Thermostabilization mechanism of bovine 

serum albumin by trehalose, J. Phys. Chem. 

B 113 (2009) 6119-26. 

5. T.Q. Faria, S. Knapp, R. Ladenstein, 

A.L. Macanita, H. Santos, Protein 

stabilisation by compatible solutes: effect 

of mannosylglycerate on unfolding 

thermodynamics and activity of ribonuclease 

A, ChemBioChem 4 (2003) 734-41. 

6. N. Borges, A. Ramos, N.D. Raven, R.J. 

Sharp, H. Santos, Comparative study 

of the thermostabilizing properties of 

mannosylglycerate and other compatible 

solutes on model enzymes, Extremophiles 6 

(2002) 209-16. 

7. E.A. Galinski, M. Stein, B. Amendt, M. 

Kinder, The kosmotropic (structureforming) 

effect of compensatory solutes, 

Comp. Biochem. Physiol. 117A (1997) 

357-65. 

8. T. Arakawa, S.N. Timasheff, The stabilization 

of proteins by osmolytes, Biophys. J. 47 

(1985) 411-4. 

9. D.W. Bolen, G.D. Rose, Structure and 

Energetics of the Hydrogen-Bonded 

Backbone in Protein Folding, Annu. Rev. 

Biochem. 77 (2008) 339-62. 

10. A. Linhananta, G. Amadei, T. Miao, 

Computer simulation study of folding 

thermodynamics and kinetics of proteins 

in osmolytes and denaturants, Journal of 

Physics: Conference Series 341 (2012). 

11. T. Arakawa, D. Ejima, Y. Kita, K. Tsumoto, 

Small molecule pharmacological chaperones: 

From thermodynamic stabilization to 

pharmaceutical drugs, Biochim. Biophys. 

Acta 1764 (2006) 1677-87. 

12. M. Sutter, S. Oliveira, N.N. Sanders, B. 

Lucas, A. van Hoek, M.A. Hink, A.J. Visser, 

S.C. De Smedt, W.E. Hennink, W. Jiskoot, 

Sensitive spectroscopic detection of large 

and denatured protein aggregates in solution 

by use of the fluorescent dye Nile red, J. 

Fluoresc. 17 (2007) 181-92. 

13. D. Shugar, The measurement of lysozyme 

activity and the ultra-violet inactivation of 

lysozyme, Biochim. Biophys. Acta 8 (1952) 

302-9. 

14. X. Xu, Y. Fu, H. Hu, Y. Duan, Z. Zhang, 

Quantitative determination of insulin 

entrapment efficiency in triblockcopolymeric 

nanoparticles by high-performance liquid 

chromatography, J. Pharm. Biomed. Anal. 41 

(2006) 266-73. 

15. C.M. Yu, C.Y. Chin, E.I. Franses, N.H. 

Wang, In situ probing of insulin aggregation 

in chromatography effluents with 

spectroturbidimetry, J. Colloid. Interface 

Sci. 299 (2006) 733-9. 

16. U.B. Ericsson, B.M. Hallberg, G.T. Detitta, 

N. Dekker, P. Nordlund, Thermofluor-based 

high-throughput stability optimization 

of proteins for structural studies, Anal. 

Biochem. 357 (2006) 289-98. 

17. V.V. Plotnikov, J.M. Brandts, L.N. Lin, J.F. 

Brandts, A new ultrasensitive scanning 

calorimeter, Anal. Biochem. 250 (1997) 

237-44. 

18. K. Huus, S. Havelund, H.B. Olsen, M. van de 

Weert, S. Frokjaer, Thermal dissociation and 

unfolding of insulin, Biochemistry 44 (2005) 

11171-7. 

19. C. Hoffmann, A. Blume, I. Miller, P. Garidel, 

Insights into protein-polysorbate interactions 

analysed by means of isothermal titration 

and differential scanning calorimetry, Eur. 

Biophys. J. 38 (2009) 557-68. 

20. A. Hawe, M. Sutter, W. Jiskoot, Extrinsic 

fluorescent dyes as tools for protein 

characterization, Pharm. Res. 25 (2008) 

1487-99. 

21. D.L. Sackett, J. Wolff, Nile red as a polaritysensitive 

fluorescent probe of hydrophobic 

protein surfaces, Anal. Biochem. 167 (1987) 

228-34. 

22. T. Ueda, M. Nagata, T. Imoto, Aggregation 

and chemical reaction in hen lysozyme 

caused by heating at pH 6 are depressed by 

osmolytes, sucrose and trehalose, J. Biochem 

130 (2001) 491-6. 

23. M.M. Santoro, Y. Liu, S.M. Khan, L.X. Hou, 

D.W. Bolen, Increased thermal stability 

of proteins in the presence of naturally 

occurring osmolytes, Biochemistry 31 

(1992) 5278-83.

summary, concludIng remarks 

and global perspectIve

& 

138 

summary 

The general objective of this thesis is to develop liquid formulations for polypeptide based 

medicines that are stable under tropical condition. 

Chapter 2 of this thesis discusses peptide instabilities in aqueous solutions and strategies 

to improve peptide stability in parenteral formulations. The most prevalent degradation 

pathways at accelerated temperature and certain pH are hydrolysis including deamidation, 

isomerization and racemization, oxidation including light and metal-induced oxidation, 

β-elimination, disulfide exchange, dimerization and further aggregation. Adequate 

knowledge about amino acid residues involved in peptide degradation can be used to 

develop strategies to improve peptide stability in aqueous solutions. The most common 

approaches to protect peptide degradation in aqueous formulations are formulating peptide 

at an optimum pH, the proper choice of buffers, antioxidants, chelating agents, removal of 

oxygen, protection from light, the use of metal ions, organic solvents, or surfactants. 

The focus of this thesis is on the stabilization of oxytocin (a nonapeptide) in formulations 

for injection. The major approach has been to investigate the stabilization of this peptide 

by divalent metal ions in combination with a certain buffer. The effect of monovalent and 

divalent metal ions in combination with buffers at a pH of 4.5 on the stability of oxytocin 

in aqueous solutions is described in Chapter 3. The chloride salts of monovalent metal 

ions (Na + and K + ) and divalent metal ions (Ca 2+ , Mg 2+ , and Zn 2+ ) in combination with 

citrate or acetate buffer were investigated. Utilizing RP-HPLC and HP-SEC the effect of 

combinations of buffers and metal ions on the stability of aqueous oxytocin solutions after 

4 weeks of storage at either 4°C or 55°C was quantified. Addition of the monovalent ions, 

Na + and K + , to acetate- or citrate-buffered solutions did not increase oxytocin stability, nor 

did the addition of divalent metal ions to acetate buffered solutions. However, the stability 

of oxytocin in aqueous formulations was improved in the presence of 5 and 10 mM citrate 

buffer in combination with at least 2 mM CaCl 2 , MgCl 2 , or ZnCl 2 and depended on the 

divalent metal ion concentration. Isothermal titration calorimetric (ITC) measurements 

were predictive for the stabilization effects observed during the stability study. Formulations 

in citrate buffer that had an improved stability displayed a strong interaction between 

oxytocin and Ca 2+ , Mg 2+ , or Zn 2+ , whereas formulations in acetate buffer did not. In 

conclusion, our study showed that divalent metal ions in combination with citrate buffer 

strongly improved the stability of oxytocin in aqueous solutions. 

In chapter 4, various degradation products of oxytocin in citrate-buffered solution 

after thermal stress at a temperature of 70°C for 5 days were identified. The changes in 

degradation pattern caused by the presence of divalent metal ions were also determined. 

Degradation products of oxytocin in the citrate buffered formulations with and without 

divalent metal ions were analyzed using liquid chromatography−mass spectrometry/mass 

spectrometry (LC−MS/MS). In the presence of divalent metal ions, almost all degradation 

products, in particular citrate adduct, tri- and tetrasulfide, and dimers, were greatly reduced 

in intensity. Formulations containing divalent metal ions in combination with citrate buffer 

suppressed the formation of degradation products N-citryl oxytocin, tri/tetrasulfide and 

dimers. The suppression occurred on the disulfide bridge between Cys 1 and Cys 6 . Cysteine is 

susceptible to oxidation and β-elimination, and degradation of oxytocin involving cysteine 

leads to dimerization, formation of tri/tetrasulfide and β-elimination followed by thioether 

formation. No significant difference in the stabilizing effects was found among the 

divalent metal ions Ca 2+ , Mg 2+ , and Zn 2+ , suggesting that divalency is the most important

property of the metal contributing to stabilization of the oxytocin-metal-citrate cluster. The 

suppressed degradation pathways all involve the cysteine residues. We therefore postulate 

that cysteine-mediated intermolecular reactions are suppressed by complex formation of 

the divalent metal ion and citrate with oxytocin, thereby inhibiting the formation of citrate 

adducts and reactions of the cysteine thiol group in oxytocin. Citrate has two opposite 

effects on oxytocin stability. First, it is reactive itself and can attack the N-terminal amino 

group from the cysteine residue to form an adduct. Secondly, it protects oxytocin from 

degradation in the presence of divalent metal ions. 

In chapter 5, the effect of divalent metal ions (Ca 2+ , Mg 2+ and Zn 2+ ) on the stability of 

oxytocin in aspartate buffer (pH 4.5) is described and their interaction with the peptide 

in aqueous solution was determined. RP-HPLC and HP-SEC results indicated that after 4 

weeks of storage at 55°C all tested divalent metal ions improved the stability of oxytocin in 

aspartate buffered solutions. However, the stabilizing effects of Zn 2+ were by far superior 

compared to both Ca 2+ and Mg 2+ . LC-MS/MS results showed that the combination of 

aspartate and Zn 2+ in particular suppressed the formation of peptide dimers. As shown by 

ITC, Zn 2+ interacted with oxytocin in the presence of aspartate buffer while Ca 2+ or Mg 2+ did 

not. In conclusion, the stability of oxytocin in the aspartate buffered-solution is strongly 

improved by the presence of zinc ions, and the stabilization effect is correlated with the 

ability of the divalent metal ions in aspartate buffer to interact with oxytocin. Aspartate 

presumably sequesters the zinc ion inside the ring structure of oxytocin protecting the 

intramolecular disulfide bridge from reactions leading to degradation. 

In chapter 6, the conformation of oxytocin in aspartate buffer in the presence of Mg 2+ 

or Zn 2+ , was investigated by 2D NMR spectroscopy, i.e. NOESY, TOCSY, 1 H- 13 C HSQC 

and 1 H- 15 N HSQC. Almost all 1 H, 13 C and 15 N resonance could be assigned using HSQC 

spectroscopy of oxytocin with neither 13 C nor 15 N enrichment. 1 H- 13 C and 1 H- 15 N HSQC 

spectra showed that Zn 2+ caused changes in the chemical shifts of almost all amino acid 

residues of oxytocin, whereas Mg 2+ only induces minor chemical shift changes in several but 

not all amino acid residues. On the other hand, NOESY spectra exhibited almost the same 

NOEs in the presence and absence of Zn 2+ . Our findings indicate the carboxylate group of 

aspartate neutralizes the positive charge of the N terminus of Cys 1 , allowing the interactions 

with Zn 2+ to become more favorable. These interactions may explain the protection of the 

disulfide bridge against intermolecular reactions that lead to dimerization and inactivation. 

Chapter 7, describes a study on the ability of various extremolytes to stabilize model 

proteins (lysozyme and insulin) in aqueous solution. The effects of the extremolytes, 

betaine, hydroxyectoine, trehalose, ectoine, and firoin on the stability of lysozyme were 

determined by Nile red Fluorescence Spectroscopy and a bioactivity assay. Insulin stability 

was determined by RP-HPLC and HP-SEC. The effects of extremolytes on the unfolding 

temperature of the proteins were analyzed using a thermal shift assay for lysozyme and liquid 

differential scanning microcalorimetry for insulin. The interaction between extremolytes 

and proteins was studied by ITC. During storage at 70°C for 10 min, firoin protected 

lysozyme against inactivation better than the other extremolytes. During storage at 55°C 

for 4 weeks, firoin also acted as a stabilizer, however, betaine, hydroxyectoine, trehalose 

and ectoine, destabilized lysozyme. These findings indicate that some extremolytes can 

stabilize proteins under certain stress conditions but destabilize the same proteins under 

other stress conditions. The increased stability caused by firoin is explained by the observed 

increased unfolding temperature of lysozyme in the presence of firoin. After storage at 40°C 

SuMMary, concludInG reMarkS and Global perSpectIve 

& 

139

& 

140 

for 4 weeks, trehalose and ectoine protected insulin against degradation, while betaine, 

hydroxyectoine and in particular firoin were found to destabilize insulin. Furthermore, 

it was shown that firoin sharply decreased the unfolding temperature of insulin. The 

interaction of firoin with lysozyme and ectoine or trehalose with insulin was negligible 

as determined by ITC. Thus, firoin appears to be an excellent stabilizer for lysozyme but 

strongly destabilizes insulin, whereas ectoine showed the opposite behaviour. This study 

clearly shows that there is not one extremolyte that can acts as a universal stabilizer for 

proteins in aqueous solution. Therefore, to develop stable protein formulations using 

extremolytes, one should consider the envisaged storage conditions besides screening the 

stabilizing effects of different extremolytes. Furthermore, for screening extremolytes for the 

stabilization of proteins, measuring the T m of the protein can be useful to predict stabilizing 

effects but only if the extremolyte induces a substantial change in the T m . 

concludIng remarks 

In this thesis an attempt is made to develop a heat-stable liquid oxytocin formulation. Several 

formulations containing specific combinations of divalent metal ions with buffers were 

found to stabilize oxytocin in aqueous solution. Different analytical tools such as LC-MS 

(MS), ITC and NMR were utilized to gain information on the mechanism of stabilization 

at a molecular level. However, it is still difficult to correlate the results with the mechanism 

by which buffers and divalent metal ions may contribute to the stabilization of oxytocin. 

Although the overall picture was rather complex and mechanisms varied between the 

different ions and buffers, it is clear that stabilization of the Cys 1,6 disulfide bridge in the 

oxytocin molecule is of paramount importance for stabilization of the peptide in aqueous 

solution. This was best illustrated in an NMR study showing that the carboxylate group of 

aspartate neutralizes the positive charge of the N terminus of Cys 1 , allowing the interactions 

with Zn 2+ to become more favorable. These interactions may explain the protection of the 

disulfide bridge against intermolecular reactions that lead to dimerization and inactivation. 

Zn 2+ causes a shift in the conformational equilibrium of oxytocin, which is evident from 

the observed changes in the chemical shifts of almost all resonances, but the exact nature 

of these changes is not clear yet. A final explanation of these observations must await a 

more detailed, dynamic description possibly from molecular dynamic simulations steered 

by distance- and angle-restraints derived from the NMR spectra. 

Molecular dynamic simulations could be an interesting approach to provide insight on 

a molecular level in the stabilization mechanisms of the different divalent metal ions and 

buffers. With this type of studies it may be elucidated which conformational changes are 

caused by the different metal ions and how these conformational changes are affected by the 

buffers in the solution. When furthermore a relation between the oxytocin conformation and 

its stability can be established, it may be feasible to define an optimum combination of metal 

ion and buffer that keeps the oxytocin in the most stable conformation. Another approach 

that may finally result in a stable oxytocin formulation is to use the knowledge gained in this 

thesis in a more or less guided extensive trial-and-error development formulation study. 

Cleavage of the disulfide bridge was found to be the major degradation pathway for 

oxytocin, and certain carboxylate buffers were found to support complex formation with 

certain divalent metal ions, which lead to stabilization. Especially the effects of different 

carboxylate buffers in combination with different divalent metal ions may be an interesting

topic in such a study. Many different carboxylate buffers exist and only three were tested in 

the work described in this thesis. The complexity of the role of the buffer is nicely illustrated 

by the dual role that was found for citrate, being both a stabilizer and an adduct former, 

the differences found for the different ions in aspartate buffer and the absence of these 

differences in citrate or acetate buffer. It is unknown whether other carboxylate buffers may 

have better stabilizing properties on oxytocin and how their effect can be further improved 

with different metal ions. In this respect it is interesting to mention here that a recent study 

in our group revealed that malonate buffer had a higher stabilizing effect than aspartate 

or citrate buffer. At this moment a real time stability study with a formulation containing 

this buffer and zinc ions is ongoing. The six months results that were obtained so far are 

promising and this formulation may be suitable for use in tropical developing countries. 

global perspectIve 

As described in the introduction of this thesis the starting point for our research was the 

insufficient access of mothers in the tropical developing countries to oxytocin, which leads 

to a yearly death rate of approximately 150.000 mothers. The availability of a suitable, 

affordable, liquid oxytocin formulation would potentially prevent their death. When starting 

the research we assumed that the fact that none of the current oxytocin formulations could 

be stored outside the cold chain was a major reason for this problem and that a heat stable 

formulation would prevent these deaths. This starting point and the results of our studies 

raise two major questions: 

1. Was a liquid heat stable oxytocin formulation developed in this study? 

2. What is the best way forward to reduce maternal death due to oxytocin insufficient 

availability in tropical regions, i.e. sub-Saharan Africa? 

Unfortunately, thus far we were not able to find a liquid-oxytocin formulation which 

meets the standard requirement of the pharmacopoeia i.e. storage stability for at least one 

year at 40°C or two years at 30°C. However, the formulations, which involve zinc ions and 

malonate buffers, did demonstrate promising results. But as long as real time stability data for 

a period of at least one to two years are not available, definite conclusions cannot be drawn. 

From a technical point of view several adequate solutions to solve the immediate need 

for a heat stable oxytocin formulation are available: 

1. In those places where a refrigerator is not available, storage of the current oxytocin 

formulation at ambient conditions is feasible, as long as the expired material is simply 

replaced every six months. 

2. Introduce refrigerators on solar energy, in those places where they are currently not 

available. 

3. A stable freeze dried oxytocin formulation (to be reconstituted before use) has been 

developed by us in collaboration with MSD and could be introduced immediately. 

4. The development of a stable spray dried oxytocin formulation for pulmonary 

administration. 

5. The introduction of the developed formulations involving zinc ions and malonate buffer 

is most likely the best solution. However, assessment on the real time stability for at least 

another six months (until November 2012) is required before drawing a definite conclusion. 

SuMMary, concludInG reMarkS and Global perSpectIve 

& 

141

& 

142 

When these five solutions are evaluated the first two solutions require funding to replace 

the oxytocin stocks every half a year and to install energy power systems. 

With regard to the third option, a fast introduction would not only depend on the 

willingness to accept the related increase in process costs and costs of goods. Much more 

important will be the willingness of the responsible authorities to accept such a product on 

their markets, as a replacement for the current product, without a full registration dossier. 

If such a dossier is to be prepared the costs will rise tremendously and they will surely be 

too high for any of the relevant countries in this project. It should in this respect also be 

realized that the proposed replacement from a scientific point of view would not have any 

relevant associated safety risk since it only replaces one parenteral formulation for another 

parenteral formulation. 

The fourth solution involves an alternative dosage form. A pulmonary dry powder 

inhalation system may be a good alternative to the parenteral administration from a 

scientific point of view. But for this kind of alternatives there is a justified request for a full 

registration dossier showing equivalence in efficacy and safety with the current injection 

product. The development studies needed to fill such a dossier would take years and the 

costs would be unacceptably high. Based on these considerations any alternative that goes 

beyond a solution for injection (or a product that is to be reconstituted to such a solution) 

would require too much development time at the cost of too many unnecessary loss of life. 

The fifth solution may potentially be the most ideal solution, but also for this solution 

authorities should be willing to limit their requirements regarding the registration dossier. 

Finally, any solutions clearly require nothing more than good will and funding. 

Oxytocin, as we described, offers the most apparent and economical solution, where the 

money should be spent on rather than for the costs of the circus of politicians and officials, 

who travel around the globe to discuss “the problem”.

samenvattIng, conclusIes, 

aanbevelIngen, 

en mondIaal perspectIef

& 

146 

samenvattIng 

De algemene doelstelling van dit proefschrift (beschreven in hoofdstuk 1) is om 

vloeibare farmaceutische formuleringen te ontwikkelen voor op polypeptides gebaseerde 

geneesmiddelen die stabiel zijn onder tropische omstandigheden. 

Hoofdstuk 2 bespreekt de instabiliteit van peptides in waterige oplossingen en 

mogelijkheden om de stabiliteit van peptides in parenterale formuleringen te verbeteren. 

Hydrolyse, oxidatie en aggregatie zijn de meest voorkomende degradatieprocessen die 

in versneld houdbaarheidsonderzoek (verhoogde temperatuur) met peptides worden 

waargenomen. Hydrolyse kan leiden tot deamidatie, isomerisatie en racemerisatie. 

Oxdatie is vaak geïnduceerd en gekatalyseerd door licht en sporen metaalionen en kan 

leiden tot β-eliminatie, disulfide uitwisseling en dimerisatie. Gedegen kennis over de 

aminozuurresiduen die betrokken zijn bij de degradatieprocessen van peptides kan worden 

gebruikt om strategieën te ontwikkelen om de stabiliteit van peptides in waterige oplossingen 

te verbeteren. Er zijn verschillende mogelijkheden om de stabiliteit van peptides in waterige 

formuleringen te verbeteren. Hiertoe behoren onder andere het bepalen van de optimale 

pH, verwijdering van zuurstof, bescherming tegen licht en de optimale toepassing van 

hulpstoffen in de formulering zoals buffers, anti-oxidanten, chelatiemiddelen, metaalionen, 

organisch oplosmiddelen, en oppervlakte-actieve stoffen. 

De focus van dit proefschrift ligt op de stabilisatie van oxytocine (een nonapeptide) in 

een formulering voor injectie. Met name onderzochten we het stabiliseren van oxytocine 

door het combineren van verschillende metaalionen en buffers. 

In hoofdstuk 3 beschrijven we het effect van monovalente en divalente metaalionen in 

combinatie met buffers bij pH 4,5 op de stabiliteit van oxytocine in waterige oplossingen. De 

chloridezouten van monovalente (Na + en K + ) en divalente metaalionen (Ca 2+ , Mg 2+ , en Zn 2+ ) 

in combinatie met citraat- of acetaatbuffer werden onderzocht. Het effect van combinaties 

van buffer met metaalionen op de stabiliteit van waterige oxytocine-oplossingen na 4 weken 

bewaren bij 4°C of 55°C is gekwantificeerd met behulp van twee analytische technieken: 

Reversed-Phase High Performance Liquid Chromatography (RP-HPLC) en High-Performance 

Size exclussion Chromatography (HP-SEC). Het toevoegen van de monovalente metaalionen, 

Na + en K + aan acetaat- of citraat-gebufferde oplossingen en het toevoegen van de divalente 

metaalionen Ca 2+ , Mg 2+ , en Zn 2+ aan acetaat- gebufferde oplossingen leidde niet tot een 

verbetering van stabiliteit van oxytocine. De stabiliteit in waterige formuleringen werd 

echter wel sterk verbeterd door in een 5 of 10 mM citraatbuffer in combinatie met ten minste 

2 mM CaCl 2 , MgCl 2 , or ZnCl 2 en was afhankelijk van de concentratie divalente metaalionen. 

Isothermal titration calorimetry (ITC) bleek een techniek te zijn met voorspellende waarde 

voor de waargenomen effecten. In formuleringen in citraatbuffer met een sterk verbeterde 

stabiliteit, werd een sterke interactie gevonden tussen oxytocine en Ca 2+ , Mg 2+ , of Zn 2+ . In 

formuleringen in acetaat buffer vonden we dit niet. Samenvattend kunnen we zeggen dat 

de combinatie van divalente metaalionen met citraatbuffer de stabiliteit van oxytocine in 

waterige oplossingen sterk verbeterde. 

Hoofdstuk 4 beschrijft een onderzoek waarin de degradatieproducten van 

oxytocineformuleringen in citraatbuffer werden geïdentificeerd in de aan- en afwezigheid 

van divalente metaalionen, na 5 dagen blootstelling aan een temperatuur van 70°C. De 

degradatieproducten werden geanalyseerd met behulp van vloeistofchromatografie-massa 

spectrometrie/massa-spectrometrie (LC-MS/MS). In de aanwezigheid van divalente 

metaalionen werd de vorming van bijna alle degradatieproducten, in het bijzonder van

citraatadduct, tri-en tetrasulfide en dimeren, sterk verminderd. Dit werd vooral veroorzaakt 

door stabilisatie van de disulfidebrug tussen de aminozuren Cys 1 en Cys 6 . Cysteine is gevoelig 

voor oxidatie en β-eliminatie, en bij de degradatiereacties van oxytocine waarbij cysteine 

betrokken was leidde dit tot de vorming van dimeren, tri- en tetrasulfide en thio-ether. We 

vonden geen substantieel verschil tussen het stabiliserende effect van de divalente metaalionen, 

Ca 2+ , Mg 2+ , en Zn 2+ . Dit suggereert dat divalentie de meest belangrijke eigenschap van het 

metaalion is die leidt tot de stabilisatie van het oxytocine-metaal-citraatcluster. Daarom 

veronderstellen we dat cysteine-gemedieerde intermoleculaire reacties worden geremd door 

de vorming van een complex tussen de divalente metaalionen, citraat en oxytocine. Citraat 

heeft twee tegenovergestelde effecten op de stabiliteit van ocytocine. Enerzijds kan het met de 

N-terminale aminogroep van cysteineresidue reageren waardoor een adduct wordt gevormd. 

Anderzijds kan citraat in aanwezigheid van divalente metaalionen oxytocine juist stabiliseren. 

In Hoofdstuk 5 wordt het effect van divalent metaalionen (Ca 2+ , Mg 2+ en Zn 2+ ) op de 

stabiliteit van oxytocine in aspartaatbuffer (pH 4,5) beschreven. Met behulp van RP-HPLC 

en HP-SEC metingen toonden we aan dat na 4 weken bewaren bij 55°C de combinatie 

van alle drie de geteste divalente metaalionen en aspartaatbuffer, leidde tot een verbeterde 

stabiliteit van oxytocine. De stabiliserende werking van Zn 2+ was veel beter dan van Ca 2+ of 

Mg 2+ . LC-MS/MS resultaten toonden aan dat de combinatie van aspartaat en Zn 2+ vooral de 

vorming van dimeren tegenging. Uit ITC metingen bleek dat Zn 2+ in de aanwezigheid van 

aspartaatbuffer interacties aangaat met oxytocine terwijl dit met Ca 2+ of Mg 2+ niet het geval 

was. Samenvattend kan worden gezegd dat de stabiliteit van oxytocine in de met aspartaat 

gebufferde oplossing sterk kan worden verbeterd door de aanwezigheid van zinkionen, 

en dat dit effect correleert met het vermogen tot interactie tussen de metaalionen en 

het oxytocine. Waarschijnlijk wordt de disulfidebrug gestabiliseerd doordat aspartaat 

complexering van zinkionen in de ringstructuur van oxytocine mogelijk maakt. 

In Hoofdstuk 6 werd de conformatie van oxytocine in aspartaatbuffer in aanwezigheid 

van Mg 2+ of Zn 2+ , onderzocht met behulp van een aantal 2D-NMR technieken: NOESY, 

TOCSY, 1 H- 13 C HSQC en 1 H- 15 N HSQC. In het HSQC-spectra van oxytocine dat niet was 

verrijkt met 13 C of 15 N konden bijna alle 1 H, 13 C and 15 N resonanties worden toegewezen. 

Uit deze spectra bleek dat Zn 2+ veranderingen van de chemical shifts van vrijwel alle 

negen aminozuren van oxytocine veroorzaakt, terwijl Mg 2+ slechts geringe veranderingen 

in chemical shifts in sommige aminozuren induceert. Anderzijds vertoonden de NOESYspectra 

nagenoeg dezelfde NOEs in de aan- en afwezigheid van Zn 2+ . Dit duidt erop aan 

dat de carboxylaatgroep van aspartaat de positieve lading van de N-terminus van Cys 1 

neutraliseert, waardoor de interactie met Zn 2+ gunstiger wordt. Deze interacties kunnen de 

bescherming van de disulfide brug in oxytocine verklaren. 

In Hoofdstuk 7 bestuderen we de stabilisatie van twee modeleiwitten (lysozyme en 

insuline) in waterige oplossing met behulp van verschillende extremolieten. Het effect van 

de extremolieten, betaine hydroxyectoine, trehalose, ectoine en firoine op de stabiliteit van 

lysozyme werd bepaald met behulp van Nile Red fluorescentiespectroscopie en het bepalen 

van de biologische activiteit van het enzym. De stabiliteit van insuline werd bepaald met behulp 

van RP-HPLC en HP-SEC. Het effect van extremolieten op de ontvouwingstemperatuur (Tm) 

van de eiwitten werd geanalyseerd met behulp van een thermal shift assay voor lysozyme 

en liquid differential scanning microcalorimetry voor insuline. Tijdens incubatie bij 70°C 

gedurende 10 minuten werd lysozyme beter gestabiliseerd door fiorine dan door de andere 

geteste extremolieten. Ook tijdens de bewaren bij 55°C gedurende 4 weken, stabiliseerde 

SaMenvattInG, concluSIeS, en MondIaal perSpectIef 

& 

147

& 

148 

fiorine het eiwit. Betaine, hydroxyectoine, trehalose en ectoine bleken lysozyme onder deze 

condities echter te destabiliseren. Deze uitkomsten geven aan dat sommige extremolieten 

eiwitten kunnen stabiliseren onder bepaalde stress-omstandigheden, terwijl dezelfde eiwitten 

juist worden gedestabiliseerd onder andere stress-omstandigheden. De verbeterde stabiliteit 

door firoine kan worden verklaard door de waargenomen verhoogde T m van lysozyme in de 

aanwezigheid van firoine. Na 4 weken bewaren bij 40°C bleek insulin te worden gestabiliseerd 

door trehalose en ectoine, terwijl het eiwit juist gedestabiliseerd werd door betaine, 

hydroxyectoine en met name door firoine. Verder vonden we dat de Tm van insuline sterk 

afnam in de aanwezigheid van firoine. De interactie van firoine met lysozyme en ectoine 

of de interactie van trehalose met insuline zoals bepaald met ITC was verwaarloosbaar. 

Samenvattend kunnen we zeggen dat firoine een uitstekende stabilisator is voor lysozyme, maar 

daarintegen een destabilisator is voor insuline, terwijl ectoine het tegenovergestelde gedrag 

vertoont. Uit onze studie blijkt duidelijk dat er geen extremoliet is die kan fungeren als een 

universele stabilisator voor eiwitten in waterige oplossing. Om een stabiele eiwitformulering 

met extremolieten te ontwikkelen, moet daarom rekening worden gehouden met de beoogde 

bewaarcondities. Voor het screenen van extremolieten op hun vermogen tot stabilisatie van 

eiwitten kan het nuttig om de T m van de eiwitoplossing met en zonder extremoliet te bepalen, 

maar alleen als de extremoliet een substantiële verandering in de T m veroorzaakt. 

conclusIes en aanbevelIngen 

Dit proefschrift beschrijft onderzoek dat gericht is op het ontwikkelen van een thermostabiele 

waterige oxytocineformulering. Diverse combinaties van divalente metaalionen met buffers 

bleken oxytocine in waterig milieuw te kunnen stabiliseren. Verschillende analysemethoden, 

zoals LC-MS (MS), ITC en NMR, werden toegepast om het mechanisme van stabilisatie op 

moleculair niveau op te helderen. Het is echter nog lastig om de resultaten hiervan direct te 

correleren aan het exacte mechanisme dat aan de stabilisatie ten grondslag ligt. 

Hoewel het algemene beeld vrij complex was en het mechanisme voor de verschillende 

gebruikte ionen en buffers anders, is wel duidelijk dat de stabilisatie van de Cys 1,6 

disulfidebrug van het oxytocinemolecuul van essentieel belang is voor het stabiliseren van 

het peptide in waterige oplossing. Dit wordt het beste geïllustreerd in het NMR-onderzoek 

waaruit blijkt dat de carboxylaatgroep van aspartaat de positieve lading van N-terminus 

van Cys 1 neutraliseert, waardoor de interactie met zinkionen beter wordt. Deze interactie 

verklaart de bescherming van de disulfidebrug tegen intermoleculaire reacties die resulteren 

in dimerisatie en inactivatie. 

Zn 2+ veroorzaakt een verandering van de conformatie van oxytocine, hetgeen blijkt uit 

van de veranderingen in de chemical shifts van bijna alle aminozuren. De precieze aard van 

deze veranderingen is echter nog niet helemaal duidelijk maar kan mogelijk opgehelderd 

worden met een dynamische beschrijving van moleculaire dynamische simulaties gestuurd 

door afstand- en hoek-beperkingen die afgeleid kunnen worden uit NMR spectra. 

Het toepassen van moleculair dynamische simulaties kan een interessante benadering 

zijn om inzicht te krijgen in het stabilisatiemechanisme van de divalente metaalionen 

en de buffers op moleculair niveau. Bij dit type studies kan worden opgehelderd welke 

conformatieveranderingen worden veroorzaakt door de verschillende metaalionen en 

hoe deze conformatieveranderingen worden beïnvloed door de buffers in de oplossing. 

Wanneer er ook een relatie is tussen de conformatie van oxytocine en de stabiliteit, is het

mogelijk om een optimale combinatie van metaalionen en buffer te bepalen die leidt tot de 

meest stabiele conformatie van oxytocine. Een andere aanpak die uiteindelijk zou kunnen 

resulteren in een stabiele oxytocineformulering is de kennis uit dit proefschrift te gebruiken 

in meer of minder door ‘trial and error’ geleide formuleringsstudie. 

Splitsing van de disulfidebrug bleek de belangrijkste degradatieroute voor oxytocine te 

zijn. Bepaalde carboxylaatbuffers in combinatie met bepaalde divalente metaalionen bleken 

te comlexeren met oxytocine hetgeen leidde tot stabilisatie van de disulfidebrug. Vooral de 

effecten van dit soort combinaties kunnen een interessant onderwerp voor verdere studies. Er 

bestaan veel verschillende carboxylaatbuffers en slecht drie werden onderzocht in het werk 

dat beschreven is in dit proefschrift. De complexiteit van de rol van de buffer wordt goed 

geïllustreerd door de dubbele rol die citraat kan spelen (stabilisator en adduc vormer), door 

de diverse effecten van de verschillende divalente metaalionen in aspartaatbuffer, en door de 

verschillen in effect van citraat- en acetaatbuffer in combinatie met divalente metaalionen. 

Het is onbekend of andere carboxylaatbuffers oxytocine beter kunnen stabiliseren en of 

de stabiliteit verder kan worden verbeterd met andere metaalionen. In dit verband is het 

interessant om te vermelden dat uit een recente studie binnen onze onderzoeksgroep 

gebleken is dat malonaatbuffer in combinatie met zinkionen oxytocine beter stabiliseert dan 

aspartaat- en citraatbuffers in combinatie met zinkionen. Op dit moment is een real-time 

stabiliteitsstudie met formuleringen die gebaseerd zijn op malonaatbuffer en zinkionen 

gaande. De resultaten na een bewaartermijn van zes maanden zijn veelbelovend waardoor 

deze formulering geschikt zou kunnen zijn voor gebruik in tropische ontwikkelingslanden. 

mondIaal perspectIef 

Zoals beschreven in de inleiding van dit proefschrift was de drijfveer voor ons onderzoek 

onvoldoende toegang in de tropische ontwikkelingslanden tot oxytocine, hetgeen leidt tot 

een jaarlijks sterfte van ongeveer 150.000 moeders. De beschikbaarheid van een geschikte 

vloeibare oxytocineformulering zou mogelijk hun dood kunnen voorkomen. Bij de start 

van het onderzoek zijn we ervan uitgegaan dat de belangrijkste reden voor dit probleem 

het feit is dat geen van huidige oxytocineformuleringen kan worden opgeslagen buiten 

de “cold-chain” en dat er met een warmte-stabiele formulering veel sterfgevallen zouden 

kunnen worden voorkomen. Dit uitgangspunt en de resultaten van ons onderzoek leidt tot 

twee belangrijke vragen: 

1. Is er in dit onderzoek een warmte-stabiele vloeibare oxytocine ormulering ontwikkeld? 

2. Wat is de beste manier om maternale sterfte als gevolg van een gebrek aan oxytocine in 

de tropen, bijvoorbeeld sub-Sahara Afrika, te verminderen? 

Helaas zijn we in onze studie niet in staat geweest om een vloeibare oxytocineformulering 

te vinden, die voldoet aan de standaardeis van veel farmacopees, zoals stabiliteit van ten 

minste één jaar bij 40°C of twee jaar bij 30°C. De formuleringen gebaseerd op de combinatie 

van zinkionen en malonaatbuffer zijn echter wel veelbelovend. Maar zolang real-time 

stabiliteitsgegevens over een periode van ten minste één of twee jaar niet beschikbaar zijn, 

kunnen er nog geen definitieve conclusies worden getrokken. 

Vanuit technologisch oogpunt is er een aantal strategieën waarmee de onmiddellijke 

behoefte aan een warmtebesteding oxytocineformulering op adequate wijze opgelost zou 

kunnen worden: 

SaMenvattInG, concluSIeS, en MondIaal perSpectIef 

& 

149

& 

150 

1. Op plaatsen waar geen koelkast beschikbaar is, is het opslaan van de huidige 

conventionele oxytocineformuleringen bij omgevingsomstandigheden mogelijk, op 

voorwaarde dat ze elke zes maanden worden vervangen. 

2. De introductie van koelkasten op zonne-energie op die plaatsen waar koelkasten op dit 

moment niet voorhanden zijn. 

3. Een warmtestabiele gevriesdroogde oxytocineformulering (te reconstitueren vlak voor 

gebruik) is door ons ontwikkeld in samenwerking met MSD en kan direct worden 

toegepast. 

4. De ontwikkeling van een stabiele gesproeidroogde oxytocineformulering voor 

pulmonale toediening. 

5. De introductie van de eerder genoemde formulering gebaseerd op een combinatie 

van zinkionen en malonaatbuffer is waarschijnlijk de beste oplossing. Een definitieve 

conclusie over de geschiktheid van deze formulering kan pas worden getrokken 

na afronding van een real-time stabiliteitsstudie van twee jaar. De resultaten van dit 

onderzoek zullen in november 2012 beschikbaar komen. 

Als we deze vijf mogelijke oplossingen voor het probleem evalueren, kunnen we 

concluderen dat voor de eerste twee alleen voldoende financiële middelen beschikbaar 

moeten komen om de oxytocinevoorraden elke half jaar te vervangen of om koelkasten op 

zonne-energie aan te schaffen. 

Een snelle introductie van een te reconstitueren gevriesdroogd poeder op de markt, 

de derde oplossing, is niet alleen afhankelijk van de bereidheid van de industrie om 

een stijging van proceskosten en materiaalkosten te accepteren. Veel belangrijker is de 

bereidheid van de verantwoordelijke autoriteiten om een dergelijk product op hun markt 

te toe te laten, ter vervanging van het huidige conventionele product, zonder dat daar 

een volledig registratiedossier voor aangelegd hoeft te worden. Als namelijk een dergelijk 

dossier moet worden voorbereid zullen de kosten enorm stijgen en zeker te hoog worden 

voor de betrokken landen. In dit verband moeten we ons realiseren dat de voorgestelde 

vervanging vanuit wetenschappelijk oogpunt niet tot relevante veiligheidsrisico’s leidt, 

aangezien alleen de huidige parenterale formulering wordt vervangen door een andere 

parenterale formulering. 

De vierde oplossing bestaat uit een alternatieve doseringsvorm. Vanuit wetenschappelijk 

oogpunt kan een droog poeder voor pulmonale toediening een goed alternatief zijn voor de 

parenterale toediening. Voor dit soort alternatieven is echter een volledig registratiedossier 

nodig waaruit de gelijkwaardigheid in werkzaamheid en veiligheid ten opzichte van het 

huidige injecteerbare product blijkt. De ontwikkelingsstudies die nodig zijn voor een 

dergelijk dossier kunnen jaren duren en de kosten zullen onaanvaardbaar hoog zijn. 

Op basis van deze overwegingen zal voor elk alternatief voor een injectie (of voor een 

reconstitueerbaar product) te veel ontwikkelstijd nodig zijn. Dit zal helaas ten koste gaan 

van te veel mensenlevens. 

De vijfde oplossing is mogelijk de meest ideale, maar ook hiervoor zullen autoriteiten 

bereid moeten zijn om hun eisen ten aanzien van het registratiedossier te beperken. 

Ten slotte, om het grote probleem op te lossen is niet meer nodig dan goede wil en financiering. 

De strategiën zoals we in het proefschrift hebben beschreven bieden de meest voor de hand 

liggende oplossing. Hier zou geld aan moeten worden besteed, in plaats van aan het circus van 

politici en ambtenaren die over de hele wereld reizen om “het probleem” te bespreken.

acknowledgements

& 

154 

Completing this thesis would not have been possible without the aid and support of 

countless people over the past four years. I must first address the sincerest gratitude to 

Professor Henderik W. Frijlink, my promotor, who allowed me to experience the life of 

a PhD student. His involvement and crucial contribution has exceptionally inspired and 

enriched my growth as a scientist. I am grateful in every possible way and hope to keep up 

our collaboration in the future. To Dr. Wouter L. J. Hinrichs, my co-promotor, I am deeply 

indebted and thankful for being patience and relentless effort throughout my research 

and during the writing of this thesis, as well as providing unflinching encouragement 

and support in various ways. I would also like to thank Dr. Herman J. Woerdenbag and 

Dr. Gerad J. Bolhuis, who initiated the first contact that led to my PhD studentship in 

Groningen. As a Dutch Top Institute Pharma fellow, I also thank Louise Lammers and John 

den Engelsman, for their valuable suggestions. 

I was extraordinary fortunate in having Professor Geny M.M. Groothuis, Professor 

Arnold J.M. Driessen, and Professor Wim Jiskoot as members of the reading committee, for 

their constructive comments that significantly improved the content of this thesis. 

The completion of this thesis was made attainable with support and contribution from 

colleagues and friends, whom provided valuable discussion, asisstance with instruments, 

molecular modelling, and production process. On these, I shall address sincere thanks to 

Jean-Piere Amorij, Andrea Hawe, Robert Poole, Bazak Kukrer, Hjalmar Permentier, Alexej 

Kedrov, Angela Casini, Frans Mulder, Ruud Scheek, Alia Oktaviani, Peter van der Moelen, 

Jamshed Anwar, Hans de Waard, Vinay Saluja, Jan Fisher, Anko Eissen, Marinella Visser, 

Andy-Mark Thunnissen, Ali Rohman, Eni Ratnaningsih, Faizah Fulyani, Ryanto Boediono, 

M. Khalid, Alexej Kedrov, Annie van Dam, Ghea Schuurman, Syarif Riyadi, Caroline Visser, 

Hans van Doorne, Wangsa Tirta Ismaya, Wouter Tonis, and Jan Ettema. 

Collective and individual acknowledgment are also owed to my colleagues at the 

Department of Pharmaceutical Technology and Biopharmacy, University of Groningen, 

whose present refreshed, helpful and memorable at any time. Many thanks address to Paul, 

Dotie, Gieta, Doreenda, Peter, Stieneke, Lida, Fesia, Parinda, TT, Taufan, Senthil, Wouter T, 

Anne, Niels, Floris, Marcell, Elham, and Maarten. For Anne de Boer, to be my company on 

the trip to the hospital and cheering me up when I had the bike accident. Also thanks to the 

talented photographer BT, who provided nice picture for the thesis cover, which I am proud 

of. And to Milica, my “roomy”, it is really at my pleasure to have her as the paranymph. 

I also benefited by outstanding works from graduate and undergraduate students, whose 

works contributed to this thesis: Dewi, Jenny, Imma, Stashek, Ilja, Marriel, Ilvy, Esther, 

Erwin, and Marieke. My special thanks to Phillips, for his “Innovative” idea. 

To Henk, Kathy, Rika, Joke, Heleen, Anneke, Annete, Tim, Lisanne, Yvonne, Rijn and 

Sonja, I am thankful for their kind help with the administration works. 

I would also acknowledge the Rector and the Dean of the Faculty of Pharmacy of the 

University of Surabaya for granting the academic leave permit during my PhD study in 

Groningen, The Netherlands. 

Furthermore, I would also like to thank members of the Indonesian Students Association 

in Groningen which has given me the opportunity to involve as well as contribute in developing 

the organization. It is my wish that our splendid relationship would remain growing. 

Sincere thanks for my cousin, FS Widoyono and his family: Indah, Pandu, Pandji who 

helped me a lot with my daily routine. My companion in arms: Mariana, Kenzie, Ono, Neng, 

Insanu, Uyung, Adit, Wisnu, Adhi, Yota, Puri, Aramel, Muiz, Robby, Astri, Rahma, Lia,

Mutia, Shinta, Kadek, Wiwin, Arvie, and my beloved best friends: Fanny, Poppy, Awalia, 

Yayok, Desti, Yuli, Aini, Klara, Tara, Sita, Ismail, Reza, Pandji, Ratna, and many others, 

whom have been becoming part of my life during my PhD studentship. The joys and the 

sorrows we share certainly help me to learn how to become a better me. 

My parents deserve my highest gratitude, for their inseparable support and prayers. 

Spiritual bond and affection between us will never end at whatever cost, even when this life 

ends, someday. Dadang, Heru, and Ichwan, many thanks for being such a supportive and 

caring siblings. Also for the in-laws with their thoughtful support. 

Words fail me to express my appreciation to my little family: my beloved husband, 

Iskandar Zulkarnain, for his support, caring and gently love. My beloved daughter, Viny, 

my ray of sunshine in the cloudiest day, thank you for being always on my side. We are one 

and hand by hand together reaching our goals. I feel blessed that, despite my limited time 

as a Doctoral student, I am still allowed to spend enough time to accompany you finishing 

your Middle Years Program at International School of Groningen. 

Lastly, I would like to thank everyone who was important to the flourishing realization 

of this thesis, as well as expressing my request for forgiveness that I would not be able to 

mention personally one by one. 

acknowledGeMentS 

& 

155

Innovative Strategies for Stabilization of Therapeutic - TI Pharma

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?